Tables and Graphs. Figure 1: a) distribution violence per month - total; b) distribution Kenema/Kailahun (orange) vs. all other districts (blue)

Transcription

1 Tables and Graphs 10% 9% 8% 7% 6% 5% 4% 3% 2% 1% 0% Figure 1: a) distribution violence per month - total; b) distribution Kenema/Kailahun (orange) vs. all other districts (blue) Figure 2: Comparison data-sets over time. 1

2 Figure 3: Main conflict actors per dataset. 2

3 Figure 4: a) TRC data month and day; b) TRC data year ; c) SLls-LED all 3

4 Tables ACLED Version 4 SLls-LED, 2014 Battle-No change of territory A battle between two violent armed groups where control of the contested location does not change Battle-Non-state actors A battle where non-state actors win control of overtake territory location (If two non-state groups fight and the group that did not begin with control acquires it, this is the Battle-Government regains territory Headquarters or base established Non-violent activity by a conflict actor Riots/Protests Violence against civilians Non-violent transfer of territory correct code) A battle in which the government regains control of a location (this includes proxies fighting for government) A rebel group establishes a base or headquarters. This event records activity by rebel groups/militia/governments that does not involve active fighting but is within the context of the war/dispute. For example recruitment drives, incursions or rallies. It also records the location and date of peace talks, militia rallies and arrests of highranking officials. When any armed/violent group attacks unarmed civilians. Civilians are UNARMED and NOT ABLE to defend themselves. It includes inflicting significant physical harm (e.g. bombing, shooting, torture, rape, mutilation etc.) or accosting victims (e.g. disappearances). When rebels or governments acquire control of a location without engaging in a violent act. When two non-state actor fight actor 2 acquires territory. When two non-state actor fight actor 1 acquires territory Not restricted to rebel groups. Any armed/violent group including DDR camps. No inclusion of peace talks, or forced to carry loads Includes destruction property, food-finding missions and ambushes when there is no violence against civilians Absent Harassment and beatings are not included When rebels (briefly) take over village/town without engaging in fighting with government forces/allies. Also; Reinforcements/Troop movements/patrols/disarmament Table I: Event type and differences in coding ACLED-v4, SLls-LED 4

5 SLls_LED Bellows & Miguel ACLED v-4 UCDP-GED TRC Coverage detail Event Totals Battles 10,6% (934) 5,7% (114) 43,9% (527) 41,9% (617) - 5,4% (473) 3,7% (340) 9.9% (198) 5,7% (105) 12.83% (154) % (126) VAC 23,5% (2061) 16,4% (328) 21,4% (257) 58,1% (857) 20,5% (8265) Looting 20,8% (1828) 2,6% (51) 8.2% (98) - 34,8% (14.020) Bases 7,2% (340) 8,4% (168) 0.4 % (5) - - Territory 28,7% (2516) 51,7% (1031) 0.7% (8) - - Miscellaneous ,6% (17.957) Actors (% allied) 77 (13%) 26 (50 %) 28 (11%) 15 (0%) 28 (6%) Main dyad RUF-GOV RUF-GOV Events ACLED-v4 Battle /Bombing, Base Execution Centre, Brutality, Attack, Troop Movement, Airfield, Control RUF-Military No-exchange Gov. overtakes territory Reb. overtakes territory Base Non-violent activity VAC Non-violent transfer territory RUF-Civilians State based Non-state based One-sided RUF-civilians Abduction, Destruct property, Extortion, Looting, Forced recruitment (Looting) Amputate, Torture, Rape, Sex. abuse/slave, Cannibal (VAC) Detention, Beating, Drugging, Forced Displacement & Labour (Miscellaneous) Time Daily Daily Daily Daily Daily peaks 1991/ 95/ / 95/ / 99/ / 95/ 98/ / 95/ 99 Location Specific Location Specific Location Specific Location Specific Location Chiefdom Bias - - Urban Partly Urban Bias - Sources NPWJ NPWJ Secondary Sources Secondary Sources (including NPWJ) Testimony Table II: Gross comparison of data-sets on Sierra Leone 5

6 TRC ACLED_v4 SLls-LED UCPD_GED B&M TRC ACLED_v SLls-LED UCPD_GED B&M Table III Kendalls Tau (KT) rank correlation for ranking five main clusters of actors All violence TRC ACLED_v4 SLls-LED UCPD_GED B&M TRC ACLED_v SLls-LED UCPD_GED B&M Table IV KT rank correlation (N=18, all p <0.01) for ranking of quarters 1997 to 2001 All violence TRC ACLED_v4 SLls-LED UCPD_GED B&M TRC ACLED_v SLls-LED UCPD_GED B&M Table V Spearman correlation for percentage of fighting per month 1997 to 2001 All violence TRC ACLED_v4 SLls-LED UCPD_GED B&M TRC ACLED_v SLls-LED *** UCPD_GED *** *** *** B&M * ** *** Table VI KT rank correlation for ranking per month 1997 to 2001 VAC only 6

7 SLls-LED UCPD_GED B&G SLls-LED UCPD_GED B&G Table VII Spearman correlation (N=13) for percentage of fighting per district 1991 to 2001 All violence NOT- NPWJ NPWJ Bo 6 6 Bombali 9 10 Bonthe 13 5 Kailahun 2 7 Kambia Kenema 4 1 Koinadugu Kono 1 2 Moyamba 11 4 Port Loko 5 3 Pujehun 7 9 Tonkolili 8 8 Western Ara 3 13 Table VIII: Seperation UCDP-GED violence based on local vs. secondary sources TRC SLls-LED UCPD_GED B&M TRC SLls-LED UCPD_GED *** *** B&M ** *** Table IX KT corr. & Spearman correlation for ranking & percentage 1991 to 2001 VAC & Looting TRC SLls-LED UCPD_GED B&M TRC SLls-LED ** UCPD_GED * *** B&M * *** *** Table X KT corr. & Spearman correlation for ranking & percentage district 1991 to 2001 VAC 7

8 Battles One-Sided Total Democratic Republic of Congo (Zaire) Liberia Central African Republic (CAR) Sierra Leone Rwanda Senegal Chad Republic of Congo Sudan Burundi Mozambique Uganda Kenya Egypt Nigeria South Africa Angola Ethiopia Somalia Algeria Table XI UCDP-GED violence, country-type ( , conflicts <150 excluded) News Human Rights Undefined Total Sierra Leone Battles One-Sided Countries One-sided > Battle Battles One-Sided Countries One-sided < Battle Battles One-Sided Table XII UCDP-GED violence by source ( , conflicts <150 excluded) 8

9 Figure5: volatility monthly event numbers Total Phase 1 Phase 2 Phase 3 Phase 4 Kenema Port Loko Kono Bo Kailahun Pujehun Moyamba Bonthe Koinadugu Tonkolili Western Area Bombali Kambia Table XIII: district victimization number of events 9

10 Figure 6: rate of in/decrease conflict intensity ( ) Model 1 Model 2 Model 3 Specifics VAC Chiefdom VAC Chiefdom/Month VAC Chiefdom/Month Battles.28(.08)***.18(.01)***.78(.07)*** Territorial Take-over.89(.07)***.72 (.01)*** 1.26 (.06)*** Constant 1.23(.80) -.06(.01) -2.77(.04) Number of Observations Number of Groups R2 Type Table XIV: VAC explanation - 38% OLS % OLS Nbreg RUF 1993 RUF/NPFL* RUF/NPFL VAC per battle 1,53 1,04 5,71 (weighted average) SD 0,69 0,69 4,28 Table XV: VAC per battle/actor, *April and May 1991 excluded. 10

11 Totals Phase 1 Phase 2 Phase 3 Phase Violence Nr Months Month Battle VAC + Looting Take-over Success Success % Battle (RUF GOV/allies) Success % Territory (RUF- GOV/allies) Actors Conflict Hierarch RUF (58%) RUF(57%) RUF(67%) RUF(51%) RUF(54%) Remaining hierarchy (all violence) SLA CDF EXT AFRC SLA NPFL CDF EXT STF SLA CDF EXT CDF EXT AFRC WSB EXT CDF SLA WSB Active General Dyads Active Specific Dyads Dominant Dyad - Battle RUF-SLA RUF-SLA RUF-SLA RUF-EXT RUF-CDF % joint operations RUF 29% 57% 0% 56% 1% % joint operations GOV/Allies 14% 28% 9% 12% 16% VAC matching Battle? Battle battle ~battle battle battle Correlation Battle-VAC Table XVI Key variables conflict environments 11

12 Figure 7: Spatial patterns four conflict environments 12