Título: Essays on impact evaluation of labour market reforms using administrative data

Autora: Ana M. Silva

 
Directores: José María Arranz y Carlos García-Serrano
 
Universidad: Universidad de Alcalá

 
Fecha de defensa: 10/07/2026

 
Comité: Emilio Congregado, Mercedes Burguillo y Catalina Amuedo-Dorantes

 
Calificación: Sobresaliente Cum Laude, Mención Internacional
 
 

This dissertation is composed of three papers that analyse how two central institutions of the labour market—the minimum wage and the unemployment benefit system—influence employment dynamics and unemployment outcomes in Spain, a country characterised by high structural unemployment and marked cyclical sensitivity. Together, the three papers examine how reforms in these areas affect job retention, labour turnover and the duration of unemployment, taking into account both heterogeneity of workers and macroeconomic context.

The first two papers focus on the national minimum wage. The first paper evaluates the impact of the substantial 22.3% (21.6% in real terms) increase implemented in 2019 on job retention and employment permanence within the same firm. To this end, robust impact evaluation techniques are used—specifically, inverse probability weighted difference-in-differences—applied to individual administrative data from the Muestra Continua de Vidas Laborales (MCVL). The results show no negative effects on job retention. On the contrary, they reveal modest and short-lived positive effects—limited to the first months after the minimum wage increase—particularly in low-wage regions and among full-time workers as well as temporary workers.
 

The second paper extends the analysis to gross employment flows—hiring and separations—at the provincial level between 2015 and 2023. To this end, the MCVL administrative records are also used, but aggregated at the provincial level. In provinces with greater exposure to the minimum wage (i.e., with a higher percentage of workers earning the minimum wage), both hires and separations decrease, thus attenuating labour turnover. However, the effect on net employment is close to zero. The results differ according to working hours: turnover increases among part-time workers and decreases among full-time workers.

The third paper evaluates two legislative changes in the unemployment benefit system and their effects on the duration of unemployment. The first reform, implemented in 2012 during the Great Recession, consisted of a reduction in the replacement rate. The second, in 2023, resulted in an increase in that rate, taking advantage of the favourable economic context. This analysis is particularly relevant due to the use of administrative records showing the entire universe of unemployed individuals (from the Servicio Público de Empleo Estatal, SEPE) and the use of quasi-experimental techniques combined with duration models. The findings show that the reduction in the replacement rate in 2012 had a negative and significant effect on the duration of unemployment. However, the increase in the replacement rate in 2023 led to a moderate increase in duration. The effects are heterogeneous in both reforms: workers with higher benefits and shorter potential duration react more intensely, while for certain population subgroups the effects are negligible.

Overall, this dissertation provides robust causal evidence on recent and highly relevant reforms of the Spanish labour market and offers an integrated assessment of how institutions operate across the hiring, retention, and separation margins. The results underscore the importance of designing labour policies that are sensitive to the economic cycle and the heterogeneity of workers, providing relevant implications for both academic research and policy design in economies with high unemployment.