Logo eprints

The use of administrative data: from the evaluation of the EPL effect to the analysis of the university graduates' labor market

Mondauto, Leopoldo (2013) The use of administrative data: from the evaluation of the EPL effect to the analysis of the university graduates' labor market. Advisor: Cambini, Prof. Carlo. Coadvisor: Pammolli, Prof. Fabio . pp. 165. [IMT PhD Thesis]

[img] Text
Mondauto_phdthesis.pdf - Published Version
Restricted to IMT staff and National library only

Download (2MB) | Request a copy
Text (Contents, Abstract, Introduction, Chapters 1, 4)
Mondauto_mixedphdthesis.pdf - Published Version

Download (787kB) | Preview


The statistical enhancement of administrative data sources has played a pivotal role in the analysis of the labor market in the last decades. They permit us to obtain a clearer picture of the reality than the common surveys. In so doing, they represent a cognitive instrument capable to support the decision-making strategies adequately and timely. This PhD thesis provides two case studies in this direction. Using unique administrative datasets for Italy, we deal with two important labor market issues. In the first part, we investigate the impact of employment protection on the composition of the workforce and worker turnover. Chapter 1 provides an overview of the main existing literature about the effect of EP on labor market. In Chapter 2, using a novel matched employer-employee administrative dataset, we adopt a regression discontinuity design (RDD) that exploits the variation in employment protection provisions in Italy between firms below 15 employees and those above 15 employees. Our newly dataset allows us to better identify the size of firms and most importantly the different typologies of labor contracts. In order to justify the use of the RDD approach we conduct three different tests. First, we show that firm-size density is actually continuous around the threshold, as proposed in McCrary (2008). Second, we follow Schivardi and Torrini (2008) and assess whether firms just below and above the threshold differ in their propensity to grow. Finally, we conduct a series of balancing tests to investigate to what extent firms just above and below the threshold differ in terms of the observable characteristics. Our RDD estimates suggest that EP increases worker reallocation, suggesting that EP tends to reduce rather to increase worker security. This can be explained by the fact that firms facing more stringent EP make a greater use of workers on fixed-term contracts. Indeed, the incidence of temporary work is 2-2.5 percentage points higher in larger firms. There is also evidence that EP reduces labor productivity. This effect is to an important extent due to the impact of EP on worker reallocation and the incidence of temporary work. In the second part, we investigate the effect of the final graduation mark on the graduates‟ probability to be hired by firms (Chapter 3). The analysis is based on a worker-level administrative dataset. By evaluating employment contracts on the basis of their expected duration, we estimate multilevel logit and multilevel ordered logit models to take into account the clustered structure of the data and the nature of the response variables. Using a random intercept and a random slope specifications, we find the existence of a non-trivial effect of the graduation mark on the probability to be hired, even after controlling for the problem of selection. Specifically, the effect is non-monotonic: it is positive below a certain threshold, and it becomes negative afterwards. In case of permanent contracts, the effect of the graduation mark is not significant, while the major chosen by the student plays a key role.

Item Type: IMT PhD Thesis
Subjects: H Social Sciences > HB Economic Theory
PhD Course: Economics, Markets, Institutions
Identification Number: 10.6092/imtlucca/e-theses/131
NBN Number: urn:nbn:it:imtlucca-27163
Date Deposited: 04 Feb 2014 14:05
URI: http://e-theses.imtlucca.it/id/eprint/131

Actions (login required, only for staff repository)

View Item View Item