Embedding environmental innovation in local production systems: SME strategies, networking and industrial relations: evidence on innovation drivers in industrial districts

Abstract

Technological innovation is a key factor for achieving better environmental performances. Its role is even more relevant in local productions system, where innovation density, knowledge spillovers and externalities are concentrated in a circumscribed territory. The paper exploits new data for a sample of manufacturing firms in Northern Italy. New evidence is provided by testing a set of hypotheses, concerning primarily the role of environmental‐devoted R&D, networking activities, quality/nature of industrial relations. The role played by environmental policy pressure, structural firm features and past firm performances is also investigated to account for more exogenous forces. We show that structural characteristics of the firm appear to matter less than R&D, induced policy costs and innovative‐oriented industrial relations. Environmental auditing schemes also show some relevant correlation to innovation adoptions. R&D efforts appear to be associated to networking activities, which substitute for size‐related economies of scale. Overall, endogenous factors driven by firm strategy or local idiosyncratic features matter more than exogenous and structural firm factors.

Keywords:

• environmental innovations
• environmental R&D
• manufacturing sector
• local production system
• environmental policy
• SME
• industrial relations
• networking
• industrial districts

JEL classifications:

• C21
• L60
• O13
• O30
• Q20
• Q58

Acknowledgement

We thank the anonymous referees and the editor for their comments, which substantially improved our work. The usual disclaimer applies. The work derives from the research activities of the Department of Economics Institutions and Territory, University of Ferrara and of CERIS CNR Milan. The research directions, focused on environmental innovation, firm strategy, complementarity and networking in local production systems, are one of the core elements of the 2005–2007 research programme on ‘Innovative dynamics in the knowledge‐based economy’ funded by the university budget for research projects of local interests (ex 60%), and PRIN‐MIUR 2003–2004 on ‘Structural dynamics: Firms, organizations, institutions’. We are grateful to the CGIL trade union of Reggio Emilia for support regarding data collection and provision of balance sheet accounts. Data derive from two surveys funded within the project Infrastructures, competitiveness and levels of government: Knowledge and development of the new economy (PRIN‐MIUR research project 2001–2002 co‐ordinated by Prof. Paolo Pini) and by CERIS CNR of Milan.

Notes

1. Aggeri (1999) calls those informal agreements innovation‐oriented voluntary agreements, where pollution is diffuse, uncertainty is high and innovation becomes the central feature.

2. Intended as a form of production organisation of small and medium enterprise where the territory plays the role of infrastructure for economic, institutional and cognitive integration (Corò and Micelli 2007, 2). On the district model, peculiar to Northern Italy but also present in other realities, we refer to the seminal contribution by Brusco (1982), Brusco et al. (1996).

3. Some other works dealing with impact of regulation on environmental performances of firms, and the effects of environmental innovation on firm performances, including employment, are: Pfeiffer and Rennings (1999); Konar and Cohen (2001); Rennings, Ziegler and Zwick (2001); Rennings and Zwick (2001); Rennings et al. (2006); Frondel et al. (2004); Rennings and Ziegler (2004); Ziegler and Rennings (2004); Doonan, Lanoie and Laplante (2005); Frondel, Horbach and Rennings (2005); Cole, Elliott and Shimamoto (2006); Cainelli et al. (2007b); Ziegler, Schroder and Rennings (2007).

4. See Getzner and Ritt (2006) for a study which exploits trade unions’ support for studying qualitative effects of eco‐innovation on employment quality.

5. The taxonomy of environmental realms is largely consistent with recent OECD studies (Darnall, Jolley and Ytterhus 2005), and with seminal conceptual papers (Kemp 2000; Rennings 2000).

6. See European comparative data in the study by de Vries and Withagen (2005) on patented innovation and environmental policy.

7. We specify in brackets acronyms used when presenting regression results in Section 4.

8. Frondel et al. (2008) provide some evidence on the effect of unions as a pressure group, finding ambiguous evidence.

9. Labour flexibility (generally conceived) indicators do not seem to be related to environmental strategies (results not reported). The effects on firm performances of new innovations, including environmental innovations, are actually scope for further research.

10. Secondly, having elicited whether emission and waste policies are being imposed on firms (policy stringency proxy), and for how many years firms they have been subject to policies, we may analyse the eventual impact of non‐monetary policy‐related indicators. We here use dummy variables for policy presence for emissions and waste (POL‐EM, POL‐WA) and the number of years since the policy was introduced (POL‐YRS), to test an eventual lagged/dynamic response of firms to environmental regulations.

11. Hansen et al. (2002) present an analysis of case studies regarding environmental innovations in small and medium sized enterprises, for five European countries. The study reveals a great variety in factors driving the process: character of environmental innovation; regulatory setting; firm strategic orientation; network relations; and sectoral influence. Innovative capability emerges as the result of the interplay between different driving forces.

12. Probit models are used when facing dummy dependent innovation variables.

13. Specified either in binary forms (specific innovation realms) or as a synthetic index. Acronyms for the various dependent variables are: INNO‐EM (adoption of process/product environmental innovation related to emissions); INNO‐WA (adoption of process/product environmental innovation related to waste); and INNO‐EN (adoption of process/product environmental innovation related to energy inputs). Then, we have INNO‐TOT (synthetic index of the adoption of the four environmental innovations).

14. Predicted values of costs are included, but are not significant.

15. Regressions concerning material innovations are not shown since they show a poor fit.

16. The correlation between such drivers is around 0.35. This highlights they are somewhat different factors, though logically correlated.

17. Linking the reasoning to the methodological debate on patent as innovation proxies in ID, we note that count data models do not here provide strikingly different evidence. We believe, in any case, that our additive index of four innovations is not a count variable in the usual sense we find in the literature: the number of categories is in fact bounded and chosen by the researcher.

18. The only somewhat counter intuitive results we find.

19. This is consistent: recent firm performances arise relevant in spurring expenditures a proxy of higher ability to pay.