The Role of Territorial Capital in Local Economic Growth: Evidence from Italy

Abstract

Territorial capital is defined as the system of territorial assets of economic, cultural, social and environmental nature that ensures the development potential of places. The potential of this concept resides in the recognition of possible interactions between factors of different nature. So far, however, very few studies have focused on the empirical verification of the links between territorial capital and economic growth. This work is devoted to the analysis of the relationship between economic growth and territorial capital in Italian NUTS3 regions between 1999 and 2008. The distribution of territorial assets across regions points out the huge gap between Italian macro-areas. These divergences do not clearly reproduce the differentials in GDP growth. The second part of the analysis is focused on the joint effect of the territorial capital components on the regional economic performance. Our findings emphasize the role of some endogenous factors in explaining the differentiation of regional growth patterns. Moreover, they point out the importance of soft assets in correspondence of an external shock, as the one represented by the recent financial crisis.

Acknowledgements

I greatly benefited from the guidance and encouragement of Roberta Capello. I also wish to thank Roberto Camagni and two anonymous referees for their valuable comments. This study was funded by the Italian Ministry of Education, Universities and Research (PRIN project 2008PP5E98 – Exploiting territorial capital – Quali-quantitative scenarios for Italian provinces to overcome the economic and financial crisis) and by the Politecnico di Milano (FARB funded project – Capitale territoriale e politiche di sviluppo locale).

Notes

1. When 1999 data are not available we employed the indicator for the nearest year (see Appendix 1).

2. Cultural heritage is approximated as the number of monuments ranked two stars in the Touring Club Green Guide Series. District economies are represented by the share of workers employed in an industrial district as defined by ISTAT based on the literature on this topic (Becattini, 1987). Agglomeration economies are measured through population density. Relational capital usually refers to the relationships and cooperation agreements between firms (Capello & Faggian, 2005). Due to the lack of data at the NUTS3 level, our proxy for relational capital pertains to the linkages between local actors outside the productive dimension of the society and it is estimated as the number of volunteers employed in social cooperatives and non-profit organizations, weighted for the resident population.

3. The three main families of splitting rules are CART (Classification And Regression Tree; Breiman et al., 1984), CHAID (CHi-squared Automatic Interaction Detector Kass, 1980) and PARTY (recursive binary partitioning; Hothorn et al., 2006).

4. Actually, the coefficients of private capital and relational services do not have the expected sign, we will come back to this point in the next lines.

5. Variable creativity is recoded as a dummy variable equal to 1 if it takes a value larger than −0.85 and equal to 0 otherwise. Agglomeration economies is transformed into a dummy equal to 1 for values above 2.3 and equal to 0 otherwise. Finally, variable education becomes a dummy equal to 1 for the values above 0.38 and equal to 0 otherwise.

6. As a further check we included as a control (not reported in Table 3) the employment growth rate in the same period. The inclusion of this regressor does not affect the significance of the other variables.

7. In all cases, we tested for spatial autocorrelation of the residuals using Moran's I test. Results fail to reject the null hypothesis that there is zero spatial autocorrelation in the residuals of the estimated models.

8. The availability of statistical sources on private capital is extremely limited. We started from some data (Appendix 1) on yearly private investment flows defined at the NUTS2 level  and classified by economic sector. As a first step, we rearranged the flows at the NUTS3 level by splitting the regional values across provinces based on the share of workers employed in each sector. Then, we computed the stock of capital by applying the perpetual inventory method to the provincial investment flows (Bonaglia & Picci, 2000).

9. Proximity is defined in terms of travel time, as the one employed for the indicators of port and airport infrastructures (see Appendix 1 for more details). Moreover, since the variables Airports and Harbours already represent a measure of accessibility, we excluded them from the set of hard assets.

10. This is due to a lack of data at the NUTS 3 level. At present, the latest and most up to date statistics refer to 2008.