1
The Value of Knowledge-Intensive Clusters for
Multinational Corporations
A study of access to knowledge through labour mobility as a
driver of the location choice of multinational corporations.
University of Amsterdam
Master Thesis Business Studies
Adriënne de Beer
Student 10454128
Supervisor Dr. W. van der Aa
Spring 2014
2
Abstract
This thesis studies the value of knowledge-intensive clusters for multinational corporations. Its aim is to discover what perceived advantages drive the decision of a multinational corporation to locate within a knowledge-intensive cluster. A model is developed, based on the discussed literature, which proposes that, next to a set of other advantages, the main driver of the decision to locate within a cluster is the perceived access to knowledge. Next to that, the model proposes that knowledge, as required by a MNC, can be acquired by capturing the knowledge flows realized by a high rate of local labour mobility. In order to ascertain the relevance of access to knowledge through a high rate of local labour mobility, a qualitative research is performed of two clusters in the Netherlands. This research consists of a desk research and in-depth interviews on both the cluster level and the managerial level. Diverse sources like publications, government presentations, consultancy reports and 12 in-depth interviews with diverse actors, are used to answer the research question of this thesis and explore its proposed model.
The qualitative study finds that access to knowledge by being able to attract (international) talent from a position within the cluster, is one of the main perceived advantages of a cluster. This thesis contributes to the literature by contradicting the knowledge acquisition mechanism that is described by its proposed model. Instead of gaining access to knowledge by drawing from a local labour mobility, MNCs seem to gain access to knowledge by recruiting international talent. Next to this main advantage, several other advantages are found that are said to also drive the decision to locate within a cluster. As a final contribution, the discussion of this thesis will argue how proximity benefits can also grant access to knowledge by enabling MNCs to engage in relevant partnerships.
3
Index
1. Introduction ... 5
1.1 A New Era ... 5
1.2 The Geographical Cluster ... 6
1.3 The Multinational Corporation ... 8
1.4 The Research Question ... 10
2. A Framework on Clusters and their Advantages for MNCs ... 12
2.1 Clusters Defined ... 12
2.2 Dissecting the Phenomenon ... 15
2.2.1 Typology ... 15
2.2.2 Actors ... 18
2.3 Cluster Location Advantages ... 21
2.3.1 Proximity... 22 2.3.2 Complementarities ... 22 2.3.3 Networks ... 23 2.3.4 Start-up Climate ... 23 2.3.5 Access to Institutions ... 24 2.3.6 Access to knowledge ... 24 2.4 Access to Knowledge ... 25
2.4.1 Knowledge Flows Within the Knowledge-Intensive Cluster ... 26
2.4.2 Knowledge Acquisition through Labour Mobility ... 28
2.5 The Model ... 30
3. Methodology ... 34
3.1 Research Design ... 34
3.2 Data Collection ... 36
3.3 Interview Analysis ... 38
4. Analysis on the Cluster Level ... 40
4.1 The Netherlands ... 40
4.2 The High Tech Campus in Eindhoven ... 42
4.2.1 Markusen’s Typology ... 44
4.2.2 Actors ... 44
4.3 Food Valley in Wageningen ... 47
4
4.3.2 Actors ... 49
5. Analysis on the Managerial Level ... 54
5.1 The Ecosystem ... 54
5.2 Labour Mobility ... 58
5.2.1 Attracting talented workers ... 58
5.2.2 Labour Mobility Compared... 61
6. Conclusion and Discussion ... 64
6.1 Conclusion ... 64
6.2 A discussion on Access to Knowledge within Clusters ... 67
6.3 Managerial Implications ... 69
6.4 Limitations and Future Research ... 71
7. References ... 74
8. Appendices ... 82
8.1 Format of Letter to Respondents ... 82
8.2 Structure of the Interviews ... 84
5
Paradoxically, the enduring competitive advantages in a global economy lie in local things; knowledge, relationships and motivation that distant rivals cannot match.
– Michael Porter
1. Introduction
This chapter will begin with introducing the concept of the geographical cluster in a global context. It will explain how knowledge-intensive clusters are arguably still relevant in today’s globalized world. Next, the notion of the multinational corporation will be introduced. This part will emphasize the special attention that needs to be devoted to the location strategy of R&D departments of multinational corporations. Finally, after this introduction, this chapter will conclude with the research question of this thesis.
1.1 A New Era
The 1970s and 1980s resemble a transition period away from a time in which technological change was mainly incremental (Mytelka, 2000). Before these decades, plenty of time was available to either amortize huge investments in new products and processes, or to catch up with dominant technologies for producing relatively stable products. Protectionism of nation states provided both the opportunity of a nursing environment for infant industries and the threat of inertia of competition and innovation. Nowadays, the competitive landscape for most firms and countries could not be more different. Globalization has pushed firms, industries and nations out of their comfort zones. Knowledge-driven and innovation-based competition emerged and the rules of the game changed. The result is a new commercial reality; the emergence of global markets for standardized consumer products on a previously unimagined scale of magnitude (Levitt, 2001).
6 Today’s global economy is fast-paced and those that keep up and profit from enormous economies of scale are able to decimate competitors that are too slow and got trapped in old convictions about how the world works. Gone are the days when firms were able to sell outdated models to less-developed countries. Gone are the days when margins, revenues and profits abroad were generally higher than at home.
Now that companies can source capital, goods, information, and technology from around the world, often with the click of a mouse, much of the conventional wisdom about how companies and nations compete needs to be overhauled (Porter, 1998). The famous strategist Jay Barney (1991), widely recognized for his resource based view, argued that a strategy can only yield a firm a sustained competitive advantage when other firms are unable to duplicate the benefits of this strategy. In this light, anything that can be efficiently sourced from a distance via global networks is available to any company located anywhere in the world. In today’s global world most firms are able to source non-locally. Since this strategy remains easy to duplicate, it seems that efficient sourcing no longer holds a source of competitive advantage. Consequently, open global markets and global infrastructures raise question marks around the relevance of location for competition. .
1.2 The Geographical Cluster
Many authors have agreed that the concept of location is losing its significance (e.g. Pla-Barber and Puig, 2009; Amighini and Rabellotti, 2006; Mariotti, 2004), some even go as far as stating that both globalization and digitization are rendering geography dead (e.g. Cairncross, 1997; De Martino et al., 2006; Morgan, 2004). Taking into account this rational economic reasoning, companies are expected to be fairly indifferent when making their location decisions. However, if location is losing its meaning, then why are the odds of finding a world-class fashion shoe company in Italy still much higher than in almost any other place? If this were a single case, it could be considered a simple outlier. Surprisingly, this is not an outlier. What is more, spatial clustering occurs in almost any region all over the world. Today’s world map is covered by what economists like Porter (1998) have named
7 ‘clusters’; critical masses in one place of unusual success in particular fields. Clusters do not appear in only one industry, but occur across many. In their report, commissioned by the European Cluster Observatory, Sölvell et al. (2009) confirm that clusters appear in any form imaginable: in both high-tech areas and more traditional industries, in handicraft industries, in manufacturing as well as services, and in small- and large- firm dominated clusters. In other words, local clusters with a global reach are easy to distinguish throughout almost any industry.
Some clusters in today’s society are famous and widely recognized: IT and Internet in Silicon Valley, movies in Hollywood and Bollywood, the automotive production in Germany, financial services in Wall Street and in the City in London, mobile communication in Stockholm, etcetera. However, it is not only in these regions that the cluster phenomenon is evident. Michael Porter, professor at Harvard Business School and world famous for his work on strategy and the competitiveness of nations and regions, has identified industrial concentrations and regional specializations all around the globe. Even though the concept has probably always existed, it was Porter’s research that was of importance to the business environment and policy makers all around the world. He wondered why some firms, based in particular regions or business environments, were able to achieve global leading positions, while firms in other environments developed less sophisticated and innovative strategies (Sölvell et al., 2009). Even for companies located in regions with comparable prosperity levels, differences in success can be striking.
It is safe to continue by saying that location is still of great importance for some companies. Seeing as economic theory predicts that location is losing value under globalization, it is important to find an explanation for the continued success of the cluster. A careful examination of the existing literature provides a potential solution: even though old claimed advantages of cluster locating do not longer hold, new conditions arose that set the term for new sorts of advantages. In this light it is not location that became irrelevant, it is the attributed advantages that disappeared and made place for new ones. The previous decades gave rise to the emergence of a new form of competition (Porter,
8 1998 & 2000; Maskell & Malmberg, 1999; Hitt et al., 1997). Before, competition was heavenly driven by input costs, and as predicted by Ricardo (1817), locations with important endowments (such as fertile lands or a skilled labour base) enjoyed a comparative advantage that was competitively decisive. Now, as explained by Porter (1998), competition is far more dynamic. Instead of relying on a comparative advantage, companies can now use global sourcing to diminish input-cost disadvantages. Creating a competitive advantage has become more dependent on knowledge and innovation and requires a more efficient use of inputs.
1.3 The Multinational Corporation
A typical company structure resulting from the newly developed global sourcing opportunities is the multinational corporation (MNC). A MNC consists of a group of geographically dispersed and goal-disparate organizations that include its headquarters and the different national subsidiaries (Ghoshal & Bartlett, 1990). The existence and success of multinational corporations lies in the fact that they are ‘efficient vehicles for creating and transferring knowledge across borders’ (Kogut & Zander, 1993). For MNCs the location decision has great relevance for their strategy. Since these firms inherently contain the possibility of locating their subsidiaries anywhere around the world, considerations and motivations behind such a choice are made strategically. One important aspect of those strategies involves the globalization of the innovative activities of MNCs (Frost and Zhou, 2005).
Following the reasoning that competitive advantage creation relies more and more on knowledge and innovation, special attention should be devoted to the location decision of the research and development (R&D) department of the MNC. It is the R&D departments for many of these MNCs that form their centres of innovative activity. Especially with regard to MNCs within a knowledge-intensive industry, R&D-based knowledge creation is of vital importance for the company as a whole. With competition more and more based on innovative activities, Senior managers’ attention is now focused on R&D’s contribution to competitive advantage (Kerssens-van Drongelen & Bilderbeek,
9 1999). In this light, with regard to the location decision, one can intuitively see the attraction of positioning such a department in an environment characterized by high levels of innovative activity. Consequently, locating the R&D department of the MNC in a knowledge-intensive cluster, where knowledge may be expected to flow freely, would seem like the most logical approach to proceed with.
Location decisions are central to the literature on international business and focuses on motives for foreign direct investment (FDI); which suggests that firms may expand abroad to access new technologies, skills, or knowledge to be applied in their home countries (Cantwell, 1989; Chung and Alcácer, 2002). In this line of thinking, it is of great importance that MNCs situate their most innovative departments, the research and development (R&D) departments, in environments beneficial to providing them such access to relevant new technologies, skills or knowledge. Seeing as the R&D department requires relevant knowledge creation in order to innovate, one might wonder how knowledge actually enters the company.
By locating the R&D department within a knowledge-intensive cluster it seems that, next to in-house knowledge creation, a MNC expects to extract knowledge from that particular area. Those studies that examined some of the traditional Italian industrial districts (Russo, 1985; Brusco, 1990; Pyke et al., 1990) found that one of the explanations for the geographical concentration of innovative activities is that knowledge developed in a cluster or industrial district flows more easily within it, but more slowly outside and across its borders. Given the tacit and complex nature of most valuable knowledge, its acquisition can be difficult (Kogut and Zander, 1992). A significant part of this tacit knowledge that MNCs are looking to acquire can be found embedded in skilled workers. Consequently, Argote and Ingram (2000) argue that the movements of these skilled workers between firms – enabling them to apply their knowledge to new settings – is one of the most important mechanisms of effectively transferring tacit knowledge across firms.
10
1.4 The Research Question
Many authors, such as Song et al. (2003) have argued that human mobility can serve as a mechanism for the acquisition of externally developed knowledge and have researched the conditions under which the mobility of R&D engineers or other skilled workers is most likely to result into inter-firm knowledge transfers. Confirmation of the relevance of labour mobility as a tacit knowledge flow vehicle can be found with MNCs located within the role model of high-tech clusters: Silicon Valley. Through in-depth interviews with the personnel managers of Silicon Valley firms, Angel (2000) for example found that workers hired away from other, local, firms are an important source of technological knowledge for the semiconductor industry. High levels of inter-firm labour mobility allow Silicon Valley firms to go beyond their own manufacturing experience and draw upon the broader stock of knowledge developed within the production complex as a whole. By moving from one company to another, workers help create an environment in which all companies involved are enabled to build cumulatively upon a common stock of technological successes and failures.
As previous literature that links location strategies – in particular strategies concerning the location decision of the R&D department of the MNC – to knowledge-intensive clusters is scarce, this thesis will make contribute to its understanding.
From this reasoning the following research question follows:
What perceived advantage(s) drives MNCs to locate an R&D department within a knowledge-intensive cluster?
This thesis will hypothesize that MNCs are attracted to knowledge-intensive clusters because of their highly innovative character, and that MNCs expect to acquire the tacit knowledge they require for new product developments by drawing from a high rate of local skilled labour mobility.
After this introduction, this thesis will proceed by establishing a theoretical framework in chapter 2. This framework will first give an overview of the emergence of the concept and then develop a
11 definition of the cluster. Next, it will describe how clusters can be structured and explore some general advantages of locating within a cluster. Once this basis is laid, special attention will be devoted to innovation and knowledge flows within the knowledge-intensive cluster. At the end of the theoretical framework, a model for the research will follow. In order to explore this model, a methodology is composed in chapter 3. Chapter 4 will consist of the analysis on the cluster level and chapter 5 will offer an analysis on the managerial level. Finally, chapter 6 will offer conclusions, followed by a discussion, including managerial implications and limitations. At the end of this thesis the reader will find a list of references and the respective appendices.
12
2. A Framework on Clusters and their Advantages for MNCs
This chapter will provide a deeper understanding of the concept of the cluster and will start with offering a definition. Next, a typology of Markusen (1996) on the cluster-level and a distinction between different actors by Sölvell et al. (2010) on the firm-level will dissect the phenomenon. Understanding the cluster phenomenon by exploring the frameworks of these authors is necessary to achieve optimal results from the research. These first two sections will set the conditions as to what an area should look like in order to be classified as a knowledge-intensive cluster. Following these conditions will prevent the research from selecting and classifying any arbitrary area as a cluster fit for this research. Advantages found in an arbitrary area could significantly differ from advantages found in an area classified as a knowledge-intensive cluster. It is the combined strength of those latter advantages – that arguably offer benefits to MNCs they cannot find anywhere else – that are of interest here. These first two sections will therefore contribute to answering the research question by defining what the area from which advantages are derived, should look like. It will help explain why MNCs perceive clusters to be relevant.
Next, the chapter will continue with synthesizing an overview of location advantages of the knowledge-intensive cluster. This overview will be derived from a number of different sources. A special section will subsequently be devoted to knowledge acquisition and labour mobility, which will be proposed to be the main driving factor of the MNC location decision. Finally, as a conclusion, this chapter will construct a model that will function as guidance for the research.
2.1 Clusters Defined
The concept of the geographical cluster has been a widely recognized phenomenon in the economic world for a long time. Geographic concentrations of trades and companies in particular industries can be found throughout centuries of history. The idea that location and specialization of production are of importance can be traced as far back as to Adam Smith (1776). In his famous groundwork The
13 cheaper than the home nation can make it, the home nation should buy it from them with some part of the produce of their own industry; in such a way that they have an advantage. It was David Ricardo who elaborated upon Smith’s idea and developed the theory of comparative advantage. Propagating free trade in his book On the Principles of Political Economy and Taxation, Ricardo (1817) writes about the advantages of relative opportunities. In his theory each nation, ceteris paribus, should specialize its production and exports in those goods for which it holds the greatest comparative advantage or minimum comparative disadvantage in cost, relative to other nations. Consequently, each nation’s imports will consist of those goods for which it holds a comparative disadvantage. According to Ricardo, this organization of production would maximize each nation’s cost-efficiency.
Years passed, and the idea of specializing production in a certain location became more prevalent. Finally, it was Alfred Marshall who laid the foundation for the concept of the cluster. However, at that point he named it an ‘industrial district’, for the popular term ‘cluster’ was to follow several decades later. In his Principles of Economics (1920), Marshall discusses the externalities of specialized industrial locations. He compares the big, highly-integrated vertical firm with groupings of small firms. In this research he shows that most of the advantages of a large scale of production can also be achieved by a population of small-sized firms concentrated in one specific area. He even subscribes more advantages to this group of clustered firms than to the large vertical firm. Marshall argues that a concentrated population of small and medium-sized enterprises must consist of companies that are specialized in the different phases of the production process of a certain commodity, and must have their consequent labour supply located in the same specific location (Becattini, 1991).
This Marshallian view inspired many nations to regard the industrial district as a framework for conceptualizing both economic and social action. During the 1990s, in the context of the wider discussion from Fordism to post-Fordism and flexible specialization, the paradigm of industrial districts dominated most economic geography and planning studies (Hadjimichalis, 2006). However,
14 it was not until Porter (2000) before one clear definition became available. After many published works with vague descriptions it was Porter’s work that was widely accepted as the norm on cluster theory . According to Porter (2000), clusters can be defined as
‘Geographic concentrations of interconnected companies, specialized suppliers, service providers, firms in related industries, and associated institutions (e.g. universities, standard agencies, trade
associations) in a particular field that compete but also cooperate’
It must be noted that ‘associated institutions’ in this definition comprises both governmental and other institutions that offer specialized education, information and technical support to the area. Other institutions that one can think of consist of universities, training centres, think tanks, standard-setting agencies and the like. Next to that, Porter adds that many clusters members are often involved in active trade associations and other collective bodies that look after the clusters’ interests and stakeholders. Finally, as a specialist in competition, Porter emphasizes the importance of the role of location for generating a competitive advantage. Seeing as his work was highly influential and is regarded as the base stone for cluster theory; and seeing as Porter was employed by several governments to assist them at policy forming to stimulate cluster development, this thesis will adopt Porter’s definition of clusters.
As a final note, as explained by Martin and Sunley (2003), over the last decades many other economic geographers have studied local industrial specialization, spatial agglomeration and regional development. This has resulted into a whole series of neologisms that were added to identify the cluster phenomenon. These include: industrial districts, new industrial spaces, territorial production complexes, neo-Marshallian nodes, regional innovation milieus, network regions, learning regions and so on. As this thesis cannot cover the specific differences between all these different notions and since their differences are hardly relevant, they are here considered as interchangeable. The terms used in this thesis will be limited to ‘industrial district’ and ‘cluster’.
15 Now that this thesis has defined the cluster or industrial district by applying Porter’s definition, the next section will explore its typology on the cluster-level and actors on the firm-level. Comprehending these characteristics will be important to carefully select the knowledge-intensive clusters for the research later on.
2.2 Dissecting the Phenomenon
A cluster does not simply emerge overnight; When an industrial district emerges it does so along a path-dependent way shaped by the region’s culture and its history. How, why and in what form the cluster is created depends on a number of variable factors. These factors can either appear spontaneously or may be planned and together form the unique mix that shape the cluster. Next to that, a cluster is always unique to its location. No two clusters are the same and each knows its own characteristics and dynamics. However, in order to get a grasp on the cluster concept it is important to categorize them in some way. Even though they will never be exactly similar, applying a typology to the concept might help to gain more insight. This section will try to dissect the concept of the cluster and explore its structures.
2.2.1 Typology
Defining the boundaries of a cluster turns out to be a highly ambiguous business. According to Porter (2000), drawing cluster boundaries often is a matter of degree and involves a creative process informed by understanding the linkages and complementarities across industries and institutions that are most important to competition in a particular field. In other words, the boundaries of a cluster can be found by determining what companies and institutions generate strong spillovers, and are of high importance for productivity and innovation and what companies and institutions are not. Porter adds that boundaries relate to the distance over which informational, transactional, incentive and other flows occur. Clusters do not concern one single industry, but contain a number of linked industries and other entities important to competition. This description however, appears to be rather vague and has caused Porter to be faced with a fair amount of criticism. Martin and Sunley
16 (2003) for example, argue that the cluster has become a ‘chaotic concept’ and equates quite different types, processes and spatial scales of economic localization under a single, all-embracing universalistic notion. In their opinion, the boundaries as described by Porter are highly generic and indeterminate. Clear boundaries are lacking on both an industrial and geographical scale and do not specify the level of aggregation, the range of associated industries, the strength of the linkages or the degree of specialization.
Consequently, Porter’s definition alone is not enough. The cluster theory requires some more distinctions. After looking in to how clusters form, now categories need to be distinguished in which they merge to. In the literature on industrial districts, many researchers like Piore and Sabel (1984), Scott and Paul (1990) and Saxenian (1994) have tried to identify formation processes and structural configurations of the cluster. It was not until Markusen’s (1996) influential work in this area before a highly quoted typology of the industrial district was offered. The factors used by Markusen to design her theory go further than the usual assessments of organizational structures and include the role of the local and national state, large-firm influences, the level of embedding in local and non-local networks, developmental trends, district sustainability and contributions to social welfare (Coe, 2001). Markusen proposes four distinctive industrial spatial types: the Marshallian industrial district (with the Italianate variant), the hub-and-spoke district, the satellite industrial platform and the state-centred district.
Table 1 below gives a clear overview of Markusen’s distinction between clusters. The table is based on both Markusen’s (1996) and Coe’s (2001) work.
Type of Cluster Characteristics Firms Interdependencies Within the Cluster
Growth Prospect Marshallian Small and locally owned
firms, low scale economies
Extensive, committed local inter-firm trade, local cultural identity, strong institutional support
Dependent on synergies and fit with the local context
Italianate Small and locally owned firms, highly specialized, low scale economies
High degree of cooperative coordination among competitors, strong trade association and local governments, strong communal
Dependent on synergies and fitting in the local context
17 feeling
Hub-and-Spoke One/several large
vertically integrated firms surrounded by suppliers, high scale economies
Firms collaborate on a global scale, no cooperation between
competitors, suppliers depend on large firms
Dependent on growth prospects of the
industries and strategies of the large hub firms Satellite
Platforms
Branch facilities of externally based multi-plant firms, high scale economies
Low intra-district trade, low cooperation for sharing
risks/innovations, low institutional support.
Dependent on ability cluster to attract and retain potential tenants State-Anchored Public/non-profit entity
as a key firm surrounded by service suppliers, high scale economies
Short-term buy-sell relationships between entity and suppliers, low cooperation for sharing
risks/innovations
Dependent on prospects/national governmental decision-making for the core facility
Figure 1.An overview of Markusen’s (1996) typology of clusters, based on Coe (2001).
As becomes clear from real-life examples and as also argued by Markusen (1996), this typology is not definite. Every region is unique, and it is very rare to have a region fitting a certain stereotype exactly. Many districts, especially in larger city areas, display characteristics of all the above types. Nowadays, clusters appear to be more and more hybrid in comparison to Markusen’s typology and some believe that this model is somewhat outdated. One must bear in mind that the world has changed significantly in comparison to Markusen’s 1990s world.
Consequently, Porter’s and Markusen’s attempts to standardize the definition of the cluster has given rise to significant criticism. For example, Coe (2001) argues that Markusen’s types are mainly constructed for manufacturing production systems with distinct networks of suppliers and customers and does not capture the dynamics of industries with a less linear production system. Even more, Morgan (2004) adds that Porter’s definition is deceptively simple and that there is no such thing as a standard cluster. Generalizing from a small number of celebrated cases therefore comes with great hazards. However, in order to write this thesis, distinctions are necessary to sketch a broad overview of industrial districts in all its forms and sizes. Regions that do not fit any of these categories and display characteristics of multiple districts will be considered a hybrid. For example, Silicon Valley is a true hybrid district and can be said to contain elements of all four types of districts. A real-life industrial district can thus be a hybrid type or may mutate over time from one type to another.
18 As a conclusion, this thesis will assume that most clusters nowadays can be considered a hybrid of Markusen’s (1996) typology. When identifying a cluster area, the research will look for different characteristics as described by both Markusen (1996) and Porter (2000) to define the type and boundaries of such an area. Differences between these characteristics might influence consequent advantages of the area, which is why they are taken into account in order to answer the research question. The next section will continue with a focus on the firm-level and define what actors are operating within the ecosystem of a knowledge-intensive cluster.
2.2.2 Actors
Seeing as most knowledge-intensive clusters these days will probably be classified as a hybrid consisting of several of Markusen’s types, it might be more useful to explore their dynamics on the cluster-level. This section will show how, as argued by Sölvell et al. (2012), on the cluster-level it becomes possible to distinguish between different parties. These parties exist within the boundaries of the geographical cluster and are argued to, together, form a distinct ecosystem.
In their report on strengthening clusters and competitiveness in Europe, Sölvell et al. (2012) from the Cluster Observatory note that well-performing clusters are particularly beneficial places for innovation. They argue that clusters are relevant for innovation, because well-functioning clusters contain a critical mass of actors that can support each other, and arrange and rearrange resources in a flexible way. In order to make clear how, they dissect a cluster into a collection of linked actors of different kinds. Figure 2 below represents a way to show the interconnections between these actors. Within this figure there are five types of actors and each type follows a path to interact with others. In an ideal world all these paths are full of traffic; people move between actors, talk with others, bring news to others, discuss with others, change jobs, and tie the system together in a thousand different ways (Sölvell et al., 2012). In this world clusters are dynamic, knowledge becomes widespread and widely-shared, and collaboration ensures the efficient use of available resources.
19
Figure 2. Five types of actors in a cluster as proposed by Sölvell et al. (2012).
Sölvell et al. (2012) do note that critical mass of these actors alone is not enough; They must be connected in various ways and there must be mobility of resources and skills including technological spillovers. First, within this ecosystem of actors, the most significant type is the firm (within the industry). It is the firms that take the risks and bring innovations to the market and put them to the test of competition. Within their ecosystem, firms interact with other actors. Small firms interact with large firms, domestic firms interact with multinationals, etc. Within the cluster a network exists in which member firms use each other as buyers, as suppliers, as technology partners, as places to find trained staff, as sources of new ideas to imitate, or simply as an inspiration to aim higher and set more ambitious goals.
Second, research organizations, such as knowledge institutes, contribute to these innovations by producing new advanced knowledge. Third, educational organizations, such as special training centres, develop skills and talent for the region. Universities are a special case within this type of actors, since they play a double role by being both a research organization and an educational institution. They consequently can employ research groups, that produce cutting-edge knowledge in
20 relevant technology with regard to the cluster, and communicate this knowledge to member firms or inspire start-ups. Next to that, they can offer specialized education programmes or endow skills to graduates that are particularly fitted to the needs of the cluster.
Fourth, capital providers, such as business angels, venture capital firms and banks, provide the financial resources that are necessary for implementing innovations or new business models. The more involved they are with the cluster, the more they become experts in providing ‘smart money’ by getting better at assessing risks and opportunities with the cluster’s business field. Fifth, public organizations consisting of governmental and public bodies, represent actors that decide and implement policies with regard to public infrastructure investment, regulations, and many other factors of importance to innovation. Many layers of government can be involved, from national to local, plus even more different kinds of public agencies. The more experience these bodies gain within the cluster, the more they will be able to make decisions that promote the cluster and remove obstacles to progress.
In all these ways actors contribute to the success of firms and enable member firms to grow and be competitive (Sölvell et al., 2012). Of course this model of thinking applies to an idealistic scenario and in the real world this system is more flawed. Cluster communication is not always that well-tailored and mind-sets of members are not always tuned to openness or collaboration. Actors can follow their own interest (researchers who are more interested in academic publishing than commercializing new findings, banks that are hesitant to invest in new innovative businesses, etc.), instead of fitting their goals to the community. Sölvell et al. (2012) argue that this mal-alignment can be largely mediated by a well-functioning cluster organization. Innovation and R&D objectives are most critical to the larger cluster organizations. Next to that, they are often involved with cluster growth and development and investment attraction from the outside.
Concluding, a well-functioning cluster should contain five types of actors (industry, public organizations, research organizations, capital providers, and education institutions). In order to align
21 these actors, a cluster organization should be present to mediate between the different actors. What is more, a critical mass of these actors should be present and between them, mobility of resources and skills should exist. This thesis will look for clusters fitting the characteristics as described by Markusen (1996) and Sölvell et al. (2012). As this thesis wants to explore the value of knowledge-intensive clusters for MNCs, the research needs a classification of how these areas should look like. Clusters that do not meet these characteristics could be functioning in different ways and therefore generate different types of advantages. Using these criteria will help create a non-ambiguous sample for this research. The results that follow from the research will therefore provide a more trustworthy answer to the research question at hand. Using data from a cluster that, for example, does not classify as a geographical cluster could distort the results.
The next section will continue with the advantages for MNCs that can be found within knowledge-intensive clusters.
2.3 Cluster Location Advantages
Now that the concept and structure of the cluster have been explored, it is important to have a look at its ‘X-factor’. The mysterious acclaimed synergy of being part of a cluster, the 1+1=3 effect, the attractiveness of an area characterized by innovation; How can these things be explained and what sorts of advantages are offered that supposedly win MNCs over to choose for a knowledge-intensive cluster? Even though most cluster participants are not direct competitors but rather serve different segments of industries, they do share many common opportunities to productivity (Porter, 2000). In this light, many authors have ascribed different advantages to locating a firm (department) within a knowledge-intensive cluster.
This section will explore what advantages could steer MNCs when considering locating an R&D department within a knowledge-intensive cluster. After exploring the advantages in general, the focus will be laid upon the concept of innovation and knowledge flows in the next part. Since there are numerous benefits mentioned throughout the literature with many of them similar to each
22 other, this thesis classifies them under some common nominators. Bear in mind that clusters form a complex and holistic matter, which, even when categorized, offers advantages that are heavily interlinked with each other. The one does not exist without the influence of the other and vice versa.
2.3.1 Proximity
The most general advantage of locating within a cluster is that clusters increase firms’ competitiveness (Martin et al., 2002). As also discussed by Robertson and Langlois (1995), the most important advantage of (Marshallian) industrial districts can be derived from the proximity of firms. First, proximity allows for easier recruitment of specialized inputs (Porter, 2000). Despite the greater competition, the availability of specialized services and components is far greater at clusters; This provides firms with a superior (to other firms) or lower cost access than firms elsewhere. Second, proximity also grants firms access to information; It enables rapid exchanges of commercial and technical information through informal channels (Robertson and Langlois, 1995). As Porter adds (2000), sophisticated buyers in the region in combination with shared information, offer important insights regarding current buyer needs. For this reason, firms located within a cluster can perceive buyer trends much faster than isolated competitors can. Finally, in clusters where companies are dealing with low economies of scale (e.g. Marshallian districts), proximity reduces transaction costs to a practical minimum (Robertson & Langlois, 1995). In these cases, industrial districts represent competitive capitalism at its most efficient, providing district members with a clear advantage over isolated competitors.
2.3.2 Complementarities
A cluster facilitates complementarities and cost savings for the activities of cluster participants. Barkley and Henry (1997) argue that cost savings arise because the concentration of firms in the district enable internal localization economies. Porter (2000) distinguishes three types of consequent complementarities: complementary products for the buyer, marketing complementarities and complementarities due to a better alignment of participant activities. First, because firms are
co-23 located and consequent internal pressures for improvement of all parts arise within the district, it is easier to offer the end-consumer complementary products and services that are of great quality. Second, the proximity of a group of related firms and industries within a district enables a collective marketing strategy; It can even improve the reputation of the district as a whole in a certain field. Third, alignment of activities is more efficient for firms within a cluster than for isolated competitors because they maintain close linkages with suppliers, channels and downstream industries. This alignment might again lead to a higher general productivity (Martin & Sunley, 2003).
2.3.3 Networks
Not only can clusters develop internal economies of scale for cost savings, they also develop internal economies in terms of specific trading links and customer-supplier relationships (Morosini,2004). Barkley and Henry (1997) explain that clusters facilitate the development of links, cooperation and collaboration among area firms; they stimulate networking. These networks can be horizontal when they connect firms that need the same specialized services or technologies, or can be vertical when they connect firms performing different functions in the same value-added chain (Rosenfeld, 1992). Consequently, networks derived from co-location of firms can provide small firms with economies of scale and access to information and markets normally only availably to larger firms.
2.3.4 Start-up Climate
Clusters are very attractive for entrepreneurs as they enjoy a high rate of new firm formation (Martin and Sunley, 2003). Porter (2000) even goes as far as stating that most new businesses are formed in existing clusters; Clusters offer entrepreneurs better information, lower barriers to entry, an existent pool of potential local customers, established networks and available resources. Next to that, in these regions it is easier for firms to attract the attention from local financial markets that are familiar with the industry’s product markets and production processes (Barkley & Henry, 1997). Even more, investors in Marshallian districts often offer patient capital to entrepreneurs on very flexible terms.
24 2.3.5 Access to Institutions
A lot of successful clusters have transformed inputs that otherwise would be too costly for private firms into public or semi-public goods (Porter, 2000); This way it is possible to recruit specialized employees from local training programmes, to access specialized infrastructures and get advice from experts in local institutions at a low cost. Barkley and Henry (1997) add that clusters can design industrial development programmes that use available resources to cater for the needs of specific industries efficiently. Investment for these goods and programmes usually originates from trade associations or other collective entities.
2.3.6 Access to knowledge
A growing group of modern authors like Bathelt et al. (2004), Malmberg and Maskell (2002) and Maskell (2001) argue that the main advantage of knowledge-intensive clusters lies in the access that these areas provide to relevant knowledge. By locating within a cluster, MNCs can access and recombine inter-disciplinary sources of knowledge. With this acquired knowledge they are consequently able to improve existing products or increase the speed of new product innovations. As this theoretical framework will follow the work of these authors and argue that access to knowledge is indeed the main advantage of spatial clustering, this topic will be discussed thoroughly in the next two sections. For now, the provided information suffices to direct the conclusions of this section.
The figure below summarizes the advantages of spatial clustering as discussed in this section:
Advantages of Spatial Clustering
Proximity benefits - specialized inputs - access to information - trust
- low transaction costs
Robertson and Langlois (1995); Porter (2000)
Complementarities Barkley and Henry (1997); Porter (2000); Martin and Sunley (2003)
Networks Morosini (2004); Barkley and Henry
(1997); Rosenfeld (1992)
Start-up climate Martin and Sunley (2003); Porter (2000); Barkley and Henry (1997) Access to institutions Porter (2000); Barkley and Henry
25 (1997)
Access to knowledge
- New knowledge acquisition and recombination
- Increased speed of new product innovations
Bathelt et al. (2004), Malmberg and Maskell (2002), Maskell (2001), Tallman et al. (2004), Oettl and Agrawal (2008), Romer (1990)
Table 3. Advantages of spatial clustering.
As a conclusion, a number of advantages can be defined under some common nominators that play a part in influencing the MNC’s location decision. None of these advantages should be regarded on a stand-alone basis; One should remember that all are interwoven and inter-dependent. First, the proximity of actors within a cluster provides easy recruitment of specialized inputs, access to relevant information through informal channels, trust in other actors and low transaction costs. Second, complementarities exist in the form of complementary products, marketing complementarities and activity alignments. Third, networks exist that link actors and stimulate collaboration. Fourth, clusters offer a beneficial start-up climate in which entrepreneurs often have access to specialized patient capital. Fifth, firms within a cluster have access to institutions tailored to the area. Last, and most important, access to relevant knowledge results in new knowledge combinations and in a consequent increased speed of new product development. The next section will elaborate upon this last advantage and explain how local knowledge flows facilitate knowledge acquisition. Next to that, it will explain how labour mobility can function as a vehicle of this knowledge.
2.4 Access to Knowledge
While considering the first five types of advantages as discussed in the previous section, one might become slightly sceptical. Sure, it seems plausible that spatial clustering provides all those benefits, but as was mentioned before, the world is different now than in the 1990s, when a lot of the previously discussed literature was written. In the light of today’s world, many of those previous advantages can be argued to be easy to source on a national, or even a global level. By means of modern global access, the same networks and relationships can be managed and the same suppliers can be reached, wherever you are. For this reason the previously discussed advantages do not add
26 up to explaining the continuing success of clusters today. There must be something else that remains highly attractive about geographical knowledge-intensive clusters.
This section will explore knowledge flows within a cluster and the advantages that can be derived from consequent knowledge acquisition and recombination. The first part will explain how this mechanism forms an important, if not the most important advantage to MNCs. The second part will explain how firms can access this knowledge by means of labour mobility.
2.4.1 Knowledge Flows Within the Knowledge-Intensive Cluster
Acknowledging the improved sourcing possibilities due to modern global access, authors like Bathelt et al. (2004), Malmberg and Maskell (2002), Maskell (2001) and Tallman et al. (2004) put their main focus on knowledge-based advantages of the cluster. They argue that the main mechanism that makes it beneficial for a firm to be located in a spatial cluster, surrounded by other similar and related firms, lies in the following: innovation, knowledge creation and learning are all best understood if seen as the result of interactive processes where actors possessing different types of knowledge and competencies come together and exchange information with the aim to solve problems (Bathelt et al., 2004). The recombination of knowledge drives innovation (Oettl and Agrawal, 2008); a wider access to knowledge facilitates more efficient innovation by reducing the need to recreate what already exists elsewhere. Some literature that focuses on knowledge flows (that occur outside market mechanisms), like for example Romer (1990), even goes as far as stating that these flows form the central determinant of economic growth.
In this line, the most important argument regarding the geographical aspects of knowledge flows has been that, the more codified the knowledge involved, the less space-sensitive these flows should tend to be. When, however, the knowledge involved is diffused and tacit, the argument is that such interaction and exchange is dependent on spatial proximity between the actors involved. Only by being in the same local environment, and by meeting in person, can and will such subtler forms of information be exchanged (Bathelt et al., 2004). One of the main distinguishing features of spatial
27 clusters of similar and related economic activity is that they provide opportunities for the transmission of sticky, non-articulated, tacit forms of knowledge between firms located there. Bathelt et al. (2004) add that when this locally embedded knowledge is combined in novel ways with codified and accessible external knowledge, new value can be created. In other words, even though codified knowledge can become widely available, it is the combination between this codified knowledge and the pool of locally available tacit knowledge that creates real value for a cluster member firm. It is the tacit part of knowledge that seems harder to transfer and access to it seems to form the source of the biggest advantage of the knowledge-intensive cluster.
Following this, the question arises as to how such a combined form of knowledge is exactly created. March and Simon (1958) proposed in their days that innovation and knowledge creation often results from borrowing, rather than innovation. Following this line, both established and recent innovation studies have insisted that radical knowledge creation follows from an interactive process across several firms (Rosenberg, 1982; Freeman, 1991; Hagedoorn and Schakenraad, 1992; OECD, 1992; Gertler, 1995). However, the idea that knowledge creation flows from a division of labour is much older and can be found as far back as one of the foundation stones of Adam Smith’s theory of economic growth in his ‘Wealth of Nations’ (1776). This means that by building a relevant differentiation, a group of firms can develop knowledge far beyond the reach of any single member of that group (Bathelt et al., 2004). With that, it starts to become clearer where the cluster synergy effect can be derived from. As Maskell and Malmberg (1999) add, a shared knowledge basis enables cluster firms to continuously combine and re-combine similar and non-similar resources to produce new knowledge and innovations. They believe that this mechanism results in an economic specialization of the cluster and the creation of localized capabilities available to cluster firms.
Concluding, according to a significant number of modern authors the most important advantage of the geographical cluster is its access to relevant knowledge. It is the combination between globally accessible codified knowledge and access to locally embedded tacit knowledge that creates value for
28 a MNC within a knowledge-intensive cluster. Acquisition of this knowledge drives the shape, and increases the pace, of innovation. By continuously combining similar and non-similar resources, member firms create a synergy effect within the cluster; By drawing from each other’s knowledge base, firms within a cluster are able to more efficiently acquire knowledge and speed up new product developments than if they were just stand-alone firms. This section has argued that this is presumably the main advantage modern clusters have to offer. The next section will build on this and argue how firms within a cluster can realize the actual acquisition of this knowledge: through labour mobility.
2.4.2 Knowledge Acquisition through Labour Mobility
It is not enough to only establish how access to knowledge offers a main advantage within a cluster; it is also very important, in order to answer the research question, to determine how such a pool of tacit knowledge can be accessed and exploited by member firms. This section will look at how knowledge exchanges take place and argue how the job-hopping behaviour of local skilled workers can realize these flows.
Economic geographers have come to see knowledge exchange as critical to defining performance and growth in regional clusters (Tallman et al., 2004; Romer, 1990). Yet, given the tacit and complex nature of most valuable knowledge, its acquisition can be difficult (Kogut and Zander, 1992). As to how such knowledge can be exchanged, Almeida and Kogut (1999) show evidence that regional labour mobility may be a significant cause of knowledge localization. These authors suggest that high-tech clusters, such as Silicon Valley, experiencing higher than average levels of inter-firm mobility, tend to also experience a greater degree of knowledge localization. These findings imply a direct relationship between labour mobility and knowledge flows (Agrawal et al., 2006). As Argote and Ingram (2000) argue, when individuals move between organizations, they can apply their knowledge to new contexts, and with that, effectively transfer knowledge across firms. In an important study, Almeida and Kogut (1999) traced the mobility of more than 400 skilled workers and
29 proved how their patterns of movements influenced the inter- and intra-regional patterns of knowledge flows and diffusion.
Of course much of the knowledge held by these skilled workers is probably easy to codify and already widely available, and of course much of the tacit knowledge held (e.g. social skills as to how to communicate with your co-workers) is irrelevant to the company’s knowledge creation strategy. However, as Almeida and Kogut (1999) have also argued, even when these skilled workers who move company contain knowledge with regard to, say, publicly disclosed patents, the physical human carriage of this knowledge is still of the utmost importance. That is, a significant part of this knowledge will be held tacitly by these skilled engineers. Agrawal et al. (2006) give as an example that certain failed experiments may be important for understanding how to modify an invention for alternative applications, but knowledge of these experiments is usually not codified since there is little incentive to do so. Skilled engineers carrying this knowledge might be of high value for the knowledge stock of a company trying to modify the invention for its own application.
Knowledge held by skilled workers matters. Available literature provides numerous examples of cases where labour mobility forms an invaluable source of relevant knowledge for member firms. Many of these cases apply to Silicon Valley, which is understandable, in view of the fact that this high-tech cluster serves like a role model to all others. A good example is provided by Angel (1991), who performed a case study of the dense cluster of semiconductor firms in Silicon Valley. She observes that the dynamics as found in Silicon Valley are due to flexible employment relations and efficiencies caused by inter-firm worker mobility that facilitate the rapid diffusion of specialized knowledge, skills and experience among member firms. While moving from one company to another, skilled workers transfer valuable information with regard to market opportunities, production processes and successful or failed technological experiments. As Hamel et al. (1989) add, this rapid spread of valuable knowledge allows firms to quickly spot market opportunities and forms the key source of stimulating innovative new product development and exploitation, ahead of competitors. It
30 is from building on this basis of experience and information sharing that member firms can significantly reduce the required investments for their innovative product developments and gain a competitive advantage.
Concluding, there seems to be a direct relationship between labour mobility and knowledge flows (Agrawal et al., 2006). Skilled workers moving between organizations are able to apply their knowledge to new contexts, and with that, effectively transfer knowledge between firms. MNCs, located in a cluster, can thus draw from the knowledge level of the cluster by hiring local skilled workers. In this line, the local labour mobility consequently forms an important driver of knowledge acquisition. Next to that, the more local skilled workers (as a human vehicle of knowledge) change employer, the faster knowledge can be assumed to spread through the cluster. When there exists a high rate of local labour mobility, firms within the cluster will be able to gain easier access to relevant knowledge. This increased access to knowledge could in its turn stimulate the development of new innovative products and processes. New knowledge combinations and re-combinations could be made quicker when knowledge moves faster and, with that, could stimulate the pace of innovative developments. This boost to the innovative capabilities of a MNC could help the firm to stay ahead of competition and could result in a consequent competitive advantage.
Now that all the aspects of the research question have been theoretically explored, the next section will construct a model based on the findings from the literature. This model will serve as guidance for the empirical research part of this thesis. Next to that, it will offer the reader an overview of what has been found in the literature and give a preliminary answer to the research question at hand.
2.5 The Model
By connecting the literature discussed in this chapter, it becomes possible to present an overview in the model below. This model will be used to give a preliminary answer to the research question at hand and will later on guide the empirical part of this research.
31
Figure 4. A model based on the theory discussed in this chapter.
This model gives a first answer on the research question as to what perceived advantages drive MNCs to locate an R&D department within a knowledge-intensive cluster. Considering the fact that the empirical research that will follow consists of an explorative, small-scale qualitative study, it will not be possible to prove or reject the validity of this model statistically. However, this model will serve as a construct to guide the direction of the research. Consequently, the analyses that will follow will derive their key focus from this model and will talk about the validity of its concepts as perceived by the acquired qualitative data.
First, before examining the advantages that drive the decision to locate in a cluster, it is necessary to determine whether a researched cluster meets the characteristics as described by Porter (2000), Markusen (1996) and Sölvell et al. (2012). Since the definition of the boundaries of the cluster
32 concept can be highly ambiguous, it becomes easy to classify many arbitrary areas as clusters. By following Porter’s definition, Markusen’s typology and Sölvell’s description of cluster actors, it can be examined whether some of the most essential elements are presents. Following these conditions will prevent the research from selecting and classifying any arbitrary area as a cluster fit for this research. Advantages found in an arbitrary area could significantly differ from advantages found in an area classified as a knowledge-intensive cluster. Of interest are the latter’s advantages that arguably offer benefits to MNCs they cannot find anywhere else. Setting these conditions for the cluster characteristics will therefore contribute to answering the research question by helping to explain why MNCs perceive clusters to be relevant (and not just any area).
From the literature that was studied for this thesis, a number of key advantages of spatial clustering were distinguished and grouped under common nominators. It was theorized that a MNC, deciding to locate an R&D department within a knowledge-intensive cluster, must be driven by one or several of these advantages. These advantages are defined as proximity benefits, complementarities, access to networks, a beneficial start-up climate, access to institutions and access to knowledge. This chapter decided to follow the work of modern authors like Bathelt et al. (2004), Malmberg and Maskell (2002) and Maskell (2001), who argued that the main advantage of knowledge-intensive clusters lies in the access that these areas provide to relevant knowledge. It is the combination between globally accessible codified knowledge and access to locally embedded tacit knowledge that creates value for a MNC within a knowledge-intensive cluster. For this reason the above model proposes that next to a number of other perceived advantages, access to knowledge forms the main driver of an R&D location choice within a knowledge-intensive cluster.
Next, access to knowledge was laid out to work according to a specific mechanism. It was established that there is a direct relationship between labour mobility and knowledge flows (Agrawal et al., 2006). Skilled workers moving between organizations are able to apply their knowledge to new contexts, and with that, transfer knowledge between firms. MNCs, located in a cluster, can thus draw
33 from the knowledge level of the cluster by hiring local skilled workers. In this line, recruitment of local labour could form an important driver of knowledge acquisition. Re-combining and applying this knowledge could consequently stimulate the development of new products. By drawing from each other’s knowledge base, firms within a cluster could be able to more efficiently acquire knowledge and speed up new product developments than if they were just stand-alone firms. This potential boost to the innovative capabilities of a MNC could help the firm to stay ahead of competitors and could result in a potential competitive advantage.
With regard to the intensity of the stimulation of new product development, it was discussed that the more local skilled workers (as a human vehicle of knowledge) change employer, the faster knowledge can be assumed to spread through the cluster. When there exists a high rate of local labour mobility, firms within the cluster will be able to gain easier access to relevant knowledge. A high rate of labour mobility could therefore stimulate the rotation of knowledge flows. Next to that, when knowledge flows increase, it could also stimulate new knowledge acquisition and re-combinations. Finally, this extra stimulation could potentially result in an extra boost to the innovative developments of firms within the cluster.
In the next section, the methodology of the empirical part of this research will be discussed. This empirical part will be guided by the model from this chapter. The empirical part will help give a final answer to the research question as to what perceived advantages drive a MNC to locate an R&D department within a knowledge-intensive cluster. It will explore whether it is access to knowledge through labour mobility, as was argued by the theory in this chapter, that forms the main driver of such a location decision.
34
3. Methodology
Now that the previous chapter has formed a relevant framework to provide an preliminary answer to the research question, the next part of this thesis will focus on the empirical verification of this framework. The concluding model that was constructed in the previous chapter will serve to guide this next part of the research. This chapter will centre around its methodology and describe how a qualitative study was conducted. The first section will first discuss the research design that was chosen and explain how the research type and its participants were selected. The second section will explore the method of data collection and will elaborate upon the procedures that were applied in order to obtain data from both a desk research and twelve semi-structured interviews. Last, the third section will illustrate how the rich data derived from the interviews was processed and prepared for the analyses sections that will follow after this chapter.
3.1 Research Design
In order to answer the research question and explore the proposed model, this thesis will employ a qualitative research approach. The in-depth nature of qualitative data will allow for a thorough exploration of this research concept. This approach is better suited than the quantitative approach, for it does not seek to collect a representative sample, but rather aims to interview those professionals who are able to shine their light on the subject and provide valuable and exhaustive information. As the research can be characterized as explorative, data derived from open-ended questions will provide more details and understanding than close-ended questions could deliver. The interviews will be semi-structured, which will be the most convenient to allow participants to elaborate when necessary or to discover differing solutions to the problem at hand.
This research will focus on knowledge-intensive clusters within the Netherlands and can therefore be considered as a case study of the model applied to the Netherlands. Within the Netherlands two knowledge-intensive clusters are selected, each with a different specialization: the High Tech Campus in Eindhoven (specialized in High Tech) and Food Valley in Wageningen (specialized in Food and
