The Technology Park Berlin-Adlershof As an Example of Spatial Proximity in Regional Economic Policy

Zeitschrift für Wirtschaftsgeographie Jg. 52 (2008) Heft 4, S. 193-208

Elmar Kulke, Berlin

The technology park Berlin-Adlershof as an example of spatial proximity in regional economic policy

Abstract: Since the 1990s spatial economic policy increasingly tries to encourage the development of innovative firms especially in selected locations like technology parks. Spatial proximity in these locations should lead to the development of competitive clusters. The technology park Berlin-Adlershof is one example for this new approach of spatial economic policy; today it is – according to the number of employees - one of the largest technology parks in Europe. This paper discusses to what extent spatial economic policy has contributed to the development of an interlinked cluster. First elements of spatial economic policy will be evaluated according to the network and cluster approaches. Then general considerations of cluster development will be con- fronted with the development path of Berlin-Adlershof. Finally the extent of network formation will be discussed. The economic dynamic in Berlin-Adlershof shows that a proximity-oriented approach of spatial economic policy has been successful. Spatial proximity has encouraged the development of networks between the units, but up to now networks between science and business are not very strong. Keywords: technology park, spatial proximity, spatial economic policy, Berlin-Adlershof, network development

Introduction processes and innovations. The innovative strength can be supported by institutions that Since the mid-nineties structural and spatial provide special qualifications (e.g. universi- economic policy in Germany has been increas- ties, technical schools) and by the use of joint ingly influenced by the idea of developing sec- infrastructure facilities. toral or regional concentrations of interlinked production units (cf. ARL 2006;FLOETING The technology park Berlin-Adlershof is one 2008;FROMHOLD-EISEBITH/EISEBITH 2005; example of this new approach to spatial eco- MOSSIG/KLEIN 2003; REHFELD 1999). On the nomic policy. Some fifteen years ago the first regional level spatial economic policy often decisions were made for the development of a tries to encourage the development of sectoral science and business park in Adlershof. The clusters following a commodity chain; it is ex- idea was to connect innovative new enterpris- pected that companies belonging to one focal es with research institutes and with university activity of economy and having intensive link- research and education; implicit in the concept ages with each other are gaining competitive was following the idea of generating – based advantages. On the local level technology cen- on the spatial proximity of the units – an inno- ters and technology parks are an important vative and interlinked cluster. But more than policy instrument to encourage economic that, the concept was following an agglomera- growth based on the spatial proximity of inno- tion-oriented approach; there was the hope that vative firms (cf. BEHRENDT 1996; GUN- simply by concentrating units at one location, DEL/LUTTMANN 2008;TAMASY 1996). The networks would develop. Explicit instruments guiding idea is that spatial proximity facili- for network management were not used. Today tates the development of linkages and commu- reports of Adlershof’s success are multiplying, nication between the units; this opens up the and one sometimes encounters almost euphor- possibility not only of reducing transport and ic statements about the dynamics of the site. In transaction costs but also of exchanging tacit his introduction to the 2005 annual report knowledge, which generates joint learning (WISTA 2006), Hardy R. Schmitz, head of 194 Zeitschrift für Wirtschaftsgeographie Heft 4 / 2008

WISTA Management GmbH, writes, “Earning are included in most model approaches in eco- money with science – Adlershof continues to nomic geography. The classic approaches dis- grow: 12.5 % more revenue, 10.8 % more em- cuss the advantages of agglomerations by con- ployment, 92 % utilized rentable space, and centrating on their cost-minimizing effects (cf. only three insolvencies out of 400 companies. GORDON/MCCANN 2000, 516 f.; KULKE 2008, This is Berlin Adlershof’s impressive balance 125 f.). The proximity of delivery and pro- sheet for the year 2005.” And the German cessing services, providers of other services, journal Der Spiegel heads its 2008 report on consumers, and political decision makers re- Adlershof: “Almost like China – in Berlin- duces not only the transportation costs for ma- Adlershof a science and technology park is terial goods, but also the transaction costs of booming; for experts it is a typical example for organizing and conducting business in net- the efficient use of subventions in the east”. works and relationships. The successful exten- sion of transportation and communication in- These statements refer to the (possible) suc- frastructures over the last few decades has, cess of a “real-life experiment” that is interest- however, led to a drastic decrease in trans- ing both for the scientific discussion in eco- portation costs; they are accordingly less im- nomic geography and for regional economic portant today as elements of spatial differenti- policy. During the last few decades – strongly ation. influenced by Porter’s discussion of the competitive advantages of clusters (PORTER 1993, At the same time, knowledge as a factor in 1999) – analysis of spatial concentrations of production is gaining importance in highly de- specialized economic activities has been a ma- veloped societies (e.g. GERTLER 2001; jor field of research in economic geography KLINE/ROSENBERG 1986). During the nineties (cf. ASHEIM/COOKE/MARTIN 2006). It is al- there was a shift in spatial economic research, ready widely accepted that these concentra- and since then technology, innovations, tions possess competitive advantages if the knowledge and creativity are seen as decisive economic units are linked with each other by factors in the spatial economic development of immaterial flows of knowledge and informa- advanced economies (SCHAMP 2007). These tion. But it remains disputed as to whether, by observations in scientific research have influ- means of spatial economic policy with initial enced economic and spatial policy and find government investments, networks, interlink- their expression in innovation-oriented ap- ages and exchange of tacit knowledge in the proaches (cf. FROMHOLD-EISEBITH 1999; sense of an innovative milieu or cluster of sci- FROMHOLD-EISEBITH/EISEBITH 2005). entific and business activities can be created successfully (cf. COOKE 2002, 157: “Can clus- The generation of innovations can be benefit- ters be built?”), or if such economically com- ed by the spatial proximity of actors in re- petitive concentrations of interlinked activities search, development, and business. The inno- – in the sense of geographical industrialization vative and competitive abilities belonging to – only develop by themselves. spatial concentrations of economic activity are especially advantageous when the actors man- The paper starts with some general considera- aging them possess strong personal and orga- tions concerning competitive clusters and on nizational linkages – usually referred to as the economic policy of Berlin. It will be dis- networks (cf. BATHELT 1998; FROMHOLD- cussed to what extent policy really follows a EISEBITH 2000; KULKE 2008, 84 f.). The ac- cluster-oriented approach. Then the develop- tors, meaning those who create, sustain and ment path and present characteristics of the use the networks, come from complementary Adlershof science and technology park will be businesses, from research and educational fa- presented. Finally, based on empirical studies, it cilities, and from various institutions. Func- will be asked to what extent Adlershof exhibits tional relationships between these actors are elements and features of a competitive cluster. based on reciprocity, trust, dialog, flexibility, and self-regulation. The networks within a spatial entity display a high degree of open- Advantages of spatial proximity ness and intensity, in contrast to a certain amount of restraint with outside actors. For the The advantages of spatial proximity in busi- successful long-term competitiveness of a net- ness activities have long been discussed and work, however, not only strong internal link- E. Kulke: The technology park Berlin-Adlershof 195 ages, “strong ties”, are essential but also the tivities have arisen (cf. DANNENBERG 2007, 18 co-existence of outside (worldwide) linkages. f.; KULKE 2008, 127 f.). Two of these cluster- Without these “weak ties,” the lock-in-phe- oriented concepts, the innovative milieu ap- nomenon can occur - that is, without a view of proach and the geographical industrialization the outside world, new trends or innovations approach, are concerned with the spatial con- that have been successful elsewhere are not centration of innovative activities and they de- sufficiently perceived and, where applicable, scribe the development path of these clusters. integrated (cf. BATHELT/DEPNER 2003). These They are suitable as explanatory substructures networks are useful (cf. GROTZ 1996) not only for the development in science and technology because of their classic economical advan- parks such as Berlin-Adlershof. Although tages (e.g. the realization of economies of these two approaches incorporate similar scale and economies of scope, advantages in characteristics and elements, they differ deci- terms of transaction costs, risk minimization), sively on the issue of the starting point and but above all because they allow the possibil- pathway of development. While the innova- ity of common processes of learning and inno- tive milieu approach assumes that the exis- vation (learning by interaction). This is aided tence of a milieu is a prerequisite for develop- by the fact that within a network not only for- ment, the geographical industrialization ap- mal knowledge is exchanged, but also knowl- proach asserts that growth industries generate edge that has not yet been codified and made their own innovative surroundings. This dif- public, which is in part based on experience ference in terms of the developmental starting (tacit knowledge). This can usually, however, point is critical for regional economic policy. only take place through personal contact, making spatial proximity therefore very According to the milieu approach, policy meaningful, and is scarcely replaceable by should aim at bringing together research com- telecommunications. In addition spatial prox- ponents, such as universities or research insti- imity opens up the possibility of observing the tutes, and support the resulting spin-off busi- innovations of competitors, thereby increasing nesses; this is essentially the approach that has the innovative dynamics of enterprises, sus- been taken at Berlin Adlershof. The innovative taining their position in competition. Inclusion milieu approach is based primarily on the in a common economic, cultural, social, and work of the French Groupe de recherché eu- political environment (embeddedness) is also ropeen sur les milieux innovateur (GREMI) an important facet of the functionality of a net- (cf. AYDALOT 1986). Aside from the basic idea work. Infrastructural facilities contribute to that creativity arises from the mutual, fre- this, as do training stations or economically quent, and intense exchange of information, connected institutions. bringing together formerly unconnected knowledge and new solutions, the approach The discussion shows that spatial proximity stresses the importance of the surrounding en- can encourage the development of networks vironment. This environment is in part created and may contribute to the generation of inno- by the actors themselves, who, because of vations. These concentrations optimally form their common technical culture, social charac- a functional cluster. According to PORTER teristics, and interests, produce a feeling of (1998) a cluster is defined as follows: “A clus- community and a site image. The milieu can ter is a geographical proximate group of inter- also be cultivated by the adoption of suitable connected companies and associated institu- economic policy instruments and the setup of tions in a particular field, linked by common- institutions suitable for a business environ- alities and complementarities.” FESER (1998, ment. This includes support in the form of re- 26) additionally points out the competitive ad- gional and scientific policy, as well as the gen- vantage of these concentrations: “Economic eration of human capital by means of training clusters are not just related and supporting in- and the development of institutions closely dustries and institutions, but rather related and tied to business interests. Using this milieu as supporting institutions that are more competi- a basis, competitive economic activities in a tive by virtue of their relationship.” specialized field of activity can then develop.

Over time, a relatively large number of indi- In contrast, according to the geographical in- vidual conceptual approaches to the spatial dustrialization approach, the settlement of concentration of industry and innovative ac- key industries would be the deciding factor, 196 Zeitschrift für Wirtschaftsgeographie Heft 4 / 2008 the development of which could, if necessary, Berlin (KULKE 2003). With reunification the be supported by complementary facilities. manufacturing sector of Berlin faced severe First, many possible alternative sites exist for problems. Many of the formerly subsidized the locational choice of innovative new indus- manufacturing enterprises of former West- tries (window of locational opportunity; STOR- Berlin moved to other locations. In former PER/WALKER 1989; cf. BATHELT 1992). Sites East-Berlin almost 80 % of the jobs in the with a minimal density just outside of already manufacturing industries were lost, as produc- existing agglomerations (which contain older tion was not competitive in either national or production areas) are especially suitable. At international markets, and the enterprises had the beginning of the development path, new to close. The transformation process led to a production areas at multiple sites arise, yet as major change in the employment structure of the expansion process progresses, only one of Berlin; today various services are the most im- these survives (selective clustering). At this portant industries, while manufacturing activ- site, through the immigration of mobile pro- ities only contribute less than 20 % to the la- duction factors, it comes to an increased con- bor force. Berlin still has weaknesses in the centration of activity. The further settlement of size structure and technology intensity of businesses, start-ups, and spin-offs takes place, manufacturing. Most notably, larger units with both of production industries and of the accom- innovative products and a high share of re- panying service industries. Additionally, a spe- search and development activities are missing. cial surrounding environment of institutions The manufacturing sector is dominated by and facilities develops, tailored to meet the smaller units that serve the local market needs of new activities of the area. In this case (KULKE 2003). spatial economic policy only has to encourage the development of key industries and later to Berlin spatial economic policy faced the prob- assist with the establishment of institutions. lem that resettlements of larger enterprises from other locations to Berlin were more or After considering the model approaches, the less an unrealistic idea. Therefore, the govern- question arises as to whether strengthening of ment concentrated its policy instruments on linkages and networks between actors in dif- encouraging the development of small, new, ferent areas, dependent on the factor of spatial technology-oriented enterprises, often estab- proximity, has taken place during the develop- lished as spin-offs from universities. With the ment path of the Adlershof site. It is also in- Business Plan Competition the policy is as- teresting to evaluate whether the observed sisting new enterprises during the establish- growth dynamic is based primarily on the in- ment phase by offering coaching services and dividual expansion of single enterprises and organizing contacts to venture capital institu- industrial segments (in the sense of geograph- tions. And the government is offering loca- ical industrialization), or rather constitutes the tions in founder/innovation centers and tech- result of networked activities and policy-led nology parks. In total, more than 900 Mio. € development of institutions (in the sense of the (with 670 Mio. € in public funding) were innovative milieu). invested to build up these locations in Berlin (Fig. 1). Today there are six general business incubation centers and eleven innovation Spatial economic policy in Berlin centers as well as five technology parks located in Berlin. They are offering more than Spatial economic policy in Berlin has long as- 400,000 m2 of space to enterprises; approxi- sisted in the development of new firms and en- mately 1,100 enterprises with more than couraged spin-offs from universities and re- 13,000 employees are located there. Estima- search institutes. In the year 1983 the first tions including companies that have success- business incubation center (BIG) of Germany fully moved out of the centers predict that was established in an old manufacturing build- more than 21,000 new jobs have been created ing in West Berlin (RAETZ/SEIFF 2003). Due to through these initiatives (RAETZ/SEIFF 2003). the positive experience, BIG has become the The most important project is located in model for several further founder and innova- Berlin’s Adlershof district; and this technolo- tion centers (TAMASY 1996). Especially after gy park today is not only the largest policy the reunification this instrument was exten- project in Berlin but, in terms of employment, sively used to overcome the economic crisis in the largest technology park in Germany. E. Kulke: The technology park Berlin-Adlershof 197

With the development of Adlershof in the Adlershof area was, at the beginning of 1990s, a change in the concept of spatial eco- the1990s , that three complementary elements nomic policy in Berlin can be observed; the – research and training at the university, re- idea of spatial proximity of complementary search in institutes, and private business activ- units (influenced by Porter’s idea of 1993 of ities – would serve as the cornerstones of a competitive advantages) was increasingly in- resulting dynamic, growing, technology-ori- fluencing the strategy. Existing innovation-oriented cluster (Fig. 2). In accordance with the ented approaches were completed by bringing GREMI innovative milieu approach, the start- together research, institutions, and enterprises ing point in Adlershof was the government- at a single location. In contrast to the network supported settlement of research institutions, and cluster models discussed earlier, the Berlin supplemented by the support of newly found- spatial economic policy since the 1990s can be ed enterprises. The concept was mainly stress- addressed as a heterogeneous proximity-orient- ing the spatial proximity factor, and, although ed approach; there is a tendency toward innov- not in an exact sense, the cluster idea. It was ative and technology-oriented firms, which is in hoped that being at one location would be the some way influenced by the idea of cluster and basis for the development of networks and for network formation. In this sense it can be a dynamic growth process. viewed as a top-down approach (FROMHOLD- EISEBITH/EISEBITH 2005), but a direct influence Historical background and long-term on network formation is missing. Instead the development path concept is following the idea of spatial proxim- The Adlershof site can look back on an almost ity of somewhat complementary units, which hundred-year long history of being orientated could, hopefully, lead to the development of toward modern technologies. This leads to the networks and competitive advantages. question as to whether a long-term development path can be observed. The development of networks, clusters, and milieus is often ex- Case study Berlin-Adlershof plained by the existence of a constant development path. The geographical industrializa- The main idea behind the development of the tion model (STORPER/WALKER 1989) describes Fig. 1: Location of founder/innovation centers and technology parks in Berlin

Source: RAETZ/SEIFF 2003 198 Zeitschrift für Wirtschaftsgeographie Heft 4 / 2008 the evolution of clusters. Later studies point Before World War II the Adlershof area was out the importance of the development of clearly oriented toward aeronautical research trust, of routines in the organizational struc- and production. The first airport for engine- ture, of trial-and-error learning processes, or powered flights in Germany was opened here of spin-off processes (e.g. BELUSSI 2006; in 1909. Thereafter the associated research fa- BOSCHMA/FRENKEN 2005;MOSSIG 2006; see cilities and numerous production companies also the case studies in COOKE/PICCALUGA settled in the area. The founding of the 2006). Looking at the situation in Adlershof Deutsche Versuchsanstalt für Luftfahrt DVL there seems to be a constant factor in the ori- (German Laboratory for Aviation) followed in entation toward modern technologies, but 1912. The DVL pursued scientifically based there have been several upheavals as a result aviation research in wind tunnels, testing fa- of changes in the economic framework. cilities, and factories, with over 500 employees in 1928 (WISTA 2006b). Alongside arose Fig. 2: Berlin-Adlershof zoning plan (2008) production companies for airplanes and their component parts. The Erich Rumpler Aircraft Construction Company (Flugzeugbau) and the Fokker Airplane Factory (Flugzeugwerke), for example, attained large market signifi- cance as producers of civilian and military aircraft, alongside many smaller operations (the so-called airplane builders, „Aeroplanbauer”).

These activities came to an end after World War II, and an extensive change in the scientific and economic orientation of the site took place. Aircraft production was no longer possible and most production-oriented activities ended. In research a strong shift towards natural sciences prevailed. The German Democ- ratic Republic (GDR) housed the facilities of its German Academy of Sciences in the former DVL buildings from the early 1950s onward. By 1989, fifteen natural science institutes, primarily in the areas of physics and chemistry, had arisen in the area. Approximately 5,400 persons were employed in these institutes at the time of the German reunification in 1989 (HU 2005). Also on the premises was the Se- curity Regiment Felix Dzerzhinsky of the Na- tional People’s Army (Nationale Volksarmee), as well as studios, production offices, and broadcasting facilities belonging to GDR television broadcasting institutions.

With the reunification came renewed upheaval in Adlershof. The military facilities were immediately closed, the television broadcasting facilities dissolved and privatized. The breakup of the erstwhile German Academy of Sciences facilities took a more differentiated path. Its sections that had been strongly oriented toward industrial development were for the most part closed, some of them re-emerg- ing as an assortment of newly founded, private-sector enterprises. Those that had focused on basic research became the basis for E. Kulke: The technology park Berlin-Adlershof 199

the construction of state-financed research in- their expression in the occasional use of some stitutes belonging to the Max Planck, Fraun- material infrastructure or in the private ex- hofer, or Leibniz scientific societies. The change of information. But they are relatively founding of new, technologically oriented pri- limited, so with the establishment of the tech- vate enterprises (particularly in technology nology park after the reunification a new start parks) was supported by the state. And, in was made; a developmental continuity can 1991, the official decision was made to relo- therefore be examined only from the 1990s cate Humboldt University’s natural sciences onward. institutes from the central district of Berlin- Mitte to Berlin-Adlershof. Development since German reunification and current structures Looking at the history of the location, no long- From the very beginning of the development term development path in the sense of an evo- of the technology park, spatial economic poli- lution of specific areas of innovation and pro- cy followed the idea of establishing three duction can be observed (cf. BOSCHMA/ complementary cornerstones: enterprises, re- FRENKEN 2005). Instead there have been sev- search institutes, and university institutes. eral upheavals with an accompanying total They were settled in the area step by step. The change in technological orientation. The tech- process began immediately after German re- nological orientation and the production struc- unification with the establishment and support ture before WWII and the research orientation of start-up companies – some of them spin- in the GDR period were completely different. offs of the former German Academy of Sci- And with the German reunification a profound ences – and with the reorganization of the re- change occurred once again. Production and search institutes. The resettlement of universi- private enterprises became a new element at ty institutes of Humboldt University needed the location, several units of the academy of several years, the first unit being established sciences were closed and new research units not until 1998 in Adlershof. For the organiza- with a different orientation were established. tion of the park’s development of the park the The only seedbed for constant relations were Berlin government established a public-pri- the still-existing personal ties between the vate company called WISTA. WISTA is the groups of persons who before had worked in operating company for the park; it was re- the academy of sciences and who afterwards sponsible for the construction, lease and oper- either established private enterprises or were ation of the technology and incubator centers. employed in new research institutes. These Today it offers rental space and maintains linkages still exist to a certain extent and find properties for sale; it promotes research and Tab. 1: Business and employment in Berlin-Adlershof (2006)

Number of Number of Turnover/budget units employees (Mio. € in 2006) (end 2006) (end 2006) Technology park Businesses 400 4,279 440 Research institutes 12 1,508 142 University institutes 6 865 55 (plus 6.434 students) Sub-total 418 6,652 637 (plus 6,434 students)

Media City and industrial/service park Businesses 138 1,432 188 Businesses 201 4,261 491 Total 757 12,345 1316 (plus 6,434 students)

Source: WISTA 2007 200 Zeitschrift für Wirtschaftsgeographie Heft 4 / 2008 business networks as well as national and in- lishment of the Innovation and Business Incu- ternational joint ventures. And it is responsible bation Center Berlin (Innovations- und Grün- for public relations, marketing, and sales ac- der-Zentrum Berlin-Adlershof, IGZ), in Adler- tivities for the entire development area (cf. shof in September 1991. The center began in a WISTA 2007). Today there are more than 750 plain, one-story building with an area of enterprises with more than 12,000 employees around 1,000m2, on which five companies working in Adlershof, and more than 6,400 and their fourteen employees settled (data, aöso students are educated there (Tab. 1). in the following, based on RAETZ 2005). Later a new building with an area of 11,800 m2 was Private enterprises and business centers erected and then extended in 1997 with the The policy for private enterprises was guided Ost-West-Zentrum (OWZ), comprising an ad- mainly by the idea of facilitating their devel- ditional 6,600 m2. The OWZ focuses primari- opment by offering well-equipped spaces at ly on international businesses. Both the IGZ low rents. On the Adlershof site locations for and OWZ offer the infrastructures typical for private enterprises were available on free the Technology and Start-Up Center (TGZ), spaces, in some existing old buildings – es- e.g. secretary services, conference rooms (cf. pecially in the so-called industrial park outside TAMASY 1996), and render manifold human the core area of the technology park (Tab. 1) – capital-intensive services (counsel in all busi- and in newly established business centers. The ness matters; contact to distributors, financial industrial park was mainly designated for clas- institutions, buyers, and service providers; sical manufacturing and service activities, team-building) in order to minimize the initial while the technology park was designed to difficulties of younger businesses. In the peri- host innovative firms and research units (Fig. od from 1991 to July 2005, 297 businesses 2). Today space is available the technology moved there, and only fifteen insolvencies oc- park in two incubation centers, four technolo- curred (Fig. 3), which can be assessed as a gy centers and in several sites for lease. success. According to the TGZ conception, businesses are to stay in these centers for only The first step for the foundation of new tech- a limited amount of time (i.e. during their ini- nologically oriented enterprises was the estab- tial phases) and then, with further growth, to

Fig. 3: Performance of IGZ/OWZ*1991 - 2005

*Innovation and Business Incubation Center Berlin/East-West-Center Source: RAETZ 2005 E. Kulke: The technology park Berlin-Adlershof 201 move to other locations. Up to now, 179 busi- and optical technologies, the environmental nesses have left the center in this manner. The technology center, the center for information technology center approach has no direct ori- technology, and the center for materials re- entation towards special areas of activities in search. The subject area focuses of the busi- the sense of a cluster formation. There is an nesses display a clear orientation toward iden- evaluation process for all units that want to tified fields of technological research in Adler- establish activities in the center, but it is main- shof, that is, toward information technology, ly based on whether the start-ups possess real- materials research, photonics, and environ- istic future perspectives (usually documented mental technology (Tab. 2). by a convincing business plan) and not on their fields of activities corresponding to the Altogether 400 private businesses with 4,279 profile of the technology park. This approach employees and a revenue of 440 Mio. € were is typical for innovation centers (cf. settled in all six centers and other locations in BEHRENDT 1996;TAMASY 1996) but also illu- technology park in the year 2006 (data accord- minates the general policy approach of Berlin, ing to WISTA 2007, also in the following, cf. which tends to follow the idea of spatial prox- Fig. 3). Another 201 businesses with 4,261 imity rather than interlinked production ac- employees are located in the industrial park. tivities. Looking back on the development since the beginning, the private economy seems to have In contrast to the IGZ/OWZ, the four other been quite successful. There has been a con- business centers offer permanently rentable tinuous increase in the number of enterprises spaces and are clearly oriented toward busi- and in the number of employees in the tech- nesses with a focus on a particular subject nology park (Fig. 4). And even though the av- area. Their establishment may be understood erage size of the enterprises (2006: 16.5 em- as a policy-led idea to generated competence ployees per unit) still is relatively small, they in fields of technologies that seem to have seem to be successful in the market. Indicators good prospects in the future. These subject ar- for this, aside from the low insolvency rate, in- eas are chosen based on the main subject areas clude a constant reduction in the share of state of the scientific and technological profile of subsidies to these enterprises (Tab. 3) and a Adlershof, and correspond with the present strong increase in turnover. In 1997 the share research facilities, above all in the areas of of state subsidies was 26.4 % of the turnover, physics, chemistry, and information technolo- but it had been reduced to 3.4 % by 2007. In gy. This is the case for the center for photonic the same period the turnover increased by

Tab. 2: Technology areas of technology park enterprises

Businesses Employees Turnover abs. % abs. % (Mio. €) %

Information and media technology 78 19.5 791 18.5 84,356 20.1

Materials and microsystems technology 39 9.8 578 13.5 44,564 10.6

Photonic and optic technology 54 13.5 867 20.3 97,503 23.3

Environmental, biological, and energy technology 65 16.3 734 16.9 43,848 10.4

Others and service industry 164 41.0 1319 30.8 149,843 35.7

Total 400 100.0 4279 100.0 420,114 100.0

Source: WISTA 2007 202 Zeitschrift für Wirtschaftsgeographie Heft 4 / 2008

86.4 % from 257 Mio. € to 479 Mio. €. From technology fields. Especially the orientation these indicators one can get the impression towards materials research, photonics, and in- that the enterprises have developed an inde- formation technology is strong; it finds its ex- pendent economic base. However, up to now, pression both in research institutes/university no very large firm has settled in Adlershof or institutes and private enterprises (Tab. 2). developed there. This is on the one hand not Contributions to material research and pho- very unusual, as the relocation potential of tonics come for the university institutes of large firms is generally low and small busi- physics and chemistry, from the electron ac- nesses do not develop so quickly. On the other celerator BESSY II (Berlin Electron Synchro- hand, cluster approaches such as that of geo- tron Particle Accelerator), the Institut für graphical industrialization stress the meaning Kristallzüchtung, IKZ (Institute for Crystal of “flagship enterprises” for growth dynamic Growth), and the Institut für Angewandte and competitive ability of a site, so that these Chemie, Adlershof, ACA (Adlershof Institute circumstances can be viewed as a still-existing for Applied Chemistry). Developments in the weakness in development. area of information technology are done in the university institutes of mathematics and infor- Research and university institutes mation sciences and the Fraunhofer Institut The development of research and university für Rechnerarchitektur und Softwaretechnik, institutes was strongly influenced by the idea FIRST (Fraunhofer Institute for Computer Ar- to establish units doing research in areas most chitecture and Software Technology). Today pertinent to the technology park. The spatial twelve research facilities are located in Adler- proximity of these units was hoped to generate shof, employing approximately 1.500 persons the exchange of ideas between research and in 2006 (Tab. 1). Their basic financing is de- production and the transfer of personnel (in- rived from state and federal budgets in the cluding students). But once again the basic ap- amount of 101 Mio. €; the facilities them- proach was spatial proximity; additional sub- selves raise additional third-party funds (from sidies for cooperation were not given. Being research foundations and contracts) in the publicly financed (usually half by the state of amount of 41 Mio. €. Berlin and half by the federal government of Germany), the influence of spatial policy on Although the first decisions about relocating the locational decision was much stronger Humboldt University’s natural sciences insti- than on enterprises of the private sector. Based tutes were made in 1991, and the concrete on concepts evaluated by the German Re- planning, including bidding for the construc- search Council (Wissenschaftsrat), units were tion work, began in 1993, the move was con- established which partly contributed to the tested for quite some time, actually until its

Fig. 4: Berlin-Adlershof: number of enterprises and employees 1996-2006

Source: WISTA 2007 E. Kulke: The technology park Berlin-Adlershof 203 completion. Those belonging to the university in 2001) and physics (move in 2003) were were aware of the absolute necessity of erected, while two former barracks were acquiring adequate facilities for the natural arranged for psychology and geography sciences institutes. In Berlin’s Mitte district, (move in 2003). The opening of the Erwin these were in buildings located in substantial Schrödinger Center in 2003 was of central im- distances from one another, which were in portance for the site as a whole and also for partly dilapidated condition, possessed no the realization of an innovative milieu. It con- suitable equipment for modern research, and, tained a central library, computer infrastruc- because of their spatial separation, impeding ture, and course rooms – and, with its cafete- intra- and interdisciplinary projects. All the ria and bookstore, it also offered the possibili- same, a great deal of resistance arose in the ty for actors to meet. Today in Adlershof there scientists affected by the decision to relocate are six university institutes with around 100 to the new site in Adlershof, which was almost professorships and over 800 employees; ap- fifteen kilometers away, southeast of Berlin, in proximately 6,400 students are presently re- what was subjectively viewed as a desert en- ceiving training there (Fig. 2, Tab. 1). vironment. It was publicly argued that, under these conditions, networks with vital research The description of the Humboldt University (e.g. joint projects) and training (e.g. with ed- resettlement process displays the individual ucational courses of study) in the economic restraints against the relocation. And even af- and social sciences institutions and with me- ter being relocated to the technology park dicine could not be maintained. Oftentimes Adlershof, cooperation with the private sector however, personal reasons, such as the consid- is often less recognized then basic research; erably longer commuting time or the loss of a for example the distribution of the university site in the historically important area along funds is based on third-party grants in basic Berlin’s famous lane Unter den Linden, research and does not consider cooperations played some role. The relocation was ulti- with private enterprises. In general there are mately decided in 1997 with a vote from the no subsidies given for network formation in scientific counsel and the approval of the fi- Adlershof, even though spatial economic pol- nancial plan by the building commission. icy is expecting these to develop. This could be addressed as a weakness in recent spatial The Institute of Information Science had to re- economic policy. Policy is more and more led locate as early as 1998, more or less compul- by the idea of cluster formation and network sorily (the lease at the old site ran out; the development (cf. FROMHOLD-EISEBITH/EISE- Berlin senate refused renewal), not to a new BITH 2005; STERNBERG/SCHÄTZL 2004), but the building, but to an empty office building in use of formal or informal systems of network Adlershof. Because of this, the basic setup of management is still limited. the relocation plan, which had actually been written with the intended timeframe of 2001 to 2007, completely changed. As relocation Tab. 3: State subsidies for private enterprises stretched out over the long timeframe from (% of turnover) 1998 to 2007, it threatened to hamper the working ability of those involved consider- 1997 26.4 ably. But, with the formulation of a new con- 1998 16.3 cept – the so-called acceleration plan – the 1999 13.5 planned timeframe of 1998 to 2003 was successfully shortened via the regrouping of in- 2000 11.0 vestments (e.g. barrack renovation rather than 2001 8.4 new construction, unification of information 2002 9.5 science and mathematics, reallocation of the 2003 8.7 annual governmental investment funds). De- 2004 5.8 spite considerable resistance (due to having to 2005 5.3 relinquish demands for a suitable new build- 2006 4.6 ing) the Institute of Mathematics followed in 2000 to the same building, which had actually 2007 3.4 been intended for service industries. New buildings for the institutes of chemistry (move Source: WISTA 2008 204 Zeitschrift für Wirtschaftsgeographie Heft 4 / 2008

Economic performance and network plementary units from the private sector, development in Adlershof from research, and from public/private insti- There is no general checklist available for the tutions are located within the area. This as- evaluation of clusters, but it is widely accept- pect analyzes the horizontal elements of a ed that not every spatial concentration of ac- competitive cluster-like concentration of activities is really a network-based cluster. For tivities and it gives an impression of the co- the existence of cluster formations, there have operation potential and the possibilities of to be intensive linkages between the units, observing new competitor developments. which have to receive innovative and competitive advantages from them. KRÄTKE/SCHE- – Second, with a more statistical viewpoint, UPLEIN (2001, 57) point out the mismatch be- the economic performance and quantitative tween theoretical approaches to and empirical success should be analyzed. This element analysis of cluster formation. They argue – tries to illuminate the overall economic based on considerations of REHFELD (1999) – strength of the location, which in the sense that locations with concentrations of success- of KRÄTKE/SCHEUPLEIN (2001) is often used ful units are often explained by the existence as an argument for the existence of a cluster. of linkages (without really analyzing them) and afterwards it is concluded that the success – Third, there should be a qualitative impres- comes from the spatial proximity. But for the sion as to what extent the single units are re- identification of cluster-like structures, an ally linked with each other within a network, analysis of linkages has to be done based on a and whether these immaterial connections survey of the networks of enterprises (e.g. information transfer) are contributing (KRÄTKE/SCHEUPLEIN 2001, 61 f.). The dimen- to the innovative and competitive strength of sions of cluster formation have been discussed the location. This element considers the ver- intensively by MALMBERG/MASKELL (2001) tical and institutional aspects of cluster for- and have been completed by BATHELT/GLÜCK- mations. LER (2002, 212 f.). They assert that a cluster is characterrized by four dimensions; horizontal, The analysis is based on different information vertical, institutional, external. The horizontal sources that were collected in various ways. dimension describes co-presence at a location, Secondary statistics are available in WISTA’s which facilitates the observation of competi- annual data publication; the annual report doc- tors, the comparison of products and process- uments data from all the enterprises and insti- es, and thereby the improvement of individual tutions in the technology park and thereby solutions. The vertical dimension is based on provides information about totals. The data the input-output linkages of complementary collection is annually done in the same way by units. The institutional dimension explains an independent research institute hired by that formal and informal institutions, regulato- WISTA, and can therefore be used as quite a ry systems, trust, and reliability promote com- reliable base for the analysis of long-term de- petitive strength. The external dimension ar- velopments. Primary statistics are based on gues that a cluster has to have additional ex- several smaller case studies that were done in ternal links to overcome the risk of the Department of Geography at Humboldt overembeddedness or lock-in. University; they focus especially on linkages between different groups of actors in the loca- In the following analysis three elements are tion, and thereby enlighten aspects of the net- discussed that are linked with the idea of clus- work formation in Adlershof. ter dimensions but are considering the special situation of policy-led developments (cf. Complementary units FROMHOLD-EISEBITH/EISEBITH 2005; STERN- The discussion above has shown that up to BERG/SCHÄTZL 2004). For the evaluation of the now all of the important components of an in- location Adlershof, and to answer the question novative cluster – research, education, institu- whether it already possesses some cluster-like tions, enterprises – have been established in structures, there seem to be at least three rele- Adlershof. It is not only that the three corner- vant aspects: stones of the private sector, the university research and education, and the research insti- – First, from a mere locational viewpoint it tutes are located there. In addition, important should be documented as to whether com- institutions for the management of the park E. Kulke: The technology park Berlin-Adlershof 205

(such as WISTA and Initiativgemeinschaft location (1.29), the choice of commercial Außeruniversitärer Forschungseinrichtungen space (1.30), the availability of a qualified in Adlershof e.V., IGAFA) and for the start-up work force (1.45), and the possibilities for co- firms (like IGZ and OWZ) have been estab- operation (1.45). Especially the possibility of lished. The Schrödinger Center opens possi- personnel transfer between university and bilities for information and communication, economy is a clear advantage of the location and there are retail facilities and restaurants and of formations of this type. And the posi- for additional activities. tive evaluation of cooperation potential gives the impression that the existing spatial prox- It also appears that a limited number of tech- imity is already used as a basis for network nological fields have successfully developed formation. In contrast to this, elements in the (Tab. 2) in which, due to the number of busi- surrounding environment such as the avail- nesses and scientific institutes involved, a cri- ability of parking (2.04) and free-time or sport tical mass for the development of a cluster has activities (2.29) were rated less favorably. The been reached. This includes above all the areas results are in accordance with the assumptions of photonics and optical technologies, infor- of the models, namely that factors such as im- mation and media technology, and materials age, personnel transfer, and cooperation con- research and technology. The currently active stitute a substantial element in the develop- businesses in these fields find complementary ment of an innovative milieu. facilities on site (physics, information technology, chemistry) with which scientific and per- Networks sonnel transfer is possible. The area of envi- The results discussed up to this point suggest ronmental, biological, and energy technology, the plausibility of the creation of an innovative in which less scientific activities on site can be milieu, but do not yet document the actual de- found, is proliferating somewhat less and also velopment of networks. Reliable data to this with below-average growth rates (WISTA purpose is exceedingly difficult to come by, as 2006); this is probably derived from the fact networks are based for the most part on per- that the Institute of Biology was actually the sonal and informal contacts between actors. only Humboldt University natural sciences in- Formalized contacts can be documented be- stitute not to be relocated to Adlershof. These tween Humboldt University institutes and the developments point to a connection between applied research facilities taking the form of science and economy. shared appointed employees (so-called s-professorships), cooperation agreements, and Performance and success common projects funded by third parties (Fig. The Adlershof growth dynamic speaks for the 5; based on information from the HU). The ab- decisive competitive ability of its private busi- solute number and intensity of these shared ness activities. The number of businesses and projects has clearly increased in the last years. their employees, as well as the amount of rev- So it was that eight shared appointments (out enue earned over the last decade, are a testa- of fifteen s-professorships in total) of the uni- ment to a sustained growth process, not only versity institutes came from scientific facili- in the centers for new businesses but also in ties in Adlershof and were newly established the remaining area of the park. The low rate of during the last decade. In six of twelve applied insolvency (Fig. 3) and the decreasing share of research facilities, contractually secured work subsidies (Tab. 3) also prove a successful de- cooperation has arisen. Alongside this, there velopment. In the publicly aided scientific fa- are common research projects in special re- cilities, the development dynamic cannot be search areas, research groups, and post-gradu- analyzed in such a manner, as their financial ate programs. A graduate school (for doctoral basis is strongly dependent on the current state candidates) is currently being established as a of the public budget. cooperation between the university institutes and applied research facilities. It can therefore That the development dynamic of businesses be asserted that strong networks between sci- in Adlershof does not alone depend on internal entific facilities exist at this point. business factors was shown in a 2003 poll of businesses contracted by WISTA (WISTA Indications of the degree of network building 2004). On a qualitative scale (1 to 3), busi- between the university and businesses were nesses very positively rated the image of the given in a 2004 field study in economic geog- 206 Zeitschrift für Wirtschaftsgeographie Heft 4 / 2008 raphy (the following data is given according to ness than basic research in natural sciences. the results of the study; cf. WEBER-BLEYLE The research also proves a discussed pre- 2008), which comprised a sample of 86 busi- sumption of the model approaches, that con- nesses (total during this time 217; return rate tacts between businesses are of special impor- 39.6 %). Of the businesses asked, 19.2 % were tance for business success; on a 5-step scale of spin-offs of the university and around half pos- importance (1 = excellent, 5 = bad), those sessed contacts with the university in various asked valued these contacts on average at 1.9, forms (e.g. personnel, projects). The share of or of more than average importance. personnel formerly trained at Berlin universities is still comparatively low (only 17.1 %); To identify the importance of networks for the this may however be a result of the founding economic success of the enterprises, a statisti- of the businesses before the time of the uni- cal cluster analysis was done (WEBER-BLEYLE versity relocation. 2008, 71 f). It evaluated the connection between business success of the interviewed en- However, the businesses in Adlershof value terprises (measured by the growth of the num- their university contacts less than their rela- ber of employees) and the type and intensity tionships with other business on location and of networks of these units. The results show with the Technical University of Berlin (TU). the highest growth rates for businesses with This is evidence of the known phenomenon intensive contacts to other businesses in that oftentimes developmental, that is techni- Adlershof, and with moderately pronounced cal solutions such as those offered by the TU, relationships to the university. Businesses that are more important for the success of a busi- focused heavily on the university rather than on other businesses were somewhat less successful in personnel growth. The least growth Fig. 5: Formalized contacts* between Hum- rates were displayed by businesses with no in- boldt University institutes and applied teractions with the university or other busi- research institutes in Berin-Adlershof nesses. The results confirm the general pre- (2008) sumption of the importance of networks; but it seems that linkages with other firms (to find technical solutions) are more important for economic success than with university research (doing more basic research).

Overall the results as far as network creation show that Adlershof forms an intensively in- terconnected scientific cluster. The university institutes have continued with their contacts to Berlin-Mitte and have developed in addition new networks with research institutes in Berlin-Adlershof. These contacts have been developing relatively fast and have reached a high degree of intensity. Networks between university and enterprises are developing more slowly. But an economic cluster – with contacts between the enterprises – has also developed with positive effect on the economic success of the businesses connected to it. The contacts that have arisen between science and business are up to this point less distinct but existent; one should be careful to note in this analysis that the university institutes only settled relatively recently in Adlershof.

*shared professorships, cooperation agreements, The discussion of the pathway of develop- ongoing shared projects ment and characteristics of the science and Source: Information of Humboldt University business network in Adlershof gives the im- E. Kulke: The technology park Berlin-Adlershof 207 pression that the site fulfils several elements References of an innovative cluster. Not only complemen- ARL (Akademie für Raumforschung und Landespla- tary units are existing, but in addition the lo- nung)(2006): Wie hell strahlen Leuchttürme? Anmer- cations shows a highly developmental dynam- kungen zur Clusterpolitik im ländlichen Raum. In: ic. And the network analysis documents that ARL, 3, 1-4. those units that are more linked to other units ASHEIM, B./COOKE, P./MARTIN, R. (Ed.) (2006): Clus- at the location display better economic perfor- ters and regional development. Milton Park/New York. mance. AYDALOT, P. (Ed.) (1986). Millieux innovateurs en Eu- rope. Paris. BATHELT, H. (1992). Erklärungsansätze industrieller Conclusion Standortentscheidungen. In: Geographische Zeitschrift, 80, 195-213. The question as to whether regional economic BATHELT, H. (1998). Regionales Wachstum in vernetz- policy may successfully develop an innova- ten Strukturen: Konzeptioneller Überblick und kritische tive milieu by means of initial investments Bewertung des Phänomens „Drittes Italien“. Die Erde, cannot be conclusively answered at present (129)3, 247-271. (see the discussion of COOKE 2002, 156 f.). BATHELT, H./DEPNER, H. (2003): Innovation, Institution The overwhelming majority of indicators for und Region: Zur Diskussion über nationale und regio- Adlershof point to the success of the project. nale Innovationssysteme. In: Erdkunde 57, 126-143. The economic dynamic of the businesses is BATHELT, H./GLÜCKLER, J. (2002): Wirtschaftsgeogra- apparent according to revenue and employee phie. Stuttgart. growth rates, as well as low insolvency rates BEHRENDT, H. (1996): Wirkungsanalyse von Technolo- and decreased shares of subsidies, now re- gie- und Gründerzentren in Westdeutschland. Heidel- duced to a very low level. And businesses that berg. are networked with other businesses operate BELUSSI, F. (2006): In search of a theory of spatial clus- especially successfully. These indicators give tering. In: Asheim, B./Cooke, P./Martin, R. the impression that the proximity-oriented ap- (Ed.)(2006): Clusters and regional development. Milton proach of spatial economic policy has been Park/New York, 69-89. successful. Similarly, networks between the BOSCHMA, R.A./FRENKEN, K. (2005): Why is economic scientific facilities are developing. The site geography not an evolutionary science? Towards an has also been noticed by the public, as the nu- evolutionary economic geography. Utrecht. (Papers in merous existing articles about Adlershof Evolutionary economic geography. show, and the assessment of the site’s image COOKE, P. (2002): Knowledge economies – clusters and the entities settled there is clearly quite learning and cooperative advantages. London/New York. positive. But the networks between science COOKE, P./PICCALUGA, A. (2006): Regional develop- and business are as yet not very strongly pro- ment in the knowledge economy. Milton Park/New nounced, so that the impression almost arises York. that a single site has been settled by both a sci- DANNENBERG, P. (2007): Cluster-Strukturen in landwirt- ence park and a business park. The large flag- schaftlichen Wertschöpfungsketten in Ostdeutschland ships of economic development are also still und Polen. Münster/Hamburg/London. (Wirtschafts- missing. However, when assessing these still- geographie, Band 43). existing limitations, the as yet comparably FESER, E. (1998): Old and new theories of industry clus- young development phase of the site should be ters. In: Steiner, M. (Ed.): Clusters and regional spe- considered. In all, the Adlershof project can be cialisation – on geography, technology and networks. classified as a clear success in regional eco- London, 18-40. nomic policy. However, with respect to the FLOETING, H. (2008): Clusterinitiativen und Netzwerke – Handlungsfelder der Wirtschaftsförderung. In: Difu- scientific question of the possibility of gener- Berichte, 2, 8-9. ating an innovative milieu with a long-term growth dynamic and a high rate of networks, a FROMHOLD-EISEBITH, M. (1999): Das „kreative Milieu“ – nur theoretisches Konzept oder Instrument der Re- conclusive answer is not yet possible. gionalentwicklung? In: Raumforschung und Raumord- nung, 2/3, 168-175. FROMHOLD-EISEBITH, M. (2000). Technologieregionen in Asiens Newly Industrialized Countries. Münster/ Note Hamburg/London. (Wirtschaftsgeographie, Band 18). FROMHOLD-EISEBITH, M./EISEBITH, G. (2005): How to Mrs. Rose Knudsen deserves thanks for linguistic editing. institutionalize innovative clusters? Comparing explecit 208 Zeitschrift für Wirtschaftsgeographie Heft 4 / 2008 top-down and bottom-up approaches. In: Research pol- PORTER, M. E. (1998): Clusters and the new economics icy, 34, 1250-1268. of competition. In: Harvard Business Review, (76)6, GERTLER, M. (2001): Best practice? Geography, learn- 77-90. ing and the institutional limits of strong convergence. PORTER, M. E. (1999): Unternehmen können von regio- In: Journal of Economic Geography, 1, S. 5-26. nalen Vernetzungen profitieren. In: Harvard Business GORDON, I.R./MCCANN, P. (2000): Industrial clusters: Manager 3, 51-63. complexes, agglomeration and/or social networks? In: RAETZ, G. (2005). Materialien zum Innovations- und Urban Studies 37, 513-532. Gründerzentrum Berlin-Adlershof. ROTZ G , R. (1996). Kreative Milieus und Netzwerke als RAETZ, G. /SEIFF, F. (2003): 20 Jahre Berliner Innovati- Triebkräfte der Wirtschaft: Ansprüche, Hoffnungen und ons- und Gründerzentrum – eine Bilanz. Berlin. Wirklichkeit. Bayreuth, 65-84. (Arbeitsmaterialien zur Raumordnung und Raumplanung, Heft 153). REHFELD, D. (1999): Produktionscluster. Konzeption, Analyse und Strategien für eine Neuorientierung der re- GUNDEL, S./LUTMANN, J. (2008): Die regionalwirt- schaftlichen Effekte von TechnologieZentrum und gionalen Strukturpolitik. München. TechnologiePark Dortmund. Münster. SCHAMP, E. W. (2007): Denkstile in der deutschen Wirt- HU (Humboldt-Universität zu Berlin (Ed.)(2005). Ad- schaftsgeographie. In: Zeitschrift für Wirtschaftsgeo- lershof Campus. Berlin. graphie, (51)3-4, 238-252. JOHNSTON, R.J./TAYLOR, P.J./WATTS, M.J. (Eds.)(1995). STERNBERG, R./SCHÄTZL, L. (2004): Clusteransätze in Geographies of global change. Oxford. der regionalen Wirtschaftsförderung. In: Zeitschrift für Wirtschaftsgeographie, (48)3-4, 164-181. KLINE, S.J./ROSENBERG, N. (1986): An overview of innovation. In: Landau, R./Rosenberg, N. (Ed.): The pos- STORPER, M./WALKER, R. (1989). The capitalist imperati- itive sum strategy – harnessing technology for eco- ve – territory, technology, and industrial growth. Oxford. nomic growth. Washington, 275-307. TAMASY, C. (1996). Technologie- und Gründerzentren KRÄTKE, S./SCHEUPLEIN, C. (2001): Produktionscluster in Ostdeutschland – eine regionalwirtschaftliche Ana- in Ostdeutschland. Hamburg. lyse. Münster/Hamburg/London. (Wirtschaftsgeogra- KULKE, E. (2003): Berlin – German Capital and Glo- phie, Band 10). bal City? In: Die Erde, (134)3, 219-233. WEBER-BLEYLE, M. (2008). Netzwerke und Wis- KULKE, E. (2008). Wirtschaftsgeographie. Pader- senstransfer in Technologieparks: Eine Bewertung der born/München/Wien/Zürich. Beziehungen zwischen universitären Forschungsein- MALMBERG, A./MASKELL, P. (2001): The elusive con- richtungen und Unternehmen aus Unternehmersicht an cept of localization economies. Towards a knowledge- den Beispielen Berlin-Adlershof und Urbana-Cham- based theory of spatial clustering. New York. (Paper for paign. Berlin. (Magisterarbeit, HU-Berlin). the AAG Annual Conference). WISTA (2004) (Ed.). Ergebnisse der Jahresumfrage MOSSIG, I. (2006): Netzwerke der Kulturökonomie. 2003. Berlin. Bielefeld. WISTA (2006a) (Ed.). Bericht über Adlershof. Berlin. MOSSIG, I./KLEIN, J. (2003): Das Produktionscluster der WISTA (2006b) (Ed.). Adlershof Magazin/ILA 2006. optischen Industrie im Raum Wetzlar. In: Raumfor- Berlin. schung und Raumordnung 61, p. 237-251. WISTA (2007, 2008) (Ed.): Bericht über Adlershof. PORTER, M. E. (1993): Nationale Wettbewerbsvorteile. Berlin. Wien.