collective effects in italian and mexican footwear industrial clusters

ABSTRACT. The focus of this article is on the analysisof the collective economic effects deriving from the intenseset of backward, forward, labor, horizontal and institutionallinkages existing within clusters of enterprises. Among theeconomic effects two main categories are distinguished:external economies, which are the spontaneous by-product ofeconomic activities undertaken within the clusters and coop-eration effects, which are the results of explicit and deliberatecooperative behaviors of the economic actors.

In the empirical investigation, these economic effects havebeen analyzed in four clusters of footwear firms in Italy andMexico. The first result of the empirical investigation is theconfirmation of the importance of collective efficiency bothin the ‘proper’ Italian districts and in the Mexican clusters.Nevertheless, there are considerable differences concerningthe intensity and quality of the collective effects between therealities studied. Those differences are explained through theimpact of the disparities in the outside environment on the corecharacteristics of the different clusters. Finally, some consid-erations about the need for moving from a static to a dynamicapproach to explain differences between stages of develop-ment and growth trajectory patterns of the districts are putforward.

1. Introduction

In recent years industrial districts generated a greatdeal of interest among development economists.Traditionally, in less developed countries smallfirms were seen as socially desirable, but theirviability remained in doubt. Their role was tocreate jobs, to be a seed-bed for indigenous entre-preneurship, to use predominantly local resources,to produce for local markets satisfying the basicneeds of the poor, to contribute to a more equi-table distribution of income, but rarely were smallfirms seen as able to become internationally com-petitive. The success of industrial districts in theWest was taken as a sign that small firms could

be economically viable and strongly contribute tothe industrial growth of their country. The conceptof the industrial district was then introduced in thestudy of industrial development in the ThirdWorld, both at the empirical and theoretical level(Schmitz, 1989; Rasmussen et al., 1992).

The aim of this paper is to contribute to this lineof investigation with a comparative study of twofootwear districts in Italy and two clusters ofspatially concentrated and sectorally specializedfootwear enterprises in Mexico.1 We intend toinvestigate backward, forward, horizontal, laborand institutional linkages and to classify thecollective economic effects deriving from them,by distinguishing between static and dynamicexternal economies and static and dynamic coop-eration effects. Ultimately, we compare the degreeof collective efficiency in the Italian and Mexicanfootwear clusters.

The article is structured as follows: in the nextsection we present the key factors characterizingindustrial districts and we discuss the conceptsof external economies and cooperation effects.Section 3 contains the main findings of the empir-ical investigation on linkages in Italy and Mexicoand section 4 provides a classification and a com-parison of the collective effects, observed in thecases analyzed. The implications of these findingsare drawn out in the last section.

2. External economies and cooperation effects

Research on industrial districts originated in thesecond half of the 1970s in Italy, where a wide-spread phenomenon of industrial developmentbased on geographically concentrated and sec-torally specialized small firms, was identified bya number of scholars.2 The increase in the numberof employees, production, exports, and per-capitaincome in north-central and north-eastern Italy,

Collective Effects in Italian andMexican Footwear Industrial Clusters

Roberta Rabellotti

Small Business Economics 10: 243–262, 1998. 1998 Kluwer Academic Publishers. Printed in the Netherlands.

Final version accepted on July 29, 1996

Economics Department, Università di Padova

also known as the ‘Third Italy’,3 attracted theincreasing interest of Italian and foreign econo-mists and sociologists,4 trying to understand thereasons for this ‘economic miracle’. During the1980s a great amount of research, both at theempirical and theoretical level, was carried out inItaly and in other Western countries, where manyagglomerations of small firms, similar to theoriginal Italian districts, were identified.

Based on the very vast body of literature avail-able, we propose to identify industrial districtsin terms of four key stylized facts (Rabellotti,1995b):

• a cluster of mainly small and medium-sizedenterprises spatially concentrated and sectorallyspecialized;

• a strong, relatively homogeneous, cultural andsocial background linking the economic agentsand creating a common and widely acceptedbehavioral code, sometimes explicit but oftenimplicit;

• an intense set of backward, forward, horizontaland labor linkages, based both on marketand non-market exchanges of goods, services,information and people;

• a network of public and private local institu-tions supporting the economic agents in theclusters.

Widespread availability of external economiesand cooperation effects, stemming from the abovefeatures and particularly from the last two char-acteristics, is commonly seen as the key factormaking small firms in the districts economicallyviable and internationally competitive, in otherwords allowing the districts to reach a highdegree of collective efficiency (Schmitz, 1992 and1995b). Schmitz defines collective efficiency asthe sum of unplanned or incidental consequencesand planned or consciously pursued effects.Taking the definition of collective efficiency a stepfurther, we introduce the distinction between staticand dynamic external economies and cooperationeffects (Rabellotti, 1995a).

The existence of external economies in indus-trial districts can be traced back to Marshall, whostressed in his Principles of Economics (1890) theimportance of those economies which ‘can oftenbe secured by the concentration of many smallbusinesses of a similar character in particular

localities’ (8th edition, 1920: 221). Externaleconomies can be defined as unpaid side-effectsof activities undertaken within the districts. Theytend to be characterized by non-rivalry and non-excludibility which can lead to under-investmentsin a market economy (Cornes and Sandler, 1986).Moreover, external economies are part of acommon pool from which every economic actorcan freely draw. In industrial districts a typicalexample can be the circulation of informationabout markets or technologies, freely available toevery economic actor located in it, who can usethe same information, adapting it at her/his poten-tially different needs. Other examples of externaleconomies are the creation of a collective com-mercial image of the district and the diffusion ofa professional and managerial culture.

Moving on to cooperation effects, they are theresult of explicit and voluntary cooperative behav-iors which differ by external economies becauseof two important characteristics: excludibility andcompensation. The existence of a mechanism forcompensation accounts for the rational interest ofeconomic actors to cooperate among each other,producing cooperation effects and increasing thecollective efficiency of the system. In a long termperspective, compensation can assume variousforms like monetary exchange, exchanges of infor-mation, technology, know how or human capital,building up a reputation which can become usefulin future interactions. Besides, the existence of amechanism for compensation allows the intro-duction of excludibility in the exploitation ofcollective advantages deriving from cooperationeffects. An example of the effects of cooperativebehaviors in industrial districts is the sharing ofspecific information within groups of firms linkedby some form of formal or informal agreements,but in any case explicit. Other examples areprojects on technology, marketing, training, designcarried out jointly by the members of the entre-preneurial associations or by groups of firmswhich explicitly decide to cooperate.

External economies and cooperation effects canbe static, when they impact the level of produc-tivity of the system, or dynamic, when they impactthe system’s capability to grow and innovate.

In what follows, we present the main findingsof an empirical investigation undertaken in fourfootwear clusters in Italy and Mexico, explicitly

244 Roberta Rabellotti

selected because they satisfy the key element ofspatial concentration and sectoral specialization.We investigate the existence, the level of intensityand the quality of backward, forward, horizontal,labor and institutional linkages as well as theexternal economies and the cooperation effects,which derive from them. Cultural and social back-ground are not explicitly considered, althoughtheir relevance indirectly comes out through theirinterconnections with the economic and institu-tional aspects of the systems.

3. The case studies

Fieldwork was carried out in four clusters offootwear enterprises: two of them located inItaly (Brenta and Marche) and two in Mexico(Guadalajara and Leon). In those clusters a sampleof firms was randomly chosen among the membersof the local associations of footwear entrepre-neurs.5 The sample included 101 enterprises: 50of them located in Italy (30 in Marche and 20 inBrenta) and 51 in Mexico (30 in Guadalajara and21 in Leon). In order to take into account firmsof different size, the sample is stratified by size,as can be seen in Table I.

A closed questionnaire was used to inter-view the sample firms, focused on investigatingbackward, forward, labor, horizontal and institu-tional linkages. The information collected withthe questionnaire within the sample of firms wascomplemented with secondary data as well aswith open-ended interviews with EntrepreneurialAssociations, sector experts, suppliers, buyers,representatives of institutions supporting thesector.

With regard to the existence of a critical massof spatially concentrated and sectoral specializedenterprises, the four areas all satisfied this condi-

tion as it was a criterion for choosing the researchareas:

• in Marche there are 2,410 firms and in Brenta680, out of a total 8,547 firms in the whole ofItaly (ANCI, 1994);

• in Leon there are about 1,700 shoe enterprisesand in Guadalajara about 1,200, out of atotal of 4,500 firms in the whole of Mexico(CANAICAL, 1993).

3.1. The Italian districts: Brenta and Marche

Before the presentation of the results of the inquiryin the two Italian districts, it is useful to intro-duce a few very essential background informationabout the Italian footwear sector.6

The recent history of the Italian footwearindustry can be divided up into two main periods:a first long period of continuous expansion fromthe beginning of the 1960s until 1985 and a secondperiod of crisis and restructuring of the sectorwhich started in the second half of the 1980s. Theoutstanding growth of the Italian footwear industrywas export-led: at the beginning of the 1950sexports represented a mere 3.7% of the totalproduction, in 1970 the proportion of exportsincreased to 63%, in 1985 it was 83% and in 199384% (ANCI, 1994).

More recently during the 1980s, other Europeancountries like Spain, Portugal, some south-eastAsian NICs like Taiwan and South Korea and alsosome developing countries like Brazil, India andChina became very competitive in the interna-tional market and greatly increased their exportshare.

Notwithstanding its attempt to counter thisincreasing competition through a process of trade-up by upgrading its exports, the Italian footwearsector had to start to compete in a market wherea large increase in the quantity of shoes suppliedfaced a stable demand. The competition becomestronger because all the new producers couldexploit a labor cost advantage with respect to Italy.

The Italian footwear industry is therefore goingthrough a very difficult time, after 30 years ofcontinuous growth: both in the domestic marketand in the international market it has to face anincreasing competition from countries with a clearadvantage in terms of labor cost.

Collective Effects in Italian and Mexican Footwear Industrial Clusters 245

TABLE IThe sample: number of firms by size

No. of firms <50 51–100 >100 Total employees employees employees firms

Italy 24 15 11 50Mexico 17 14 20 51

Source: author’s survey

(a) Backward linkages. The existence of a verywell developed system of suppliers working forthe footwear sector represents one of the mainassets of the Italian shoe producers: it has allowedto maintain some competitiveness in a marketrecently characterized by stagnant demand. Theability of suppliers to manufacture a wide varietyof products with short delivery times, allows theshoe producers to postpone to the last momenttheir purchases of inputs. This has several advan-tages: first of all, it reduces the stocks required forproducing shoes; secondly, it leads to the pro-gressive shortening of the period between orderand delivery which characterizes the sector andfinally, it increases the capacity of shoe producersto diversify their products and to satisfy marketdemand.

This advanced system of production includestanneries, producers of components and acces-sories, suppliers of machines and service firms.From the results of our interviews, as well as fromother available studies (Varaldo, 1988; Gaibisso,1992), it clearly appears that in many cases thecollaboration between suppliers and shoe pro-ducers is facilitated by spatial proximity and bythe existence of long-term relationships (often 15to 20 years old). These are crucial factors whichallows continuous interaction, facilitate compre-hension and provide important sources of learningfor all actors.

Altogether the relationships between shoe firmsand their suppliers seem to be rather satisfactoryand in fact 75% of the sample firms have notencountered any particular problems and theremaining 25% sometimes have to face someavailability difficulties, namely delays in deliv-eries. The firms interviewed declare that normallywhen there are problems with the suppliers theytry to solve them together. In most cases it worksbecause, as seen before, they usually have wellestablished relationships, which due to physicalproximity and the duration in time have oftenbecome friendships. Moreover, if something goeswrong in either direction, i.e. the producer of com-ponents does not supply the quality required or theshoe producer does not pay the orders, the infor-mation immediately circulates within the district.The consequence is a loss of reputation for thefirm which has broken the agreement or ultimatelythe extreme sanction can be the exclusion from the

community. This informal, but widely accepted,system of social and business rules contributes tocreating a very efficient and cooperative systemof production, able to satisfy a very diversified andvolatile demand.

Wide circulation of information takes also placeabout products, because component manufacturersare the first to see the novelties and it is quitecommon for them to offer other firms modelssimilar to what they are already making for oneclient. The diffusion of information helps smallerfirms which do not have the capabilities toproduce their own original sample collection tosurvive. On the other hand, larger firms whichmay invest a lot of resources in the developmentof new products try to defend themselves fromimitation by other firms in the area through thestability of their relationships with suppliers,based on trust and long-standing collaboration.Naturally, the component suppliers have their owninterests in maintaining stable and good relation-ships with their best clients and will thereforeavoid making exact copies of the same productsin the same season for other firms.

In the two areas analyzed, Italian shoe pro-ducers may also count on the existence of a widenetwork of small subcontracting enterprises spe-cialized in one or two phases of the productioncycle. According to a survey (Varaldo, 1988),more than 80% of the Italian footwear firmsdecentralize the production of bottoms to subcon-tractors, more than 70% decentralize the phases ofedging and sewing of uppers and more than 50%externalize the cutting phase.

Results from the sample firms (Table II)confirm Varaldo’s study (1988) that the produc-tion of bottoms is the phase more frequentlydecentralized. This happens because the produc-


TABLE IIThe extent of externalizing production tasks: Italy

(% of sample firms)

% of pairs None <50% 51%–90% >90%externalized

Cutting 26 20 20 34Sewing 06 16 20 58Bottoms 10 10 02 78Finishing 96 02 02 00


tion of soles, insoles and heels is a separatephase from the rest of process, using specializedmachines, requiring particular labor force skills,and with larger scale economies than shoe pro-duction itself. The edging and sewing phases alsotend to be decentralized to subcontractors becausethey are highly labor intensive. However, in firmsproducing top quality products they may be inter-nalized to maintain a greater control over theprocess. The cutting phase is important both withrespect to quality and cost of the final product, andit depends much on the ability of the labor force.For the production of sample sets every firm needsto have its own internal cutting department. Thisexplains why cutting is somewhat less frequentlydecentralized than the production of bottoms.Finally, fitting and finishing are usually internal-ized because it is crucial to have a strict controlon these final phases. Decentralization to subcon-tractors is often adopted only in case of orders toolarge for the internal capacity of the firm.

According to the enterprises interviewed, themain reasons for decentralization are the reductionof costs (according to 72% of the firms), increasedflexibility (54%), the certitude of costs (28%) andincreased specialization (26%). The cost advan-tages of externally decentralizing part of the pro-duction arise partly from greater tax evasion andhigher exploitation of labor in small subcon-tracting enterprises. In addition, subcontractingenterprises, being more specialized, are able tobetter exploit the different economies of scalein the diverse phases of production, efficientlyemploying the production capacity available. Theresults being that they produce better products atlower costs, with a shorter lead-time.

Analyzing the relationships among the shoefirms and the process specialized firms (e.g. firmsspecialized in the sewing or the cutting phases),they are mainly hierarchical, characterized by astrong dependence of the subcontracting firms onthe shoe enterprises. Most of the subcontractingfirms interviewed stress that their activity dependsheavily on the demand of shoe producers, so thatthey move from period of extra-work to periods inwhich they do not have enough work to occupytheir workers. They are therefore obliged todemand extremely high flexibility of the workers.Often, they employ relatives or very youngwomen, who are normally paid on a piece-rate

basis. Most of the subcontracting firms suffer fromtheir dependence on shoe producers, particularlyduring periods of crisis, because their workloadcan be significantly reduced and it becomes diffi-cult to make any organizational plans; they canonly wait and hope to be given work by the shoeproducers.

More rarely, in these relationships there are alsosome aspects of cooperation. Some of the footwearenterprises interviewed stated that they oftensupply raw materials, workforce training, techno-logical assistance and sometimes even credit tosome of their subcontractors. They have stablerelationships with their subcontractors in order tohave more guarantees in terms of quality andservice. From the interviews, it appears that thisstrategy is common mainly among firms whichproduce a medium-high or high quality product;in this case the stability of the subcontractor rela-tionships is an essential condition for maintainingthe required quality level.

(b) Forward linkages. The Italian footwear indus-try led the international market until the mid ’80sand for a long time shoe manufacturers faced amarket characterized by excess demand. Many ofthe entrepreneurs interviewed in Italy emphasizedthat until a few years ago foreign buyers, mainlyfrom Germany, used to go and visit their firmsregularly, and all they needed to do was to produceshoes. It was not necessary to make any effortto sell them. Therefore, traditionally the Italianshoe manufacturers did not care much about com-mercialization and marketing. According to theEntrepreneurial Association (ANCI, 1992), mostof the entrepreneurs still identify the quality oftheir product only from a purely manufacturingpoint of view, according to the quality of pro-cessing and raw materials. Rarely they are awareof the importance of other factors like the imageof their products or their brand name.

In Italy, it has been argued that in the footweardistricts the efficient way of organizing the pro-duction phase is not complemented by an equallyefficient system of commercialization (Gaibisso,1992). In the footwear areas the role of mid-dlemen, who have been so important in other dis-tricts, specialized in different sectors, amongwhich we can quote the famous case of Pratoand its ‘impannatori’, has been very limited. Only


recently in Brenta have some wholesalers begunto play an important role in organizing the pro-duction of a number of small firms, but they allcome from outside the district. They are foreignor Italian wholesalers, who take advantage of theproduction skills of some local firms unable to selltheir own products on the market and are there-fore obliged to become subcontractors. Among thefirms interviewed, the few that sell their productsto buyers stressed the risk of the dependence on afew, or more often only one customer. Only onefirm, interviewed in Brenta, was satisfied with thisway of selling its products because it producesvery high quality shoes for two well knownfashion houses and it believes that the quality ofits products guarantees a high degree of bargainingpower in front of its customers, who have everyinterest in maintaining successful and stablerelationships.

Commercialization through non-exclusiveagents took place in 76% of the sample firms,through direct sales to customers in 22%, throughexclusive agents in 16%, through trading compa-nies or in shops owned by the same firm in 8%and in the remaining 8% through consortia withother firms. The relationship with non-exclusiveagents often generates discontent among footwearfirms because they do not control the activity oftheir agents and receive very little informationabout the market from them. Many of the enter-prises interviewed realized they would need amore active commercial strategy to be successfulin an increasingly competitive market. However,in marketing, economies of scale are importantand very few among the Italian firms have thefinancial capability to invest in advertising or inmarket research, to set up their own shops or toemploy exclusive agents.

A collective solution, rather diffused in thetwo area analyzed, is the establishment of exportconsortia, set up for supporting the export activityof firms with an initial public financial contribu-tion (see beyond in the section on institutionallinkages).

(c) Horizontal linkages. Evidence about formalagreements of horizontal cooperation is minimal.The only formal agreements among shoe firmsfound in the sample are agreements for commer-cial cooperation. Specifically, 8 firms have created

a trading company with some partners, in mostcases aimed at entering in new markets.

The most common type of linkages among shoefirms are informal relationships: in our sample42% of the firms have frequent informal contactswith other firms in their area, 28% have somecontacts and the remaining 30% very rarely haverelationships. Considering size, small (42% of thesmall firms in the sample) and medium firms(47%) have more frequent informal contacts thanlarge firms (27%).

As regards the nature of these informal ties, themost important one being the meetings organizedby the entrepreneurial associations (42% of thesample), followed by spatial proximity (24%).Moreover, 71% of small firms, 60% of mediumand only 18% of large enterprises believe thatinformal contacts are very important. This dif-ferent approach towards informal contacts of thedifferent categories of firms may be interpretedas a tendency of small, losing firms to relyon informal cooperation as a sort of solidaritynetwork, allowing them to survive beyond marketrules (Rabellotti, 1995a).

(d) Linkages with the labor market. The existenceof a reservoir of skilled labor is always assumedto be one of the externalities of production orga-nization in highly specialized clusters of firms.The Italian industrial districts are described in theliterature as areas where specialized jobs aretaught to children by their parents and where skillsare accumulated and transmitted from one gener-ation to the other.

This process of collective learning, of knowl-edge accumulation among people, and therefore ofknow how circulating among firms through laborforce mobility, enhances local innovation capa-bility. In other words, the innovation process,which usually takes place inside the firm, becomesa collective process in the industrial districts,based on common knowledge accumulated inpeople rather than firms. This mechanism wasconfirmed by the inquiry: the majority of the firmsinterviewed employ people who have already beentrained in other firms. This strategy is particularlyfrequent among small firms, which are thus ableto save on training costs.

Nevertheless, the labor market has changed inthe last ten years: a majority of the firms inter-


viewed (55%) stated that the availability of skilledlabor was good during the period from 1980–1985,while in the second half of the 1980s, 70% of thefirms said it had dropped. In our survey, accordingto 46% of the sample firms availability of skilledlabor is low, according to 36% is good and to 18%is very good. The entrepreneurial associationsand a number of firms interviewed believe thatimproved welfare and better education, due to theexpansion of the footwear sector in the areas,results in an increasing tendency for local youngpeople, who are better educated than their parents,to abandon the sector if they can find alternative,mainly non manual jobs. The resulting interrup-tion of the accumulation and transmission ofskills from parents to children, from one genera-tion to the other could, in the future, underminethe collective learning effect which has been soimportant in determining the competitiveness ofthe districts.

Concerning the availability of unskilled labor,it seems to be rather good according to the firmsinterviewed: 44% of them said that availability isvery good, according to another 44% it is rathergood and in 10% is low.

In the two areas labor climate is also charac-terized by a high flexibility of the labor force, interms of the willingness to work extra-hours,week-ends etc. and by generally good labor rela-tionships. All but two of the firms intervieweddeclared that there is a high availability amongtheir staff to work extra-hours when needed.This flexibility is crucial in an industry such asfootwear where production is almost seasonal andthe orders tend increasingly to arrive at the lastmoment. Finally, the existence of very friendly andeasy relationships with the workforce is one of themain advantages of being located in the two areas,stressed by a large majority of the entrepreneursinterviewed.

(e) Institutional linkages. The importance of insti-tutional support in the growth process of Italianindustrial districts has been much emphasized inthe literature. This has fueled the myth of an effi-cient local government able to intervene to supportthe needs of local industries, creating public orsemi-public centers for real services, technolog-ical development, commercial promotion and soon (Brusco, 1989 and 1992). However from a

careful investigation of the available literature onindustrial districts in Europe, Schmitz and Musyck(1994) conclude that the picture of institutionalintervention coming out is patchy and there arefew evaluations of the services supplied from theusers’ viewpoint.

From the investigation carried out in this study,it was clear that central government interventionwas of limited importance. Only 20% of thesample firms said they have received any finan-cial incentives. Moreover, the majority of the firmscomplained about the enormous bureaucratic dif-ficulties to access the incentive schemes, whichmainly consist of easy credit terms for small firmsor enterprises with innovative projects.

Regarding local governments, they mainly con-tribute with financial support for some of theinitiatives promoted in the two areas by theentrepreneurial associations. Both in Marche andBrenta, in fact, the main institutional actors are thelocal entrepreneurial associations, which supply anumber of services to the members and play animportant role in promoting initiatives to supportthe sector. Among the sample firms the most usedservices are tax counseling and the organizationof trade fairs.

In Brenta, the local association ACRIB(Associazione Calzaturifici della Riviera dellaBrenta) was established by some local entrepre-neurs more than 30 years ago and since then it hassupported several initiatives like a large exportconsortium and a center for technological assis-tance and training. The export consortium (namedConsorzio Maestri Calzaturieri del Brenta),created in 1967 with an initial public financing,manages collective advertising campaigns, assistsfirms in their export activities, and runs a databank on 20,000 customers all over the world.Moreover, the consortium organizes a collectiveexhibition in Dusserdolf twice a year andpromotes researches and market surveys. In 1986with public money provided by a law aimed atsupporting the crafts, ACRIB promoted a new ini-tiative: the Consorzio Centro Veneto Calzaturiero.The main areas of intervention of this institutionare: technology assistance, quality control, envi-ronmental control and training.

Besides ACRIB, in Brenta there is also a quiteactive association of artisan firms with morethan 500 members in the footwear filière which


supplies several services like book-keeping, laborand tax assistance, consultancy for facilitatingaccess to credit.

In Marche, there are two entrepreneurial asso-ciations, in Fermo and in Macerata, which supplyservices like training, book-keeping, fiscal andfinancial assistance to their members. In Fermothere are 650 members, and only 140 in Maceratabecause the footwear area in this province issmaller than in the area around Fermo. The twoassociations have recently launched a new initia-tive: a specialized center (SCAM – Società per laCalzatura Marchigiana) supplying technologicalassistance, training, fashion information andpromotional activities. The services are usuallysupplied at very low costs because they are par-tially financed by public grants (i.e. the trainingcourses are financed through EC grants). An inter-esting project of the center is a monthly tradefair, where local producers exhibit and sell theirproducts, outside the traditional seasonal appoint-ments of the main national and international tradefairs.

During the 1980s thanks to an initial publicgrant several export consortia were created inMarche, to supply services like translations, faxfacilities, assistance with export procedures.Recently, some of them have moved towards amore specialized variety of services like marketsurveys and promotional activities in new markets(i.e. Japan, Arabian countries, Australia). Amongthe services supplied there is also the access tointernational data banks about customers, inven-tories of unsold products and subcontractingopportunities.

In Marche there are also a few consortia tofacilitate access to credit. One of them, promotedby the entrepreneurial association in Fermo, has300-odd members and provides common guaran-tees, easy credit terms and access to long andmedium-term credit.

Both in Marche and Brenta most of the firmsprefer to deal with local banks because they oftenknow the people who work there and usually thishelps in obtaining credit. However, it must be saidthat the easier access to credit in local banks,based on trust, friendship and high circulation ofinformation, as opposed to national credit institu-tions, may become an important constraint whenfirms go through periods of crisis. Local banks

immediately know about the difficulties and makeaccess to credit even more difficult. This happenedrecently, according to a large part of the firmsinterviewed both in Marche and Brenta, when thereaction of local banks to the difficulties in thefootwear industry was to drastically reduce creditavailability to shoe firms.

To conclude, the activities of Marche andBrenta entrepreneurial associations were, ingeneral, assessed positively by the entrepreneursinterviewed, who also participate in them quiteactively. However, the tradition of institutionalintervention is more firmly established in Brentathan in Marche and this is confirmed by the wellorganized activities supplied by the specializedcenters. In Marche, according to many of theentrepreneurs interviewed, there is little traditionof collaboration and institutional intervention andthis is confirmed by the only recent establishmentof a service center in an area of specializationwhich is the most important in Italy. The need forgreater institutional intervention is also confirmedby the general opinion among the entrepreneurs inMarche, who very often quote the Brenta case asa good example of how institutions can help thesector.

3.2. The Mexican clusters: Guadalajara andLeon

Mexico was a protected market for a long time,having adopted an import-substitution strategy,which was drastically abandoned in 1988 with theopening up of the economy to foreign competition.The acceleration of the opening up of the Mexicanmarket to international competition, through theelimination of import licensing and tariff reduc-tion, had a big impact on the footwear industry.The market was flooded with imports whichincreased from US$ 13.7 million in 1987 to US$148.2 million in 1991.

For many decades the domestic producers tookadvantage from a protected market where therewas excess demand; to make money in the sectorwas easy because every kind of product was sold,no matter what its quality, design and cost.Nowadays, international competition is becomingstronger, and the Mexican footwear enterprises arestarting to realize that they must increase their


efficiency if they want to survive and grow(Rabellotti, 1995a).

(a) Backward linkages. In Mexico the relation-ships between suppliers and shoe producers areless collaborative than in Italy, being based mainlyon a pure market mechanism. Many sample firmscomplain about the low quality of components andraw materials, the scarce attention to fashionchanges and the bad service provided by theirinput suppliers. In their turn, suppliers do notaccept responsibility for their low developmentand accuse shoe entrepreneurs of having alwaysadopted a strategy more focused on price thanquality. Suppliers complain of unstable demand,of small order sizes, of continuously changingproducts, and of delays in payments. Both thesuppliers of components and the manufacturersblame each other and the main deficiencies are ofcommunication and collaboration among the twolinked sectors.

However, with the opening of the market andthe increase in shoe and component imports, therelationships between shoe producers and sup-pliers are improving and becoming more collabo-rative. The majority of the suppliers interviewedagree that they now try to work close to their cus-tomers in developing products and they are alsoavailable to adapt products to the needs of theshoe producers. On the other hand, some of thefootwear firms interviewed are now more satisfiedabout their relationships with suppliers and theyare trying to set up stable linkages, building upan ability to cooperate in defining fashion trendsand product characteristics.

To increase the stability of demand, somegroups of shoe producers have tried to organizecommon purchases. Some wholesalers are alsoworking in this direction, organizing the purchasesof components and raw materials for all theirclients and therefore guaranteeing large and stableorders for the suppliers. Moreover, both in Leonand Guadalajara the local Credit Unions of shoeentrepreneurs have recently launched a programfor pool purchasing; their objective for the nearfuture is to import components from abroadjointly.

Some efforts towards a more cooperativeattitude between shoe producers and suppliershave also been undertaken by the Camara del

Calzado – the footwear entrepreneurial associa-tion – and the association of suppliers in Leon,which have begun to organize meetings to discussand jointly elaborate fashion trends. Together withthe Camara del Calzado in Guadalajara, the twoassociations in Leon have also begun to work onthe standardization of the measurement system.The lack of a standard, commonly accepted mea-surement system is a major obstacle for the devel-opment of an efficient system of specialized firms.The case of lasts is well representative of thesituation: in Mexico every footwear firm requiresits own models, based on a measurement systemwhich is different for every firm.

Concerning subcontracting of phases of the pro-duction cycle, in Mexico, according to severalrecent studies (Boston Consulting Group, 1988;Concalzado, 1991; Dominguez-Villalobos andGrossman, 1992) and according to the results ofour own survey, the level of the division of laboris generally low (Table III) and certainly muchlower than in Italy (Table II). Mexican shoe firmsgenerally decentralize some phases of productioncycle only when they receive orders too large fortheir internal capacity. About 50% of the samplefirms do not externalize any stage of the produc-tion process. This low level of the division of laborcan be explained by two main factors:

• backward linked industries have remained at alow level of development with regard to design,fashion quality of components and servicebecause the former protection of the domesticmarket has limited competition and, therefore,the incentive to innovate. We know from AdamSmith that the division of labor is limited bythe extent of the market and this seems to applyto our Mexican cases;


TABLE IIIThe extent of externalizing production tasks: Mexico

(% of sample firms)

% of production None <50% 51%–90% >90%externalized

Cutting 094 06 0 00Sewing 080 20 0 00Bottoms 059 04 4 33Finishing 100 00 0 00


• the sector lacks a standard technical languageand a common, universally accepted, measure-ment system and this strongly increases trans-action costs.

The poor level of development and the diffi-culties of communication with backward-linkedindustries induce many shoe firms to internalizeas many phases of the production cycle as possiblein order to reduce their dependency on an unstable,low quality supply. Vertical integration meansdifferent phases of production are carried outinside the firm, with a lot of problems of organi-zation of the production cycle, because the dif-ferent phases are characterized by very differenteconomies of scale, different degrees of laborintensity and different processing times. This pro-duction structure is even common among verysmall enterprises.

More recently, there is an increasing numberof firms which, after having interrupted theiractivity as shoe producers because they were hitby fierce competition, began to work as subcon-tractors specialized in the production of uppers.Most of the subcontracting firms interviewedstated that they strongly depend upon and sufferfrom the still very unstable flow of work fromthe shoe producers. According to them, this insta-bility limits possibilities for the developmentof a well-organized and efficient system of pro-cess specialized enterprises, like in the Italiandistricts.

To conclude, in the Mexican clusters notwith-standing the local concentration of suppliers, thelack of competition has often reduced the advan-tages in terms of price and service deriving fromthe good local availability. Nevertheless, with theopening of the market the quality of linkagesbetween suppliers and manufacturers have begunto improve. Besides, there is also a trend towardsan increasing decentralization of some phases ofproduction. Nonetheless, to increase the degree ofdivision of labor and the competitiveness of theindustry, the two clusters need to become well-organized systems, where shoe producers may relyon quality and prompt deliveries from subcon-tracting firms and their linkages with suppliers arenot only based on a price factor, disregarding someimportant aspects such as fashion and design,material quality and delivery service. From this

point of view the Italian footwear districts mayactually provide some interesting lessons.

(b) Forward linkages. The Mexican footwearindustry traditionally neglected commercializationand marketing because the domestic market wasclosed to international competition for long timeand this allowed the domestic producers toproduce shoes which were easily sold on the pro-tected market, no matter what the quality, design,fashion content were. Generally speaking there-fore, the Mexican footwear firms suffer fromproblems similar to those of the Italian enterprises:they have limited control over their market andlittle knowledge of it; they depend on non-exclu-sive agents and they are not used to adoptingactive commercial strategy to sell their productson a competitive market. Among 63% of thesample firms the main factor of competition isprice, only 24% of the firms interviewed believethat design is an important factor and 12% thatquality is important.

Considering marketing channels, 74% of thesample firms sell part of their production throughnon-exclusive agents, 51% through wholesalersand finally 16% through trading companies orshops owned by the same firms and another 16%directly to retailers.

One of the main consequences of a very frag-mented distribution system dominated by inde-pendent retailers is the small size of orders:according to the Boston Consulting Group study(1988) the most frequent size of orders in thedomestic market is around 300 pairs, but some-times shoe producers receive orders for very fewpairs. The possibility of working on the basis oflarger orders and therefore to be able to planproduction process on a longer term basis and toexploit the economies of scale deriving from spe-cialization in a few products, is the main reasonthat has pushed an increasing number of firms tosell to retail chains and supermarkets. Anotherimportant reason stressed by some of the firmsinterviewed is related to better payment termsguaranteed by retail chains and supermarkets ascompared with small family shops.

According to the firms interviewed which sellto retail chains, the relationships are definitely ofdependency because these chains generally adopta strategy based on the search for the lowest


possible price on the market. Many stories are toldabout some of these chains contacting some firmsto buy their models and then commissioning otherfirms, on the illegal market, to manufacture thesame products at a lower price. Some of thesechains are also believed to be responsible for mostof the illegal imports of very cheap synthetic orfabric shoes, which have invaded the low-pricemarket during the last few years.

However, a few firms interviewed told us that,although most of the retail chains still maintain avery traditional strategy of commercializationbased on price, some of them and particularlysupermarkets, are progressively increasing theirattention to quality and therefore some of the shoeproducers try to limit their dependency on them,increasing the stability of their relationship bysupplying good service and a stable quality ofproducts.

Recently, an interesting experiment was startedaimed at developing a stable and constructiverelationship thanks to a few wholesalers. Theyhave adopted a new strategy aimed at selling aquality product and have therefore selected agroup of shoe firms with which to collaborate.Technical staffs employed by the wholesalersregularly call on the shoe enterprises, controllingquality and giving advice on technological andorganizational matters; moreover the wholesalersorganize a system of pool purchases of some keycomponents or raw materials like leather, in orderto guarantee a stable quality level, better pricesand good service. Among the firms interviewed,the ones linked to wholesalers through such arelationship were generally very satisfied, notonly with the sales, but mainly with the comple-mentary services, like technological and organi-zational assistance, they receive. Many of themhave been able to introduce important improve-ments in the organization of the productionprocess and in the quality of their products thanksto the collaboration with these wholesalers.Naturally, they realize that depending on one maincustomer, who can always find another enterpriseable to make the same product at a lower price, isvery risky; but most of the firms also realize thatthey do not have enough marketing and commer-cialization skills to compete on the market ontheir own. Therefore they try to build up a linkagewith the wholesalers where both parties have an

interest in collaborating, at least partially offset-ting dependency.

(c) Horizontal linkages. In the two Mexican foot-wear clusters, cooperation among sample firms isaimed at the exchange of technological informa-tion and sometimes of machines in 12 firms, atsubcontracting and farming out part of the pro-duction when they have excess orders in 9 firmsand at the establishment of commercial agree-ments in 6 cases. In the majority of the firms inter-viewed the interactions take place among smallgroups of firms linked by family ties and/or long-standing friendship.

In Guadalajara, linkages among firms outsidefamily groups have been favored by an initiativepromoted by the local Camara del Calzado,inspired by a UNIDO methodology initially aimedat favoring the growth of small rural enterprisesto reduce emigration towards urban centers. Theproject consists of promoting the creation ofagrupamientos industriales, initially based on acourse for entrepreneurs who had to agree toorganize a visit to their own firms by the otherentrepreneurs in the group and to allow an outsideexpert (a business student in her/his final year) tomake diagnoses of their firms. In 7 years 7 groups,involving about 120 enterprises, began in this wayand afterwards some of these groups have carriedon with regular meetings to discuss problemsrelated to technology, marketing, suppliers and soon, and to exchange information about clients,machines, workers, and orders.

Another interesting initiative has very recentlybeen promoted at national level for the creation ofempresas integradoras. The program, which wasinspired by the Ministry of Trade after a visit tothe Italian industrial districts, aims at creatingcompanies established by groups of firms forjointly selling products or buying inputs or anyother common objective. The new companiesenjoy financial and tax incentives and facilitatedbureaucratic procedures. In the survey we foundthree projects for the creation of empresas inte-gradoras: one is for the creation of a new commonsite for a group of 7 enterprises linked by familyties, another is a group of 15 very small firms inLeon (with fewer than 7 employees in each firm),which after a successful experience of joint par-ticipation in the local trade fair, would like to


create a trading company with a single brand namefor all their products and finally there is a groupof 5 tanneries in Leon, which would like to createa company for joint buying and selling, and sup-plying common services like book-keeping andtraining.

Apart from these initiatives aimed at favoringinteractions among firms, informal relationshipsare frequent among 80% of the enterprises inter-viewed and they are considered an important assetby almost all. The role of the entrepreneurial asso-ciations seems to be very important in inducinginformal contacts among firms because, accordingto 37 firms, events organized by the Camara arethe most important opportunities for informalexchange with their colleagues. Family ties alsoplay an important role in 17 firms and, finally,social events are important for 11 firms. Severalforms of cooperation in many different aspects ofa firm’s life, arise from these informal contacts.These include exchanges of technological andmarket information, exchanges of machinery andworkers, sub-contracting of orders in case ofexcess demand, joint commercialization, joint pur-chases of inputs, and joint recovery of credits.

Notwithstanding a lot of informal linkagesare going on within the two clusters analyzed, sig-nificantly 65% of the interviewed firms expresseda desire for more stable, organized and focusedon specific objectives relationships with otherfirms.

(d) Linkages with the labor market. Analyzing thecharacteristics of the labor market, the mainproblems emphasized by the sample firms werein order of importance: the low availability of askilled workforce and the high turnover of thelabor force. According to 65% of the sample firms,skilled labor is a scarce resource and qualifiedworkers demand relatively high salaries and good-working conditions. This explains why small andmedium firms, which usually can only afford topay lower salaries and offer more unstableworking conditions, because of the irregularity oforders, have more difficulty in hiring. Unskilledlabor represents a less important problem (only 4firms believe that locally there is a low avail-ability), because there is a large reservoir ofyoung people, mainly women, available to work.According to most of the sample firms, the main

problem with unskilled labor is the high turnover,because workers move away from the footwearsector as soon as they find a job in other, moreremunerative sectors or, in the case of women,when they get married.

Training was also investigated. In most of thesample firms training takes place inside the samefirm. External training is generally reserved forsupervisors, technicians or designers; less skilledworkers like edgers are very rarely trained outsidethe firms. A problem emphasized by many firmsis the risk of losing their workers sent to be trainedoutside, because they find better paid jobs. Skilledworkers are heavily sought after and frequently‘pirated’ from one firm to the other. This is a verycommon problem among small scale enterprises,which usually pay lower salaries than larger firms:a typical case of market failure which justifiessome kind of extra-market intervention to makeup for the firms’ negative attitude about externaltraining for their workers.

Finally, the rate of unionization is very low bothin Guadalajara and in Leon, because the sectorgrew up in a family environment with entrepre-neurs who, in most cases, started as manualworkers in other firms and therefore establish rela-tionships based on friendship and solidarity withtheir employers.

(e) Institutional linkages. At national level thereare no specific policies addressing the footwearsector, nevertheless some of the sample firmswhich export part of their production have accessto the scheme of financial incentives for exportingenterprises. Other firms have obtained easy creditterms from National Financiera, the Mexicanbank for industrial development. A recent policyinitiative which seems to have some potentialitiesfor supporting the footwear industry is theprogram to favor the creation of empresas inte-gradoras, described above.

At local level, a number of important institu-tions have been identified both in Guadalajara andin Leon, some are associations of firms, whileothers provide special services to the footwearindustries. Entrepreneurial associations (Camaradel Calzado) are the most important institutionssupporting the footwear industry in Mexico; thereare three local associations, in Guadalajara, Leonand Mexico City, and a national one aimed at


coordinating the activities of the three localCamaras. The role of these associations, financedthrough the membership fees and profits from theorganization of the trade fairs, is to supply serviceslike fiscal, legal and labor advice, managerialtraining, organizing of trade fairs and lobbyingactivities at a political level. Entrepreneurial asso-ciations also exist for the tanning and componentsectors.

In Leon there is also a center for research andtechnological assistance (CIATEG – Centro deInvestigacion y Asistencia Tecnologica del estadode Guanauajato), created as an agency of thecentral government, and which can be consideredan arm of the Mexican National Science andTechnology Council. The center was created withthe objective of supplying technical and qualitycontrol services and specialized training to thefootwear industry at national level; in fact, it ismainly used by Leon’s enterprises, leaving shoefirms in the rest of Mexico almost without anytechnological support.

A similar institution in Guadalajara is theInstituto Tecnologico del Calzado (ITC), estab-lished in 1984 with support of the local Camaraand a grant from the World Bank and devotedto training and technological research. Its activi-ties include a diversified program for trainingdesigners, skilled manual workers like edgers, andalso managers and entrepreneurs.

Both in Guadalajara and Leon there is a creditunion aimed at obtaining credit for its membersat more favorable conditions than the market ratesbecause it can borrow directly from NationalFinanciera. In Guadalajara the credit union islinked with the Camara and it was promoted byan initiative of some entrepreneurs taking part inthe agrupamientos industriales. Its activity hasrecently increased thanks to the active participa-tion of a number of small entrepreneurs who gotinvolved in 1991 when the union was suffering abad crisis due to the decrease in credit availabilityin the country. In Leon, the credit union is fairlyrecent (established in 1992) and it does not haveany linkage with the local Camara.

The existence of a relatively well developedinstitutional support network for the footwearfilière is a very important condition in the caseof adoption of a growth strategy based on anapproach at system level. Institutions like the

Camara del Calzado can have a very relevant rolein diffusing among entrepreneurs a systematicvision of their business, in other words the ideathat the survival and growth of their own firmsstrongly depends on the development of the wholesystem of shoe firms, suppliers, buyers, marketagents, service firms and supporting institutions.

4. A classification of collective effects

Section 3.1 and 3.2 describe the linkages amongeconomic actors found in the case studies in Italyand Mexico. The aim of this section is to classifythe collective effects deriving from those linkagesin terms of the concepts introduced in section 2:external economies and cooperation effects. Inorder to classify the collective effects, a two-dimensional classification may be usefully intro-duced: in Table IV the collective effects aremapped out on one side according to their beingincidental and excludable and on the other sideaccording to their static or dynamic objective.

Let us begin with external economies whichderive from unplanned, incidental relationshipsamong the economic actors who interact withinthe system. The most typical incidental relation-ships usually going on in every district and there-fore also in the Italian and Mexican footwearclusters are the frequent social occasions, likecasual meetings in the streets, cafes, clubs andbusiness occasions, like meetings at entrepre-neurial associations and at suppliers’ facilities. Aflow of information about products, markets,fashion trends, bad and good customers, suppliers,process specialized firms and technology is gen-erated from these continuous interactions andfreely circulates within the districts. This circula-tion of information generates both static anddynamic external economies.

With regard to static external economies, thefree circulation of information may allow firmsto be productive because they get access to infor-mation which otherwise they could not afford.This type of external economy is particularly sig-nificant for small firms, which can rarely affordmarket studies, participation in foreign exhibitionsor subscriptions to data banks. The free avail-ability of information also contributes to the birthand survival of small firms; this allows a contin-uous renewal of firms and recycling of the human


and capital resources available in the districts.Nevertheless, it also permits the survival of someinefficient firms, which only have a very superfi-cial market knowledge and which base their com-petitive edge on their ability to make shoes andtake advantage of low entry barriers. This categoryof firms can be defined as followers or, even free-riders of the district, because they usually tend toexploit resources created by other more innova-tive firms, for instance they imitate successfulproducts at lower costs and lower quality. It canbe said that these firms ‘use’ the district to com-pensate their structural and strategic shortcomings.Furthermore, it should be added that, althoughduring the period of excess demand they couldeasily find a market for their products both in Italyand in Mexico, nowadays facing increasing com-petition they are usually the first to suffer. So, forinstance, in Italy many small firms unable todevelop an independent product and marketingstrategy become subcontractors for other enter-prises, or even close down. In Mexico, the samephenomenon occurs, with an increasing number of

firms closing down in the formal market tobecome informal activities, non-officially regis-tered and therefore able to survive without payingtaxes and paying less than the legal minimumwage. We can conclude that these externaleconomies contribute to the collective efficiencyof the systems analyzed, but firms which rely toomuch on such incidental effects have shown lowperformances.7

The free circulation of information can alsohave dynamic effects, generating a sort of demon-stration effect on attitudes and motivations whichmay induce economic actors to introduce innova-tions in processes, products or forms of organiza-tion and contributes to the growth of the system.This positive spontaneous effect of clustering iscommon in Brenta and Marche, in Guadalajaraand Leon, as well as in many other clusters offirms in industrialized and developing countries.

The local concentration of firms which produceor sell inputs, machinery and services and firmsspecialized in some phases of the productionprocess at competitive prices, at a high level of


TABLE IVCollective efficiency: a classification of effects

External economies Cooperative effects

Static Marche and – high availability of free information; – cooperation with process specialized firms;Brenta – high availability of inputs at competitive – cooperation in export consortia, in credit

– prices, at great speed, at low – consortia, entrepreneurial associations;– transaction costs;– high division of labor; – collective reputation;

Guadalajara – high availability of free information; – cooperation in export consortia, in creditand Leon – good availability of inputs at market – consortia, entrepreneurial associations;

– conditions; – cooperation in agrupamientos industriales and – low division of labor; – empresas integradoras;– collective reputation; – cooperation with buyers;

Dynamic Marche and – demonstration effects on attitudes – strong cooperation with suppliers of raw Brenta – and motivations; – materials, components, machinery;

– high collective learning; – cooperation with process specialized firms;– rare cooperation in export consortia, credit – consortia, entrepreneurial associations;

Guadalajara – demonstration effects on attitudes – very little cooperation with suppliers of raw and Leon – and motivations; – materials, components, machinery;

– low collective learning; – rare cooperation in export consortia, credit – consortia, entrepreneurial associations;– rare cooperation in agrupamientos – industriales and empresas integradoras;– cooperation with buyers;

specialization generates some important staticexternal economies. Every firm located within thedistrict can in fact save on input costs because ofthe high competition among suppliers, becausetransaction and transport costs are lower due to thespatial and cultural proximity and because firmscan maintain very little inventories, being able tobuy what they need rapidly. Moreover, shoe firmscan easily put out some phases of the productionprocess to highly specialized firms, able to exploitthe different economies of scale which charac-terize the phases of the production cycle. Theseexternal economies definitely characterize the twoItalian footwear districts and this represents oneof the main competitive advantages that promptedthe growth of the Italian footwear industry duringthe 1970s.

With regard to the Mexican clusters, the situa-tion is different because the long closure of thedomestic market has not encouraged the growth ofa competitive industry of suppliers and processspecialized firms. Mexican shoe producers inGuadalajara and Leon can therefore locally buymost of the inputs they need and they may saveon transportation costs, but the lack of competi-tion has greatly reduced advantages in terms ofprice, quality and service. Moreover as seen insection 3.2, Mexican shoe firms are usually morevertically integrated and they tend to put out somephases of the production process only when theyhave excess orders, without any important gainsin terms of specialization. The low degree ofdivision of labor is also partly due to the hightransaction costs generated by the lack of a stan-dardized measurement system. From the empiricalinvestigation it appears that the advantages interms of face-to-face contacts, typical of clus-tering, are partially offset by the high transactioncosts deriving from the lack of standardization.This makes the relationships of shoe producerswith suppliers of components and process spe-cialized firms particularly difficult and costly andlimits the gains from clustering in Mexico.

Another type of static external economies,typical of industrial districts, is the effect of col-lective reputation, in other words the attraction ofcustomers achieved by a large concentration ofspecialized producers. So for instance, Germanbuyers know very well that they can find highquality women’s shoes in Brenta and they go there

to buy. The same is true in Mexico: Guadalajarais the largest market in the country for women’sshoes and Leon for men’s and children’s shoes.This enormously facilitates access to distantmarkets.

There is a negative side of these externaleconomies, too. The reputation of the district isin fact usually rather homogeneous and therefore,for instance, shoe makers from Marche are knownfor producing a medium-low quality product. Thiscollective reputation may have a negative effecton firms trying to differentiate their productiontowards different segments of market.

Finally, the high local availability of skilledlabor is a dynamic external economy from whichthe firms located in Marche and Brenta and tosome extent the firms in Leon and Guadalajaratake advantage. The accumulation of know how inpeople moving from one firm to the other gener-ates a process of collective learning whichenhances the system’s innovation capability. Thistypical characteristic of Italian industrial districtsis less evident in the Mexican clusters, where theavailability of skilled labor is more limited. It mustbe added that some changes may soon occur alsoin the Italian districts with respect to collectivelearning as a consequence of the increasingtendency of well-educated young people to seeknon-manual jobs, mainly in sectors other than thefootwear industry.

Moving to cooperative effects, in Italy explicitcooperative linkages between the shoe producersand their suppliers of raw materials, componentsand technology have been frequently found in theempirical investigation. Shoe firms and their sup-pliers often work together to develop new productsand the cooperation is based on the self-interest ofboth parties in maintaining a stable relationship.These cooperative linkages represent an importantcontribution to the system’s capability of innova-tion and growth and their effects can be thereforedefined as dynamic. Thanks to those linkages,shoe firms are able to supply a very diversified,quality product in a time-to-market which isshorter when compared with the shoe industry inmany other countries.

The links between shoe producers and processspecialized firms are quite frequently character-ized by a clear dependence of subcontractorson shoe enterprises, which define the payment


conditions and can choose another firm if theyare not satisfied of the quality or the service.Nevertheless, in some cases there is also someform of cooperation, because some shoe firmsprefer to have stable relationships with their sub-contractors and therefore they try to solve inci-dental problems and supply the needed trainingand sometimes even financial or technical help.Moreover, these linkages are facilitated by the lowtransaction costs within the districts due to theeasy face-to-face contacts. The cooperative effectsderiving from these linkages can be either staticor dynamic.

In Mexico, similar cooperative linkages withprocess specialized firms are very rare becauseshoe firms tend to be highly vertically integrated.Only a few firms have been recently pushed byincreasing competition to move towards a higherdivision of labor and more cooperative linkages.But, this is still an in fieri tendency.

Also the linkages with suppliers are less coop-erative than in Italy and they can be defined asmarket relationships, based on a price factor. Thismeans that thanks to clustering the Mexican shoeproducers can locally buy raw materials, compo-nents and machinery but as opposed to the Italiandistricts they rarely cooperate developing theseinputs and they must buy what is available on themarket, at the market terms. The weakness ofbackward cooperative linkages can be explainedby the long closure of the domestic market, whichhas not favored competition based on productquality, fashion contents and design, either in theshoe market or in the component and accessorymarket. Only with the recent change in tradepolicy, namely with the opening up of the market,have some of the shoe firms and some of theirsuppliers begun to realize that they belong to asystem in which the success of one firm stronglydepends on its interactions with other firms.Therefore, relationships are becoming more coop-erative and firms are trying to build stable linkagesbased on a mutual self-interest in improvingquality and service.

In Italy, some cooperative linkages have beenfound among firms which belong to export con-sortia and in other commercial agreements.Linkages in export consortia may be defined asdynamic if they are aimed at discovering newmarkets; they are otherwise static. In Mexico,

these type of consortia have not been found.Relationships within credit consortia, entrepre-neurial associations and other service centersaimed at supporting the footwear industry can bealso defined as cooperative linkages, and they canbe found both in the Italian districts and in theMexican clusters. All these relationships are char-acterized by an explicit decision of a selectedgroup of firms to cooperate for several aims likejointly selling their products, dealing with banksor utilizing joint services supplied by entrepre-neurial associations. Their effects are thereforestatic, apart from when they involve cooperationfor carrying out some joint innovative projects.

In Mexico, other forms of cooperative agree-ments are the agrupamientos industriales andthe empresas integradoras. These are two experi-ences promoted precisely with the aim of favoringthe development of cooperation among selectedgroups of firms. Their effect is more frequentlystatic, when they regard cooperation among firmswhich exchange machinery, labor force or ordersor buy inputs jointly. They became dynamic when,more rarely, they involve cooperation on innova-tive projects. Besides, we may add that theseforms of cooperation can be also found amonggroups of firms linked by family ties.

Finally in Mexico, a few cases of linkages havebeen reported between shoe producers and buyers,based on cooperation to produce good qualityproducts and sometimes to introduce innovationson products, processes and organizational forms.However, relationships based on a strong depen-dence from the shoe producers’ viewpoint arealso common, because they often sell most of theirproduction to only one customer. In Italy asidefrom a few exceptions, buyers do not play a reallysignificant role in commercialization and they tendto have a market relationship with their customers,without intervening in the phase of product devel-opment and production.

The linkages with non-exclusive agents canbe also defined as pure market relationships bothin Italy and in Mexico. From the empirical inves-tigation it appears in fact that the relationshipswith non-exclusive agents very rarely involve anexchange of information, in most cases they areinstead pure market linkages in which agents sellthe products of different firms, without a real con-tribution to the development of new products


based on market needs. This represents an impor-tant limitation both for the Italian and the Mexicanfootwear industry, which have developed veryweak forward linkages with the market and haveone of their biggest weak points in the commer-cial function.

In the Mexican case the reason for thisweakness may be found again in the lack of com-petition, which has made it unnecessary to developthe marketing and commercial function for sellingproducts in the protected market. The reason forthe weakness of forward linkages in the Italiandistricts is more complicated. In this case thereason can be found in the capability of the Italianfootwear firms, during the 1970s and the first halfof the 1980s, to impose their products on the inter-national market without the need to develop acommercial and marketing function. Before theadvent of the international competitors likeBrazil, Spain or Portugal, Italian shoes were in factbought for their quality and fashion content andfor the service supplied in terms of speed, varietyof products, flexibility. This characteristic of themarket has favored the development of districtsthat are highly developed on the production sidebut weak in commercial and marketing functions.Since the mid 1980s, the increasing competitionin the international market, due to the ‘arrival’ ofnew producers, mainly from newly industrializingcountries, has generated some difficulties in thetwo footwear industrial districts analyzed, as wellas in many other Italian areas specialized infootwear. The number of firms has begun todecrease, the exports have slowed their continuousgrowth and the domestic market has been invadedby imported shoes competing with the Italianproducts. As a result of these changes, the needto develop a marketing function has become clearto an increasing number of firms, which react tothese changes in several ways, influencing theorganizational structure of the industrial districts(Camagni and Rabellotti, 1995).

To conclude some general considerations aboutthe degree of collective efficiency in the four casesanalyzed can be drawn from Table IV:

• external economies emerge both in Italy and inMexico. However, there are some main differ-ences concerning the availability of inputs, thedegree of labor division and the availability of

skilled labor. In all these cases the productionof external economies is higher in the Italiandistricts than in the Mexican clusters;

• cooperative effects are definitely more commonamong the Italian firms than the Mexican ones.The main differences are in the relationshipswith suppliers and process specialized firms,which are based on cooperation in Italy andmostly on market rules in Mexico.

As expected from the comparison of very dif-ferent realities – two proper districts in Italy andtwo clusters in Mexico – we may conclude that thedegree of collective efficiency, to which externaleconomies and cooperative effects contribute, ishigher in the Italian districts than in the Mexicanclusters, where firms can exploit fewer clusteringadvantages. In the final section, we elaborate onthis conclusion.

5. Some final considerations

From the results of the empirical investigations wecan derive a confirmation about the importanceof collective efficiency, not only in the two Italiandistricts, but at some extent, also in the twoMexican cluster. The analysis of the collectiveeffects deriving from the linkages among theeconomic actors within the Italian districts andthe Mexican clusters has in fact confirmed theexistence of a certain degree of collective effi-ciency.8 Nevertheless, although collective effi-ciency matters both in Italy and Mexico, theempirical investigation has clearly highlightedsome important differences concerning theintensity and quality of collective effects betweenthe two realities studied. Besides, distinctionshave been also identified between the empiricalcases and some of the key stylized facts usuallycharacterizing industrial districts, as stated insection 2.

The first important difference is the scarcityboth in Italy and in Mexico of forward linkages aswell developed as the backward linkages in theItalian districts. Moreover, in the Mexican clustersthe intensity and the quality of the linkages withsuppliers and with process specialized firms isdefinitely lower than in Italy. Concerning externaleconomies, their production in the Italian districtsand in the Mexican clusters is quite high but some


of them do not only contribute to the level of effi-ciency and the growth rate of the systems, ratherthey allow the survival of some low-performingfirms on the market. In Mexico, the low avail-ability of skilled labor generates a weaker effectof collective learning than in Italy. The generalconclusion which can be drawn is the existenceof different degrees of collective efficiency in thecases analyzed. In Italy, collective efficiency ishigh but not as high as expected and in Mexico itis even lower than in Italy.

The differences in the degree of collectiveefficiency between the Italian districts and theMexican clusters may be in part explained by theexistence of differences in external conditions. Themost evident difference is trade policy: in Italy theexistence of a competitive market has favored thedevelopment of a highly efficient system of pro-duction, based on a high degree of division oflabor among specialized enterprises and intensecooperative linkages with suppliers, while inMexico the long closure of the domestic market tointernational competition has induced the devel-opment of vertically integrated firms linked totheir suppliers through pure market linkages. Theimportant lesson, which can be drawn, concernsthe opportunity to carefully take into considera-tion differences in external conditions when indus-trial districts are used as a reference point foranalysis in different contexts.

It must be also stressed that Mexican firms haverecently tried to change and improve their rela-tionships with suppliers on a more collaborativebasis in response to increasing competition. Thesame is true for forward linkages both in Italy andin Mexico: a few firms try to develop new mar-keting strategy and new links with the market toface their commercial weakness. A similar processof change in the quality of relationships wasstressed in a study on the footwear industry in theSinos Valley in Brazil (Schmitz, 1995a).

This issue introduces the question of how thesystems react to changes in external conditionsand the possibility that differences between dis-tricts can be explained by the existence of differentstages and paths of development of clusters. Thequestion of the capacity of clusters to respond toexternal changes transforming their internal orga-nization is central. The literature on industrialdistricts has usually adopted a static approach,

neglecting the possibility that an external radicalchange, like the change in the competitive positionin the cases analyzed, may represent a sign ofrupture in the evolutionary path of the districts andeventually result in a new organizational form,distinct from the original one, and possibly remotefrom the traditional industrial district.

In another paper (Camagni and Rabellotti,1995) we tried to identify some typologies ofchanges and trajectories of evolution of the dis-tricts, based on the recent Italian footwear expe-rience.9 The main result is that the reaction ofthe Italian districts hit by increasing internationalcompetition has not been homogeneous within thesystems, but some leading firms have begun toface competition by developing aggressive mar-keting strategies, for which they have sought somecomplementary resources outside the districts. Theimpact of these new strategies on the structure ofthe districts is a tendency towards hierarchizationwith the leading firms, which on one side organizethe activities of a number of subcontracting firmswithin the districts and on the other interact strate-gically with other economic actors on the outside.To capture the realities of the Italian footwear dis-tricts it would be worthwhile to study how thestructures of the systems evolve from one stageto the other, along their trajectory of development,as a reaction to external events and this involvesexplaining the different behaviors of the economicactors, facing change within the districts.

Similarly, in Mexico with the opening up of themarket to international competition a few leadingfirms have begun to change their relationshipswith suppliers and buyers, trying to set up coop-eration linkages, aimed at improving the productquality, fashion content, design and service. Someattempts to increase the division of labor amongprocess specialized firms have been also registeredin the two clusters studied. Therefore, in theMexican clusters too it is necessary to understandthe evolution of the system, generated by a changein the trade policy and led by the actions of someeconomic actors.

What is common among all the experiencesanalyzed is the replacement of generic inci-dental relationships by strategic cooperation withselected partners, aimed at providing specific com-plementary resources, in other words an upgradingof linkages. The new partners tend to be located


in the districts but they may also be outside, if theresources were not available within the districts.Most of the firms continue to gain from cluster’sadvantages but they are also able to overcome thelimitations of the districts, networking with theoutside world.

To conclude, the more general indication whichemerges for future research is that the comparisonof different clusters at a given point in time is anexercise which has severe limitations. A morefruitful exercise should appropriately comparetrajectories of development, searching for commonpatterns in these trajectories. Are the trajectoriesof development of LDCs clusters characterized bysome common patterns? And do these patternsdiffer from those which characterize districts indeveloped countries? These are all questions forfurther research.

Notes

* This paper presents some of the results of my D.Phil. thesis.During my work, I received constructive comments and sug-gestions from many people; I am particularly grateful forinsightful suggestions from Roberto Camagni, Meine Pietervan Dijck, Gioacchino Garofoli, John Humphrey and HubertSchmitz.1 In what follows, the terms ‘industrial district’ and ‘cluster’are sometimes interchangeable, but it is worth recalling thatwhile an industrial district is always a cluster, as it is evidentfrom the definition proposed in this paper, the reverse is notalways the case. Since Marshall, ‘industrial district’ impliesat least inter-firm division of labor and in most contemporaryanalysis it also implies some cooperation.2 Among the first studies in which the phenomenon was iden-tified are: Bagnasco (1977), Becattini (1979), Brusco (1982),Fuà (1983),Garofoli (1981) and Goglio (1982).3 The term ‘Third Italy’ (Terza Italia) was first coined by theItalian sociologist Arnaldo Bagnasco (1977) to define and dis-tinguish the industrializing areas of north-central and north-east Italy from the underdeveloped south of the country andfrom the traditionally industrialized north-western region. Itconsists of the regions of Friuli-Venezia Giulia, Veneto,Trentino-Alto Adige, Emilia-Romagna, Toscana, Marche andUmbria.4 The book which first brought the success of the districtsto the attention of English-speaking audiences was the SecondIndustrial Divide by Piore and Sabel (1984). Later a numberof articles, some of which had previously appeared in Italian,were translated into English and published in Goodman andBamford (1989), Pyke, Becattini and Sengenberger (1990) andPyke and Sengenberger (1992).5 The choice to include only members of the Associations isjustified by the focus of our inquiry on the different forms ofexplicit and implicit cooperation among firms. We assume that

firms taking part in the Associations are more likely thanothers to entertain relationships with other firms; to becomemember of the Associations may in fact be interpreted as asign of interest in getting in touch with other firms.6 For a more detailed presentation of background informa-tion on the footwear Italian and Mexican industries seeRabellotti (1995a).7 This is an important result of correspondence analysispresented in Rabellotti, 1995a.8 This result is also confirmed by factor and correspondenceanalyses presented in Rabellotti, 1995a.9 Specifically, the research is focused on three districts:Marche, Brenta and Montebelluna (specialized in the produc-tion of ski-boots).

Bibliography

ANCI, 1994, ‘L’industria calzaturiera Italiana’, mimeo, Milan:ANCI.

Bagnasco, A, 1977, Tre Italie: La problematica territorialedello sviluppo italiano, Bologna: Il Mulino.

Becattini, G., 1979, ‘Dal “settore industriale al ’distretto indus-triale”. Alcune considerazioni sull’unità di indaginedell’economia industriale’, Rivista di Economia e PoliticaIndustriale (1), 35–48 (reprinted as: ‘Sectors and/or dis-tricts: some remarks on the conceptual foundationsof industrial economics’ in E. Goodman and J. Bamford(eds.), 1989, 123–135.

Boston Consulting Group, 1988, ‘Industria del Calzado’,Mexico City: mimeo.

Brusco, S., 1982, ‘The Emilian Model: Productive Decen-tralisation and Social Integration’, Cambridge Journal ofEconomics (6), 167–184.

Brusco, S., 1989, ‘A Policy for Industrial Districts’, in E.Goodman and Bamford J. (eds.), 259–269.

Brusco, S., 1992, ‘Small Firms and the Provision of RealServices’, in F. Pyke and W. Sengenberger (eds.), 177–196.

Camagni, R. and R. Rabellotti, 1995, ‘Alcune riflessioni sulladinamica dei milieux calzaturieri in Italia’, in G. Gorlaand O. Vito Colonna (eds.), Regioni e sviluppo: modelli,politiche e riforme, Milan: Franco Angeli, 161–184.

CANAICAL, 1993, ‘Perfil de la Industria del Calzado’,Camaras de la Industria del Calzado, Mexico City: mimeo.

CONCALZADO, 1991, ‘Perfil de la Industria del Calzado’,Calzavance, September.

Cornes, R. and T. Sandler, 1986, The Theory of Externalities,Public Goods and Club Goods, Cambridge: CambridgeUniversity Press.

Dominguez-Villalobos, L. and F. Grossman, 1992, ‘Employ-ment and Income Effects of Structural and TechnologicalChanges in Footwear Manufacturing in Mexico’, WorkingPaper (224), World Employment Programme Research,Geneva: International Labour Office.

Fuà, G., 1983, ‘L’industrializzazione del Nord Est e nelCentro’, in G. Fuà and C. Zacchia (eds.), Industrializ-zazione senza fratture, Bologna: Il Mulino, 7–46.

Gaibisso, A. M. (ed.), 1992, Struttura e competitività delsettore calzaturiero in Italia, Milan: Franco Angeli.


Garofoli, G., 1981, ‘Lo sviluppo delle aree periferichenell’economia italiana degli anni settanta’, L’Industria2 (3), 391–407.

Goglio, S., 1982, ‘Sviluppo regionale e dimensioned’impresa’, in S. Goglio (ed.), Italia: centri e periferie,Bologna: Il Mulino, 175–220.

Goodman, E. and J. Bamford (eds.), 1989, Small Firms andIndustrial Districts in Italy, London and New York:Routledge.

Marshall, A., 1920, Principle of Economics, 8th edition,London: Macmillan (1st edition, 1890).

Piore, M. and C. Sabel, 1984, The Second Industrial Divide.Possibilities for Prosperity, New York: Basic Books.

Pyke, F., G. Becattini and W. Sengenberger (eds.), 1990,Industrial Districts and Inter-Firm Co-operation in Italy,Geneva: International Institute for Labour Studies.

Pyke, F. and W. Sengenberger (eds.), 1992, IndustrialDistricts and Local Economic Regeneration, Geneva:International Institute for Labour Studies.

Rabellotti, R., 1995a, ‘External Economies and Cooperationin Industrial Districts: A comparison of Italy and Mexico’,D.Phil. thesis, Institute of Development Studies, Universityof Sussex.

Rabellotti, R., 1995b, ‘Is There an ‘Industrial District Model’?Footwear Districts in Italy and Mexico Compared’, WorldDevelopment 23 (1), 29–42.

Rasmussen, J., H. Schmitz and M. P. van Dijk (eds.), 1992,‘Flexible Specialisation: A New View on Small Industry?’,IDS Bulletin 23 (3).

Schmitz, H., 1989, ‘Flexible Specialisation – A New Paradigmof Small Scale Industrialisation?’, IDS Discussion Paper(261), Brighton: Institute of Development Studies, Univer-sity of Sussex.

Schmitz, H., 1992, ‘On the Clustering of Small Firms’, IDSBulletin 23 (3), 64–68.

Schmitz, H., 1995a, ‘Small Shoemakers and Fordist Giants:Tale of a Supercluster’, World Development 23 (1), 9–28.

Schmitz, H., 1995b, ‘Collective Efficiency: Growth Path forSmall-scale industry’, The Journal of Development Studies31 (4), 529–566.

Schmitz, H. and B. Musyck, 1994, ‘Industrial Districts inEurope: Policy Lessons for Developing Countries?’, WorldDevelopment 22 (6), 889–910.

Varaldo, R. (ed.), 1988, Il sistema delle imprese calzaturiere,Florence: Giappichelli.


collective effects in italian and mexican footwear industrial clusters

Documents