1The structure and development of urban spaces are increasingly subject to global economic dynamics, among which number not only the global movement of people, financial capital, information and services, but also the global production and circulation of goods. The latter are now predominantly organized in transnational production networks that rest on the global division of labour and the unequal conditions prevailing in different regions of the world. The “Made in ...” label usually informs consumers about the country of export, but the numerous, concrete, tangible places that a product passes through in the course of its manufacture, and likewise the people therein involved, remain completely invisible (as well as unimaginable).
2In fashion industries in particular, owing to ethical concerns about current working conditions, efforts are now being made to gain knowledge on international supply chains and assure greater transparency, either through scientific research (Brooks 2015, Plank and Staritz 2009, Pickles and Smith 2011, Rossi et al. 2014), or journalistic enquiry (Klein 1999, Rivoli 2005, Uchatius 2010), the work of NGOs (such as Clean Clothes Campaign or Fair Labour Association), or companies’ own corporate social responsibility programmes. Some brands and retailers even publish their global suppliers lists online (H&M Supplier list n.d.), or allow selected items to be traced on their website (dm Pfad-Finder n.d.). But despite these attempts – often just promotionally effective acts of greenwashing – and despite increasing demands and efforts to assure decent wages in “fair trade” production relations (Ditty 2015, Musiolek and Luginbühl 2014, Hughes et al. 2008, Raynolds and Bennett 2015), the actual supply chains of the vast majority of products remain in the dark.
3Moreover, the types of architecture and urban context in which globalized production takes place and the impact thereof on the built environment are still largely unexplored. If, however, local places are increasingly constituted by their transnational links, then, as Massey (2004) has argued, the players at powerful points within global networks bear responsibility, because of their practices, for the constitution of other places in the world that are somehow connected with these.
4How globalized production relations determine the economic development of individual production locations has been the object of economic and social sciences research into global commodity chains for some time. Yet the extent to which they also shape the architectural profile and urban environment of such locations is a largely neglected topic. For one the urban dimension of industrial activity is only rarely the focus of urban research and, secondly, it is seldom examined in the light of transnational production processes.
5However, this not yet fully fathomed potential, to further urban research by studying global commodity chains, has recently been recognised and addressed, in particular in the field of research devoted to transnational urban spaces. Based on a review of these approaches, the present paper, firstly, outlines a research agenda for studying the built environments of transnational clothing production, and, secondly, offers exemplary insights into a still on-going research project that draws on a single clothing industry case study. Turning the spotlight on multi-local production relations, first and foremost from a spatial perspective, can shed new light on the transnational constitution of urban spaces and bring into focus places and urban dynamics that tend to be neglected in urban research, or not to be examined in context. In the following, I introduce the research agenda by briefly discussing the interfaces established to date by commodity chain literature and urban research, and by outlining my specific research concept and case study. The main section offers insight into the on-going fieldwork by tracing a typical supply chain from the major highstreet retailers in Berlin back to home-based work in Istanbul. The specific focus will be on aspects of the visibility and architectural representation of each activity in its respective urban space.
6In response to the surge in increasingly complex production relations that criss-cross international borders, economics and social sciences research has crystallised around specific issues since the early 1990s – global commodity chains (GCC, Gereffi and Korzeniewicz 1994), global value chains (Gereffi et al. 2005) and global production networks (Coe et al. 2008, Henderson et al. 2002) – in order to deepen understanding both of the division of labour in the capitalist world order and of global disparities between the north and south. Such research draws on case studies in order to highlight the various stations and transactions in which the products of complex, globally organised production processes are implicated as well as to discuss both the unequal distribution of power and value creation and the production regions’ respective chances of upgrading their status within these chains or networks (Bair 2005, Fischer et al. 2010).
7As Bair (2005: 158-159) argues, “Tracing the path of a commodity – be it a pair of blue jeans sewn in China from Indian-made denim, a cocoa bean grown in Ghana and processed into a chocolate bar in the Netherlands, or the assembly of a laptop computer in Mexico from components produced in East Asia – provides a grounded way to study and operationalize the global-local nexus. The GCC method permits one to analyze globalization in situ, directing our attention to the specific locations where particular production processes occur, while simultaneously illuminating how these discrete locations and activities are connected to each other as constituent links that collectively comprise the commodity chain”.
8To trace the global circulation of commodities and thereby heighten the visibility both of the respective region of production and the various cultural contexts and interfaces encountered underway is meanwhile an established approach in cultural anthropology and cultural geography, too (Brooks 2015, Cook 2004, Knowles 2014, Miller and Woodward 2010, Pliez, 2012). While some of these studies methodologically approximate the qualitative approach of this present research, the economics-oriented and, often, quantitative case studies found in commodity chain research are fundamental to an understanding of production structures and value distribution in the clothing industry.
9The debate on “World City Networks and Global Commodity Chains” served to establish the first interfaces with urban research (Brown et al. 2010). However, inquiry here – into how global urban networks and production networks overlap in the service or logistics sectors of certain exemplary cities, for example – proceeded at the macro level (Hesse 2010, Parnreiter 2010). The systematic spatial description of production networks or commodity chains thereby “remain[ed] an unfulfilled task” (Brown et al. 2010: 19). But moves were made soon afterwards to merge these two fields of research. In “The Transnationality of Cities”, Krätke et al. (2012) describe an increasingly relevant field of interdisciplinary urban research and sketch out an agenda for the future investigation of transnational urban spaces. The integration of commodity chain research in urban research is imperative, in their view, since previous global and world cities research has failed to address the significance of industrial production in global economic relations. Another item on their agenda is to integrate in this research both physical urban spaces and various scales from the urban network to the single building. Parnreiter (2012), in his contribution to the debate, likewise argues that the transnational constitution of urban spaces should be assessed specifically in terms of its physical, architectural and planning dimensions – and then demonstrates how, in explicit reference to a case study in Mexico City.
10Accordingly, the present research project endeavours to integrate in urban research the global commodity chain approaches pursued to date and thus foster a shift in perspective. Instead of considering solely the changes wrought in a single location by globalised production, it broadens the scope by examining a typical production network, respectively several locations within it at which various steps in the production and distribution process are carried out. The case study here is a clothing industry transnational production network that interlinks fashion retailers and markets in Western Europe and various production locations on the Southeastern margins of Europe. The analysis focuses on selected stations of the network, from the production of clothing in Turkey, particularly Istanbul, to its consumption in Germany. Aware of the fact that there is no natural beginning and end to commodity chains and networks, the research deliberately limits itself to the actual steps of clothing production and distribution, which form a rather complex network already due to its dispersal to many different firms and locations. Of course, including also the production chains of raw materials, their extraction or agriculture, processing and trade, and the consumption, disposal, recycling and second-hand-market of clothes (see Brooks 2015) would create an even more complex and, definitely, more globalized picture.
11The clothing industry has relied on the spatial division of labour since the 17th century (Komlosy 2010), and it now ranks among the most thoroughly globalised sectors (Bair 2006, Dicken 2008). Since clothing manufacture is especially labour-intensive and pay therefore a significant factor, outsourcing to low-wage countries is commonplace, as are extreme variations in value creation at different points in the production chain. The clothing industry is accordingly treated in commodity chain research as a typical example of buyer-driven production relations (Gereffi 1994), since leading retailers and brands, which focus on design and marketing as their lean core business, dominate production (Dicken 2008, Klein 1999). At the lower end of the production chain one often finds badly paid and unregistered labour, operating in precarious circumstances (Brooks 2015, Dedeoǧlu 2008, Staritz, 2010). Production for the global market is based mainly in Southeast Asia, but has also boomed in East Europe and the Mediterranean over the last twenty-five years, manufacturing first and foremost for the EU fashion market, which benefits from the short lead times (Bair 2006).
12While many East European countries have developed since the 1990s primarily as low-pay locations for labour-intensive production phases (Pickles et al. 2006, Plank and Staritz 2009), Turkey produces cotton and textiles as well as clothing and has therefore managed to secure a foothold in broad sections of the textile chain and upgrade its standing within the global value chain (Evgeniev 2008, Neidik and Gereffi 2006, Tokatli and Kızılgün 2009). Since liberalising its economy in 1980, Turkey has come to rank among the world’s leading exporters of clothing. Istanbul, the heart of Turkish clothing production and trade, is a telling example both of how production networks may spread across vast areas of a city and the spatial division of labour put its particular stamp on each location (Figure 1). Yet the urban impact of the city’s clothing industry – which not even a tourist could fail to notice – is rarely mentioned in the literature (Eder and Öz 2010, Erder 2005, Tokatli et al. 2011), despite the popularity of themes such as the urban transformation, globalisation and neoliberal restructuring of Istanbul in more recent urban research (see Angell et al. 2014, Enlil 2011, Eraydin 2008, Esen and Lanz 2005, Güvenç and Işık 2002, Keyder 1999, Türkün 2011).
Figure 1. Distribution of clothing companies in the Istanbul urban area
Source: Chamber of Commerce / Kadir Has University, Istanbul Studies Center, 2009. Mapping by the author and Edin Zaim.
13The goal of the present research is to comprehend the spatial logic of a specific transnational system of production and thereby render visible, and characterise, the architecture and urban context of its individual stations, in particular by recourse to mappings and graphical analyses. The core question of the project is how the architectural and urban properties of each individual station correspond to the economic hierarchy of the production network. Moreover, it will examine the impact of globalized industrial production on the built urban environment, and, vice versa, the influence of local urban dynamics and planning on production locations. Thus, the project intends to better understand contemporary built environments and processes of urban transformation by studying the transnational economic processes and forces that shape them.
14The ideal, conceptually appealing methodology, namely to trace the entire trajectory of a single item of clothing from its raw to its consumable state, proved problematic precisely owing to the thorny matter of restricted visibility. Despite increasing demands and claims of greater transparency in supply chains (see Doorey 2011, Egels-Zandén et al. 2015), many large German fashion companies and retail chains were unwilling to reveal their supply structures for the purpose of this research project. Some production companies were open to interviews and factory visits, however, and gladly named their major clients as good references. Yet they had rather less to say on the subject of outsourcing certain steps in the production process, even though the practice is commonplace in Turkey. Evidently, to trace the trajectory of a specific item of clothing over several stations would call for considerable investigative input. To do so would be to focus, moreover, on a single, ephemeral moment within a highly dynamic and complex web of production relations. It was accordingly decided instead, to focus in this research on selected locations and players within a broadly defined production network, each of which or whom represents a step in the production and value chain as well as a typical form of urban development, even though not all of the players in this case study are connected through direct production relations (Figure 2).
Figure 2. Selected stations of clothing production and their locations in the Istanbul region
Source: the author
15Spatial analysis of the selected locations is pursued at various scales, from their overall geographic constellation and the transport routes that interlink them, to their urban setting and the way(s) they are embedded in it, to their actual architecture. The primary research tools are urban and architectonic analyses of the built environment based on drawings, for instance, in the form of morphological and typological studies, as well as functional and infrastructural analyses. Qualitative interviews and materials research likewise propel assessment of the forces and conditions driving the constructive development of the selected locations: the spatial requirements of each respective production phase, the spatial practices, interests and conflicts of the players involved, local circumstances and dynamics, and transnational economic and regulatory conditions.
16While the ongoing research project thus tackles manifold aspects and dimensions, this article gives particular attention to the specific aspect of visibility and architectural representation in urban space of the various stations of production and distribution. What is the visual impact of each individual station given its architectural idiom and urban setting? How does the position of a certain step in production within the value chain correspond with its material presence in the city and the representative function of its architecture? What is made visible? What remains hidden? Using this as a recurrent theme, the following section provides fragmentary insights into the present field work. It sheds light on some exemplary stations of transnational clothing industry – between the poles of German retail outlets and home-based workers in Istanbul – and the production networks these places are connected to, and it examines the architectures and urban situations involved. The information presented here was garnered in 2013 and 2015 from both personal observation in situ and interviews conducted with local players (producers, sourcing offices, home-based workers, urban planners) and experts (scientists, activists, inspectors, urban planners).
17The area around Hackescher Markt in Berlin-Mitte has become a major retail location for fashion and accessories over the past fifteen years. In the early 1990s, it was the new pulse of subcultural nightlife yet it soon turned into a tourism hot-spot full of cafés, restaurants and boutiques. Over the past decade these have ceded in turn, mainly to major international fashion brands and chain stores. Also, the shopping strip has stretched eastwards and now extends to Alexanderplatz. Since branches of chain stores have popped up (H&M, Mango and Tom Taylor) and brand stores have spread their ubiquitous logos in the streets (adidas, BOSS, Tommy Hilfiger, Levi’s), the area has acquired the retail profile typical of pedestrian zones and shopping malls the world over. The building structure, a small-scale heterogeneous mix of rather simple late nineteenth century houses, infills of GDR prefab-housing, and recent new construction, has been adapted to the needs of high-street fashion chain stores. In order to maximise surface area in the medium-sized houses, the original street front stores have often penetrated deep into the side wings and basements – as in the case of the H&M Group’s somewhat more exclusive offspring COS and & Other Stories. This greatest possible extension of commercial use is hardly visible from outside. Likewise, the advertisement on the façades is modest and apparently subject to official regulation. Nevertheless, fashion retail and globalized brands get increasingly visible in the streets and replace other businesses and functions, but only who enters the shop steps inside the fully designed ‘brandscape’ that make the single garment part of a lifestyle.
18When searching the clothing on sale there, one can often find the label “Made in Turkey”, at times also the vaguely more euphemistic “Made in Europe”. All the major fashion chains now follow the example set by the Swedish giant H&M and the Spanish Inditex-Group (Zara, Mango et al.), constantly ushering in new collections, so that customers can find new models at every visit (Tokatli 2008). The previously standard rhythm of two or, at most, four collections a year has disappeared. Another recent innovation is to test-run new models before embarking on its large-scale production. This system called “fast fashion” relies on short lead times and a quick response to consumer demand (Tokatli 2008). China and Bangladesh are at times too far away for lean deliveries, so production hubs in Turkey, Eastern Europe and Arabic countries have gained significance for the EU market (Bair 2006). H&M and Inditex are meanwhile the biggest buyers in Turkey and almost all major fashion chains buy at least a part of their collections there. Turkish producers have been able to consolidate their market position vis-à-vis Asian producers thanks, firstly, to fast contractual processing and short lead times and, secondly, because, as “full-package providers”, they can generally cover every step in the production process, from design to pattern making, to materials purchase, to cutting, sewing and finishing (Neidik and Gereffi 2006, Tokatli and Kızılgün 2009) (Figure 3). Fashion houses therefore often buy “ready-mades” from Turkish producers’ own collections, or request only the slight modification of colours and detailing. In the case of bulk production, however, they tend still to place their orders in Southeast Asia.
Figure 3. Typical production chain in Turkish clothing industry: The producer takes care of all production steps, but outsources the labor-intensive procedures to subcontractors
Source: the author
19In order to be able to recurrently view production companies’ collections, locate new suppliers or supervise on-site production, the major chains either maintain their own liaison office in Istanbul or work with local agencies. Many of them are in the Yenibosna district of West Istanbul, a half-hour drive from the city centre and only five minutes from Ataturk Airport. The motorway to the north of the airport is flanked to its east and west by an industrial zone that is flanked in turn by densely built residential districts. This is where A, a subsidiary of a German mail order company, has its sourcing office. In the vicinity are the liaison and sourcing offices of other German chain stores and fashion houses as well as the Istanbul reps of Hong Kong-based Li&Fung, the world’s biggest global sourcing agency. A’s sourcing office is located in a solitary rectangular, four-storey structure of blue-tinted mirrored glass that reveals nothing of its function (Figure 4). However, the great depth of the building and the six loading bays to its rear are clues to its commercial purpose: A uses its base here not only for offices but also for goods delivery and quality control. The circulation areas around the building are fenced in and the front and rear gates are guarded by a small porter’s lodge in the form of a white container of the sort common all over the city. These somewhat symbolic security arrangements, the vertical strips of mirror-glass curtain-wall – a popular feature for office buildings in Turkey, also due to its sun protection capacity –, and the absence of a logo or firm sign, make the building appear representative but not very welcoming.
Figure 4. A’s sourcing office is located in a rectangular building of blue-tinted mirrored glass that reveals nothing of its function
Source: the author
20In the surrounding buildings of a similar size and shape there are numerous clothing production companies. Many of them have outsourced sewing to subcontractors and now deal only with certain tasks in-house: cutting, quality control and packaging, for example. Machinists’ workshops are increasingly being pushed out of Yenibosna. Being relatively central, close to the airport and convenient on account of its high concentration of clothing production companies, it is an advantageous location for sourcing offices and company headquarters. While certain buildings still evince the simple architecture of a modest factory or warehouse with concrete or rendered façades, others are clad in glossy or tinted-glass panels, too, so as to more emphatically underline their upgraded status in the production hierarchy. Evidently, the industrial area is under pressure to open up to more profitable activities. Luxurious high-rise apartments, office towers and an outlet shopping mall have shot up, lately, on the neighbouring former industrial sites.
21The counterpart to the sourcing and liaison offices of the generally West European fashion companies are the showrooms of the Turkish production companies and wholesalers. As a rule, the bigger production companies have their own design department and produce their own collections, which they show in their headquarters to buyers. In order that clients can easily reach them, they too have their premises in Yenibosna, Güneşli or Merter, districts of West Istanbul formerly dominated by manufacturing. One example is B, a producer of high-end circular knitwear headquartered in Güneşli, some two km to the north of company A’s sourcing office. Among B’s clients number the H&M subsidiaries COS and & Other Stories, premium brands such as Armani, and smaller fashion houses from Germany and West Europe. Textiles and models for B’s own collection are developed in Güneşli and presented in the in-house showroom. Pattern making takes place here and likewise the planning of everything from yarn purchase to textile knitting and dyeing, to sewing and eventual finishing. Knitting and dyeing are outsourced to various subcontractors in or near Istanbul, all the cutting is done in Güneşli, and the sewing, quality control and packaging are carried out in a small, largely rural town in the Tekirdağ region, some 180 km to the west of Istanbul, partly in the company’s own factory, partly in those of a few exclusive subcontractors.
22A few years ago, B decided to sell his own headquarters building in an industrial area near Güneşli – probably with a high profit – and to outsource more functions to their production facilities in Tekirdağ. In addition to the management, the design and modelling department, and spaces for cutting, storage and the quality control of yarns and textiles, B has held onto one small production unit in his Güneşli facility. The company now rents 4500 square meters on several floors in a detached, four-storey commercial building, a largely abstract big box, which is set back from the street and surrounded by unbuilt areas (Figure 5). The landlord’s firm, a wholesaler of bathroom fittings and accessories, has plastered the main façade with his product range logos, while B uses an entrance on the side of the front façade which is marked by his company sign. Additionally, he positioned a signboard next to the access at the street. The building with the parking lot in front thus evokes the image of a suburban superstore.
Figure 5. The clothing producer B rents 4500 square meters in a box-shaped commercial building in Güneşli
Source: the author
23In the main, the heterogeneous industrial area is the preserve of factories, company headquarters and wholesale markets, with smaller factories and workshops dotted about the modest residential district on its eastern flank. The vast array of outlet stores along the street attests the high concentration here of clothing and footwear producers. At the same time, some large unbuilt areas, brand new apartment and shopping complexes in the neighbourhood are indicative of land speculation and urban upgrading.
24In general, the major West European companies leave it to their liaison office or a local agency to make contact with Turkish export companies. By contrast, buyers from other regions mainly place orders for limited quantities of stock with the smaller producers and traders operating in the centrally located wholesalers’ districts, Merter, Laleli and Osmanbey. Here, one finds an increasingly professional version of the “suitcase trade” first established in Laleli in the 1990s (Eder and Öz 2010, Yükseker 2002, Piart 2012), which today mainly targets small retailers and traders from East Europe, the post-Soviet countries, the Arab world and Africa. Merter will serve us here as a prime example of a district that clearly bears the stamp of the clothing industry’s recent evolution. The industrial zone that was built here in the early 1980s, on open countryside to the west of Istanbul’s historical city limits, was officially planned as a manufacturing and wholesale hub of the furniture industries with a dense and regular street network. Its three- to five-storey commercial units have been built by individual companies, packed in side-by-side, on narrow, deep plots.
25Turkey’s market liberalisation and export orientation as of 1980 triggered a boom in its clothing industry that subsequently left its mark on Merter, which by the early 1990s was as a thriving hub of clothing production. Many Turkish clothing producers started their business here (such as the aforementioned knitwear producer B), but with their expansion they left the district or moved at least a part of their operations elsewhere. In addition, over roughly the last fifteen years, fashion wholesale has increasingly pushed out manufacturing. Nowadays, on the streets in the southern half of the district, one mostly finds showrooms, set up either in converted factories and warehouses or in new buildings and giving rise thus to a specific, new, extrovert architecture. In the northern half of the district, the main roads are given over to wholesalers both of fabrics and accessories, such as buttons, studs and zippers, the less prestigious side streets to small factories and warehouses (Figure 6).
Figure 6. Large parts of the former production district Merter have turned into wholesale areas for accessories and clothing
Source: the author
26C has quit the textile sector at the end of the 1990s, and turned towards real estate and construction business. He transformed his former factory building in the showroom-district of Merter into a çarşı, a bazaar, by inserting a passage into the building on three floors. Now, the interior is organised like a shopping mall, in which many small traders and producers each rent a small surface area on which to present their latest lines. On the street, these types of buildings vie for attention with extravagantly designed or fully glazed façades and numerous logos for aspiring wannabe brands, often imitating their more established counterparts. The eclectic styles of the façades include historical references, trivial decorations and more contemporary designs. Like many other building owners, C has made the basement accessible through broad staircases in front of the building to enlarge showroom space with direct sight contact with the street (Figure 7). These changes and also the new construction of bigger and more spectacular wholesale malls are the visible effects of the rising demands and real estate values.
Figure 7: C has transformed his former factory in Merter into a wholesale mall
Source: the author
27C is chairman of a local association of businesspeople, which is striving to turn Merter into a leading global centre of fashion and wholesale, one that will be able to hold its own against London and Milan. Already, banners on the streets herald the imminent re-branding of Merter as a fashion quarter. The upgrade of public spaces is another step in this strategy. Merter’s businessmen convinced the Greater Istanbul Municipality to initiate a top-level urban design project, which currently turns the streets of the southern showroom district into a properly designed pedestrian zone. Besides, municipality and the commissioned architects are not pleased with the wild potpourri of the street façades, in particular, the uncontrolled logos, billboards and air condition units which were individually attached (Figure 8). So they are developing guidelines and proposals for the adjustment of each individual façade, although it will be hard to convince showroom owners to actually follow these specifications.
Figure 8. The buildings in Merter’s wholesale district vie for attention with logos and extravagantly designed façades
Source: the author
28Thus while buyers, headquarters and wholesalers are primarily to be found in Istanbul’s more central districts, the actual production of clothing, first and foremost, the task of sewing, has been moved largely to the outskirts of the city or even further a-field. First-tier suppliers either outsource production to subcontractors or operate their own factories, preferably in regions with low-cost labour. Thus, the multinational group of textile companies D, who only recently moved its headquarters from Merter to the more peripheral district of Beylikdüzü, has various factories in the western hinterland of Istanbul but outsources the bulk of its production to its own subsidiaries in Bulgaria, Jordan and Egypt (Figure 9). Since the early 1990s, numerous large textile and clothing producers settled in the hitherto mainly agricultural region of Tekirdağ, to the west of Istanbul. The area between the towns of Çorlu and Çerkezköy, conveniently located by the main EU-Istanbul motorway and barely 100 km from the centre of Istanbul, has since become a major industrial hub, because, in addition to the good transport links, land for development was available, gas, water and electricity were subsidized, and wages were lower than in Istanbul. The area particularly appealed to companies such as dyeworks and jeans manufacturers, which require lots of water and need to dispose of their highly polluted wastewater. As these dirty industries were gradually pushed out of Istanbul, many settled in the region around Çorlu in a rather uncontrolled way, and started to contaminate the rivers in the area.
Figure 9. The D group has built up a transnational network of companies and factories
Source: the author
29The D group runs a gigantic yarn, knitting and dyeing factory on 45 hectares to the west of Çorlu, plus another textile dyeing and printing factory nearby. Like many other large textile companies in Turkey, D meanwhile owns a string of businesses also in other profitable sectors, such as tourism, construction and energy. A further operation is its free zone management. In the late 1990s, the group acquired some 200 hectares of agricultural land near Çorlu, alongside the EU-Istanbul motorway, and then requested the Turkish government to designate this a free trade zone. It thus effectively procured comprehensive customs and tax exemption for any company engaged there in export production and trade. Construction of this European Free Zone was launched in 1998 on the basis of an urban development plan drawn up by a big American architecture firm. Officially inaugurated in 2003 by Prime Minister Erdoğan, the free trade zone provides companies with a broad range of building plots, office spaces, industrial hangars, warehouses and services as well as a comprehensive technical infrastructure.
30Due to initial tax-exemptions for income taxes, the D group moved large parts of its labour-intensive clothing production here, but when these exemptions expired, most of it was relocated to Jordan and Egypt instead. One clothing factory of D, however, stayed in the free zone, which now accommodates a mix of international firms from electronics, automobile and chemistry sectors. The subsidiary of D produces men’s trousers and suit jackets of woven fabric for clients in Denmark and the UK. Its factory was built in 1999, with a second storey added later, owing to expansion. The two hangar-like factory floors accommodate the entire production process, from cutting to packaging. Every day, more than 600 workers – bussed into the zone from the surrounding villages and towns – produce as many as 3,500 items of clothing.
31The factory is constructed of precast concrete elements, its wall panels are painted pink on the outside and pierced by regular patterns of square and circular windows of blue-tinted glass with dark red frames. The entrance and office section is accentuated by a curtain-wall of blue-tinted mirror glass and a company sign next to it. In the clean and neat external area, only the pink ashtrays next to the doors, and a roofed area with wooden benches in a setback of the façade attract attention and point to its use during tea breaks. The factory looks simple and modest, but well-kept. There’s an obvious intention of a proper appearance, despite the slightly awkward colour and the unambitious design (Figure 10). Likewise, the architectural form and appearance of most buildings in the free zone is dominated by a functional structure and cost-effective building materials, such as precast concrete or corrugated sheet panels, which usually makes them look like generic boxes that you can find in industrial zones all over the world. The major difference, though, is that the free trade zone is completely fenced in and a customs office monitors the sole access road. The zone is bordered to its south by the motorway but otherwise surrounded by fields. The nearest villages and industrial zones are several kilometres away.
Figure 10. The clothing factory of the D group in the European Free Zone consists of two stacked hangars
Source: the author
32Not all of the clothing factories in the Tekirdağ region are truly typical of the Turkish clothing industry, since many of them are vertically integrated, i.e. their workers accomplish a variety of production tasks under one roof. Thus, a single location, like D’s suit factory, can produce complete garments. Producers here often deny they ever outsource, or claim to do so only exceptionally. But it is often hard to believe they completely forego the evident benefits of being able to fall back, as and when necessary, on the subcontractors so ubiquitous in the Turkish supply chain (Evgeniev 2008, Neidik and Gereffi 2006, Tokatli et al. 2011), given that to do so would enable them to respond more flexibly to shifts in demand. One labour rights advocate describes the major first-tier suppliers’ factories as the “shop windows” of the Turkish garment industry. For while only half a million workers are officially employed in the clothing and footwear sector, the Clean Clothes Campaign speaks of an additional one and half million unregistered workers (Seckin et al. 2014).
33The vast majority of this informal labour is employed in small businesses – commonly called ateliers (atölye) in Turkey – or they are home-based workers. It is common knowledge that the Turkish clothing industry is a modular system comprised of numerous, small, specialised operations, and that its competitiveness rests on the backs of countless underpaid workers deprived of social benefits (Dedeoǧlu 2008, Evgeniev 2008, Tokatli et al. 2011). Many of these are poorly educated people from rural regions of Anatolia, but also, there is an increasing number of migrants from other countries, especially Syrian refugees, who find informal work in Turkey’s clothing industry (Johannisson 2016).
34Officially, West European buyers stipulate that work cannot be subcontracted without their express consent, since the quality and social standards of every company involved in a supply chain must now be certified. The manager of the sourcing office A indeed acknowledged that each of their Turkish suppliers works with at least three subcontractors. Yet how legitimate subcontractors really are, is hard to say: auditors mostly lack the capacity to thoroughly inspect all the official ones and the unofficial ones remain difficult to pin down. Many such small second- or third-tier sewing operations can be found in working-class and outlying districts of Istanbul, and also increasingly near the Syrian border in Southeast Anatolia, in towns like Batman or Karamanmarash.
35The clothing producer E represents a more typical case in this small-scale modular system. He has specialized on embellished jersey products for German and other West European clients. His headquarters and showroom are located in Merter, where he deals only with cutting and finishing, and outsources all other operations to subcontractors (Figure 11).
Figure 11. A production chain connecting the clothing producer E in Merter and the sewing atelier F in Sultangazi
Source: the author
36The company buys dyed fabric from a factory in Çorlu, for example, makes the cutting in-house, sends the cut parts to a printer’s workshop in Güneşli, sends it afterwards for embroidery to the workshop of E’s brother, also located in Merter, and then sends all the parts to the sewing atelier F in Sultangazi. Afterwards, the pieces are being checked, ironed and packaged back in the headquarters and then directly shipped to their clients (Figure 12).
Figure 12. The clothing producer E uses only the top and basement floors of his building in Merter
Source: the author
37Two of E’s sewing subcontractors are located in Sultangazi, a workers’ district at the northern edge of the city. Since the clothing industry continues to swallow up a large amount of Anatolian migrants in the city, it is no coincidence that small subcontractors and ateliers have expanded precisely in the districts they inhabit, first in the once illegally built and meanwhile densely overbuilt gecekondu districts, such as Zeytinburnu, Bağcılar or Sultangazi (gecekondu literally means “houses built overnight”, millions of which were thrown up between the 1960s and 1980s by migrants newly arrived in Istanbul from rural Anatolia). Sultangazi started to develop in the late 1970s. Today it is a dense district of half a million inhabitants, and majorly characterized by a narrow street grid with four- to six-storey houses of reinforced concrete and brick, thrown up cheaply, as a rule. Its Sultançiftliği neighbourhood extends on both sides of a high street with a major tram line.
38The high street is mainly lined by commercial buildings with six or seven floors (iş hane in Turkish) (Figure 13). While there is usually retail shops or restaurants on the ground floors, the upper floors often reveal their function through banners and ads on the façades (“Merve Tekstil”, “Hira Tekstil”, “Favori Tekstil”) with the list of jobs they offer (operators, quality controllers, ironers), and the services they provide (meals, bus service, social security), some ads also list the daily payment (€14 for unqualified, and €22 for qualified workers). Indeed, a large part of the commercial buildings seems to be occupied by sewing ateliers and related workshops. According to enquires by Clean Clothes Campaign Istanbul and the author’s observations, working conditions in many of these places are rather bad, there’s a lot of unregistered workers, including Syrian refugees and underage persons.
Figure 13. Sultangazi’s main street is lined by commercial buildings with the ads of sewing ateliers on their façades
Source: the author
39The side streets, in contrast, are more dominated by residential buildings. Nevertheless, they are also interspersed with small companies on the ground floors and in the basements, some of which are visible through their signs, open doors or a glimpse of harsh neon light through the windows, and audible by the rattling machines and popular Turkish music. E’s subcontractor F, a small family business, has been renting a workshop here since four or five years. Several brothers and sisters had moved to Istanbul in 2007 and soon decided to establish their own business instead of being workers in other sewing ateliers. Their workshop is located in the basement of an apartment house, which is not completely underground, due to the building’s location on a slope. So, the atelier is directly accessible from the pavement by a few steps and can get some fresh air through basement windows. Here, usually 20 people are working in one production line with around 10 sewing tables and the preparatory procedures. All workers are living in the neighbourhood, half of them are family members. The typical Istanbul apartment house is clad in cream white and light blue mosaic tiles, its four upper floors protruding over the pavement. The atelier becomes visible from the street by grey window grilles, a grey metal door, and, most apparently, a big red banner above the door: “Looking for workers: overlock operators, quality controllers, cleaners”. Although some other workshops may be deliberately hidden from sight by plastic film or coating of paint applied to window panes, clothing industry is conspicuously present here, owing to its high concentration in this neighbourhood (Figure 14).
Figure 14. The sewing atelier F is located in the basement floor of a five-storey apartment building in Sultangazi
Source: the author
40Because the location in the basement is so typical for small sewing workshops in Istanbul, the Turkish term merdivenaltı for “below stairs” has established as a synonym for informal production. Although these are usually legal companies, many of them hire a part of their workers unofficially to save social security costs. Dedeoğlu (2008) has shown how heavily Turkish export production relies on the unregistered labour of women in particular. This huge yet largely invisible reservoir of cheap labour primarily works in local family businesses or at home, not infrequently because traditional social codes do not countenance regular employment for women beyond the family circle. It is often middlemen or middlewomen who organise local networks of home-based workers on behalf of the neighbourhood’s ateliers. The tasks carried out there, almost the last in the production chain, include cutting off loose threads or the manual embroidery and embellishment of otherwise complete garments.
41The group of home-based workers organized by G has rented a space of its own, so as to be able to work together (Figure 15). These women live on the northern margins of the Ümraniye district, on the Anatolian side of Istanbul. The urban structure here, mainly one to three-storey detached houses, evokes a village or small-town ambience. Between this district and a motorway one finds a cluster of typical gecekondus; on the far side of the motorway, by contrast, luxury housing complexes sprang up more recently on former forest land. The first houses in this neighbourhood were built in the late 1980s by working-class families, and many have been successively extended, vertically or horizontally. G’s group has a workspace on the ground floor of a freshly renovated five-storey apartment building accessible directly from the street. The ground floor façade, however, maintains unfinished with raw brick walls and some uneven plaster. A downpipe is running diagonally across the wall. New windows have been inserted in the front and side walls to transform the former fuel storage into a workspace.
Figure 15. Home-based workers have rented this workshop in the north of Ümraniye district to be able to work together
Source: the author
42Inside, the concrete ceilings and girders are exposed, the brick walls unplastered, and the PVC-flooring is meagre and makeshift. The centrepiece is a small wood-burning stove, with old sofas and armchairs clustered around it. Some ten to twenty women work there. All of them are middle-aged and live no more than fifty metres away. They don’t see themselves as workers but as housewives out to make a little pin money or assure their children’s education. They work for three knitwear factories located in the Dudullu industrial zone in the south of Ümraniye. All three factories produce for export, inter alia for international chain stores such as Zara, as well as for German retailers.
43G is the middle-woman: it is she who organises the factory orders and allocates work to the other women. Mostly, they are engaged in sewing seams, cutting off loose threads, and sewing on buttons or other appliqués. The women are paid on a piecework basis (EUR 0.03–0.06) and make up to €6.80 a day. The finished pieces are returned to the factory, ironed and packaged, and then delivered to the client. The women are worried to think their unregistered labour might come to light yet, at the same time, they agree that consumers of clothing should know where it was made and by whom. This is why many Turkish home-based workers have been organising over the last ten years, campaigning nationwide – in the newly founded trade union Imece, for example – for greater visibility, legal status and better working conditions.
44The future focus of this research project, as outlined in the research agenda above, will be the systematic, spatial analysis of selected locations typically found in the clothing production and export trade, specifically, of the respective architectures and urban settings, their development dynamics and the forces that shape them. Although it is possible to identify typical places for certain stages of production and their locational factors, field research to date has not given rise to a simple hypothesis about the correlation between the economic hierarchy and the building structures. With regard to the various players in the supply chain, their company size and geographic scope, one clearly has an inverted pyramid, with hugely profitable, globally active retail chain stores and brands at the top, followed by large Turkish garment producers, that operate on a national or regional level, then smaller ateliers and individual home-based workers who are limited to the district or neighbourhood scale. Such nuances are certainly reflected in the architectures and urban settings of production locations: the value of land and building investment obviously declines along this gradient (although not straight proportionally), from top-ranked Berlin retail areas to poor working-class neighbourhoods of Istanbul, so does the building quality, from high-end design to make-shift solutions.
45However, one cannot come to a straight conclusion about the centrality, urban grain, or degree of planning at different levels of the production chain. Indeed, there is an obvious difference of urban grain between the large-scale factory and the small workshop, but not so much between the retail store and the atelier. In contrast to the European Free Zone, the workshops sprang up in an unplanned, bottom-up process, but so did many other big textile factories in the industrial areas around Çorlu.
46It turned out to be easier, in comparison, to describe the change of production locations caused by the dynamic clothing industry, on the one hand, and by local urban development, on the other. The actual manufacturing of clothing is gradually pushed out of Istanbul’s central districts by planning efforts, rising real estate and labour costs, while the Turkish industry continues to upgrade and starts outsourcing to other regions and countries. In the individual locations, this may be reflected by the upgrade towards higher-rank production stages (as witnessed in Merter), by the displacement of the industry through other functions (as seen for example in Yenibosna), or by the influx of clothing companies that have relocated from other places (as the examples of Çorlu region and Sultangazi have shown).
47But instead of providing comprehensive answers to the bigger research questions, this article has focussed on the visual impact and material presence of the selected production stages within urban space. In the beginning, there was a self-evident working hypothesis regarding the visibility of various production tasks. Just like the incidence of unregistered labour generally goes hand in hand with the diminished visibility of operators and workers both in buyers’ supply chains and official statistics, the physical presence and visibility of built spaces – and also the representative function required of architecture – decreases proportionally, too, from the flagship stores of international fashion retailers at the “top” end of a value chain to the ateliers and home-based workers at the “bottom” end. That the biggest slice of the value creation pie is generated by branding and marketing is evident from retail stores’ conceptual brand-image designs as well as from the efforts made by wholesalers in Merter to create a brand identity for their products and then emblazon it for all to see on a personally customised façade. At the other end of the production chain, precarious work is hidden in urban leftover spaces and niches.
48But things turned out to be more complicated. The actual visibility and material presence of certain production locations is depending on many individual aspects and cannot be explained with a simple model. Merter’s wholesale, for example, is much more strongly affecting the streetscape than the Berlin retail stores (also due to different building regulations and the extent of their enforcement). On the other hand, Merter’s hysterical façades are rarely seen by ordinary passers-by, because it is majorly business people who come into this area. Laleli, in contrast, another wholesale district located in close vicinity to Istanbul’s tourist hot spots, is of higher visibility for a general urban public.
49Compared with the wholesale’s strident presence in urban space, the architectures of the described sourcing office and producer’s HQ tend to be more discreet in character, representative and yet cool – qualities that are often expressed in Turkey by means of mirrored or tinted glazed façades. Their mainly confidential trade relations don’t require urban visibility, clients are usually brought here by the firm’s own drivers. Their showrooms are open only to invited guests and it is in the interiors that the actual representation takes place. The huge industrial zones and clothing factories on the outskirts of Istanbul, for their part, may be hard to overlook from the motorway and yet, for the urban public, they are a virtually invisible presence, banished to the margins of the city. For auditors, however, these proper factories are often the “shop-windows” of Turkish clothing production which they may certify and inspect in detail.
50Meanwhile, at the lower end of the supply chain, where one would expect more clandestine situations, the sheer concentration of subcontractors’ workshops creates a strong and visible presence in urban space. But again, most inhabitants or visitors of Istanbul are not aware of this presence, as they hardly ever come to Sultangazi. Thus, the most visible means of communication here, the ateliers’ ads, are majorly addressing the potential workers who live or work in this area.
51For many home-based workers, invisibility is actually a condition of their employment in the clothing industry due to the traditional family structures which restrict women to the domestic realm. Despite this lack of visibility at the “bottom” end of the production chain, access for research purposes appears to be sometimes easier than at the “top”. While globally active fashion companies remain largely unwilling to show their hand, many home-based workers are obliged to fight for greater visibility.
52Even though the cases presented here do not clearly indicate a coherent “visibility hierarchy” that parallels the hierarchy of the value chain, nearly all of them yet reveal, how noticeably clothing industry has shaped urban spaces: through new constructions, some of huge dimensions, as well as through the visible transformation of single buildings and entire neighbourhoods. But the examples also show that the question of visibility always depends on the viewer and his or her viewpoint. Just as the supply chain is not equally visible for different actors involved – the consumer, the researcher, the fashion firm, the sourcing manager, the producer, the worker – its different stations and their material presence in urban space certainly become visible only for those who have a reason to actually visit these places. To examine more deeply the remaining contradictions and the questions raised will be the task of further research.