Back to overview
labour; wages; skills; employment; education
Lay Summary (Italian)
Questo progetto si propone di analizzare come le persone scelgono il proprio percorso formativo e come queste scelta influenzano di conseguenza le loro competenze e il loro successo sul mercato del lavoro.
Questo progetto si propone di analizzare come le persone scelgono il proprio percorso formativo e come queste scelta influenzano di conseguenza le loro competenze e il loro successo sul mercato del lavoro. Ci soffermeremo in particolare sul confronto tra percorsi formativi di carattere scientifico (STEM) e umanistico. Grazie all'utilizzo di una speciale banca dati dove abbiamo informazioni sulle competenze linguistiche e matematiche di un campione rappresentativo di residenti in Italia, riusciremo a studiare l'effetto delle scelte d'istruzione sul processo di sviluppo di queste competenze. Avendo inoltre nella stessa banca dati anche informazioni su occupazione e salari, riusciremo anche a calcolare l'effetto dell'istruzione e delle competenze su questi processi.
L'analisi sarà svolta separatamente per maschi e femmine e permetterà di capire perché in molti paesi sembra esserci una scarsità di laureati in materie scientifiche, soprattutto tra le donne e nonostante una domanda apparentemente molto forte per questo tipo di figure.
Parole chiave: istruzione, competenze, occupazione, salari.
Responsible applicant and co-applicants
A large literature in labour economics and in a number of other disciplines explores the link between education and labour market success. The most natural way through which such a link is established is by education improving the human capital of the workers. Unfortunately, measuring human capital is problematic and, as a consequence, a number of important empirical and policy questions remain unanswered. For example, while demand for STEM (Science, Technology, Engineering and Mathematics) graduates seems strong, in many countries students, especially girls, prefer to enrol in non-STEM programs. Understanding why this happens requires investigating the effects of STEM education on skills and in turn how these skills are valued in the market. In this project we aim at answering these and a variety of other important questions by analysing a unique dataset of Italian workers containing detailed information on education and labour market performance, combined with direct tests of literacy and numeracy skills and measures of college proximity. The data come from the original sample of the Italian OECD-PIAAC (Programme for the International Assessment of Adult Competences) survey that, thanks to a formal agreement with INAPP (Istituto Nazionale per l'Analisi delle Politiche Pubbliche), the Italian institute responsible for the national PIAAC survey, have been merged with an historical archive of Italian universities (the Histories of Italian Universities - HIU). In these data we observe for each respondent the usual labour market variables, such as employment and wages, but also their tested literacy and numeracy and the distance to the closest university that existed at the time when they completed compulsory education. In addition, we can distinguish whether the closest university offered programs in STEM or not.We develop a model of educational and occupational choice, where college proximity can be used to generate arguably exogenous variation in the cost of attending college and taking a STEM or non-STEM subject. The model allows different educational tracks to affect the development of skills, both literacy and numeracy and a variety of other observable and unobservable skills. In turn, skills feed into productivity and wages. Individuals optimally choose their schooling and occupation based on expectations about employment probabilities and wages. The model can be structurally identified with our data using a combination of instrumental variable and maximum likelihood techniques.The results of this estimation exercise will allow us to shed light on a variety of important issues. For instance, we will be able to say whether the low enrolment rates in STEM subjects are due to the geographical distribution of universities offering STEM programs or from the fact that the labour market returns for these subjects are not large enough to compensate for the effort required to obtain a STEM degree. By estimating the model separately for men and women, we will also contribute to the debate on gender differences in educational choices, skills and labour market outcomes. Finally, we will also be able to perform policy experiments manipulating the geographical distribution of universities and their offer of STEM programmes.