Does ‘workfare’ work? The Norwegian experience

The aim in this article is to examine the recruitment process to workfare programmes in the Norwegian municipalities and determine whether these programmes actually enhance self-sufficiency . The design of the study is quasi-experimental. The programme group consists of 300 people and the comparison group of a 10% sample of 1,559 non-participating social assistance recipients from 40 local social service administrations in 1995. In the Norwegian workfare schemes, recruitment according to 'need', i.e. labour market problems and lack of human resources, appears to be the dominating approach. The workfare schemes do not produce significant effects, either on employment or on earnings. These results are supported by analyses of a number of observed variables and of two models applied to deal with unobserved selection bias: the parametric Heckman model and the semi-parametric maximum score model.