Does Social Assistance Recipiency Influence Unemployment?

Abstract

Has the frequency of unemployment a tendency to increase the number of social assistance recipients, or does the relationship work the other way round? This article utilizes Swedish monthly data on aggregated open unemployment and means-tested social assistance recipiency in the period 1991—2004 and proposes a multiple time-series approach based on vector error-correction modelling in order to distinguish between theories about the direction of influence. First, we show that rates of unemployment and receipt of social assistance are co-integrated. Second, we demonstrate that adjustments to the long-run equilibrium are made through adjustments to the receipt of means-tested social assistance. This indicates that the frequency of social assistance recipiency reacts to changes in unemployment rather than vice versa. It is also shown that lagged changes in the number of social assistance recipients do not predict changes in rates of unemployment in the short term. Together, these findings demonstrate that the number of social assistance recipients does not increase the number of unemployed.