ABSTRACT
In this chapter, the authors present a joint Markov model for employment and poverty, conditional on educational attainment. They estimated in several stages. The authors observed schooling levels are modelled using an ordered logit model. They consider the state’s employment and unemployment and estimate them with a state-dependent logit model. The authors resort to ex-post microsimulation of three basic strategies against poverty: increasing the coverage of the minimum income, activation of the unemployed poor and raising the educational level of vulnerable groups. They estimate Markov model, they simulate the impact of these anti-poverty policies for the representatives of the respective target groups present in our sample over the period they were observed. Given the high degree of mobility into and out of poverty, the net effects of anti-poverty measures appear to be much smaller than a static model would predict. Moreover, depending on the type of policy adopted, long-term effects may be much greater or smaller than short-term effects.
