Predictors of permanent emigration in a long-term cohort of Spitak earthquake survivors in Armenia

This study tests the hypothesis whether long-term out-migration was self-selective among survivors of the devastating 1988 Spitak earthquake in Armenia. Migrants are usually not a random sample of their original population, but self-selected in some systematic way for better competitiveness in the host country's labor-market. This phenomenon is known as favourable self-selection, which may lead to loss of human capital, economic stagnation and demographic ‘swamping’ of the sending country. While typical for economic migration, self-selection is considered less common for post-disaster emigration. In this study, characteristics of emigrants and non-migrants were compared in a post-earthquake cohort and the determinants of emigration identified using Cox proportional hazards regression analysis. A follow-up study in 2012 successfully traced 80.3% (n = 1423) of the initial cohort. The cumulative rate of permanent emigration during 23 years of follow-up was 22.1%. The main destination country was Russia (81.0%). The fitted model identified that younger age, male gender, higher education, and sociability were among factors independently predicting out-migration, thus, confirming the hypothesis concerning the favourable self-selectivity of the emigration. Importantly, unemployment and loss of housing were strong push factors for out-migration. Hence, providing proper housing and job opportunities to disaster survivors could reduce their likelihood of emigrating and prevent detrimental effects of the loss of human capital for small countries with limited human resources, like Armenia. In addition to short-term activities for disaster relief, governments must build migration-preventing policies into their medium- and long-term disaster-area rehabilitation programmes if the country in question is negatively affected by out-migration.