SONEDDA DANIELA

Early-career training contracts are increasingly used by governments to promote skill acquisition, yet take-up rates remain low. We examine how the type of labour contracts and the process of learning about workers’ productivity shape lifetime earnings and utility. Exploiting a reform in Italy that encouraged the use of training contracts, we provide reduced-form evidence on its impact on the types of contracts available to workers at the start of their careers. These estimates, together with other empirical moments, are then used to calibrate a structural model via the simulated method of moments (SMM). We also estimate a dynamic model of labour supply in which workers update their beliefs about productivity through Bayesian learning. Our results indicate that early-career training contracts have persistent effects on lifetime earnings and utility, driven primarily by the information revealed through learning. Such contracts benefit workers with intermediate or high productivity, but may penalise low-productivity workers, as the learning component can reveal their true type.

Leveraging a major Italian reform enacted in July 2012 that eroded employment protection to workers on permanent contracts, we use detailed worker level administrative data to estimate how this reduction affected the cost of job loss. We employ a stacked event-study research design, comparing treated and untreated workers as they move from employment into nonemployment. Weakening employment protection led to additional penalties in terms of lower re-hiring earnings and lower re-employment probabilities. Heterogeneous effects of the reform deepened pre-existing divides, penalizing labor market outsiders, such as young workers, and curbing employment opportunities for individuals in poorer regions, such as the South.

This paper assesses the impact of auditing final high school grades on educational and labour market outcomes, leveraging a 2007 reform in Italy that introduced external examiners. We extend the combined fixed effects method to estimate how the reform affected the earnings returns to higher education. We find that 90\% of the reform effect is across occupations and larger for women. We show this impact was mediated by changes in higher education choices, unlocking opportunities in occupations and earnings they wouldn't have had otherwise. We conclude that grade inflation related to internal teacher assessment has detrimental effects on both educational and labour market outcomes.

We assess how coal phaseout affects local employment. After the industry’s rapid dissolution in Great Britain (GB) post-1980, the number of total jobs rebounded in many affected areas -- defined as those with an initial high coal employment share -- even surpassing pre-phaseout levels despite the loss of coal jobs. Notably, 47\% of the affected areas saw employment growth in the four decades post-phaseout outpacing the GB average. These findings indicate that a significant portion of affected areas successfully adapted to the coal phaseout’s negative impact. We then explore why certain affected areas rebounded, while others did not. Our analysis reveals that the impact of coal phaseout on local employment varied based on how close areas were to tertiary education institutions (TEIs). In the decade prior to the coal phaseout, employment growth rates were comparable between areas near and far from TEIs. However, following the phaseout places closer to TEIs saw markedly faster employment growth. The growth of advanced manufacturing until the late 1990s, combined with the rise of skill-intensive service sectors, explains this trend by offsetting employment declines in the coal industry.

We examine how agglomeration economies shape the effects of training-based labour market policies. Exploiting the 2012 Italian Fornero Reform and a sharp age-eligibility cutoff for training contracts, we implement a difference-in-discontinuity design using administrative job-flow data from Piedmont. This approach isolates the causal impact of the reform by comparing workers just below and just above the eligibility threshold before and after the policy change. We find pronounced spatial heterogeneity: the effects are concentrated in thick local labour markets characterised by high firm density and strong occupational pooling. These results provide causal evidence that matching efficiencies, worker complementarities, and coordination advantages in dense labour markets amplify the effectiveness of training policies, highlighting a key interaction between labour market institutions and urban agglomeration forces.

We study the interaction between need- and merit-based university grants in a non-selective higher education system. Using administrative data from a northern Italian university, we analyse how eligibility criteria affect enrolment, academic performance, and labour market outcomes. We document a trade-off between the two criteria, with merit requirements acting as endogenous screening. We rationalise this trade-off with a three-period model predicting that merit thresholds increase effort among students with higher expected ability but may discourage effort among students at risk of falling short, as losing the grant reduces expected utility. We support these predictions using a difference-in-differences estimator for multiple treatments, separately analysing students switching into and out of need- and merit-based eligibility. Our results show that grants target disadvantaged but academically strong students, generate perverse incentive effects that vary by gender, and fail to retain a substantial share of initial recipients.