56 Clusters and Competitive Advantage advantage created. The main steps in the methodology employed are sum- marized in Figure 3.4. 18 The qualitative analysis of the case studies is based on a review of the litera- ture, secondary data analysis and in-depth interviews with cluster participants. The review of the literature provided information on the clusters’ historical development and their main local players, as well as recent changes and challenges faced by the cluster members. Academic journals, government publications, trade magazines, trade associations and local newspapers were amongst the major sources. Secondary data on given clusters provided such information as a cluster’s size and growth over the years, its employment specialization, the distribution of each firm in the cluster by number of employees, the type and distribution of occupations in the cluster, and the cluster’s export and productivity performance. A total of 72 in-depth interviews and numerous shorter ones were conducted. The interviewees were managers of firms, government officials and other industry experts from organizations linked to the cluster (such as universities, sectoral associations, research centres and institutes). The inter- views were semistructured and took 1.5–2 hours, although some took as long as 4–5 hours. The respondents were asked for their overall evaluation of the cluster, the reasons for geographic concentration and the possible Analyse industrial statistics Analyse trade statistics Identify internationally competitive industries Seek a statistical link between geographic concentration and competitiveness Select and conduct cluster case studies Clarify the link between geographic concentration and competitiveness Identify geographically concentrated industries Figure 3.4 Methodological process followed in the study Industrial Clusters in Turkey 57 contribution of attributes of the location to the international competitiveness of the cluster. The interviews were then transcribed and the main themes were identified. This information served as the basis for the reports prepared on each of the cases studied. The information derived from the interviews, the literature review and the secondary data were used to construct a cluster map for each cluster. 19 Geographic clusters and competitiveness: which cases to study? To understand the complex role that the attributes of the local environment play in shaping the competitive advantages of firms, a detailed analysis of specific clusters is needed. Only a full analysis of the historical development and current structure of a cluster can reveal the true nature of the relationship between geographic clustering and competitiveness. This challenging task will be the subject of the next four chapters. The clusters chosen were the furniture cluster in Ankara, the towel and bathrobe cluster in Denizli, the machine-made carpet cluster in Gaziantep and the leather clothing cluster in Istanbul. The furniture cluster in Ankara served as a pilot study, since my residential proximity to this cluster made it much easier to go back and forth and reinterview many of the participants, and to revise the method of investigation accordingly. Although the cluster is one of the leading areas of economic activity in Ankara and is highly concentrated geographically, its degree of international competitiveness has remained low. This is interesting in that furniture making is a traditional, labour-intensive industry and in this respect it resembles many of the competitive clusters. Given that there is an over- emphasis on competitive clusters in the literature and the relatively less competitive ones are understudied, the furniture industry in Ankara makes an ideal case for detailed study. The second case study was of the towel and bathrobe cluster in the Aegean city of Denizli. This choice was well justified given that this cluster was arguably the one that brought the concept of geographic clusters to the attention of academics and policy makers in Turkey. Its success in the world market has attracted considerable attention. A comprehensive analysis of that success and its resilience in the face of recent challenges at home and abroad can substantially improve our understanding of the relationship between clustering and competitive advantage. The process Denizli has gone through is also different from that in other provinces with significant textile industries, such as Tekirdag which owes most of its development impetus to its geographical proximity to Istanbul and Western Europe. Moreover the public sector does not have much of a presence in Denizli, which has forced it to rely on its own capabilities. Finally, the textile-town of Denizli provides us with a natural and unique laboratory to investige the 58 Clusters and Competitive Advantage geographic concentration of specific industries given that the impact of localization economies is very clear in this city in the absence of urbaniza- tion economies. The south-eastern city of Gaziantep’s success in several areas, including machine tools, carpets, yarn and pasta, has also attracted considerable atten- tion. Of these the carpet cluster, with its high geographic concentration and rapidly improving export performance, deserves detailed analysis for two reasons. First a case study of this cluster will serve to reinforce the findings obtained from the study of the equally successful towel and bathrobe cluster in Denizli. Second, Gaziantep is not among the major historical carpet-weaving centres in Turkey. This adds another dimension to the study in the form of an exploration of why and how the carpet cluster came to develop in this city rather than somewhere else (for instance in one of the traditional carpet- weaving centres in Anatolia). An additional reason for studying this cluster is that it is located in south-eastern Turkey, an area otherwise associated with poor economic development. Thus in addition to contributing to our understanding of locational factors in competitiveness, this cluster study is likely to provide valuable insights into regional economic development more generally. Finally, the leather clothing cluster in Istanbul provides an opportunity to investigate a very interesting set of relations. First, the nature of its link to international markets is different from that of many other clusters, given the important role assumed by the so-called ‘luggage trade’ in this cluster. Second, in response to recent changes in the business environment there has been a shift in the strategic orientation of Istanbul’s leather producers, which provides interesting insights into the changing patterns of competitive advantage in clusters. Third, within Istanbul there are two separate concen- trations of activity that are related to the cluster: production units and networks are concentrated in Tuzla on the Anatolian shore, and the sales and marketing functions are concentrated in the Laleli/Zeytinburnu districts in the historical, European section of the city. Such an interesting network of relations merits special consideration. The analysis of this cluster in Chapter 7 also contains a discussion of the Istanbul economy in more general terms, providing the reader with knowledge of the range of activities that prevail in the economic capital of Turkey. In addition to the detailed studies outlined above, this book briefly discusses other clusters that exist in the cities examined, including the construction cluster in Ankara, the footwear, cutlery and marble clusters in Denizli, the machinery cluster in Gaziantep, and the footwear, motion picture and jewellery clusters in Istanbul. All the case studies presented in the following chapters have the same structure. Each chapter starts with an overview of the cluster, outlining its general characteristics and importance. Next the origins of the cluster and Industrial Clusters in Turkey 59 the key events in its history are presented. This is followed by an investigation of the cluster’s sources of competitive advantage/disadvantage, and then the specific reasons for its geographic concentration in the area in question are identified. The concluding section discusses the future prospects of the cluster, given the preceding analysis. 60 4 The Furniture Cluster in Ankara In many industrialized countries the furniture sector accounts for about 2–4 per cent of the production value of the manufacturing sector. The EU furniture industry is particularly strong, accounting for about half of the world’s production. Table 4.1 shows the market shares of the top ten exporters of furniture (SITC 821) in 1991–2000. As can be seen, Italy had the highest share with an impressive 17 per cent. Moreover Italy maintained its leading position throughout the period, while the export share of the runner-up, Germany, dropped from 15 per cent to around 9 per cent. The latter figure matched the US share, which remained fairly stable. The share of Canada, on the other hand, increased considerably in the second half of the 1990s and reached a significant 9 per cent. The shares of the two other important furniture producers, France and Denmark, fell slightly, each stabilizing at about 4 per cent. Belgium and Luxemburg also suffered a gradual decline from 4.5 per cent to 3 per cent. A similar market share (3 per cent) was captured by the United Kingdom, whose stake remained relatively stable. Meanwhile China and Poland emerged as new players and gradually increased their shares to 8 per cent and 4 per cent respectively. The export performance of the Turkish furniture industry, on the other hand, has not been particularly impressive. Despite an increasing trend in recent years, exports remain insignificant and the industry is mainly domestically oriented. The value of exports rose from about US$20 million in 1990 to $180 million in 2000. Over the same period imports increased from $25 million to $190 million. Turkey’s share of the world furniture market has remained at 0.2–0.3 per cent in recent years, which is below the calculated cut-off rate of 0.52 per cent for the Turkish manufacturing industry as a whole (Chapter 3), suggesting that the industry is not competitive in respect of SITC 821; that is, the general category ‘furniture’. Turkey does, however, hold competitive positions in some of the subsections of the industry. Of these, the export performance of manufacturers of seats and convertible beds (SITC 82115), motor vehicle seats (SITC 82112) and mattresses of other materials (SITC 82125) was relatively The Furniture Cluster in Ankara 61 strong in 2000, with world export shares of 4.2 per cent, 3.8 per cent and 2 per cent respectively (ITC, 2002). The major export destinations are Germany, Israel, the Netherlands, France, Russia and the CIS countries, especially Azerbaijan, Kazakhstan and Turkmenistan. In recent years exports to the Balkans in general and Greece in particular have also increased. Italy, Germany, France, the United States, the United Kingdom and Spain, on the other hand, are the leading exporters of furniture to Turkey. Within Turkey, the export performance of furniture manufacturers in Ankara is particularly weak, especially when compared with those in Istanbul and Bursa-Inegöl. 1 The geographic concentration of furniture manufacturers in some EU countries is striking. German enterprises, for example, are concentrated in three regions: North Rhine-Westphalia, Bavaria and Baden-Württemberg. Several districts in northern Italy are home to more than two thirds of the total number of furniture manufacturers and related employment in the country, with about 55 per cent being concentrated in the regions of Lombardy, Veneto and Tuscany. In France the concentrations are in the Ouest/Vendee/Bretagne, Ile-de-France, Rhone-Alpes and Normandy regions. Almost all Belgian furniture manufacturers are located in the western part of the Flemish region (Engels, 1999). Other well-known examples are Jutland in Denmark and Smaland in Sweden, and, outside the EU, Sasaki in Japan (Erzurumluoglu, 1991). As in the countries mentioned above, furniture manufacturing in Turkey is highly localized, with Ankara, Istanbul, Izmir, Bursa and Kayseri being the leading centres (Table 4.2). Although the manufacturers in Istanbul have recently replaced those in Ankara in terms of share of total employment in the industry, the LQ for Ankara is much higher than that for Istanbul Table 4.1 Market shares of the leading furniture exporting countries (per cent) Sources: ITC (2002); UN (1999). Country 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Italy 17.7 17.0 16.6 16.7 17.6 17.8 16.7 18.1 16.6 15.3 Germany 15.4 13.9 11.7 10.8 10.5 9.7 8.6 10.0 10.0 8.7 USA 7.9 8.5 9.5 9.2 8.0 8.0 9.1 9.2 8.5 8.7 Canada 3.7 4.0 4.8 5.4 5.5 6.1 7.0 8.9 9.0 9.5 France 5.6 5.5 4.7 4.5 4.4 4.3 4.2 5.0 4.8 4.4 Denmark 4.9 4.9 4.6 4.4 4.6 4.0 3.9 4.2 3.8 3.5 China 1.5 2.4 3.1 3.7 3.7 3.7 4.7 5.9 6.8 8.4 Belgium/ Luxemburg 4.5 4.5 4.0 3.7 3.6 3.4 3.0 3.4 3.6 3.1 United Kingdom 3.3 3.3 2.6 2.7 2.8 3.0 3.1 3.3 3.1 2.7 Poland 1.1 1.2 1.7 2.2 2.8 3.2 3.3 4.0 3.8 4.0 62 Clusters and Competitive Advantage (2.57 and 0.85 respectively), indicating that in relative terms the industry is much more localized in Ankara. Indeed the firms in Ankara are concen- trated in one district (Siteler) on the outskirts of the capital. Origins and historical developments Ankara was once a significant centre of mohair production, and this was the core economic activity in the city until the late nineteenth century and the coming of the railway. This enhanced the strategic location of the city and attracted additional economic activities. The main goods exported from Ankara at the time were grain, mohair, grapes, honey, leather, carpets/rugs and agricultural tools. Interestingly, although there was little in the way of forestry in the environs of Ankara, it appears that wood working was relatively developed in the region (Önsoy, 1994). For example wooden looms for the weaving of mohair were produced locally in the sixteenth and seventeenth centuries, pointing to the existence of wood-related business activities (Tuncer, 2001). In fact the craft of furniture making has been practised in central Anatolia for centuries. Impressive exhibits of eighth century BC royal wooden furniture Table 4.2 Provincial shares of employment in the Turkish furniture industry (ISIC 3610) 1 Notes: 1. Top ten provinces in terms of share of national employment and LQ greater than one. 2. Concentration ratios for the first four and eight provinces, respectively. Province Share of national employment in the sector (%) Location quotient Ankara 24.17 2.5698 Istanbul 20.75 0.8475 Izmir 8.57 1.1798 Bursa 6.88 1.7617 Kayseri 3.82 2.8844 Adana 3.06 0.9712 Hatay 2.46 1.4040 Eski sehir 2.32 1.8171 Samsun 2.27 1.3281 Içel 2.09 0.8367 Sakarya 1.10 1.0796 Trabzon 1.03 1.0026 Isparta 0.52 1.0332 Burdur 0.31 1.0100 Van 0.10 1.8051 C4EMP 60.36 2 C8EMP 72.02 2 The Furniture Cluster in Ankara 63 from Gordion, the ancient Phrygian capital, located near Ankara, are displayed in the Museum of Anatolian Civilizations in Ankara (Simpson et al., 1992). Much of the area’s historical magnificence was, however, lost over time, as evidenced by the way in which Major Robert Imbrie, a special representative of the United States, described the Ankara of the early 1920s: ‘The shops are tiny affairs with the most primitive of stocks. There is no bazaar worthy of the name. It is impossible to buy a chair, a table or a bed, or the simplest household article.’ According to him, it seemed that ‘the swing of the pendulum through the arch of centuries had brought little change to Angora’ (Cross and Leiser, 2000, pp. 143–4). Despite historians’ bleak picture of the general economic situation in Ankara at that time, the city took on strategic importance – both politically and economically – when the newly established Turkish Republic declared Ankara its capital in 1923, a decision that hardly received a warm welcome. Illustrative of this and the underdevelopment of Ankara is the fact that most diplomatic missions were reluctant to move from Istanbul to Ankara. For instance it took the American ambassador more than 10 years to move his residence permanently from Istanbul to Ankara (ibid.) The plan of the new Turkish government was to preserve the heart of old Ankara and build a completely new city round it. The transformation of Ankara from a small, underdeveloped, Oriental town to a large Western- styled city and the consequent flurry of construction activity inevitably influenced the structure of local industry and marked the emergence of two related sectors: construction and furniture. 2 Towards the end of the 1920s, factories and workshops began to spring up around Akköprü, specializing in ceramics, timber and furniture. 3 A notable development in the 1950s was the relocation of small producers and tradesmen outside the city centre. This started with the construction of small industrial estates around Ankara, including Yeni Sanayi, Büyük Sanayi, Ata Sanayi, Demir Sanayi and Keresteciler Sitesi. Although a number of industrial zones housed furniture firms and enterprises in related areas, one in particular, Siteler, is important for our purposes due to its currently high concentration of furniture manufacturers. 4 Written and oral evidence suggests that Siteler was established in the late 1950s by timber merchants who, after a large fire, were encouraged to move from their central location in Akköprü and Sogukkuyu to a more peripheral one, mainly because of the risk of fire but also due to the noise and pollution associated with their activities. The head of the municipality in Ankara, Atif Benderlioglu, with the support of Prime Minister Menderes, was highly instrumental in organ- izating this move. About 60 acres of land were allocated to the purpose, and the seeds of the subsequent furniture cluster were sown in Siteler by the arrival of the timber merchants. In the following years more timber merchants set up business in the district, and towards the end of the 1960s furniture makers formed a subdistrict of their own. Later the construction of residential 64 Clusters and Competitive Advantage buildings began, and production activities gradually diffused into the new residential areas, blurring the boundaries of the district. Finally, the entry of larger firms triggered a transformation in Siteler in that the emphasis on the local market shifted towards regional, national and even international markets, though to a lesser extent (Tekeli, 1994). Very few final customers visited Siteler in the early years of its develop- ment. Rather they bought their furniture from retailers in the city centre, where the prices included a considerable mark-up. Seeing the business potential in this, many furniture makers in Siteler set up showrooms often on the ground floor of their premises. Thus Siteler was no longer just a production centre but also a specialized shopping centre. This proved very profitable for the furniture makers, especially given the fact that their rents were much lower than those in the centre. Moreover ‘the former workshop owner now acquired the status of a businessman who had an office in the showroom’ (Hazar, 1983, p. 17). The success achieved by Siteler during these years resulted in nationwide recognition of ‘made in Ankara’ furniture and a cosy life for the furniture makers. However they failed to channel their earnings into investments, and when similiar furniture districts began to emerge in other parts of the country, coupled with the liberalization of imports, they found it difficult to sustain their dominant position since they had little export experience and few contacts in international markets to turn to in order to fight the competition. As a consequence the good reputation Siteler had built up over the years began to erode. A prominent manager interviewed in the course of this study, whose company was amongst the first to move to Siteler, stated that what they dreamt of in the early days could not be realized. They had envisaged that Siteler would come to be endowed with well-functioning institutions, including sector-specific education and research institutions and even a forest of its own. He thought that over time the district’s highly skilled craftsmen had gradually been replaced by people who had the required capital but were not necessarily equipped with sufficient skill. This, in his view, had changed the nature of the district. ‘If it had developed as envisaged, Siteler would have been a leading export centre; but now only 10–15 firms are engaged in exports on a regular basis’, he lamented. Siteler now occupies about 1250 acres of land. According to the Small and Medium-Sized Industry Organization (KOSGEB), approximately 10000 firms are active in furniture making and related businesses in Siteler. The number of workers employed by these enterprises is estimated to be around 100000. Although the cluster has managed to survive the recent challenges, Istanbul has replaced Ankara as the leading location for furniture making and the export performance of the cluster is still rather poor. The following pages examine the reasons for this decline, which will enable us to comment on the competitiveness of clusters more generally. Before proceeding with this The Furniture Cluster in Ankara 65 analysis, however, we shall summarize the general economic outlook of Ankara in order to put the situation of the furniture cluster into that context. As can be seen in Table 4.3, which shows the structure of Ankara’s economy by economic activity and share of employment, the current structure is quite similar to that observed historically: construction and related activities, Table 4.3 Employment in Ankara, by economic activity 1 ISIC Sector Location quotient Ankara’s share of national employmen t in the sectors (%) 8022 Technical and vocational secondary education 8.2230 77.33 9220 News agency activities 7.7405 72.80 4020 Manufacture of gas; distribution of gaseous fuels 5.6415 53.06 7210 Hardware consultancy 5.0227 47.24 7493 Building cleaning activities 4.9979 47.00 4520 Construction of buildings; civil engineering 4.2882 40.33 7010 Real estate activities with own or leased property 3.5444 33.33 6591 Financial leasing 3.4805 32.73 2694 Manufacture of cement, lime and plaster 3.3328 31.34 2924 Manufacture of machinery for mining and construction 3.1600 29.72 6412 Courier activities other than the national mail service 3.1314 29.45 7421 Architectural and engineering activities and consultancy 2.9847 28.07 2923 Manufacture of machinery for metallurgy 2.6363 24.79 3610 Manufacture of furniture 2.5698 24.17 3000 Manufacture of office and computing machinery 2.5244 23.74 9219 Other entertainment activities n.e.c. (not elsewhere classified) 2.4805 23.33 7250 Maintenance and repair of office machinery 2.4669 23.20 6420 Telecommunications 2.4608 23.14 8532 Social work 2.4494 23.04 1422 Extraction of salt 2.3852 22.43 3311 Manufacture of medical and surgical equipment 2.3068 21.69 7310 Research on and experimental development of NSE (Natural Sciences and Engineering) 2.2911 21.55 4540 Building completion 2.2308 20.98 6301 Cargo handling 2.2249 20.92 5150 Wholesale of machinery, equipment and supplies 2.2110 20.79 8519 Other human health activities 2.1374 20.10 6719 Activities auxiliary to financial intermediation n.e.c. 2.1184 19.92 2925 Manufacture of machinery for food etc. 2.0585 19.36 2699 Manufacture of other non-metallic mineral products 1.9742 18.57 2915 Manufacture of lifting and handling equipment 1.8827 17.71 2222 Service activities related to printing 1.8692 17.58 [...]... leave the firm as a result of a conflict with the owner-manager and then establish his own business As such people tend to leave before they fully acquire the necessary skills, this practice is 70 Clusters and Competitive Advantage seen as damaging rather than contributing to the dynamism of the cluster (Porter, 1998) Family firms dominate the cluster These small firms are typically managed by the owner,... apparel, and materials and metals The industry is also associated with the construction industry, which is internationally competitive, and with motor vehicles, whose export performance has improved in recent years On the other hand Turkey is not competitive in many subsectors of the timber industry in terms of both price and quality (SPO, 1995) For example the export performance of the fibreboard and particleboard... towards the end of the 1990s Germany and former republics of the Soviet Union, especially the Russian Federation, became the most important foreign customers (Koç et al., 1999) This shift from the Middle Eastern and North African markets to Russia and former Soviet republics closely matches the path taken by Turkish construction firms in their international operations (Öz, 2001) The development of the. .. requires the expertise of master craftsmen 72 Clusters and Competitive Advantage Institutions that provide technical and marketing information and arrange seminars and trade fairs (the latter can have a strong influence on furniture design and public taste, as do the Triennale exhibitions in Milan) either do not exist or are weak in Ankara Thus information on foreign markets is limited, and acquiring... subsectors is rather poor The local machinery sector is weak as well, and state-of -the- art, sector-specific machinery is imported from Europe, especially from Germany and Italy, which together hold more than half of the world market in this product category (OAIB and KOSGEB, 1996) As mentioned above, the Turkish construction industry has become internationally competitive and many of the major construction... painted and preserved Materials and metals Figure 4.1 Internationally competitive subsectors of the Turkish furniture industry, plus related sectors and institutions Note: Industries/institutions with a presence in Ankara are shaded The Furniture Cluster in Ankara 73 The furniture industry is linked to three of the strongest sectors of the Turkish economy: housing and household goods, textiles and apparel,... strengthening the links between the two industries Also, when the construction industry is booming, this has a positive impact on the Ankara cluster, especially with respect to kitchen and bathroom furniture and interior decoration In summary, of the competitive industries with links to the furniture industry, only construction and metal-working are located in Ankara It therefore cannot be said that the. .. early development of the furniture cluster More recently, in the 1980s, the liberalization of imports served to educate both producers and consumers, and prompted some furniture firms to look for export markets KOSGEB-Siteler is the main local governmental institution to support the development of the furniture cluster.9 Another governmental institution, 74 Clusters and Competitive Advantage IGEME, concentrates... enterprises usually have their own export departments, while very small firms and workshops are either unaware of the services provided or have little interest in them as they do not intend to export The high levels of bureaucracy and red tape constitute major drawbacks for firms in this regard Another government-related problem is the absence of stringent quality and service standards The managers interviewed... pay in the centre, the district has limited opportunities for expansion In summary, of the three major operating costs – raw materials, wages and finance – cluster firms have an advantage only in labour cost and suffer severe disadvantages in the others A comparison of the price of a typical piece of bedroom furniture in Turkey and Germany reveals that low wages are not enough to compensate for the high . 4.5 4.0 3. 7 3. 6 3. 4 3. 0 3. 4 3. 6 3. 1 United Kingdom 3. 3 3. 3 2.6 2.7 2.8 3. 0 3. 1 3. 3 3. 1 2.7 Poland 1.1 1.2 1.7 2.2 2.8 3. 2 3. 3 4.0 3. 8 4.0 62 Clusters and Competitive Advantage (2.57 and 0.85. engineering 4.2882 40 .33 7010 Real estate activities with own or leased property 3. 5444 33 .33 6591 Financial leasing 3. 4805 32 . 73 2694 Manufacture of cement, lime and plaster 3. 332 8 31 .34 2924 Manufacture. classified) 2.4805 23. 33 7250 Maintenance and repair of office machinery 2.4669 23. 20 6420 Telecommunications 2.4608 23. 14 8 532 Social work 2.4494 23. 04 1422 Extraction of salt 2 .38 52 22. 43 331 1 Manufacture