Employment and unemployment in Italy: measurement issues and behavioural analyses


Research Project supported by the Italian Ministry of University and Scientific Research, 1999

[Home] [Versione italiana] [Printer-friendly version]

Projects | 1997 | 1999 | 2001 | 2003 | 2005 



The programme builds on an ongoing project, started in 1998, supported by a grant from MURST, involving the research units in Padova, Siena and Torino (and two other units not involved here). Besides, the unit in Padova has been working some ten years on measuring and modelling participation at work via data from surveys; the unit in Siena is experienced in 'school-to-work' transitions and living conditions analysis; the unit in Torino since early '80s pioneered the use of Italian Social Security files for economic analysis purposes (workers mobility, job turnover, wage mobility); the unit in Venezia started working on the Veneto case-study three years ago focusing on labour mobility as documented by Social Security files. In the following we place the programme in the relevant scientific background, with reference to each of the five broad topics we plan to deal with.
Potentials of and drawbacks from using administrative data sources for studying the labour market are by now well recognised. An up-to-date assessment is in Autori Vari (1998). In Italy most of the contribution to the issue came from the work by Contini and his associates on linked employer-employee Social Security files. See Contini and Revelli (1992) for a comprehensive review; an updated list of references is in the research proposal of the unit in Torino.
NETLABOR started as a pilot system run by the labour exchanges in the Veneto region. The resulting automatic protocol is currently used by some 70% of the Italian Labour Exchanges. Within the 1998 programme, we are completing an assessment of data quality and are reshaping the archive to enhance its use for economic analyses (Gambuzza and Rasera, 1999; Bassi, 1999). In particular, NETLABOR seems worthwhile to work with, because of three main reasons: (i) it yields a remarkable amount of information on jobs features; (ii) it provides quite updated, timely information; (iii) it allows us to establish the employer-employee links.
Measuring labour force participation, and in particular unemployment, is a long-standing issue on which the debate is far from being settled (Green, 1995, provides an assessment). Over the years, some members of the research unit in Padova have got a well established research experience on the issue (Trivellato, 1991), their recent research interests being on the analysis of labour market attachment (Rettore, Torelli and Trivellato, 1993; Rettore and Trivellato, 1993 and 1998). Besides, a supplement to the current Labour Force Survey has been designed within the 1998 programme, based on the 'time-use' methodology, to improve the measurement of participation at work and in particular to detect hidden participation at work (Campostrini, Giraldo, Parisi and Trivellato, 1999). The pilot survey will be run by ISTAT (the Italian central statistical agency) next fall.
As for the analysis of the dynamics of participation at work, attention will be focused on the following topics:
(i) unemployment duration. Recent contributions to the field by members of the research unit paid specific attention to model specifications consistent with the peculiar features of available information - heaping in reported durations, large fraction of long-term unemployed, etc. - (Torelli and Trivellato, 1993a e 1993b; Paggiaro, 1999);
(i) relationships between participation at work and poverty/social exclusion. Recent contributions to the field are Jarvis and Jenkins (1997) and Hills (1998). Contributions by members of the research unit are Rettore and Rizzi (1996) and Trivellato (1998).
(ii) evaluations of selected policies. Recent contributions to the field are Manski (1995) and Mohr (1995). Contributions by members of the research unit are Battistin, Gavosto and Rettore (1998) and Trivellato and Franceschini (1998).
School-to-work youth transitions will be studied by using longitudinal data from a survey on four cohorts of students attaining the compulsory level. A recent contribution to the field is Andrews and Bradley (1997). Contributions by members of the research unit in Siena are Ghellini and Trivellato (1996), Bernardi and Ghellini (1997) and Bernardi, Ghellini and Penello (1999).
The problem of combining data from complementary sources arises in several instances. At an aggregate level, calibration methods are needed to estimate a parameter of interest using data from one source subject to the constraints coming from information available from an independent source. A point in case is Steel (1997), which uses calibration methods to estimate employment and unemployment. Calibration methods have also been used to account for non-sampling errors (Skinner, 1998), and to estimate micro-econometric models in the presence of macro information (Imbens and Lancaster, 1994).
At the micro level, a relevant issue is the linkage of individual records from complementary sources. Within the 1998 programme, an experiment is in progress, jointly with researchers at ISTAT, to link individual records from the Labour Force Survey to the Social Security files.
The peculiar conditions and behaviour of the weak segments are a distinguishing feature of the Italian labour market, as compared to other EU countries. In Italy, individuals and households under the poverty line, and broadly speaking females and young people, are more hit by unemployment than in other countries.
As for the relationships between participation at work and poverty/social exclusion, recent contributions are Jarvis and Jenkins (1997) and Hills (1998). Contributions by members of the research units are Ghellini, D'Agostino and Neri (1996), Rettore and Rizzi (1996), Trivellato (1998) and Lemmi (1999).
As for segmentation of the labour market and discrimination of weak segments, see Biffl (1997) and Rubery, Fagan and Meier (1997). Contributions from members of the unit in Venezia are Occari, Tattara and Volpe (1997) and Tattara and Volpe (1999).

Info: anna.giraldo@unipd.it