Unemployment Dynamics Across Countries PRELIMINARY DRAFT

In this paper, we compare aggregate unemployment persistence across countries and over two decades covering the 1980s and the 1990s. We Þnd evidence that individual persistence is by no means the only cause of aggregate unemployment persistence. Its variations only explain 29 % of the variations in our measure of aggregate per-sistence across countries and across time. We explain why we might expect it to be so in a context where the unemployment ßow rates are not constant with respect to the unemployment rate, at a given level of economic activity. The more sensitive the unemployment ßow rates to the unemployment rate the higher aggregate unemployment persistence will be. Using our small panel, we carry out some regressions of the vari-ous components of unemployment and unemployment persistence on several measures of labour market institutions and labour supply vari-ables. We Þnd that, at a given level of individual persistence, aggre-gate persistence is decreased by tight employment legislation and by a high fraction of male employment in part-time jobs. 1