Clusters and Competitive Advantage_8 doc

This is still valid for Istanbul’s leather industrialists, as repeatedly emphasized bythe interviewees, who stated that Istanbul differs from Izmir Istanbul’smajor competitor in leather

of leather works in Kazliçesme was the ready availability of water there.Moreover, although raw leather was available almost everywhere in Anatolia,high-quality tannin, an essential ingredient of the tanning process, could beobtained only in a limited number of places, one of which was Istanbul.26The initial factors that favoured the location, therefore, were factor conditionssuch as the availability of water, and historical events such as the Turks’conquering of Istanbul Tanneries were concentrated in certain locations inthe Ottoman era as well, but were encouraged to move outside the citycentres because of the unpleasant smell and pollution caused by their activities.Environmental concerns continue to be one of the main reasons why tanneriestend to concentrate in space, in that they are encouraged or required tooperate in places permitted by the government These are usually outsidethe city centres and the tanneries share the cleanup costs

The Istanbul leather producers were located in Kazliçesme for more than

500 years until their move to Tuzla.27 Subcontracting was very common(Yelmen, 1992, pp 153, 161), and recommendation by key actors in thedistrict was of great importance: ‘I used to try hard to convince a well-knownmaster craftsman to use and recommend the machinery that I produced Hisopinion was valued very much in Kazliçesme’ said one interviewee The factthat they were in tough competition with each other did not prevent manu-facturers from spending time together socially They attended the samelocal cafes and restaurants, which helped them to keep up to date with whatwas going on and provided them with motivation

Although leather industrialists in Kazliçesme had a history of solidarity,28the competitive pressures that existed paralleled those identified in the clustersexamined in the previous chapters One of our interviewees stated that

‘partnerships were rare in Kazliçesme and even these collapsed sooner orlater’ since ‘the atmosphere of the place did not tolerate partnerships’ This

is still valid for Istanbul’s leather industrialists, as repeatedly emphasized bythe interviewees, who stated that Istanbul differs from Izmir (Istanbul’smajor competitor in leather clothing) in this regard.29 Leather industrialiststhink that the ‘artistic nature’ of leather processing heightens the intensity

of competition and is a major factor in the dominance of family businesses:

‘The craft, that is, the particular way of processing leather, is only taught toclose members of the family who will take over the business’ Leather is anorganic and diverse material, and the expertise of craftsmen with the necessaryskills to ensure its proper treatment is a very valuable asset for leather firms

As a result they compete fiercely for such craftsmen

Another commonality of the Istanbul leather clothing cluster with theclusters previously examined is that specialization is a relatively recentphenomenon, despite the long history of the overall industry According toour interviewees, the beginnings of this specialization date back to the late1950s Prior to that period, leather jackets were not commonly worn and

The Leather Clothing Cluster in Istanbul 153

were mainly associated with policemen, railway workers and lorry drivers

It was only after the introduction of a new dyeing method, which involvedthe use of chromium, that leather garments were seen as desirable by othercustomers One of our interviewees, Hasan Yelmen, who is a chemical engineerand regarded as the father of the contemporary Turkish leather industry,was the first to use this technique in Turkey and paved the way for thespecialization in leather garments In the 1970s a growth in demand for theseproducts, especially from Europe, prompted others to enter the sector Thiscoincided with the ready availability of high-quality but cheap sheepskin,which was the favoured leather for manufacturing garments Meanwhile thefirm Derimod pioneered the incorporation of design features into leatherclothing The cluster’s specialization in leather garments, therefore, was initiallytriggered by an entrepreneur and subsequently reinforced by changes ininternational demand

Demand patterns also played a crucial part in the development of thefootwear sector This time, however, the push was in the opposite direction

in that the domestic demand for shoes was large and the footwear sector didnot come under pressure to internationalize until the 1990s, when it began toface significant foreign competition in the domestic market As a consequencethe export-oriented leather clothing sector now dominates the leather industry,unlike in Italy, where the industry has a rather different structure

Some of the benefits of clustering are independent of the location inquestion and are related to the phenomenon of clustering itself, as analysed

in the previous chapters First, it is beneficial to customers since it enablesthem to see a wide range of alternative options in a short time ‘This is wherethe heart of the industry beats’, one interviewee said, paralleling the com-ments of his counterparts in the other clusters examined, who emphasizedthe ‘market-like’ nature of clusters in the eyes of the customers This factoralso ensures a good price–quality ratio since customers can easily compareproducts in terms of value for money

The second benefit is the proximity of related and supporting industries.One interviewee, for instance, argued that ‘isolated’ firms in provincessuch as Konya and Elazig were not so successful because ‘They could noteven find a technician to repair broken machinery Besides, they had tobuy chemicals from here, plan everything in advance accordingly and bearthe associated costs, whereas we can sometimes buy chemicals daily.’Another interviewee was of the opinion that it was not only handy to haveeverything needed nearby, but also knowing that they were within reachgave added confidence Yet another interviewee made an analogy with thehuman body: ‘All parts should function properly and only then can youtalk about a healthy system.’ Overall, these points reinforce the finding ofthe previous cases examined, that the presence of related and supportingindustries contributes positively to the development and competitiveness

of clusters

that this benefit had been much more pronounced in the original location

of the industry, Kazliçesme, where the factories had been even closer to eachother According to him, when rival firms are located close it is possible tosee what is going on, whether new machinery has been bought and the types

of dye competitors are using

The fourth benefit is that if success is achieved during the early stages ofcluster formation, this initiates a self-reinforcing dynamic For instance inIstanbul there were only a few firms specializing in leather-related chemicals

in the 1950s, but with the leather cluster’s growing success their numberincreased considerably in the subsequent decades The Istanbul leathermachinery sector underwent a similar development In the 1950s the sectorconsisted of nothing more than a number of craftsmen manufacturing some

of the tools used in leather processing From the 1970s onwards, however,machines imported from abroad were copied by Istanbul engineers, whocontinued to hone their skills and formed the foundation of today’s leather-related machinery sector In the meantime the founders of the familybusinesses that dominated the cluster were taking steps to ‘professionalizethe family business via a professionalization of the family’ (Bugra, 1994).This included sending members of the second and third generations abroad(usually to France or Britain) to complete their education in leather-relatedareas Another factor in the self-reinforcement of the cluster was newbusiness formation by spin-offs from leading firms, as well as by newcomers

to the area who had heard about the opportunities it offered An interestingfinding in this regard, and one that also applies to the previous casesexamined, is that a noteworthy number of cluster participants came fromthe same town, Egin, located near the Anatolian cities of Malatya and Sivas.(Recall that the town in question was Babadag in the case of the towel andbathrobe cluster and Besni in the case of the carpet cluster.) As with theother towns, Egin was a place with few opportunities and many of theyounger generation were forced to leave to seek their fortunes This not onlysuggests that entrepreneurship is triggered by difficulties and challenges butalso illustrates the importance of contacts and fellow townsmanship, as well

as the magnetic nature of clusters

The final observation on clustering is unique to Istanbul and is not found

in the other clusters examined: there are certain costs that can be linked to

‘diseconomies of urbanization’ Despite its obvious advantages, Istanbul, beingthe largest city in Turkey, can be a very difficult place in which to conductbusiness For instance the cost of living and therefore the cost of labour ishigher than in many other locations in Turkey Moreover leather firms aresubject to stringent regulations, for example in terms of environmental pro-tection, which is not necessarily the case for firms elsewhere Nonethelessthe disadvantages tend to be outweighed by the advantages There is a pool

of qualified workers to draw upon and no shortage of supporting industries

The Leather Clothing Cluster in Istanbul 155

and services, including transportation and logistics According to the viewees, what is even more important is that Istanbul has an ‘aura’ (echoingMarshall’s (1949) ‘something in the air’ hypothesis), a refined culturalenvironment and well-established international links All this encouragesfirms to upgrade their activities and motivates people to realize their potential

inter-As one interviewee put it, ‘it is a school’ in that ‘even while walking on thestreets of Istanbul you can see a clothing item that inspires you’ The factthat leather firms do not contemplate relocating is related not only toIstanbul’s unbeatable advantages but also to their expectation that today’sfew disadvantages will turn into tomorrow’s advantages (see the next sectionfor more on this issue) Lastly, the presence of diseconomies of urbanization

in Istanbul prompts the more general question of whether clusters in smallcities are different in nature from clusters in metropolises but this matterwill be left to Chapter 8

Future prospects for the leather clothing cluster in Istanbul: stuck

in the middle?

As mentioned at the start of the chapter, Turkish exports of leather garmentshave levelled off in recent years, and there has been a considerable fall inTurkey’s world export share Although some of the loss in registered exportscan be explained by the activities of the informal economy in Laleli, the sale

of leather garments via shuttle trade has also fallen considerably since thepeak in the mid-1990s In the meantime China, India and Pakistan haveemerged as the new leaders of the leather industry in the price-sensitivesegments, while the world export share of Italy, a noteworthy producer

of design- and fashion-oriented leather garments for the upper end of themarket, has remained relatively stable

The fall in Turkey’s relative position can be explained by the fact that theindustry had to import more than half of its requirement of raw leather, bythe rise in labour costs, and by the emergence of new competitors offeringmore competitive prices In other words, Turkey’s heavy reliance on cheapraw materials and low-cost labour to provide it with competitive advantagewas unsustainable in the long term As a consequence the Istanbul leathergarment producers came under pressure to find a more sustainable source ofcompetitive advantage, otherwise they might well become ‘stuck in themiddle’ (Porter, 1985)

In order better to understand the gradual change in the strategic orientation

of the Istanbul leather clothing firms, it is informative to look at the impact

of the Russian market on the cluster Prior to the emergence of the informalshuttle trade the firms mainly served the European market, where customerswere highly demanding This forced producers to improve the quality oftheir products, but when these customers were replaced by informal traderswho required low-cost goods and were less concerned about quality, the latter

the beginning and the charm of an easy life was hard to resist, this created

a poor image for Turkish products that would prove difficult to counter(SPO, 2000) Moreover in the late 1990s, consumer demands in the Russianmarket started to change and the informal traders began to pay more attention

to quality In the meantime wholesalers and transportation companies weregaining a hold over the organization of the shuttle trade (Yükseker, 2003).Apart from the difficulty of re-establishing the neglected contacts withEuropean customers, the leather clothing firms now had to improve theirimage in the Russian market, where in some of the more upmarket shoppingcentres signs stating ‘Turkish products are not sold here’ were placed in shopwindows The temporary shift to the ‘easy life’, therefore, proved costly forthe cluster firms Indeed some of them switched to Italian-sounding brandnames in an effort to reverse the damage

Returning to the subject of new sources of competitive advantage, afterthe setbacks described above the leather clothing manufacturers drew ontheir accumulated experience and know-how and engaged in marketing andpromotion efforts to improve the image of Turkish leather garments abroad

As a result, in the early 2000s the fortunes of the cluster firms started to pick

up again and they managed to erase their low-quality image in the Russianmarket Today the larger firms are targeting the upper end of the market andare second only to Italian firms in this sector (SPO, 2000; IGEME, 2002).Thus although it is still possible to cater to the lower end of the market,larger producers are producing and exporting high-quality fashion items as

it is impossible for them to compete with Chinese, Indian and Pakistanileather garment producers in terms of price

A related issue is that currently the existence of stringent regulations onenvironmental pollution and working conditions puts the cluster firms in

a disadvantageous position vis-à-vis firms in other regions that do not

have to bear the associated costs The two leather industrial zones inTurkey (Tuzla in Istanbul and Menemen in Izmir) have full cleanup facilit-ies, the cost of which is inevitably reflected in the prices of the productsmanufactured in these zones Although this situation is problematic inthe short term as it hinders competition with other regions, it will in factprovide the manufacturers with a significant advantage in the long term,given that being ‘green and competitive’ is now viewed as the way forward(Porter and Van der Linde, 1995)

In summary the analysis of the leather clothing cluster in Istanbul hashighlighted three important points First, it has given us the opportunity

to consider the distinguishing characteristics of an informal, transnationaleconomy (the case of Laleli), which is a rather understudied form of inter-nationalization in the literature on clusters Second, the analysis has revealedthat the benefits of clustering seem to outweigh the diseconomies ofurbanization currently faced by the cluster, given that the cluster persists

The Leather Clothing Cluster in Istanbul 157

and relocation is not on the agenda of the cluster firms, which moreoverexpect these disadvantages to turn into advantages in the near future.Accordingly it is preferable to work in a demanding environment providedthat this channels the industry in the right direction, as in the case of thestringent environmental standards mentioned above Finally, the examination

of the leather clothing sector has provided us with the opportunity to take

a closer look at ‘a cluster in transition’ We have seen that the manufacturers’low price strategy resulted in their being ‘stuck in the middle’ when newcompetitors began to threaten them in price-sensitive markets Moreoverthe pattern of demand has changed in their largest market, the RussianFederation These developments, coupled with the 1998 Russian crisis, haveforced producers to reconsider their low quality, low price strategy, andtaught them the importance not only of product design and quality but also

of diversifying their export markets The changes observed in the competitivedynamics in this leading cluster of the Turkish economy are therefore veryinformative, but arguably some aspects of this, such as the pressure caused bythe emergence of new competitors, are not specific to the Istanbul cluster

Conclusions

This chapter will first discuss the theoretical implications of the key findings

of the cluster analyses with respect to the ongoing debate in the literature

on the competitiveness of clusters (see Chapters 1 and 2), and then considerthe implications of the study in terms of policy

Theoretical implications

The emergence of clusters

At the beginning of this book we discussed some alternative theoreticalapproaches to how and why a cluster emerges in a particular location Themain triggers might be historical accidents, demand patterns and the priorexistence of related and supporting industries and special inputs, includinglabour (Marshall, 1949; Krugman, 1991a; Porter, 1998) In general the emer-gence of the clusters examined in this book can be related to one or more ofthese, but additional factors also play a part In the case of Ankara it was thedeclaration of the city as the capital of the Turkish Republic, its central locationand climate, and the presence of a related industry (construction) that triggeredthe emergence of the furniture cluster For the Denizli towel and bathrobecluster it was the city’s good location, historical circumstances, the availability

of inputs, the prevailing demand patterns and the presence of an preneurial spirit A similar set of forces triggered the development of theleather clothing cluster in Istanbul, for which a specific chance event, SultanMehmet II’s particular interest in the sector, was also very important Finally,the emergence of the carpet cluster in Gaziantep was linked to its key location,the availability of inputs, the presence of related industries (yarn and textilemachinery) and an entrepreneurial spirit, as well as a ‘historical accident’:the long-established border trade with Syria.1

entre-Although these findings are generally in line with those in the literature,additional factors have been revealed For example an examination of othercentres of production shows that similar conditions, such as a long history

Conclusions 159

in textiles or leather and the presence of inputs and several related industries,pertain in other locales in Turkey, prompting the question of whetherchance events were the determining factors in the emergence of the com-petitive clusters This contention is argued against in Chapter 6, whichdiscusses how Gaziantep emerged as the leading location for machine-wovencarpets when there was no tradition of carpet weaving, apart from kilims

In this particular case the importance of historical accidents is clear, sincethe event that triggered the emergence of the cluster was the growingdemand for silk carpets imported via Syria It is notable that this chanceevent was capitalized on in Gaziantep and not in another Eastern Anatoliancity that also engaged in cross-border trade This can be attributed to theentrepreneurial spirit of Gaziantepians, which was in turn reflected in thecity’s favourable business environment As discussed in Chapter 6, it maywell be that such historical accidents happen all the time, but they triggerthe emergence of a cluster only if the local business environment is conducive

to growth

Another interesting issue revealed by the case studies is that although theproduction of textiles and leather had a long history in the clusters examined,specialization had only taken place recently The real impetus began in the1970s in all three cases, triggered by different combinations of circumstances(on top of those which favoured the general area of activity), namely entre-preneurship and demand patterns (mainly international) in the towel andbathrobe cluster; chance events, entrepreneurship and demand patterns(both national and international) in the carpet cluster; and chance events,entrepreneurship and demand patterns (mainly international) in the leatherclothing cluster This observation adds a new insight to the contributions inthe literature, although the general theme is in line with the notion of ‘theprocess of becoming specific’ (Storper, 1999)

Another issue that has been highlighted is the possibility that once a clusterbegins to form a self-reinforcing dynamic comes into play and locks theactivity into the location even if the initial conditions change (Arthur, 1985).This is exemplified by the cases of the towel and bathrobe and leather clothingclusters in that the basic conditions that triggered their emergence no longerprovided them with a clear advantage Apparently the benefits of clustering(see below) are strong enough to ensure the continuity of the cluster even ifthe initial advantages diminish in importance or disappear

The final point on the emergence of clusters is that it is often a odological challenge to identify the historical triggers The present studybenefited from the fact that specialization in the products in questionwas a relatively recent phenomenon, so the scarcity of written documen-tation was countered by the availability of people who had witnessed theearly stages of the clusters’ development This underlines the importance oforal history and the necessity of documenting this history while it is stillpossible

meth-The costs and benefits of clustering

The findings from the case studies in respect of the costs and benefits ofclustering are generally in line with those in the literature in that the proximity

of related and supporting industries has been found to be very beneficialand the benefits of clustering for customers are large, while the proximity ofrival firms generates both costs and benefits When it comes down to thedetails, however, some interesting issues emerge

For instance clusters are convenient not only for customers since theyreduce the likelihood of arbitrage as well as search costs, but also for thefirms themselves since clusters are treated as markets to be visited, ratherthan the firms having to visit the customers Similarly, the beneficial effects

of the proximity of related and supporting industries include not onlystatic aspects such as reduced transportation and transaction costs but alsomore dynamic ones, such as the contribution it makes to innovation andgrowth

The perceived costs and benefits of in-cluster rivalry, on the other hand,are far from uniform Before analysing the details of this issue, the perceptions

of cluster firms with regard to competition from other regions of Turkeyshould be mentioned The managers interviewed in the furniture cluster, forinstance, stated that they competed with Istanbul producers in the market forquality furniture and with the other locales such as Inegöl in the price-sensitivesegments The Istanbul leather industrialists saw Izmir as their main competitor

in leather clothing, but other regions could cause trouble in price-sensitivemarkets by means of unfair competition Finally, managers in the toweland bathrobe and carpet clusters perceived competitors in other regions as

‘different’ in the sense that they were isolated and very large, and thereforewere not direct competitors

In-cluster rivalry is very intense in all four cases examined in this study Infact, competitive forces are much more pronounced than cooperative ones,supporting the finding by Saxenian (1994) and Porter (1998) As noted inthe literature, in-cluster rivalry can be personal and emotional, and is some-times triggered by envy (Schoeck, 1966; Klein, 1975) and a desire to lookgood in the local community (Porter, 1998) A repeated theme raised by theinterviewees is that the success of rival firms acts as a motivator and thusstimulates growth Since the rival firms are working under the same condi-tions, firms have little excuse ‘if the neighbouring company is exporting,say, to France, but you cannot’ This not only stimulates motivation but canalso increase confidence since, as one manager stated, ‘you learn not onlyabout the opportunities themselves but also about the fact that they cangenerate fruitful results’ These observations provide support for the emphasis

in the literature on the motivating nature of competition (Porter, 1990, 1998)

as well as the importance of learning, tacit knowledge and informationspillovers (Camagni, 1991; Brusco, 1996)

Conclusions 161The fact that information spreads rapidly in a cluster, on the other hand,raises the issue of what information should be shared with competitors andwhat should not In the case of the Turkish clusters examined, we have seenthat information on products and markets is crucial for success, and hence

it is considered inappropriate to share such information with competitors.The greatest disadvantage of the proximity of rivals, according to the managersinterviewed, is the rapid pace of imitation Imitating best practices is easier

in a cluster, being facilitated by the rapid flow of information through socialnetworks In-cluster competition, therefore, has advantages (it brings dyna-mism and stimulates innovation) and disadvantages (for example imitation

by competitors is easier) However, even the problems associated with the ease

of imitation might provide benefits in the long term since firms are inevitablypushed to develop sustainable competitive advantages

It is also possible that the cluster environment prevents destructive pricecompetition, given that the likely costs of this are obvious to all and socialsanctions may be imposed on those who start a price war (Dei Ottati, 1994).This points to the self-disciplining aspect of clustering, as observed in thecase of the towel and bathrobe cluster in Denizli, where there is an unspokenconsensus on the approximate price that should be charged for a typicalbathrobe In other words competition is good, provided that it is ‘sweet’

As the case of the furniture cluster in Ankara illustrates, however, there is also apossibility that it can turn into destructive competition The challenge is toidentify the circumstances that cause in-cluster competition to evolve towardseither of these variants, which will be discussed in the next subsection The analysis in this study has provided some insights into the nature oflocalized competition as well According to Baum and Haveman (1997), localcompetition can lead to the differentiation of some product dimensions Thisfinding derives from a study of the hotel industry in Manhattan, wherehotels are different in size but similar in terms of their price range The leatherclothing cluster in Istanbul confirms this finding since the firms there areconcentrated in terms of their strategic orientation The analysis of theGaziantep carpet cluster, on the other hand, has revealed that carpet firms

of similar size tend to concentrate in space and that smaller ones dream aboutupgrading their location to the industrial zones that house the larger carpetfirms

Another related issue is the assertion that clustering is linked to duction divisions (Steinle and Schiele, 2001) Both the Istanbul leather clothingcluster and the Gaziantep carpet cluster provide evidence against this conten-tion in that most of the firms are vertically integrated but remain clustered.The furniture cluster in Ankara, though in line with this contention giventhat the firms specialize in different areas of the furniture business (each

pro-of which is concentrated on a street-by-street basis), is not competitive ininternational markets Apparently clustering offers benefits in addition toreducing transaction costs

‘the logic that it is a better idea to establish a new firm in another location sincethere are too many firms here does not work’, as one of the interviewees put it.Accordingly, ‘it is costly to be outside, remote from the center of activity’,the associated costs usually being in the form of unanticipated productivitypenalties (Porter, 2000) The self-reinforcing dynamic in clusters, which alsohas the potential to turn the location into a magnet, is fed by informationspillovers In the Turkish clusters, spillovers related to the strategic orientation

of firms can be even more important than technological spillovers This is inline with the finding by Beal and Gimeno (2001) that information spillovers

on marketing strategies are more important than technological spilloversfor US software firms In a similar vein, the Turkish clusters underline theimportance of strategic innovation as well as technological innovation Whenthese thoughts are brought together with the fact that the major sources oflearning are competitors, customers and firms in related industries, theimportance of tacit knowledge (Becattini and Rullani, 1996; Brusco, 1996)and collective learning (Camagni, 1991) becomes apparent

A related observation is that a disproportionate number of entrepreneurs

in each of the competitive Turkish clusters examined came from a singletown in the regions in which the clusters are located: Babadag in the case ofthe Denizli towel and bathrobe cluster, Besni in the case of the Gaziantepcarpet cluster and Egin in the case of the Istanbul leather clothing cluster.The fact that these small Anatolian towns offer few opportunities for theyounger generation, and therefore the latter have to leave to seek their fortune,suggests that entrepreneurship might be triggered by harsh conditions.What is more important for our purposes, however, is that it also illustratesthe key role played by social relations and contacts – in the form of fellowtownsmanship in our example – in the process of new business formationthat feeds cluster development

A much-debated topic in the literature on clusters is the part played bytrust Two of the cases examined in this book have provided interestinginsights in this regard In Denizli, trust is clearly manifest in the so-called

‘Babadag banknote’ (see Chapter 5) This type of trust, which is culturallyembedded in the locale, has been well studied in the literature (Granovetter,1985) and is mostly exemplified by clusters in ‘Third Italy’ It is also clearthat contract-based trust can function well, too, if supported by the properenforcement of law (Cohen and Fields, 1999) However the type of trustwitnessed in Laleli, the centre of informal trade in the Istanbul leather clothingcluster, does not fit either of these two types of trust Given that in Lalelitrust-based relationships are established between two parties who start off

as total strangers to each other and come from different countries, oftenTurkey and the Russian Federation, it is obvious that such relationships arenot culturally embedded in the locale At the same time, the fact that Laleli

is an informal market means that the kind of trust observed there is not the

Conclusions 163same as that in Silicon Valley either, where written contracts and properenforcement of the law substitute for the lack of culturally embedded trust.Among the available theoretical perspectives, the most suitable for explainingthe type of trust that exists in Laleli is game theory (Dixit and Nalebuff, 1991).That is, trust is built up and reinforced by repeated interactions between twoparties, whose previous contacts and the results of these determine theoutcome of subsequent ones The fact that the parties involved come fromdifferent countries does not appear to matter (see Chapter 7)

The case of Laleli is also informative in respect of a relatively less studiedsubject in the literature: the informal international business activities in thislocale are conducted exclusively by small informal entrepreneurs (Yükseker,2003), indicating that such activities might also tend to cluster in space Theimportance of entrepreneurial activity more generally is emphasized by thecase studies in this book In particular, it has been entrepreneurs who havemobilized the potentials that exist in the location for their particular area ofbusiness This brings us back to the importance of the quality of the localbusiness environment, which will be discussed in the next subsection

Clusters and competitive advantage

The clusters examined in this book are largely dominated by small andmedium-size enterprises, although a few larger firms also exist, especially inthe furniture and carpet clusters There is a large degree of specialization inthe furniture and towel/bathrobe clusters, whereas the key processes in thevalue chain have been internalized by many of the firms in the carpet andleather clothing clusters.2 The part played by entrepreneurship has beenimportant, especially in the more competitive clusters All four clusters havehad the opportunity to benefit from the export incentives provided by theTurkish government, and the firms located in Denizli and Gaziantep fromthe assistance offered to less developed regions The firms in the Istanbulleather clothing cluster and the Denizli towel and bathrobe cluster havemade good use of these incentives, but those in the Ankara furniture clusterhave been unable to do so because of their low level of internationalization.The Gaziantep carpet cluster firms have also benefited little from the incen-tives as their development started too late There are competitive forces inall four clusters, and differing degrees of cooperation and competition Forexample small firms in Denizli joined forces to form a sectoral foreign tradecorporation that has enabled them to break export performance records, butsuccessful entrepreneurs in Gaziantep are sceptical about cooperating witheach other All but one of the clusters have integrated themselves intointernational markets, but the forms of integration differ The most interestingcase in this respect is that of Laleli, as discussed above

One general conclusion that can be drawn from the analysis is that thecontention that clustering is a widespread phenomenon in both developed anddeveloping economies finds support from the case of Turkey (see Chapter 3),