Analysis of youth underemployment in Macedonia, Montenegro and Serbia
The objective of the study is to describe underemployment and investigate its effect on personal wellbeing in the three countries. Special reference to gender will be made. By so doing, it opts to bring the issue to policymakers’ agenda and formulate ideas for policies tackling underemployment.
Methodologically, we first set underemployment a function of personal and local-labor-market characteristics. We rely on Heckman models – binary- and ordered-outcome ones – to overcome the potential selectivity bias into underemployment. For the ordered case, we use a CMP estimator. We then regress personal welfare on underemployment and worker’s characteristics. As welfare is latent variable, we rely on MIMIC model. To overcome potential endogeneity of underemployment, we specify third regression in the model whereby observables and exogenous instrument determine underemployment. The average local/municipal wage at the time when the person became (under)employed is used as instrument.
The School to Work Transition Surveys (2014-2015) are used.
Donor: Partnership for Economic Policies