4.1. Why develop competitiveness poles?

Up to this point, we have considered innovation and production ecosystems as processes which develop and acquire legitimacy from the perspective of public policy and from that of the actors involved. We will now consider the effective policies applied by public authorities in France in the specific context of competitiveness poles, seen as a possible incarnation of these ecosystems. Note, however, that this assimilation is somewhat erroneous: competitiveness poles participate in the development of the local ecosystem (in the broadest sense of the term), applying a logic of subsidiarity in their relations with other actors in order to avoid replicating existing offers of service. Issues relating to efficient articulation between different actors may be encountered on three levels: within competitiveness poles, between these clusters and other organizations present in the same area, and between these clusters and clusters located in other countries.

Competitiveness poles may be considered as elements of a new industrial policy and “as a means of working around both European constraints with regard to industrial policies and the prohibition of direct subvention. In this case, geography does not only support competitiveness, but also serves to further the role of the State in the economy” [DUR 08].

The cited document specifies that the new industrial policy:

“constitutes a necessary improvement to methods of combining territory, innovation and industry. The creation of closer links between industrial actors, scientists and training establishments, including anchoring in a single territory, following the cluster model, constitutes:

• – a source of innovation: proximity stimulates the circulation of information and skills, thus facilitating the emergence of innovative projects;
• – a source of attractiveness: the concentration of actors in a single territory increases international visibility;
• – a hindrance to delocalization: the competitiveness of a company is linked to its territorial implantation, due to the presence of essential skills and strategic partners.

Based on a vision shared by the different actors involved, each competitiveness pole develops its own five-year plan, allowing it to:

• – consolidate partnerships between different partners with recognized and complementary skills;
• – foster the emergence of strategic collaborative R&D projects, which may benefit from public financial aid, notably via the fonds unique interministériel (FUI, unified interministerial fund);
• – promote a global environment which favors innovation and promotes the interests of the actors involved in the cluster, via actions intended to boost, mutualize or support the activities of cluster members in areas such as access to private funding, international development, industrial property, provisional management of skills and human resources, etc. These different actions are supported by the cluster’s innovation and growth ecosystem” [DUR 08].

The competitiveness pole design is based on a total of four elements: a central objective of innovation, a partnership-based logic, the development of dynamic and growth-creating ecosystems, and territorial anchoring.

The industrial cluster is presented, in the literature, as a permanent feature and a significant characteristic of the organization of economic activities. In the French example, this dynamic is driven by public policies; through the fonds unique interministériel, which provides funding for the best R&D projects and innovation platforms, public action supports and channels R&D efforts within competitiveness poles. This action is strengthened in the cases where the state supports development, alongside territorial authorities, via the participation of different institutions such as the Agence nationale de la recherche (National Research Agency) or the Caisse des Dépôts et Consignations (Deposits and Consignments Fund). There is a clear focus on the production of new knowledge, with the aim of reaching critical R&D thresholds by fostering partnerships and collaborations. The basic idea is that clusters possess emerging properties, i.e. when businesses, public actors and institutions form a coordinated ecosystem, their collective resources are greater than the sum of their parts, seen in terms of returns for each actor.

4.2. Competitiveness poles and the legacy of systèmes productifs locaux (SPL)

Competitiveness poles are only indirectly connected with their predecessors, the systèmes productifs locaux (SPL, local productive systems). SPLs were intended to promote inter-company collaboration but, unlike competitiveness poles, they did not directly include research and training elements. The aims of SPL were more modest, relating to the creation of a shared brand or logo, the establishment of collective export structures, etc. Finally, “the SPL policy was transformed into a permanent call for projects, with an increasingly restrictive specification” [DUR 08].

The authors focused on the economic performances obtained in SPLs over the period from 1996 to 2004. Before joining an SPL, the selected actors displayed comparable economic performance to those of other companies (the control group). The study showed paradoxical results: integration into the SPL resulted in a significant decrease in productivity which was lower than the average obtained for other French companies in the sample group¹. The explanation put forward was that “companies entering SPLs are located in départements [counties] and belong to sectors with lower than average productivity” [DUR 08]. Taking account of these characteristics and of the situation of these companies in relation to the averages for the counties and sectors in question, the result is somewhat different: the positive difference in productivity in favor of candidate companies for inclusion in an SPL disappears in relation to the average for a given sector and county. There was no visible cluster effect and no observable incidence on productivity. There was no effect on employment. The only positive observation related to small, single-site companies; however, this effect was only seen on a short-term basis and was no longer visible two years later. “If the beneficial effect of cluster policies for small companies were confirmed, this would raise questions as to the relevance of focusing on larger groups in establishing competitiveness poles” [DUR 08].

In this light, it is hard to find much continuity between SPLs and competitiveness poles, in that the two forms of organization have significantly different objectives. SPLs were notably more closely linked to territorial development policies.

4.3. Analyzing

The first notable feature shown in Table 4.1 is an absence of scattering. Over the period 2006–2011, the fifteen clusters receiving the greatest support represented 81% of the FUI’s public finance efforts for collaborative R&D projects across all 71 clusters². This concentration of support becomes even more apparent when we observe the first five clusters, which received almost 50% of this same funding pool over the period in question. Considered in terms of typology (international clusters, clusters with international aims and regional clusters), international clusters and, to a lesser extent, potentially international clusters display a greater capacity to concentrate funding and to mobilize actors in important R&D projects. However, the issues remain in terms of technological specialization, in that each cluster covers an average of twelve key technologies.

Although there is an observable concentration of funding (taking account of not only the FUI but also the ANR and the OSEO), this does not translate to a strategic vision at the national level [BEA 12] or to a clear hierarchy of priorities (official government documentation highlights the existence of collaborative R&D projects in all economic sectors). Furthermore, as the report shows, the effectiveness of the chosen typology in directing results is questionable, particularly as the number and variety of national contacts (at the ministerial level) and local (regional) contacts assigned to each cluster does not promote efficiency or clarity within the mechanism.

The administrative complexity and instability of certain structures is clearly highlighted by the following comment:

“For poles certified in 2005, a R&D zone was identified at communal level. Delimitation contributed to the sustained mobilization of territorial collectives, particularly regional councils, in supporting. However, the R&D zoning of poles does not seem to have produced the expected results. The associated fiscal advantages have almost all been cancelled. Funding matching for R&D projects is not a decisive factor in company implantation decisions. Moreover, the absence of R&D funding for clusters certified since 2007 has resulted in inequalities in the treatment of clusters. The adaptation of current R&D zones would require a decree to be issued by the Conseil d’État [State Council], which makes evolution more difficult. The current regulatory mechanism for R&D zoning of competitiveness poles therefore needs to be simplified, for example through the adoption of a contractual definition” [BEA 12].

In terms of performance over the period 2008–2011, “2,500 of the R&D projects carried out within the clusters resulted in innovations, almost three-fourths of which related to products and processes. Innovations relating to services, organization and marketing constitute a minority… Around 25% of the projects resulted in an innovation” [BEA 12]. The report specifies that approximately 1,000 patents were submitted (mostly in the fields of ICT, biotech and health, and energy), representing between 1 and 1.5% of patents submitted in France; over the same period, the clusters in question accounted for 4.5% of R&D spending. These R&D projects resulted in the creation of 93 start-ups, 5% of total annual creations of innovative companies in France. However, a slight improvement can be seen in more recent times. A total of 1,526 collaborative R&D projects were certified and funded over the period from 2005 to 2014, producing 100 start-ups. R&D spending was €6.5 billion, accounting for 13.5% of total R&D spending. Furthermore, in 2013, 42% of R&D spending on projects associated with these clusters and supported by the FUI went to small- or medium-sized companies.

More recent research has highlighted other interesting results [FRA 16b] concerning the effect of cluster membership on company R&D activities, measuring the difference between the R&D behaviors of companies belonging to clusters and those of companies outside of the clusters (the control group). The selected sample concerned companies with total annual R&D spending of less than €16 million. The study revealed three beneficial effects:

• – a leverage effect generated by public funding: “companies receiving an average of €103,000 of public funding in 2012 increased their own research spending for the same year by €474,000” [FRA 16b];
• – an effect on employment, in terms of recruitment of R&D personnel with, on average, 2.5 additional staff members in 2007 and 6.5 additional staff members in 2012;
• – an effect on R&D output, visible from 2010 onwards, with two additional patents per company within a cluster.

In total, the report estimates that cluster membership has a significant effect on the R&D activities of small- and medium-sized companies: “a positive effect can be seen from 2007 onwards in terms of R&D personnel and from 2008 onwards in terms of self-funding, public funding and increased patent submissions” [FRA 16b].

However, the combined evolution of several R&D support systems means that the results are hard to interpret. Over the period 2006–2009:

“Whilst participation in poles seems to increase R&D expenditure [of SMBs and TIEs], we have not been able to precisely isolate the effects of the pole system from that of the CIR [crédit impôt-recherche] reform of 2008. Over the latter part of the period in question, companies active within poles made heavy use of the CIR, greatly reducing their R&D costs. However, it is possible that the pole companies found it easier to use the CIR than other companies, as eligibility for the tax credit was easier to prove for companies already in receipt of direct aid. This direct aid generally did not cover all of companies’ intended spending on a project (for FUI projects, funding is calibrated to cover around 30% of planned spending); the increase in R&D spending within poles may be due to the fact that companies were able to obtain further funding by making use of the CIR.

Furthermore, the evaluation is based on a comparison between companies in poles and companies outside of poles. This evaluation does not allow us to eliminate the hypothesis that all R&D subsidies, notably CIR, has a substitution effect for all companies, whereby private spending is replaced by funded spending” [DOR 13].

Two additional weaknesses should be noted. The launch phase of R&D projects reflects a genuine collaborative dynamic; however, this element is prioritized over the development and commercialization phases, i.e. bringing products to market. In other terms, collaborative projects are strongly focused on the creation of new technology, with insufficient attention to usage³. The final stages of the innovation cycle are not covered by public funding mechanisms, creating the impression of a break between the production of innovative ideas and consideration of their applications. Nevertheless, as we saw in Chapter 1, problems relating to scaling and the commercialization of innovative ideas are critical, and public or semi-public funding is needed to support the final phases of innovation.

In terms of collaboration, empirical studies on American clusters have shown that the dynamics of knowledge creation through collaboration evolve over the lifecycle of an industry or cluster [AUD 96]. Companies prioritize relations with universities and research bodies during the first phases of the cycle but this collaboration is then attenuated by the emergence of congestion effects, before undergoing significant reductions as the point of maturity is reached. The management of cooperative R&D projects is thus subject to an additional constraint, relating to the position of companies within the lifecycle of an industry or cluster.

Furthermore, the cluster development mechanism and the development of legitimacy by producers focus more on the extension of existing businesses (notably small- or medium-sized companies) than on the endogenous development of these organizational forms. The idea that clusters can develop even in the absence of agglomeration economies requires further investigation [GOL 13]. The innovation process also relies on knowledge and skills accumulated by companies; while big businesses generally take a Schumpeterian approach in prioritizing innovations which correspond to their own trajectories, they often struggle to recognize the importance of innovative ideas which do not directly relate to their own programs but which might open up interesting market perspectives. In this context, dynamics may self-sustain via the spin-off mechanism. Potential innovations appear as by-products of previous innovations: “in this sense, the process of opportunity recognition is serendipitous, i.e. the opportunity was discovered as an unintended outcome of activities with another purpose” [DEN 03]. Exploitation of an opportunity leads to a spin-off movement if the employee(s) contributing to the innovation are able to leave their host company to create a new business and if they can mobilize the other resources which are required, including the creation of more specific resources. Companies created in this way produce knowledge within the territory in question and opportunities for creation are often to be found at the point of intersection between existing clusters.

Work carried out in the USA in the automobile and semiconductor sectors, on tires and hard disks, and in bio-therapies shows that the creation of new firms is principally linked to the existence of positive external factors produced by major operators. These new firms are rapidly able to reach significant levels of performance, proving Gibrat’s law according to which the growth rate of big businesses is lower than that of small units. This is not a Marshall-style externality, but one that relates directly to entrepreneurial vigor [GOL 13]. Why is the performance of Silicon Valley so much higher than that of the Dallas cluster? Because the companies present in the first cluster have created more spin-offs, creating “fertile ground for innovation and explosive growth” [GOL 13]. The externalities of knowledge initially benefit spin-off companies, before having a more limited effect on incoming businesses. The arrival of businesses in a cluster may even result partly from the spin-off mechanism. Public authorities need to be attentive to these mechanisms and to develop the means of facilitating their development, notably in the form of fiscal measures.

More generally, clusters have a dual effect on entrepreneurship [DEL 14]. First, the growth rate of new establishments belonging to established companies increases with the presence of associated industries within the cluster. Second, the development of a cluster around an industry increases the pool of shared competitive resources and lowers the entry barriers for new firms. These two aspects partially confirm the spin-off effect mentioned previously.

4.4. Conclusion

Our first remark is of a methodological nature: political decision makers and specialists are faced with a problem of endogeneity. The dynamic driven by public authorities strengthens an autonomous and self-sustaining dynamic. Hence, companies and institutions, on the one hand, and collective resources, on the other hand, develop conjointly to form ecosystems which are a product of previous actions, the result of a co-development process of which “causality is difficult to attribute” [FEL 14]. Positive feedback loops, observed after the event, are not the result of chance or of an historical accident; they cannot be attributed to a single cause, as each component is necessary but not sufficient. Each cause forms part of the puzzle of localized economic development. This connects back to our remark on the emerging properties of clusters, according to which the resources created over the course of the process are greater than the sum of the contributions made by individual actors.

Duranton et al. [DUR 08] highlighted the difficulty of measuring cluster effects due to a number of technical problems. Firstly, some specific locations may be particularly well-endowed with specific resources necessary for an activity (as in the case of perfume manufacturing in Grasse, France). There is thus a natural effect that must be taken into account. Furthermore, geographic concentration may result in increased costs for companies. In these cases, only the most productive companies are likely to remain in the area, as firms with lower returns will be unable to meet the necessary costs. In other terms, the positive relationship between location and productivity is the result of a selection effect but it is impossible to establish the causal connection between these two variables. Public authorities must ensure they are well-informed concerning growth mechanisms before attempting any intervention aimed at reinforcing clusters.

The authors raise a further interesting point: the large number of certified clusters may potentially create a problem in terms of critical mass. As we saw in the previous chapter, many clusters in the USA failed to reach the necessary critical mass for sustainability, despite belonging to sectors which demonstrate significant growth. This issue requires further investigation. In spite of the concentration of assistance around certain clusters, the establishment of these ecosystems appears to respond to two contradictory objectives: competitiveness and equity. According to official documentation, collaborative projects may be found across the whole of the territory.

The perspective offered in Chapter 1 respects the logic of innovation, moving away from an overly regulatory understanding of clusters and an excessively administrative interpretation of innovation, often limited to the initial stages of the process. Innovation also creates opportunities which may be exploited, allowing the emergence of spin-offs and the development of clusters. While public policies are essential, they only offer a partial view of the creation and development of competitiveness poles, which are, moreover, excessively dependent on public funding. The approach taken in this work highlights the relational and organizational aspect, the need to develop legitimacy and the importance of exploiting opportunities, as a mechanism which is endogenous to ecosystem dynamics and which constitutes an ingredient for growth.