Clustering and the location of energy industries

Clustering and the location of energy

industries

By

Björn Oortman

University of Groningen

2007

Clustering and the location of energy industries

B.Oortman

International business Management

University:

University

of

Groningen

Faculty

of

International

Business

Management

Landleven 5

9700

AV

Groningen

The Netherlands

University supervisors:

Drs. M.A.G. van Offenbeek

Drs.

A.R.M.

Gigengack

Company:

Protiviti

Amsterdam

Delflandlaan

1

1062

EA

Amsterdam

The Netherlands

Company supervisor:

Drs., Ing. A.F. Schreij

Management summary

Research Background

In order to keep the economy of the Netherlands competitive. The Dutch government constructed the Dutch National Reform Program. The most important goals of the program were to increase innovation, productivity and price competitiveness. After an assessment in 2006 by the European Union it became clear that the program was not accomplishing in the areas of innovation and productivity increases. It was found that clusters have promising effects on both innovation and productivity. Clusters have the characteristic that they increase productivity, innovation and the capacity for productivity and innovation. These corresponded perfectly with the goals set in the DNRP. Therefore the objective of the research was established as: This thesis is aimed at providing further insights into the underlying factors of the clustering and location of industry to stimulate the accomplishment of the goals set in the DNRP. After careful elaboration, it seemed impossible to

deepen the insights in causes for industrial agglomeration for every industry. The energy industry was subsequently selected as the subject for this research. Resulting in the following research question:

What are the initial factors that enable energy industry clustering to develop?

Method

Three methods were used to get the results and information needed for this thesis. An historical analysis was preformed to establish whether there is a common industrial history. Interviews and surveys were performed to gather the remaining information needed.

Facts and Findings

The above described methods led to the following findings; the presence of a large market together with the presence of skilled labour are the most important factors for the emergence of an energy cluster. Although the presence of natural resources plays an important role as well. Surprising was the little importance that was placed on the availability of distribution channels and governmental incentives.

Conclusion

The variables mentioned in the matrix on the next page can be considered as the initial factors that have to be present for energy clustering to emerge. Positive fulfilment of other variables as defined in the research will increase the probability of an energy cluster to emerge but are not the initiators that induce energy cluster formation.

Matrix: initiating factors for clustering

(industrial)

History

Regional characteristics

Strategic

objective

Entry mode

Not influecable Long-term influencable Short- term influencable - there is no common industrial history that underlies energy cluster formation - (Natural) resources or - Presence of large market (customers and suppliers) - Geographic location (e.g.. sea access, location on world map) - skilled labour - Infrastruc-ture - political stability - industrial organisation - a facilitating governmental involvement - Knowledge development, Market development and Resource access all seem to be equally important strategic objectives.

-

• This research could well be used by governments to assess whether the initiation of an energy cluster in a region is feasible or not.

• Governments can also use this research to reform policies to further energy cluster development

For further research

• Research into how a government can facilitate the energy clustering without intervening • Research into market driven energy clusters how they emerge and what their needs are

Preface

In the light of my study International Business and Management Science at the University of Groningen, writing a thesis is the final hurdle to take in becoming an academic. During an eighth month period you are supposed to make use of gained academic knowledge in a business environment. On November 6th 2006, I started my final research project at Protiviti in Amsterdam. This thesis is written on the subject of clusters, and especially what the initiating factors are for energy clusters to emerge.

The first month was used to explore the possibilities of the research topic handed to me by the organisation. It was difficult to find out what their real goal was with the topic and after a month it was rejected. During this period, and after, I learned to take responsibility and to defend my choices, thereby not forgetting to listen to others.

Next to writing my thesis I participated in presentations and sometimes helped people with their assignments. All in all, it was a very good experience to work in a real business environment.

I want to thank my company supervisor Aart Schreij who has been a great help in being a sparring partner for discussing ideas and in the guidance of writing the thesis. Special thanks also goes to Maria Bautista without her Spanish translating skills I could not have managed to pull it all off. Further more I want to thank all people within Protiviti for their support and making it a very nice environment for writing my thesis.

At the University of Groningen, my gratitude goes out to my supervisors drs. M.A.G. van Offenbeek en drs. A.R.M. Gigengack for their advice and instructive support.

Finally I want to thank Jonathan Reynolds (EEEGR), Allard van Dijk (NOM), Per Moller Pedersen (energiparken), the interviewees and all the respondents for their help.

Groningen, The Netherlands August 2007

5

Table of contents

Introduction ... 7

Background and development of the assignment ... 7

Lay out ... 9

Chapter 1: Problem Statement ... 10

1.1 Research objective ... 10

1.1.1 The current state of the DNRP. ... 10

1.1.2 Influencing the economy; The Governmental Toolbox ... 11

1.1.3 Lessons from other countries ... 12

1.2 Main research question ... 15

1.2.1 What is a cluster? ... 15

1.2.2 The lifecycle of (specialised) clusters ... 16

1.2.3 Cluster formation theory ... 18

1.2.4 Industry selection ... 20

1.3 Conceptual model ... 21

1.4 Pre-conditions and assumptions ... 23

Chapter 2: Methodology ... 25

2.1 Research design ... 25

2.2 Case selection ... 26

2.3 Data-sources ... 28

2.4 Methods of measurement and observation ... 28

2.4.1 Questionnaire development ... 30

2.4.2 Sample ... 35

2.5 Methods of analysis and reporting ... 35

Chapter 3: Historical analysis ... 37

3.1 Aberdeen: an industrial overview ... 37

3.2 The Basque country ... 39

3.3 East Anglia ... 41

3.4 Houston ... 42

3.5 Northern Part of The Netherlands ... 43

3.6 Stavanger ... 45

3.7 Common aspects ... 46

3.8 Conclusion ... 47

Chapter 4 Interviews ... 48

4.1 Interview Energiparken ... Fout! Bladwijzer niet gedefinieerd.

4.2 interview Gas Terra ... Fout! Bladwijzer niet gedefinieerd.

4.3 Interview Calash ... Fout! Bladwijzer niet gedefinieerd.

4.4 Interview Royal Nedalco ... Fout! Bladwijzer niet gedefinieerd.

4.5 Interview Protiviti ... Fout! Bladwijzer niet gedefinieerd.

4.6 Common aspects ... 48

Chapter 5 The survey results ... 49

5.2 Reliability ... 51

5.3 Results ... 51

5.3.1 Region characteristics ... 51

5.3.2 Strategic objective ... 53

5.3.3 Entry mode ... 54

5.3.4 Inter case differences ... 56

5.3.5 Results of open questions and remarks ... 57

Chapter 6 Conclusions ... 58

Chapter 7 Discussions, limitations and Recommendations ... 62

7.1 Discussion ... 62

7.1.1 Cluster Theory ... 62

7.1.2 Assumptions ... 62

7.1.3 Conclusion ... 63

7.2 Limitations ... 63

7.2.1 Survey data ... 63

7.2.2 Interviews ... 64

7.2.3 Historic analysis ... 64

7.2.4 Sample Case selection ... 64

7.3 Recommendations ... 65

References ... 66

Annexes

Annex I Company profile

Annex II Article, case analysis

Annex III Interview energyvalley

Annex IV Questionnaires

Annex V English and Spanish mail

Introduction

This report is about clustering and the location of energy industries. The initial factors that have to be present for energy industry clustering to occur will be presented. In this section an overview is given of the development of the assignment. It is explained why the initial assignment was rejected and how the final assignment took shape. Further on the lay-out of the thesis will be unfolded.

Background and development of the assignment

Protiviti1 was interested in the possibility of establishing a Free trade Zone within the Netherlands. The reason for their interest was mainly attributable to articles published during that time concerning this subject. In these articles2 Free trade Zones were seen as promising tools to promote business activity and economic growth within the Netherlands and to help reaching the goals as stated in the Dutch National Reform Program (DNRP). The presumptions made in these articles were mainly based on experiences in China (e.g. Shengzhen) and Dubai.

To keep the level of prosperity at a high standard, the Dutch government has launched an extensive and ambitious programme of structural reforms. The main aim of this programme is to address the challenges posed by globalisation and the ageing population, and to strengthen the country’s economic structure. This programme is formalized in the Dutch National Reform Program (Ministerie economische zaken, 2005).

Protiviti was interested in the risks concerning the establishment of a business in a Free trade Zone. In the preliminary study the author discovered that a “Free trade Zone” can take a large variety of forms. Each of these forms involves different risks. So it had to be assessed which form would be most likely would be most likely for the Netherlands. However before starting research into this subject it is imperative to know whether it is even likely that any form of free trade zone would bring benefit to the Netherlands.

The author tried to find empirical evidence that supported the presumptions made in the respective articles, with a special emphasis on Free trade Zones (FTZ) established in developed countries. During the search the author reached the conclusion that there were many types of “free trade zones” ranging from, export processing zone’s (mainly China), Foreign trade area

(US)

The author concluded that the common features were that “Free trade Zones” were

1

Annex I

2

The examples and presence of these zones around the world are numerous. Although they are far more common in developing nations then they are in developed countries. The effect of these zones on business activity and economic growth seemed to be very limited. To take one example Ireland experienced its steep economic growth not until the nineties while the Shannon free zone had already been established in 1959 for example (OECD 2006;Rondinelli et al. 2000;Tahir, 1999).

For developing nations they seem to be a helpful tool to attract business and reduce unemployment. The FTZ’s in Chile3 have been remarkable successful and positive effects were also experienced in Mexico (de Armas and Sadni-Jallab, 2002) and Mauritius (YeungLamKo, 1998) for example. The type of employment and activity created through FTZ’s is mainly low-end production activities (Pacheco-Lopez and Thirlwall, 2003) that get attracted by the favourable legislative circumstances and abundance of cheap labour. Two conditions which are not present or wanted within the Dutch boarders. Moreover difficulties would probably arise with the European Union when the Netherlands would adjust tax rates and legislative measures regarding corporate law for one particular area.

This led to the conclusion that it is highly unlikely that a “Free trade zone” would further business activity within the Netherlands. Moreover, if it would enhance business activity it seems highly unlikely that it would bring high-end innovative industries in which the Netherlands can reach a long-term competitive advantage. This is underlined by the conclusions reached in the article in the annex4. In this article the relation between governmental policies (toolbox) economic growth, innovation and productivity is assessed by analysing four economies that have experienced steep economic growth over the last years (Singapore, UAE, Ireland South Korea). Finally, lots of potential problems could arise with the EU. This conclusion however does not mean that FTZ’s cannot stimulate trade. The numerous zones popping up around world, especially around ports5, have certainly facilitated trade (McCalla, 1990). It is merely concluded that it is not reasonable to assume that an FTZ would bring the sorts of industry and economic development the Dutch economy needs (DNRP 2005).

On a completely other stand, it’s highly questionable whether the Dutch labour market would be able to support a boom in company presence. At present the economy is already in danger of overheating as a result of demand exceeding production capacity6. The level of job vacancies is at a 6 year high, for

The Port of Rotterdam also has a free warehousing policy, Goods not destined for the Dutch marked can be stored here outside of customs.

6

now inflation and wages are in check but this may change if the gap between demand in respect to capacity widens7.

The above conclusion led to a change in focus of my thesis and subscribes the conclusions reached in the DNRP that innovation and productivity must be enhanced and not mere company presence. The focus was now; how to enhance productivity and innovation within the Netherlands. In the problem definition the conclusion was reached that this can be done by stimulating cluster activity.

Lay out

The problem definition is described in chapter 1. Herein I will also come to my final research question and sub-questions. Ongoing in chapter two with the Methodology designed to find the answers on the main question and the four sub-questions designed in the problem definition.

In chapter three a historical analysis is performed of the cases that were selected during the Methodology. In this chapter answers are found for the first two sub-questions. Chapter four presents summaries of the performed interviews. At the end of this chapter common aspects are presented that were distilled from the interviews. These will help to answer sub-questions two and three. In chapter five the results of the survey are given. These will contribute to answer sub-questions two, three and four

The conclusions will be presented in chapter six and the discussion, limitations and recommendations are outlined in chapter seven.

7

Chapter 1: Problem Statement

In this chapter the core of the research is given, i.e. the problem definition, consists of the research objective and main research question. Furthermore the conceptual model is presented. The conceptual model results from the problem definition. This chapter also contains the pre-conditions and assumptions to the thesis.

1.1 Research objective

In this paragraph a short analysis is given of the Netherlands in respect to its current situation, the DNRP and possibilities to influence the situation. At the end of this paragraph the research objective of this thesis will be presented.

1.1.1 The current state of the DNRP.

The DNRP was constructed on the basis of the Lisbon Treaty. The Lisbon Treaty was agreed upon during the 2000 Lisbon conference of EU members. In this Treaty the member states committed themselves to the goals of reinvigorating the European economy and boost the level of employment. The premium goal set in this treaty, is to make the European economy the most competitive in the world.

In their analysis of the Lisbon treaty in 2006 Gelauff and Lejour identified five of the most important Lisbon goals:

- Opening up of the service market; - Reduction of administrative burdens; - Improving human capital;

- 3% target of GDP on R&D expenditures; - 70% labour participation;

Gelauff and Lejour calculated that when these goals are accomplished overall GDP in the European Union could rise anywhere between 12 to 23 percent and the labour participation rate with 11 percent. These outcomes are in line with the most important objectives defined in the Lisbon Treaty, being:

1) Increase GDP of member countries

2) Increase competitiveness of member countries

As a refinement of the Lisbon goals the Dutch government also defined three further objectives that fit their specific situation (DNRP, 2005):

2) Increase productivity

3) Increase price competitiveness

The goal for meeting these challenges is to attract more innovating businesses and stimulate economic growth. A recent analysis however has indicated that with the current measures the challenges are not met in full, particularly in the field of innovation (1) and productivity growth (2). Labour productivity has slowed down over the last decade (measured in hours) coinciding with a deceleration in total factor productivity growth (TFP) 8 across all industries (European commission 2006).

This led to the conclusion that policies aimed at increasing labour participation and research and development, while necessary, were likely insufficient to increase both trend growth and productivity growth9. And although productivity is still rising the growth rate has slowed down severely. In order to increase productivity and innovation, additional measures have to be taken.

1.1.2 Influencing the economy; The Governmental Toolbox

Following the former paragraph it is useful to give a small insight into the policies a government can employ to influence its countries business environment.

Dicken (2003) identifies several kinds of industrial policies that belong to the toolbox of a government. The most obvious policy measures are those in the field of financial measures. Commonly one can distinguish two kinds of financial measures. First governments may provide capital grants or loans to firms to supply part or all of the investment required for a particular productive venture (subsidies). The other major financial employed by governments is that of tax reductions. These incentives can take various forms. From favourable corporate tax rates to enabling companies to write off their assets against an accelerated rate or offering them tax–free periods (tax-holidays).

Next to the financial measures a government also has the possibility to employ all kinds of labour and employment policies to stimulate presence and development of business within their boundaries. Governments may employ schemes to reduce the cost of labour for companies or set-up industry specific training facilities. A government can also exert power on the unions through labour legislation and even ban them altogether10.

8

TFP is a measure of the overall evolution of the efficiency of the economy. TFP growth measures the growth in the economy that cannot be accounted by the measured increases in capital and labour. In other words, the part of the overall economic growth that cannot be accounted for by increases in physical capital and labour must be accounted by the change in the overall efficiency of capital and labour.

9

http://www.imf.org/external/pubs/ft/scr/2006/cr06284.pdf

10

To the extent that the governmental toolbox influences the location decision of businesses, a region created with an optimum policy (from an economic business perspective)11 concerning these matters would have a high concentration of business activity.

The question to be answered then was the following: Is there evidence that economic growth in

terms of GDP coincides with enhanced innovation and increasing productivity, and to what extent can this be influenced by governmental policies?

1.1.3 Lessons from other countries

To answer the above stated question, four growth economies were analysed (Singapore, U.A.E., South Korea, Ireland) leading to the following conclusions:

Based on the TFP scores there is no evidence that economic growth coincides with enhanced innovation or productivity. Although, in Ireland high levels of productivity were achieved during the expansion period. This might be ascribed to Ireland already being a developed country having in place a good infrastructure, education system and financial markets. In the other countries the catch up process is not captured in TFP.

In general, economic development occurs in a three-step process. And in each step the role of the government is different. The first step for being a growth economy seems to be based on factors that lay outside the influence sphere of governments. These consist of great availability of labour (thereby making it cheap), geographic location and potentially good access to a large market. Natural resources did not seem to have played a major role in attracting business activity to the selected cases. In the UAE oil caused economic growth after its discovery, but initially did not substantially enhance business activity in the region. Afterwards however, the money earned with oil was invested to make the UAE more attractive for businesses to invest. Without the oil money this would not have been possible.

During the second step the role of the government becomes increasingly important in attracting foreign direct investment and redirecting domestic investments. In the selected cases lowering taxes and relaxing legislative environment played a large role in attracting foreign direct investment to the countries. Parallel to this, is the need for infrastructural progress to attract and support the increased activity. The refinement of this infrastructural progress increases with the number of companies

11

From a social economic perspective, labour perspective or consumer perspective an optimum decision on the variables would look completely different.

12

present and the processes being performed by the companies. Meaning an infrastructure supporting multiple industries and multiple functions (e.g. R&D, Services, and Manufacturing) within these industries needs to be more sophisticated then an infrastructure only supporting manufacturing.

Next to the tax regulation measures all four governments played a role in wage control. In the case of Singapore, Korea and the UAE a more intrusive position was chosen. In essence, strikes and unionization were prohibited in these countries. In Ireland the government played a more stimulating role by facilitating dialogue and social arrangements. During this stage it is also important to create financial markets and services that are able to absorb the increased flows of money, otherwise evaporation of productive potential is likely to occur.

In the third step the economy has to shift to a knowledge base with innovating entrepreneurs, institutes and companies. A government can stimulate this, but money-based incentives seem to have little effect in this phase (Karakaya, 1998). Creating a good educational system and enabling companies to create/import knowledge seem to be more important. Already during prior phases stimulation of the education system and upgrading knowledge and skills of nationals has to be started. This final step has to be made because the comparative advantage of low wages in low-end processes is likely to be diluted by the increase in national income and wages.

During the last two stages it is important to keep the economy sufficiently diversified to limit the effects of downturn in one industry on the economy as a whole. This was indicated by the case of Korea, and indicators for negative effects of single industry economies were also found for Ireland. During the whole process, indifferent of the stage the economy is in, stability and continuous policy adaptation to economical and business needs are very important. The three steps in the economic development could also be described as a recognition of comparatives advantages, utilizing these comparative advantages and turning them into competitive advantages.

We could safely say that the Dutch economy is in the third stage of the process. As mentioned before, the main concern for the economy is that productivity and innovation has to be enhanced. Although in this analysis no indication was found that enhanced innovation and productivity coincides with economic growth. The role of the government has been large in all four cases. During the first step government involvement is limited to signalling potential. In step two the role of the government changes, Initially infrastructure (legal, telecomm, roads, ports, airfields) has to make fit to support industrial and business activity. In congruence companies can be attracted to the country by tax breaks, other financial incentives or favourable labour legislation. Later on in the second stage, a properly functioning financial system has to be established. At stage three the role of the government differs again and is not so much steering, but facilitating. At this stage, it has to facilitate sophistication of the

second step. Basically offering an optimum business environment, not only t based on tax or labour legislation, but on a broad scale.

The measures belonging to the Governmental Toolbox to directly influence the economy seem to have little effect in the third phase of development (annex II). Ireland, however, has success by targeting specific industries and trying to attract them to their country by emphasising the advantages the country offers to the targeted industries. In this process Ireland targeted industries that would fit with the country and the advantages it offered. Ireland in essence then tried to stimulate aggregation of these industries in their country. This aggregation of companies (clustering) of selected industries may well be one of the reasons why TFP growth was large in Ireland in respect to other countries. Two of the effects of clustering are according to Porter (2000) increased productivity and innovation. Later on in the thesis this will be explained further.

Matrix 1.1: Stages of economical development and the role of the Government Stage 1: recognition of comparative

advantage

Stage 2: utilisation of comparative advantage by proper design of

Stage 3: competitive advantage

Comparative in the area of: - Natural resources - Geographic location - Labour force

-

Physical infrastructure: e.g. - roads

- airports Legal system: e.g.

- labour laws

- ownership regulations Financial system: e.g.

- Tax system - Financial markets

- stimulate clusters of industries

(agglomerations)

- sophisticate the systems designed in stage 2 - Stimulate knowledge

creation

Diversification of economy

Educational system and knowledge creation

Political stability

Based on the above it seemed sensible to presume that; when targeting the right industries and keeping the economy well enough diversified, cluster stimulation can contribute to the accomplishment of the goals as envisioned in the DNRP. This resulted in the following research objective

research objective:

- This thesis is aimed at providing further insights into the underlying factors of the clustering and location of industry to stimulate the accomplishment of the goals set in the DNRP.

1.2 Main research question

Now the research objective of the thesis has been unfolded. The theories from which the main research question and sub-question are drawn will be presented in the following paragraph. Subsequently a conceptual model was developed which will be presented in chapter 1.3.

1.2.1 What is a cluster?

Porter (1998) defines clusters as a geographic concentration of interconnected companies, specialized suppliers, service providers, and firms in related industries, and associated institutions (e.g., university standards, agencies, trade associations).

Dicken (2003) further divides clusters into two types of clusters; generalized clusters and specialized clusters. Both types are based on the notion of externalities, the positive spillovers created through locating facilities in close proximity of each other, thereby constituting the idea that the whole of the cluster is bigger then the sum of its parts.

- Generalized clusters simply reflect the fact that human activities tend to agglomerate and form urban areas. Hence, such benefits have traditionally been labelled urbanization economies. General clustering of activities creates the basis for sharing the costs of a whole rang of services. Larger aggregate demand in, say, a large city encourages the emergence and growth of a variety of infrastructural, economic, social and cultural facilities that cannot be provided where their customers are geographically dispersed. The larger, the greater the variety of available facilities and vice versa.

- Specialized clusters, on the other hand, reflect the tendency for firms in the same or closely related, industries to locate in the same places to form what are sometimes termed ”industrial districts” or “industrial spaces”. Such benefits have been called localisation economies. The basis of specialized a cluster arises from the geographical proximity of firms performing different but linked functions, in particular production networks.

In today’s economy of globalisation and the disappearance of boarders of all kinds (technical, tariffs, quotas, etc) it seems a paradox that geographically localised clusters are forming. Even more so that companies situated in such a cluster often retain competitive advantages over companies situated outside of the cluster. When one looks at the geo-economic map one finds tendencies of both concentration and dispersal, but still with a very strong propensity for economic activities to agglomerate into localised geographic clusters (Dicken, 2003).

In this thesis the author will focus on the specialised clusters. Within this thesis the author is especially interested in the emergence and location factors of specialized clusters.

1.2.2 The lifecycle of (specialised) clusters

Clusters follow a path of emergence, growth, transformation or decline. Although every cluster develops differently, in the following the path for cluster development will be outlined. (Observatory of European SME’s, 2002)

The emergence of a cluster can often be traced to history. These can exist of specific knowledge of raw materials, the specific needs of a certain group of customers or firms for example. Coincidence may also so initiate the emergence of a cluster. However according to Pinch and Henry (1999) the growth is often offset by some explicit location factors in particular long term development of specific knowledge that can be turned into new productive use. The first stage in cluster development often involves new firm spin-offs leading to a geographical concentration of firms in nearly the same production stage. The agglomeration is followed by local competition that is an essential driver of innovation and entrepreneurship (Porter, 1998). Furthermore when an agglomeration of companies becomes established, it creates a virtuous circle. The first external economies often include the creation of a set of specialised supplies and service firms, frequently originating from vertical disintegration of firms, and the creation of a specialized labour market (Storper 1995). The development can lower the cost of shared inputs as savings in production costs are passed from specialised suppliers to client firms. The client firms will then derive a benefit not available to similar firms in less localised settings (Harrison, Kelley et al, 1996).

The next step is the formation of new organisations that supply the growing number of organisation with additional services; knowledge organisations, specialised education establishments and business associations. These organisations promote further collaboration, learning processes and technological spillovers, as well as the creation of localised forms of knowledge by employees in the region’s industries. These organisations hold special competences and can supply companies with professional competence that small firms seldom acquire themselves, but which is often necessary in accomplishing larger innovations.

The formation of these new “development organisations” the development of additional externalities and the emergence of new local organisations, increases the visibility, prestige and attractiveness of a cluster. This may result in even more firms and skilled employees moving to the cluster. Thereby again enhancing the potential breeding ground for new companies.

Now the creation of non-market, relational assets that foster an untraded circulation of information and knowledge13 and help with coordinating activity will become apparent. Mature regional clusters may contain ensembles of specific, differentiated, and localised relations between persons and organisations that are coordinated by routines or conventions that often work in a context of proximity (Storper, 1995). Communication that contains flows of non-codified knowledge, and which is complex and uncertain, frequently involves dense human relations, which in turn are stimulated by proximity between individuals, firms, and organisations.

Finally, although a cluster can renew its success for decades or become part of a new cluster, many regional clusters sooner or later enter a period of decline. The decline is often seen to reflect a situation of technological, institutional, social and/or cultural “lock in business” behaviour. Regional industrial development may become locked in by the very socio-economic conditions that once made the region a core region in a specific industry. The initial strength of a cluster in the past, e.g. a well educated or experienced workforce, a highly developed and specialised infrastructure of firms, knowledge organisations can turn into an inflexible obstacle to development (Grabher, 1993). Clusters can fall into the trap of “rigid specialisation”. Cluster development sometimes tends to reinforce old behaviours and suppress new ideas, which in particular is a danger for the continued survival of a cluster when technological and global economic conditions change (Porter, 1998).

Although some indications can already be found in the above for why companies would want to locate in a cluster. Dicken (2003) states that operations tend to be established for one of the following reasons; to exploit a localized resource, to serve the host market itself, or to use the location as a platform for exporting either finished products or components.

Companies do not set-up facilities in places were they do not expect to achieve an advantage or further their profits. According to Porter (1998) being in a cluster improves the companies’ position in three broad ways: It increases the productivity, it supports innovation and it increases the capacity for productivity and innovation.

The reasons for new businesses to set-up facilities in a cluster are various. One of the biggest inducements may be the availability of better information about opportunities. Individuals working somewhere in or near the cluster more easily perceive new gaps in products, services, or suppliers to fill. Having had these insights, these individuals more readily leave established firms to start new ones aimed at filling the perceived gaps (Porter, 1998).

13

When the cluster becomes a success more and more multinationals will enter the cluster (Porter, 2000). Large companies often face numerous of constraints and impediments to innovating. Spin-off companies then often pick up the slack, sometimes with the blessing of the former companies. It is not uncommon to see larger companies in a cluster develop relationships with innovative smaller ones.

So according to Porter clustering helps to realize increased productivity, innovation and also increases the capacity for productivity and innovation. These effects show almost a perfect match with the goals set in the DRNP. But why do these clusters tend to emerge in specific locations and what are the specific characteristics of these locations? In the next paragraph these questions will be tried to be answered.

1.2.3 Cluster formation theory

In unmodified form the traditional explanations of patterns of trade and of the location of production and investment are inadequate to explain the complexities of the modern global economy (Dicken, 2003). Especially with respect to the location and clustering of economic activities there remains a wide area of unexplored terrain. Most of the theories existing today are an extension or upgrade of the Ricardo theory14. These theories tend to explain location and clustering by comparative advantage15 (Ricardo theory) and factor endowments theory16. Although these should not be entirely rejected they do not take into account the complex interaction of companies of different sizes and national governments. National productivity is largely set by the sophistications by which companies compete. By sophisticated forms of competition is meant, competition on a broader scheme then mere cost and differentiation strategies. It means competition on the basis of total quality of processes, in the supply chain and organisational context. When companies do not become more productive, it is difficult for an economy to become more productive17. The sophistication of companies’ approaches to competing determines the prices that their products and services can command and the efficiency with which they produce. Cluster thinking thus assumes an important role in economic policy as well as company strategy.

The context for strategy and rivalry can be divided into two primary dimensions. One is the climate for investment in its various forms. A rising investment intensity of competition is necessary for support

14

www.lse.ac.uk/collections/geographyAndEnvironment/whosWho/profiles/storper/pdf/GlobLocalization.pdf

15

Comparative advantage theory states that any country or geographic area should specialize in producing and exporting those products in which it has a comparative, or relative cost, advantage compared with other countries and should import those goods in which it has a comparative disadvantage.

16

Also known as Heckscher- Ohlin model, the theory argues that the pattern of international trade is determined by differences in factor endowments. It predicts that countries will export those goods that make intensive use of locally abundant factors and will import goods that make intensive use of factors that are locally scarce

17

The economy could become more productive if the people become more productive because they are more happy for example.

of more sophisticated forms of competition and higher levels of productivity. The other dimension of the context for competition is local policies affecting rivalry itself. Openness of trade and foreign investment, government ownership, licensing rules, antitrust policy, and the influence of corruption, among other things, has a vital role in setting the intensity of local rivalry (Porter, 2000).

However in a global economy macroeconomic policies are necessary but not sufficient to stimulate a country’s economy18 or its clusters. Governments also need to focus on the microeconomic level. They need to focus on removing obstacles to the growth and upgrading of existing and emerging clusters.

Porter (2000) draws a concept of a cluster that sees clustering as an independent process. Regardless of location it has its own laws of development that can be copied and success will arise. Sceptics of this strategy acknowledge the benefits associated with developed clusters; however they question whether this is a realistic industrialisation strategy for many regions. The development of a cluster requires specific conditions that may be attainable only at significant costs. They state that for areas deficient in these necessary conditions, the promotion of industry clusters will be unproductive19.

In this thesis the latter view is adopted and it also stresses the importance of building on comparative advantages. Clusters are seen as excellent phenomenon’s to turn a comparative advantage into a competitive advantage, thereby granting the area a significant advantage over an extented period of time.

As stipulated in the above, clustering holds a lot of potential advantages and plenty of research has been done into the results of clustering. However, there has been done very little research in to the

factors that causes it to emerge to be productive in a specific location. Therefore I aimed to deepen the understanding of the connection between location and industrial agglomeration.

The former resulted in the main research question and sub questions:

- What are the initial factors that enable industry clustering to develop?

Sub-questions:

1. What is the industrial history of the cluster areas?

2. What are the regional characteristics that companies prefer in a cluster? 3. What do companies want to achieve by locating in an cluster?

4. What mode of entry do energy companies choose to become part of an cluster?

18

Annex II

19

1.2.4 Industry selection

After careful elaboration it seemed impossible to deepen the insight in causes for industrial agglomeration for every single industry. Between industries an enormous ambivalent picture is expected with such an overload of factors and processes that any conclusive answer would be impossible to present. Additionally, not every industry is interesting for the Netherlands. labour intensive industries such as textiles for example are not likely to be valuable to the Dutch economy20. Therefore, I choose to select one single industry. I choose to select the energy industry. For this I have two reasons.

First energy is becoming increasingly important and focuses are shifting gradually from non-sustainable to non-sustainable energies. Gas and oil resources are mainly located in areas with low political stability and more people become depended on them. Potential results of these developments are signalled by Russia actively seeking contact with Iran. These countries have the largest gas resources in the world and are also dominant players in the oil market. Iran has even suggested to form a cartel organisation like OPEC directed at the gas supply 21. Poetin has, for the time being, rejected this proposition but has certainly not been reluctant to use Russia’s gas dominance as a political pressure tool (e.g. Belarus and Ukraine)22. When Western countries become more reliant and dependent on these resources it undermines their bargaining and political position. As an effect, countries try to become increasingly self-sufficient in the area of their energy needs. Resulting in nuclear power being back on the agenda and more effort put into research on alternative energy resources. Another reason for the later developments to arise, are treaties and a common coming of understanding to reduce the polluting gas’s strain on the environment. This means that reductive measures have to be taken. Developing and innovating energy must bring CO2 levels in line with the agreements such a Kyoto23. As a result of these developments the energy market not only becomes interesting for reducing harmful gasses or political uncertainties. It also creates possibilities for the development of innovating businesses.

Second, In general it requires large sums of money to be invested in knowledge and facilities for energy companies to achieve presence in an area. Therefore an energy cluster may be quite sustainable and causing competitive advantages to hold and sort effect over an extented period of time.

An understanding of the factors that are particularly important to energy companies for deciding in which location to set-up facilities, is in this respect very important. This led to a modification of the main research question and sub-questions:

main research question and sub questions:

- What are the specific factors that enable energy industry clustering to develop?

Sub-questions:

1. What is the industrial history of the energy cluster areas? 2. What are the regional characteristics of an energy cluster?

3. What do energy companies want to achieve by locating in an energy cluster?

4. What mode of entry do energy companies choose to become part of an energy cluster?

1.3 Conceptual model

The conceptual model was designed to picture the influence of price competitiveness, productivity and capacity to innovate on a regional level. Influencing these factors for a region within a country will have effect on the level of price competitiveness, productivity and capacity to innovate on the national level.

Figure 1.1: Conceptual model, clustering.

Industry

Objective

Entry mode

history

clusters

Regional characteristics

Short-term

influencable

Long-term

influencable

Not influecable

Referring to the conceptual framework of figure 1.1 the aspects are; History, Where (regional characteristics), Why (objective) and How (entry mode) a company chooses to locate in a particular area (Lasserre, 2003).

The industrial history of the researched region is often of importance to the formation of a cluster as established in the section about cluster development24.

Where does an energy company want to locate its business, what are the regional characteristics? These were split up in three categories based on the extent to which they can be influenced by governments and other institutions. This was done to be able to assess how region bound the energy industry is. When many of their location decisions are based on factors that cannot be influenced, it is impossible to get a flourishing cluster in an area where these characteristics are not present.

The question Why they want to set-up facilities in a specific area is directed at the objective of the expansion or creation of facilities.

The last aspect is How an energy company want to locate its business, what mode of entry does it prefer. Do energy companies rather make a Greenfield25 investment or do they prefer to takeover (smaller) companies already present in the area.

When the questions concerning these aspects are answered, interlocking these answers brings forth a holistic picture. For example the industry determines learning as an important aspect and this goes in congruence with taxes and a preferred mode of entry by joint venture. The conclusion can be drawn that suitable joint venture partners have to be present in the area that are knowledge leaders and that formation can be stimulated by tax advantages directed at joint ventures and knowledge creation.

The conceptual model is predominantly based on the location factors identified by Lasserre. I did not however completely take over Lassere’s theory. In this research, strategic components are added in the form of the objective and preferred entry mode of the firm. The main objective gives useful insides regarding the incentive of the company for wanting to expand or being set-up. Furthermore, connections can be made between regional characteristics and company objective. Adding entry mode (How) to the conceptual model and research will bring forth useful information regarding commitment for example. A company making a Greenfield investment ties itself more to a region then a company

24

Chapter 1, paragraph 1.2.2

25

A Greenfield Investment is the investment in a manufacturing plant, office, or other physical company-related structure or group of structures in an area where no previous facilities exist.

making joint venture arrangements; in general a bigger investment in the region is needed. These advantages exist next to the advantages of combining the factors as described previously.

Cluster theory tends to explain that clustering happens, what the advantages and disadvantages are (Barkeley and Henry, 2001). But does not explain why it happens in a specific location. My view is that every individual company makes an individual decision to set-up facilities in a specific area. However, other companies being present in the region selected can definitely influence or shape the decision. Notwithstanding that companies are not forced and do not make agreements with other companies to move in all at once. Additionally, the early movers are probably not aware of a cluster forming and therefore not able to assess the potential cluster benefits. Hence, they will assess the location in a cluster in the same way as they will assess a location outside of a cluster. Thus, combining the most important factors for companies in choosing a specific location will create a picture of the underlying factors needed for cluster emergence.

Therefore I choose to focus on location theory. I want to know what the initial factors were for setting up business in the selected area. By selecting companies from within the cluster, the cluster variable will be accounted for.

I choose to use location theory because I want to distil the answer to the main question from a company’s individual location decisions.

1.4 Pre-conditions and assumptions

The pre-conditions set the boundaries and reflect the context wherein the research is being conducted. There are two different kinds of conditions. Pre-conditions that focus on the product and pre-conditions that focus on the process (De Leeuw, 2001). Next to the pre-pre-conditions, six assumptions are given for the research.

Product pre-conditions

1. The product has to have relevance to the industries Protiviti is incorporated in 2. The product has to be of such a standard to pass university requirements

Process pre-conditions

1. The costs of the research will stay within the boundaries of reason 2. The research will be finished before the 31st of august

3. The research will be directed at clustering and not elaborate on other mean to enhance innovation and productivity

4. The complete research will only be accessible for Protiviti the University of Groningen and the people involved in the research.

Assumptions

1. The Netherlands have committed themselves to the achievement of the Lisbon Treaty 2. Clusters have positive effect on the accomplishment of the Treaty

3. Netherlands is not achieving the set goal at the moment

4. The challenges set in the DNRP are the main challenges faced by the Netherlands

5. A case study will deliver useful information for cluster stimulation/formation within the Netherlands

6. Chaos theory does not explain the clustering of energy industries in a specific location26

26

When clustering of energy companies explained by chaos-theory this research would be redundant as it would be impossible to find the answer to the main question as it is far to complex to understand.

Chapter 2: Methodology

In this chapter the study’s research design is explained. The research design was primarily based on the problem definition as stated in chapter 2.

2.1 Research design

The usefulness of research is guaranteed when coherent decisions are taken on the five aspects of research given in figure 2.1. (De Leeuw, 2000) For this research one aspect is added and integrated in the model of De Leeuw.

• Problem definition (chapter 1)

• The concepts used during research (chapter 1) • Case selection

• The data-sources

• The methods of measurement and observation • The method of analysis and reporting

Model 2.1: Research design.

The core of this figure represents the problem definition and the concepts that were used to deal with the problem at hand. Additionally, case selection, data –sources, methods of measurement and observation, methods of analysis and reporting all influence the core of the figure.

Hereafter all these elements will be described. Thereby first addressing the case selection, then data-sources ongoing with the methods of measurement and observation, methods of analysis and reporting and in the former chapter the problem definition and conceptual model (i.e. conceptual frame work) were depicted.

2.2 Case selection

To develop an answer to the main question, a case study was conducted. To this end I selected six energy industry clusters. The selection of these clusters will be explained in this chapter. There were two selection criteria identified. These were based on the situation in the Energy valley area, Netherlands.

- Three kinds of businesses must be present in the selected cluster 1. Energy transition organisation

2. Conventional energy organisations

3. Knowledge and innovation energy organisations

- The selected cases must bear relevance to the geographic and demographic conditions of the North of the Netherlands

4. Gas reserves 5. Petroleum reserves 6. Developed economy 7. European

Using the internet and contacts from Energy Valley27 clusters were identified. This means that clusters were identified on the basis of being named as an energy cluster by either the contacts in Energy Valley or by the website of the governing organisation. The selected clusters all have governing organisations in which the member companies have organized themselves, to be able to act as a group for attracting new businesses or to exert political pressure for example. Pre-selecting clusters with governing organisations is also helpful for data availability and access. Applying these criteria for pre-selection led to the following clusters: Basque country (Spain), Aberdeen (Scotland), Oldenburg

27

(Germany), Linz, (Austria), Houston (US), East Anglia (England), Stavanger (Norway), and Fiona (Denmark).

The first cluster selected is the Energy Valley Cluster. The other clusters were selected by placing them in the following matrix.

Matrix 2.1: Selection matrix28

selection- Transition Convent. K&I Gas Petroleum Developed European

criteria 1 2 3 4 5 6 7 Clusters Basque Country, Spain + + + -# - + + Aberdeen, Scotland + + + + + + + Oldenburg, Germany +/- +/- + +/- - + + Linz, Austria +/- - + - - + + Houston, USA + + + + + - + East Anglia, England + + + + + + + Stavanger, Norway + + +/- + + + + Fiona, Denmark +/- - + +/- - + +

On the basis of the scores applied in the selection matrix the following clusters were selected:

Basque country (Spain), Aberdeen (Scotland), Houston, (U.S.), East Anglia (England), Stavanger (Norway). Within these cluster all relevant industries were present and no other impairments such as language barriers seem to be present. Although Spain has no gas and petroleum resources, a substantial part of the Basque cluster derives revenue from this sector29. Therefore the cluster is viewed as particular interesting. In the pre-selection Fiona and Oldenburg seemed to be very suitable candidates. But the focus of the cluster in the respective areas is much more on the development of sustainable energies and especially wind-energy. Although they have both gas and petroleum reserves, these are very small. Linz in Austria is a cluster completely dedicated to renewable and sustainable energy industries.

28

https://www.cia.gov/cia/publications/factbook/, http://www.clusterenergia.com/, www.energyvalley.nl, www.energiparken.no, www.eeegr.com, http://www.aberdeenrenewables.com/,

http://www.oldenburg.de/stadtol/index.php?id=2181,

29

From the selected case studies, companies to be interviewed were selected on the conditions described in Methods of analysis and reporting. The companies are selected randomly. Information about company presence in the selected clusters was reviewed by consulting websites and cluster organisations and the companies were picked. The main issue is to interview the people, who know about and had a hand in, setting up facilities in the selected areas.

2.3 Data-sources

The main data-sources to be used in this thesis will be employees of companies that have facilities in the selected cluster cases. Websites will also be used to find members, companies within the cluster and to find background information.

2.4 Methods of measurement and observation

Due to the fact that this research has examined a relatively unknown field and tried to establish the factors behind clustering in a specific location, it was concluded that this thesis has a more explorative character (De Leeuw, 2000). Another characteristic of explorative research is that the research questions have an open character.

The main method of measurement for this kind of research is interviews. Next to interviews a web-based survey among companies within the selected benchmark and a historic analysis was conducted. The interviews were conducted by phone and had a semi-structured character.

Semi structured interviews are partly standardised. Semi structured interviews are best used in qualitative research in order to reveal and understand the what, how and why of a certain phenomenon. Essentially when it’s necessary to understand the reasons for the decisions that your research participants have taken it will be necessary to conduct a qualitative interview. As it was wanted for this thesis to establish a relationship between location factors and the clustering of companies this is a relevant form of conducting research. Jankowicz (2000) states that a semi-structured interview will be most appropriate for the following two situations:

- When the questions to be answered are either complex or open-ended - When the order and logic of questioning may need to be varied

Both situations apply to this research. The questions asked in the interview are to a large extent open-ended. And because the interview is semi-structured a participant may answer a question before it is

asked. Apart from this every company is different and other forms of questioning may be necessary, however the same main question has to be answered in the end.

For this research it was imperative to question managers with decisional power. Managers are more likely to agree to be interviewed, rather than to complete a questionnaire, especially where the interview topic is seen to be interesting30. On the downside many of these interviews had to be conducted by phone, because of a lack of funds. It would have added too much costs to the project to physically visit the different clusters. The biggest disadvantage of interviews by phone is the limited personal contact. Limited personal contact makes it harder to develop a relationship of trust. Developing a relationship of trust is especially important when asking sensitive questions. In this research the questions asked were not assessed as being particularly sensitive as answers were not to be connected to the specific company but were to be used to develop a generic picture. Next to this the companies were contacted by mail, explaining the purpose of the interview and the research.

Potential research participants receiving a questionnaire by email may be reluctant to complete this for a number of reasons. They may feel that it is not appropriate to provide sensitive and confidential information to someone they have never met. They may also not completely trust the way in which the information provided is to be used. The use of personal interviews, where appropriate, may therefore achieve a higher response rate. (Healey, 1991). This underpins the method chosen for the first part of this thesis but undermines the surveypart.

All selected clusters had governing organisations in which the energy companies in the selected cluster have organized themselves. To achieve a high response rate and to get access to a large potential pool of e-mail addresses these organisations were contacted and asked to cooperate with the research in return for access to the results. It is estimated that results of the research are useful to them because of their role in enhancing the quality and “density” of a cluster. Therefore it was expected that they will provide the support needed. These contacts together with contacts within Protiviti were used to review and command on the questionnaire.

In Matrix 2.2 the outline of the methods and their purpose is given. In the matrix the method and the sub-question to be answered by that method are connected by means of a *.

Matrix 2.2: Methods and purpose

Q131 Q2 Q3 Q4

Historic analysis

* *

Interview

* *

Survey

* * *

2.4.1 Questionnaire development

In the following sections the structure and content of the questionnaire is explained. A proper design is essential for the quality of the survey and thereby the quality of the research being preformed.

Structure

In order to develop a questionnaire the goal and variables of the questionnaire had to be established first. The goal of the questionnaire is to help find the answer to the main Research question: What are

the most important factors for an energy company in choosing a location to set-up facilities?

In the conceptual model it was established that three factors play an important role in the site selection process, being:

- The regional characteristics that energy companies prefer - The objective of the regional presence for energy companies

- The sort of investment energy companies makes for achieving presence in a region

These three factors were based on Lasserre’s (2003) work in the area of global strategic management.

The importance of a factor was established by measuring the variables that comprise the factor. The variables to be measured can be established by looking at prior research and theories. The downside of this method is that it may not be completely industry relevant and up to date. However it does provide a scientific base to the research. The same goes for the questions to be asked to quantify the variables. Furthermore it is essential that the relationship between variables and questions are clear. By using prior research this condition is more likely to be fulfilled. In the end though it is impossible to envisage every relevant relationship or interdependency. These eventually will come forth out the questionnaire data.

To be able to locate the interdependencies and compare answers a questionnaire was used with closed questions and pre-structured answers. Thereby easing the quantitative assessment of interdependencies and it also “helps” the respondent. By limiting the possibilities it helps to a certain extent the

31

respondent to assess what is expected and to reply on the subjects the questionnaire is designed for. On the other hand one has to be careful that all logical answer possibilities are provided for. It was also chosen to provide for an option “other” in combination with a remark box. This offers the possibility to the respondent to elaborate whether to his opinion something is missing or lacking. Thereby the negative effect of pre-structured answers not reflecting the opinion of the respondent was limited.

Next to the questions and type of questionnaire to be used it is also essential within the designing process to make decisions regarding the answer categories and scale. In this questionnaire the answers can be filled out on a four point-scale, ranging from unimportant to extremely important. An even number of options has been selected for the scale to force the respondents to reveal the score most applicable to them. In this questionnaire I aim to explain why companies are in a particular location. For this the company must have had good reasons. When they are neutral it’s considered not important to their location decision. This “filtering” is enhanced by using a skewed scale. As in this research is looked for factors of importance it is of irrelevance how neutral or not important questions are. Therefore focussing on the degree of importance of relevant questions.

The level of measurement will be on an ordinal scale. This comes forth out of the foregoing four-point scale and answers ranging from unimportant to extremely important. These answered are valued as 0 for not important and 3 for extremely important. An ordinal scale implies that there is a distinct order in the answers but the distances between the answers are not absolute (However when using multiple item scales, under certain conditions it is possible to make correlation analysis of an interval type). Calculation with the variables is as a consequence only possible under certain conditions. It is possible to establish modus, median and independent values. These are also the values that can be used as input for the analysis to be performed.

Content

The regional characteristics can be measured using location variables. These are variables established by prior research into location decisions of companies. To a large extent I expect that general variables for the location of companies as established by Karkaya and Canel (1997) also play a role for energy companies. On the basis of their research, that of Lasserre (2003) and the Protiviti risk model32 preferences regarding regional characteristics of energy and energy related companies were established. On this basis I created a conceptual model for measuring the location. The conceptual model (2.2) brings forth 6 variables to be measured; Costs, Infrastructure, Resources, Business environment, Risks and Quality of living. These variables are defined as follows:

32

Costs: Under costs, factors are placed that cause a direct outflow of money from the organisation.

Infrastructure: Under infrastructure, factors are placed that can facilitate the physical and technological transactions of products, data and personnel.

Resources: Under resources, factors are placed that form natural, human or economic input into the production of goods and services.

Business Environment: Under business environment, factors are placed that form the external conditions in a region that could significantly affect the success or failure of a business enterprise.

Risks: Under risks, factors are placed that can impose direct threats to the processes or the existence of the organisation.

Quality of living: Under Quality of living, factors are placed that can be of influence on the liveability of a region for the employed by the company.

Model 2.2: regional characteristics

The importance of each variable is expected to come forth out of questions regarding the categories and subjects as described in the conceptual model33. Further analysing the answers lead to conclusions on the variables and in the end to the factor; regional characteristics.

33