Role of Industrial District Externalities in Export and Value-added Performance: Evidence from the Population of Italian Firmsстатья из журнала
Аннотация: Abstract Becchetti L., De Panizza A. and Oropallo F. (2007) Role of industrial district externalities in export and value-added performance: evidence from the population of Italian firms, Regional Studies 41, 601–621. The paper tries to shed light on the export and value-added performance of industrial districts by integrating information on the population of Italian firms (from the ASIA database of the Italian Statistical Institute) with balance sheet data for all limited-liability companies in selected manufacturing industries. The findings show that firms located in industrial districts (specialized or not in the leading manufacturing activity of the district) export more and have higher added value than firms located elsewhere, net of the impact of relevant controls. The robustness check undertaken on the agglomeration–performance nexus detects, as expected, a significant relationship between dichotomous district indicators and (global and industry-specific) continuous measures of agglomeration and specialization. Export and value-added performance results are confirmed when proxying location in industrial districts with these continuous variables and when considering the weight of subcontracting and outsourcing activities of productive units as additional controls. Becchetti L., De Panizza A. et Oropallo F. (2007) Le rôle des effets externes industriels dans les performances des exportations et de la valeur ajoutée: des preuves auprès des entreprises italiennes, Regional Studies 41, 601–621. Cet article cherche à éclaircir les performances des exportations et de la valeur ajoutée des districts industriels en intégrant des données auprès du parc d'entreprises italiennes (la base de données ASIA de l'Institut italien de la statistique) à un bilan pour toute société à responsabilité limitée dans certains secteurs industriels. Il s'avère que les entreprises situées dans des districts industriels (qu'elles se spécialisent dans l'activité phare industrielle du district ou non) exportent plus et jouissent d'une valeur ajoutée plus élevée que ne le font les entreprises situées ailleurs, compte tenu de l'impact des contrôles en cours. Le contrôle de solidité fait sur la connexion agglomération-performances laisse voir, comme prévu, un important rapport entre des indicateurs dichotomiques des districts et des mesures d'agglomération et de spécialisation continues (à la fois globales et spécifiques). Si l'on assume des emplacements dans des districts industriels à partir de ces variables continues, et que l'on considère le poids de la sous-traitance et des approvisionnements externes des établissements productifs comme des contrôles supplémentaires, les performances des exportations et de la valeur ajoutée sont confirmées. Mesures d'agglomération géographique Districts industriels Performances des exportations Becchetti L., De Panizza A. und Oropallo F. (2007) Die Rolle der Externalitäten von industriellen Bezirken für die Export- und Wertsteigerungsleistung: Belege aus der Bevölkerung italienischer Firmen, Regional Studies 41, 601–621. In diesem Beitrag untersuchen wir die Exporte und die Wertsteigerungsleistung industrieller Bezirke, indem wir Informationen über die Bevölkerung von italienischen Firmen (aus der ASIA-Datenbank des Statistikinstituts Italiens) mit den Bilanzdaten aller Gesellschaften mit beschränkter Haftung in ausgewählten produzierenden Branchen kombinieren. Unsere Ergebnisse zeigen, dass Firmen in industriellen Bezirken (unabhängig davon, ob sie auf die führende produzierende Aktivität in diesem Bezirk spezialisiert sind) mehr exportieren und eine höhere Wertsteigerung erfahren als Firmen an anderen Standorten, auch wenn die Auswirkung der relevanten Kontrollen berücksichtigt wird. Die Robustheitsprüfung der Ballungs-Leistungs-Verknüpfung zeigt wie erwartet eine signifikante Beziehung zwischen den dichotomen Bezirksindikatoren und den (globalen und branchenspezifischen) kontinuierlichen Maßstäben der Ballung und Spezialisierung. Die Ergebnisse hinsichtlich des Exports und der Wertsteigerungsleistung werden bestätigt, wenn man den Standort in industriellen Bezirken durch diese kontinuierlichen Variablen ersetzt und das Gewicht der Zulieferung und externen Vergabe von produktiven Einheiten als zusätzliche Kontrolle berücksichtigt. Geografische Ballungsmaßstäbe Industrielle Bezirke Exportleistung Becchetti L., De Panizza A. y Oropallo F. (2007) El papel de los efectos externos en comarcas industriales para el rendimiento de exportación y de valor añadido: ejemplo de la población en empresas italianas, Regional Studies 41, 601–621. Con este ensayo pretendemos arrojar luz sobre el rendimiento de exportación y valor añadido de las comarcas industriales al combinar información sobre la población de las empresas italianas (base de datos ASIA del Instituto Italiano de Estadísticas) con datos de balances financieros para todas las sociedades de responsabilidad limitada en selectas industrias manufactureras. Nuestros resultados indican que las empresas ubicadas en comarcas industriales (especializadas o no en actividades líderes de fabricación en la comarca) exportan más y tienen un valor añadido superior que las empresas ubicadas en otros lugares, aunque se tenga en cuenta el impacto de los controles pertinentes. Al comprobar la solidez de los nexos entre aglomeración y rendimiento, se detecta, tal como se preveía, una relación significativa entre los indicadores dicotómicos de comarcas y las medidas continuas (para una industria o para todas) de aglomeración y especialización. Se confirman los resultados del rendimiento de la exportación y del valor añadido al sustituir la ubicación en las comarcas industriales por estas variables continuas y al considerar el peso de los subcontratos y servicios externos de gestión de las unidades productivas como controles adicionales. Medidas de aglomeración geográfica Comarcas industriales Rendimiento de exportación Keywords: Geographical agglomeration measuresIndustrial districtsExport performance Notes 1. 'Industry's secrets are ceasing to be secrets: they are, as it where, in the air and children are unwittingly learning many of them. Work well done is immediately recognized and people discuss right away the merits of inventions and improvements made to machines, processes and the general organisation of industry: if somebody comes up with a new idea, it s at one taken over by others and combined with their own home-made suggestions; it this becomes a source of other new ideas' (Marshall, Citation1920, p. 148). 2. Two typical Italian districts with this structure are those of Matera (furniture) with Natuzzi and of Montebelluna (sport shoes) with Lotto and Diadora as leader firms (Paniccia, Citation1998, Citation2002). 3. Models of patent races for breakthrough innovations are typical examples of this kind (Dasgupta and Stiglitz, Citation1980; Flaherty, Citation1980; Beath and Ulph, Citation1990; Futia, Citation1980). 4. The empirical findings of this paper illustrate that, between 1989 and 1992, district firms had a preferential relationship with local banks, which reduces loan rates and the likelihood of credit rationing. This relative advantage in the access to external finance vanished in the post-1992 recession period. 5. With a factor analysis, Paniccia Citation(2002) reclassifies Italian IDs into four types of clusters: craft/based or urbanized/tertiarized; canonical or quasi canonical; concentrated and organizationally upgraded; and peripheral/embryonic or Shumpeterian. 6. LLSs are aggregations of elementary administrative units (municipalities) defined according to a methodology developed by ISTAT, Irpet and the University of Newcastle upon Tyne (UK), which takes into account Census data on commuting for work purposes. Municipalities are aggregated according to self-containment, contiguity and space–time relationships criteria. This methodology when applied to the 1991 Census data led to a partition of Italy into 784 LLSs. 7. These are as follows. (1) An index of manufacturing activity (manufacturing employment as a share of total employment) 30% higher than the national average. Regions with an employment average lower than the national average are allowed to consider their own average. (2) An index of entrepreneurial density in manufacturing (local units/resident population ratio) higher than the national average. (3) An index of productive specialization (share of employment in a given manufacturing sector on total manufacturing) 30% higher than the national average (the sector is to be defined according to ISTAT classification, and the area has to be clearly administratively identified). (4) A share of employment in the manufacturing sector of specialization of at least 30% of total manufacturing employment of the area. (5) A share of employment in SME higher than 50% of total employment in the sector of specialization. 8. An administrative definition of IDs may contain upward and downward biases. On the one hand, regions might tend to overestimate the number of their own district areas (Piemonte identified 25 districts against the 16 found by ISTAT; and Lazio six against none) in order to maximize political consensus. On the other hand, other regions did not yet define any ID for administrative delays. 9. Iuzzolino Citation(2004) proposes an alternative algorithm for the identification of IDs. It does not start from LLSs; while it identifies a municipality with a significantly high level of industry specialization and builds endogenously around this municipality the IDs space. 10. The present paper adds these industries to check whether district location affects also service industries whose quality, according to part of the literature (Dobkins, Citation1996) is a value added of district areas. 11. This picture is confirmed when one adopts the Sole 24 Ore classification; while the regional classification leads to different results showing that district firms have no improved access to bank lending. This result is paralleled by relatively higher returns on investment. 12. Evidence on this point is omitted and available from the authors upon request. 13. For the determination of industry specialization in IDs in the ISTAT definition, see the second section. 14. A district may be considered as relatively specialized in a given industry if the relative share of firms belonging to that industry is above the national average. 15. The variable is the Tagliacarne index of infrastructural endowment, calculated at province level, widely used in empirical research on Italian data. The index is a weighted average of ten different groups of indicators (the number of elementary variables composing each of them is shown in parentheses): road infrastructure (10); railway infrastructure (9); energy and environmental resources (17); telecommunications (11); marine infrastructure (18); airport infrastructure (14); credit and services (16); education (25); culture and recreation (35); and health and assistance (44). Endowment ratios are calculated for each group by taking into account the gap between demand and supply of infrastructure in order to eliminate geographical effects. 16. Separate industry regressions show that the district/export nexus is positive and significant in all of the three industries examined. The industry-specific effect is robust to changes in the definition of IDs in the textile industry (less in the other two industries). These findings are omitted for reasons of space and available from the authors upon request. 17. Consider also that, in the present case, endogeneity cannot be ruled out just by introducing lagged regressors. This is because it is possible, in principle, that time-invariant firm characteristics affect both localization in IDs and performance creating the 'illusion' of a 'district effect' on performance. The experiment may rule out at least one hypothesis of endogeneity: that which assumes that firm-invariant characteristics affect both localization in an LLS or in a (non-specialized district) and firm performance. 18. Conditions under which an LLS becomes a district are described in the second section. 19. The findings on this point will be compared with a more straightforward test of industry-specific externalities by using in the third section the specialization index of Glaeser et al. Citation(1992). 20. According to Paniccia Citation(1998), by using the LLS as an observation unit, one is close to what Becattini Citation(1990) describes as a fundamental trait of the ID, i.e. 'the merging of people and economy in a self contained area'. To take into account Tattara's (2002) and Iuzzolino's (2004) criticism on the definition of the LLSs, a further robustness check in which all the indicators which follow in the paper are calculated on municipality and not on an LLS basis has been performed with the results substantially unchanged, which are available from the authors upon request. 21. The global specialization (GS) index is calculated as: where L ij is employment of the ith sector in the jth LLS. Alternatively, the same index is calculated as: where PU is productive unit. 22. The index is calculated as: where where N ij is employment of the ith sector in the jth LLS. The alternative in which employment is replaced by the number of local productive units is also provided. 23. Inter-firm externalities may arise for different reasons and might be fostered by the decision to cooperate for the provision of local public goods (i.e. export services). Another channel for the relevance of inter-firm externalities may be determined by the fact that trust works as a substitute for legal enforcement of contracts in business controversies given the well-known relative weakness of the Italian judicial system in this field. 24. The measure is: where X kj is the relevant performance measure (export or value added per worker according to the relevant equation) in all industries, with the exception of the one to which the specific productive unit under observation is affiliated. 25. The problem here is that it is not possible with the present data to know whether the division of labour classically occurs in the geographically delimited ID area or involves delocalization in other areas or countries. Results omitted for reasons of space are available from the authors upon request. A robustness check on findings from Tables 7–9, in which the control variable of the overall competitiveness of the labour system (Svil-sll) is replaced by the number of bank branches per inhabitants at provincial level, confirms all the results on the impact of agglomeration measures. Results omitted for reasons of space are available from the authors upon request. 26. Results omitted for reasons of space are available from the authors upon request. 27. A robustness check on findings from Tables 7–9, in which the control variable of the overall competitiveness of the labour system (Svil-sll) is replaced by the number of bank branches per inhabitants at provincial level, confirms all the results on the impact of agglomeration measures. 28. An example of them are small and medium-sized subcontractors in LLSs specialized in agricultural machinery (Suzzara and Guastalla) (Paniccia, Citation2002).
Год издания: 2007
Авторы: Leonardo Becchetti, Andrea Panizza, Filippo Oropallo
Издательство: Routledge
Источник: Regional Studies
Ключевые слова: Italy: Economic History and Contemporary Issues, Regional Economics and Spatial Analysis, Global trade and economics
Открытый доступ: closed
Том: 41
Выпуск: 5
Страницы: 601–621