Selective Migration, Regional Growth and Convergence: Evidence from Italy

Abstract

Fratesi U. and Percoco M. Selective migration, regional growth and convergence: evidence from Italy, Regional Studies. This paper studies the link between regional disparities and migration flows, focusing on the skill content of migration. Disparities may lead to migration, which should reduce them; but at the same time if migration is skill-selective, it may have an opposite effect and reinforce the richest regions. The object of this empirical analysis is Italy, a country where unskilled interregional migration flows were a large and very well-known phenomenon during the 1950s and 1970s, whilst in recent years, after three decades of very low labour mobility, thousands of Southern graduates have been moving to Northern regions. What is the economic impact of those flows of selective migration? Using data covering the period 1980–2001, it is found that although a slight process of convergence occurred between Italian regions, the loss of human capital in the South was detrimental to regional growth.

Fratesi U. and Percoco M. 选择性移民，区域成长与聚合：来自意大利的证据，区域研究。本文研究区域不均与移民流动之间的连结，并聚焦移民的技术内容。不均会导致移民，并进而降低不均；但移民若同时是具有技术选择性的，则会出现相反的效应，反而强化了最为富裕的区域。本文的经验分析对象是意大利，该国在 1950 以及 1970 年代时，非技术移民的区域流动，是相当广泛且众所皆知的现象，但经历了三十多年来非常低的劳动流动性之后，近年来，数以千计的南方毕业生不断地迁徙至北部区域。这些选择性移民的流动具有什麽样的经济影响呢？本研究运用涵盖1980至2001年的资料，发现儘管意大利各区之间发生少部分的聚合过程，但南方的人力资本流失对于区域成长而言，是相当不利的。

Fratesi U. et Percoco M. La migration sélective, la croissance régionale et la convergence: des données obtenues en Italie, Regional Studies. Cet article étudie le lien entre les écarts régionaux et les flux de migration, focalisant le niveau de qualification des migrants. Il se peut que les écarts amènent à la migration, qui devrait les réduire. Mais, en même temps, si la migration s'avère moins axée sur la main-d'oeuvre qualifiée, il se peut qu'elle ait l'effet contraire et renforce les régions les plus riches. La présente analyse empirique a pour but l'Italie, un pays où les flux de migration interrégionaux des travailleurs non-qualifiés étaient un important phénomène très connu au cours des années 1950 et 1970, alors que ces dernières années, suite à trois décennies de ralentissement de la mobilité du travail, des milliers de diplômés en provenance du sud se sont déplacés à destination des régions du nord. Quel est l'impact économique de ces flux de migration sélective? À partir des données pour la période allant de 1980 jusqu’à 2001, il s'avère que la perte de capital humain dans le sud était préjudiciable à la croissance régionale, bien qu'un processus de convergence limitée ait eu lieu entre les régions italiennes.

Fratesi U. und Percoco M. Selektive Migration, regionales Wachstum und Konvergenz: Belege aus Italien, Regional Studies. In diesem Beitrag untersuchen wir den Zusammenhang zwischen regionalen Disparitäten und Migrationsströmen unter besonderer Berücksichtigung des Aspekts der Qualifikation im Zusammenhang mit der Migration. Disparitäten können zu Migration führen, woraufhin sie sich verringern sollten; wenn aber die Migration zugleich auf qualifikationsselektive Weise erfolgt, kann der gegenteilige Effekt eintreten, und es können die finanzkräftigsten Regionen gestärkt werden. Gegenstand dieser empirischen Analyse ist Italien, wo interregionale Ströme von unqualifizierten Migranten in den fünfziger und siebziger Jahren des 20. Jahrhunderts ein umfangreiches und wohlbekanntes Phänomen darstellten; nach drei Jahrzehnten mit einer äußerst geringen Mobilität der Arbeitnehmer sind in den letzten Jahren wiederum Tausende Hochschulabsolventen aus dem Süden in nördliche Regionen gezogen. Welche wirtschaftliche Auswirkung haben diese Ströme von selektiver Migration? Anhand von Daten aus dem Zeitraum von 1980 bis 2001 stellen wir fest, dass zwischen den italienischen Regionen ein geringfügiger Konvergenzprozess auftrat, wobei sich aber der Verlust von Humankapital im Süden nachteilig auf das regionale Wachstum auswirkte.

Fratesi U. y Percoco M. Migración selectiva, crecimiento regional y convergencia: el ejemplo de Italia, Regional Studies. En este artículo analizamos la relación entre las desigualdades regionales y los flujos de migración, prestando atención a la migración de trabajadores cualificados. Las desigualdades pueden llevar a emigrar y luego tal emigración debería reducirlas, pero al mismo tiempo, si la migración depende de las cualificaciones, podría tener el efecto contrario y reforzar a las regiones más ricas. Para este análisis empírico utilizamos Italia, un país donde los flujos migratorios interregionales de trabajadores no cualificados fueron un fenómeno a gran escala y bien conocido durante los años cincuenta y setenta, mientras que en los últimos años, tras tres décadas de muy baja movilidad laboral, miles de licenciados del sur se han desplazado a las regiones del norte. ¿Qué impacto económico tienen estos flujos de migración selectiva? A partir de datos entre 1980 y 2001, observamos que aunque ocurrió un ligero proceso de convergencia entre las regiones italianas, la pérdida de capital humano en el sur tuvo un efecto perjudicial en el crecimiento regional.

• Interregional migration
• Selective migration
• Regional growth
• Convergence

• 跨区域迁徙
• 选择性移民
• 区域成长
• 聚合

• Migration interrégionale
• Migration sélective
• Croissance régionale
• Convergence

• Interregionale Migration
• Selektive Migration
• Regionales Wachstum
• Konvergenz

• Migración interregional
• Migración selective
• Crecimiento regional
• Convergencia

JEL classifications:

• J, J6, J61, R, R1, R11

Acknowledgements

The authors wish to thank the participants at the ‘The Causes and Effects of Interregional Migration’ workshop (Alghero) and at the North American Meetings of the Regional Science Association (San Francisco); several referees for useful comments; and Marco Maffezzoli and Romano Piras for sharing their data. The usual disclaimer applies.

Notes

1. This is a long-lasting trend, and the same consideration was shared by Greenwood (1975).

2. However, these estimates are not adjusted by the cost of living, which is notably higher in the North than in the South.

3. Cingano and Cipollone (2009) found that the return on human capital investment in the South was higher. However, they used data from the Survey on Household Income and Wealth, while Guagnini and Mussida (2009) used data from the European Union Statistics on Income and Living Conditions (EU-SILC) survey, which is highly representative at the regional level (especially the 2006 wave, which was over-sampled in most of the Italian regions). In order to increase the reliability of the results, Cingano and Cipollone (2009) pooled the Survey on Household Income and Wealth waves between 1987 and 2000, an operation which may have considerably biased the estimates. On the other hand, Guagnini and Mussida (2009) used the EU-SILC 2004, 2005 and 2006 waves, but estimated the rate of return separately for each year.

4. For a complete review of the interregional migration issues, see Faggian and McCann (2009a).

5. For comprehensive studies on Italian international migration, see Audenino and Tirabassi (2008) and Del Boca and Venturini (2005).

6. For a review, see Ramella (2009).

7. Interestingly, a low level of schooling was a persistent feature also of second- and even third-generation migrants from southern regions (Ramella, 2009). This was probably due to poor neighbourhood conditions in terms of education supply, spatial segregation (Monfort and Ottaviano, 2002), and adverse selection in the labour market.

8. Descriptions of all the variables used in the paper are provided in Appendix A.

9. The fact that the human capital of out-migrants is higher than that of in-migrants is not a sufficient condition for migration to reduce the human capital per person of a region, when both flows are more skill-endowed than residents. In fact, the effect also depends on the absolute sizes of inward and outward flows, and there is anecdotal evidence that Lombardy, for example, is attracting a large number of skilled young persons from the rest of Italy and losing a relatively low number of highly skilled people to higher level technological poles, not rarely abroad.

10. The choice of including the lagged value of the explanatory variables was also made by Grogger and Hanson (2008) and Ortega and Peri (2009). However, both articles present the results from gravity models, while the present data allow only net migration in regions to be considered.

11. NUTS = Nomenclature des Unités Territoriales Statistiques.

12. The present measure of human capital gain from migration is slightly different from the measures proposed in equations (1) and (2). In these latter, and are expressed in terms of in- and out-migrants, whereas the dependent variable in equation (11) is in per capita terms. This change was necessary to make the variable consistent with the dependent variables (per capita GDP and its growth rate) used in the next section to estimate the effect of migration on regional growth.

13 Card's (2001) units of analysis are cities rather than regions.

14 Ortega and Peri (2009) propose a similar instrument with similar arguments, although they use origin–destination flows, which, in the present case, are not available for the time period considered.