To evaluate the effects of a universal program, i.e. all targeted get access, is associated with the problem that there are no individuals who represent the hypothetical or counterfactual state of being eligible to the program but not participating in it. In this study of a Swedish universal labor market program – the “Activity Guarantee” – we show how regional differences in the implementation of instructions on assignment to the program were utilized to create a control group representing the counterfactual state.

After having gained access to a treatment and a control group, the authors evaluated the effects of the program on the probability of leaving unemployment and on the duration of unemployment. The effects estimated were statistically significant and indicate a clear positive effect of program participation.