Statistically assisted programme selection (SAPS)

The effects of labour market programmes are heterogeneous among persons. In Switzerland, as well as in other European countries, active labour market policies contain several different programmes. The principal objective of active labour market policy is to re-integrate unemployed people in the labour market. To obtain this goal, unemployed individuals should be assigned to the most effective programme, taking into account the available ressources. In most countries the assignment is done by caseworkers. First results show that this is not always optimal and could be improved.
Microeconometric evaluation of programmes is able to estimate treatment effects on a very disaggregated level, even for single persons. These results can be used for predicting the best programme for persons on the basis of their individual characteristics. Statistically assisted programme selection (SAPS) is an expert system which predicts the effectiveness of different programmes and recommends a programme on the basis of these statistical results.
The goal of this project is to improve evaluation methods on a disaggregated level and to develop SAPS as a dynamic expert system for treatment assignment.