Navigation – Plan du site

AccueilNuméros129-130Modes de coordination, réseaux et...Towards a framework for the evalu...

Modes de coordination, réseaux et politiques industrielles

Towards a framework for the evaluation of policies of cluster upgrading and innovation

Marco Bellandi et Annalisa Caloffi
p. 259-276


Aujourd’hui, un nombre important et croissant de politiques de développement local et de modernisation incorporent directement l’idée d’innovation comme un processus systémique, encastré dans des contextes socioculturels et institutionnels spécifiques, et entremêlé à des challenges, opportunités et stratégies internationales. Ces politiques conduisent à de nouveaux challenges pour les activités d’analyse et d’évaluation : en dépit de la diffusion d’une approche systémique à la fois pour la théorie et les politiques d’innovation, un cadre d’analyse adéquat pour l’analyse et l’évaluation de ces politiques est loin d’être réalisé (Bellandi and Caloffi, 2010). Pour avancer dans cette perspective, on propose différents exemples dans le cadre de contextes précis, ceux des districts industriels (italiens en particulier), caractérisés par des clusters composés de PMEs confrontés à la globalisation contemporaine. Centré sur les politiques d’innovation visant à promouvoir la modernisation de districts et clusters confrontés à des filières et chaînes de valeurs internationales en mutation, cet article discute la signification des politiques industrielles quand une perspective systémique est considérée. Sur ces bases, des exemples aident à comprendre les caractéristiques des méthodes d’évaluation adéquates. Les aspects méthodologiques relatifs aux évaluations sont finalement discutés.

Haut de page

Notes de la rédaction

A first version of this paper was prepared as base for the presentation by Marco Bellandi both at the TCI 2008, 11th Annual Global Conference of the Competitiveness Institute, 29-31 October, Southern Sun Cape Sun, Cape Town, South Africa, and at the REI Colloque, 13-14 November 2008, Juan-les-Pins (Nice) France.

Texte intégral

I. — Introduction

1Nowadays, in industrialized and industrially emerging countries and regions, policies and strategies for local development and cluster upgrading explicitly incorporate the idea of innovation as a systemic process, embedded in specific socio-cultural and institutional contexts and intermingled with international challenges, opportunities, and strategies. The analyses of the systemic contexts (places, times, and rules of interaction) in which the innovation processes take place have fed a large and growing body of literature. This is exemplified by the widespread reference to systemic units of combined public actions and private strategies, such as the milieux innovateurs, the local innovative networks, the dynamic industrial districts, the innovative clusters or local innovative systems, the regional innovation systems, the clusters and the governance of international value chains.

2As the innovative process takes place in a systemic context, the risk of inconsistencies increases if policies are not based on an appropriate systemic unit of analysis and intervention. However such a frame is not easy to get, for at least two reasons. The first is general and comes directly from the evolutionary systemic nature of the phenomena to which policies are targeted. Unexpected novelties or ruptures may lie under a thick web of feed-backs and business and social relations. The second is related more specifically to phases of quick change and transition in the systems themselves. This complexity explains why the actors playing strategic roles within the systems need to be embedded in fabric of relations of the system itself, in order to understand and to be understood and followed by the local producers; on the other hand, too much closeness generates collusion, inertia and myopia. As a consequence, the assessment and evaluation of these policies is as important as puzzling. It is important because it not only adds to the accountability of policy-makers against collusion and inertia, but also combines with context sensitive step-wise policies. It is puzzling because it is very difficult to figure out what could be the state of the system with different sets of policies undertaken by the same policy-makers or by different ones.

3We propose here some reflections and exemplifications on a quite delimited set of conditions and contexts, i.e. those of industrial districts (Italian, in particular), characterized by SMEs clusters facing contemporary globalization challenges, as defined in section 2. Sections 3 to 5 introduce other conceptual bases, with particular regard, first, to innovation policies aimed at supporting functional upgrading of districts and clusters soaked in changing international filières and value chains; and, second, to the meaning of evaluation of industrial policies when a systemic perspective is considered. On such premises sections 6 and 7 illustrate a couple of exemplifications on the features of appropriate evaluation methods. Section 8 concludes with notes on a systemic approach in the evaluation of policies for local development and cluster upgrading.

II. — Globalization and SMEs clusters: A view based on Italian districts experiences

4The last phase of globalization has posed big challenges to industrial districts in older industrialized countries, whilst offering opportunities for growing new industries in old and new localities of industry, some of them related to emerging economies. The current burst of a planetary crisis of finance, industry and trade will probably modify the scenario, but it is too early as we write (March 2009) to understand how much and in what directions. Here we rather keep on reflecting on some important factors emerged in the phase ante crisis, and maintain that they will still play a role when international relations will stabilize again.

5A number of studies have proven the peculiar capacity of some traditional industrial districts, for example in Italy, to internationalize through various individual business actions and relations. As these studies have shown: i) international subcontracting (sometimes started at first as temporary exports) is not difficult to set up for district entrepreneurs who are used to manage links external to the firm; ii) relations with foreign suppliers are favored by a variety of contractual and personal relationships, as for instance when district entrepreneurs that choose to live abroad intervene directly or indirectly in the activities of their partners; iii) active foreign direct investments (FDI) controlled by larger district firms can take a district form when for instance FDI are concentrated in a particular territory and the leading district firm induces other district firms to invest in the same territory; and iv) some district firms may also take part in the governance of international value chains as when they produce locally an intermediate product that is sent in a foreign country to be transformed and sold by a foreign partner into the final market.

  • 1 See for example Abernathy et al. (2006), Belussi and Sammara (2009), Gereffi et al. (2005); Humphre (...)

6The aggregate result of such individual actions is difficult to grasp1. We can see it as the combination of three components related to the territorial configuration of a particular activity (or set of activities). The first component is quantitative: the size of the economic activity may increase or decrease in a place where it was originally localized without any change elsewhere; or it may increase or decrease in the original place with a related decrease or increase in another place. The second component concerns local value chain relations: in the place where the activity decreases it may be replaced by other types of activities more or less related to the first one; or where the activity increases, other types of activities more or less related may close down or start up. The third component concerns cluster and district relations: the moved activity may have been embedded in a network of industrial and social relations promoting local productivity, learning and creativity in the original place, or may have kept instead a footloose relation with it; and the moved (or the replacing) activity may embed in such a type of network in the new place, or it may be localized there just for generic spatial advantages (transport costs, cheap labor and land, tax rebates, polluting rights, etc.). Of course many intermediate cases could be contemplated.

7Coming back to aggregate results, we would say that:

8a) an embedded activity which moves and becomes foot-loose is « de-localized »; while an embedded or foot-loose activity, which moves and is embedded in the new place, is « re-localized »;

9b) a cluster where an embedded activity is replaced by a foot-loose activity is functionally down-graded in its systemic content;

10c) a cluster where a foot-loose activity is replaced by an embedded activity is functionally up-graded in its systemic content;

11d) local decline is characterized by quantitative shrinking and by functional down-grading in one (or more) not compensated by up-grading in other clusters in the same locality;

12e) local development is characterized by quantitative expansion and by functional up-grading in local clusters (and marginal down-grading processes if any).

13Positive aggregate results, like re-localization, functional up-grading, and local development need the insertion of individual business international actions in a consistent framework of private and public strategies and investments, supported by and feeding-back positively into the inner local structure of social relations. This framework presents (at least) two sides. On one part, without revamped sources of inner local productivity and creativity, international investments and relations do not lend positive solutions to clusters and localities hit by globalization challenges, more probably making the problems worse. On the other part, without appropriate international strategies, the value produced by inner productivity may be predated by multinational agents, or spill over too easily to global markets. Both local and international strategies should lie on peculiar drivers of local competitive advantage on international markets and, of course, they change under the influence of many conditions and tendencies.

14Considering for example the case of those many Italian industrial districts characterized by SMEs clusters producing highly differentiated goods for the person and for the house, revamped sources of inner local productivity and creativity resides nowadays in upgraded capacities of teams of district firms to produce and sell products of medium-to-high quality, with highly personalized and craft content, increasing the investments not only in craft, industrial and R&D capacities, but also in quality, environmental and ethic certificates, and generally in all the instruments promoting association with taste for beauty, good quality of life, love for well done craft jobs, creativity, etc. (Becattini 2004; Cooke and Lazzeretti, 2007). On the other part, international strategies appropriate to local up-grading reside in enhanced capacities of teams of district firms to join: a) international production fragmentation with the preservation of both local high value added manufacturing capacity, and control of more standardized or less specific complementary products and operations realized in foreign localities and clusters; and b) international trade and distribution channels with common platforms characterized by rich catalogues of high quality complementary or diversified products.

15The internationalization capacities should combine, through international management skills, the peculiar factors of inner productivity, creativity and innovative potential with foreign resources and opportunities. Recent contributions and cases suggest that an effective combination needs appropriate forms of international relations, for example specific to the type of entrepreneurial and craft skills and attitudes characterizing successful SMEs clusters and districts (Bellandi and Caloffi 2008a). These latter are characterized essentially by individual energy, creativity and simple internal managerial structures, interacting in teams of independent specialized firms, with the help of shared trust bases, of cognitive proximity, and of various types of specific public goods providing indivisible assets and rules. Going abroad with a networked structure is consistent with the use of such peculiar features together with the deployment of rich pools of resources; keeping contact with home resources through networked relations reduces the temptation of lonely or predatory strategies, whereby for example single entrepreneurs buy cheap and undersell home competitors also by means of phony brands and misleading advertising; getting in touch with foreign partners who have or may learn to have similar networked structures and skills helps the building of collaboration and joint infrastructural, production, and trade projects on a basis of organizational and institutional similarity.

16All this has suggested, also beyond the cases of Italian industrial districts, that appropriate international strategies, helping the up-grading and development of SMEs clusters and their home localities, should take the form of re-localization management supported by international cross-cluster and trans-local collaborative relations (Bellandi and Caloffi, 2008a; De Propris, 2008). On the other hand, coming back to the revamping of inner bases of local productivity and creativity, it may be recalled that the constitution and maintenance of some among such bases demand high levels of urban infrastructure (Trullen and Boix, 2001). The local scale and urban quality of an industrial district are rarely adequate to this purpose. A regional (sometimes a national) milieu hosting dynamic regional cities and districts can offer a solution, if the collaboration among networks of firms, research centers, knowledge services, high level trade facilities, and pro-active local and regional policy-makers find the way to develop and interact (Cooke, 2005). This means that networked projects, cross-cluster and trans-local relations at a regional level are also an important part of the recipe for inner levers and drivers.

17The vision of local and cluster upgrading through regional and international cross-cluster and trans-local relations and structures, in face of globalization challenges and opportunities, sets the stage for the following reflections on some aspects of policies for local development and cluster upgrading, and their evaluation.

III. — Innovation policies as support to local upgrading

18Public policies and governance processes supporting the revamping of local productivity and creativity extend necessarily to various aspects. One of this is the local capacity to incorporate and promote innovation processes. We take again, as a field for exemplifying our reflections, the case of SMEs clusters characterizing various types of localities of industry, where competitive advantage is driven by networked and clustered business and social resources and activities. This is potentially conducive to fruitful intersections with the development and working of innovative systems.

19An innovative system, as observed by Lane and Maxfield (1996), is the result of social processes of interaction repeated over time, through which specific agents, leveraging their knowledge and competencies, launch and implement streams of relations with other agents in order to compare, manipulate, combine and re-combine (pieces of) existing knowledge and artifacts, or create new ones, at the same time adapting/modifying their own set of knowledge and competencies.

20Let us consider, for example, the case of a firm that produces precision lasers (low power diode lasers), normally used in the sphere of medicine, for the manufacture of devices used in the treatment of painful muscular symptoms and in tissue stimulation. This firm is part of a system within which a series of relations exists between the developers of the laser source (research centers, universities and firms), the inventors and developers of the instruments used to channel the source into specific devices, and of the devices themselves (research centers, universities and firms), the developers and manufacturers of the specific components of these devices (other firms), the experimenters of the instrument (medical research institutes, hospitals, clinics) and those performing its validation (hospitals, and medical clinics). Some of these relations demand a high degree of territorial and cognitive proximity between the agents, evolving within specific clusters, more or less rooted in specific territories (Hendry et al., 2000). Let us assume now that in this optoelectronics centered net a significant part of the relations (as well as a significant proportion of the agents) be rooted in a local territory, which is also a repository of manufacturing traditions, knowledge and specific competencies. The localized part of the net may be seen as an optoelectronics cluster with local systemic features, a local production system.

  • 2 Here « region » means both a delimited regional milieu hosting a set of localities and clusters sho (...)

21Policies spanning a region2 can support both the development of such a cluster, and together the emergence of cross-fertilization patterns of growth with other clusters of the region. Two main levels of strategy may be identified. The first one is the creation of (or contribution to) public goods with features which are specific to the innovative processes and to the reproduction of knowledge and competencies crucial for the cluster and local context: investment in specific (specialized and applied) research centers, funding of existing research projects, bridging organizations between the world of industry and that of research, incubators for new enterprises having related technological/sectoral/territorial targets. The need of public intervention increases with the size of the intervention, the presence of coordination and normative problems, the risks and time scale involved. Public resources add to private funds coming from the local systems within which much of the design process takes place.

  • 3 In the case of the Tuscany Region, in Italy, which is home of an important optoelectronics cluster, (...)

22The second level concerns the processes of cross-fertilization between clusters incorporating potentially complementary features. The public intervention here is targeted to pull the application of certain technologies or organizational solutions, generated or developed within specific local and business contexts, to other technological/sectoral contexts of the region. And it intersects more or less deliberately the revamping of specific sources of inner local productivity and creativity. For example, in the case recalled previously, the target could be the application of the optoelectronics technologies to other traditional industries (or contexts of economic activity) rooted in the region which are hit by globalization challenges3.

23This kind of intervention is not simply aimed at the diffusion of specific innovations within the regional borders, but at the embeddedment of the same innovations into local systems and clusters included within the regional milieu but different from those in which the new solutions were developed. Embedding is fostered by appropriate relations amongst key actors of the systems involved: actors having different competencies but sharing a common language, capable of mobilizing networks of relations at the local and regional scale. Presumably not all the relations will lead to success, but the intervention is nevertheless aimed at the creation of channels of interaction previously non-existent, which the various agents may take in and develop or not. Once applied to other contexts, the initial solutions (the new technologies, the organizational solutions…) can give rise to new streams of innovation.

IV. — Policies for active district and cluster internationalization

24Let us come now to active district and cluster internationalization in the form of re-localization management supported by international cross-cluster and trans-local collaborative relations. Various types of specific public support have a direct role here.

  • 4 Promising prospects for twinning are related to general economic and social foundations, in particu (...)

25Firstly, even if the twinning among localities of industry (and relative clusters) often results just from random scouting, trials and errors, public support is crucial nonetheless in funding some early steps of the process, giving institutional protection and representation to sets of smaller firms, and providing coordination in order to concentrate efforts on more promising opportunities of relation4.

26Secondly, local and supra-local policy facilitates the creation of a set of learning structures, business incentives and life facilities for the circulation, among the localities, of « Argonauts » who feel at home in different localities, and know how to communicate with those who, on the different sides, have just one homeland (Saxenian and Sabel 2009). They favor the translation of specific pools of knowledge and their incorporation in innovation capabilities.

27Thirdly, and related to the previous point, especially in developing areas, district-like relations are sometimes potentially feasible, but not yet fully emerged. If one of those areas is targeted by a developed cluster – say from an Italian industrial district – for investments in complementary activities, then public support helps district business agents to implant in the targeted area the knowledge of technical and trade standards necessary for specialization among local firms, and of rules for improving job conditions and the participation of workers to strategies of quality and innovation (Mehrotra and Biggeri, 2007).

  • 5 It is still possible for strong districts and clusters – say Italian ones with respect to clusters (...)

28Fourthly, trans-local collaboration for investments and re-location needs not only the presence of appropriate local forces in both parts, but also the definition of relations and agreements that have a prospect of equity and reciprocal empowerment. If the partners are not able to reach an understanding on such bases then the re-location fails as such (Sugden et al., 2005)5.

29Fifthly, only few larger district firms may have the internal capacity to learn, adopt and monitor systems for managing relations at a trans-local scale. They can play as bridging business actors (De Propris, 2008). However, in order to limit the growth of monopolistic positions within the clusters involved, and to help a wider direct engagement of smaller but dynamic firms, public support has to be directed to collective platforms (communication facilities, technical and trade standards, and complementary skills) complementary to the limited international management resources of the smaller firms (Sabel, 2004).

30Finally, and again related to the previous point, as always when adaptation in systemic conditions and high uncertainty is involved, the knowledge and the resources to bet on the development of the process, if any, tend to be concentrated in a small set of entrepreneurs and innovators, from either private or public ranks. But if the consensus of the larger local communities, and in particular if complementary investments of other local business agents are needed, then networks of business and social relations between the internationalization actors and a larger group of local cluster and district agents have to be preserved. Public agencies should provide participatory methods, public seed funds, and third party assurance, which make easier this interaction (Meyer-Stamer, 2004; Enright and Ffowcs-Williams 2001).

V. — Perspectives on methods of evaluation of industrial policies

31We come now and in what follows to some reflections and exemplifications on the evaluation of policies which target, more or less explicitly, objectives like those recalled in the previous two sections. Evaluation has been seen originally as an ex post component which gives some measure of the results achieved by policies. In fact it is much more, and within a systemic approach it becomes a necessary though complex component.

32Let us take innovation policy as an example. In traditional models, innovation is something happening within the boundaries of the firm, where the results of basic research activities are applied and transformed, by means of relatively unexplored mechanisms, into new products or services. Here public funds are supposed to be granted to the enterprises (in the form of tax allowances, capital accounts, etc.) for the development of their internal R&D capacities, the acquisition of basic research results incorporated within patents or other artefacts, the development of services useful for the introduction of the new products on the market. In this context, the evaluation of public policies is focussed on the additionality of the support measure, and in particular on the measurement of the effects achieved by the beneficiaries (be they quantified, for instance, in terms of new products introduced on the market, sales or patent applications) thanks to the existence of the public intervention. A large body of literature has focused on the exploration and the measurement of the additionality effect. Among them, the quasi-experimental approaches try to improve the identification of counterfactual samples of « non-treated » enterprises, and their comparison with the « treated » enterprises benefiting from the public funds.

33The methodologies developed under these approaches are not fully suitable for evaluating policies targeting systemic character as those aiming at promoting local and regional clusters. On one hand, it is quite obvious that the results of a system-based policy such as the promotion of industry-research linkages (e.g. through various kind of innovative networks or community of practices) within a specific locality cannot be captured by the simple observation of a possible increase in sales or patent applications made by the single enterprises. On the other hand, the identification of the counterfactual can be particularly challenging, the additionality of the support measure being difficult to evaluate through the comparison with a theoretical regional system showing the same static and dynamic properties as the « real » one, but for the public intervention under observation.

34The most innovative experiences in policy evaluation undertaken by some European countries such as Sweden (Klofsten et al., 1999), Finland (Rajahonka e Valtakari, 2005), Austria (Priedl et al., 2008), and Germany (Eickelpasch e Fritsch, 2005), emphasize the growing necessity of integrating the evaluation of the impact of public intervention on the different components of the system (single enterprises, universities and research centres, etc.), with the analysis of the innovative system as a whole. This requires a careful combination of languages and tools of different study’s fields: the evaluation of public policies, the analysis of innovation and development processes, and the analysis of the international fragmentation of production chains. The evaluation is made more complex by the necessity of considering the specificities of the different policy tools, as the promotion of joint university-industry research projects, networks of innovators, venture capitalists, start-up and spin-offs of innovative enterprises.

35The following directions of methodological development may be considered:

36a) the identification and testing of methods and tools for the quantitative and qualitative measurement of some systemic outcomes and impacts, such as the networking effects (Russo and Rossi, 2008), the development of sustainable university-industry relations (Barnes et al., 2002; Siegel et al., 2003), and the interactive learning effects (Kuhlmann, 1998);

37b) the contextualization of the results of the policy intervention in the framework of the broader regional innovation system (Lengrand, 2006), with particular regard to: i) the analysis of possible changes in the relations between beneficiaries and non-beneficiaries of the public intervention; ii) the effects in terms of the strengthening/weakening of the crucial relations of the system; iii) the possible impacts in terms of changes in the functions performed by the system actors;

38c) the evaluation of the involvement of the various local and regional stakeholders, as stressed by a set of approaches developed during the last decade: the participatory evaluation (Diez, 2001), fourth-generation evaluation (Guba e Lincoln, 1989), utilisation-focused evaluation (Patton, 1997), learning evaluation (Autio, 1998) or empowerment evaluation (Kuhlmann, 1998).

39Finally, the evaluation needs to evolve from an activity implemented at a particular point in time to a process strongly integrated within the policy cycle. On one side, the results of the evaluation activities provide valuable inputs for the management of a specific intervention (ongoing monitoring and evaluation) or the design of new policies. On the other side, in order to set the stage for the development of sustainable policy making activities, the design of the policy intervention is accompanied by the design of the evaluation activities.

40Next two sections illustrate the systemic challenge to evaluation policies with two types of examples which relate to the fields of conditions and policies defined in the first part of the paper, that is the support to local and cluster upgrading through regional and international cross-cluster and trans-local relations and structures. The first exemplification concerns local and regional innovation; the second concerns international relations.

VI. — Evaluating regional policies for innovation systems: an example from Tuscany

  • 6 What is illustrated here is extracted from Bellandi and Caloffi (2008a), which is the result of DSS (...)
  • 7 The database includes a set of interventions implemented within the DOCUP-SPD 2000-2006 and the RPI (...)
  • 8 The database includes all the projects funded in the period under consideration, i.e. 122 projects (...)

41Policies on networks of innovators have been implemented recently by the regional government of Tuscany (Italy), in the form of a set of programs financed by European funds, and managed by the same Tuscany Region in the period 2000-2006 (Bellandi and Caloffi, 2010)6. They are aimed at supporting innovative projects implemented by networks of heterogeneous economic actors7. The specific programs encouraged networking among actors belonging to the worlds of industry, research and services for the purpose of projecting joint R&D or innovation diffusion actions. More specifically public intervention was aimed on the one hand at supporting the innovative potential of clusters and local systems of the region (introducing technological/sectoral targets consistent with their specializations), and on the other at favoring the emergence of relations among the same systems8. The whole set of projects – considered here as a network of networks – involve 908 agents, classified as follows: i) innovation centers, business development service centers, technology parks and similar infrastructures; ii) departments of universities and units of research centers; iii) chambers of commerce, business associations and other kinds of local/regional association; iv) enterprises; v) other (various public bodies).

  • 9 The enterprises show a degree of local containment of the relations which is above the average (41 (...)

42A fairly significant element is that a high share of the relations that take place within the total network has a local dimension (approximately 37 % of the total relations, within the same « Provincia », i.e. a county). Therefore, the web of relations we are observing has specific territorial roots. The links between the various local systems take place thanks to the activity of applied research centers or universities, innovation centers and trade associations, which entertain relations with a number of agents operating in different technological/sectoral and/or territorial contexts. The generally limited size of the firms involved in these programs does not favor their role as trans-local bridging actors9.

43Beyond the representation of the network as a whole, significantly stable (in the time span under consideration) relations emerge. They involve a sub-set of agents that exchange information, competencies and build strategies along non-episodic timescales. Here we can presume to find the actors that are capable of generating, producing and reproducing rules of interaction, competencies and strategies that are the basis of the system. Hence we consider their sub-network as a projection of a web of relations, supporting integration between innovation’s actors variously embedded in the territories of the regional milieu.

  • 10 As regards the « traditional industries », we refer to the definition adopted in the various fundin (...)

44Within such sub-network of significant relations it is possible to identify some « islands », that are parts of the sub-network which shows a relatively high degree of self-containment. The main islands have sectoral/technological features: optoelectronics, biotechnologies, mechanics-robotics and traditional industries10. The analysis of the whole set of relations developing around the four islands (or sectoral/technological sub-networks), is directed at identifying what type of structure of relations characterizes these islands. More specifically: do they display a structure and an organization of the relations that can be considered useful for the exploration, exploitation and dissemination of innovations within and between the clusters and localities of the region milieu?

45Within the islands the density of local relations is obviously more elevated than in the average of the general network, simply because the islands have specific territorial roots within the industrial clusters of the region. The structure of the relations that characterize the four islands is nevertheless different. Given the large number of nodes and relations involved, a clear visualization of these four technological sub-networks proves difficult; figure 1 above synthesizes the typical architecture of their central relations, i.e. the relations among the most central actors.

46The optoelectronics sub-network, composed of 178 agents localized mainly within the regional capital (Florence), has universities and research centers at its core, and a set of enterprises linked to such core, while other kinds of agents play a less central role. The firms involved (not only SMEs) are highly innovative. Some of them are university spin-offs, others have been founded by former graduates from the Faculty of Engineering of the University of Florence. Most of them have long-term connections both with universities and research Institutes of the CNR (National Research Council) which are located in the Florence area. While the dyadic relationship between research and production seems here quite strong, the « third dimension » represented by innovative services (such as incubators, technology parks or similar infrastructures) which could support innovation diffusion processes at the local and regional level appears to be weak at the moment.

Figure 1 : Architecture of the central relations of the sub-networks considered

Figure 1 : Architecture of the central relations of the sub-networks considered

Key : Black boxes are universities and research centres; Black circles are firms; White boxes are service centres; White circles are business associations; Black triangles are incubators; Black box with a white circles are other public organizations. Black lines are direct relations.

47The biotech sub-network is fragmented in three components, with specific local roots and around different fields of biotech applications (green biotech for the largest component, and white biotech for the other two components). The central architecture of the main component is displayed. At the core is a particular kind of bridging actor such as public/private agencies for the protection of the environment, which are often inserted between the dyadic relation linking research centres and universities. However, the relational architecture seems to be too weak to have a real impact on the activity of the cluster agents: it is participated only by a small amount of biotech agents and it develops outside of the main centre of biotech production. This web should be carefully monitored in order to understand its diffusive potential for enterprises operating within specific niches of the green biotech.

48The relational architecture that most clearly characterizes the fragmented body of the traditional industries’ island is centered on the relation between SMEs, trade associations and innovation centers. The sub-networks are clearly centered on the traditional and well-known industrial districts of the region. Here the service centers have a specific sectoral target and are often the result of a collective action aimed at creating local innovation support infrastructures (Brusco, 1994). The (scarce) relations among the different district sub-networks are managed by local governments, business associations or (generic) service centers. Universities or research centers are almost completely absent from the group of the more central actors.

49The mechanics-robotics sub-networks, rooted within a specific territorial area of the region (the Pisa urban area, with its important universities, and some adjacent localities), reveals a balanced relational architecture, connecting the worlds of industry, of innovation and business services, and of research. Here SMEs (often high-tech SME) play a central role, being well connected to the other types of actors and frequently positioned at important crossroads of relations. Relations involving SMEs extend also at the regional level. The core of the technological competencies (and part of the productive competencies) is centered on an urban area which hosts an important part of a high tech cluster.

50The architectures here illustrated come from the combination of two spaces, one formed by relations among agents autonomous (i.e. not strictly dependent) from the regional programs referred above; the other formed by the projects presented for funding and financed by such programs. Of course a full evaluation of the results of the programs would ask a verification of a positive influence on innovation capabilities and performances embedded in the clusters and localities of the regional milieu and in the relations among them. We will come back to this point in the final section. What we try to assert here is preliminary. That is, the analysis of the architectures shown in the programs gives suggestions on possibilities of policies’ success, insofar as the description of the same architectures is complemented by an understanding both of the original state of the system (which maps directly on the first space), and of the type of innovation focused relations which would be necessary for supporting upgrading and development in the regional territories (which should map on the second space).

  • 11 It is a « Technological district » (Bonaccorsi e Nesci, 2006) as defined by the regional policies.

51For example, referring to what observed before, the form of relations within traditional industries reflects quite directly a state of the system which is characteristic of innovation in many Italian industrial districts: innovation with a low intensity of direct investments in scientific knowledge through R&D. However, in face of current globalization challenges, it can be argued that the traditional bases should be complemented by a significant inflow of scientific knowledge in ways consistent with embeddedment, mobilization and upgrading of the professional capacities and entrepreneurial attitudes of the district producers. A clearly successful regional innovation policy, throughout its funding programs, should have promoted the constitution of stable networks between SMEs, business services, and research units. We have seen instead that the sign of the existence of such networks, among those funded by the regional programs, are quite weak. Symmetrical remarks could be done as regards to the optoelectronic island. While the seemingly satisfactory robustness of networks in the mechanics-robotics island is related also the existence of stronger high tech clusters in the core area11. In this case the programs have mirrored well a good conformation of relations, and the evaluation should move on the verification of what has been done, thanks to public funding and promotion, in terms of innovation (capabilities and performances), which could have not been done without.

VII. — Evaluating policies for cross-clusters and trans-local international relations: an E.U. example

52Local and cluster upgrading through international cross-cluster and trans-local relations and structures, in face of globalization challenges and opportunities, is something which has been conceptualized quite recently. Therefore it is not yet easy to find apposite policies and evaluate them. However there is an old and diffused tradition of public action, provided by various regional, national, and international agencies, to help international trade (in particular exports), cultural and research exchanges connected to business field, international industrial good practices’ exchanges, technological and organizational aid to developing, backward, or transition areas. Within these traditions and experiences, signs of cross-cluster and trans-local actions can be found. We do not go after a systematic account of such signs here. It is presented just an example which helps to illustrate the detection of international cross-cluster and trans-local actions amid traditional international business and cooperation policies.

53The example is taken from an European Union INTERREG 3C project, called DISTRICT (Developing Industrial Strategies Through Innovative Clusters and Technologies). It has funded trans-regional collaboration under E.U. « Knowledge Economy and Technological Innovation » targets, in the period 2005-07, with 4M€ funding. The project has involved 50 public bodies from four European regions: Tuscany (Italy, Lead Partner); West Midlands (UK); Saxony (Germany); Västra Götaland (Sweden). The substantive aim assigned to the DISTRICT project was to help improving « innovation strategies and policies for specific measures created for regions whose economies have traditionally been based on industry », providing « its partners a continuous cooperation framework to exchange information and best practices, and jointly develop innovative methodologies for improved regional strategies, strategic planning and pilot actions » ( Under the DISTRICT frame, 11 sub-projects have been selected and funded along three thematic components: a) connecting clusters and business networks to innovation, within the global context; b) SME’s innovation projects, in collaboration with Universities and Research and Transfer Centers; c) innovation in Financial Engineering, seed venture capital, start-ups and spin-offs.

54The analysis of sub-projects’ reports makes evident that a good part of the activities undertaken falls under the traditional categories of exchange of information and good practices. However some sub-projects or parts of single sub-subprojects go beyond that, suggesting attempts for creating relationship-specific assets, expanding mutual learning, combining complementary capabilities, establishing superior governance structures and lowering transactions costs (District, 2008). For example a sub-project (INTECHTEX), claiming as its objective to develop different strategies for new technologies, products and markets in the sector of technical textile, has involved various types of clusters (in agriculture, building, automotive, clothing) in the common use of some basic technologies. The sub-project has promoted the development of various partnerships around new technical textile processes and products and the constitution of a joint venture between two European firms. More generally the regional public partners (universities, public business centers, local authorities) have contributed, with their business associates, to the constitution of a specialized international (mainly cross-cluster) data-base for partner searching within technical textiles fields of demand and supply. An evaluation of policies for local upgrading through international collective actions should look after such types of results. But before and under the results, the evaluation should have, as in the previous section, a structural side concerning the architecture of collaborative relations spanning the various types recalled in section 4.

VIII. — Some conclusions on appropriate evaluation methods in a systemic perspective

55In section 5 we have argued that the international debate on policy evaluation is shifting the focus on the integration of such activity within a learning-based approach to policy making. This methodological evolution comes not only from the accumulation of theoretical reflections within the disciplines of policy evaluation, but also from the specific needs of the new policies. The applications discussed in this paper concern two types of policies of local industrial development. They contribute in particular public support to processes of district and cluster upgrading, through regional and international cross-cluster and trans-local relations and structures, in face of globalization challenges and opportunities.

56The exemplifications have shown one feature of the new policies and of the related evaluation methods. They do not target individual firms, but instead networks of various types of business, institutional, and research agents. The ways in which these networks are constituted and managed, and the architectures they take, impinge directly on the dynamic capabilities of districts and clusters. They are systems composed of networks, and may appropriately interact with larger innovation systems and international value chains by means of sub-sets of their networks. The analysis of such architectural features, and of how they are possibly influenced by public programs and actions, is part and parcel of the evaluation job. Focusing on dynamic capabilities, and not only on specific business performances at the end of specific public programs, is also necessary because districts and clusters, being complex social systems, have roundabout, step-wise, and partly surprising ways to show results from new embedded capabilities.

  • 12 We would just refer here to the evaluation activity illustrated in Lazzeretti et al.(2007).

57Furthermore this suggests that the evaluation job is a process which should intermingle not only with the policy process but also with the action of at least some of the operative networks directly involved. In the previous exemplifications we did not go in depth on this second suggestion, even if the cases illustrated could have allowed some direct illustration too. In fact all this could not have been contained in a single paper12. Finally, we would say that local and cluster upgrading in the flux of globalization asks, perhaps not peculiarly but evidently, upgraded policies and policy evaluation methods.

Haut de page


Abernathy F.-H., Volpe A., Weil D. (2006), « The Future of the Apparel and textile Industries: Prospects and Choices for Public and Private Sectors », in Environment and Planning A, 38, pp. 2207-2232.

Autio E. (1998), « Evaluation of RTD in Regional Systems of Innovation », European Planning Studies, vol. 6, n° 2, pp. 131-140.

Barnes T., Pashby I., Gibbons A. (2002), « Effective University -Industry Interaction: A Multi-case Evaluation of Collaborative R&D Projects », European Management Journal, vol. 20, n° 3, pp. 272-285.

Becattini G. (2004), Industrial Districts: A New Approach to Industrial Change, Cheltenham: Edward Elgar.

Bellandi M., Caloffi A. (2008a), « District internationalization and trans-local development », Entrepreneurship and Regional Development, vol. 20, November, pp. 517-532.

Bellandi M., Caloffi A. (2008b), The promotion of innovative networks in a regional innovation system perspective, The DISTRICT Regional Framework Operation at work: Innovation Pathways and Knowledge Economy, Bruxelles: DISTRICT INTERREGIIIC - DG Regional Policy, pp. 34-51.

Bellandi M., Caloffi A. (2010), « An analysis of regional policies promoting networks for innovation » European Planning Studies, vol. 18, n° 1, pp.67-82.

Bellandi M., Lazzeretti L., Zanni L., (eds.) (2008), Metodi e parametri di monitoraggio, accompagnamento e valutazione sostanziali su partnership per l’innovazione entro comunità di pratica e sistemi di relazioni industria-ricerca, Progetto DSS-TEC, Decision Support Systems on TEchnological Clusters, INTERREG IIIC SUD - Operazione Quadro Regionale DEPURE, Regione Toscana.

Belussi F. and Sammarra A. (eds.) (2009), Business Networks in Clusters and Industrial Districts. The Governance of the Global Value Chain, London and New York: Routledge.

Bonaccorsi A., Nesci F., (eds.) (2006), Bacini di competenze e processi di agglomerazione. I distretti tecnologici in Europa, Milano: Franco Angeli.

Brusco S. (1994), « Servizi reali, formazione professionale e competenze: una prospettiva », in Bellandi, M., Russo, M., (1994) (eds.), Distretti industriali e cambiamento economico locale, Torino: Rosenberg e Sellier.

Cooke P. (2005), « Regionally asymmetric knowledge capabilities and open innovation. Exploring “Globalisation 2”-A new model of industry organisation », Research Policy, vol. 34, n° 8, pp. 1128-1149.

Cooke P, Lazzeretti L. (eds) (2007), Creative Cities, Cultural Clusters and Local Economic Development, Cheltenham: Edward Elgar.

De Nooy W., Mrvar A., Batagelj V. (2005), Exploratory Social Network Analysis with Pajek, New York: Cambridge University Press.

De Propris L. (2008), « Trust and social capital in glo-cal networks », in P. Bianchi M.-D. Parrilli and R. Sugden (eds.), High Technology, Productivity and Networks: A Systemic Approach to SME Development, Palgrave Macmillan: Basingstoke.

Diez M.-A. (2001), « The Evaluation of Regional Innovation and Cluster Policies: Towards a Participatory Approach », European Planning Studies, vol. 9, n° 7, pp. 907-923.

DISTRICT (edited by) (2008), The DISTRICT Regional Framework Operation At Work: Innovation Pathways and Knowledge Economy, Bruxelles: INTERREG-IIIC.

Eickelpasch A., Fritsch M. (2005), « Contests for cooperation-A new approach in German innovation policy », Research Policy, vol. 34, n° 8, pp. 1269-1282.

Enright M.-J. and Ffowcs-Williams I. (2001), « Local Partnerships, Clusters and SME Globalization », inEnhancing SME Competitiveness, Paris: OECD.

Gereffi G., Humphrey J., and Sturgeon T. (2005), « The Governance of Global Value Chains », Review of International Political Economy, vol. 12, n° 1, pp. 78-104.

Guba E., Lincoln Y. (1989), Fourth-Generation Evaluation, London, New Delhi: SAGE, Thousand Oaks.

Hendry C., Brown J. and De Fillippi R.-J. (2000), « Regional Clustering of High Technology-based Firms: Opto-Electronics in Three Countries », Regional Studies, vol. 34, n° 2, pp. 129-144.

Humphrey J., Schmitz H. (2004) « Governance in global value chains », in Schmitz H., (ed.), Local Enterprises in the Global Economy, Cheltenham: Edward Elgar.

Klofsten M., Jones-Evans D. and Schärberg C., (1999), « Growing The Linköping Technopole- A Longitudinal Study of Triple Helix Development in Sweden », vol. 24, n° 2-3, pp. 125-138.

Kuhlmann S. (1998), « Moderation of policy making? Science and Technology Policy evaluation beyond impact measurement. The case of Germany », Evaluation, vol. 4, n° 2, pp. 130-148.

Landabaso M., Oughton C., Morgan K. (2001), « Innovative Networks and Regional Policy in Europe », in Koschatzky K., Kulicke M. and Zenker A. (eds.), Innovation networks: Concepts and Challenges in the European Perspective, Heidelberg: Physica-Verlag.

Lane D.-A., Maxfield R. (1996), « Strategy Under Complexity: Fostering Generative Relationships », Long Range Planning, vol. 29, n° 2, pp. 115-131.

Lazzeretti L., Bellandi M., Pilotti L., Zanni L. (2007) (eds.), MARI -Gruppo di Monitoraggio e Analisi Reti di Innovazione. Mis.1.7 Docup Ob. 2 2000-2006 - Reti di Benchmarking and Foresight, Rapporto di Ricerca, Regione Toscana, Firenze.

Lengrand L. (ed.) (2006), Smart Innovation: A practical Guide to Evaluating Innovation Programmes, DG Enterprise and Industry, Bruxelles.

Longhi C. (2005), « Local systems and Networks in the globalization process », in Quadrio Curzio A. and M. Fortis (eds.), Research and technological innovation. The challenge for a new Europe, Heidelberg: Physica-Verlag.

Mehrotra S. and Biggeri M. (eds.) (2007), Asian informal workers. Global risks, local protection, Abingdon UK: Routledge.

Meyer-Stamer J. (2004), « Paradoxes and Ironies of Locational Policy in the New Global Economy », in Schmitz H. (ed.), Local Enterprises in the Global Economy: Issues of Governance and Upgrading, Cheltenham: Edward Elgar.

Patton M.-Q. (1997), Utilization-Focused Evaluation, London: SAGE Publications.

Porter M., and Ketels C. (2009), « Clusters and Industrial Districts: Common Roots, Different Perspectives », in Becattini G., Bellandi M. and De Propris L. (eds.), The Handbook of Industrial Districts, Cheltenham: Edward Elgar.

Priedl I., Hochmayer K., Jaeger H. (2008), Innovation Policy Impact Assessment at Regional Level: Lower Austria’s Regional Deployment, Impactscan, Ideum, Lower Austria.

Rajahonka M., Valtakari M. (2005), NETS - Networks of the future 2001-2005. Evaluation Report, TEKES, Helsinki.

Rullani E. (2003), « Il ruolo dei distretti industriali in Europa: la lezione dell’Italia », in Formez e TeDis, La governance dei processi di internazionalizzazione dei sistemi produttivi locali, Roma: Formez.

Russo M., and Rossi F. (2008), « Cooperation partnerships and innovation. A complex systems perspective to the design, management and evaluation of a EU regional innovation policy programme », in Evaluation, vol. 14, n° 3.

Sabel C.-F. (2004), Districts on the Move. Note on the Tedis Survey of the Internazionalization of District Firms, Columbia Law School, mimeo.

Saxenian A. and Sabel C. (2009), « The New Argonauts, Global Search and Local Institution Building », in Becattini G., Bellandi M. and De Propris L. (eds.), The Handbook of Industrial Districts, Cheltenham: Edward Elgar.

Siegel D.-S., Waldman D., Link A., (2003), « Assessing the impact of organizational practices on the relative productivity of university technology transfer offices: an exploratory study », Research Policy, vol. 32, n° 1, pp. 27-48.

Sugden, R., Wei P. and Wilson, J.-R. (2006), « Clusters, Governance and the Development of Local Economies: A Framework for Case Studies », in Pitelis C., Sugden R., Wilson J. (eds.), Clusters and Globalization: The Development of Economies, Cheltenham: Edward Elgar.

Trullen J., and Boix R. (2001), « Città creative nell’era della conoscenza », Sviluppo locale, n° 13, pp. 41-60.

Wasserman S., Faust K. (1994), Social Network Analysis: Methods and Applications, New York: Cambridge University Press.

Haut de page


1 See for example Abernathy et al. (2006), Belussi and Sammara (2009), Gereffi et al. (2005); Humphrey and Schmitz (2004), Longhi (2005), Porter and Ketel (2009).

2 Here « region » means both a delimited regional milieu hosting a set of localities and clusters showing common languages and traditions, and the lower territorial level of regulatory power influencing the coordination and distribution of key public resources among that set of localities and clusters.

3 In the case of the Tuscany Region, in Italy, which is home of an important optoelectronics cluster, specific public funds has been allocated to experimenting the potential of the application of laser technologies to the restoration of cultural assets, to the manufacturing of textiles and to other industries in the region.

4 Promising prospects for twinning are related to general economic and social foundations, in particular: a) some productive activities distributed between the (potentially) twinned contexts (cluster/locality) are (partially and potentially) complementary, and each of them is embedded in its context in such a way as to promote productivity, learning and creativity; and b) the twinned contexts have similarities in terms of institutional and organizational structures and approaches, and they are such as to help the building of reciprocal understanding and trust on specific projects (segmented virtual proximity): see Bellandi and Caloffi (2008).

5 It is still possible for strong districts and clusters – say Italian ones with respect to clusters in areas of developing countries – to build internal joint action for constituting manufacturing and trade enclaves in developing areas, and accessing cheap factors of production and growing markets. Indeed, this is a replica of MNE strategies. It can work, but MNEs are more skilled at controlling opportunities and risks in this field.

6 What is illustrated here is extracted from Bellandi and Caloffi (2008a), which is the result of DSS-TEC project funded under the DEPURE – INTERREGIIIC RFO and implemented during the period March 2007 – June 2008, and Bellandi and Caloffi (2008b) realized within the DISTRICT program initiatives – INTERREG-IIIC RFO and implemented during the period April 2006 – September 2007. See also Lazzeretti et al. (2007).

7 The database includes a set of interventions implemented within the DOCUP-SPD 2000-2006 and the RPIA-ITT programme (Regional Programme of Innovative Actions « Innovazione Tecnologica in Toscana » 2001-2004, funded within the ERDF Innovative Actions framework), implemented by the Tuscan regional government during the programming period 2000-2006. As documented by the studies of Eickelpasch and Fritsch (2005), these types of initiative, eliciting the growth of self-organised co-operation networks in research and development, have been promoted in several European regions.

8 The database includes all the projects funded in the period under consideration, i.e. 122 projects implemented by 122 networks of innovators. A social network analysis has been applied to the dataset (Wasserman and Faust, 1994; de Nooy et al., 2005). The various agents are connected through co-membership relations in innovation projects. In particular, it is assumed here that two agents are directly connected when they participate in the same innovative project. The web of agents participating in two different innovative projects may be also indirectly connected by the activity of agents operating within both projects.

9 The enterprises show a degree of local containment of the relations which is above the average (41 % against 37 % of the total network).

10 As regards the « traditional industries », we refer to the definition adopted in the various funding specifications: textiles/clothing, footwear, marble, furniture and light (instrumental) mechanics. The public funding was aimed at promoting IT technology within these sectors.

11 It is a « Technological district » (Bonaccorsi e Nesci, 2006) as defined by the regional policies.

12 We would just refer here to the evaluation activity illustrated in Lazzeretti et al.(2007).

Haut de page

Table des illustrations

Titre Figure 1 : Architecture of the central relations of the sub-networks considered
Légende Key : Black boxes are universities and research centres; Black circles are firms; White boxes are service centres; White circles are business associations; Black triangles are incubators; Black box with a white circles are other public organizations. Black lines are direct relations.
Fichier image/jpeg, 87k
Haut de page

Pour citer cet article

Référence papier

Marco Bellandi et Annalisa Caloffi, « Towards a framework for the evaluation of policies of cluster upgrading and innovation », Revue d'économie industrielle, 129-130 | 2010, 259-276.

Référence électronique

Marco Bellandi et Annalisa Caloffi, « Towards a framework for the evaluation of policies of cluster upgrading and innovation », Revue d'économie industrielle [En ligne], 129-130 | 1er et 2e trimestres 2010, document 11, mis en ligne le 15 juin 2012, consulté le 19 mai 2022. URL : ; DOI :

Haut de page


Marco Bellandi

Università di Firenze: (, Dipartimento di Scienze economiche, Via delle Pandette 9, 50127 Firenze, Italia

Articles du même auteur

Annalisa Caloffi

Università di Firenze: (, Dipartimento di Scienze economiche, Via delle Pandette 9, 50127 Firenze, Italia

Articles du même auteur

Haut de page

Droits d'auteur

© Revue d’économie industrielle

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search