Economic Liberalization and the Computer Industry: Comparing Outcomes in Brazil and Mexico
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
World Development Vol. 29, No. 7, pp. 1199±1214, 2001
Ó 2001 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
www.elsevier.com/locate/worlddev 0305-750X/01/$ - see front matter
PII: S0305-750X(01)00038-9
Economic Liberalization and the Computer Industry:
Comparing Outcomes in Brazil and Mexico
JASON DEDRICK, KENNETH L. KRAEMER
University of California, Irvine, USA
JUAN J. PALACIOS
University of Guadalajara, Mexico
PAULO BASTOS TIGRE
Universidade Federal do Rio de Janeiro, Brazil
and
ANTONIO JOSE JUNQUEIRA BOTELHO *
Ponti®ca Universidade Catolica do Rio de Janeiro and Universidade Estadual do Rio de
Janeiro, Brazil
Summary. Ð Market liberalization has been carried out by many developing countries in the hopes
of stimulating trade, investment and technology transfer. In order to analyze the impacts of
liberalization on a speci®c industry sector, this paper compares the experiences of Brazil and
Mexico in liberalizing the computer industry in the 1990s. The authors conclude that liberalization
leads to lower prices and more rapid diusion of computer use throughout the economy, but at a
cost to domestic computer ®rms who were harmed by foreign competition. Both countries saw an
increase in computer production, but Mexico's production was mainly for export to the United
States, while Brazil was producing for the domestic market. The dierences between outcomes in
the two countries have been determined more by environmental factors than by the nature and pace
of liberalization. Ó 2001 Elsevier Science Ltd. All rights reserved.
Key words Ð Latin America, Brazil, Mexico, computer, information technology, liberalization,
outcomes of liberalization
1. INTRODUCTION the vertically-integrated mainframe computer
industry structure (Grove, 1996). This devel-
Market liberalization has been embraced by
most developing countries around the world
since the 1980s, driven partly by neoliberal * This research has been supported by grants from the
ideology and partly by pragmatism in an Sloan Foundation; the Computation and Social Systems
increasingly interconnected global economy. Program in the Computer, Intelligent Systems and
Developing countries who once feared Engineering (CISE) Division and the International
economic dependency now fear being left out of Program of the Social and Behavioral Sciences (SBE)
the dynamic process of globalization. Division, US National Science Foundation; the Univer-
The Information Technology (IT) industry sity of California Institute for Mexico and the United
has become perhaps the most globalized of all States, and the Consejo Nacional de Ciencia y Tecno-
industrial sectors, especially since the growth of logia of Mexico (UC Mexus-CONACyT) Collaborative
the personal computer (PC) industry whose Grants Program. Final revision accepted: 3 February
horizontally specialized structure has replaced 2001.
11991200 WORLD DEVELOPMENT
opment led to the growth of a global computer role of the state in promoting national indus-
production network focused on East Asian trial development, based on the experiences of
countries which specialized in particular East Asian countries (Amsden, 1989;
industry segments such as PC assembly, Anchordoguy, 1989; Wade, 1990). They have
components and peripherals (Borrus, 1997; credited the governments of these countries
Dedrick & Kraemer, 1998; Ernst, 1994; Wong, with providing domestic industries with selec-
1995). Motivated in part by the success of these tive protection to compete internationally
countries, newcomers such as China, India, and while introducing limited competition to
several others in eastern Europe and Latin national markets. Their rationale is that
America have liberalized their computer sectors simply opening up a country to international
in hopes of joining the global production competition without a period of adjustment
network and gaining access to technology and and without providing support to domestic
capital. producers will result in the destruction of
This paper analyzes the liberalization process national industries.
in two large developing Latin American econ- This paper addresses the following issues
omiesÐBrazil and Mexico. Both of these raised by the debate over liberalization:
countries pursued protectionist policies in the ÐIs promotion of a domestic computer
computer industry during the 1980s, and both industry behind protective barriers a tenable
liberalized their markets in the early 1990s. strategy?
While the general outlines and timing of the ÐDoes liberalization lead to optimal results
liberalization process are similar for the two (in terms of output, trade, employment,
countries, there are important dierences in technology diusion) or are there tradeos
national environments, the pace of liberaliza- between dierent outcomes? For example,
tion, and the nature of related IT policies. As a does liberalization lead to lower prices and
result, the outcomes and impacts of liberaliza- faster diusion, but harm domestic produc-
tion have been quite dierent for the two ers?
countries. This paper contrasts the two experi- ÐDoes national environment matter or will
ences and assesses the respective results. similar policies lead to similar outcomes in
dierent contexts?
ÐCan government policy intervention lead
2. THEORY, CONCEPTUAL to better results than simply liberalizing
FRAMEWORK AND METHODOLOGY and allowing market forces to work?
ÐWhat are the elements of a comprehensive
Countries around the world have been policy approach to computer industry liber-
implementing economic liberalization programs alization?
in a variety of forms over the past 15 years. The analytical framework used to address these
These have ranged from deregulation and issues is presented in Figure 1, which posits that
privatization programs in developed countries the impacts of liberalization are in fact in¯u-
such as the United States, the United Kingdom enced by national context, including: (a) the
and New Zealand, to outright market reforms respective country's economic and political
such as those in China and the former Soviet environment, its geographical location, its
bloc. Latin American countries have likewise information infrastructure and national capa-
been undertaking various liberal economic bilities such as human resources and technol-
reforms. The reform process was guided in ogy, and the structure and composition of
many casesÐChile and Mexico in particularÐ computer production and the extent of
by technocrats trained in the economics computer use at the time of liberalization; (b)
departments of North American universities. the nature of the liberalization process in
Their ideas on liberalization policies have thus question, including the timing, pace and extent
been based on the theoretical underpinnings of of liberalization, and related IT policies, such
neoclassical economics (e.g., Friedman & as investment incentives, export promotion,
Friedman, 1980; Little, 1982), which argues that incentives for IT use, training in IT skills and
the market is the best mechanism for both telecommunications policies.
allocating resources and determining invest- Key outcome variables include the impact of
ment decisions in virtually every context. liberalization on industry structure, computer
In contrast to neoclassical theory, another production, trade, the domestic IT market and
school of analysts have argued for a stronger the nature of IT and Internet use.LIBERALIZATION AND THE COMPUTER INDUSTRY 1201
Figure 1. Framework for comparison.
This paper is based on research by the have never succeeded beyond the domestic
authors in both Mexico and Brazil, reported in market. Japan's PC market was isolated by the
several individual country papers (Botelho, adoption of technology standards incompatible
Dedrick, Kraemer, & Tigre, 1999; Dedrick, with the global standard, and its PC makers
Kraemer, & Palacios, 2001; Palacios, 2000; have had very limited success outside Japan
Tigre & Botelho, 2001). The primary research (West & Dedrick, 2000). By contrast, successful
included extensive secondary literature reviews, PC-producing countries such as Taiwan,
and interviews with over 50 people in each Singapore and Ireland have integrated them-
country, representing the computer industry, selves into the global production networks led
IT user organizations, industry associations, by multinational computer makers (Dedrick &
government and academia. The authors met for Kraemer, 1998; Tallon & Kraemer, 2000).
two days in September 1999 to review and By the early 1990s, a global production system
compare the ®ndings of the two cases and draw was well established for computer hardware,
conclusions as presented in the present paper. with the United States, Japan and East Asia
dominating the industry. Newcomers faced dim
prospects for breaking in unless they had a
3. ENVIRONMENT unique asset, such as China's large potential
market, to attract investment by multinational
(a) Global environment corporations (MNCs). But the industry has seen
a shift toward build-to-order production in the
The global environment in¯uences the poli- late 1990s, requiring PC assemblers and some of
cies of national governments and de®nes the their suppliers to locate production close to the
opportunities and challenges facing policy ®nal market. This trend has led to regional
makers and business leaders. The computer production strategies, with major PC makers
industry is highly globalized, with common having separate assembly plants in Asia, North
technology standards used around the world. America and Europe in order to serve the three
This means that thousands of companies are major markets. The result has been new
doing research, designing and manufacturing opportunities for countries such as Ireland,
products, and developing software for a few Scotland, Hungary and as we shall see, Mexico.
standard technology platforms, particularly the
IBM/Wintel PC. In the face of such global (b) Economic environment
economies of scale and technological dyna-
mism, it has proven impossible to develop a Brazil and Mexico are the largest economies
competitive computer industry in an isolated of Latin America, with Brazil's GNP about
national market. Examples of countries that twice that of Mexico. The countries have similar
tried and failed to develop competitive PC levels of GNP per capita and industrialization.
industries in protected markets include India Brazil has managed to grow about twice as fast
and China, each of which have partially liber- as Mexico since 1980, although it was plagued
alized their PC markets in the 1990s (Dedrick & by high in¯ation until the mid-1990s.
Kraemer, 1993; Kraemer & Dedrick, 1995). While Mexico became more open during
Korea protected its PC industry with a ban on 1980±96, with trade increasing from 24% to
imports during 1982±88, and its PC makers 42% of GDP, Brazil actually became more1202 WORLD DEVELOPMENT
self-contained, with trade dropping from 20% stability brought about by the end of in¯ation
to 15% of GDP (World Bank, 1999). Privati- with the 1994 Real Plan. Even before then, the
zation of state enterprises in Brazil gained Brazilian State had begun reducing its
impetus from 1996 and in some sectors, such as economic role. The Brazilian Congress, domi-
telecommunications, has been completed. Both nated by a multiparty government alliance, has
countries' mix of exports has shifted from supported the government's strategies of liber-
primary products to manufactured goods, but alization and privatization, although it has
much more dramatically in Mexico. resisted more profound changes in the critical
Both countries endured foreign debt crises areas of social security and public de®cit
during the 1980s, leading to severe economic control (Tigre & Botelho, 2001).
downturns. In response to these crises and Mexico was governed by the Institutional
resulting pressure from international lending Revolutionary Party (PRI) from 1929 until
agencies, both countries began to open up their 2000, the longest continuing rule of any politi-
protected domestic markets. By the late 1990s, cal party in the world. The PRI's hegemony
each had privatized state enterprises, opened ®nally ended with the election of opposition
major industry sectors to foreign investment, candidate Vicente Fox in 2000, but the
and lowered taris and other trade barriers. In conservative Fox is not expected to backtrack
addition to beginning earlier, in Mexico, the on the liberalization eorts of the past 15 years.
liberalization process was swifter and broader,
while in Brazil it unfolded more gradually.
Both countries have joined the World Trade (d) National capabilities
Organization and both have entered free trade
agreements in the 1990s. Mexico joined the The potential impacts of liberalization on a
United States and Canada in NAFTA, while country are aected by a set of national capa-
Brazil spearheaded the creation of the bilities needed for production and use of
MERCOSUR (MERCOSUL in Portuguese) computers and information technology. Most
free trade area, which includes Argentina, important of these are the quality of the tele-
Paraguay and Uruguay. For Brazil, MERCO- communications infrastructure and human
SUR is signi®cant, but the three other countries resources, and the size and competitiveness of
have a combined GNP of only US$335 billion, the IT industry before liberalization. Telecom-
or less than half that of Brazil. On the other munications infrastructure is vital to eective
hand, NAFTA gives Mexico access to the US use of IT, especially in the era when Internet
and Canadian markets, with a combined GNP and other network technologies are coming to
of US$8.3 trillion (World Bank, 1999). This dominate. Human resources include everything
asymmetry, as well as its close proximity to the from literate workers for advanced manufac-
United States, make Mexico a logical place to turing facilities to engineers and technicians
locate high-volume manufacturing for the needed by the IT industry, to scientists and
North American market, while production in researchers who can conduct research and
Brazil is still mostly for the domestic or development (R&D) and develop new tech-
MERCOSUR markets. nologies.
The economic dierences between Brazil and The capabilities of the existing IT industry
Mexico also help explain the dierent approa- can also aect the results of liberalization. If
ches taken to liberalization. Brazil, with its domestic companies are already internationally
larger economy and distance from other major competitive, they are likely to survive and even
markets, has been more inclined to protect and thrive under liberalization. If they have
promote local industries, while Mexico has succeeded only by taking advantage of the
been willing to open its market in return for the protected domestic market, they are unlikely to
promise of access to the large North American survive the onslaught of foreign competition.
market. Even if existing companies do not survive,
however, there may be individuals who have
(c) Political environment gained experience starting and managing
companies, and these skills will still prove
Brazilian politics have been in a period of valuable after liberalization. They can shift to
relative stability since the early 1990s. The dierent industry sectors, start new companies
ongoing democratization, begun in the late or help foreign MNCs carry out higher value-
1970s, was further consolidated by economic added activities within the country.LIBERALIZATION AND THE COMPUTER INDUSTRY 1203
Table 1. Comparative telecommunications indicators, managers of IT companies in both countries
1998a con®rm the high quality of engineering skills
Main lines Cellular mobile for the hardware industry, and a good supply
Per 1,000 Subscribers per 1000 of software skills as well. These skills are
inhabitants inhabitants available at a much lower cost than in indus-
Mexico 104 35 trialized countries, an additional attraction for
Brazil 121 47 IT companies.
Korea 432 302
Taiwan 524 216
Malaysia 202 101
(g) IT industry
a
Source: ITU (2000). Prior to liberalization, Brazil had a larger
and more technologically advanced IT indus-
The ability of a country to participate in a try, which produced a wide range of systems,
global, post-liberalization environment and peripherals and even some components (Tigre,
realize the potential bene®ts of that participa- 1992). But, the industry produced mostly for
tion depends to a great extent on whether it has the protected local market and had not honed
or can develop all of these capabilities. itself against foreign competition. Mexico's IT
industry, by contrast, was dominated by
foreign MNCs even during the protectionist
(e) Telecommunications infrastructure 1980s. Locally-owned companies had initially
tried to develop their own PC designs and
The telecommunications infrastructures of brands, but once the restrictions on component
Brazil and Mexico are quite similar in terms of imports were loosened, they shifted to simple
wired and wireless penetration. As Table 1 assembly of PCs based on foreign designs and
shows, however, both countries lag signi®cantly components (Borja, 1992).
behind Asian countries such as Malaysia, Korea These dierences help to explain Brazil's more
and Taiwan in terms of main lines and cellular active local entrepreneurs after liberalization.
phone subscribers per 1,000 people. Both Latin Both countries do have well-educated entrepre-
American countries have seen improvements, neurs who have entered the IT industry, but
thanks to privatization and deregulation of the there is more activity in Brazil. This could be due
telecommunications sector in recent years, and to the larger role of foreign multinationals in the
further improvement is likely. Mexican market, oering more job opportuni-
ties and also more competition to potential
(f) Human resources entrepreneurs. In addition, the larger Brazilian
market oers more opportunities for startups to
The general level of education in Mexico and target niche markets at home, and Brazil's
Brazil is about average for their income levels. government has more actively supported local
In terms of IT skills, both countries have high- entrepreneurs through export assistance and
quality personnel and quite large numbers of incubator programs. But the fact that Brazil had
professionals, as Table 2 shows. Interviews with a domestically-owned computer industry before
Table 2. Human resource indicators
Country Mexico Brazil Korea Taiwan Malaysia
Population (millions)a 96.5 159 44.9 21 20.1
Adult literacy (%)a 90 83 98 na 84
Mean years of educationb 4.7 3.9 8.8 NA N/A
Secondary enrolment ratio (%)a 58 45 101 NA 57
Masters and Ph.Ds in science and engineering 5916 N/A 7070 4011 N/A
awarded, 1990c
R&D scientists and technicians per 1,000 peoplea 0.3 0.2 2.9 N/A 0.2
Number of software professionalsd 321,482 549,840 340,168 140,070 53,389
a
UNDP (1998).
b
UNDP (1993).
c
Dedrick and Kraemer (1998). For Mexico, CONACyT (1995).
d
Jones (1998).1204 WORLD DEVELOPMENT
liberalization meant that many professionals global production network for high-volume PC
gained experience in ®nancing, launching and production (Dedrick & Kraemer, 1998).
managing their own companies, a set of skills These two developments led to a shift in
developed by fewer Mexicans. strategy in 1985 when IBM asked permission to
produce PCs in Guadalajara, but balked at the
joint venture requirement in the Computing
4. LIBERALIZATION OF THE Program. The Mexican government was
COMPUTER SECTOR opposed to granting an exemption since other
foreign ®rms had already formed joint ventures
(a) Pre-liberalization policies with Mexican partners. But keeping IBM out of
Mexico's PC market would isolate the country at
During the 1980s, both Brazil and Mexico a time that it hoped to become more integrated
employed protectionist policies in order to into the emerging industry (Dedrick et al., 2001).
promote their respective computer industries. Ultimately, IBM and the Mexican govern-
The policies diered in scope, with Brazil's ment came to an agreement that gave IBM the
``market reserve'' policy causing more conster- right to retain full ownership of its new plant in
nation among trading partners, especially the return for concessions on R&D and develop-
United States. Brazil's policies placed a heavier ment of local suppliers. Most important, IBM
emphasis on local ownership, while Mexico agreed to export a large share of the plant's
initially favored joint ventures and ultimately output, helping Mexico reduce its trade de®cit
allowed full foreign ownership. Both countries in PCs from $200 million in 1985 to $78 million
accepted the de facto presence of IBM in the in 1987 (Borja, 1992). After the IBM decision,
mainframe computer market, and targeted the the government shifted its emphasis from
emerging minicomputer and microcomputer industry protection to encouraging IT use and
(PC) markets for development. eventually to liberalization.
In 1981, the Mexican government formulated Brazil's computer industry policies in the
the so-called Program to Promote the Manu- 1970s and 1980s followed a ``greenhouse''
facture of Electronic Computing Systems, strategy of protecting domestic producers to
Their Central Processing Units and Their allow them to grow and innovate. The goal was
Peripheral Equipment, more commonly refer- to limit linkages between local and foreign
red to as the Computing Program. The key ®rms, so that local companies could develop
goals of the plan were to generate local capabilities for autonomous innovation with-
production of microcomputers, peripherals and out depending on foreign technology (Evans,
components, to promote exports, and to 1992). This policy, dubbed the market reserve,
achieve greater autonomy in computer tech- focused on the lower end of the marketЮrst
nologies (Borja, 1992). minicomputers, and then microcomputers as
As part of the program, access to the well. Foreign competitors were kept out of
domestic computer market was limited to those segments of the Brazilian market through
companies that would produce according to the restrictions on imports and foreign investment.
Program's provisions, and foreign ownership Local companies were prevented from acquir-
was limited to 49% in the production of PCs ing foreign technology without government
and peripherals. Foreign companies were permission.
required to invest between 3% and 6% of gross This policy gave local ®rms the space to grow
sales in R&D, and include a minimum and develop capabilities, and by the end of the
proportion of Mexican-made parts and 1980s, Brazil had a set of diversi®ed IT
components in their systems. Preferential companies with signi®cant presence in the local
treatment was given in government procure- market. The output of local computer hard-
ment to companies registered in the program. ware producers grew from less than $200
Almost as soon as the policy was put in place, million in 1979 to more than $4 billion in 1990,
the environment began to change. First, the debt and some local ®rms produced impressive
crisis of 1982 shifted the government's emphasis results in R&D (Evans, 1992). But, the industry
from industrial policy to ®nancial policy, with was largely isolated from the dynamism of the
the balance of payments becoming the critical global PC industry, and Brazilian computer
concern. Second, the introduction of the IBM- companies were not competitive outside the
PC led to standardization of the PC industry on protected domestic market. Moreover, many
the IBM architecture and the creation of a local ®rms made no eort to innovate andLIBERALIZATION AND THE COMPUTER INDUSTRY 1205
chose instead to pirate or surreptitiously license cies for the computer sector more in line with
foreign technology while taking advantage of those of the United States and Canada and
the market reserve. thus, minimized the changes needed when
The market reserve policy also raised the cost NAFTA was implemented. The continuing
of computers to users, although, as Tigre (1992) reticence to adopt any computer industry
points out, the price gap between Brazil and the policies or incentives is in¯uenced by concerns
United States dropped signi®cantly in the late that such policies would be seen in the United
1980s, so that Brazilian prices were more in line States as attempts to pull jobs across the border
with prices in Europe and Latin America. Still, to Mexico.
there was a delay in introduction of new tech-
nologies that ranged from one to several years.
(c) Nature of liberalization
(b) Forces driving liberalization
Liberalization of Mexico's computer sector
took place in a sweeping one-time move when
Brazil began to reduce the heavy role of the
the Computing Program was abandoned in
state in its economy in the early 1990s, in the
1990. The only remaining barrier to the Mexi-
wake of a decade of economic stagnation. In
can market was a 20% tari on hardware,
the computer industry, the market reserve was
which fell to 12% by 1994, and was phased out
abandoned in 1992 and replaced by a more
by 1998 within NAFTA. In 1999, the Mexican
market-oriented policy. The shift was driven in
government established the Sectoral Promotion
part by fears that Brazil was being left behind
Program for producers and nontrading
technologically in computers and IT. In addi-
companies in the electric±electronics sector to
tion, pressure for change was brought to bear
import components and machinery from
by the US government, which threatened Brazil
outside the NAFTA region at lower taxes
with Super 301 trade sanctions for its market
(Palacios, 2000).
reserve policy. There was also concern over
Brazil likewise removed most barriers to its
high levels of smuggling and gray market
computer market in 1992, but left in place a
activity (Tigre & Botelho, 2001).
complex mix of taris and national and local
While these factors pushed Brazil toward
taxes, which could total over 30% of the cost of a
liberalization, the government did not want to
computer. By maintaining these taxes, Brazil
lose the capabilities developed by domestic
could oer exemptions as a means of promoting
companies under the market reserve. More-
domestic production. The goal was to allow
over, it wanted foreign companies to do more
foreign companies into the market with the
than simply import products to serve the
latest technologies and lower prices, but to give
Brazilian market, especially, given the country's
them an incentive to locate production in Brazil.
balance of payments problems. As a result, the
new policies adopted in 1992 included incen-
tives for companies to produce in Brazil, and (d) IT promotion policies
ensuing programs provided support for local
companies entering the industry. There have been signi®cant dierences in the
Mexico's decision to liberalize the computer degree to which the two countries have
sector was driven as well by the pressures of promoted IT production and use after liberal-
globalization in the industry, but it was ization. Brazil has replaced industry protection
supported by a stronger ideological commit- with an active policy of industry promotion,
ment on the part of Mexico's leaders. Presidents while Mexico has taken a laissez faire
Miguel de la Madrid (1982±88) and Carlos approach.
Salinas (1988±94) were technocrats with a Brazil's post-liberalization IT policies were
neoliberal bent. Salinas opened the computer de®ned in 1991 by Law 8248/91, aimed at
industry to trade and foreign investment in 1990 establishing alternative mechanisms to preserve
without apparent reservations as to the fate of local manufacturing and R&D activities in the
the small local computer industry. Since then, computer sector. The approach taken was to
the prevailing belief, in the words of one federal oer exemptions to various taxes if companies
government ocial, is that ``the best policy is to would commit to certain levels of local
have no policy'' (Dedrick et al., 2001). production, local content and R&D. There
Mexico's unilateral market opening in 1990 were no barriers to imports or foreign invest-
actually brought its trade and investment poli- ment; however, computer makers that simply1206 WORLD DEVELOPMENT imported products for sale had to forego the Investment Board, the foreign investment oce bene®ts provided under the law. of the Ministry of Commerce and Industrial The policy consisted of four types of incen- Promotion (SECOFI), and two development tives. First, ®scal bene®ts consisted of a waiver banksÐBanco Nacional de Comercio Exterior on the industrialized goods tax resulting in a (Bancomext) and Nacional Financiera reduction of 15% in the ®nal cost of production. (NAFIN). In addition, the national electronics, Second, a discount of 50% on income tax for telecommunications and informatics industry R&D expenditures was available to all indus- chamber (CANIETI) has implemented a strat- trial sectors. Third, in order to provide support egy to develop a local supply base for the for new capital investment, a discount of 1% of computer and electronics industries. There is, the income tax payable by companies investing however, little coordination of eorts, limited in IT ®rms was available until 1997. Fourth, funding and no overarching strategy aimed at government procurement policy favored the developing long-term capabilities (Dedrick et acquisition of IT goods developed and produced al., 2001). in Brazil, as long as they had similar prices to To summarize, both Mexico and Brazil imported equipment. By 1997, 248 ®rms had pursued protectionist policies in the computer taken advantage of the program, including most sector in the 1980s, but under internal and major domestic and foreign computer makers. external pressure, both liberalized in the early A program to promote the software industry, 1990s. Mexico opened up more quickly and called Softex, was introduced in 1993. The completely, taking a hands-o approach to trade program includes the formation of regional and investment in the IT sector. Brazil main- centers and incubators to stimulate cooperation tained a set of taris and taxes, for which it could among small software ®rms, the installation of oer exemptions as a tool to promote local marketing oces in foreign markets to support production and R&D. Both countries imple- Brazilian ®rms' export eorts, and provision of mented policies to promote the IT industry, but incentives for training IT professionals within Brazil's were more extensive and better funded. ®rms (Tigre & Botelho, 2001). In both cases, policies were mostly developed ad While laissez faire has been the trend in hoc, with no guiding long-term goals or coordi- Mexico, there have been some scattered nation mechanisms to link production, use and attempts at industry promotion in the 1990s. creation of national capabilities. When Ernesto Zedillo became president in December 1994, he called for the development and exploitation of information technology as a 5. IMPACTS OF LIBERALIZATION national goal. This goal was given form in the Plan for the Development of Informatics Given these dierences in policy, one might (PDI), which targets: (i) promotion of IT use; expect large dierences in impacts. Yet in some (ii) human resource development; (iii) research respects, the impacts of liberalization have been and development; (iv) development of a local similar in the two countries: consumers have IT industry to exploit niche opportunities; (v) bene®ted from lower prices and greater choice in improvement of the telecommunications infra- computer products; IT use has expanded structure; and (vi) creation of the necessary throughout the economy; and many local legal framework to support IT (e.g., intellectual computer makers have been driven out of busi- property rights). ness by foreign competition. There have been Despite its broad objectives, the PDI has had signi®cant dierences as well, partly due to little impact, for two reasons. First, there are no dierences in the liberalization process, but funds set aside to pay for new projects, so more importantly due to dierences in environ- funding must come from the relevant govern- ment such as location, domestic market size, ment agencies' already tight budgets. Second, existing national capabilities, and also to IT there is little coordination among the institu- policies. To analyze these impacts, we look at the tions involved, and no pilot agency to ensure following: domestic computer market, computer that policies are designed to complement each production, trade, IT use, and employment. other. The government also is encouraging devel- (a) Computer market opment of a local supplier base to support the multinationals producing in Mexico. The The impacts of liberalization on the compu- organizations involved are the Mexican ter market in Mexico and Brazil were dierent
LIBERALIZATION AND THE COMPUTER INDUSTRY 1207
in the short-run, as Mexico saw somewhat surpassed its pre-crisis level and continued to
faster growth in computer sales after the liber- grow solidly in 1998 and 1999.
alization of 1990, while Brazil's market at ®rst In retrospect, there is evidence that liberal-
stayed rather ¯at after its 1991 liberalization ization accelerated the rate of IT spending in
(Figure 2). Later, however, Brazil's market each country, although not as dramatically as
began to take o, mainly due to the successful might have been expected. This could be due to
eort to control in¯ation through the Real Plan the fact that both countries had actually taken
in 1994. Once prices were stabilized, businesses steps to push prices downward on domestically-
could make more realistic long-term investment produced computers even before liberalization
decisions and began to invest heavily in infor- by allowing imports of many components at
mation technology. It was only then that the lower tari rates. Thus, IT spending was
impacts of liberalization were fully felt. The already growing in the late 1980s.
market declined in 1999 due to a ®nancial
crunch and devaluation that were triggered by (b) Computer production
the Asian ®nancial crisis.
In Mexico, the rapid expansion of IT Computer production in Brazil ®rst dropped
spending was interrupted by the peso crisis and in response to liberalization, which brought in
deep recession of 1995. Growth resumed, imported computer hardware to compete with
however, in 1996 and by 1997, the market had local production (Figure 3). After a one-time
Figure 2. IT markets for Brazil and Mexico 1985±99. Source: IDC, data provided to authors.
Figure 3. Computer hardware production in Brazil and Mexico, 1989±99. Source: Reed (various years).1208 WORLD DEVELOPMENT
drop in 1992, production began to grow as A similar, but smaller electronics hub is
MNCs took advantage of Brazil's industry found in the state of Campi~ nas, Brazil, near
promotion policies by producing in Brazil for Sao Paulo. Campi~ nas is home to computer
the local market. manufacturers such as Compaq and IBM, as
In Mexico, liberalization initially had little well as a growing number of contract manu-
impact on production. But, when the NAFTA facturers. These companies use Brazil as a
agreements were signed in 1992, production production base for both the domestic market
began to grow steeply as new companies moved and the MERCOSUR free trade area (Botelho
into Mexico and existing foreign producers et al., 1999).
increased production there in anticipation of
NAFTA. The combination of low labor costs, (c) Trade
NAFTA membership, and close proximity to
the large US market made Mexico an attractive The dierences between Brazil and Mexico in
location for assembly of PCs and components the post-liberalization era are most vivid in the
such as printed circuit boards. By 1999, Mexico area of foreign trade (Figure 4). Brazil has seen
had actually equaled Brazil's total output for a steady increase in imports after opening its
computer hardware, after trailing by a 4:1 market, but exports have languished and a
margin in 1991. signi®cant trade gap has opened. Mexico initi-
Production in Mexico has continued to ally saw imports rise quickly, leading to a trade
expand, recently with the in¯ux of contract de®cit in computers, but after the peso deval-
manufacturers (CMs) into the western state of uation of December 1994, that de®cit has
Jalisco. Some of the world's largest CMs have turned into a surplus. The shift was due to a
come to Guadalajara and its metropolitan decline in imports, while exports continued to
region. These include SCI, Solectron, NatSteel, grow as they had ever since NAFTA was signed
Flextronics, Jabil Circuit and Dovatron, which in 1992. In 1997±98, export growth accelerated
joined brand name vendors IBM, Hewlett- even further as production by the newly arrived
Packard, NEC and Lucent, who were already contract manufacturers came on line.
operating in the area. IBM also has continued The dierence between the two countries is
to increase its production of PCs, disk drive that while both have seen substantial growth in
components, and software in its large Guad- computer production, in Brazil it was mostly
alajara facility, which employs around 10,000 for the domestic market, while in Mexico it was
people. This concentration of electronics mostly for export. The reasons have to do with
production has led Guadalajara to be dubbed, trends in the computer industry, geography and
the Mexican Silicon Valley, although a more NAFTA, more than with liberalization of the
accurate comparison would be with manufac- computer sector per se.
turing clusters such as Singapore or Penang, In recent years, the PC industry has become
Malaysia (Palacios, 1995; Dedrick et al., 2001). extremely time sensitive, thanks to the rapid
Figure 4. Trade in computer hardware for Brazil and Mexico, 1989±99. Source: Reed (various years).LIBERALIZATION AND THE COMPUTER INDUSTRY 1209
depreciation of products and the build-to-order as Brazilian-owned ®rms have sustained a
production model pioneered by Dell Computer larger share of their local market than have
(Curry & Kenney, 1999; Kraemer, Dedrick, & their Mexican counterparts. Table 3 shows that
Yamashiro, 2000). This has pushed PC makers six Brazilian-branded PC companies controlled
to locate production close to the market. For 25.6% of the market in 1997, while three
the US market, this often means producing in Mexican-owned ®rms controlled only 12.2% of
North America, and for labor-intensive activi- the market in 1998. Moreover, ``white box''
ties such as assembly of circuit boards, cables, clones made by local assemblers accounted for
connectors and monitors, Mexico is an ideal over half of Brazil's market, compared to 21%
location due to low wages and proximity to the in Mexico. The situation in Brazil may change
United States. NAFTA's provisions helped in the future as PC heavyweights such as Dell
further by creating greater con®dence among and Gateway enter the market, but for the ®rst
foreign multinationals that Mexico's market decade of liberalization, Brazilian companies
liberalization was ®rmly rooted. The result has were better able to survive in the competitive
been a shift of contract manufacturing in PC market.
particular from Asia to Mexico (Goad, 1999). Beyond the PC industry, most Brazilian-
While most of Mexico's production is owned hardware makers who had been
exported, most of Brazil's production is for the producing minicomputers and peripherals
local market. The other MERCOSUL markets either disappeared after liberalization, moved
are so small that they provide limited export into other markets such as services and distri-
opportunities. Compaq is now using Brazil to bution or were bought out by foreign ®rms
supply all of South America, but the whole entering the Brazilian market. In the compo-
continental computer market, including Brazil, nents market, Brazil has lost much of its
is less than one-tenth the size of the US market. production capacity as companies such as
Brazil cannot easily export to the United States Itautec have abandoned the components busi-
because its industry centers are too far from the ness and scaled back R&D in the face of
United States to ship time-sensitive products by international competition. Brazilian ®rms have,
truck or sea. Once air shipment is required, however, been able to sustain a larger role in
Brazil has no advantage over Mexico and large niche segments of the IT industry, especially in
Asian producers such as Taiwan, Malaysia, banking automation, where Brazil had been an
China and Singapore in serving the US market. early adopter as banks responded to the
demands of a high-in¯ation environment (Tigre
(d) Industry structure & Botelho, 2001).
Few Mexican-owned ®rms have entered the
Since liberalization, both countries' IT components market, although a small Guad-
industries have become increasingly dominated alajara ®rm, Compuworld has become a
by foreign (mainly US) ®rms. There is a supplier of disk drive components to IBM.
signi®cant dierence between the two countries, Otherwise, computer makers and contract
Table 3. PC market share for Brazil and Mexicoa
Brazil, 1997 Mexico, 1998
Company Market share (in %) Company Market share (in %)
Compaq (US) 10.4 Compaq (US) 20.0
Itautec (Brazil) 6.8 Acer (Taiwan) 11.3
IBM Brasil (US) 5.6 IBM (US) 12.8
UIS (Brazil) 4.9 HP (US) 10.2
Tropcom (Brazil) 4.7 Alaska (Mexico) 8.9
Byte On (Brazil) 3.4 Dell (US) 3.8
Hewlett-Packard (US) 3.1 Lanix (Mexico) 2.9
Microtec (Brazil) 2.9 Toshiba (Japan) 2.9
Fivestar (Brazil) 2.9 Apple (US) 2.3
Acer (Taiwan) 2.5 Printaform (Mexico) 0.4
Other int'l brands 2.8
Others (mostly white boxes) 52.8 White boxes 21.3
a
Source: Brazil, IDC cited in Crespo (1998). Mexico, IDC provided to authors.1210 WORLD DEVELOPMENT
manufacturers mostly import components such systems or have worked with foreign and
under either the maquiladora or PITEX domestic IT companies to develop their own
(Program of Temporal Importation for Export) capabilities.
regimes that allow for duty free import as long After a slow start, the Internet has caught on
as the ®nal products are exported (within up to in both countries in recent years. One reason
60 days). has been the availability of lower cost PCs to
On the other hand, local Mexican ®rms have access the Internet. Another has been the
been more successful in custom programming liberalization of telecommunications in each
and information services, led by Softek, based country, leading to lower prices and improve-
in Monterrey, which had over 2,000 employees ments in infrastructure. The growth in number
and over US$50 million in sales in 1997. Some of Internet users has supported an increase in
of the leaders in this segment have been taken Spanish and Portuguese language content,
over by US giants such as IBM (which bought which has in turn drawn more users.
software developer TecnoSys) and GE Capital
(which bought 80% of system integrator EDM). (f) Summary and analysis
(e) IT use In summary, liberalization has led to lower
prices and greater diusion of computer tech-
Falling prices and a wider array of available nologies in both countries. It also has created a
products and services helped expand IT use better environment for foreign investment,
throughout both countries' economies. As of which has brought with it leading edge tech-
1998, Brazil was slightly ahead of Mexico in nology and knowledge of how to use the tech-
terms of PCs per capita, and both were well nology more eectively. On the other hand,
ahead of poorer developing countries such as while both countries have seen growth in
China and India. But, compared to wealthier computer production, the growth has been
Asian economies such as Korea and Taiwan, much faster in Mexico. While Brazil's industry
Brazil and Mexico still lagged with less than produces almost exclusively for the domestic
one-third the number of PCs per 100 inhabit- market, Mexico has become a major export
ants (Table 4). Brazil compares favorably to platform for the North American market.
other developing countries in total IT spending Returning to the conceptual framework
as a percentage of GDP, while Mexico is about developed earlier in this paper, a comparison of
average. Brazil and Mexico shows that dierences in the
IT use in both countries has been spurred as impacts of liberalization in the two countries
well by increased competition across large can be related back to the twin factors of
segments of the economy. As foreign banks, environment and policy.
retailers and manufacturers have entered
previously closed markets, they have brought in (g) Impact of environmental factors
advanced information systems and exposed
Brazilian and Mexican managers to those ÐIdeology and political dynamics. In Mexi-
technologies. Domestic companies have either co, powerful Presidents with a strong neolib-
looked for foreign partners to gain access to eral ideology have implemented more rapid,
Table 4. IT use indicatorsa
Country PCs (millions) PCs/100 IT spending as % Internet users/ Hosts per 1,000
1998 inhabitants 1998 of GDP ('98) 1,000 people, 1998 people, 2000
Brazil 5 3.0 1.6 15 3
Mexico 4.5 4.7 1.1 14 4
USA 124 45.6 4.2 220 193
Japan 30 23.8 2.3 132 20
China 11.2 0.9 1.0 2 0.05
India 2.7 0.3 0.5 1 0.02
Malaysia 1.3 6.2 1.7 36 3
Korea 7.2 15.7 1.5 66 6
Taiwan 3.3 15.9 1.1 142 28
a
Source: ITU (2000) (PCs and Internet users); IDC (IT spending); Internet Software Consortium (2000) (hosts).LIBERALIZATION AND THE COMPUTER INDUSTRY 1211 wide ranging liberalization programs with ÐIndustry trends. The PC industry's trend little eort to protect the domestic computer toward increased inventory turnover and industry from foreign competition. Brazil's build-to-order production has favored Mex- more pluralistic government has taken a ico as a production location for North more gradual approach to liberalization, America, and has attracted production to maintaining some protection and incentives Brazil for South America. While Mexico for local industry. This is also because the clearly bene®ts more due to the size of the computer industry in Brazil has greater North American market, Brazil should enjoy political in¯uence due to the involvement increased production as the South American of large banks, industry groups, universities market, particularly within MERCOSUR, and the military. By contrast, Mexico's grows. domestic computer companies are mostly small independent startups with little politi- (h) Impacts of policy cal clout. ÐInternational political environment. Pres- ÐPre-liberalization policy dierences. Bra- sure from foreign governments has had an zil's market reserve policies put greater impact on liberalization and IT policy for emphasis on promoting local ownership both countries. Brazil faced strong US oppo- and domestic content than Mexico's sition to its market reserve policies and Computing Program, so by the time of liber- decided to liberalize at least partly in re- alization, there were a number of vertically- sponse to that pressure. Mexico liberalized integrated computer ®rms in Brazil making unilaterally in advance of negotiations to a range of systems and components. Some join the GATT and later to join NAFTA. of these ®rms survived in certain segments Mexico also has been limited in its ability of the systems market (such as PCs), while to oer investment incentives that other others shifted to software and services. By countries commonly provide due to two fac- contrast, Mexico failed to develop strong tors: (i) concerns that Mexico can be seen as locally-owned computer makers, or even trying to lure jobs south of the border, espe- suppliers to MNCs, during the 1980s. After cially since NAFTA has been in eect; and liberalization, foreign companies grabbed (ii) congruence with the Salinas and Zedillo market share in sectors that had been closed administrations' decided advocacy of free to them previously. On the other hand, Mex- market ideology which required the govern- ico's maquiladora program had helped at- ment to follow a hands-o policy orienta- tract foreign investment in electronics tion. As a result, Mexico was passed over assembly, helping it to develop skills and by Intel when it decided to locate an assem- experience that attracted computer makers bly plant in Costa Rica instead of Guadalaj- and contract manufacturers. ara, given the incentives the company was ÐDierences in the pace of liberalization and oered by the Costa Rican government. remaining internal barriers. The continuing ÐSize and location. Brazil's larger domestic presence of various taxes, slow customs market and distance from major IT markets processing and other costs (referred to as have encouraged a greater focus on produc- the ``Brazil cost'') in Brazil discouraged many tion for the local market. It has also sup- foreign ®rms from entering the market until ported more local entrepreneurship in recently. These costs also make Brazil less market segments such as banking automa- competitive as a producer for global markets. tion, software and Internet services and con- On the other hand, domestic companies have tent. On the other hand, Brazil has run up a been given time to adjust, often by switching large trade de®cit in IT as its imports have from hardware production to other industry grown, but it has failed to grow exports. By segments. By contrast, Mexico's liberaliza- contrast, Mexico's proximity to the United tion was more sudden and comprehensive, States and membership in NAFTA have and foreign companies quickly ¯ooded into made it more of an export platform for multi- the Mexican market. This competition has nationals. It also has been easier to serve the driven local companies out of the market Mexican market with products from the and possibly discouraged local entrepreneurs United States, so local entrepreneurs have from entering the IT industry. had more limited opportunities and less ÐDierences in complementary IT policies. incentives to enter the PC making business. IT policies have had an impact, but less so
1212 WORLD DEVELOPMENT
than environmental factors. Brazil's various perhaps inevitable process for both Brazil
tax incentives have helped increase local and Mexico, given the globalizing trends in
production by foreign MNCs selling to Bra- the computer industry in the early 1990s.
zil and MERCOSUR, and have begun to at- Prior to liberalization, IT users in Brazil
tract contract manufacturers. Perhaps more and Mexico were paying relatively high
important have been programs to promote prices for computers, and local computer
the software industry and develop Internet makers were not competitive outside their
infrastructure. These eorts go beyond protected domestic market. India's similar
promotion of commodity hardware produc- eorts to develop a domestic computer
tion and move Brazil toward the most dy- industry behind protective barriers was even
namic segments of the IT industry. Mexico, less successful, and Korea's ban on PC im-
by contrast, has done little to provide ports in the 1980s isolated it from global
government support to promote the devel- markets (Dedrick & Kraemer, 1998). For
opment of software and Internet services, Brazil and Mexico, the question was no
sectors with a large potential market of longer whether to liberalize, but rather when
Spanish speakers in Latin America and the and how to do so.
United States. It also has failed to oer ÐLiberalization has had mixed results in
incentives to compete with other countries both countries. It has been bene®cial in
for foreign investment. Industry executives terms of reducing the cost of computers
have stated in interviews with the authors and expanding diusion of IT throughout
that Mexico could attract much more invest- the economy. Both countries also have seen
ment in components production if it oered growth in production and have attracted
incentives comparable to those of Asian investment from leading multinational com-
and European countries. puter makers. On the other hand, there have
ÐDevelopment of capabilities. Both coun- been tradeos, particularly, the demise of
tries have made important gains in quality domestic ®rms in the face of foreign compe-
of telecommunications infrastructure as a re- tition, and in Brazil's case, an expanding
sult of privatization, deregulation and intro- trade de®cit in computers. So while liberal-
ducing competition into the sector. Brazil ization has lived up to many of the promises
has done more to help entrepreneurs gain of its advocates, it also has led to painful
the knowledge and resources they need dislocations as feared by skeptics.
through programs such as Softex. Mexico ÐNational environments do make a dier-
has done little to develop such resources to ence. Brazil and Mexico took somewhat
support local entrepreneurs. It has, however, dierent approaches to liberalization, and
done a good job of meeting the human re- these variations did aect outcomes, but ulti-
source needs of the growing IT industry in mately it was dierences in size and geo-
places such as Guadalajara, mainly through graphic location that had a bigger impact.
the eorts of local government and educa- Mexico's proximity and close economic ties
tional institutions. to the United States were enhanced by liber-
alization, and resulted in its becoming an ex-
port platform for the US market. Brazil's
6. CONCLUSIONS AND IMPLICATIONS larger domestic market and its distance from
major world markets led to the creation of an
While a comparison of two countries is not industry oriented to serving local and regio-
sucient to draw broad conclusions, the cases nal demand, with little export production.
of Mexico and Brazil do provide insights into ÐGovernment can play a positive role be-
the ®ve questions raised earlier in this paper. yond simply liberalizing and taking a
Based on those questions, we draw the follow- hands-o approach to the industry. Policies
ing conclusions: that promote IT production and use can suc-
ÐThe idea of promoting a domestic compu- ceed if they take into account both global
ter industry behind protective barriers was market forces and national environment. It
not tenable in an industry marked by rapid is possible to learn from the experiences of
technological change and dominated by others, but those lessons must be applied
multinational companies who set and con- with a clear understanding of local market
trol global technology standards. We would conditions, national capabilities and chang-
argue that liberalization was a necessary and ing global forces. What worked for MexicoLIBERALIZATION AND THE COMPUTER INDUSTRY 1213
will not necessarily work for other countries national capabilities such as human re-
located far from major markets, and what sources, high-quality and low-cost telecom-
worked for Brazil will not work for countries munications networks and Internet
with very small domestic markets. connections, as well as ®nancial systems
ÐA comprehensive policy approach to the capable of supporting local entrepreneurs
computer sector needs to go beyond the de- (see Dedrick & Kraemer, 1998; Tallon &
bate over whether to liberalize, as that pro- Kraemer, 2000). By doing so, a country
cess is already far advanced in most can take advantage of opportunities in the
developing countries. Liberalization is now global market while developing national
seen as a ®rst step to ensure that countries applications of information technology.
have access to international markets, tech- These countries also show the value of
nology and foreign capital. Beyond that, having a coordinated policy approach that
the issue is how to realize the potential bene- explicitly links computer production, use,
®ts from increased competition at home and and the creation of capabilities, supported
from participation in global markets and by a clear understanding of global market
global production networks. What has been and technology factors. Such a coordinated
made clear by the mixed experiences of Bra- approach can help a country adjust to
zil and Mexico, and the success of countries the dislocations caused by liberalization,
such as Taiwan, Singapore and Ireland, is and to take advantage of the opportunities
that countries need to focus on developing created.
REFERENCES
Amsden, A. H. (1989). Asia's next giant: South Korea States and the world? New York: Oxford University
and late industrialization. New York: Oxford Univer- Press.
sity Press. Dedrick, J., Kraemer, K. L., & Palacios, J. J. (2001).
Anchordoguy, M. (1989). Computers, Inc., Japan's Impacts of liberalization and economic integration
challenge to IBM. Cambridge, MA: Harvard Univer- on Mexico's computer sector. The Information Soci-
sity Press. ety, 17(2).
Borja, A. (1992). The state and industrial development in Ernst, D. (1994). Carriers of regionalization: the East
the NICs. The Mexican computer industry in compar- Asian production networks of Japanese electronics
ative perspective. Unpublished Ph.D. dissertation. ®rms. Berkeley Roundtable on the International
Duke University: Department of Political Science. Economic, Working paper 73, Berkeley: University
Borrus, M. (1997). Left for dead: Asian production of California.
networks and the revival of US electronics. In: B. Evans, P. (1992). Greenhouses and strategic nationalism:
Naughton (Ed.), The China circle: economics and A comparative analysis of Brazil's informatics
electronics in the PRC, Hong Kong and Taiwan. policy. In: P. Evans, C. Frischtak, & P. B. Tigre
Washington, DC: Brookings Institution. (Eds.), High technology and third world industrializa-
Botelho, A. J. J., Dedrick, J., Kraemer, K. L., & Tigre, tion: Brazilian computer policy in comparative
P. B. (1999). From industry protection to industry perspective (pp. 1±37). Berkeley, CA: IAS/University
promotion IT policy in Brazil. University of Califor- of California at Berkeley.
nia, Irvine: Center for Research on Information Friedman, M., & Friedman, R. (1980). Free to choose: A
Technology and Organizations http://www.crito.u- personal statement. New York: Harcourt, Brace and
ci.edu/publications/pdf-ppt/brazil-case-10-99.pdf. Jovanovich.
Crespo, R. (1998). Fabricantes nacionais disputam Goad, P. (1999). Mexican wave. Far Eastern Economic
mercado de PC. Gazeta Mercantil, 23(C-3). Review, 162(45), 10±13.
CONACyT (1995). (Consejo Nacional de Ciencia y Grove, A. S. (1996). Only the paranoid survive: How to
Tecnologia) Indicators of scienti®c and technological exploit the crisis points that challenge every company
activities. Mexico City: CONACyT. and career. New York: Doubleday.
Curry, J., & Kenney, M. (1999). Beating the clock: Internet Software Consortium. (2000). http://www.
Corporate responses to rapid change in the PC isc.org/ds/WWW-200001/dist-bynum.html.
industry. California Management Review, 42(1), 8±36. ITU (2000). Yearbook of statistics. Geneva: ITU.
Dedrick, J., & Kraemer, K. L. (1993). India's quest for Jones, C. (1998). The year 2000 software problem:
self-reliance in information technology: costs and Quantifying the costs and assessing the consequences.
bene®ts of government intervention. Asian Survey, New York: Addison-Wesley.
33(5), 463±492. Kraemer, K. L., & Dedrick, J. (1995). From nationalism
Dedrick, J., & Kraemer, K. L. (1998). Asia's computer to pragmatism: IT policy in China. IEEE Computer,
challenge: Threat or opportunity for the United 28(8), 64±73.You can also read
