how do italian footwear industrial districts face

HOW DO ITALIAN FOOTWEAR INDUSTRIAL DISTRICTS FACE GLOBALISATION?

Alessia Amighinia and Roberta Rabellottib a

United Nations Conference on Trade and Development (UNCTAD) [email protected] b

CORRESPONDING AUTHOR

Department of Economics and Quantitative Methods Università del Piemonte Orientale Via Perrone 18, 28100 Novara, Italy Ph. +39.0321.375317 - Fax +39.0321.375305 [email protected]

(forthcoming in European & Planning Studies) Abstract This paper examines the impact of globalisation of production on the Italian footwear sector and investigates the structural changes taking place in some shoe districts as they join international production networks. The following questions are discussed: Are Italian footwear districts specialising in particular phases of the production cycle? Is there a trend towards the reduction of activities carried out within districts? Or are different patterns emerging according to the districts’ main market segment and to the value chains (e.g. luxury fashion market or mass market) they belong to? The study explores these issues using data on outward processing trade (OPT) collected by Associazione Nazionale Calzaturieri Italiani (ANCI) to analyse the fragmentation of production in the footwear sector at “provincia” level. The available disaggregation of data allows us to investigate the different outsourcing strategies and emerging trends within the district. Two case studies are presented; one on Riviera del Brenta in Veneto and the other on Barletta in Puglia. In the footwear districts investigated, we find evidence of different international delocalisation strategies. We argue that these different patterns of specialisation are closely related to the clusters’ market position and suggest that these patterns influence the clusters’ potential for future competitiveness.

1

1. Introduction1 This paper is concerned with the effect of globalisation of production in the Italian footwear sector and is aimed at investigating the changes occurring in the organisational structure of some shoe districts as they enter into international production networks. International fragmentation of production processes, i.e. the splitting of the production cycle over production sites located in different countries, as a cost-reducing strategy, is one of the most pervasive and disruptive effects of globalisation in the footwear industry, as in many other traditional manufacturing sectors. Fragmentation allows producers to take advantage of differences in factor prices among countries, thereby obtaining a reduction in costs by setting up an international production network (Jones and Kierzkowski, 2000). Many Italian industrial districts are deeply affected by this phenomenon, mainly those in traditional manufacturing sectors (such as textiles, clothing, apparel, leather, leather goods and footwear), because firms are facing increasing competitive pressure from low labour-cost producers in emerging economies (Corò and Grandinetti, 1999). Nevertheless, the outsourcing of low value-added activities is not the only form of globalisation taking place in the Italian footwear industry. The dominance of large fashion groups, which are increasingly transnational corporations, and the growing concentration in distribution are also responsible for changes in the traditional organisational structure of many footwear industrial districts. Overall, these trends are urging to reduce the range of activities carried out within the districts. On the one hand, empirical evidence at national level shows that increasing international competition is spurring a massive fragmentation of production processes through outsourcing of low value-added 1

The authors would like to thank ANCI – Associazione Nazionale Calzaturifici Italiani- for kindly providing the data on outward processing trade. Drafts of the paper were presented at conferences at the University of Modena and ICE (Istituto Italiano per il Commercio Estero) in Rome. We acknowldege all the useful comments received in both occasions as well as the remarks provided by two anonymous referees. Financing from PRIN 2003 - Capabilities dinamiche tra organizzazione d'impresa e sistemi locali di produzione - is gratefully acknowledged.

2

activities abroad (mainly towards Eastern European countries) (ANCI, 2001 and 2003). On the other hand, evidence from a previous study of one of the most important Italian footwear clusters – the Riviera del Brenta – suggests that local producers’ control over key activities such as design, branding, marketing and sales shrinks when they enter into global fashion production networks (Rabellotti, 2004). This paper addresses the following questions: Are Italian footwear districts specialising in some particular phases of the production cycle? Is there any common trend towards a reduction of activities carried out within the districts? Or, instead, are different patterns emerging according to districts’ main segment of market and according to the value chains (luxury fashion market or mass market) they belong to? These issues are explored using data on outward processing trade (OPT) collected by Associazione Nazionale Calzaturieri Italiani (ANCI) to analyse the fragmentation of production in the footwear sector at “provincia” level. The available disaggregation of data enables us to investigate the differences in outsourcing strategies emerging at district level. Two very different districts are studied: one, Riviera del Brenta, in Veneto, specialising in the luxury segment of the market, and the Barletta district of Puglia, which is oriented towards the lowprice market. We provide further empirical evidence in order to see whether specialisation in different segments of the market is in some way related to different forms of internationalisation, exploring both outsourcing strategies and linkages with global production networks. The paper is organised as follows. Section 2 describes the data and methodology used and analyses the outsourcing strategies of some footwear districts in Italy. Section 3 discusses and compares the two case studies – Riviera del Brenta and Barletta – which represent two contrasting patterns of internationalisation. Section 4 contains the paper's conclusions.

2. Fragmentation of production in the Italian footwear sector

3

Of the traditional “made in Italy” sectors, the footwear industry accounts for around 3% (and falling) of national exports. In Italy, footwear production is typically organised in specialised industrial districts, highly concentrated in a few regions: Veneto in the North-East, Tuscany and The Marches in the Centre, Campania and Puglia in the South (Rabellotti, 1997). In the past two decades, the footwear industry – like other traditional labour-intensive sectors – has been undergoing major global changes, as emerging economies have gradually displaced industrialised countries as producers of labour-intensive goods. As a result, industrial countries are slowly losing their comparative advantage in the production of such goods. Rising competitive pressure from emerging low labour-cost countries has forced the footwear sector in Italy to specialise increasingly in the high quality segment of the market (vertical product differentiation)2. Over time, this strategy has widened the quality gap with the exports of other industrialising countries operating in the same sector (Chiarlone, 2001; De Nardis and Trau, 1999)3. Consequently, the quality of exported shoes (measured by the unit value of exports) has increased over the last few years (Figure 1). Figure 1 here Moreover, to counter competitive pressure from low labour-cost producers many Italian footwear firms have recently increased their outsourcing of low-value added activities, making then part of international production systems. As a result, footwear production has become increasingly fragmented across different countries. International outsourcing of production can take different forms: when all segments of a production process are delocalised, foreign direct investment (FDI) and international subcontracting are the more common ways to internationalise4, while when firms find it technically and/or economically more advantageous to outsource only certain phases (e.g. pre-assembling/intermediate or 2

Vertical product differentiation as a competitive strategy was originally modelled by Flam and Helpman (1987). A similar strategy has been undertaken in other traditional “Made in Italy” industries. 4 For instance, this is what happens in the textile industry, characterised by high levels of production automation in most segments of the production process (Baldone et al, 2002). 3

4

assembling/final phases), this gives rise to fragmentation of production across countries, i.e. to outward processing (OP). This paper focuses on the latter form of international delocalisation of production, i.e. on international outsourcing operations.5 We analyse the international fragmentation of production in some of Italy's main footwear districts in order to investigate differences in their competitive strategies and emerging patterns of delocalisation of production according to the international production networks and the different market segments they belong to. The analysis is based on Outward Processing Trade (OPT) data at the provincia level, which record trade flows as temporary exports and imports.

2.1 An overview of OPT data Since 1996, ANCI, the Italian footwear association, has collected data on outward processing trade. The data available on regions and provinces (“provincia”)6 provide information about bilateral flows to each partner country. Until the first half of 2001, the source of these data was the Ministry of Foreign Trade; since July 1st 2001, the Customs authorities have been in charge of reporting on temporary trade flows under the special OPT regime. This has implied several changes in data availability and comparability. The biggest change relates to the quality of the available data: the ‘old’ source provided data on the estimated amounts of OPT operations, whereas the ‘new’ registers actual amounts. Under the old system, data was collected on the number of imported uppers and shoes, the value of exported production for processing and the value of processing outsourced abroad; since July 2001, data are much more detailed and include both the quantity and value of export and import flows of shoes

5

The other form of international delocalisation (FDI and international subcontracting involving all the segments of the production process) is also a cost-reducing strategy in traditional sectors, as clothing and textiles, but it is not very common in the footwear sector. 6 The districts considered are located in the following provinces: Riviera del Brenta in the provinces of Venezia and Padova, Verona in the province of Verona, Montebelluna in Treviso, Fermano in Macerata and Ascoli Piceno, Barletta in Bari and Foggia, Salento in Lecce.

5

and components (not just uppers, but also heels, soles, accessories, etc.). However, these major improvements have created a problem of the comparability of the old and the new data set. Because of this, our analysis reviews the major trends in the last half of the 1990s using the old data and goes on to discuss the latest available evidence in greater detail.

2.2 Empirical analysis This section analyses the international delocalisation (ID) strategies of some of the main Italian footwear districts: Riviera del Brenta, Verona and Montebelluna in Veneto; Fermano in the Marches; Barletta and Salento in Puglia. Table 1 summarises certain stylised characteristics of the districts investigated. Table 1 here The districts above are the ones that rely the most on ID within the Italian footwear sector: in 2001 they together accounted for over 85% of the import value of Italian OPT and almost 87% of Italian OPT export value (Table 2). The analysis is based on the following indicators: the degree and pattern of ID, the quality of outsourced goods and the relationship between ID and export performance. Table 2 here Degree of ID The degree of ID is measured by the value of outward processing as a fraction of the total value of exported production for processing (Figure 2). Among the districts investigated, in the period from 1996 to 2000, Barletta registered the highest percentage of value added abroad; particularly in 2000, more than 70% of the value was added abroad. On the other hand, all the remaining districts had much lower percentages (around 30% or less). Overall, there is no clear-cut trend towards an increase or a decrease in the value of outsourced processing with respect to intermediate production

6

carried out at home, except in Verona, which shows a slight but steady increase. Figure 2 here Pattern of ID Looking at the different phases of the production process, one can distinguish between two main outsourcing strategies: the subcontracting of intermediate phases (measured by the importance of components and parts in the OPT) and the subcontracting of assembling and final processing phases (measured by the importance of shoes in the OPT). Figure 3 shows the different types of delocalisation strategies in the districts investigated in 2001. In Barletta and Salento shoe firms focus heavily on outsourcing the intermediate phases of production and are therefore strong net importers of parts. Barletta ranks first in outsourcing of intermediate processing (more than 2,300 thousand kilos of imported parts), with a unit value of imported parts which is the lowest among the districts under analysis (6.6 €/kg), suggesting that Barletta is pursuing a strategy of massive delocalisation of intermediate phases and imports of low quality parts. It is worth adding that in the outsourcing of final processing Barletta ranks last among the districts considered (Table 3). Table 3 here Figure 3 here In Veneto, the footwear districts considered are all net exporters of parts and net importers of finished goods. Verona stands out as being the first district by number of parts exported (3,700,000 kilos) and shoes imported from foreign subcontractors in 2001 (more than 4,400,000 pairs). Montebelluna also outsources a great volume of assembling operations abroad (900,000 kilos), importing almost 700,000 pairs of shoes. Finally, Brenta and Fermano do not rely so heavily as the previous districts on international delocalisation in both intermediate and final processing as a competitive strategy.

7

If one analyses the partner countries in intermediate processing, there appears to be a rather high concentration of imported intermediate items from a group of Eastern European countries (Table 4). In Veneto, the preferred partner country for intermediate processing is Romania, while in Puglia it is Albania. By contrast, in the Marches, the composition of partner countries for intermediate processing is much less concentrated, with Romania and Bulgaria appearing the two preferred destinations for outsourcing. In final processing (Table 5), Romania is by far the preferred location of footwear firms in Veneto. Montebelluna mainly outsources to Romania and also, even if not continuously, to Albania and Croatia. Recently, it also begun delocalising to Asia, as confirmed by other recent studies (Parolini and Visconti, 2003). Riviera del Brenta used to delocalise towards Ukraine in the mid-1990s. It then switched to Romania in 1999, importing more than 75% of pairs of shoes as OPT in 2000. Verona has mostly delocalised final processing mostly to Romania, except in 1997 and 2000, which suggests that it might have started to diversify its partner countries mainly in favour of Ukraine. In the Marches too, the preferred partner country in final processing is Romania, but recently there have been increasing imports of finished goods from Hungary, which now accounts for more than 55% of the pairs of shoes imported by Fermano. In Puglia, where outsourcing of final processing is less relevant with respect to intermediate processing, the favourite partner country for imports of finished goods remains Albania. A final general consideration drawn from the data reported in Tables 4 and 5 regards the wide variation in the amount of production delocalised to each country, from one year to the next. These sudden changes are confirmed by recent studies on the delocalisation strategies of Italian firms (Mariotti, 2003). It may be concluded that Italian firms tend to adopt “soft” forms of internationalisation, such as subcontracting or trade agreements, in order to minimise risks and sunk costs. Tables 4 and 5 here

8

Relative quality of OPT The literature on international trade points out that one of the main effects of globalisation is greater international competition achieved by improving the quality of exports: old-industrialised countries would upgrade by increasing the quality of their exported goods, in order to reduce the competition with those exported by developing countries (Flam and Helpman, 1987). This type of upgrading can be pursued either by leaving the production of low-quality goods to other countries or by outsourcing low value added phases of production abroad, maintaining high-value added phases at home7. As a matter of fact, the long-standing Italian specialisation in traditional manufacturing sectors – including the footwear sector – notwithstanding the increasing competition by emerging economies has been possible through the quality upgrading of exports and the delocalisation of some production stages abroad (Baldone et al., 2002; Chiarlone, 2001; De Nardis and Trau', 1999). A common index used to measure the relative quality of exports and imports is the Quality Index (QI)8. Given data availability, we have calculated the QI for finished goods in 2001 only in the three districts located in Veneto. The results suggest that goods produced domestically in Brenta and Montebelluna and exported for further processing are of a higher quality than the shoes imported as OPT (respectively Qis are 2. and 1.7); whereas in Verona differences in quality on OPT of finished goods are not significant (QI is 1.03). Relationship between ID and export performance As regards the relationship between international delocalisation and export performance, trends at the provincial level are not clear cut. Nevertheless Figure 4 shows that shoe districts in Puglia 7

For a recent discussion about the different possible forms of upgrading within international production networks or global value chains see Schmitz (2004) and Giuliani et al. (2004). 8 The index is computed as the share between unit value of exports and unit value of imports (Fontagné and Freudenberg, 1997). Quality differences are bigger as much as QI is far from 1. Flows are usually classified as vertically differentiated (i.e. with relevant differences in quality) if QI is significantly different from 1, i.e. it is external of an exogenous interval, commonly fixed from 0.85 to 1.15. If QI is greater than 1.15 the quality of exports is relatively higher than the quality of imports, whereas if QI is smaller than 0.85 then the quality of imports is relatively higher than the quality of exports.

9

export less than those in Veneto and the Marches. Moreover, it is also clear that, of the districts investigated, Barletta has the highest percentage of foreign processing (more than 72% of outsourced production in 2000) and the lowest value of final exports. On the contrary, the Fermano district in the Marches has the highest value of final exports and outsources a much lower percentage of production exported as OPT (34%). Overall, we can say that in Brenta, Montebelluna and Fermano the need to maintain high quality standards to be internationally competitive in the high segments of the market is likely to result in limited international delocalisation of intermediate stages of production to foreign subcontractors. On the other hand, in Barletta, producers of lower quality items outsource a higher percentage of intermediate processing, orienting their sales mainly towards the domestic market, given that international competition in their segment of market is unsustainable. Figure 4 here 2.3 The emergence of different patterns of international outsourcing This section summarises the empirical results presented in Section 2.2 and compares the different patterns of international outsourcing emerging from the analysis of OPT data. The first main finding is that the shoe districts investigated follow different strategies of delocalisation according to the degree and pattern of outsourcing and the quality of exports relative to imports in the OPT regime. To explain this result we may refer to differences in the market segment in which each district specialises and in the competitive strategies adopted. High-quality, export-oriented producers in Veneto outsource fewer intermediate parts than those produced domestically and are therefore net exporters of intermediate goods. On the other hand, these districts are net importers of shoes and the quality of shoes exported for further processing is higher than the quality of those assembled abroad. This confirms that high-quality producers are actually upgrading their local production process by delocalising low-value added activities abroad. At the same time, they continue to perform high-value added operations, which are either capital or

10

skill intensive, at home. At the district level, Verona relies heavily on both intermediate and final outsourcing. Brenta is less reliant on outsourcing of assembling operations, possibly because the very high quality of its final products does not allow for a massive delocalisation of operations. Finally, in Montebelluna outsourcing as a competitive strategy refers more to final assembling than to production of intermediate goods. The situation is entirely different for low-quality producers in Puglia, which outsource a high percentage of production abroad (mainly intermediate processing) and are net importers of parts from foreign subcontractors. These districts attempt to cope with fierce price competition from emerging economies and they are forced by adopting a massive delocalisation strategy, displacing significant part of their production process to low labour-cost countries. Moreover, they cannot compete internationally and therefore target their production mainly to the low-price segments of the domestic market. In these cases, upgrading strategies are urgently needed to increase competitiveness on both the domestic and foreign markets, and to counter the decline in local employment (see 3.2). As regards Fermano in the Marches, this district stands out as being less involved in intermediate outsourcing than districts in Puglia, and at the same time less involved in final outsourcing than districts in Veneto. In the next section, we investigate in detail the different upgrading strategies adopted by districts oriented towards very diverse market segments, focusing on Brenta at the top end of the market and Barletta at the low end.

3. Italian footwear districts in global value chains: the cases of Brenta and Barletta In the previous section we argued that differences in the delocalisation strategies of Italian footwear districts are likely to be related to their market position and to the international production networks 11

they belong to. In this section we explore in some detail two of the districts examined above to find further supporting evidence for this argument. These cases represent two very diverse districts: the Riviera del Brenta produces high-quality women's shoes and is strongly exported-oriented, while Barletta specialises in low-quality shoes, that are mainly sold on the domestic market. Furthermore, the OPT analysis has shown that the two districts have chosen very different delocalisation patterns: Brenta relies relatively little on international delocalisation and is a net exporter of intermediate inputs, whereas Barletta has displaced more and more intermediate processing abroad and is progressively outsourcing assembling operations. This may suggest that the overall production process is likely to be dismantled and completely transferred to lower labour cost countries. In this section we suggest that these opposing strategies are likely to have a different impact on the future growth of these footwear districts.

3.1 The Brenta shoe district9 The origins of the footwear industry in Brenta date back to the beginning of last century. During the footwear industry boom after Second World War, the sector progressively absorbed most of the rural workforce in the area. In the 1960s, local enterprises expanded and increased their exports, specialising in the upper segments of the market. In 2003, almost 90 per cent of the shoes produced in the area were medium-high and high priced women's shoes with an average ex-factory price of 58 Euro. Since the second half of the 1980s, the area has suffered from increasing international competition and sales have stagnated, fluctuating between 7.9 and 8.8 million pairs (mainly due to exchange rate fluctuations). In value terms, however, sales continued to increase in most years (ANCI, 2003). The main market for Riviera del Brenta has traditionally been Europe, particularly Germany, and to 9

This section is based on Rabellotti (2004).

12

a more limited extent France, Great Britain and the rest of the EU. Recently, as demand in the German market has fallen, many firms in Brenta have begun working as subcontractors for leading global fashion firms, as increasingly important category of customers for the district. In order to understand the impact of the rapid expansion of these top brand firms in Brenta, it is useful briefly to investigate some of the characteristics of top global brand leaders. In recent years, the luxury fashion system has gone through significant changes, which have turned it into an oligopoly dominated by a few multi-product giants. The growth strategy of many companies follows a similar pattern: first, successful firms establish their brand names in specific product lines (for example three of the most important and largest companies in the industry, LVMH, Gucci and Prada, began by producing and selling leather goods); second, they capitalise on their brand names and diversify to other segments (in the cases in point, clothing, footwear, glasses, perfumes, wines) and finally, they begin to grow through acquisitions of other well-known brands. The economic logic behind these growth strategies is a search for scale and scope economies in activities other than manufacturing, such as branding, marketing, advertising, opening of monobrand shops in the most exclusive and expensive streets in the world. The increasing concentration of fashion enterprises in the intangible phases of the value chain may be explained by the growing concentration of rents in these activities. According to Kaplinsky (2000), in the past decade the barriers to entry in manufacturing have begun to fall and consequently the rent going to production activities has shrunk in favour of rents accruing to activities outside the area of production. This explains the growing concentration of investments in areas like branding, advertising, marketing and sales, where capital cost barriers to entry are high. Therefore, leading luxury firms are expanding through mergers and acquisitions, capitalising on their core competencies such as design, advertising, marketing, brand naming, which are no longer sector specific. At the same time, world top luxury companies search for highly skilled manufacturing capabilities to manufacture the different products sold under their brand name. For this reason, the Riviera del

13

Brenta has been identified as a preferred area in which to find subcontractors in the footwear sector. It may be worth adding that in Brenta this trend began at a difficult time for firms, as the positive impact of the 1992 devaluation of the Lira on exports was coming to an end. The findings of a survey undertaken by Rabellotti (2004) show that almost half of the sample firms work as subcontractors for high fashion companies producing shoes with top global brand names. In a limited number of cases, firms work exclusively as subcontractors, while the majority also produce directly for the market. This result is confirmed by the local business association, which estimates that 50% of the total output of enterprises in the Brenta district is subcontracted by high fashion companies. In most of the cases analysed, fashion companies provide the design and Brenta manufacturers take care of all production phases, including product development and purchase of raw materials and components. The shoes are then sold with the brand names of the fashion companies. It may be added that fashion companies are becoming increasingly involved in other phases, directly selecting suppliers, sometimes even through acquisitions of firms, and extending their control on quality and delivery conditions backwards along the chain. On the basis of the foregoing, it would appear that Brenta has been undergoing a process of functional downgrading. Traditionally, the design and acquisition of inputs were controlled locally and carried out inside the firms or the district. More recently, with the advent of luxury fashion companies, local enterprises are moving out of the typical core cross-sector competencies of luxury fashion companies, namely design, branding and sales. This choice is not an impoverishing strategy. Rabellotti (2004) provides evidence of a statistically significant positive relationship between performance and the share of production sold to high fashion companies. In fact, top brand companies exploit final consumers’ willingness to pay very high prices for luxury goods, earning high profits above production costs and to some extent sharing them within the chain in order to guarantee high and consistent quality and compliance with

14

delivery conditions. The high rents earned in the top brand value chain also explain why Brenta enterprises are increasingly using their internal production capacity to make shoes as subcontractors to high fashion companies and outsourcing abroad the lower value added stages of their production. This is confirmed by the significant amount of imported shoes for further processing registered in Brenta (see Table 3). Decentralisation to Romania and other Eastern European countries is needed to reduce costs, given that price competition is severe even in high quality markets. As discussed in Section 2, outsourcing is a functional upgrading strategy for Brenta firms: moving low value added activities and products abroad and focusing on production for the rent-rich luxury market at home. Again, with regard to upgrading, Rabellotti (2004) provides evidence of a positive and statistically significant relationship between the amount of production made for high fashion companies and the degree of product upgrading. We may conclude that, in Brenta, top brand companies are definitely setting the parameters with which local firms have to comply, in many cases they are also co-operating with their highly skilled partners to obtain top quality products. Very importantly, they are also willing to share part of their rent with them in order to acquire their production skills. Therefore although local firms have abandoned some key functions, their performance is still very positive, because they share the high rents of the luxury industry with the chain’s leaders. However, it is not clear what this means for Brenta's future. While local firms are gaining a place in the rent-rich global top brand market focusing on their production skills, what they offer can increasingly be found in other clusters in the world and this may slowly erode their competitiveness and independence. To date, the need for highly skilled subcontractors capable of producing for the top quality market and the higher transaction costs of establishing relations with more distant and less qualified subcontractors still discourage luxury companies from looking for alternative

15

subcontractors in countries like Romania or Brazil. Nevertheless, these conditions could quickly change and the dynamic comparative advantages of Brenta’s producers may one day vanish.

3.2 The Barletta shoe district10 Barletta is the second largest footwear district in the south of Italy and the seventh largest in the country (ISTAT, 1996). The origins of the district date back to the years before the Second World War when migrants returned from Vigevano, Italy's oldest footwear district, and set up the first small shoe enterprises. Up to the 1940s, production was carried out by hand and consisted of leather shoes. It was only after the war that local producers began introducing innovations in the production process, abandoning craftsmanship in favour of industrial production. The main innovations were the partial mechanisation of production and the use of rubber together with leather. A major reason for this change in production techniques was the severe shortage of leather in the aftermath of the war. From the 1960s onwards, rubber as a raw material was replaced by plastic. The introduction of this new material brought about a radical change in the production process, as well as the establishment of new linkages among local producers and certain actors outside the district, namely specialised machinery suppliers located mainly in Veneto and Tuscany. With the new specialization in low-market plastic shoes, Barletta’s producers acquired a large share of the international market and international buyers began to visit Barletta regularly. German buyers were the first to come, having abandoned the Marches, where local producers moved up market. Among the major local competitive factors were the low production costs, given the initial informality of many firms, and the ability rapidly to adapt production to market changes and to introduce new machinery. Indeed, since the mid 1960s, local producers have invested heavily in machinery, making strong productivity gains. 10

In this section we draw information on D’Ercole (2000).

16

Since the early 1970s, with the introduction of new machines, local firms have specialised in the production of sports shoes with plastic soles. To satisfy the increasing demand for low priced shoes in the European market, the district has specialised in highly standardised products with huge volumes of production. As a consequence, the growth of each firm has depended on the growth of the district as a whole. In other words, while single firms are too small to satisfy alone the huge quantities of standardised goods demanded by international buyers, the whole district can collectively provide adequate quantities. Therefore, local firms have had to homogenize quality and their profitability has depended on the production capacity of the whole district. With this model of production, Barletta was an international leader in its market segment until the end of the 1980s. However, since the 1990s, Barletta has begun to face increasing international competition, mainly due to the entry into the international market of the countries of South-East Asia, particularly China. Barletta had also had to face a shift in international demand for sports shoes, in favour of high quality shoes with a high technological content on the one hand and of low quality fashion shoes on the other hand. As attempt to respond to the huge differential in wage costs between Italian and Asian producers, Barletta producers successfully lobbied for greater market protection against ‘unfair’ imports from Asian countries. In 1994, the European Commission introduced quotas on imports from China and in 1998 it fixed a minimum price on imported goods and tariffs of between 2 and 20.3% on imports from China, Indonesia and Thailand. Nevertheless, notwithstanding the protection of the European market, the Barletta model has recently shown severe vulnerabilities with respect to international competitors. The district faces a structural problem of progressive loss of competitiveness and has started drastically to reduce production, profitability and employment (Parolini and Visconti, 2003). This should lead to a radical re-organisation of production and of investment in different segments of the markets. Up to now, the main production re-organisation strategy has been the displacement of some firms and a

17

massive outsourcing of production to Albania (see Section 2). In a sense, Barletta has replicated its own production model by displacing part of its production capacity to the eastern side of the Adriatic Sea. As a result, there has been a significant reduction in employment and in average firm size. While Barletta's externalisation of lower value-added operations may be interpreted as a functional upgrading strategy, the problem is that there is no real local investment in higher valueadded functions. A very different strategy was chosen in Montebelluna, the major Italian footwear district specialising in sports shoes and ski boots. Here, the outsourcing of low value added segments of the production process began a long time ago, taking advantage of international production costs differences, but local firms, which include some international leaders, such as Diadora, Lotto, Tecnica, have moved to the high quality segment of the market, retaining locally high value-added phases such as R&D, marketing, design (Camagni and Rabellotti, 1992; Mariotti, 2003). To conclude, we should like to mention a recent agreement between Brenta and Barletta to increase the outsourcing of production phases from Veneto to Puglia. This may represent an attempt by producers in Barletta to enter a new market segment, albeit only as subcontractors (Il Sole 24 Ore, 2003 a and b). As in Brenta, producers in Barletta may attempt to pursue a strategy of functional downgrading to the extent that it allows for a repositioning into different – higher quality – value chains.

4. Conclusions This paper studies the impact of global transformations on some of Italy's main footwear districts. It analyses the pattern of international delocalisation of production in relation with the position of these districts in international production networks. Data on OPT and qualitative empirical evidence is used to address the following questions: Are Italian footwear districts specialising in certain phases of the production cycle? Is there a trend towards a reduction of activities carried out within

18

the districts? Or, instead, are different patterns emerging according to districts’ main segment of market (top brand, low brand, mass market) they belong to? In what follows we present our main conclusions. First, districts vary in their outsourcing strategies, depending on their market position. Districts like Barletta and Salento in Puglia, which specialise in the low-price market segments, outsource a higher percentage of intermediate production abroad than districts in Veneto and the Marches. They are net importers of parts from foreign subcontractors. Faced with a very strong price competition mainly from Asian countries, shoe producers in the low segment of the market react by outsourcing a large part of the production cycle abroad. Nevertheless, international delocalisation as a strategy to reduce production costs is unlikely to be sufficient to counter competition from low cost producers in less developed countries. For instance, Barletta has exported its production model to Albania, but this has resulted in a strong reduction in local employment and in the number of firms, because the loss of low value added functions has not been accompanied by a shift of resources to other higher value added activities. Footwear districts specialising in high-price segments of the market do outsource some production phases as a cost-reducing strategy, but mainly final rather than intermediate processing. In this case, the decentralisation of low-value added activities to foreign subcontractors can be interpreted as a form of functional upgrading, as lower value added activities are externalised and high-value added phases are kept at home. In particular, the bulk of intermediate processing remains at home, while a larger volume of assembling operations is delocalised. This might be interpreted as a way to reduce production costs on assembling operations (which are usually labour intensive), while maintaining high quality standards on intermediate inputs. Nevertheless in the case of Brenta, external leading fashion firms are playing an increasingly significant role in the top segment of market, assuming control of the highest value added phases of the production process, such as design and sales. To be part of the top brand chain, Brenta's shoe

19

producers accept a functional 'downgrading', focusing on production and losing some of their key competencies in favour of the chain leaders. It must be added that local firms, which become subcontractors, perform better than other local producers in terms of sales and profits, as well as process and product upgrading. It appears that the luxury brand companies share some of their high profits with their skilful subcontractors. It is not yet clear what these trends imply for the future of Italian footwear districts. On the one hand, all the districts investigated are involved in an extensive overhaul of the production process through international delocalisation of low value-added phases. This phenomenon obviously implies the substitution of local firms by international subcontractors. Will this endangers the survival of the districts or is it rather a process of natural selection eliminating the least competitive firms? The question remains open. On the other side of the value chain, footwear districts are losing their independence in high value added phases such as design, branding and marketing in favour of luxury oligopolies. So far, the chain leaders have needed the highly skilled manufacturing capabilities of their Italian subcontractors. However, conditions may change, Romanian or Brazilian shoe producers may rapidly improve their skills and the dynamic comparative advantage of high quality producers in Italian shoe districts may one day vanish.

20

References ANCI (2001), Il Traffico di Perfezionamento Passivo nel settore calzaturiero (8° Rapporto semestrale), Associazioni Nazionali Calzaturifici Italiani: Milan. ANCI (2003), Il Traffico di Perfezionamento Passivo nel settore calzaturiero (10° Rapporto semestrale), Associazioni Nazionali Calzaturifici Italiani: Milan. ANCI, various years, L’industria calzaturiera italiana, Associazione Nazionale Calzaturifici Italiani: Milan. ACRIB, various years, Rilevamenti statistici, Associazione Calzaturifici della Riviera del Brenta: Strà. Baldone, S., Sdogati, T. and Tajoli, L. (2002), “Moving to the Central-Eastern Europe: fragmentation of production and competitiveness of the European textile and apparel industry”, Rivista di Politica Economica. Camagni R, Rabellotti R., 1992, "Technology and organisation in the Italian textile-clothing industry", Entrepreneurship and Regional Development, 4, 271-285 Chiarlone, S. (2001), “Evidence of Product Differentiation and Relative Quality in Italian Trade”, Rivista Italiana degli Economisti, No. 2 (August). Corò G, Grandinetti R., (1999), “Strategie di delocalizzazione e processi evolutivi nei distretti industriali italiani”, L’Industria, XX, n.4 De Nardis, S. and F. Traù (1999), “Specializzazione Settoriale e Qualità dei Prodotti: Misure della Pressione Competitiva sull’Industria Italiana”, Rivista Italiana degli Economist,. Anno IV No. 2 D’Ercole, M. (2000), “Il distretto barlettano della calzatura”, in Viesti, G., ed., Mezzogiorno dei distretti, Meridiana Libri: Roma. Flam. H. and Helpman, H. (1987), “Vertical Product Differentiation and North South Trade”, American Economic Review, 77, pp. 810-822. Fontagné, L. and Freudenberg, M. (1997), “Intra-Industry Trade Methodological Issues Reconsidered”, CEPII Working Paper, 1997-2001. Il Sole 24 Ore (2003a), “Calzature, patto Veneto-Puglia”, April 4th. Il Sole 24 Ore (2003b), “Calzaturieri sempre a corto di addetti”, May 27th. Istat (1996), Industrial Census, preliminary results available at www.istat.it. Giuliani E., Pietrobelli C., Rabellotti R., (2004), “Upgrading in global value chains: Lessons from Latin American Clusters”, forthcoming in World Development. Jones, R. and Kierzkowski, H. (2000), “A Framework for Fragmentation”, in Arndt, S. and H. Kierzkowski, eds, Fragmentation in International Trade, Oxford University Press: Oxford. Kaplinsky R. (1998,) “Globalisation, Industrialisation and Sustainable Growth: The Pursuit of the Nth Rent”, IDS Discussion Paper 365, Institute of Development Studies: Brighton. Kaplinsky R. (2000), “Spreading the gains from globalisation: what can be learned from value chain analysis?”, IDS Working Paper, 110, Institute of Development Studies: Brighton. Kaplinsky R., Readman J., (2001), “How can SME producers serve global markets and sustain income growth?”, University of Sussex and University of Brighton, mimeo.

21

Mariotti I.(2003), “Le strategie di delocalizzazione delle imperse del Nord Est nei paesi dell’Europa Sud Orientale: reti lunghe o fabbriche con le ruote?”, in Rapporto ICE, Istituto del Commercio Estero, Roma, pp.316-328. Pambianco (2002), “Osservatorio delle principali operazioni di Merger & Acquisition avvenute nei settori del Made in Italy a livello mondiale”, mimeo, available at www.pambianco.com . Pambianco, various years, “I bilanci del Made in Italy”, mimeo, available at www.pambianco.com . Parolini

C., Visconti F., 2003, “Footwear manufacturing districts: the effects of internazionalisation, delocalisation and new technologies”, paper presented at the Conference “Clusters, industrial district and firms: the challenge of globalization”, September, Università di Modena: Modena.

Rabellotti R. (1997), External Economies and Cooperation in Industrial Districts. A Comparison of Italy and Mexico, Macmillan: London. Rabellotti R. (2004), “How globalisation affects Italian industrial districts: The case of Brenta” in H. Schmitz (2004) Schmitz H., (2004), ed., Local Enterprises in the Global Economy: Issues of Governance and Upgrading, Edward Elgar Publishers: Cheltenham.

22

Figure 1: Unit value of Italian footwear exports (Euro) 20.00 19.00 18.00 17.00 16.00 15.00 14.00 1996

1997

1998

1999

2000

2001

2002

2003

Source: elaborated on ISTAT database (http://www.coeweb.istat.it) Figure 2: Value of required processing in third countries (% on value of exported production for processing), 1996-2000 80,00%

70,00%

60,00%

50,00%

40,00%

30,00%

20,00%

10,00%

0,00% Montebelluna

Verona

Fermano

Barletta

Salento

Brenta

Source: elaborated on OPT database (ANCI, 2001) Figure 3 Imports and exports of parts, 2001

Exports of parts

4000

Barletta

3000

Salento Brenta

2000

Montebelluna Verona

1000

Fermano 0 0

1000

2000

3000

4000

Im ports of parts

Source: elaborated on OPT data from ANCI (2001)

23

Figure 4 Value of outsourced processing and value of final exports, 2000 3500000 3000000 2500000 2000000 1500000 1000000 500000 0 rm Fe

value of final exports

o an B

tta le ar

l Sa

t en

o en Br

value of outsourced processing

ta

M

t on

eb

lu el

na

n ro Ve

a

Source: elaborated on OPT data from ANCI (2001)

Table 1: The shoe districts analysed Brenta Verona Montebelluna Fermano Barletta

N° of shoe N° of Product of firmsa employeesa specialization 295 4,796 Women shoes 417 6,450 Mixed 362 8,152 Sport shoes-sky boots 2,554 21,156 Mixed 453 5,872 Mixed

145 6,935 Salento a Istat, intermediate industrial census (1996)

Mixed

ExportOrientation High High High High Decreasing from high to low Low

Main segment of market Top Medium Medium to high. Medium to high Low Low

Table 2 The relevance of OPT in Veneto, Puglia and The Marches (% on Italy, 2001, 2nd qr.) Veneto Puglia The Marches Total above Total Italy (mln €)

OPT Imports

OPT Exports

53.93% 24.64% 6.57% 85.14%

58.73% 23.21% 5.01% 86.95%

198 62 Source: elaborated on OPT data from ANCI (2003)

Table 3: OPT flows in selected Italian footwear districts (2001, 2nd qr.)

Barletta Salento Brenta Montebelluna Verona The Marches

Mi ‘000kgs 2340 1125 509 78 703 425

€/kg 6,6 21,5 24,0 30,1 26,0 10,5

Mf ‘000pairs €/pair 25 10,6 714 12,1 465 13,4 687 9,8 4404 13,6 448 15,3

Xi ‘000kgs €/kg 2077 4,9 635 6,6 521 13,00 915 9,2 3694 5,6 366 7,3

Xf ‘000pairs 1,0 6,7 1,4 -

€/pair 32,28 16,52 14,01 -

NXi ‘000kgs -263 -491 16 836 2991 -59

NXf ‘000pairs -464 -680 -4402 -

Legenda M: Imports; X: Exports; NX: Net Exports; i: intermediate goods; f: finished goods.

Source: elaborated on OPT data from ANCI (2003)

24

Table 4: Number of upper pairs required as OPT, 1996-2000 (% by partner country) Albania 96,90% 86,05% 92,65% 100,00% 93,56% 66,10% 80,55% 73,44% 90,81% 86,19%

Romania

Bulgaria

1996 1997 7,40% 1998 7,31% 1999 0,00% 2000 5,73% 1996 14,36% 17,65% Salento 1997 1998 2,95% 1999 4,60% 4,60% 2000 7,44% 6,37% 1996 Brenta 58,23% 1997 64,64% 1998 69,75% 1999 87,19% 2000 92,88% 1996 1,55% 12,31% Montebelluna 1997 18,18% 1998 86,73% 1999 71,55% 2000 38,54% 1996 Verona 77,34% 1997 10,61% 75,46% 1998 12,59% 62,43% 1999 0,59% 81,81% 2000 3,99% 74,57% 1996 10,43% 21,34% 37,17% Fermano 1997 14,95% 11,07% 34,79% 1998 3,77% 37,02% 38,85% 1999 22,75% 6,93% 54,37% 2000 7,14% 29,49% 54,52% Source: elaborated on OPT data from ANCI (2001) Barletta

Hungaria

Poland

Yugoslavia 3,10% 6,55%

Croatia

Ukraine

Slovakia

Bosnia Erz.

Others

0,04% 0,72% 1,72% 12,15% 23,61%

0,16% 7,29%

9,94% 13,76%

13,18% 7,93% 14,53%

11,59% 0,19% 36,92% 2,03% 6,01% 29,34% 15,63% 6,13% 6,71% 5,54% 2,55% 13,32% 17,97% 13,54% 9,71% 8,85%

1,31% 2,22% 5,04% 5,21%

3,05% 4,30% 0,88%

0,79% 1,18% 1,22% 1,04% 36,92% 56,25% 10,66% 22,27% 5,14% 3,25% 3,55% 5,82% 3,27% 3,21% 3,91%

14,92% 11,89% 14,53%

3,73% 0,99%

4,72%

1,18% 10,83%

1,48% 25,57% 0,58% 0,17% 14,39% 0,84% 1,28%

1,20% 4,17% 4,85%

4,16% 5,16% 1,69% 1,82% 1,39%

10,61% 3,27% 6,26%

1,77%

12,59% 2,94% 2,96% 0,52% 5,52% 6,35% 1,95% 2,35%

Total 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% 100%

% on Italy 8,27% 19,10% 22,60% 10,94% 20,03% 45,43% 21,80% 8,75% 15,55% 20,25% 3,75% 8,91% 5,69% 7,30% 7,61% 3,79% 0,31% 7,13% 10,23% 1,40% 1,49% 9,65% 19,69% 15,04% 25,89% 7,78% 13,57% 17,55% 6,44% 7,30%

Table 5: Number of pair of shoes required as OPT, 1996-2000 (% by partner country) Romania Albania Croatia Yugoslavia Bulgaria Hungaria Brenta 1996 25,00% 1997 32,26% 1998 16,13% 1999 100,00% 5,79% 2000 75,67% 26,44% Montebelluna 1996 73,56% 1997 60,45% 23,05% 14,54% 14,71% 1,56% 1998 78,83% 24,39% 4,86% 1999 63,15% 1,67% 8,36% 2000 78,93% Verona 1996 100,00% 9,00% 5,31% 8,49% 25,44% 1997 41,67% 9,72% 1,68% 4,49% 3,93% 1998 77,51% 13,81% 1999 51,34% 1,44% 8,26% 30,59% 0,82% 11,84% 8,22% 2000 47,70% Fermano 1996 10,43% 32,10% 12,04% 7,98% 1997 69,15% 6,65% 11,09% 1998 53,42% 13,31% 1999 93,75% 5,97% 2000 38,72% 55,31% Barletta 1996 1997 1998 1999 100% 2000 16,67% Salento 1996 83,33% 1997 26,92% 59,62% 1998 66,67% 1999 100% 31,85% 2000 68,15% Source: elaborated on OPT data from ANCI (2001)

Ukraine Bosnia E. Slovakia Poland Others Total % on Italy 100% 2,39% 100,00% 100% 2,26% 75,00% 100% 3,96% 51,61% 100% 2,22% 18,54% 100% 8,24% 100% 0,91% 1,96% 100% 24,50% 1,87% 3,04% 100% 16,39% 3,20% 1,88% 2,51% 100% 16,41% 3,34% 1,00% 6,69% 100% 11,42% 100% 4,79% 10,10% 100% 31,93% 2,68% 100% 40,98% 12,21% 1,62% 2,51% 8,83% 100% 35,86% 0,82% 100% 23,22% 21,67% 23,75% 100% 9,94% 13,30% 9,57% 100% 10,62% 11,09% 4,44% 100% 5,76% 6,25% 100% 4,12% 100% 3,45%

100%

13,46% 33,33%

100% 100% 100% 100% 100%

28

0,39% 57,46% 14,68% 3,83% 0,13% 17,98%