The effect on skill and search cost

The employment effect of activation has been analysed for different types of activation measures and separately for non-insured and insured unemployed in activation (Graversen and Weise 2001 Bach 2002). 

Private job training has incomparably the largest direct employment effect. Individuals who participate in this type of activation will on average reduce their dependence on social security by 16 percentage points, which is equivalent to about 2 months a year. However, only one out of ten in activation participates in private job training, and the effects of other measures are significantly lower. 

The most frequently apphed measure is municipal employment projects. Participation in this type of programme reduces the dependence on social security by only three percentage points or about one fifth of the effect of private job training. Individual job training in pubhc workplaces is another measure frequently used. This type of measure reduces the participants dependence on social security by six percentage points, equivalent to 3 weeks a year. 

In a study of rehabilitation, Filges et al. (2002) found that the effects of this measure are very small and may be even smaller than the effect of activation. This is quite surprising since rehabilitation is offered to persons who are expected to be able to regain working ability. However, a significant part of the group would probably have left the labour market if they were not offered rehabilitation. 

In Denmark it has been the general belief that education saves people from unemployment. Therefore, it is a wonder why effects on using education in activation for the insured unemployed and the cash benefit recipients on the whole cannot be measured. One reason could be that it is difficult to measure the effect on education because the effect first shows up in the very long view and works very differently for different target groups. Another reason could be that the education received in an activation course is not designed and dosed well enough. 

---