Which kind of economic networks boost innovation?

Social and economic networks affect innovation and economic growth through facilitating diffusion of knowledge and information (Granovetter 2005). However, networks are associated with externalities, as their initial costs are borne mostly by network organisers while their benefits can be shared with network members. Because of these externalities, policy interventions may be necessary to develop networks to the optimal level and maximise social welfare.

What are effective policy targets for creating firm networks?

In practice, policies supporting development of firm networks – such as subsidies for trade exhibitions and research collaborations – have been implemented. However, which policy measures are effective in creating firm networks is not evident because it is not clear what types of networks contribute to knowledge diffusion and innovation. For example, some researchers have found that strong and dense networks with trust within regional or organisational communities lead to more effective knowledge diffusion and thus more innovation (Ahuja 2000, Phelps 2010). In other studies, however, weak ties with outsiders were found to be more effective sources of information because the knowledge of outsiders is more diversified than that of partners within the community (Burt 1992, Granovetter 1973).

Based on our recent work (Todo et al. 2014, 2015), this column examines which structures of supply chain networks are effective for firms’ performance and resilience and provides some policy implications. In both studies, we use a large unique firm-level data set that includes information on major supply chain ties in the whole Japanese economy.

Knowledge diffusion through supply chain networks

In recent research (Todo et al. 2015), we estimate how different structures of supply chain networks selectively affect firms’ productivity and innovative capability as measured by sales per worker and the number of registered patents, respectively.

In particular, we focus on the following two characteristics of network structures. First, we distinguish between supply chain partners (buyers and suppliers) within and outside of the same prefecture to find knowledge diffusion within and across regions through supply chains. We presume that the knowledge of distant partners is more diversified than that of neighbouring partners, based on the argument of Burt (1992) and Granovetter (1973). In fact, this presumption is supported by Figure 1, which shows a weighted share of patents in each technology class for each of the nine most-advanced prefectures in Japan, constructed from data for all registered patents there. This figure indicates that major technology classes vary across regions.

Second, we examine how a firm’s performance is affected by the density of its ego network, measured by the ratio of the number of actual ties between the firm’s supply chain partners to the number of all possible ties between the partners. Because the density of a firm’s ego network can indicate the strength of supply chain networks, this analysis can investigate whether such strength matters to knowledge diffusion.

Figure 1. Geographic diversity of knowledge

Source: Todo et al. (2015).

We start with estimation of the effect of ties with suppliers, using simultaneous-equation models that incorporate dynamics of firms’ performance and networks.

• Our results show that ties with distant suppliers outside of the prefecture improve firms’ performance while ties with neighbouring suppliers within the same prefecture do not.

These results are illustrated in panels A and B of Figure 2, which shows simulation results predicting the percentage changes in sales per worker and total sales when the number of neighbouring suppliers (A) or distant suppliers (B) increases by 50%. We presume that because intermediates from distant suppliers embody more diversified knowledge than those from neighbouring suppliers, using intermediates from distant firms promotes firms’ productivity due to the benefits from diversity.

Figure 2. Percentage change in sales per worker and total sales from simulation

Source: Todo et al. (2015).

The results are opposite for the links with clients.

• Ties with neighbouring buyers improve productivity while those with distant buyers do not.

We argue that the geographical role of networks with suppliers and buyers is different because the knowledge of suppliers diffuses to buyers in an embodied fashion whereas the transfer of knowledge from buyers to suppliers requires face-to-face communication. For example, Toyota Motor Corporation often organises associations of its suppliers in which it provides them with valuable technical and managerial assistance (Dyer and Nobeoka 2002). In this case, geographic distance can be a barrier to diffusion of disembodied knowledge.

• Furthermore, ties with distant suppliers and clients improve innovative capability as measured by the number of registered patents, while ties with neighbouring suppliers or clients do not.

This evidence suggests that in innovation activities, diversified knowledge from distant partners is more important than in production activities.

The other focus of our study was the density of a firm’s ego network that measures the degree of interaction between the firm’s partners. We find that the density has a negative effect on productivity and innovative capability, as shown in panel C of Figure 2.

• This result implies that the knowledge of densely interconnected firms is largely redundant, probably because they have already shared information with each other.

Overall, our results emphasise the importance of diversified partners in knowledge diffusion through supply chain networks. Face-to-face communication with neighbouring partners works well to promote diffusion of knowledge for production activities, but this is not sufficient for innovation activities as the knowledge of firms within the same region overlaps. Ties with firms outside of the region are necessary to bring fresh information to the region.

Economic resilience through diversified networks

In another study (Todo et al. 2014), we examine the effects of diversity of supply chain partners on economic resilience to natural disasters. Using the data described above, merged with firm-level data collected after the great east Japan earthquake in 2011, we estimated how the short- and long-term recoveries of firms in the impacted areas are affected by supply chain ties within and outside of the areas.

We found that ties with distant partners promoted short-run recovery from the earthquake, i.e., the restart of firm operations after the earthquake (Figure 3). This is probably because distant partners were more likely to be unaffected by the earthquake and thus could provide support to damaged firms. Toyota, for example, dispatched thousands of workers to support its factories and suppliers in the impacted areas.

Figure 3. Distant ties and recovery from natural disasters

Source: Todo et al. (2014).

• This result implies that geographically long supply chains contribute to economic resilience.

This conclusion is contrary to the argument that firms should not rely on long supply chains because they disseminate negative shocks. The pessimistic view of long supply chains is over-emphasised, ignoring the positive aspects of distant partners. Because we also find that ties between firms in the impacted areas facilitated longer-term sales growth, we confirm the importance of diverse supply chain networks for firm performance.

Industrial cluster policies

Our results provide implications for policies promoting the development of industrial clusters, which are generally considered an important driver of growth in regional economies (Rosenthal and Strange 2004). Although the policies of national and regional governments are often aimed at fostering industrial clusters, their outcomes have been mixed. For example, policies such as the high-tech offensive Bayern in Germany and the industrial cluster plan in Japan stimulate innovation in their target regions (Falck et al. 2010, Nishimura and Okamuro 2011), while the local productive systems in France and the technopolis in Japan failed to raise regional productivity (Martin et al. 2011, Okubo and Tomiura 2012).

Our results indicate that promoting firm networks should be an important part of cluster policies, besides other standard policies such as tax incentives to establishments of plants. In addition, creating ties with outsiders, e.g., firms outside of the region, is at least as important as – or probably more important – than creating ties within the region. Examples of policy measures for this purpose include supporting business matching with distant (including foreign) buyers and financial institution and research collaboration between firms and universities across regions. The importance of cluster policies for creating such ties with outsiders that promote diffusion of diversified knowledge should not be underestimated.

References

Ahuja G (2000), “Collaboration Networks, Structural Holes, and Innovation: A Longitudinal Study”,Administrative Science Quarterly 45, 425-455.

Burt R S (1992), Structural Holes: The Social Structure of Competition, Harvard University Press: Cambridge.

Dyer J, and K Nobeoka (2002), “Creating and Managing a High Performance Knowledge-Sharing Network: The Toyota Case”, Strategic Management Journal, 21, 345-367.

Falck O, S Heblich S, Kipar S (2010), “Industrial Innovation: Direct Evidence from a Cluster-Oriented Policy”, Regional Science and Urban Economics, 40, 574-582.

Granovetter M (2005), “The Impact of Social Structure on Economic Outcomes,” Journal of Economic Perspectives, 19; 33-50.

Granovetter M S (1973), “The Strength of Weak Ties”, American Journal of Sociology, 78; 1360-1380.

Martin P, T Mayer, F Mayneris (2011), “Public Support to Clusters: A Firm Level Study of French ‘Local Productive Systems'”, Regional Science and Urban Economics 41,108-123.

Nishimura J, H Okamuro (2011), “Subsidy and Networking: The Effects of Direct and Indirect Support Programs of the Cluster Policy”, Research Policy 40, 714-727.

Okubo T, E Tomiura (2012), “Industrial Relocation Policy, Productivity and Heterogeneous Plants: Evidence from Japan”, Regional Science and Urban Economics, 42, 230-239.

Phelps C C (2010), “A Longitudinal Study of the Influence of Alliance Network Structure and Composition on Firm Exploratory Innovation”, Academy of Management Journal, 53, 890-913.

Rosenthal S S, and W C Strange (2004), “Evidence on the Nature and Sources of Agglomeration Economies”, In: Henderson JV, Thisse J-F (Eds), Handbook of Regional and Urban Economics, vol. 4. Elsevier BV, Amsterdam, 2119-2171.

Todo Y, P Matous, and H Inoue (2015), “The Strength of Long Ties and the Weakness of Strong Ties: Knowledge diffusion through supply chain networks“, RIETI Discussion Paper, No 15-E-034.

Todo Y, K Nakajima, and P Matous (2014), “How Do Supply Chain Networks Affect the Resilience of Firms to Natural Disasters? Evidence from the Great East Japan Earthquake”, Journal of Regional Science, 55, 209-229.

This article is published in collaboration with VoxEU. Publication does not imply endorsement of views by the World Economic Forum.

To keep up with the Agenda subscribe to our weekly newsletter.

Authors: Yasuyuki Todo is a Professor in the Faculty of Political Science and Economics, Waseda University. Petr Matous works in the Complex Systems Research Group in the Faculty of Engineering and IT, University of Sydney.

Image: A man walks past buildings at the central business district of Singapore February 14, 2007. REUTERS/Nicky Loh.