Skip to main content

Advertisement

Log in

Structural differences across macroregions: an empirical investigation

  • Original Paper
  • Published:
Empirica Aims and scope Submit manuscript

Abstract

Using a macro–micro econometric framework that allows studying the labour market dynamics, this paper offers an in-depth investigation of the structures of both national and macro regional labour markets in Italy. The simulation results reveal structural differences between regions in the short as well as the long run. Regional gaps represent one of the main components of the natural unemployment rate in Italy. The results may help regional and national policy makers in the European Union to formulate strategies tailored to the specific needs of regional labour markets.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4

Similar content being viewed by others

Notes

  1. According to the standard classification and on the basis of the NUTS 1 codes, Italy can be divided into three main macro-areas: North (Northwest Italy—ITC code—and Northeast Italy—ITD), Centre (Central Italy—ITE) and South (South Italy—ITF—and Insular Italy—ITG).

  2. As explained in the Sect. 2, the Italian labour market is characterised by geographic discrepancies in labour market (and economic) indicators. For additional details on the statistics presented in the second section, see the online data warehouse of the ISTAT, http://dati.istat.it/. For details on the relevance of estimating a steady-state unemployment rate by geographic area in Italy, see Salituro and Scorcu (2001).

  3. For further details, see Attanasio and Padoa-Schioppa (1991) and Bodo and Sestito (1991).

  4. For details also see Manacorda and Petrongolo (2006).

  5. For details see Contini and Rapiti (1994).

  6. The Italian Association for the Development of Southern (SVIMEZ) is a non-profit organization established in 1946 whose primary task is the analysis of southern Italy’s economic situation in order to present policy actions aimed at stimulating industrial activities.

  7. GDP deflator gathered by ISTAT.

  8. For details on the Italian LFS definitions, see Brandolini et al. (2004) and ISTAT (2002).

  9. For a detailed discussion of the characteristics of the Italian LFS, see ISTAT (2006).

  10. The measure of employees used, i.e., full-time equivalent labour units, is more appropriate than the measure used in earlier models in which labour demand was specified in terms of hours worked, and the number of employees was derived through an estimate of hours per worker. This latter approach, as emphasised by the literature—e.g., Baussola (2007) for Italy, and Fauser (2011) for Germany, could not completely account for the labour-hoarding effect determined by, for instance, hours spent within the Cassa Integrazione Guadagni, which allowed workers to remain employed (instead of being laid off) when firms were restructuring their plants, particularly during recessions. These factors must be taken into account when analysing the Italian economy because in the 40-year period examined in this paper, there were two recessions (at the beginning of the 1990s and since 2008).

  11. It is worth highlighting that the decrease in the regional gaps and in general in the differences observed for gender and other characteristics may have been affected by a change in the Labour Force Survey methodology that occurred in 2004. This change significantly reduced the overall unemployment rate by excluding that part of the population that was discouraged and although available to work, did not consistently search for employment. For more detail, see Commission Regulation (EC) No. 1897/2000 of 7 September 2000 implementing Council Regulation (EC) No. 577/98 on the organisation of a labour force sample survey in the European Community concerning the operational definition of unemployment.

  12. Given the small number of observations (the time series dimension T of the sample is equal to 40), which is insufficient to obtain robust results in an application of time series tests, economic theory is used to directly derive relationships and to impose an error correction mechanism as an auxiliary adjustment rather than interpreting the results of the time series relationships modelled in the data (see Kennedy 2008, p. 307).

  13. The discouraged workers effect hypothesis is originally attributed to Tella (1964), who estimated a significant discouraged worker effect for both males and females in the US using employment-to-population rates as the explanatory variable. The hypothesis has been extensively used in the literature. For instance, Benati (2001) reviews numerous studies that attempted to estimate the discouraged worker effect, both at the aggregate level and for disaggregated sex-age groups.

  14. For a detailed description of the labour market transition probability matrix, see Baussola and Mussida (2014a).

  15. The features of the data employed in the present work (see the Sect. 2) allow the use of Markov chains models. Because individuals are observed at discrete time points, the exact transition times are not observed, and all that is known is the state occupied at the time of each scheduled survey. The Markov chain approach, as also explained by Aeschimann et al. (1999), allows one to describe the evolutions of the labour market transition probabilities for individuals who are not continuously observed over time.

  16. For example, for the state of (E)mployment, there are the permanence rate (ee) in the condition and two outflows, the transition from employment to unemployment (eu) and the transition from employment to inactivity (en). The same criteria apply for the states of (U)nemployment and (N)on-labour force or inactivity.

  17. A detailed technical description of the Maximum Likelihood method in this context can be found in Gourieroux (1989, ch. 5) and Cameron and Trivedi (2005, ch. 1).

  18. It would have been interesting to detail the investigation at a regional level, but there are problems of relative sample sizes (the sample sizes of smaller regions, e.g. Aosta Valley or Abruzzo, are too small to obtain statistically significant estimates).

  19. Obtained by multiplying the MNL coefficient estimates of the variable for each individual characteristic used in the model and their means.

  20. The MNL coefficient estimates of the regional unemployment rates multiplied by \(UR_t\), which is the unemployment rate (regional) computed by using identity (7.8) of the macro-level identities module of the model.

  21. In fact the behavioural equation residuals turn out to be significantly correlated. In this sense, a SUR estimation allows the estimation efficiency to be significantly improved (see also Fauser 2011).

  22. For details on the increase in the use of the Wage Supplementation Fund Scheme during the last crisis, see Addabbo et al. (2012) and Baldini and Ciani (2011).

  23. For the sake of brevity and illustrative purposes Tables 3 and 4 only report the coefficient estimates of MNL models for Italy and its main geographic areas for 2007–2008. The estimates for the overall period are available upon request.

  24. Education is divided into four levels: absence of or low education, lower secondary education, diploma or secondary education, and degree or tertiary education (baseline category).

  25. As in the macro model, the use of full-time equivalent labour units appears to be more appropriate than a simple head count of employees.

  26. Wald tests for the null hypothesis of the equality of coefficients across geographic areas and years were calculated.

  27. For detailed statistics on the Italian labour market, see the data warehouse of statistics produced by ISTAT, available on the Internet at http://dati.istat.it.

  28. OECD data confirm such evidence. These are available at http://stats.oecd.org.

  29. Disadvantaged workers are defined by the European Commission (2012) (Commission Regulation (EC) No. 2204/2012 of 12 December 2002 on the application of Articles 87 and 88 of the EC Treaty to State aid for employment [article 2]) as “any person who belongs to a category which has difficulty entering the labour market without assistance”.

  30. The simulations and the impacts of the policy exercise for all of the economic and labour market indicators, developed by the EViews software, are available upon request.

References

  • Addabbo T, García-Fernández R, Llorca-Rodríguez C, Maccagnan A (2012) Poverty and unemployment: the cases of Italy and Spain. In: Parodi G, Sciulli D (eds) Social exclusion short and long term causes and consequences. Springer, Berlin

    Google Scholar 

  • Aeschimann G, Antille G, Carlevaro F, Chaze JP, Ferro-Luzzi G, Flückiger Y, Gilli M (1999) Modelling and forecasting the social contributions to the Swiss Old Age and Survivor Insurance scheme. Swiss J Econ Stat 135:349–368

    Google Scholar 

  • Allen K, Stevenson A (1974) An introduction to the Italian economy. Martin Robertson & Co., Ltd., London

    Google Scholar 

  • Attanasio O, Padoa-Schioppa F (1991) Regional inequalities, migration and mismatch in Italy. In: Padoa-Schioppa F (ed) Mismatch and labor mobility. Cambridge University Press, Cambridge, pp 1980–1986

    Google Scholar 

  • Azmat G, Guell M, Manning A (2006) Gender gaps in unemployment rates in OECD countries. J Labour Econ 24:1–37

    Article  Google Scholar 

  • Bachman R, Bechara P, Kramer A, Rzepka CJ (2015) Labour market dynamics and worker heterogeneity during the great recession: evidence from Europe. IZA discussion paper, vol 9233. Bonn, Germany

  • Baldini M, Ciani E (2011) Inequality and poverty during the recession in Italy. Polit Econ J Econ Policy 3:297–322

    Google Scholar 

  • Barbieri L (2010) The Italian labour market. A ‘global’ regional econometric model. VDM - Verlag Dr. Muller, Saarbrucken

  • Barbieri L, Baussola M, Mussida C (2013) A macro–micro econometric model for the Italian labour market. Riv Int Sci Soc 1:5–30

    Google Scholar 

  • Barnichon R, Nekarda CJ (2012) The ins and outs of forecasting unemployment: using labor force flows to forecast the labor market. Brook Pap Econ Act 1:83–131

    Article  Google Scholar 

  • Baussola M (2007) Modelling a regional economic system: the case of lombardy. Reg Stud 41:19–38

    Article  Google Scholar 

  • Baussola M, Fiorito R (1994) Regional unemployment in Italy: sources and cures. J Policy Model 16:497–527

    Article  Google Scholar 

  • Baussola M, Mussida C (2014a) Transitions in the labour market: discouragement effect and individual characteristics. Labour 28:209–232

    Article  Google Scholar 

  • Baussola M, Mussida C (2014b) Disadvantages workers in the Italian labour market: gender and regional gaps. In: Malo MA, Sciulli D (eds) Disadvantaged workers: empirical evidence and labour policies. Springer, Berlin

    Google Scholar 

  • Benati L (2001) Some empirical evidence on the “discouraged worker” effect. Econ Lett 70:387–395

    Article  Google Scholar 

  • Bettio F, Corsi M, D’Ippoliti C, Lyberaki A, Samek Lodovici M, Verashchagina A (2012) The impact of the economic crisis on the situation of women and men and on gender equality policies. European Commission

  • Bertola G, Garibaldi P (2003) The structure and history of Italian unemployment. Journal of Policy Modeling, CESifo working paper series no. 907

  • Blanchard O, Jimeno JF (1995) Structural unemployment: Spain versus Portugal. Am Econ Rev 85:212–218

    Google Scholar 

  • Bodo G, Sestito P (1991) Le vie dello sviluppo. II Mulino, Bologna

    Google Scholar 

  • Brandolini A, Cipollone P, Viviano E (2004) Does the ILO definition capture all unemployment? Temi di discussione Banca d’Italia No. 529

  • Brunello G, Lupi C, Ordine P (2001) Widening differences in Italian regional unemployment. Labour Econ 8:103–129

    Article  Google Scholar 

  • Cameron AC, Trivedi PK (2005) Microeconometrics: methods and applications. Cambridge University Press, New York

    Book  Google Scholar 

  • Checchi D, Peragine V (2010) Inequality of opportunity in Italy. J Econ Inequal 8:429–450

    Article  Google Scholar 

  • Contini B, Rapiti F (1994) Young In Old Out: nuovi pattern di mobilità nell’economia Italiana. Lavoro e Relazioni Industriali 1:51–76

    Google Scholar 

  • Daly MC, Hobijn B, Sahin A, Valletta RG (2012) A search and matching approach to labor markets: did the natural rate of unemployment rise? J Econ Perspect 26:3–26

    Article  Google Scholar 

  • de la Rica S, Rebollo-Sanz YF (2015) Gender differentials in unemployment ins and outs during the great recession in Spain. IZA discussion paper, vol 9135. Bonn, Germany

  • Elsby MWL, Hobijn B, Sahin A (2013) Unemployment dynamics in the OECD. Rev Econ Stat 95:530–548

    Article  Google Scholar 

  • European Commission (2008) Employment in Europe 2008

  • Fauser S (2011) Modeling regional labor markets in Germany: insights not only for Germany policy makers. Empirica 38:169–201

    Article  Google Scholar 

  • Gourieroux C (1989) Econométrie des variables qualitative, 2nd edn. Economica, Paris

    Google Scholar 

  • Graziani A (1980) L’Economia Italiana dal 1935 ad oggi. II Mulino, Bologna

    Google Scholar 

  • ISTAT (2002) Forze di Lavoro, Media 2001, Annuario no. 7

  • ISTAT (2005) Conti Economici Regionali (Anni 1980–2004), Annuari

  • ISTAT (2006) La rilevazione sulle Forze di Lavoro: contenuti, metodologie, organizzazione, Metodi e Norme no. 32

  • ISTAT (2012) Conti Economici Regionali (Anni 1995–2012), Annuari

  • Kennedy P (2008) A guide to econometrics. Blackwell, London

    Google Scholar 

  • Limosani M (2004) Beyond regional institutions: widening unemployment differentials in Italy. Labour 18:503–514

    Article  Google Scholar 

  • Manacorda M, Petrongolo B (2006) Regional mismatch and unemployment: theory and evidence from Italy, 1977–1998. J Popul Econ 19:137–162

    Article  Google Scholar 

  • OECD (2007) OECD Economic Outlook 2007

  • Ricciardi L (1991) La disoccupazione di lunga durata in Italia: un’analisi dell’evidenza empirica nel periodo 1977–1989. Econ Lavoro 25:69–94

    Google Scholar 

  • Şahin A, Song J, Topa G, Violante GL (2011) Measuring mismatch in the U.S. labor market. Federal Reserve Bank of New York, New York

  • Salituro B, Scorcu AE (2001) Il tasso di disoccupazione di equilibrio nelle regioni Italiane. Polit Econ 2:129–158

    Google Scholar 

  • SVIMEZ (1998) I Conti economici delle regioni italiane. Il Mulino, Bologna, Italy

  • Tella A (1964) The relation of labor force to employment. Ind Labor Relat Rev 17:454–469

    Google Scholar 

  • Tronti L, Gatto R (2012) Measuring the long wave. Unemployment, discouragement and semi-employment in Italy, during and after the crisis. In: Parodi G, Sciulli D (eds) Social exclusion short and long term causes and consequences. Springer, Berlin

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Laura Barbieri.

Appendix: Variables and equations

Appendix: Variables and equations

See Table 5.

Table 5 List of variables

1.1 Macro block

1.1.1 Stochastic equations

$$\begin{aligned} \Delta \log (EEIND)_{i,t} &= c_i(1)+c_i(2)\Delta \log (EEIND)_{i,t-1}+c_i(3)\Delta \log (VAIND)_{i,t} \\&\quad+\,c_i(4)\Delta \log (WIND)_{i,t}+c_i(5)\log (EEIND)_{i,t-1} \\&\quad+\,c_i(6)\log (VAIND)_{i,t-1}+c_i(7)\log (WIND)_{i,t-1}+\varepsilon _{i,t} \end{aligned}$$
(7.1)
$$\begin{aligned} \Delta \log (EESER)_{i,t}&= {} c_i(8)+c_i(9)\Delta \log (EESER)_{i,t-1}+c_i(10)\Delta \log (VASER)_{i,t} \\&\quad+\,c_i(11)\Delta \log (WSER)_{i,t}+c_i(12)\log (EESER)_{i,t-1} \\&\quad+\,c_i(13)\log (VASER)_{i,t-1}+c_i(14)\log (WSER)_{i,t-1}+\nu _{i,t} \end{aligned}$$
(7.2)
$$\begin{aligned} \Delta \log (PR)_{i,t}&= {} c_i(15)+c_i(16)\Delta \log (PR)_{i,t-1}+c_i(17)\Delta \log (SERATE)_{i,t} \\&\quad+\,c_i(18)\Delta \log (EERATE)_{i,t}+c_i(19)\Delta \log (IMMIG)_{i,t}+c_i(20)\log (PR)_{i,t-1} \\&\quad+\,c_i(21)\log (SERATE)_{i,t-1}+c_i(22)\log (EERATE)_{i,t-1}+c_i(23)\log (IMMIG)_{i,t-1}+\zeta _{i,t} \end{aligned}$$
(7.3)
$$\begin{aligned} \Delta \log (SE)_{i,t}&= {} c_i(15)+c_i(16)\Delta \log (SE)_{i,t-1}+c_i(17)\Delta \log (PROFSE)_{i,t} \\&\quad+\,c_i(18)\Delta \log (YUR)_{i,t}+c_i(20)\log (SE)_{i,t-1} \\&\quad+\,c_i(21)\log (PROFSE)_{i,t-1}+c_i(22)\log (YUR)_{i,t-1}+\xi _{i,t} \end{aligned}$$
(7.4)

1.1.2 Identities

$$TE_{i,t}= {} EEIND_{i,t} + EESER_{i,t} + EEOTH_{i,t} + SE_{i,t}$$
(7.5)
$$TEI_{i,t}= {} \gamma TE_{i,t}$$
(7.6)
$$LF_{i,t}= {} PR_{i,t} * POP_{i,t}$$
(7.7)
$$UR_{i,t} = {} \frac{LF_{i,t}-TEI_{i,t}}{LF_{i,t}}*100$$
(7.8)
$$\begin{aligned} PROFSE_{i,t}=\frac{PROF_{i,t}}{SE_{i,t}} \end{aligned}$$
(7.9)

1.2 Micro block

$$\begin{aligned} pne_{i,t}&= {} \frac{ne_{i,t}}{ne_{i,t}+nu_{i,t}} \end{aligned}$$
(7.10)
$$\begin{aligned} ue_{i,t}U + ne_{i,t}N&= {} (eu_{i,t} + en_{i,t})E \end{aligned}$$
(7.11)
$$\begin{aligned} eu_{i,t}E + nu_{i,t}N&= {} (ue_{i,t} + un_{i,t})U \end{aligned}$$
(7.12)
$$\begin{aligned} N&= {} \frac{(eu_{i,t} + en_{i,t})}{ne_{i,t}E}- \frac{ue_{i,t}}{ne_{i,t}U} \end{aligned}$$
(7.13)
$$\begin{aligned} N&= {} \frac{-eu_{i,t}}{nu_{i,t}E}- \frac{(ue_{i,t}+un_{i,t})}{nu_{i,t}U} \end{aligned}$$
(7.14)
$$\begin{aligned} e_{i,t}E&= {} d_{i,t}U \end{aligned}$$
(7.15)
$$\begin{aligned} e_{i,t} &= {} eu_{i,t}+(1-pne_{i,t})*en_{i,t} \end{aligned}$$
(7.16)
$$\begin{aligned} USS_{i,t}&= {} \frac{e_{i,t}}{e_{i,t}+ue_{i,t}+un_{i,t}*pne_{i,t}}*100 \end{aligned}$$
(7.17)
$$\begin{aligned} DIFFUR_{i,t}&= {} UR_{i,t} - USS_{i,t} \end{aligned}$$
(7.18)
$$\begin{aligned} num\_ue_{i,t}&= {} \exp \left( \alpha ^{[ue]}_{i,t} + \beta ^{[ue]}_{i,t} UR_{i,t}\right) \end{aligned}$$
(7.19)
$$\begin{aligned} num\_un_{i,t}&= {} \exp \left( \alpha ^{[un]}_{i,t} + \beta ^{[un]}_{i,t} UR_{i,t}\right) \end{aligned}$$
(7.20)
$$\begin{aligned} ue_{i,t}&= {} \frac{num\_ue_{i,t}}{num\_ue_{i,t}+num\_un_{i,t}+1} \end{aligned}$$
(7.21)
$$\begin{aligned} un_{i,t}&= {} \frac{num\_un_{i,t}}{num\_ue_{i,t}+num\_un_{i,t}+1} \end{aligned}$$
(7.22)
$$\begin{aligned} p_{ab, t(h)}&= {} Pr = (X_{t,h}= b|X_{t-1,h} = a,z_{t,h}) , \end{aligned}$$
(7.23)
$$\begin{aligned} P_{ab, h}&= {} \frac{\exp {z^h_t\beta _b}}{\sum ^2_{l=0}\exp {(z^h_t\beta _l)}}, \end{aligned}$$
(7.24)

1.3 Micro–macro block link

$$\begin{aligned} ab_{i,t} = \frac{exp(\alpha ^{[ab]}_{i,t} + \beta ^{[ab]}_{i,t} UR_{i,t})}{exp(\alpha ^{[ab]}_{i,t} + \beta ^{[ab]}_{i,t} UR_{i,t}) + exp(\alpha ^{[ac]}_{i,t} + \beta ^{[ac]}_{i,t} UR_{i,t}) + 1} \end{aligned}$$
(7.25)

where \(i={\textit{Italy, North, Centre, South}}\) and \(t=1970,1971,\dots ,2010\).

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Barbieri, L., Mussida, C. Structural differences across macroregions: an empirical investigation. Empirica 45, 215–246 (2018). https://doi.org/10.1007/s10663-016-9356-0

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10663-016-9356-0

Keywords

JEL Classification

Navigation