The main research question is: Does participation in pre-vocational measures (short: pre-voc measures) influence the probability of subsequently entering regular vocational education and training (VET) programs and if so, why? For whom do these measures generate new opportunities or reinforce disadvantages, and which measures do so? To study this research question properly, we have to unravel the relationship between selection into different pre-voc programs, individual’s development during participation in pre-voc programs, and the variation in participant’s subsequent success of entering apprenticeships.
Hence, we need to examine the following sub-questions. (1) Who enters pre-voc programs after leaving school, and who enters which program type? (2) Do pre-voc measures indeed initiate a positive development of low-achieving youth’s individual skills characteristics? And if so, in which types of pre-voc measures do we find such improvement, and for whom? Taking selection and heterogeneity into account (question 1), our main research question on variation in the returns to participation in pre-voc measures, too, includes sub-questions: Do the chances of entering regular VET programs (and of higher-level VET programs) after pre-voc participation vary between pre-voc participants? And why do these chances vary between pre-voc participants—because of differences in pre-voc program types (differences in signaling values), differences in skills developments during participation (question 2), and/or differences in the extent to which pre-voc measures offer opportunities for gaining work experience (firm linkage)? In addition to these variations in returns to pre-voc measures within the group of participants, we will also study if participation in pre-voc measures generates scar effects—and ask: Do participants’ chances of entering regular VET programs (and of entering higher-level VET programs) differ from the chances of a comparable group at the time of leaving school?
We will study these transitions for youth who left regular schools and special schools for learning disabilities without a degree (subgroup 1) and those who only completed a lower secondary school degree (subgroup 2). We will use the data of NEPS starting cohort 4. We will employ ideas of the counterfactual approach to causal analysis. This enables us to isolate the impact of pre-voc measures on developmental trajectories of individual characteristics and on VET entry chances, respectively. We will use Coarsened Exact Matching (CEM), A
reatment effects on the respective T
reated (ATT), and Difference-in-Difference (DiD) estimations. In cooperation with the IAB, the individual NEPS data will be linked with fine-graded, time-varying regional information (at the level of district codes/Gemeindekennziffern) on apprenticeship and entry labor markets in order to appropriately account for selectivity in transitions into pre-voc measures.