State dependence in Swedish social assistance
Published: October 2013
Keywords: social assistance, state dependence, country of birth, marital status, mixed couples, unobserved heterogeneity, initial condition, dynamic probit mode
JEL classification: I30, I38, J18
This study estimates the state dependence in social assistance in Sweden during 1990–1999 by different types of households, which were grouped by the country of birth of the sampled individual and his/her partner. The results show that the structural state dependence differed extensively across different household types. Although Swedish-born partners who separated are one of the groups with the lowest receipt of social assistance (i.e., 1.08–1.76%), these individuals exhibit the highest state dependence (24.4 percentage points). Foreign-born singles have almost the same value for the state dependence, but these individuals also have the highest receipt of social assistance (18.47%). Surprisingly, the group with the lowest receipt of social assistance (0.27–3.06%) and the lowest state dependence (4.7 percentage points) are the foreign-born women living together with a Swedish-born man. However, for the mixed samples, there are only few parameters for the country of origin that are statistically significant. For example, compared to the Nordic-born group, men born in Eastern Europe who lived with a Swedish-born woman during 1990–99 have a lower propensity (1 percentage points) to receive social assistance, whereas men born in the Middle East or the rest of the world who lived with a Swedish born woman in 1990 have a higher propensity (1.1–1.2 percentage points).