The impact of damage to buildings on displacement during the 2014 Israel–Gaza war: a propensity score matching study

Abstract

BackgroundThe July 2014 war in Gaza caused more devastation in the area than at any time since the Israeli occupation began in 1967. The 50-day war resulted in more than 2200 Palestinian deaths, over 11 000 injuries, and large-scale destruction of infrastructure and displacement of over a quarter of the Palestinian population in Gaza. In this study, we examined the causal effect of damage to buildings within a neighbourhood on displacement behaviour in Gaza. Identifying demographic and socioeconomic predictors of displacement, given ongoing exposure to chronic violence and precarious living conditions in Gaza, will allow for action to be taken to avert the consequences of modifiable risk factors of displacement in this population. MethodsThe study was based on data from a 2015 cross-sectional survey including a representative sample of households in the occupied Palestinian territory. The study sample included 10 017 Gazans aged 18 years and above from whom complete data were available for 9285 adults, who were included in matching analysis. Through a quasi-experimental design, propensity score matching estimators were used to evaluate causality between exposure and outcome. The exposure effect was whether 5% or more of the buildings in a respondent's neighbourhood experienced damage. The outcome variable was self-reported displacement from residence as a result of the war. Score matching was conducted using age, sex, education, marital status, refugee status, Gaza residence, employment, household age composition, household crowding (≥3 people per room, excluding kitchen and bathrooms), income, injury status, chronic disease status, and car possession status. FindingsOf the 9285 adults in the analysis, 5304 (57·12%) reported displacement and 3005 (32·36%) lived in an area that experienced damage to buildings during the war. Being a refugee in Gaza and having higher levels of educational attainment were found to be protective against experiencing displacement. An analysis of model-based estimates and average treatment effect showed a positive effect of neighbourhood damage on displacement (average treatment effect 0·35; 95% CI 0·33–0·36), indicating that living in an area that experienced damaged due to war caused displacement. InterpretationDamage to buildings had a substantial impact on population displacement during the 2014 Israel–Gaza conflict, but was not the sole predictor of displacement. The substantial proportion of individuals who were displaced did not live in areas that experienced damage directly, indicating that they were potentially influenced by perception of danger and fear in general. A limitation of this study is the possibility that not all relevant covariates in the matching stage are accounted for, which would reduce comparability of treatments and control groups, and thus limit assertions of causality. Findings from this study can be used to help inform public health programmes and policies designed to protect and serve displaced civilian populations during humanitarian emergencies such as wars, in Gaza and beyond. FundingThis project was supported, in part, by a research grant from the Yale MacMillan Center for International and Area Studies, and an award from the Yale Center for the Study of Race, Indigeneity, and Transnational Migration.