The fight against unemployment remains a top priority in Europe: in most countries it is high and shows little sign of falling. An examination of different countries shows clearly that a large number and wide variety of policies have been implemented. Moreover, a new drive followed the Luxembourg Job Summit of November 1997, which proposed targets for EU member countries in terms of participants in various training programmes. In view of the large costs of operating these programmes it is of paramount importance that sound research is conducted in the field of policy evaluation to provide a framework for identifying best practices within the current menu of policies.

Although there has been a revolutionary shift in the methodology used in evaluating social programmes in recent years, especially in the US, little is known about the effectiveness of these programmes in Europe. Part of the explanation for why Europe lags behind the US in this area of research has been the lack of adequate data sources where experimental designs that control for selection biases can be developed. As is well known, endogenous choice of programme participation can lead to incorrect scientific inferences and very misleading policy conclusions.

Fortunately, the situation in Europe is changing. A growing number of researchers are becoming well trained in applying the new methodology to both experimental and non-experimental data. New micro datasets, with detailed information on unemployment experiences, programme participation and earnings, are increasingly becoming available. This project brings together eight leading institutions from seven EU countries to undertake five connected lines of research related to the evaluation of labour market policies: youth programmes and transition from school to work; training programmes and adult education; unemployment insurance; subsidies and tax breaks for the less skilled; and working time reductions. The teams in this network include the leading researchers in Europe in the field, both in terms of the new methodology and in the manipulation of the new datasets. The network will thus build a strong community of researchers trained in state-of-the-art techniques and with the capability of exploiting the data now available.