Globalisasi

INTERNATIONAL INSTITUTE FOR LABOUR STUDIES
DECEIVED
2 5 APR 1994
MfwnationaJ
Discussion Papers
Labour Ofltee
jLO_BlBL
BJT
Labour Institutions and New

Industrial Organisation Programme
Footwear industrial districts

in Mexico and Italy:
A comparative study
by
Roberta Rabellotti
Institute of Development Studies
University of Sussex (United Kingdom)
and Bocconi University, Milan (Italy)
39859
DP/65/1994
The Discussion Paper Series presents the preliminary results of research undertaken by the ELS. The
documents are intended for limited dissemination with a view to eliciting reactions and comments before they
are published in their final form in the Research Series or as special publications.
Copyright International Labour Organization (International Institute for Labour Studies) 1994
Short excerpts from this publication may be reproduced without authorization, on condition that the source
is indicated. For rights of reproduction or translation, application should be made to the Editor, International
Institute for Labour Studies, P.O. Box 6, CH-1211 Geneva 22 (Switzerland).
ISBN 92-9014-549-8
First published 1994
The responsibility for opinions expressed in this paper rests solely with its author, and its publication does
not constitute an endorsement by the International Institute for Labour Studies of the opinions expressed.
Requests for this publication should be sent to: IILS Publications, International Institute for Labour Studies,
P.O. Box 6, CH-1211 Geneva 22 (Switzerland).
A ckno wledgements
The author wish to thank Frances Stewart for her comments on an earlier version, Valeria
Severini for her collaboration on statistical analysis and the Cdmara del Calzado of Guadalajara
and Leon for their organizational assistance during thefieldwork in Mexico. Financial support
from the Instituto Tecnol6gico de Monterrey and Regione Emilia-Romagna is gratefully
acknowledged. Finally, my thanks go to the shoe manufacturers of Mexico and Italy who gave
generously of their time.
Table of contents
Page
Abstract
vii
1. Introduction
n. The Italian and Mexican footwear industry

1. Italy
A. Marche
B. Brenta
2. Mexico
A. Guadalajara
B. Le6n
2
2
4
4
5
6
7
m. The sample
1. The footwear clusters
A. Backward linkages
B. Forward linkages
C. Industrial atmosphere
D. Institutions
7
8
8
10
11
13
IV. The dominant economic structures
13
V. The profiles of sample

firms
1. The statistical analysis
2. Somefindingsof the correspondence analysis
VI. Conclusions
Bibliographical references
17
17
22
23
24
Abstract
This paper presents the results of empirical research carried out in two footwear clusters
located in Italy, the "country" of districts, and two clusters located in an industrializing
country, Mexico.
The study is based on empirical material collected with surveys carried out by the author
in the two countries, testing with a questionnaire the main theoretical hypotheses behind the
industrial district model.
In the first section a synthetic description of the two samples and some methodological
notes are given; in the second a descriptive picture of the analysed clusters is presented and the
following sections show the results of factorial and correspondence analyses in order to obtain
a more synthetic and powerful picture of the clusters. Finally, some considerations on the
comparison are indicated.
I.
Introduction
The research on industrial districts originated in the second half of the 1970s in Italy,
where a widespread phenomenon of industrial development based on small firms, geographically concentrated and sectorally specialized, was first identified by a number of scholars.1
The outstanding increase in the number of employees, die rate of production and export and
the per capita income in north-central and north-eastern Italy (known as the "Third Italy"2
attracted increasing interest from Italian and foreign economists and sociologists3 trying to
understand the reasons for this "economic miracle". During the 1980s, a large amount of
research, both at the empirical and the theoretical level, was carried out in Italy and in other
Western countries, where many agglomerations of small firms, similar to the "original" Italian
districts, were identified.
The main stylized facts of the industrial district meta-model can be summarized as follows:
(a) there is a cluster of mainly small- and medium-sized enterprises spatially concentrated and
sectorally specialized;
(b) there is an intense network of forward and backward linkages among the economic agents
acting within the cluster, based both on market and non-market exchanges of goods,
information and people;
(c) a common cultural and social background links the economic agents and creates a
behavioural code, sometimes explicit but often implicit;
(d) a network of public, semi-public and private local institutions supports the economic
agents within die cluster.
The interplay of these elements is supposed to bring about a competitive edge for firms
belonging to the district in comparison with isolated firms. This is explained by the generation
within the district of the following economic effects [Camagni, 1989]:
(a) external and "district" economies, both in a static (e.g. reduction of costs) and a dynamic
perspective (e.g. training process, accumulation of know-how and knowledge taking place
in a spontaneous and socialized way);
(b) proximity economies, reducing transaction costs and costs of using die market, through
easier circulation of information and personal contacts;
(c) synergy elements, enhancing local innovation capability through imitation processes,
interaction among local agents, private and public institutions, customer-supplier
cooperation.
1
The first studies in which the phenomenon was identified include Bagnasco [1977], Becattini [1979], Brusco
[1982], Goglio [1982] and Fua [1983].
2
The term "Third Italy" (Terza Italia) was first coined by the Italian sociologist Arnaldo Bagnasco [1977] and
distinguishes the area of north-central and north-east Italy from the underdeveloped south of the country and the
traditionally industrialized north-west region. It consists of the regions of Friuli-Venezia Giulia, Veneto, TrentinoAlto Adige, Emilia-Romagna, Tuscany, Marche and Umbria.
3
The book which first brought the success of the districts to the attention of English-speaking readers was The
second industrial divide by Piore & Sabel [1984]. Later, a number of articles, some of which originally appeared
in Italian, were translated into English and published in Goodman & Bamford [1989], Pyke et al. [1990] and Pyke
& Sengenberger [1992].
DISCUSSION PAPER SERIES NO.
65
The existence and intensity of the main elements characterizing the industrial district metamodel, as specified above, and its related economic effects, have been tested in the empirical
investigations in Mexico and Italy.
In the following sections, the main results of the research are presented, focusing on the
comparison. Section II presents some background information on the Italian and Mexican
footwear industry and on the four clusters selected. Section III gives a synthetic description of
the two samples and some methodological notes. Section IV presents a descriptive picture of
some characteristics of the clusters analysed and Section V shows the results of factorial and
correspondence analyses, to obtain a more synthetic and powerful representation of the clusters.
Finally, some considerations on the comparison are indicated in Section VI.
//. The Italian and Mexican footwear industry

In this section, some background information on the Italian and Mexican footwear industry
is presented and a few facts characterizing the four districts is studied. This first descriptive
picture will help in analysing and understanding the results of the statistical analyses.
1.
Italy
The outstanding growth of the Italian footwear industry began in the 1960s and went on
continuously until 1985, led by exports. At the beginning of the 1950s, exports represented a
mere 3.7 per cent of total production; in 1970 the proportion of exports increased to 63 per
cent; in 1985 it was 83 per cent and, in 1992, 81 per cent [ANCI, 1992].
The competitive factors explaining the success of the Italian footwear industry have
followed a rather typical evolutionary pattern. In the first phase of development, during the
1960s, Italy exploited a labour-cost advantage with respect to other European competitors.
Subsequently, the high specialization of the Italian footwear system, based on the division of
the production cycle among several enterprises specialized in the different phases of production,
and on the existence of a very well-developed network of backward-linked firms producing
components and raw materials for the sector, became the main source of comparative
advantage. More recently, during the 1980s, other European countries like Spain, Portugal,
some south-east Asian newly-industrializing countries (NICs) like Taiwan and South Korea and
also some developing countries like Brazil, India and China became very competitive in the
international market and greatly increased their export share. In order to face this increasing
competition, the Italian footwear system strengthened its advantage in terms of image, fashion
content and design to reduce the price elasticity of demand for its products.
Notwithstanding its attempt to counter this increasing competition with a process of tradeup by upgrading its exports, the Italian footwear sector had to start competing in a market
where a large increase in the quantity of shoes supplied faced stable demand. From the Second
World War onwards, Italian exports gradually increased, reaching their highest level in 1985.
However, from 1986 onwards, exports decreased continuously, except in 1990 when a small
increase was registered (Table 1). In the international market, Italy shifted from second place
in 1986 to fifth place in 1990, among the world's leading exporters, after China, Hong Kong,
Taiwan and South Korea.4
According to SARTRA [1992] in value terms China and Hong Kong would fall behind Italy among the leading
exporters because of their low unit values.
FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY
The difficulties in the international market were compounded by some difficulties in the
domestic market due to a large increase in imports, mainly from Asian countries (in 1992, 46
per cent of the value of total imports) and other EEC countries (35 per cent). In fact, from
1980 onwards, Italian shoe imports have continuously increased from 36.8 million pairs in
1980 to 127 million pairs in 1992 (Table 1) [ANCI, 1993].
The Italian footwear industry is therefore going through a very difficult time, after 30
years of continuous growth. In both the domestic and the international market, it has to face
increasing competition from countries with a clear advantage in terms of labour costs.
Table 1 : The Italian footwear industry, 1970-1992
Year
Employees
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
132 608
141 360
135 923
130 050
135 791
130 791
130 806
120 806
127 967
142 819
132 475
139 636
142 288
135 128
134 317
133 914
128 825
122 513
115 886
114 123
113 980
111701
108 350
Source:
Pairs
(mill.)
345 898
372 878
380 834
356 967
383 601
366 306
410 581
398 129
429 344
515 388
415 743
468 692
531 300
487 718
496 198
524 509
499 285
464 581
436 162
406 935
424 916
433 424
418 827
Value
(billions
lire)
793.0
908.7
990.4
1 076.8
1 371.0
1 683.5
2 370.2
2 890.3
3 385.2
4 577.8
4 899.7
5 554.7
7 009.8
7 420.5
8 554.2
9 590.7
9 783.7
9 648.9
9 144.1
10 345.6
10 946.9
11 463.5
11 531.4
Index
Exports
pairs(ml)
value
1970=100
100.0
114.6
124.9
135.8
172.9
212.3
298.9
364.5
426.9
577.3
617.9
700.5
884.0
935.8
1 078.7
1 209.4
1 233.7
1 216.8
1 153.1
1 304.6
1 380.4
1 445.6
1 454.1
217 666
235 848
246 704
226 412
251 483
232 424
264 701
264 986
294 803
374 351
314 609
338 581
387 325
374 237
393 134
434 753
411 022
383 872
378 191
339 858
360 022
347 820
338 657
Imports Imports/
pairs(ml) exports
(%)
3 370
3 081
5 650
7 472
7 398
6 977
9 512
13 543
20 044
27 384
36 761
42 032
38 389
54 318
53 359
56 073
63 917
89 048
100 675
92 207
73 668
104 060
127 524
1.5
1.3
2.2
3.3
2.9
3.0
3.6
5.1
6.8
7.3
11.7
12.4
9.9
14.5
13.6
12.9
15.5
23.2
26.6
27.1
20.4
29.9
37.6
ANCI [1993].
The Italian footwear industry is characterized by a very strong spatial concentration: the
two regions, Marche and Veneto,5 in which the two clusters analysed are located, represent
the two main areas of production with, respectively, 23.6 per cent and 23.4 per cent of the total
production [ANCI, 1993]. Other regions of specialization are Tuscany (19 per cent), Lombardy
(10 per cent), Apulia (7.5 per cent), Emilia-Romagna (6.4 per cent) and Campania (4.9 per
cent).
The sector is characterized by a very low level of concentration: only 0.5 per cent of
companies have more than 200 employees and those with a workforce of less than 50 account
for 62 per cent of all employment. There is a considerable number of firms employing fewer
than 10 people, representing 63 per cent of all companies in the sector and 16 per cent of all
The district analysed in Marche is the area around Civitanova and Fermo. Brenta is located in Veneto.
DISCUSSION PAPER SERIES NO. 65
employees. In 1991, the average size of Italian footwear firms was 9.3 employees [Gaibisso,
1992].
The survey areas selected represent two of the most important footwear clusters in Italy:
Marche contains the largest concentration of firms producing leather street shoes and Brenta
is an area highly specialized in the production of high and medium-high quality women's shoes,
strongly export-oriented (Table 2).
Table 2: The two footwear clusters analysed
No. of firms
Brenta
Marche
No. of employees
Pairs of shoes
(millions)
1986
1991
1986
1991
1986
1991
775
832
10 008
9 419
8 420
8 100
2 629
2 410
29 704
25 437
124 821
108 536
Source: ANCI [1992].
Marche
The shoe industry was established in Marche in the 1920s, when some inhabitants from
Montegranaro, one of the main centres of the district, emigrated to the North of Italy to work
in footwear firms and eventually came back to their villages to set up the first local enterprises.
At first, the local industry worked mainly as subcontractor for the old-established enterprises
located in the North. Then, the sector started to grow thanks to a rapidly-increasing demand,
with a first boom after the Second World War when some firms in the North closed down, and
a second real boom during the 1970s when the footwear industry in Marche became the largest
concentration of firms in Italy. In the period 1971-1981, the regional rate of growth of
employment in the footwear industry was 90.7 per cent, while the average national rate in the
sector was 37.5 per cent. From 1951 to 1991, the share of workforce in Marche over the
national workforce in the sector rose from 4.3 per cent to 23.3 per cent [ANCI, 1992].
From Montegranaro, the footwear industry progressively spread to the neighbouring rural
areas and today the strongest concentration of firms is around Fermo, Montegranaro, Monte
San Giusto, Monte Urano, Porto Sant'Elpidio, Sant'Elpidio a Mare and Civitanova Marche,
an area located between two provinces Macerata and Ascoli Piceno. All these are small
towns whose economic system is completely specialized in the footwear sector and some related
industries.
In the second half of the 1980s, footwear enterprises in Marche, as in the rest of the
country, began to suffer from increasing competition in the international market and this caused
a decrease in the number of firms from 2,629 in 1986 to 2,410 in 1991; in the same period,
the number of employees decreased from 29,704 to 25,437 and shoe production decreased from
about 125 to 108 million pairs (Table 2).
B. Brenta
In Brenta, the origin of the footwear industry goes back to the establishment of
Calzaturificio Voltan in 1898, founded by Giovanni Voltan, an artisan who, after spending
some time in the United States to learn the profession, came back to his little village (Stra) with
machines for making shoes to set up his own business. In 1904, the firm employed 400 people,
producing 1,000 pairs of shoes per day. It was the first shoe factory in Italy. At the beginning
of this century, many workers from Voltan's enterprise began to set up their own businesses
in Stra and in other nearby villages, such as Vigonovo, Fosso, Fiesso d'Artico, Noventa,
Saonara and Vigonza. This was the beginning of the shoe district called Brenta, from the name
of the river which flows through it. From an administrative point of view, the district is divided
up into two provinces: Venice and Padua.
During the boom in the footwear industry after the Second World War, the sector
progressively absorbed most of the rural workforce available in the area. In the 1960s, the local
enterprises expanded, further increasing their penetration in the export market, specializing in
the high and medium-high segments of the market.
Nowadays, 85 per cent of the shoes produced in the area are women's shoes in the
medium-high and high segments of the market and 74 per cent of total production is exported.
This area has recently suffered from increasing competition in the international market and
from 1986 to 1991 its production decreased from 8.4 to 8.1 million pairs. Employment in the
sector also decreased from about 10,000 people to 9,400, while the increase in the number of
firms from 775 to 832 may be ascribed entirely to an increase in producers of components,
because in the same period the shoe firms themselves decreased from 306 to 274 (Table 2).
2.
Mexico
Mexico was a protected market for a long time, having adopted an import-substitution
strategy, which was drastically abandoned in 1988 with the opening up of the economy to
foreign competition. Therefore, the Mexican footwear industry produced mainly for the
domestic market for many decades, with a rate of growth strictly linked to the increase in
domestic demand (Table 3.)
Table 3: The Mexican footwear industry, 1970-1991
Years
1970
1975
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
Manufacturing
industry GOP
(th. of ml.
pesos)'
539.1
718.9
847.9
934.5
988.9
1 052.7
1 023.8
943.5
990.9
1 051.1
995.8
1 026.1
1 059.0
1 135.1
1 201.2
1 245.3
Leather &
footwear GDP
(th. of ml.
pesos)1
16.1
21.8
25.5
28.9
29.6
32.6
32.6
27.7
29.3
30.2
28.6
24.6
24.0
24.9
25.4
24.5
Note: At 1980 prices.

Source: Institute National de Estadistica, Geografia e Lnformatica.
Number of
employees
(leather &
footwear)
97 555
118 292
121 670
138 155
138 909
152 303
155 706
130 521
131 478
135 390
135 666
128 875
119 578
117 529
114 323
n.d.
Exports
USSmill.
Imports
US$mill.
3.4
12.1
25.2
30.8
31.0
24.5
14.9
12.1
19.3
13.7
18.4
53.5
67.1
67.4
80.6
104.0
15.5
20.8
19.0
17.1
62.0
67.3
13.1
4.6
11.9
15.7
9.4
13.7
54.3
75.0
86.0
148.2
65
The acceleration in the opening up of the Mexican market to international competition,

through the elimination of import licensing and tariff reduction6 had a very heavy impact on
the footwear industry. The market was flooded with imports which increased from 2.2 million
pairs in 1987 to 38.2 in 1991 [SECOFI, 1992].
Whilst on the one hand the Mexican footwear industry is threatened by increasing
international competition, on the other hand it has a unique opportunity to massively enter the
United States market.7 In fact, Mexico has just signed a Free Trade Agreement with the USA
and Canada and if the Mexican shoe producers are able to take advantage of the progressive
elimination of duties and of their proximity to the US market, Mexico could possibly become
one of the world's leading exporters of footwear in a few years.
The challenge faced by the Mexican footwear industry today is therefore extreme: the
alternatives are a progressive reduction in the number of firms, because of the increasing
competition in the domestic market of imported shoes from countries like Taiwan, China and
Korea, or the acquisition of a leading position as an exporter in the international market.
The footwear industry in Mexico is highly concentrated in three different areas: 37 per
cent of the officially registered 4,500 firms is located in Le6n, 27 per cent in Mexico City and
its outskirts (Distrito Federal), 26 per cent in Guadalajara, while the remaining 10 per cent is
spread out in the rest of the country [CICEG, 1990].
A. Guadalajara
Guadalajara is a town of about four million inhabitants, capital of the state of Jalisco,
north-west of Mexico City. Since the last century, mere was a tradition in Guadalajara of shoe
production in small workshops to satisfy local demand. In the second decade of this century,
thanks to the continuous increase of the local market and to the improvement of
communications with the rest of the country, the sector started to grow and was transformed
from an artisanal to an industrial activity. In 1927, there were 34 plants producing shoes, one
producing lasts and one tannery; about 100 small shoe workshops were also located in the
town. In that period, the footwear industry was the most important industrial sector in the
town.
From the 1930s, the footwear industry began to develop rapidly thanks to the presence of
a United States enterprise, United Shoe Machinery, which contributed in diffusing know-how
among local enterprises. In 1942, me Regional Chamber of Footwear Industry (Cdmara
Regional de la Industria del Calzado) was established and in 19S9 Guadalajara was chosen as
the location for the National Footwear Trade Fair [Arias, 1980; 1985; 1992].
At the end of the 1950s, the footwear industry was one of the most important sectors in
the economic system of Guadalajara and, with 757 officially registered firms, was the sector
with the largest number of employees in the town. In the following years, the manufacturing
structure of Guadalajara increased its diversification and in 1988 the footwear industry, together
with the textile and clothing industry, was the second most important sector for its contribution
(5.5 per cent) to Jalisco's GDP after the food industry (11.1 per cent).
With regard to size distribution, 15 per cent of the footwear enterprises have fewer than
five employees, 34 per cent have fewer than 25, 48 per cent fewer than 100 and only 3 per
cent are large firms with more than 100 employees.
For the footwear industry, the tariff was almost halved from 35 per cent to 17 per cent of import value and
all import licenses were eliminated [CANAICAL, 1991].
7
In 1989, the USA consumed 1,393 million pairs of shoes, of which 82 per cent were imported [ILO, 1992].
B.
Le6n
Ledn is a town of about one million inhabitants, located in the state of Guanajuato (north
of Mexico City). Its industrial development has been favoured by its central location along one
of the main axes connecting Mexico City to the North and therefore the United States. In 1900,
the footwear industry was already the second most important industrial sector after textiles in
terms of number of employees. In 1940, it became the most important and some related sectors
started to develop. During the Second World War, the footwear industry in Le6n expanded
significantly thanks to a large increase in exports to the USA [de la Cruz Labarthe, 1985].
Since then, the footwear industry has dominated the industrial system of Le6n, where 55 per
cent of employees in the industrial sector in the state of Guanajuato are concentrated. In 1985,
according to the last industrial census, 45 per cent of employment in the manufacturing sector
in the whole state was in the footwear, textile and clothing industry.
Finally, concerning size distribution, 45 per cent of die enterprises have fewer than five
employees, 43 per cent fewer than 25, 8 per cent fewer than 100 and only 3 per cent more man
100 people.
///. The sample

The field work was carried out in the four clusters selected, in a sample of firms which
had been randomly chosen among the members of the local associations of footwear
entrepreneurs.8
The sample included 101 enterprises: 50 of them located in Italy (30 in Marche and 20 in
Brenta) and 51 in Mexico (30 in Guadalajara and 21 in Le6n). In order to take into account
firms of different sizes, the sample was stratified by size: in Italy, 24 enterprises employing
fewer than 50 employees, 15 between 51 and 100 employees and 11 more than 100 employees,
while in Mexico 17 enterprises employing fewer than 50 people, 14 between 51 and 100
employees and 20 more than 100 employees were selected (Table 4).
Table 4 : The sample
No. of firms
< 50 employees
51-100 employees
> 100 employees
Total firms
Italy
24
15
11
50
Mexico
17
14
20
51
As already mentioned, the first condition to be satisfied for the existence of an industrial
district is an agglomeration of spatially concentrated and sectorally specialized enterprises. In
all the districts analysed a significant critical mass was found:
(a) in Le6n there are approximately 1,700 shoe enterprises and in Guadalajara approximately
1,200, of a total of 4,500 shoe firms in the whole of Mexico [CANAICAL, 1991];
(b) in Marche there are 2,410 firms and in Brenta 680 of a total 8,547 enterprises in the
whole of Italy [ANCI, 1992].
8
The choice to include only members of the Associations was justified by the focus of the inquiry on the
different forms of explicit and implicit cooperation among firms. It was assumed that firms belonging to the
Associations were more likely than others to entertain relationships with other firms; to become a member of the
Association can in fact be interpreted as a sign of interest in getting in touch with other firms.
65
Moreover, in all the clusters analysed, there are backward- and forward-linked firms and
institutions devoted to the assistance of the footwear sector.
Given the existence of a significant critical mass, on the basis of a closed questionnaire
which was used to interview die sample firms, the existence of other conditions for the
presence of a district were identified, i.e. die existence of backward and forward linkages and
of a common cultural, social and institutional background.
The results of die questionnaire were analysed with three statistical instruments:
(a) frequency tables to identify the existence of relationships among variables;
(b) factor analysis to summarize the main structures characterizing the universe analysed;
(c) correspondence analysis to identify the profiles of sample firms according to their
performance.
The information collected wim the questionnaire within the sample of firms was
complemented by several open interviews wim entrepreneurial associations, sector experts,
suppliers, buyers and representatives of institutions supporting die sector. In addition, die
results of me interviews were complemented and cross-referenced widi the available official
statistics and other studies carried out on the Italian and Mexican footwear industry.
1.
The footwear clusters
In mis section, die existence and intensity of die diree aspects specified above as the
stylized facts of die industrial district model in die four selected footwear clusters are analysed.
A.
Backward linkages
Widi regard to backward linkages, in Italy die existence of a very well developed system
of suppliers, located within the footwear clusters in a situation characterized by strong
competition due to die large numbers of firms specialized in the same typologies of products,
but at die same time by cooperation between footwear firms and their suppliers, represents one
of me main assets of the industry.
Strong competition pushed suppliers to produce a very wide variety of products with very
short delivery times, allowing die shoe producers to postpone to die very last moment their
purchases of inputs. This has several advantages: first of all, it reduces die stocks required for
producing shoes; secondly, it leads to die progressive shortening of die period between order
and delivery which characterizes die sector and, finally, it increases die capacity of shoe
producers to diversify their products and to satisfy market demand.
This very advanced system of production includes tanneries, producers of components and
accessories, suppliers of machines and service firms. From the results of the interviews, as well
as from odier available studies [Varaldo, 1988; Gaibisso, 1992], it is clearly demonstrated that
die cooperation between suppliers and shoe producers is a crucial asset for the competitiveness
of the sector, because fashion trends are decided by suppliers and shoe producers together. The
capability of die footwear sector to satisfy very volatile demand is in fact due to die capacity
of die backward-linked industries to follow its production rhythms.
In Mexico, the relationships between suppliers and shoe producers are much less
collaborative, being based mainly on a pure market mechanism. Among the sample firms,
many have complained about the low quality of components and raw materials, insufficient
attention to fashion changes and die bad service provided by input suppliers. In turn, suppliers
do not accept complete responsibility for their low level of development and accuse shoe
entrepreneurs of having always adopted a strategy more focused on price than quality. Suppliers
complain about very unstable demand, very small batches of orders of continuously changing
products, and about delays in payment. We seem to be faced with a two-sided problem: both
the suppliers of components and the manufacturers blame each other and the main deficiencies
are of communication and collaboration between the two linked sectors.
Both footwear enterprises and suppliers are now beginning to realize the importance of a
systemic view of the production process: only a strong collaboration between them will in fact
permit the production of an internationally competitive product. In Leon the associations of
suppliers and footwear firms are starting to work together on fashion content and
standardization of measurement systems. There are also cases of continuous and productive
collaboration among some shoe firms and some tanneries: a stable demand permits a stable
level of quality of the leather supplied. Some wholesalers are also working in that direction,
organizing the purchase of components for all their clients and therefore guaranteeing large and
stable orders to the suppliers. In both Leon and Guadalajara, the local credit unions of shoe
entrepreneurs have also launched a programme for common purchasing.
These are a few examples of initiatives trying to foster the systemic approach, showing
a real effort towards building up a system of production in some way inspired by the
"industrial district" model. In Mexico, a relatively high intensity of linkages between suppliers
and manufacturers within the districts has been found; however, in many cases, these linkages
need a requalification because they are based only on the price factor, disregarding important
aspects like fashion, quality of materials, delivery time, design.
From the survey, it emerges that in both Mexico and Italy the vast majority of linkages
with suppliers takes place within the same district. However, a substantial difference in terms
of the quality of those linkages has been clearly identified. In the Mexican footwear industry,
the relationships between suppliers and manufacturers are mainly market linkages, based simply
on price, while in Italy a network of relationships much closer to the ideal typology of
cooperative linkages was found, based on a common effort of suppliers and shoe producers to
work together to produce new products [Rabellotti, 1993b].
Another aspect strongly differentiating the Italian and the Mexican footwear industry is the
degree of division of labour. According to a recent survey [Varaldo, 1988], in Italy more than
80 per cent of footwear firms decentralizes the production of bottoms, more man 70 per cent
decentralizes the phases of edging and sewing and more than 50 per cent externalizes the
cutting phase. In Mexico, according to several recent surveys [Boston Consulting Group, 1988;
Concalzado, 1991; Dominguez-Villalobos & Grossman, 1992] and to the results of this survey,
the degree of division of labour is very low in all the different phases of the production cycle
[Rabellotti, 1993a] (Table 5).
Table 5: Division of labour in the Italian and Mexican footwear industry (% of sample firms)
% of pair
externalized
0%
Italy
Phases of production:
Cutting
Sewing
Bottoms
Finishing
26
6
10
96
<50%
Mexico
94
80
59
100
Italy
20
16
10
2
Mexico
6
20
4
-
51%-90%
>90%
Italy
Mexico
Italy
20
20
2
2
4
-
34
58
78
-
Mexico
33
-
Source: Author's survey.
To facilitate the relationships with enterprises in other phases of production, in Italy the
shoe firms tend to decentralize mainly within their area and to maintain stable and continuous
10
65
linkages with them. According to most of the entrepreneurs interviewed, decentralization is one
of the main strategic assets of the Italian footwear industry and depends on three main factors:
(a) technological factors, because the different phases of the production cycle are
characterized by different economies of scale;
(b) structural factors which can be represented as a trade-off between production and
transaction costs: decentralization allows limits on the size of the firm and reduces labour
costs but, of course, increases the costs of using the market;
(c) strategic factors such as the capability to control quality and to avoid the circulation of
confidential information on the market.
In contrast, in the Mexican footwear industry the very low degree of division of labour
can be explained by two main reasons:
(a) backward-linked industries have reached a very low level of development with regard to
quality, design, fashion of components and service because die past closure of the
domestic market has limited the competition and therefore the incentive to innovate. We
know from Adam Smith that "the division of labour is limited by the extent of the market"
and this seems to apply very well to the Mexican cases;
(b) in the sector it lacks a standard technical language and a common, universally accepted,
measurement system and this, of course, strongly increases transaction costs and costs of
using the market.
Therefore, the poor level of development and the difficulties of communication with backwardlinked industries have induced many shoe firms to internalize as many phases of the production
cycle as possible in order to reduce their dependency on an unstable, low-quality supply.
A very interesting result is the positive relation between decentralization and profits among
the sample firms in Mexico: 47 per cent of the decentralizing firms have obtained a regular
profit; 21 per cent a very good profit, while only 10 per cent no profit at all and 13 per cent
a loss. For firms which do not decentralize, the percentages are, respectively, 33 per cent, 16
per cent, 16 per cent and 33 per cent.
Decentralization therefore seems to have a positive effect on performance, flexibility and
specialization and from that it was concluded that firms should be encouraged to decentralize
more. However, a system of production based on a strong division of labour such as, for
instance, the Italian one, necessarily requires a very well-developed network of suppliers, and
this is definitely not present in Mexico.
B.
Forward linkages
In both Italy and Mexico, relationships with the market are very weak, hardly under the
control of shoe producers, for different reasons. In Mexico, it is because the domestic market
has been closed to international competition for a long time and in Italy because the domestic
footwear industry led the international market until the end of the 1980s, the shoe
manufacturers faced a market characterized by demand exceeding supply and therefore, until
recently, they did not adopt an active commercial strategy. Many of the entrepreneurs
interviewed in Italy emphasized that until a few years ago foreign buyers, mainly from
Germany, used to regularly visit their firms and all they needed to do was to produce shoes:
it was not necessary to make any effort to sell them.
The same is true in Mexico, where the long absence of international competitors allowed
the domestic manufacturers to produce shoes which were easily sold in the closed market,
11
regardless of their quality, design and fashion content. Therefore, in Mexico the footwear
industry also neglected commercialization and marketing for a long time.
In Italy, it has been emphasized [Gaibisso, 1992] that in the footwear districts the highly
efficient way of organizing the production base is not complemented by an equally efficient
system of commercialization. In the footwear areas the role of middle-men, who have been so
important in other districts in different sectors among which we can quote the famous case of
Prato and its impannatori, has been very limited.
In 76 per cent of the sample firms, commercialization takes place through non-exclusive
agents, in 22 per cent through direct sales to customers, in 16 per cent through exclusive agents
and in 6 per cent through trading companies or in shops owned by the same firm.
The relationship with non-exclusive agents often generates discontent among footwear
firms because they do not control the activity of their agents and receive very little information
about the market from them. Many of the enterprises interviewed realized they would need a
much more active commercial strategy to be successful in an increasingly competitive market.
However, in marketing, economies of scale are important and very few of the Italian firms
. have the financial capability to invest in advertising or market research, to set up their own
shops or to employ exclusive agents. Some experiences of export consortia were developed in
the two districts analysed; however, they seem to play a minor role because of the difficulties
firms encountered in cooperating in a crucial area like commercialization. A clear proof of
these difficulties is the large number of projects, which, after an enthusiastic start, fell through
because of personal interests of the firms involved.
In Mexico, 73 per cent of the sample firms sell part of the production through nonexclusive agents, 51 per cent through wholesalers and 16 per cent through trading companies
or shops owned by the firms.
Generally speaking, the Mexican footwear firms suffer from very similar problems to the
Italian enterprises: they have very limited control over their markets, very little knowledge of
them, depend on non-exclusive agents and do not adopt any commercial strategy for selling
their products. Recently, an interesting experience was started thanks to a few wholesalers who
have adopted a new strategy aimed at selling a quality product and have therefore selected a
group of shoe firms with whom to develop a stable and constructive relationship. Among the
firms interviewed, the ones linked through such a relationships with wholesalers were generally
very satisfied, not only with the sales realized but especially with the complementary services
they receive, such as technological and organizational assistance. Many of them have been able
to introduce important improvements in the organization of the production process and in the
quality of their products thanks to collaboration with these wholesalers.
C. Industrial atmosphere
The concept of "industrial atmosphere" can be very vague. It will be used to capture
explicit and implicit forms of collaboration and interaction among local economic agents within
the districts, enhancing local production and sometimes innovation capability.
The first aspect to be emphasized is related to the characteristics of the local labour market
in the areas analysed. The existence of a large reservoir of skilled labour has always been
assumed to be one of the externalities to the organization of production in highly specialized
clusters of firms. The Italian industrial districts have been described in the literature as areas
where specialized jobs are taught from father to son and where skills are accumulated and
transmitted from one generation to the other. This process of collective learning, of
accumulation of personal knowledge and therefore of circulation of know-how among firms
through labour force mobility has traditionally enhanced local innovation capability. In other
12
65
words, the innovation process, usually taking place inside the firm, in the industrial district,
becomes a collective "affair", based on common knowledge, which is accumulated more in
people than in firms.
Surprisingly enough, in the Italian footwear districts, there was a rather low availability
of skilled labour, due to the increasing tendency of young people to seek alternative jobs in
other more attractive sectors, even outside their area of origin. Moreover, a larger share of
young people than before stays at school after primary education and subsequently they seek
more qualified, non-manual jobs.
The result of the diffused increase in welfare, due to the expansion of the footwear sector
in the areas, is therefore an increasing tendency for local people to abandon the sector if they
can find alternative jobs. The resulting interruption of the accumulation and transmission of
skills from father to son, from one generation to the other, could negatively affect the collective
learning effect, which has been so important in determining the competitiveness of the districts,
in the future.
With regard to Mexico, skilled labour is also a rather scarce resource (according to 73 per
cent of the firms interviewed), but this result is less surprising in the Mexican context than in
Italy.
Innovation capability within the districts is also enhanced by technological cooperation
among firms and with technology suppliers. In the production of leather shoes, technological
advance has been very limited and therefore cannot be used in the sector as a barrier against
competition from less developed countries.
In Italy there is intense cooperation between manufacturers and producers of machines
which has definitely contributed to the development of a mechanical sector in which Italy is the
leading world exporter [ASSOMAC, 1992]. However, the difficulties in the footwear sector
have recently had a negative impact on the tendency to collaborate among technology suppliers
and shoe manufacturers and this may negatively influence the process of technological
development.
In Mexico, the domestic production of machinery is very small and therefore machines are
bought from local retailers or sometimes directly abroad, mainly from Italy and Brazil.
Among Mexican firms, the main sources of information are: participation in technology
fairs (57 per cent of the interviewed firms), other innovative firms (27 per cent) and the
technology suppliers (20 per cent).
Within the districts analysed there is also an intense network of informal relationships
amongfirmswho exchange information about technology or the market, machines and workers,
subcontract orders in case of excess demand, jointly sell products, jointly purchase inputs or
jointly recover their credits.
Rather surprisingly, it was found that informal cooperation is more frequent in Mexico
than in Italy: in 80 per cent of the interviewed Mexican firms, informal contacts are very
frequent, while in Italy this happens only in 40 per cent of the sample firms: 24 per cent of
them sometimes have informal relationships and the remaining 28 per cent do not have any
contacts at all.
The "industrial atmosphere" and in general the existence of an intense network of
relationships among the economic agents is supposed to be favoured by the dominant social
climate within the districts. In both Mexico and Italy, a diffused sense of belonging to an
old-established community was evident among most of the entrepreneurs interviewed.
Moreover, most of the entrepreneurs agreed that this favourable social climate is very important
because exchanges and relationships with other firms are facilitated.
D.
Institutions
From the investigations in Italy, it was clear that in both Marche and Brenta, the main
institutional actors are the local entrepreneurial associations, supplying a number of services
to the members and playing an important role of promoting initiatives to support the sector.
There are also centres for technological assistance and training, technical schools, export
consortia and credit unions.
In Mexico, entrepreneurial associations are also active both in Guadalajara and Le6n,
providing services like fiscal and labour consultancy, commercial assistance, managerial
training, organization of trade fairs. Moreover, in both the districts, centres for technological
assistance and training and credit unions exist.
In both Italy and Mexico, it is therefore important to emphasize the existence of an
institutional framework based on local "from below" institutions, directly involved with the
economic agents and having, as we will see in Section V, an important influence on the
performance of the firms located in the districts.
IV. The dominant economic structures

In this section, the results of factor analysis, a statistical technique used to identify a
relatively small number of factors which best represent relationships among sets of many interrelated variables, are presented [Kim & Mueller, 1978].
A huge number of variables can be used to describe the economic systems analysed in
Mexico and Italy: size of firms, degree of innovativeness, segment of market, frequency of
relationships with other firms, etc. However, the description might be greatly simplified if it
were possible to identify a few underlying dimensions or factors.
The fundamental assumption behind factor analysis is that the underlying factors, which
are smaller in number than the observed variables, are responsible for the co-variation among
the variables. Factor analysis therefore has two goals: first of all to represent relationships
among sets of variables parsimoniously and, secondly, to obtain interpretable factors. With
these objectives, factor analysis was adopted, selecting a few variables from the questionnaires
in Mexico and Italy.
In Mexico, a set of 20 variables was selected and, with factor analysis, three principal
factors were extracted, explaining 76 per cent of the total variance of the observed variables.9
In Table 6 the matrix of factor loadings which are the coefficients used to express a
standardized variable in terms of the factors is presented: factors with large coefficients (in
absolute value) for a variable are closely related to that variable. To interpret the factors, the
variables appearing in each factor with the highest factor loadings are taken into account.
The first factor, explaining 39 per cent of the total sample variance, is composed of the
following variables, listed in decreasing order of importance according to their factor loading:
Bl
H57
H7A
H55
H56
number of employees
investments in managerial training
quality of product as one of the two main competitive advantages of the firm
investments in commercialization
investments in technical training
9
The PROC FACTOR procedure in the SAS system (6.03) has been used and the prior communality estimates
have been computed from the maximum value of partial correlation.
13
14
C3512
LI
H78(-)
Fl
good or very good technological level

location in Leon
institutional assistance as one of the two main competitive advantages of the firm10
equity shares in other shoe firms or in other firms connected with the sector, located
within the cluster.
The first factor is therefore dominated by the size of firms (Bl). With size, other
variables, representing the strategies of internal (H57, H55, H56) and external investments
(Fl), the technological level (C3512) and the quality of product (H7A), are correlated. In
addition, from a spatial point of view the factor is characterized by the location in Leon and
by the limited importance attributed to institutional assistance.
Table 6: Factor loading matrix: Mexico
Variables
Bl
H57
H7A
H55
H56
C35
LI
H78
D78
Factor 1
Factor 2
Factor 3
Fl
G131
H76
B1212
D16
NH8B
H73
Gl
C222
NH84
0.64
0.63
0.57
0.47
0.46
0.46
-0.46
-0.51
0.18
0.33
0.38
0.01
-0.22
-0.008
-0.007
0.19
0.21
0.32
0.16
0.12
-0.33
0.11
0.06
0.27
0.31
-0.05
0.01
0.29
0.54
0.39
-0.57
0.15
0.24
0.06
0.31
-0.15
0.13
0.24
-0.29
0.24
0.09
0.13
-0.19
0.14
0.02
0.46
0.43
0.19
-0.04
0.06
0.21
0.49
0.37
-0.20
-0.30
-0.08
-0.02
0.03
-0.10
-0.18
Tot. Explain.Variance
38.75
21.57
15.50
Dim
In economic terms, the factor can be interpreted as a confirmation of the importance of

scale economies: a number of variables related to investment strategies, process and product
characteristics, are, in fact, correlated wim size. Moreover, size is negatively correlated with
institutional assistance and this can be interpreted as indifference of large firms to district
effects and external economies generated within the cluster.
The second factor explains 22 per cent of the total sample variance and is composed of the
following variables:
D7812 original design
DIII1
high quality products
The negative sign means that the correlation with the other variables is negative.
Fl(-)
Bl(-)
15
equity shares in other shoe firms or in other firms connected with the sector, located
within the cluster (negative correlation)
number of employees (negative correlation).
This second factor is characterized by variables related to quality of products (Dim) and
design originality (D7812), negatively correlated with size (Bl) and external investments (Fl).
The emphasis is therefore on quality, which is independent of scale economies. This could
be seen in contrast with the first factor, where the variable H7A appears; however this last
variable represents the opinion of entrepreneurs about their products, while DIII1 and D7812
indicate, in a more unbiased way, that the product belongs to a high segment of market.
Finally, the third factor, explaining 15 per cent of the total sample variance, is composed
of the following variables:
G131
H76
LI
frequent use of more than one service supplied by the Entrepreneurial Association
information access as one of the two main competitive advantages of the firm
location in Guadalajara.
In this last factor, two variables related to the spatial environment are identified (G131 and
H76) together with the location in Guadalajara (LI). The factor can therefore be interpreted as
the collective dimension of the analysed system, emphasizing the role of supporting institutions
and information access.
Going on to analyse the results of factor analysis on the Italian sample, three factors can
be identified, explaining 87 per cent of the total sample variance (Table 7). The first factor (33
per cent of the total sample variance) is composed of the following variables:
B612
increasing trend of employment in the last five years
Mil
investments in technology
G142
forms of commercialization other than agents
B1012(-) high availability of skilled labour force (negative correlation)
LOC(-) location in Brenta
G31
high or medium-high segment of market
D3(-)
percentage of the sewing phase put out to other firms (negative correlation)
H21
equity shares in other firms.
The first factor characterizes expanding firms: growing employment, high investment in
technology, a mixed strategy of commercialization, a high or medium-high quality product and
therefore difficulty in finding skilled labour and a limited tendency to decentralize the
production process and, finally, the tendency to own equity shares in other firms.
The second factor (30 per cent of sample variance) is composed of the following variables:
Bl
B2S
F112
M18
G31(-)
L21(-)
number of employees
percentage exported
good or very good technological level
investments in other sectors
high or medium-high segment of market (negative correlation)
use of more than one service supplied by the entrepreneurial association (negative
correlation).
This factor can be interpreted as a size-quality factor: firms of large size, exporting a
product of low quality and producing with advanced technological processes. They tend to
diversify their investments towards other sectors and do not utilize the services supplied by the
16
entrepreneurial association. This category of firms seems to be independent of the district

effects.
Table 7: Factor loading matrix: Italy
Variables
Factor 1
Factor 2
Factor 3
B612
Mil
G142
B1012
LOC
Bl
B25
F112
M18
G31
L21
H41
H612
D3
M15
H21
0.57
0.41
0.38
-0.30
-0.45
0.11
0.06
0.33
-0.25
0.32
0.28
0.02
-0.14
-0.34
0.24
0.30
0.06
0.40
0.06
-0.20
0.27
0.45
0.42
0.42
0.39
-0.38
-0.44
0.08
-0.02
0.02
-0.26
0.11
-0.23
0.03
0.17
-0.15
-0.38
-0.26
0.05
0.28
-0.12
0.15
-0.07
0.47
0.46
0.39
-0.31
-0.05
Tot. Explain.Variance
33.10
29.75
23.75
Finally, the third factor (24 per cent of sample variance) is composed of the following
variables:
H41
non-equity agreements with other firms
H612
frequent informal relationships within the district
D3
decentralization of more than 50 per cent of the sewing phase
MIS(-) investments in commercialization (negative correlation).
This last factor emphasizes the collective behaviour of firms: high decentralization of the
production process and frequent cooperation based on both formal and informal contacts with
other firms within the district.
With factor analysis, the three main dimensions of the two economic systems studied have
been identified: in Mexico they are, in order of importance, scale economies, quality of
products and collective behaviour, while in Italy the factors are expansion, size-quality and, as
in Mexico, collective behaviour.
Comparing the results obtained in Mexico and Italy, thefirstimportant issue to emphasis
is the relevance of economies of scale: the size of firms emerges as the most important element
in the first factor in Mexico, while it appears only in the second factor in Italy.
In Mexico, it is evident that investments and technological innovation are strongly related
to size: therefore enterprises within the district are not all the same and the existence of
external and district economies does not entirely eliminate the advantages of size.
In Italy, size appears to be related to a good technological level and production for a low
segment of the market, with a high share of exports while, in the first factor, a tendency to
increase employment is related to investments in technology, the search for alternative ways
of trading products, such as the creation of trading companies or the ownership of shops in the
main cities and production for the high or medium-high segment of market. This first factor
FOOTWEAR INDUSTRIAL DISTRICTS IN Mexico AND ITALY
17
is not related to size and it can therefore be concluded that, in Italy, the most advanced
evolution of the district and its effects limit the importance of size in carrying out investments
and commercial diversification.
With respect to size, it is also important to emphasis that in both Italy and Mexico large
firms appear rather indifferent to institutional support and less embedded in the spatial
environment than smaller enterprises.
In the second factor, in both Italy and Mexico, the quality of products has been
emphasized as one of the main dimensions of the economic system. In the footwear industry,
quality differentiates the structure of firms: in the high segment of the market, the production
process is highly labour-intensive, labour force skill assumes an essential role and scale
economies can be insignificant.
Finally, the third dimension identified is the collective one, emphasizing institutional
support and information access in Mexico and informal relationships and division of labour
among specialized firms in Italy. Therefore, in both cases, spatial environment comes out as
one of the three principal underlying dimensions of the economic systems analysed, confirming
the validity of the district approach.
V. The profiles of sample firms

1.
The statistical analysis
The objective of mis section is to classify the sample firms according to their performance,
identifying different profiles of structure and of internal and external conduct. This is done with
correspondence analysis, a multivariate descriptive statistical method that represents graphically
the rows and columns of a contingency table in the same low-dimensional space [Greenacre,
1984].
In Mexico, the trend of net profits during the last five years was adopted as an indicator
of performance: increasing (COL1), stable (COL2) and decreasing (COL3). With correspondence analysis, a contingency table has been created where, along the columns, the three
different performance profiles are represented and along the rows a number of variables,
indicating the structure of the firms and their internal and external conduct, are presented.
Because there are only three column profiles, two dimensions give a complete
representation of the profiles, with 76 per cent of total inertia" "explained" by the first axis
and the remaining 24 per cent by the second one.
The first significant result is the location of the three column profiles (COL1, COL2 and
COL3) in three separate sub-areas of the graphic display, identifying diree very different sets
of characteristics for each profile (Figure 1).
The partial contributions to inertia of column points indicate that the first principal axis
is defined by the clear opposition of COL1 (increasing profits) (0.63 of total inertia) from
COL2 (stable profits) (0.19) and COL3 (decreasing profits) (0.18), while the second axis is
completely defined by the opposition of COL2 (0.49) from COL3 (0.50).
With the analysis of partial contribution to inertia of row points it is possible to define the
first axis as given by:
11
A profile is the distribution of conditional frequencies of each category. The correspondence analysis problem
is to measure the distance between two generic row (column) profiles as chi-square distances and to calculate the
total inertia, defined as the weighted sum of squared distances from each row (column) profile and its respective
centroid.
18
Figure 1 : Correspondence analysis: Mexico
AX I S = I
-0.3
0.2
-0.1
0.0
0.1
AX IS=2
19
(a) institutional support (H78) (0.26 of total inertia), which is an important competitive
advantage for firms with increasing profits with respect to firms with stable or decreasing
profits;
(b) commercial strategy as a weakness of firms (NH8B) (0.15), appearing to be a more
significant difficulty for firms with stable or decreasing profits than for firms with
increasing profits;
(c) quality of product (H7A) (0.14): a competitive advantage more important for COL2 and
COL3 profiles than for those in COLl.
With regard to the first dimension, therefore, firms with increasing profits are
differentiated from firms with stable and decreasing profits because they attribute a greater
importance to institutional support, while in the other two categories their commercial weakness
and the quality of their products are given greater emphasis.
The second axis is characterized by:
(a) a high degree of technological innovation (C3512) (0.21), characterizing firms with stable
profits (COL2) in comparison with those with decreasing profits (COL3);
(b) small size (fewer than 50 employees) (B1N1) (0.13) characterizing COL3 from COL2;
(c) commercial strategy as a weakness of the firm (NH8B) (0.12) representing a more serious
problem for the profile with stable profits than for the one with decreasing profits.
After the interpretation of the two axes, the next step is to analyse the comparative profiles
according to the three typologies of performance, considered as poles of attraction of the row
characteristics. From Figure 1 the profile of firms with increasing profits (COLl) can be seen
as being characterized by:
(a) frequent use of use of more than one service supplied by the entrepreneurial association
(G131);
(b) commercialization of more than 50 per cent of the products through wholesalers (D1550);
(c) medium size (50-100 employees) (B1N2);
(d) low availability of components as a weakness of the firm (NH84);
(e) availability of skilled labour force (H73), access to information (H76) and institutional
support (H78) as competitive advantages of firms;
(f) high degree of technological innovation (C3512).
Firms with stable profits (COL2) are characterized by the following profile:
(a) large size (more than 100 employees) (B1N3);
(b) medium-high segment of market (DIII2) and original design (D7812);
(c) commercial strategy as a weakness of the firm (NH8B).
Finally, firms with decreasing profits (COL3) are characterized by:
(a) investments in commercialization (H55);
(b) high segment of market (DIII1);
20
65
(c) commercialization of more than 50 per cent of products through non-exclusive agents
(D1650);
(d) frequent informal cooperation with other firms within the district (Gl).
For Italy, three separate sub-areas have been identified in this case, too, illustrating three
very differentiated profiles of performance. It should be noted that the indicator of performance
has changed with respect to the Mexican sample; in the Italian case the trend of production
during the last five years was adopted, because the data on profits were not available for the
whole sample.
In the Italian sample, the first principal axis accounts for 61 per cent of total inertia and
the second accounts for the remaining 39 per cent.
The first dimension separates COL3 (stable trend of production) from COL2 (increasing
trend of production) and COM (decreasing trend of production). The second dimension
separates COL2 and COL3 from COM.
However, from the analysis of the partial contributions to inertia of column points, it is
evident that the first dimension is dominated by COL2 (0.23 of total inertia) versus COL3
(0.73), while the second one is dominated by COL2 (0.26) versus COM (0.73).
The first axis is determined by:
(a) availability of skilled labour force: high in the category COL2 (B1012: 0.11) in
comparison with COL3, characterized by low availability (B1013: 0.25);
(b) investments in technology (Mil) (0.11), characterizing COL2 with respect to COL3.
The second axis is determined by:
(a) frequency of informal relationships within the district (H612) (0.24), characterizing COM
with respect to COL2;
(b) low technological level (F113) (0.15), characterizing COM with respect to COL2;
(c) use of more than one service supplied by the entrepreneurial association (L21) (0.13),
characterizing COL 2 with respect to COM.
From Figure 2, we can identify the profile of firms increasing their production (COL2)
as characterized by:
(a) medium size, between 50 and 100 employees (B1N2);
(b) a good or regular technological level (Fill and F112) and investments in technology
(Mil);
(c) a commercial strategy based both on agents and direct sales to customers (G142 and G144)
and investments in commercialization (Ml5);
(d) rare informal relationships within the district (H613).
(e) use of more than one service supplied by the entrepreneurial association (L21).
(ft equity shares in other firms (H21).
The profile of firms with stable production (COL3) is characterized by:
(a) large size, more than 100 employees (B1N3);
(b) more than 50 per cent of exported production (B25N1);
21
Figure 2: Correspondence analysis: Italy
AX!S=1
0.51
0.0
-0.5H
-0.5
0.0
AX IS=2
0.5
22
65
(c) decentralization of less than 50 per cent the sewing phase (D31);
(d) low availability of skilled labour force (B1013).
Finally, the profile of firms decreasing their production is composed of:
(a) small size, fewer than 50 employees (B1N1);
(b) frequent informal relationships within the district (H612) and non-equity agreements with
other firms (H41).
From the comparison of the different identified profiles in the next section, several very
interesting insights are highlighted.
2.
Some findings of the correspondence analysis
In this section, a summary of the results of the correspondence analysis is presented,

leaving out all the technical details.
First of all, institutional support in Italy (L21) and in Mexico (H78 and G131) and other
external economies like access to information (H76) and availability of a skilled labour force
(H73) in Mexico appear to be significant elements of the successful performance of firms.
Therefore, external and district economies have a positive effect on the performance of firms
located within the district.
However, in Mexico a negative aspect of the district has also been emphasized by
successful enterprises, complaining about the low availability of components (NH84), a
problem stressed in Section m.l above.
The second important consideration is related to the size of firms: both in Italy and in
Mexico medium size (61N2) characterizes successful enterprises, large size (B1N1) stable ones
and small size (B1N3) losing firms.
Medium-sized firms are therefore able to take advantage in the best way of district
economies and, at the same time, they reach a scale large enough to invest in technology
(C3512 in the Mexican sample and F i l l , F112 and Mil in Italy). Moreover, in both Italy and
Mexico, alternative strategies of commercialization other than the traditional one are also
attempted: in Mexico selling through wholesalers (D1550) and in Italy selling directly to their
clients to increase their knowledge of the market (G142) and investing in commercialization
(M15).
In Mexico, the profile of stable firms is characterized by large firms producing for the
medium-high segment of market (DIII2), original designed products (D7812), finding
difficulties in the commercialization of their products (NH8B), notwithstanding their size and
therefore confirming the structural weakness on the commercial side of the Mexican footwear
industry. In Italy, stable firms are also large and tend to be quite vertically integrated (D3) and
export-oriented (B25N1); in addition they encounter some difficulties in employing a skilled
labour force.
Finally, the last profile of losing firms is characterized by small size in both Italy and
Mexico and by frequent informal cooperation within the district (Gl in Mexico and H612 in
Italy). Informal relationships in both countries therefore assume an important role mainly
among enterprises in crisis, unlike relationships with supporting institutions which are important
for successful firms. Informal cooperation may therefore be interpreted as a kind of solidarity
network, allowing firms to survive beyond market rules.
In Mexico, small losing firms are also characterized by the production of high quality
products (DIII3) and investments in commercialization (H55), notwithstanding that most of their
products are sold through non-exclusive agents (D1550). The conclusion which can be drawn
is that small firms, strongly oriented towards high quality, do not reach the scale needed to
create a commercial capability to distribute their products.
VI. Conclusions
Some significant differences in the macroeconomic environment have influenced the
evolutionary path of the clusters analysed. In Italy, a favourable international situation allowed
an outstanding growth of the footwear clusters led by exports, until changes in the international
market generated some difficulties in the districts, which have not yet been overcome.
In Mexico, the protection of the domestic market has favoured the growth of an industry
which was unable to compete in the international market after the liberalization of the economy,
with a number of consequences for the structure of the clusters and the linkages among firms.
With regard to the impact of macroeconomics on the evolution of the districts, two main
conclusions can be drawn: first of all, districts are necessarily different if embedded in different
macro contexts and this must be remembered when policy lessons are transferred from one
country to another.
Secondly, the districts analysed seem to be good at exploiting existing comparative
advantages but less so at dealing with more radical changes, like transformations in the market,
when the creation of new dynamic comparative advantages is needed.12 This last issue must
necessarily be addressed in the perspective of elaborating a strategy of industrial development
at district level in developing countries.
If districts are different because they are inserted into different macro contexts, also within
the same district economic agents are not all equal between themselves: one of the main
findings of the inquiry is, in fact, the relevance of size. First of all, it must be emphasized that,
in both countries, best performance is not equal to large size but to medium size and this is
indeed a very important finding. Medium-sized enterprises, both in Italy and Mexico, have a
scale large enough to be able to invest in technology and commercialization, with the capability
to efficiently exploit some of the external economies available in the cluster. The district
method of organizing production is therefore a competitive organizational form with respect
to vertical integration, not only a way to overcome scale disadvantages.
Notwithstanding that in both Italy and Mexico medium sized enterprises appear to perform
best, it must be noted that in Italy, where the districts and their related economic effects are
more advanced (e.g. the existence of collaborative backward linkages instead of market ones,
as in Mexico), size assumes a less significant role than in Mexico, where it comes out in factor
analysis as the most important element of the first extracted factor.
The last point to emphasize is related to the collective dimension of the systems analysed,
which can be divided into two different parts: a more formal, institutionalized cooperation and
an informal one. A significant result of the analysis is the role of the institutional support on
the positive performance of enterprises, while informal relationships among firms assume a
relevant role mainly for losing (small) firms, as a solidarity network, allowing enterprises
located in the districts to survive beyond market rules.
A final general conclusion which can be drawn is the usefulness of adopting the concept
of the industrial district to study clusters of spatially concentrated and sectorally specialized
12
The question of the capability of the districts to react to radical changes was the focus of an inquiry
undertaken in three footwear districts in Italy by the author [Camagni & Rabellotti, 1993].
23
24
65
firms in developing countries, comparing the results with similar clusters in advanced countries.
In this analysis of footwear clusters, differences related to different macro contexts and some
similarities of structures and strategies within the analysed clusters have been found. The
comparison has emphasized the importance of adopting a dynamic concept of the industrial
district, identifying stages of development and different connected typologies of districts, in the
perspective of learning industrial lessons from the experience of some developed countries to
elaborate an industrial policy at "industrial district" level.
Bibliographical references
ANCI {Associazione Nazionale Calzaturifici Italiani). 1992. Relazione Economica Statistica, Milan.
. 1993. Relazione Economica Statistica, Milan.
Arias, P. 1980. "El proceso de industrializacidn en Guadalajara, Jalisco: Siglo XX", in Relaciones,
Vol. 1, No. 3, Zamora, Michoacan, El Colegio de Michoacan.
. 1985. Guadalajara, le gran ciudad de la pequena industria, El Colegio de Michoacan, Zamora,
Michoacan.
. 1992. El calzado en la regidn jalisciense: la industria y la cdmara, Guadalajara, Camara de la
Industria del Calzado de Jalisco.
ASSOMAC (Associazione Nazionale Costruttori Italiani Macchine e Accessori per Calzature e
Pelletteria). 1992. Rapporto Congiunturale, Vigevano.
Bagnasco, A. 1977. Tre Italie: Laproblematica territoriale dello sviluppo italiano, Bologna, II Mulino.
Becattini, G. 1979. "Dal 'settore industriale' al 'distretto industriale'. Alcune considerazioni sull'unita
di indagine dell'economia industriale", inL'industria, No.l.
Boston Consulting Group. 1988. Industria del calzado, Mexico City (mimeo).
Bmsco, S. 1982. "The Emilian model: Productive decentralisation and social integration", in Cambridge
Journal of Economics, No. 6, pp. 167-184.
Camagni, R. 1989. "Cambiamento tecnologico, milieu locale e reti di imprese: verso una teoria dinamica
dello spazio economico", in Economia e Politico Industriale, No. 64, pp. 209-235.
Camagni R.; Rabellotti, R. 1993. Footwear milieux in Italy: A dynamic comparative analysis. Paper
presented at the IV Gremi Conference, held in Ascona, 29-31 October.
CANAICAL (Cdmaras de la Industria del Calzado). 1991. Estudio tecnico sobre el dano a la industria
national del calzado por el incremento en las importaciones, Mexico City (mimeo).
CIGEC {Cdmara de la Industria del Calzado de Guanajuato). 1990. "Perfil de la industria del calzado
en M6xico", in Calzavance, September.
CONCALZADO. 1991. "Perfil de la industria del calzado", in Calzavance, September.
de la Cruz Labarthe, M.R. 1985. Notas sobre el proceso de industrialization de Leon. Autobiografia de
un obrero del calzado, Cuadernos de Investigaci6n, El Colegio del Bajo, Le6n.
Dominguez-Villalobos; Grossman. 1992. Employment and income effects of structural and technological
changes in footwear manufacturing in Mexico, World Employment Progamme Research, Working
Paper No. 224, Geneva, International Labour Office.
Fua, G. 1983. in Fua, G.; Zacchia C. (eds.): Industrializzazione senza fratture, Bologna, II Mulino.
25
Gaibisso, A.M. (ed.) 1992. Struttura e competitivita del settore calzaturiero, Milan, Franco Angeli.
Goglio, S. (ed.) 1982. Italia: centri e periferie, Bologna, II Mulino.
Goodman, E.; Bamford, J. (eds.) 1989. Smallfirmsand industrial districts in Italy, London & New
York, Routledge.
Greenacre, M.J. 1984. Theory and correspondence analysis, London, Academic Press.
ILO. 1992. Recent developments in the leather and footwear industry, Fourth Tripartite Technical
Meeting for the Leather and Footwear Industry, Reports I and II, Geneva, International Labour
Office.
Kim, J.; Mueller, C. 1978. Introduction to factor analysis, Sage University Papers, Series: Quantitative
Applications in the Social Sciences, No. 07-013, Beverley Hills and London.
Piore, M.; Sabel, C , 1984. The second industrial divide: Possibilities for prosperity, New York, Basic
Books.
Pyke, F. et al. 1990. Industrial districts and inter-firm co-operation in Italy, Geneva, International
Institute for Labour Studies.
Pyke, F.; Sengenberger, W. (eds.) 1992. Industrial districts and local economic regeneration, Geneva,
International Institute for Labour Studies.
Rabellotti, R. 1993a. "Industrial districts in Mexico The case of the footwear industry", in Small
Enterprise Development, Vol. 4, No. 3, pp. 26-36.
. 1993b. Is there an "industrial district" model? A comparison between footwear districts in Italy and
Mexico, Paper presented at the IDS Industry Cluster Workshop on Intra-firm and Inter-firm
Reorganisation in Third World Manufacturing Industry, Brighton, University of Sussex, 14-16th
April.
SARTRA, 1992. World Footwear Markets, SARTRA Footwear Technology Centre, Kettering, March.
SECOFI, 1992. "Programa para promover la competitividad e internacionalizacion de la industria dc la
curtiduriay del calzado", Mexico City (mimeo).
Varaldo, R. (ed.) 1988. // sistema delle imprese calzaturiere, Giappichelli Editore, Florence.
Titles still available in the Discussion Paper Series

L 'analyse du marche du travail urbain en Afrique, by Jean-Pierre Lachaud. No. 2. 1987.
ISBN 92-9014-420-3.
Non-standard employment practices in Latin America, by Adriana Marshall. No. 6. 1987.
ISBN 92-9014-427-0.
Latin America's labour market research: A state of the art, by Jorge Jatoba. No. 9. 1988 (rev. 88).
ISBN 92-9014-448-3.
Les politiques du temps de travail dans 13 pays industrialises, by G. Bosch and F. Michon. No. 16. 1989.
ISBN 92-9014-460-2.
Empleo temporario y trabqjo a tiempo parcial en la Argentina, by Adriana Marshall. No. 17. 1989.
ISBN 92-9014-480-7.
Flexible specialisation in Third World industry: Prospects and research requirements, by Hubert Schmitz.
No. 18. 1990. ISBN 92-9014-481-5.
The role of labour standards in industrial restructuring: Participation, protection and promotion, by Werner
Sengenberger. No. 19. ISBN 92-9014-510-2.
Subcontracting as an alternative organizational form to vertically-integrated production, by Mark Lazerson.
No. 20. 1990. ISBN 92-9014-483-1.
The secrets of industry are in the air. Cooperation industrielle et equilibre organisationnel de la firme
innovatrice, by Dominique Foray. No. 21. 1990. ISBN 92-9014-484-X.
State employment in the fiscal crisis: The Latin American case, by Adriana Marshall. No. 22. 1990.
ISBN 92-9014-485-8.
Recent trends of industrial subcontracting in Japanese manufacturing industry, by Naoki Kuriyama. No. 24.
1990. ISBN 92-9014-487-4.
Public sector labor market and privatization in Korea, by Young-bum Park. No. 25. 1990.
ISBN 92-9014^88-2.
Formas precarias de trabqjo asalariado: dos estudios en el drea metropolitana de Buenos Aires, by Adriana
Marshall. No. 26. 1990. ISBN 92-9014-489-0.
Developments in the reorganisation of the Swiss economy: The small and medium-sized enterprise structure as
an opportunity and a threat, by Beat Hotz-Hart. No. 27. 1990. ISBN 92-9014-491-2.
Bilan et propositions de recherches concernant I 'approche du developpement despetites et moyennes entreprises
en Afrique sub-saharienne, by Jose" Trouve\ No. 28. 1990. ISBN 92-9014-490-4.
Small enterprises and business work partnerships in Hungary, by Peter Galasi and Gyorgy Sziiiczki. No. 29.
1990. ISBN 92-9014-492-0.
Labour market flexibility. No. 31. 1990. ISBN 92-9014-494-7.
Laflexibilite des marchis du travail. No. 32. 1990. ISBN 92-9014-495-5.
Le marche du travail urbain au Maroc, by M. Salahdine. No. 33. 1991. ISBN 92-9014-496-3.
Le marchi du travail urbain en Algirie, by Chantal Bernard. No. 34. 1991. ISBN 92-9014-497-1.
El empleo atlpico en Lima-Metropolitana, by F. Verdera V. No. 36. 1991. ISBN 92-9014-499-8.
Review of the developments in the organisation and structure of small and medium-sized enterprises in Norway,
by Kristin Dale. No. 37. 1991. ISBN 92-9014-500-5.
The impact of labour law on employment practices: Temporary and part-time employment in Argentina and
Peru, by Adriana Marshall. No. 38. 1991. ISBN 92-9014-501-3.
Pauvrete et qffre de travail a Abidjan (Cote d'lvoire): concepts, methode et analyse, by Jean-Pierre Lachaud.
No. 39. 1991. ISBN 92-9014-502-2.
Priorities for urban labour market research in anglophone Africa, by William J. House. No. 40. 1991.
ISBN 92-9014-505-6.
Labour institutions and economic development, by G. Rodgers. No. 41. 1991. ISBN 92-9014-504-8.
Growth model, state social intervention and labour regulation, Argentina, 1890-1990, by Rosalia Cortes and
Adriana Marshall. No. 42. 1991. ISBN 92-9014-506-4.
Impact du disengagement de I'Etatsur lemarchi du travailau Togo, by Dodgi V. Silete-Adogli. No. 43.1991.
ISBN 92-9014-507-2.
Industrial districts and the Valencian community, by J.-A. Ybarra. No. 44. 1991. ISBN 92-9014-508-0.
Trade unions and decentralised production: A sketch of strategic problems in the West German labour
movement, by Horst Kern and Charles F. Sabel. No. 45. 1991. ISBN 92-9014-509-9.
Globalisation and strategic choices in tripartite perspective: An agenda for research and policy issues,
by Duncan Campbell. No. 46. 1991. ISBN 92-9014-511-0.
Recent developments in the small enterprise sector in India: Economic and social aspects, by Surendra Pal
Kashyap. No. 48. 1992. ISBN 92-9014-513-7.
Wage-setting institutions in Japan and the Republic of Korea, by Tsuyoshi Tsuru and Young-bum Park. No. 51.
1993. ISBN 92-9014-517-X.
Pauvreti et marchi du travail au Mali: le cas de Bamako, by J.-P. Lachaud and B. El Hadji Sidibe'. No. 52.
1993. ISBN 92-9014-518-8.
Les icarts de salaires entre les secteurs public et prive en Afrique francophone: analyse comparative, by J.-P.
Lachaud. No. 53. 1993. ISBN 92-9014-519-6.
The creation of employment in segmented labour market: A general problem and its implications in India,
by Gerry Rodgers. No. 54 (rev. 1). 1993. ISBN 92-9014-520-X.
Pauvreti et marchi du travail urbain en Afrique au sud du Sahara : analyse comparative, by Jean-Pierre
Lachaud. No. 55. 1993. ISBN 92-9014-521-8.
L'ajustement structurel et le marchi du travail en Afriquefrancophone, by Jean-Pierre Lachaud. No. 56. 1993.
ISBN 92-9014-522-6.
Pauvreti et marchi du travail a Ouagadougou (Burkina-Faso), by Ourobe" Mathias Sanou. No. 57. 1993.
ISBN 92-9014-523-4.
Economic consequences of labour protection regimes: A comparative study of Latin American countries, by
Adriana Marshall. No. 58. 1993. ISBN 92-9014-543-9.
Trends in deprivation in the London labour market: A study of low incomes and unemployment in London
between 1985 and 1992, by Peter Lee and Peter Townsend. No. 59. 1993. ISBN 92-9014-544-7.
Employment deprivation and poverty: The ways in which poverty is emerging in the course of economic reform
in Russia, by Natalia Tchernina. No. 60. 1993. ISBN 92-9014-545-5.
Labour institutions and technological change: A framework for analysis and a review of the literature,
by M. Carnoy, S. Pollack and P. Lindquist Wong. No. 61. 1993. ISBN 92-9014-546-3.
Social exclusion and Africa south of the Sahara: A review of the literature, by Charles Gore. N3 62. 1994.
ISBN 92 9014-547-1.
Some paradoxes of social exclusion, by Marshall Wolfe. No. 63. 1994. ISBN 92-9014-547-1.
Mexican maquiladoras: Origins, operations and outlook, by Philip L. Martin. No. 64. ISBN 92-9014-548-X.
A complete list of IILS publications can be obtained upon request from the International Institute for Labour
Studies, P.O. Box 6, CH-1211 Geneva 22 (Switzerland).

Globalisasi

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Globalisasi

Transféré par

Droits d'auteur :

Formats disponibles

INTERNATIONAL INSTITUTE FOR LABOUR STUDIES

Labour Institutions and New

Footwear industrial districts

n. The Italian and Mexican footwear industry

IV. The dominant economic structures

V. The profiles of sample

DISCUSSION PAPER SERIES NO.

//. The Italian and Mexican footwear industry

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

DISCUSSION PAPER SERIES NO. 65

Source: ANCI [1992].

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

Note: At 1980 prices.

DISCUSSION PAPER SERIES NO.

The acceleration in the opening up of the Mexican market to international competition,

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

///. The sample

> 100 employees

DISCUSSION PAPER SERIES NO.

The footwear clusters

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

Source: Author's survey.

DISCUSSION PAPER SERIES NO.

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

DISCUSSION PAPER SERIES NO.

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

IV. The dominant economic structures

DISCUSSION PAPER SERIES NO. 65

good or very good technological level

In economic terms, the factor can be interpreted as a confirmation of the importance of

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

DISCUSSION PAPER SERIES NO. 65

entrepreneurial association. This category of firms seems to be independent of the district

FOOTWEAR INDUSTRIAL DISTRICTS IN Mexico AND ITALY

V. The profiles of sample firms

The statistical analysis

DISCUSSION PAPER SERIES NO. 65

Figure 1 : Correspondence analysis: Mexico

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

DISCUSSION PAPER SERIES NO.

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

Figure 2: Correspondence analysis: Italy

DISCUSSION PAPER SERIES NO.

Some findings of the correspondence analysis

In this section, a summary of the results of the correspondence analysis is presented,

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

DISCUSSION PAPER SERIES NO.

FOOTWEAR INDUSTRIAL DISTRICTS IN MEXICO AND ITALY

Titles still available in the Discussion Paper Series

Vous aimerez peut-être aussi