A data mining approach for the monitoring of active labour market policies

The paper addresses the problem of evaluation of the effectiveness of Active Labour Policies in the province of Bologna, a manufacturing district in Northern Italy, during the period 2004/2006. Using surviving analysis through Kaplan Meier filter and a new approach to propensity score computation, the Authors shows that the policies run by the Labor Market Authorities are able to compensate the disavatanges that secondary labor forces such as migrants, old age or less educated workers have in getting a job when fired. Moreover, they put new light on the transitions from temporary job to permanent jobs, and show that the probability of transitions is very low.