Collective efficiency, policy inducement and social embeddedness: Drivers for the development of industrial districts

Abstract

Where is the future of traditional industrial districts in global markets where competition is fiercer every day? This paper presents the case of the furniture district of Forlí, Italy, as a means to explain the development process, the constraints and the growth prospects that involve this industrial district and, perhaps, a wider variety of districts and SME-based clusters. We hypothesise that development is more likely to be generated when three main drivers, taken from the main bodies of literature on districts and clusters, are taken together: ‘collective efficiency’, ‘policy inducement’ and ‘social embeddedness’. The case study of Forlí helps to identify the trajectory of one among many Italian industrial districts and its solutions to deal with the new competition. Yet, our approach highlights some of the main difficulties that this district is facing nowadays and the related challenges for future development. The general lesson derived from this analysis is that traditional ways of regarding cluster development on the basis of collective efficiency need to be supplemented with an adequate weighing of the social embeddedness driver, as well as of the national and local policy environment. This approach delivers strategic analytical tools to interpret the reality of districts and to target effective development actions.

Keywords:

• industrial districts
• furniture
• collective efficiency
• policy inducement
• social embeddedness

Acknowledgements

I would like to thank entrepreneurs and officers – in particular Remo Ruffilli – associated to the National Confederation of Artisans (CNA) in Forlí, who helped us to obtain relevant primary information about their competitive development processes. I would also like to thank colleagues that helped me refine this research; in particular Peter Knorringa, Institute of Social Studies, The Hague; Roger Sugden and Lisa De Propris at the University of Birmingham; Carlo Pietrobelli, University of Rome 3; Sandrine Labory, University of Ferrara who gave me useful suggestions on earlier versions of this paper, as well as the former editor of this journal, Bengt Johannisson, and two anonymous referees that contributed both broad and insightful critiques and comments on this paper and that helped me convert it to the current development stage. The usual disclaimers apply.

Notes

Notes

1. We adopt a different conceptualisation of ‘district’ and ‘clusters’. ‘Cluster’ is taken as a general concept: a large number of firms specialised in the same sector and agglomerated in close geographical proximity, such as a municipality or a set of very close municipalities (Schmitz 1995). ‘Clusters’ may be more or less successful forms of agglomerations of firms that may include also large companies that lead subcontracting chains (the case of hub-and-spoke clusters, the satellite clusters and the state-anchored clusters; for a revision see Markusen 1996); in other cases the division and specialisation of labour is very low and public support minimal (the case of ‘survival clusters’ in developing countries, see Altenburg and Meyer-Stamer 1999). ‘Industrial districts’ are a specific kind of cluster; they are successful clusters based on small and medium enterprises only, which do not only benefit from intensive division and specialisation of labour but, also, from significant support from public institutions. In addition, industrial districts benefit also from a very supportive social environment where trust and social cohesion favour cooperative initiatives that strengthen collective scale and scope economies (Becattini 1990). The theoretical and policy objectives of this paper, however, address issues that are relevant to IDs as well as to those clusters that are composed only by small and medium firms and that represent a very specific geographical locality. In this paper we will call them, ‘SME-based clusters’.

2. Clustering is interpreted as the process of firms agglomeration. It is a concept that we use as synonymous for the coming together of a large number of firms in both IDs and clusters.

3. Over time the literature identified a number of cluster typologies, which include ‘marshallian industrial districts’ (Marshall 1918; Brusco 1982; Becattini 1990), ‘hub-and-spoke’ clusters, ‘satellite clusters’ and ‘state-anchored clusters’ (Markusen 1996), ‘artisanal clusters’ (Brusco 1990; Parrilli 2004) and ‘survival clusters’ (Altenburg and Meyer-Stamer 1999; Knorringa 2002; Parrilli 2007).

4. These clusters are led by one or a few large firms that determine the growth prospects of the locality and of local firms; similar cases are represented by ‘state-anchored clusters’, where the large company is a national institution/company, and ‘satellite clusters’ where the large firm is situated outside the local cluster (Markusen 1996).

5. In particular, this stream led to the formation of public programmes and projects targeting the benchmarking of best practices and the transfer of new technologies across SMEs, localities and countries (such as, the new EU programme ‘INTER-REG’).

6. Working on a range of experiences, some theorists followed a similar path of analysis and come to emphasise the importance of stages in the context of cluster development (Knorringa 2002; Guerrieri and Pietrobelli 2004). Among these the ‘cluster's life-cycle’ theory (Swann, Prevezer, and Stout 1998) defines the development of clusters in terms of birth, growth, maturity, and decline. We would rather propose an open-ended approach that recognises the wide margins for the continuous regeneration of the ID/cluster's development process in any moment of its life.

7. The interesting analysis by Lorenzen (1998b, 160) of industrial districts in Denmark focuses on their relevant social strengths. He emphasises both ‘trust’ and ‘entrepreneurship’ since: ‘Norms of entrepreneurship and honesty are the basis for conventions of business practice. Conventions of the prestige in entrepreneurship and in helping new entrepreneurs ensure the continuous process of starting up new firms in spite of the periodically low survival rate of old ones’. Simultaneously he also stresses that ‘conventions of cooperation without opportunism help create social trust’.

8. Other aspects can also be taken into account, such as geographic localisation (Losch 1940), size of urbanisation (Krugman 1998), historic specificities (Putnam 1993), and so forth. All of these factors matter and are likely to explain part of the development of local production systems. Yet, these seem to be rather specific and, in a quantitative type of analysis, would be more likely to be considered dummy variables. These explain some processes, but not in all cases.

9. Analyses based on other furniture districts in Italy show similar tendencies to relocation abroad and downsizing of operations (Bambi 1998; Lojacono and Lorenzen 1998).

10. This percentage refers to the number of current firms from which at least one worker left in order to set up his/her own firm.

11. It is low if compared with the 93% that is shown above with respect to the spin off trend in the 1970s and 1980s.

12. This kind of analysis and indicators were previously realised by Maskell in his study of Danish furniture industries (Maskell 1998, 43). He and Macaulay (1963, in Maskell 1998) also emphasise the importance of ‘not going through formal contract and other formal legal framework for a dynamic interaction among local actors’.

13. In the literature on industrial districts, with special reference to historic roots and social embeddedness, there are theorists who describe the process of local homogenisation as a century-long process (Putnam 1993; Platteau 1994; on similar prospects in Danish districts see Lorenzen 1998b).