Final Activity Report Summary - LABOR POLICY WOMEN (Evaluating the Effects of Labour Market Policies on the Employment Prospects of Women) The European Employment Strategy is a joint effort of European countries to promote employment. A specific goal of this joint strategy, which rests on the concept of gender mainstreaming, is to increase the employment rate of women. In particular, the European Employment Guidelines which concretise the employment strategy define women who want to return to the labour market after a family break as a major target group which should be assisted in finding gainful employment. Since the systematic evaluation of employment and labour market policies is still not common practice in many member states, as mentioned by the European Commission in1998, little was known about which measures of active labour market policies (ALMP) would be successful instruments to achieve this goal. Therefore, this project acquired new insights on which programs were helpful to increase the employment prospects of women with a special emphasis on reintegrating women. A literature survey summarised the current knowledge on employment effects of ALMP for women and identified the most necessary and fruitful new lines of research. The few evaluation studies for Europe which distinguished between men and women along with the broad evidence from the United States showed that programmes which increased the direct attachment to the labour market could be effective tools to raise the employment prospects for women, including reintegrating women. However, this result was not unambiguous, given that, for example, some employment programs displayed negative effects. This raised questions over what was the optimal design of active labour market programmes. Recommendations concerning this were possible in case structural parameters were estimated, i.e. parameters which were time and space invariant. In order to receive such estimates I developed a search model where the decision to return to the labour market was included, as well as the availability of specific labour market programmes. The different programme types, such as training and employment programmes were considered, as well as their degree of closeness to the labour market. Moreover, psychological and sociological determinants were incorporated in the model. The model was accessible for estimation with the likelihood estimation technique. On the basis of empirical results policy conclusions concerning the optimal program design for reintegrating women could be given. This model was the most important scientific achievement of my project. In the third project part I investigated the employment effects of employment programmes with the conditional difference-in-differences approach which had recently become popular. Next to the commonly estimated treatment-on-the-treated effect for the actual participants I also estimated the potential effect on the average population, i.e. the population average treatment effect. Both were separately conducted for men and women. In order to allow the population average treatment effect to vary according to the time the treatment took place I proposed a method which set hypothetical treatment times for the nonparticipants. In accordance with previous findings, the estimates showed positive employment effects for women which were often significant, whereas the effects for men were sometimes negative. Another important finding was that, in case the group of employment programme considered wage subsidies for regular jobs, there seemed to have the most favourable effects for women. This was the case for the success criterion reemployment probability, and even more so for the probability of the remaining employed. A potential superiority of wage subsidies for men could only be observed for the success criterion probability of the remaining employed. Furthermore, the estimated average treatment effects for the entire population showed that, with respect to this parameter, women gained slightly more than men from participating in employment programmes. Finally, these average treatment effects indicated that a different targeting with respect to men, either in the direction of the average male population or towards more women, could have improved the average treatment effect.
