NO SANCTUARY IN TIMES OF CONFLICT A systematic

NO SANCTUARY IN TIMES OF CONFLICT: A systematic analysis of security incidents affecting MSF’s medical mission from 2009 -2013 in four highly insecure contexts Maya Sibley AAU analyst for MCUF June 13, 2014 www. msf. be

BACKGROUND • MSF works in highly unstable contexts, and faces frequent acts of violence against its beneficiaries and medical mission • Typology of violence against MSF is poorly understood – increasing violence is perceived, but not evidence-based Þ Launch of three-year project:

THREE RESEARCH QUESTIONS 1. How many and what are the trends in security incidents in Afghanistan, DRC, South Sudan and Kenya? 2. What are the consequences of security incidents in these four contexts? 3. What are the opportunities and limitations of SINDY reports with regards to the data needs of the MCUF project? SINDY reports & semi-structured interviews for Q 3 Data collection

RESULTS & FINDINGS

MCUF INCIDENTS IN SINDY REPORTS

INCIDENT TYPOLOGY Trend over time Baseline not established, but no increase observed over study period MCUF hotspots Most perpetrators are unknown Locations From those identified, the most common perpetrator is Reflect security strategy taken State Security Forces (25% overall) Most frequent types of aggression Afghanistan Armed entry DRC Robbery South Sudan Threats Kenya Robbery/Death threats 42% 34% 38% (each)

VICTIM TYPOLOGY Total number of victims Incidents without victims Afghanistan DRC South Sudan Kenya N % 32 123 167 9 10 20, 83% 5 8, 62% 15 18, 07% 1 11, 11% Incidents with unspecified victim # 14 29, 2% 13 22, 4% 30 36, 1% 3 33, 3% National staff are the most common victims in absolute numbers but international staff are 5. 1 times more at risk (p-value <0. 0001, CI: 3. 9 -6. 7) • • Males are more frequently victims in Afghanistan (59%) , DRC (73%) & S. Sudan (6%) Women are more frequenty victims in Kenya ( 67%, but sample size is small)

INCIDENT TRIGGERS & CONSEQUENCES The most common triggers and consequences of incidents were mapped (N, %) Trigger: • Financial constraints • Active conflict & civil unrest 60 (30%) 30 (15%) General consequence: • Financial loss • Compromised quality of care • Interruption of services • Compromised neutrality 64 (32%) 45 (23%) 37 (19%) 32 (16%) Types of consequence per victim: • Psychological 206 (62%) • Physical 65 (20%)

IMPACT HEALTHCARE DELIVERY & USE Trigger Not impacted Active conflict/unrest Military/police activity Financial constraints Lack of respect HR & IHL Avengement Perceived lack of respect Other Unknown N 16 12 50 12 3 8 19 28 Perpetrator Security forces & public authority Armed non-state actors International armed forced Ex-MSF General civilian Intentional Lack of Precaution % 53, 3% 63, 2% 83, 3% 75, 0% 50, 0% 100, 0% 82, 6% 77, 8% N 14 7 10 4 3 0 4 8 Not impacted Intention Impacted N 39 11 4 3 31 % 69, 6% 55, 0% 80, 0% 100, 0% 86, 1% Not impacted N 139 7 % 76, 8% 50, 0% % 46, 7% 36, 8% 16, 7% 25, 0% 50, 0% 17, 4% 22, 2% Impacted N 17 9 1 0 5 Impacted N 42 7 P-value % 23, 2% 50, 0% RR 95% CI 0, 002 0, 06 0, 4 0, 05 0, 2 0, 9 0, 5 2, 8 n. s. 1 n. s. 3, 0 n. s. 1. 4 -5. 5 1. 1 -8. 0 - P-value RR 95% CI 0, 2 0, 6 0, 3 0, 07 1 n. s. - % 30, 4% 45, 0% 20, 0% 13, 9% P-value RR 95% CI 0, 03 1 2, 2 1. 2 -3. 9

SINDY REPORTING Half had not reported to SINDY a recent MCUF incident (4/8) TOP 4 REASONS FOR NOT REPORTING “Incidents occur security lack if was “SINDY isdemotivated frequently that ofwe diddirectly “I“Managing the soreporting situation. SINDY feel used only for due to incidents which a systematically priority an impact on MSF activities. ” report, we would be constantly or feedback andoversecurity analysis from indirectly have no reporting” reporting!” report HQ. ” MAIN SINDY SHORTCOMINGS FOR MCUF • Descriptive narrative • What should and shouldn’t be reported unclear • SINDY fields are vague

BIAS & LIMITATIONS

BIAS & LIMITATIONS • Data quality / Partial data • Inconsistent reporting • Lack of denominators • Victims • Time sensitive • Highly subjective • Limited capacity & know-how

CONCLUSIONS 1. First evidence-based mapping of violence directed against MSF and its beneficiaries. 2. Context-specific and general patterns of violence were identified, allowing development of more appropriate tools for the monitoring of “Medical Care Under Fire”. 3. Such monitoring is vital to sustain MSF operations in such contexts.

THANK YOU