Time to degree: students' abilities, university characteristics or something else? Evidence from Italy

Abstract

We use a representative sample of Italian graduates drawn from the Consorzio AlmaLaurea to assess the impact of individual and family characteristics, university inputs and the labour market on the time taken to attain a degree. Our estimates highlight that all these dimensions drive the outcome analysed. Weak labour market prospects contribute particularly to length time to degree. Our results suggest that a comprehensive policy intervention is needed to increase the number of students graduating within the minimum period.

JEL classifications: :

• J24
• I23

Keywords: :

• tertiary education
• time to degree
• labour market

Acknowledgements

We are grateful to Consorzio AlmaLaurea for their data set and assistance. We acknowledge financial support from the Fondazione Alfieri CRT and PRIN-2007. Opinions, conclusions or recommendations in this paper are those of the authors and do not necessarily reflect the views of the Consorzio AlmaLaurea and of the Fondazione Alfieri-CRT. We thank participants in SIEP (2009), SMYE (2009), AIEL (2009), IWAEE (2010), AEDE (2010) and a seminar at the Turin University. Individual thanks for useful comments are due to Gilberto Antonelli, Giuseppe Bertola, Massimiliano Bratti, Andrea Cammelli, Furio Camillo, Arnaud Chevalier, Damon Clark, Angelo di Francia, Giorgio Di Pietro, Ross Finnie, Aldo Geuna, Paolo Ghinetti, Aldo Goia, Gian Piero Mignoli and Simone Moriconi. The authors also gratefully acknowledge the comments of two anonymous referees. The usual disclaimers apply.

Notes

A comprehensive description of the university education system in Italy is beyond the scope of this paper, for overviews, see Perotti (2002) and Cappellari and Lucifora (2009).

According to the data on tuition fees for the academic year 2008/2009, 67% of Italian university students pay annual university fees of less than 1000 euros; 20% pay between 1000 and 1500 euros, and only 13% pay more than 1500 euros (MIUR Statistical Office). Tuition fees charged to students enrolled beyond the prescribed duration of their degree course are approximately a half of those charged to ‘regular’ students.

A 2009 reform of the Italian university funding system has introduced a monetary penalty according to the percentage of students enrolled beyond the legal duration.

See Table 1 for further details on the methodology used to define the dependent variable, and the potential problems arising from the use of an outflow sample.

It should be noted that in Italy students are considered ‘regular’ if they graduate by the last graduation session of their third academic year. According to our organization of graduation sessions, this means that those who graduate by the end of the second session should be considered regular, even if they were entitled to graduate in the summer session.

Each student is at risk of graduation from session 0 onwards.

There are several dimensions related to these regional fixed effects, for example, the quality of the high school that the student attended and potential structural inefficiencies characterizing certain areas of the country which also affect the university organization.

Under the assumption of ‘proportional hazard’, the duration profile of the hazard is the only function of the time variable and, therefore, is the same for all individuals where the profile is shifted upwards or downwards by the explanatory variables.

We performed several checks to test the robustness of our results. In particular, we ran the analysis using different measures of unemployment. However, since unemployment rates show great variability across regions, but not over time within regions during the sample period, all the tests confirmed the estimates.

According to the Italian Ministry of Education, some 80.4% of the students in the academic year 2007/2008 enrolled in a university in their home region.

The hazard ratio associated with a characteristic j is the relative risk of graduating for individuals with this characteristic in the reference group. For example, if the estimated hazard ratio is 0.6, then individuals with characteristic j have a 40% lower probability of graduating than the reference group; if the hazard ratio is 1.5 individuals with characteristic j have a 50% higher probability of graduating than the reference group.

For example, the quota of professors respect to the academic staff is about 67% and 63%, respectively, for the centre-north and south of Italy.

See Section 2.1.