Clusters of High Technology SMEs: The Dutch Case

11
This article was downloaded by: [University of Bath] On: 19 November 2014, At: 04:36 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Regional Studies Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/cres20 Clusters of High Technology SMEs: The Dutch Case Egbert Wever a & Erik Stam a a Urban Research Centre Utrecht (URU), Faculty of Geographical Sciences , University of Utrecht , PO Box 80115, Utrecht, 3508 TC, The Netherlands Published online: 18 Aug 2010. To cite this article: Egbert Wever & Erik Stam (1999) Clusters of High Technology SMEs: The Dutch Case, Regional Studies, 33:4, 391-400, DOI: 10.1080/713693556 To link to this article: http://dx.doi.org/10.1080/713693556 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Transcript of Clusters of High Technology SMEs: The Dutch Case

Page 1: Clusters of High Technology SMEs: The Dutch Case

This article was downloaded by: [University of Bath]On: 19 November 2014, At: 04:36Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office:Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Regional StudiesPublication details, including instructions for authors and subscriptioninformation:http://www.tandfonline.com/loi/cres20

Clusters of High Technology SMEs: The DutchCaseEgbert Wever a & Erik Stam aa Urban Research Centre Utrecht (URU), Faculty of Geographical Sciences ,University of Utrecht , PO Box 80115, Utrecht, 3508 TC, The NetherlandsPublished online: 18 Aug 2010.

To cite this article: Egbert Wever & Erik Stam (1999) Clusters of High Technology SMEs: The Dutch Case,Regional Studies, 33:4, 391-400, DOI: 10.1080/713693556

To link to this article: http://dx.doi.org/10.1080/713693556

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)contained in the publications on our platform. However, Taylor & Francis, our agents, and ourlicensors make no representations or warranties whatsoever as to the accuracy, completeness,or suitability for any purpose of the Content. Any opinions and views expressed in thispublication are the opinions and views of the authors, and are not the views of or endorsedby Taylor & Francis. The accuracy of the Content should not be relied upon and should beindependently verified with primary sources of information. Taylor and Francis shall not be liablefor any losses, actions, claims, proceedings, demands, costs, expenses, damages, and otherliabilities whatsoever or howsoever caused arising directly or indirectly in connection with, inrelation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantialor systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, ordistribution in any form to anyone is expressly forbidden. Terms & Conditions of access and usecan be found at http://www.tandfonline.com/page/terms-and-conditions

Page 2: Clusters of High Technology SMEs: The Dutch Case

Regional Studies, Vol. 33.4, pp. 391± 400

Clusters of High Technology SMEs:The Dutch Case

EGBERT WEVER and ERIK STAMUrban Research Centre Utrecht (URU), Faculty of Geographical Sciences, University of Utrecht, PO Box 80115, 3508

TC Utrecht, The Netherlands

(Received July 1998; in revised form October 1998)

WEVER E. and STAM E. (1999) Clusters of high technology SMEs: the Dutch case, Reg. Studies 33, 391± 400. This paper

examines the relevance of high technology clusters in the speci ® c Dutch spatial context: a relatively small and homogeneous

country. In this context the spatial pattern of high technology ® rms is congruent with, or closely follows, the general spatialpattern of all ® rms. Regional clusters, characterized by innovation linkages with other ® rms and knowledge centres, hardly

exist. Most high technology small and medium sized enterprises (SMEs) have nationwide linkages.

The Randstad High technology SMEs High technology services Innovative activities Regional clustersCo-operation

WEVER E. et STAM E. (1999) Les regroupements de PME WEVER E. und STAM E. (1999) Kluster hoch technisierterde pointe: e tude de cas de la Hollande, Reg. Studies 33, kleiner und mittlerer Unternehmen: der Fall der Nieder-

391± 400. Cet article cherche aÁ examiner l’importance des lande, Reg. Studies 33, 391± 400. Dieser Aufsatz untersucht

regroupements de PME de pointe quant aÁ la ge ographie de die Relevanz hoch technisierter Kluster im spezi® sch nieder-la Hollande, un pays relativement petit et homogeÁ ne. Dans laÈ ndischen raÈ umlichen Zusammenhang, d.h. einem relativ

ce contexte-laÁ , la distribution ge ographique des entreprises kleinen und homogenen Lande. In diesem Zusammenhang

de pointe est en harmonie avec ou correspond e troitement entspricht das raÈ umliche Muster der Hochtechnologie® rmenaÁ la distribution ge ne rale des entreprises. Rares sont les dem allgemeinen raÈ umlichen Muster aller Firmen, oder folgt

regroupements re gionaux, caracte rise s par des relations avec ihm weitgehend. Es gibt kaum regionale Kluster, die durch

d’autres entreprises et d’autres centres d’ e tudes. La plupart InnovationsverknuÈ pfungen mit anderen Firmen und Wis-des PME de pointe jouissent des liens sur l’ensemble du senszentren charakerisiert werden. Die meisten hoch techni-

territoire. sierten kleinen und mittleren Unternehmen unterhalten

landesweit Verbindungen.

Randstad PME de pointe Services de pointe

Activite s innovatrices Regroupements re gionaux Randstad

Coope ration Hoch technisierte kleine und mittlere UnternehmenHoch technisierte Dienstleistungen

Innovative TaÈ tigkeiten Regionale Kluster

Zusammenarbeit

INT ROD UCT I ON for realizing economic growth and creating many new

jobs. Since Michael Porter published The Competitive

It is generally agreed that, besides the well-known Advantage of Nations (1990), clusters of (innovative)

factors from traditional neo-classical theory, other fac- ® rms have become a popular policy objective, not only

tors like technology, or more general knowledge, are in the European Union, but in the Netherlands as well.

According to PORTER, 1990, p. 149 a cluster consistsbecoming more and more important in explaining the

economic performance of ® rms and, as a consequence, of industries linked through vertical (buyer/supplier) or

horizontal (common customers, technology, channelsthe economic development of regions (LAMBOOY,

1997). Or in the words of FREEMAN , 1982, p. 169, etc.) relationships. PORTER, 1990, p.157, also notes

the importance of geographical concentration for`not to innovate is to die’. This explains why know-

ledge-intensive activities, for example those carried out innovation. In this paper we concentrate speci® cally

on the innovative links of high technology small andby high technology ® rms, take a central position in the

economic policy of most European countries. It is medium sized enterprises (SMEs) with customers, sup-

pliers and knowledge centres.assumed that such ® rms are an excellent starting point

0034-3404/99/040391-10 ©1999 Regional Studies Association

Dow

nloa

ded

by [

Uni

vers

ity o

f B

ath]

at 0

4:36

19

Nov

embe

r 20

14

Page 3: Clusters of High Technology SMEs: The Dutch Case

392 Egbert Wever and Erik Stam

The Randstad, or more generally the Netherlands, sector, even when no innovation aspects are involved.

Such diVerences make the Randstad case as regardscan be considered an interesting case concerning the

location and the spatial clustering of high technology high technology services of particular interest.

® rms. This is, ® rstly, because of its speci® c spatialcontext, which is quite diVerent from that in which

HI G H T E CHNOL OG Y CL US T E RS INthe Cambridge Phenomenon, the `Third Italy’, the

T H E D UT CH S PAT IA L CONT E X TCatalonian or Sophia-Antipolis high technology clus-

ters operate. The Dutch context is characterized by its Although the Dutch spatial context is speci® c, we often

use the results of research in which notions like local/restricted national size and consequently relatively smalldiVerences between regions. Although the Netherlands regional’ or `high technology’ have quite another

meaning than in the Netherlands. So our research isas a whole is comparable in size with Baden-

WuÈ rttemberg, 12 provinces and 40 statistical areas also directed towards the identi® cation of `Third Italies’ ,

`Porter clusters’ , `Cambridge phenomena’ or innovative(Corop regions) can be distinguished. The Dutch pub-

lic educational system is extremely deconcentrated. milieux in general. We argue that the relatively homo-

geneous Dutch spatial context in¯ uences the locationUniversities and polytechnics are evenly spread overthe country, implying that there are no large variations of high technology ® rms, and the spatial clustering of

high technology ® rms, in a very speci® c way.in regional education levels. Information is easily avail-

able from local Chambers of Commerce or Innovation As the literature abounds with theories dealing with

clusters of high technology ® rms and/or innovativeCentres, which are distributed evenly over the country

as well. Moreover, there are no regional institutional milieux, we will not dwell at length on them. The idea

behind resource muni® cence theory, a more or less neo-diVerences. Employees and employers are primarilyorganized nationally. Wage negotiations are conducted classical approach, is that conditions for ® rms to innov-

ate diVer regionally. It is assumed that some regionsat the national level, implying that regional diVerences

in labour costs per sector hardly exist. There are hardly have a more favourable resource base than others: better

educated labour; more and better knowledge centres;any regional or local taxes. As a consequence diVerences

in regional business conditions are relatively small. more venture capital ® rms; more high-quality services;etc. Such a resource base might stimulate ® rms toA second argument makes the Randstad an interest-

ing case. The literature about clusters of high techno- innovate (OERLEMANS, 1996; KEEBLE, 1997).

The learning economy or learning region approachlogy ® rms deals primarily with high technology

manufacturing ® rms (O AKEY, 1995; OERLEMANS, (LUNDVALL, 1992; NELSON, 1993; STORPER, 1995;

MORGAN , 1997), building on Schumpeter and much1996). However, the Dutch economy as a whole can

hardly be characterized as high technology. According to more evolutionary in character, is based on the idea thatinnovation is the outcome of a process of interactionthe OECD, 1994, the employment share of high and

medium technology industrial activities as a proportion between customers, suppliers and knowledge centres.

As a consequence, proximity of these actors and anof all manufacturing employment in the Netherlands

was only 40% in 1994 against 54% in Germany. Perhaps institutional framework that favours interaction will

stimulate the formation of regional clusters of innova-this is one of the reasons why attempts to attract foreign

high technology ® rms to the Netherlands often fail, tive activities.This aspect of interaction can also be found in acompared with Dutch success in attracting distribution

companies (JACOBS et al., 1990; D E L IGT and number of other approaches, for example in the network

approach of the IMP research group (HAKANSSON ,WEVER, 1998). More generally, the Dutch high

technology sector is heavily dominated by high techno- 1993; KAMANN , 1998). In this approach the role of

sustainable, stable, interpersonal relations is stressed,logy business services, especially activities in the ® eld

of information and communication technology: the implying again the relevance of institutional and prox-imity aspects. In Nooteboom’s dynamic transaction costsICT sector. High technology business services diVer

in several respects from high technology manufacturing approach (NOOTEBOOM , 1992, 1996) innovation is

linked to transaction costs and distance. According to® rms. Compared to manufacturing, customers of (high

technology) services generally account for more Nooteboom, regional clustering is especially attractive

when the activities involved are in the ® rst phase ofinnovation impulses than suppliers, although someauthors claim the opposite (e.g. EVANGELISTA and their product life cycle. In this phase, particularly, the

exchange of tacit knowledge, which requires distance-S IRILL I, 1997). Often these customers ask for services

made to measure, implying that something new has to sensitive face-to-face contacts in order to avoid high

transaction costs, is important.be added to a standard service, leading to incremental

innovation (FROWEIN et al., 1998). At the same time The last approach we will deal with is the well-

known theory of regional innovative milieux, con-it is, again compared to manufacturing, much morediYcult to de® ne notions like product and process nected with the GREMI research group (CAMAGNI,

1991; HANSEN , 1992). It includes notions like `net-innovation in the service sector. And face-to-face con-

tacts are always essential in the (high technology) service works’, transaction costs’ , trust’ and regional embed-

Dow

nloa

ded

by [

Uni

vers

ity o

f B

ath]

at 0

4:36

19

Nov

embe

r 20

14

Page 4: Clusters of High Technology SMEs: The Dutch Case

Clusters of High Technology SMEs: The Dutch Case 393

dedness’ . In this approach it is assumed that the relative number of high technology ® rms are small.

We use the notion region’ in the Dutch sense, that is(manufacturing) SMEs, being unable to innovate on

their own, are enabled to generate innovations by in terms of very small regions. We hypothesize in the

second place that we will not ® nd regional clusters ofworking together intensively. Such co-operation isstimulated by the presence of a common set of norms high technology ® rms in Dutch regions. A regional

cluster is de® ned here as a geographical concentrationand values (the institutional element), but this often

presupposes proximity of the actors involved. A con- of ® rms which exhibit a signi® cant degree of intrare-

gional linkages. We will focus especially on linkagescentration of rivals, customers, and suppliers promotes

eYciencies and specialisation and, even more impor- with customers, suppliers and knowledge centres. Wewill concentrate most on the innovation impulses thattant, stimulates innovation’ (HANSEN, 1992, p. 97).

How should we evaluate these approaches in the such a regional cluster exercises on the ® rms belonging

to that regional cluster.Dutch context? We should realize that they all stress in

one way or another ® rm-external stimuli for innovation. In relation to both hypotheses we do not deny

that there are clusters of economic activities in theBut this presupposes that substantial regional diVerences

exist in resource conditions, institutional frameworks, Netherlands, even regional and local clusters. Since the1950s and 1960s, the port of Rotterdam has containedand values and norms. In the Randstad context, with

relatively homogeneous regional business environ- an example of CHARDONNET ’s, 1953, oil and petro-

chemical complex. Horticultural clusters, for examplements, this assumption can be seriously questioned. In

such a context we should not underestimate the role vegetable cultivation in the Dutch `Glass City’ (between

Rotterdam and The Hague) and ¯ ower growing in theof ® rm-internal factors: the individual entrepreneur’s

inclination to innovate; the way in which an entrepren- Aalsmeer region, are well known examples. In thesmall municipality of Urk we have a traditional ® shingeur organizes his ® rm’s internal human resources

(VAESSEN , 1993; V AESSEN and WEVER, 1993; cluster. For some speci® cally modern activities, such as

the multimedia sector, a cluster can be identi® ed inVAESSEN and KEEBLE, 1995; OERLEMANS, 1996).

This implies, at the same time, that we should not the Northern wing of the Randstad, in and near

Amsterdam. However, we decided not to concentrateoverestimate the role of geographical’ factors such asdistance and regional institutional networks within our research on such existing (small) clusters. Our

research was focused on the issue of the existence ofsmall and relatively homogeneous regions like the

Randstad. The Randstad incorporates the three west- regional clusters of high technology activities. Given

the speci® c Dutch spatial context we assume that forern provinces of North Holland, South Holland and

Utrecht, an area of 8,922 km2, i.e. only one-fourth of high technology activities such clusters will primarily

be national clusters, functioning within a homogeneousthe size of Baden-WuÈ rttemberg. Generally speaking,within such a context distance is not a real problem. national institutional setting.

In the next section we will present some of theAnd what about the institutional context? Without

doubt, in a context in which resource elements hardly results of our empirical research. We deal ® rst with

the location hypothesis, followed by the regional clustervary spatially, it is appealing to assign a signi® cant role

to the impact of diVerences in institutional frameworks hypothesis.

or regional Wirtschaftsgeist as CHESHIRE, 1996,pp. 1± 2, has already noted: `Regions which develop

H IG H T E CH NOL OG Y S M E s IN T H Emore successful territorially competitive policies will

NE T HE RL A ND S : T HE L OCAT IONgrow at the expense of those that do not’ . But here as

PAT T E RNwell one can doubt whether big diVerences in at least

formal regional institutional frameworks exist within In our empirical research and mainly for pragmatic

reasons we have used BUTCHART ’s, 1987, de® nitionthe Randstad. One should realize that in the Randstad,in addition to numerous polytechnics, there are ® ve of high technology industries. Based on this de® nition,

the EIM, the Dutch economic institute for SMEs,universities and until recently eleven Chambers of

Commerce. Perhaps it is not by accident that HANSEN , prepared a database of all high technology ® rms which

employ less than 100 employees. This reveals that 8´1%1992, p. 95, while giving examples of regional high

technology clusters in Italy, Spain, France, Portugal, of all such ® rms in the Netherlands can be classi® ed ashigh technology according to the Butchart de® nition:Denmark, Germany, and Switzerland, did not mention

any example from the Netherlands. 1´5% are high technology manufacturing ® rms and

6´6% high technology service ® rms, most of whichIn this paper we are not going to deny the potentially

important role of the factors stressed in the various belong to the ICT sector. In 1996 the share of this

ICT sector in Dutch GNP was 5 4% (BOOZ , ALLENtheoretical approaches. But bearing the speci® c Dutch

context in mind, and on the basis that geography & HAMILTON, 1998). So we are certainly not dealingwith a substantial part of the Dutch economy. Whenmatters’ , we may formulate two hypotheses. We assume

in the ® rst place that, within the Netherlands and the number of high technology ® rms is expressed per

100 ® rms in all sectors,1 the province of Zeeland hasespecially within the Randstad, regional diVerences in

Dow

nloa

ded

by [

Uni

vers

ity o

f B

ath]

at 0

4:36

19

Nov

embe

r 20

14

Page 5: Clusters of High Technology SMEs: The Dutch Case

394 Egbert Wever and Erik Stam

Fig. 1. Number of high technology ® rms and total number of ® rms in the Corop regions, 1995

the lowest score nationally (3 5), and the new province

of Flevoland the highest (10 8). Within the Randstad

the province of Utrecht registers the highest score: 9 7

per 100 ® rms, against 8 7 for South Holland and

7 9 for North Holland. Together the three Randstad

provinces contain a relatively large number of hightechnology ® rms: 54% of the national total. However,

this share becomes less impressive when we realize that

these three provinces also account for 51% of all ® rms.

For the smaller Dutch Corop regions diVerences

are, not surprisingly, larger. But, here as well, these

diVerences more or less disappear if we relate thenumber of high technology ® rms to the regional stock

of ® rms (Fig. 1).

The few outliers in Fig. 1 can easily be explained.

For example, the high score of Flevoland (this Corop

coincides with the province of Flevoland) is related tothe fact that in this new province until now only a

relatively small number of ® rms have existed. The low

score of Zeeland is related to the fact that this small

regional economy is for many service activities depend-

ent on firms located in nearby Corops, such as Rotter-

dam. Within the Randstad the diVerences are relativelysmall. However, Fig. 1 also clearly shows that three

(Randstad) Corops contain, in absolute terms, the most Fig. 2. Share of high technology service ® rms in total numberhigh technology SMEs: Utrecht; Greater Amsterdam; of ® rms in Corop regions, 1995and Greater Rotterdam.

Some interesting minor diVerences emerge when we

diVerentiate between high technology service and high Summarizing, where the manufacturing sector tradi-technology manufacturing ® rms. Within the national tionally dominates (outside the Randstad) we ® ndcontext, the Randstad is less important for the latter relatively more high technology manufacturing ® rms.(Fig. 3) than for the former (Fig. 2). But this is again Where the service sector traditionally dominatesless surprising when we link these numbers to the total (within the Randstad) we ® nd relatively more high

number of ® rms in manufacturing and services as we technology service ® rms. This is certainly not new.did in Fig. 1. The Randstad (the three western prov- Research in the 1980s, although using somewhatinces) contains 48% of all Dutch high technology diVerent de® nitions of high technology, yielded virtu-manufacturing ® rms and 56% of all Dutch high techno- ally the same picture (KOERHUIS and CNOSSEN ,logy service firms, compared to 51% of all ® rms. High 1982; DRENTH , 1990). In a recent publication,technology manufacturing ® rms are over-represented STOREY and TETHER, 1998, p. 942, show the same

outside the Randstad, where the manufacturing sector outcome for new technology-based ® rms in the Euro-

pean Union.generally is over-represented.

Dow

nloa

ded

by [

Uni

vers

ity o

f B

ath]

at 0

4:36

19

Nov

embe

r 20

14

Page 6: Clusters of High Technology SMEs: The Dutch Case

Clusters of High Technology SMEs: The Dutch Case 395

(293) and in high technology services (1,439). The

latter total is greater than that for Greater Amster-

dam (1,287) or Greater Rotterdam (1,001).

2. The province of Utrecht is highly oriented to ser-vices: 83% of all its high technology ® rms can be

found in business services, most of them in the

information communication sector (ICT). As a

contrast, it was therefore decided to study high

technology SMEs in the municipality of Rotterdamas well, since this region is traditionally more ori-

ented to manufacturing. Both regions are endowed

with many relevant resources; both have a university

(including an academic hospital), several polytech-

nics, a Chamber of Commerce, an Innovation

Centre, etc.

We assume that it is generally unlikely that regionalhigh technology clusters are to be found elsewhere in

the Netherlands when such clusters are totally lacking

in our research regions. Of course this assumption

might be wrong. It is theoretically always possible that

a cluster could be found even in a region with arelatively small number of high technology ® rms. In

order to ascertain whether clusters exist in our two

study areas we interviewed 94 high technology ® rms,Fig. 3. Share of high technology manufacturing ® rms in total partly by telephone. We used a slightly modi® ed version

number of ® rms in Corop regions, 1995 of the TSER Network’s common survey question-

naire.2 In use, it soon became clear that the question-naire, oriented particularly towards high technologyRE G IONA L CL US T E RS OF HIG Hmanufacturing ® rms, raised some problems for ourT E CH NOL O G Y S M E s IN T H Eservice ® rms. Notions like process and product innova-NE T HE RL A ND S ?tions and radical and incremental innovations are well

The conclusion ± that most Dutch high technology known in manufacturing, but for a high technologymanufacturing and service SMEs are to be found in service ® rm it is very diYcult to answer questionsthose areas where manufacturing or services dominate about aspects such as process innovations. Followingthe regional economy ± does not of course automa- LUNDVALL, 1988, 1993, we used the notions ® rmtically imply that regional clusters are lacking. Clusters internal’ (process innovations), for example computer-as we de® ned them earlier might exist, especially when ization, and external’ (product) innovations, mostlywe realize that in the Netherlands business services are incremental innovations resulting from the needs orhighly dominant within Butchart’s high technology speci® cations of customers or suggestions from sup-classi® cation. It is generally agreed that such ® rms are pliers of software. In this report we will concentrate onrelatively sensitive to distance from their customers. In the entrepreneurs’ opinions about the role of externalorder to test the second hypothesis about the presence partners or economic actors (customers, suppliers,of regional clusters, a research project was carried out knowledge centres, intermediate organizations) forin the Randstad, focused on two regions containing a innovation within and modernisation of their ® rms.large absolute number of high technology SMEs (see

Fig. 1): Corop Utrecht (which coincides with the

province of Utrecht); and the municipality of Rotter- S OME CH A RACT E RIS T ICS OF T H E

RE S E A RCH P OP UL AT I ONdam (an important part of Corop Greater Rotterdam).

This choice was based on several arguments:Some 80 out of all 94 interviewed SMEs belonged to

the sectors of computerization, software development1. The province of Utrecht is one of the biggest Dutch

Corop regions, by area as well as number of ® rms. and programming services, the so-called computer ser-

vice ® rms. The number of manufacturing SMEs,Where the total stock of ® rms is concerned, Utrecht

ranks second after Greater Amsterdam. However, it which tend to be heavily emphasized in the literature

on innovative milieux, is very limited, namely 14. Thiscontains the largest total number of high technologySMEs (1,732) in the Netherlands, even more than outcome was not that surprising, given the orientation

of the Dutch economy to services. However, even theGreater Amsterdam (1,512), as well as the biggest

number of high technology ® rms in manufacturing Rotterdam sample contained only four manufacturing

Dow

nloa

ded

by [

Uni

vers

ity o

f B

ath]

at 0

4:36

19

Nov

embe

r 20

14

Page 7: Clusters of High Technology SMEs: The Dutch Case

396 Egbert Wever and Erik Stam

high technology ® rms out of a total of 39. Of all

interviewed ® rms 72 were very small (10 employees or

fewer), 53 having even fewer than ® ve employees. This

® rm size distribution had one advantage: in nearly allcases we interviewed the owner. Often this was the

original founder of the ® rm, since nearly 50% of all

® rms have been in existence for less than ® ve years.

Ten ® rms were founded as branch plants by existing

® rms, 29 were spin-oVs and 55 could be characterizedas normal SMEs, established as a conventional start-up

by one or more entrepreneurs. Most of the spin-oVs,

which comprise the bigger ® rms in our sample, came

from large incubator companies such as Apple Europe,

Bull, Alcatel, IBM and Digital. These incubator organ-

izations were often located in the Randstad, althoughnot all (only 48%) were sited in the Utrecht or Rotter-

dam regions. In those cases where the incubator ® rm

was elswhere in the Randstad, the spin-oV entrepreneur Fig. 5. The importance3 of formal and informal contacts forwas often resident in the Utrecht or Rotterdam region. innovative activitiesThe entrepreneurs behind the interviewed high

technology SMEs display Schumpeterian entrepren-eurial spirit’ , going by the reasons which they gave for

starting their ® rms; these focus chie¯ y on the wish to lower score for suppliers. Since our respondents are

very small ® rms we had expected that they would justbe creatively engaged, the wish to be independent and

identifying a market niche. buy standard software from normal wholesalers. This

did happen. However, in many cases they receivedinteresting tips from their suppliers. This explains the

HI G H T E CHNOL OG Y CL US T E RS positive rating of suppliers. There are hardly any diVer-

ences in the scores between spin-oVs and normal highFor our high technology SMEs by far the most impor-technology SMEs. Only the very small ® rms receivedtant external innovation impulses came from otherfewer innovation impulses from their customers and® rms (Fig. 4). This outcome is in line with many othersuppliers. But some of these hardly saw themselves asresearch ® ndings, including research among non-highbeing innovative. They are more similar to traditionaltechnology businesses (JACOBS and VAN DER

jobbers’ in manufacturing.MEIJDEN , 1995; TAMAÂ SY and STERNBERG, 1998).How do external innovative impulses reach the ® rmsThe diVerence between customers (businesses and insti-

we studied, or in other words how does the interactiontutions like schools, hospitals, government depart-between our respondents and their customers andments) on the one hand and suppliers (wholesalers andsuppliers take place? In many cases this takes the formdistributors of basic software) on the other hand wasof one-sided interaction: the customer wants to besmaller than we had expected. We had expected asupplied according to his speci® cations. This ® nding is

certainly in¯ uenced by the size of our respondents. We

are dealing with very small ® rms which obtain orders

mainly from much bigger companies or organizations.

This may explain why many high technology SMEsstress the importance of formal contacts (Fig. 5). This

holds especially for contacts with public organizations.

Nevertheless, sometimes there is real co-operation,

with a possibility of collective learning. In these cases

the customer has a speci® c (software) problem that canonly be solved when both partners co-operate. The

outcome of this problem is sometimes a new pro-

gramme. But genuine interaction of this kind, with

its potential for collective learning, is certainly not

dominant. In such a situation our high technology

SMEs stress the importance of formal contacts ± evenwhen informal contacts, which often go hand in hand

with formal activities, exist.Fig. 4. The importance3 of external partners for innovative

activities Compared to vertical co-operation there is much

Dow

nloa

ded

by [

Uni

vers

ity o

f B

ath]

at 0

4:36

19

Nov

embe

r 20

14

Page 8: Clusters of High Technology SMEs: The Dutch Case

Clusters of High Technology SMEs: The Dutch Case 397

less horizontal co-operation (see Fig. 4). The role

of knowledge centres (universities, polytechnics, large

public laboratories) is very limited, although each

respondent is in the vicinity of one or more suchknowledge centres. This again is not a surprising out-

come. In the Netherlands the role of such institutions

is relatively unimportant for SMEs as a whole. This is

also the case for bigger manufacturing ® rms in relatively

large areas in Germany (TAMAÂ SY and STERNBERG ,1998). Only 20 of our SMEs regarded co-operation

with a knowledge centre as being of importance. Even

in these cases we should not overestimate the innovative

impulses generated by the institutions involved. The

most important goals of co-operation with knowledge

centres as reported by our respondents involved thevisiting of a centre’s library and the temporary recruit-

Fig. 6. The importance3 of regional customers and suppliersment of students. There were no diVerences in responsefor innovative activitiesrates between spin-oVs and normal high technology

SMEs. But very small (up to 10 employees) ® rms made

less use of knowledge centres than small SMEs (11± 100

employees). Not surprisingly, it seems to be somewhatnon-regional customers. There is hardly any diVerenceeasier for larger small ® rms to realize contacts with

in these ® ndings between size categories. On theknowledge centres. Much more relevant, however,question of whether they were explicitly looking forwere respondents’ opinions about the relevance of suchco-operating partners in their own region, the respond-innovation impulses with respect to regional origin:ents mostly answered: `No, why should I?’. One entre-the most important impulses were transmitted frompreneur explicitly answered: `The only geographicalknowledge centres located outside the region. In nearly

factor is the national border’. On the other hand, manyall cases this means elsewhere in the Netherlands. In

respondents con® rmed that existing personal relations’the Netherlands, distance to a knowledge centre doeswere important for establishing contacts with othernot therefore seem to be a problem.® rms. But in the Dutch spatial context `existing per-Only 17 SMEs stressed the relevance of regionalsonal relations’ seemingly extend beyond the bound-institutions (Chambers of Commerce, Innovationaries of Utrecht, even though this is certainly not theCentres, labour unions, employers’ unions). Contacts

smallest Dutch Corop region.with these institutions were not considered particularlyHorizontal partners, mainly other SMEs, were morerelevant for the innovative development of the ® rm.

often found in the region itself. But the role of theseThey have more to do with daily problems, althoughregional partners for the innovative development of thethey include information about training and retraining® rms involved was considered much less important thanschemes as well. Surprisingly, innovation centres, whichthe role of vertical partners, customers and suppliers.have been especially created by the Dutch Government

A similar picture was found for co-operation withto stimulate the innovation activities of SMEs, wereknowledge centres (Fig. 7). Although the Utrechtnot mentioned at all. This may have to do with theregion is endowed with many such centres, most offact that these centres are focused on technology andthe contacts that were judged important took placemanufacturing ® rms. It may also be related to the smallwith centres located elsewhere. This is especially truesize of the ® rms in our study.for research activities. Only three of the ten Utrecht

SMEs that co-operated formally with knowledgeRE G IO NA L CL US T E RS

centres in the ® eld of R&D collaborated with a partner

located within the province of Utrecht. Four SMEsFor testing the hypothesis about regional clusters weeven collaborated with partners from abroad. Andshould explicitly concentrate on the role of regionalalthough we concluded earlier that it is easier for lessactors. This part of the paper therefore concentrates

small SMEs to establish informal co-operation linkson the Utrecht region. This region con® rms ourwith knowledge centres, this does not result in thehypothesis. Most of the customers and suppliers thatUtrecht SMEs forging strong links with their ownare considered relevant for the innovative developmentregional centres. However, our Rotterdam respondentsof the interviewed SMEs are located outside the regionreported more regional informal contacts which they(Fig. 6).considered not very important for their innovativeNone of the SMEs interviewed in fact co-operated

activities. Of course, this outcome may be in¯ uencedexclusively with regional customers. On the other

hand, six received innovative impulses exclusively from by the fact that we included only the bigger knowledge

Dow

nloa

ded

by [

Uni

vers

ity o

f B

ath]

at 0

4:36

19

Nov

embe

r 20

14

Page 9: Clusters of High Technology SMEs: The Dutch Case

398 Egbert Wever and Erik Stam

The ® rst relates to the use of the relatively old

BUTCHART, 1987, classi® cation. This classi® cation,

published in 1987 and based on yet older data, may

have in¯ uenced our results. We all know that the lifecycle of new products is shortening drastically. It may

be hypothesised (NOOTEBOOM , 1996) that the life

cycle of regional clusters of high technology SMEs is

shortening as well, in the sense that spatial clusters of

new activities will be confronted with deconcentrationtrends much more rapidly than new activities in the

1960s and 1970s. If this argument is correct, we might

perhaps have found regional clusters of BUTCHART ’s,

1987, high technology SMEs if we had carried out

our research in the 1970s or if we had concentrated

our research on ® rms carrying out really new activitieswhich may well not be included in the relatively old

Butchart classi® cation. The multimedia sector mightFig. 7. Formal and informal co-operation withbe a case in point. VAN ECK VAN DER SLUIJS et al.,knowledge centres1998, found a clear concentration of ® rms in this

sector in the northern wing of the Randstad. This

concentration may perhaps include inter-® rm clus-centres. We know (VAN WEESEP and WEVER, 1996),for example, that by far the largest share of all contract tering as well. On the other hand, we should of course

avoid the danger of drawing general conclusions onresearch at the University of Utrecht is not carried out

for small regional business ® rms, but for big inter- the basis of research involving rather exceptional cases.

The second question has to do with some character-national pharmaceutical firms such as HoVman-La

Roche, Solvay-Duphar and Glaxo-Wellcome. istics of our research scheme. We concentrated onrelatively small regions: the municipality of RotterdamThe relative absence of links between our high

technology ® rms and regional knowledge centres is and the province of Utrecht. The latter is a region that

is centrally located within the Netherlands, accessiblenot surprising. Studies stressing such regional links are

often focused on ® rms that are more R&D-based than from any location in the country within two hours

by car. Although other research results (GLAS, 1996)the ® rms we are dealing with. Thus L AWTON SMITH ,

1998, found that, in Oxfordshire, 66% of all R&D- provide little support for this, our outcomes might havebeen diVerent if we had carried out our research in abased ® rms had links with the local university and local

public sector laboratories. But there as well, for ® rms bigger region, that is oriented less exclusively towards

business services and is less centrally located within thein the computer and business services sector this per-

centage was much lower, at only 29%. Netherlands.

The third question has to do with our de® nition of

SMEs. We only included ® rms with fewer than 100employees, so it is not surprising that we found a large

S OME CONCL UD I NG RE MA RK Snumber of very small ® rms. The size of our respondents

may explain why links with universities or other institu-Our research drew upon the theoretical literature about

innovative regional milieux and has argued that we tions are rather weak, and why formal and one-sided

contacts with their relatively big customers dominate.should be careful about transferring research ® ndings

from one spatial context to another. On the basis Certainly some of our results would have been quitediVerent if we had interviewed much larger highof EIM’s national database we tested the hypothesis

concerning the existence of regional (in the Dutch technology service ® rms. In that case we might have

found more links with universities and a more impor-sense) clusters (that is, a concentration of ® rms with

intraregional linkages) of high technology SMEs within tant role for informal contacts. There is, however, no

obvious reason why such ® rms’ regional links, in thethe Randstad, especially in the province of Utrecht,the most important region for such SMEs, and the Dutch sense, would be much more important.

municipality of Rotterdam. Although only a small

number of interviews (94) could be achieved, our

outcomes are in line with both hypotheses: the spatialNOT E Spattern of high technology SMEs in the Netherlands

closely follows the general spatial pattern of ® rms and 1. Using the number of high technology ® rms per 10,000regional clusters hardly exist. However, three quali® ca- inhabitants gives nearly the same results.tions to, or comments on, these ® ndings should perhaps 2. This questionnaire was developed as a collective enterprise

by the members of the TSER Network at its 1996be made.

Dow

nloa

ded

by [

Uni

vers

ity o

f B

ath]

at 0

4:36

19

Nov

embe

r 20

14

Page 10: Clusters of High Technology SMEs: The Dutch Case

Clusters of High Technology SMEs: The Dutch Case 399

3. The importance of co-operation with external partnersSophia-Antipolis meeting. It covers seven themes: charac-

teristics of the ® rm; characteristics of the local labour in innovative activities was measured using a ® ve-stepLikert scale. The range of the scale varied frommarket; relationships with customers; suppliers; know-

ledge centres; other actors; and sources of collective 1 5 nonexistent to 55 essential, with the range 3± 5 being

considered as denoting importance.learning.

RE F E RE NCE S

BOOZ , ALLEN & HAMILTON (1998) Netherlands’ ICT Twinning Centers and Investment Funds: Building the Mind-set and the

Skill Base for the Information Society. Booz, Allen & Hamilton, Amsterdam.BUTCHART R. L. (1987) A new UK de® nition of the high technology industries, Econ. Trends 400, 82± 88.

CAMAGNI R. P. (Ed) (1991) Innovation Networks: Spatial Perspectives. Belhaven, London.

CHARDONNET J. (1953) Les Grands Types de Complexes Industriels. Sirey, Paris.CHESHIRE P. (1996) The spatial impact of European integration, paper presented at the 36th European Congress of the RSA,

ZuÈ rich.

DE L IGT T. and W EVER E. (1998) European distribution centres: location patterns, Tijdschr. Econ. Soc. Geogr. 89(2), 217± 23.DRENTH D. (1990) De informatica-sector in Nederland tussen rijp en groen: een ruimtelijk-economische analyse, Netherlands

Geographical Studies 108. Royal Dutch Geographical Society, Utrecht.

EVANGELISTA R. and S IRILLI G. (1997) Innovation in services and manufacturing: results from the Italian survey, WorkingPaper 73, ERSC Centre for Business Research, University of Cambridge.

FREEMAN C. (1982) The Economics of Industrial Innovation. Frances Pinter, London.

FROWEIN J. C., PENNINGS L. J. and VERSTAPPEN M. J. F. (1998) De ICT Kennisinfrastructuur in Nederland. Een BenchmarkingStudie in het Kader van het Actieprogramma Elektronische Snelwegen. TNO-STB, Den Haag.

GLAS G. (1996) IndustrieÈ le netwerken, Netherlands Geographical Studies 201. Royal Dutch Geographical Society, Utrecht.

HAKANSSON H. (1993) The network as a governance structure; inter® rm co-operation beyond markets and hierarchies, inGRABHER G. (Ed) The Embedded Firm: On the Socio-economics of Industrial Networks. pp. 35± 51. Routledge, London.

HANSEN N. (1992) Competition, trust and reciprocity in the development of innovative regional milieux, Pap. Reg. Sci. 71,

95± 105.

JACOBS D., BOEKHOLT P. and ZEGVELD W. (1990) De Economische Kracht van Nederland. Een Toepassing van Porters Benaderingvan de Concurrentiekracht van Landen. SMO, Den Haag.

JACOBS D. and VAN DER MEIJDEN R. (1995) Waar haalt het MKB zijn kennis vandaan?, Econ. Stat. Ber., 25 January,

pp. 80± 83.KAMANN D.-J. F. (1998) Modelling networks: a long way to go, Tijdschr. Econ. Soc. Geogr. 89, 279± 97.

KEEBLE D. (1997) Small ® rms, innovation and regional development in Britain in the 1990s, Reg. Studies 31, 281± 93.

KOERHUIS H. and CNOSSEN W. (1982) De Software- en Computer-service Bedrijven. Geogra® sch Instituut, Groningen.LAMBOOY J. G. (1997) Knowledge production, organization and agglomeration, GeoJournal 41, 293± 300.

LAWTON S MITH H. (1998) Institutionalisation of informal networking and collective learning’ in the Oxfordshire high

technology economy, in KEEBLE D. and LAWSON C. (Eds) Collective Learning Processes and Knowledge Development in theEvolution of Regional Clusters of High Technology SMEs in Europe, pp. 85± 97. ESRC Centre for Business Research, University

of Cambridge.

LUNDVALL B. A. (1988) Innovation as an interactive process: from user-producer interaction to the national system ofinnovation, in DOSI G., FREEMAN C., NELSON R., S ILVERBERG G. and SOETE L. (Eds) Technical Change and Economic

Theory, pp. 349± 69. Frances Pinter, London.

LUNDVALL B. A. (1992) National Systems of Innovation: Towards a Theory of Innovation and Interactive Learning. Routledge, London.LUNDVALL B. A. (1993) Explaining inter® rm co-operation and innovation; limits of the transaction-cost approach, in

GRABHER G. (Ed) The Embedded Firm: On the Socio-economics of Industrial Networks. Routledge, London.

MORGAN K. (1997) The learning region: institutions, innovation and regional renewal, Reg. Studies 31, 491± 503.NELSON R. (Ed) (1993) National Innovation Systems. Oxford University Press, Oxford.

NOOTEBOOM B. (1992) Towards a dynamic theory of transactions, J. Evolutionary Economy 2, 281± 99.

NOOTEBOOM B. (1996) Innoveren, Globaliseren, Econ. Stat. Ber., 9 October, pp. 828± 30.OAKEY R. (1995) High Technology New Firms: Variable Barriers to Growth. Paul Chapman, London.

OECD (1994) The OECD Jobs Study. OECD, Paris.

OERLEMANS L. A. G. (1996) De Ingebedde Onderneming: Innoveren in IndustrieÈ le Netwerken. Tilburg University Press, Tilburg.PORTER M. (1990) The Competitive Advantage of Nations. Free Press, New York.

STOREY D. J. and TETHER B. S. (1998) New technology-based ® rms in the European Union: an introduction, Res. Policy

26, 933± 46.STORPER M. (1995) The resurgence of regional economics, ten years later: the region as a nexus of untraded interdependencies,

Europ. & Urban Reg. Studies 2, 191± 221.

TAMAÂ SY C. and STERNBERG R. (1998) Informal networking and collective learning by innovative SMEs in Germany, inK EEBLE D. and LAWSON C. (Eds) Collective Learning Processes and Knowledge Development in the Evolution of Regional Clusters

of High Technology SMEs in Europe, pp. 123± 34. ESRC Centre for Business Research, University of Cambridge.

Dow

nloa

ded

by [

Uni

vers

ity o

f B

ath]

at 0

4:36

19

Nov

embe

r 20

14

Page 11: Clusters of High Technology SMEs: The Dutch Case

400 Egbert Wever and Erik Stam

VAESSEN P. M. M. (1993) Small Business Growth in Contrasting Environments, Netherlands Geographical Studies 165. Royal

Dutch Geographical Society, Utrecht.VAESSEN P. and KEEBLE D. (1995) Growth-oriented SMEs in unfavourable regional environments, Reg. Studies 29, 489± 505.

VAESSEN P. and WEVER E. (1993) Spatial responsiveness of small ® rms, Tijdschr. Econ. Soc. Geogr. 84, 119± 31.

VAN ECK VAN DER SLUIJS P. A., HULSHOFF H. E. and PRINCE Y. M. (1998) Jonge, Kleine, Innovatieve ICT-Bedrijven inNederland. EIM, Zoetermeer.

VAN W EESEP J. and WEVER E. (1996) Een Universiteit in Utrecht: De Banden Tussen Universiteit Utrecht, Stad en Regio. University

of Utrecht, Utrecht.

Dow

nloa

ded

by [

Uni

vers

ity o

f B

ath]

at 0

4:36

19

Nov

embe

r 20

14