A longitudinal analysis of local innovation in Italy: How do proximity measures matter?

ABSTRACT

This paper investigates the factors influencing local innovation from a longitudinal perspective while assessing geographical, economic and technological proximity. The research hypotheses concern spatial interactions, spillover effects and proximity measures that best fit innovation patterns and territorial interactions in Italy. The estimation strategy is the spatial Durbin panel model. The optimal specification to handle cross-sectional dependence in the data was derived from statistical tests evaluating (i) individual-specific effects, (ii) time-specific effects and (iii) both individual and time effects. The model was estimated using data from 107 Italian provinces over 2010–2019. The results show that both endogenous and exogenous interaction effects drive innovation processes and the underlying spillovers are global. Economic proximity explains local innovation patterns more effectively than geographical contiguity and technological proximity.

KEYWORDS:

• Local innovation
• knowledge production
• spatial panel model
• spatial spillover

JEL:

• C23
• O30
• R58

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the author(s).

Notes

1 The model was estimated using the maximum likelihood estimator via the R spml routine (Millo et al., 2012). Additional insights into its shortcomings and underlying assumptions can be found in Elhorst (2003).

2 ATECO is the Italian version of the European nomenclature (NACE), which classifies economic activities hierarchically from general to particular by means of alphanumeric codes: $\mathrm{sections}$ (capital letters), $\mathrm{divisions}$ (two numerical digits), $\mathrm{groups}$ (three digits), $\mathrm{classes}$ (four digits), $\mathrm{categories}$ (five digits) and $\mathrm{subcategories}$ (six digits).

3 As pointed out in Gupta (2019), since the elements of $(W_2)$ and $(W_3)$ were determined by economic indicators, it is reasonable to consider both matrices as stochastic.

4 This may arise from the possibility that time fixed effects may partly cover common factors (Halleck Vega & Elhorst, 2016).

5 Alternative specifications accounting for individual effects or both time and individual effects are reported in Table A2 and Table A3 in the Appendix in the online supplemental data, respectively.