Effect of an armed conflict on relative socioeconomic position of rural households: case study from western C？te d'Ivoire

We conducted a secondary analysis of data pertaining to risk factors for malaria and neglected tropical diseases. Standardized questionnaires were administered to 182 households in a rural part of western C？te d'Ivoire in August 2002 and again in early 2004. Between the two surveys, the area was subject to intensive fighting in the Ivorian civil war. Principal component analysis was applied at the two time points for constructing an asset-based wealth-index and categorizing the households in wealth quintiles. Based on quintile changes, the households were labeled as 'worse-off', 'even' or 'better-off'. Statistical analysis tested for significant associations between the socioeconomic fates of households and head of household characteristics, household composition, village characteristics and self-reported events associated with the armed conflict. Most-poor/least-poor ratios and concentration indices were calculated to assess equity changes in households' asset possession.Of 203 households initially included in the first survey, 21 were lost to follow-up. The population in the remaining 182 households shrunk from 1,749 to 1,625 persons due to migration and natural population changes. However, only weak socioeconomic dynamics were observed; every seventh household was defined as 'worse-off' or 'better-off' despite the war-time circumstances. Analysis of other reported demographic and economic characteristics did not clearly identify more or less resilient households, and only subtle equity shifts were noted.However, the results indicate significant changes in livelihood strategies with a significant return to agricultural production and a decrease in the diversity of socioeconomic activities.Situational constraints and methodological obstacles are inherent in conflict settings and hamper conflict-related socioeconomic research. Furthermore, sensitive methods to assess and meaningfully interpret longitudinal micro-level wealth data from low-income countries are lackin