M
CHAPTER 6
Trade and Development I: Import
Substitution Industrialization
exico has experienced an economic revolution during the last 20
years. Until the mid-1980s, Mexico was one of the most heavily
protected and highly directed nonsocialist economies in the world.
Importing anything into the country required formal government approval.
Even with such approval, tariffs were very high, averaging over 25 percent
and rising as high as 100 percent for many goods. Moreover, Mexico did
not belong to the General Agreement on Tariffs and Trade (GATT), and it
was hard to imagine any conditions under which Mexico would seek a
free-trade agreement with the United States. Behind these high tariff walls,
the Mexican government intervened deeply in the domestic economy.
Government-owned financial institutions channeled investment capital to
favored private industries and projects. The government created stateowned
enterprises in many sectors of the economy (about 1,200 of them
by 1982) that together attracted more than one-third of all industrial
investment (La Porta and López de Silanes 1997). Today, by contrast,
Mexico is one of the most open developing countries in the world. Mexico
entered the GATT in 1987 and the North American Free Trade Agreement
(NAFTA) in the early 1990s. The Mexican government has retreated
sharply from involvement in the domestic economy. It has sold stateowned
enterprises, liberalized a wide variety of market-restricting
regulations, and begun to integrate Mexico deeply into the global
economy. In less than 10 years, the Mexican government opened Mexico
to foreign competition and drastically scaled back its role in managing
Mexican economic activity.
Mexico’s experience is hardly unique. Governments in India, China,
much of Latin America, and most of sub-Saharan Africa opted out of the
164
global trade system following World War II. Most governments erected
very high trade barriers, and to the extent that they participated at all in the
GATT, they sought to alter the rules governing international trade.
Convinced that the GATT was biased against their interests, developing
countries worked through the United Nations to create international trade
rules that they believed would be more favorable toward industrialization
in the developing world. Like Mexico, most governments intervened
extensively in their economies in an attempt to promote rapid
industrialization. Drawing on the logic of the infant-industry case for
protection, governments used the power of the state to pull resources out
of agriculture and push them into manufacturing. And, like Mexico, these
policy orientations have changed fundamentally since the late 1980s. Most
developing countries have dismantled the protectionist systems they
maintained in the first 30 years of the postwar period, have become active
participants in the World Trade Organization (WTO), and have abandoned
the quest to institute far-reaching changes to international trade rules. Most
have greatly reduced the degree of government intervention in the
domestic economy.
This chapter and the next examine how political and economic forces
have shaped the adoption and evolution of these new trade and
development policies. This chapter examines why governments in so many
developing countries intervened deeply in their domestic economies,
insulated themselves from international trade, and sought changes in
international trade rules. The next chapter focuses on why so many
governments have dismantled these policies during the last 30 years. We
look first at how economic and political change throughout the developing
world brought to power governments supported by import-competing
interests. We then examine the economic theory that guided policy during
those times. As we shall see, this theory provided governments with a
compelling justification for transforming the protectionism sought by the
import-competing producers that supported them into policies that
emphasized industrialization through state leadership. Having built this
base, we turn our attention to the specific policies that governments
pursued during that period, looking first at their domestic strategy for
industrialization and then examining their efforts to reform the
international trade system.
DOMESTIC INTERESTS, INTERNATIONAL
PRESSURES, AND PROTECTIONIST
165
COALITIONS
Developing countries’ trade policies underwent a sea change in the first
half of the twentieth century. Until World War I, those developing
countries that were independent, as well as those regions of the world held
in colonial empires, adopted liberal trade policies. They produced and
exported agricultural goods and other primary commodities to the
advanced industrialized countries and imported most of the manufactured
goods they consumed. Governments and colonial rulers made little effort
to restrict this trade. But by the late 1950s, these liberal trade policies had
been replaced by a protectionist approach that dominated the developing
countries’ trade policies until the late 1980s, and whose remnants remain
important in many countries today. We begin our investigation of
developing countries’ trade and development policies by looking at this
initial shift to protectionism.
Trade and development policies in developing countries have been
strongly shaped by political competition between rural-based agriculture
and urban-based manufacturing. Developing countries pursued liberal
trade policies prior to World War I because export-oriented agricultural
interests dominated politics. In general, developing countries are
abundantly endowed with land and poorly endowed with capital (Lal and
Myint 1996, 104–110).
The relative importance of land and capital in developing countries’
economies can be appreciated by examining the structure of those
economies, together with exports, as presented in Table 6.1 and Table 6.2.
For the time being, we will focus on 1960, as this will allow us to put to
the side the consequences of the development policies that governments
adopted during the postwar period. With a few exceptions (particularly in
Latin America), between one-third and one-half of all economic activity in
developing countries in 1960 was based in agriculture, whereas less than
15 percent was based in manufacturing. By contrast, agriculture accounted
for only 5 percent of gross domestic product (GDP) in the advanced
industrial economies. If we include the “other industry” category, which
incorporates mining, then in all regions of the developing world other than
Latin America, agriculture and nonmanufacturing industries accounted for
more than half of all economic activity.
A similar pattern is evident in the commodity composition of
developing countries’ exports (Table 6.2). In 1962, developing countries’
exports were heavily concentrated in primary commodities: agricultural
products, minerals, and other raw materials. Roughly speaking, in each
166
developing country, primary commodities accounted for more than 50
percent of exports, and in more than half of the listed countries, primary
commodities accounted for more than 80 percent of exports. In addition,
each country exported a narrow range of primary commodities. Some
countries were monoexporters; that is, their exports were almost fully
accounted for by one product. For example, more than 80 percent of
Burundi’s export earnings came from coffee, and cocoa accounted for 75
percent of Ghana’s export earnings (Cypher and Dietz 1997, 339). Similar
patterns were evident in Latin America: in 1950, coffee and cocoa made
up about 69 percent of Brazil’s exports, and copper and nitrates constituted
about 74 percent of Chile’s exports (Thorp 1999, 346). The structure of
their economies and the composition of their exports thus underline the
central point: developing countries are abundantly endowed with land and
have little capital.
TABLE 6.1
Economic Structure in Developing Countries (Sector as
a Percent of Gross Domestic Product)
Notes: Figures may not sum to 100 because of rounding.
Other Industry Includes mining, construction, gas, and water.
Sources: Data for 1960 from World Bank, World Tables, 3rd ed. (Washington, DC:
The World Bank, 1983). Data for 1980 and 1995 from World Bank, World
Development Indicators (Washington, DC: The World Bank, 1997).
TABLE 6.2
Developing Countries’ Export Composition (Sector as a
167
Percent of Total Exports)
Note: n.a. = not available.
Sources: Data for 1962 from World Bank, World Tables, 3rd ed. (Washington, DC:
The World Bank, 1983). Data for 1980 and 1993 from World Bank, World
Development Indicators (Washington, DC: The World Bank, 1997).
The precise form through which landowners dominated politics prior to
World War II differed considerably across regions. In Latin America, an
indigenous landowning elite dominated domestic politics. In Argentina
and Chile, for example, the landowners controlled government, often in an
alliance with the military. Even though these political systems were
constitutionally democratic, participation was restricted to the elite, a
group that amounted to about 5 percent of the population, in a system that
has been characterized as “oligarchic democracy” (Skidmore and Smith
168
1989, 47). In other Latin American countries such as Mexico, Venezuela,
and Peru, dictatorial and often military governments ruled, but they
pursued policies that protected the interests of the landowners (Skidmore
and Smith 1989, 47). With landowners dominating domestic politics, Latin
American governments pursued liberal trade policies that favored
agricultural production and export at the expense of manufactured goods
(Rogowski 1989, 47). As a result, most Latin American countries were
highly open to international trade, producing and exporting agricultural
goods and other primary commodities and importing manufactured goods
from Great Britain, Europe, and the United States.
In Asia and in Africa, export-oriented agricultural interests dominated
local politics through colonial structures. In Taiwan and Korea, for
example, Japanese colonization led to the development of enclave
agriculture—that is, export-oriented agricultural sectors that had few
linkages to other parts of the local economy (Haggard 1990). Agricultural
producers bought little from local suppliers and exported most of their
production. In both countries, agricultural production centered on the
production and export of rice; in Taiwan, sugarcane was a staple crop as
well. India produced and exported a range of primary commodities,
including cotton, jute, wheat, tea, and rice. In exchange, India imported
most of the manufactured goods it consumed from Britain. In Africa,
colonial powers encouraged the production of cash crops and raw
materials that could be exported to the mother country (Hopkins 1979;
Ake 1981, 1996). In the Gold Coast (now Ghana), the cocoa industry was
a small part of the economy in 1870. Under British rule, Ghana became the
world’s largest cocoa producer by 1910, and cocoa accounted for 80
percent of its exports. In Senegal, France promoted groundnut (the
American peanut) production, and by 1937 close to half of all cultivated
land was dedicated to this single product (Ka and Van de Walle 1994,
296). Similar patterns with other commodities were evident in other
African colonies (Hopkins 1979).
These political arrangements began to change in the early twentieth
century. As they did, the dominance of export-oriented interests gave way
to the interests of import-competing manufacturers. In many instances, the
most important triggers for this change originated outside of developing
societies. In Latin America, international economic shocks beginning with
the First World War and extending into World War II played a central role
(Thorp 1999, Chapter 4). Government-mandated rationing of goods and
primary commodities in the United States and Europe during the two
World Wars made it difficult for Latin American countries to import many
169
of the consumer goods they had previously purchased from the
industrialized countries. In addition, falling commodity prices associated
with the Great Depression and the disruption of normal trade patterns
arising from World War II reduced export revenues. The interruption of
“normal” Latin American trade patterns led governments in many
countries to introduce trade barriers and to begin producing many of the
manufactured goods that they had previously imported. The rise of
domestic manufacturing in turn produced a growing urban middle class as
workers and industrialists began to move out of agricultural production
and into manufacturing industries.
The emergence of manufacturing industries gave rise to interest groups,
industry-based associations, and labor unions that pressured the
government to adopt economic policies favorable to people working in the
import-competing sector. The creation of organized groups to represent the
interests of import-competing manufacturing generated its own political
logic. On the one hand, the groups that saw their incomes rise from
protection had a strong incentive to see protectionist policies continued in
the postwar period (see Rogowski 1989; Haggard 1990). On the other
hand, the emergence of new organized interests and a growing urban
middle class created an opportunity for politicians to construct new
political coalitions based on the support of the urban sectors. In Argentina,
for example, Juan Perón rose to power in the late 1940s with the support of
labor, industrialists, and the military. A similar pattern was evident in
Brazil, where Getúlio Vargas was elected to the presidency in 1950 with
the support of industrialists, government civil servants, and urban labor.
Nor were Argentina and Brazil unique: throughout Latin America, postwar
governments were much less tightly linked to landed interests than
governments had been before World War I. Instead, governments rose to
power on the basis of political support from interest groups whose incomes
were derived from import-competing manufacturing (Cardoso and Faletto
1979). Such governments had a clear incentive to maintain trade policies
that protected those incomes.
A similar dynamic is evident in India. The global economic collapse of
the 1930s forced India to become increasingly self-reliant. Markets for
Indian exports constricted sharply, thereby greatly constraining Indian
export revenues. Unable to earn foreign exchange, India had to reduce
imports of manufactured goods as well. Under this forced self-reliance,
India began to create an indigenous manufacturing sector. By the end of
World War II, India had emerged as “the tenth largest producer of
manufactured goods in the world” (Tomlinson 1979, 31). The indigenous
170
urban manufacturing sector then fused with the burgeoning nationalist
movement during the late 1930s to lead the push for Indian independence
and to supplant the predominantly foreign-owned export sector at the
center of the Indian political system. By the time India achieved
independence in 1947, it was committed to a strategy of autonomous
industrialization.
In Pacific Asia, the shift in political power came about as a product of
de-colonization. In Korea and Taiwan political change resulted from the
defeat of Imperial Japan in World War II (see Haggard 1990). In South
Korea, Japan’s defeat transferred power from a foreign colonizer to
indigenous groups. Although the landowners initially dominated postwar
politics, the Korean War of the early 1950s and a series of land reforms
implemented during that same decade greatly reduced the landowners’
power and increased the relative power of the emerging urban sector. On
mainland China, Japan’s defeat was followed by the defeat of the
nationalist Chinese government and the migration of the Chinese
nationalists to the island of Taiwan. Once installed in Taiwan, the Chinese
nationalists instituted land reforms to assert their authority over indigenous
landowners and to prevent a repeat of their experience on the mainland,
where the rural sector had supported the Communists. As in South Korea,
land reforms reduced the power of landowners and increased the power of
the urban–industrial sector.
Africa’s transition came later, as decolonization began only in the
1950s, and it took a slightly different form. The push toward
decolonization was led by a coalition of indigenous professionals who had
been educated by the colonial powers and had then acquired positions in
the administration of colonial economic and political rule. One factor
motivating Africa’s push for independence was dissatisfaction with the
discriminatory practices of colonial administration. Colonies were run for
the profit of the colonists, with colonial economic enterprises staffed and
managed by men from the colonial power. The local population had
limited opportunities to participate in these economic arrangements other
than as workers. The nationalist struggles for independence that emerged
in the 1950s sought to transfer control over existing economic practices
from the colonial governments to indigenous elites.
The period demarcated by the start of World War I and the end of
decolonization in sub-Saharan Africa thus brought a fundamental change
to patterns of political influence in developing countries. Political
structures once dominated by export-oriented agricultural interests were
now largely under the control of import-competing manufacturing
171
interests. Consequently, governments beholden to the import-competing
sector had a clear incentive to abandon liberal trade policies and to
continue the protectionist arrangements they had built during the 1930s.
As we will see, the political interest in protectionism was reinforced by an
elaborate theoretical structure that argued that protectionism was the only
path to the establishment of industrialized economies.
THE STRUCTURALIST CRITIQUE: MARKETS,
TRADE, AND ECONOMIC DEVELOPMENT
Although protectionism reflected the interests of the politically influential
import-competing manufacturing sector, it did not represent a coherent
economic development strategy. And most governments were committed,
at least rhetorically, to the adoption of policies that would promote
economic development. Most governments wanted to shift resources out of
agricultural production and into manufacturing industries because they
believed that poverty resulted from too heavy a concentration on
agricultural production. Higher standards of living could be achieved only
through industrialization, and according to what was then the dominant
branch of development economics, called structuralism, the shift of
resources from agriculture to manufacturing would not occur unless the
state adopted policies to bring it about (see Lal 1983; Little 1982).
The belief that the market would not promote industrialization provided
the intellectual and theoretical justification for the two central aspects of
the development strategies adopted by most governments throughout much
of the postwar era. Because structuralism played such an important role in
shaping developing countries’ trade and development policies,
understanding the policies governments adopted requires us to understand
the structuralist critique.
Market Imperfections in Developing Countries
Structuralists argued that market imperfections inside developing countries
posed serious obstacles to the reallocation of resources from agriculture to
manufacturing industries. Structuralists argued that markets would not
bring about the necessary shift of resources because developing economies
were too inflexible.
Most important, according to the structuralists, was the belief that the
market would not promote investment in manufacturing industries
(Scitovsky 1954). The structuralists pointed to two coordination problems
172
that would limit investment in manufacturing industries. The first problem,
called complementary demand, arose in the initial transformation from
an economy based largely on subsistence agriculture to a manufacturing
economy (Rosenstein-Rodan 1943). In an economy in which few people
earned a money wage, no single manufacturing firm would be able to sell
its products unless a large number of other manufacturing industries were
started simultaneously. Suppose, for example, that 100 people are taken
out of subsistence agriculture and paid a wage to manufacture shoes,
whereas the rest of the population remains in non-wage agriculture. To
whom will the new factory sell its shoes? The only workers earning money
are those producing shoes, and these 100 workers are unlikely to purchase
all of the shoes that they make. In order for this shoe factory to succeed,
other factories employing other people must be created at the same time.
Suppose instead, that 500,000 workers are taken out of subsistence
agriculture and simultaneously employed in a large number of factories
producing a variety of different goods; some make shoes, others make
clothing, and still others produce refrigerators or processed foods. With
this larger number of wage earners, manufacturing enterprises can easily
sell their goods. Shoe workers can buy refrigerators and clothes, workers
in the clothing factory can purchase shoes, and so on. Thus, a
manufacturing enterprise will be successful only if many manufacturing
industries began production simultaneously.
Structuralists doubted that uncoordinated market behavior would
produce simultaneous investment in multiple manufacturing industries. No
single entrepreneur has an incentive to invest in a manufacturing enterprise
unless he or she is certain that others will invest simultaneously in other
industries. People willing to invest will thus wait until others invest and, as
a consequence, no one will invest in manufacturing unless all potential
investors could somehow coordinate their behavior to ensure that all will
invest in manufacturing at the same time. The problem of complementary
demand thus meant that if investment were left to the market, there would
be little investment in manufacturing industries.
The second coordination problem, called pecuniary external
economies, arose from interdependencies among market processes
(Scitovsky 1954). Think about the economic relationship between a steel
plant and an automobile factory. Suppose that the owners of a steel factory
invest to increase the amount of steel they can produce. As steel
production increases, steel prices begin to fall. The automobile factory,
which uses a lot of steel, begins to realize rising profits as the price of one
of its most important inputs falls. These increasing profits in the
173
automobile industry could induce the owners of the car plant to invest to
expand their own production capacity. Such a simultaneous expansion of
the steel and auto industries would raise national income.
The two firms face a coordination problem, however. The owners of the
steel plant will not increase steel production unless they are sure that the
auto industry will increase car production. Yet, the owners of the auto
plant will not increase auto production unless they are certain that the steel
producer will make the investments needed to expand steel output. Thus,
unless investment decisions in the steel and auto industry are coordinated,
neither firm will invest to increase the amount it can produce. Once again,
structuralists argued, the market could not be expected to solve this
coordination problem.
The structuralists’ assertion that coordination problems would prevent
investment in manufacturing was a serious problem for governments intent
on industrialization. Fortunately, the structuralists offered a solution to the
problem. Structuralists argued that the way to overcome these coordination
problems was with a state-led big push. The state would engage in
economic planning and either make necessary investments itself or help
coordinate the investments of private economic actors. Thus, what the
market could not bring about, the state could achieve through intervening
in the economy. The structuralist critique of the market therefore provided
a compelling theoretical justification for state-led strategies of
industrialization.
Market Imperfections in the International Economy
Structuralists also argued that international trade provided few benefits to
developing countries. This argument was formulated during the 1950s,
principally by Raul Prebisch, an Argentinean economist who worked for
the United Nations Economic Commission for Latin America (ECLA), and
Hans Singer, an academic development economist. According to the
Singer-Prebisch theory, participation in the GATT–based trade system
would actually make it harder for developing countries to industrialize by
depriving them of critical resources.
The Singer-Prebisch theory divides the world into two distinct blocks—
the advanced-industrialized core and the developing-world periphery—and
focuses on the terms of trade between them. The terms of trade relate the
price of a country’s exports to the price of its imports. An improvement in
a country’s terms of trade means that the price of its exports is rising
relative to the price of its imports, but a decline in a country’s terms of
174
trade means that export prices are falling relative to its import prices. As a
country’s terms of trade improve, it can acquire a given amount of imports
for a smaller quantity of exports. Thus, an improvement in its terms of
trade makes a country richer, but a decline in its terms of trade makes it
poorer.
The Singer-Prebisch theory argues that developing countries’ terms of
trade deteriorate steadily over time. When they developed this theory,
developing countries exported primary commodities and imported
manufactured goods. Singer and Prebisch argued that primary commodity
prices steadily fell relative to manufactured goods prices, thereby steadily
reducing the incomes of developing countries. The periphery’s terms of
trade deteriorate, according to this theory, in large part as a result of
differences in the income elasticity of demand for primary commodities
versus industrial goods (see Lewis 1954; United Nations 1964; Gilpin
1987, 275–276).
The income elasticity of demand is the degree to which a change in
income alters demand for a particular product. For a product with a low
income elasticity of demand, a large increase in income produces little
change in demand for the good. For a product with a high income elasticity
of demand, a small increase in income produces a large change in demand
for a particular good. Structuralists argued that the income elasticity of
demand for primary commodities was quite low, but income elasticity of
demand for manufactured goods was relatively high. Thus, as incomes rise
in the core countries, a smaller and smaller percentage of those countries’
income will be spent on imports of primary commodities. But as incomes
rise in the periphery countries, a larger percentage of those countries’
income will be spent on manufactured imports from the core. Falling
demand for primary commodities will cause the periphery countries’
export prices to fall, whereas rising demand for manufactured goods will
cause the periphery countries’ import prices to rise. Rising import prices
relative to export prices yields deteriorating terms of trade.
Most research disputes the claim that developing countries face a
continuous decline in their terms of trade (see, for example, Borensztein et
al. 1994; see also Bloch and Sapsford 2000). Yet, the objective validity of
the Singer-Prebisch hypothesis is not the central consideration. What
mattered was that governments in developing countries believed the
hypothesis. Governments of developing countries were convinced that
industrialization would not occur if they participated in the GATT–based
international trade system. This conviction played an important role in
shaping the trade and development policies that developing countries
175
adopted.
DOMESTIC AND INTERNATIONAL ELEMENTS
OF TRADE AND DEVELOPMENT STRATEGIES
Structuralism enabled governments to transform the protectionist trade
policies that benefited their principal political supporters into
comprehensive state-led development strategies. The trade and
development policies that most governments adopted following World
War II had both a domestic and an international dimension. At home, the
desire to promote rapid industrialization led governments to adopt state-led
development strategies that were sheltered by high protectionist barriers.
In the international arena, concern about the distributional implications of
international trade led developing countries to seek far-reaching changes to
the GATT–based trade system. We examine each dimension in turn.
Import Substitution Industrialization
Structuralism provided the intellectual justification for a state-led
development strategy. Confidence that the state could achieve what
markets would not was based in part on evidence of the dramatic
industrialization that the Soviet Union had achieved between 1930 and
1950 with an approach based on centralized planning and state ownership
of industry. In developing societies outside the Soviet bloc, this statecentered
approach to development came to be called import substitution
industrialization, or ISI. The strategy of ISI was based on a simple logic:
countries would industrialize by substituting domestically produced goods
for manufactured items they had previously imported.
Governments conceptualized ISI as a two-stage strategy (see Table 6.3).
Its initial stage was “wholly a matter of imitation and importation of tried
and tested procedures” (Hirschman 1968, 7). Easy ISI, as this first stage
was often called, focused on developing domestic manufacturing of
relatively simple consumer goods, such as soda, beer, apparel, shoes, and
furniture. The rationale behind the focus on simple consumer goods was
threefold. First, there was a large domestic demand currently satisfied by
imports. Second, because these items were mature products, the
technology and machines necessary to produce them could be acquired
easily from the advanced industrialized countries. Third, the production of
relatively simple consumer goods relies heavily on low-skilled labor,
allowing developing societies to draw their populations into manufacturing
176
activities without making large investments to upgrade their skills.
Governments expected to realize two broad benefits from easy ISI.
Initially, the expansion of manufacturing activities would increase wagebased
employment as underutilized labor was drawn out of agriculture and
into manufacturing. In addition, the experience gained in these
manufacturing industries would allow domestic workers to develop skills,
collectively referred to as general human capital, that could be applied
subsequently to other manufacturing businesses. Of particular importance
were the management and entrepreneurial skills that would be gained by
people who worked in and managed the manufacturing enterprises
established in this stage. Success in the easy stage would therefore create
many of the ingredients necessary to make the transition to the second
stage of ISI.
TABLE 6.3
Stages of Industrialization in Mexico and Brazil, 1880–
1968
Commodity
Exports,
1880–1930
Primary ISI,
1930–1955
Secondary ISI, 1955–
1968
Main
Industries
Mexico:
Precious
metals,
minerals, oil
Brazil: Coffee,
rubber, cocoa,
cotton
Mexico and
Brazil: Textiles,
food, cement,
iron and steel,
paper, chemicals,
machinery
Mexico and Brazil:
Automobiles,
electrical and
nonelectrical
machinery,
petrochemicals,
pharmaceuticals
Major
Economic
Actors
Mexico:
Foreign
investors
Brazil:
National
private firms
Mexico and
Brazil: National
private firms
Mexico and Brazil:
State-owned
enterprises,
transnational
corporations, and
national private firms
Orientation
of the
Economy
World market Domestic market Domestic market
Note: ISI = import substitution industrialization.
Source: Gereffi 1990, 19.
177
Easy ISI would eventually cease to bear fruit. The domestic market’s
capacity to absorb simple consumer goods would be exhausted, and the
range of such goods that could be produced would be limited. At some
point, therefore, governments would need to shift from easy ISI to a
second-stage strategy characterized by the development of more complex
manufacturing activities. One possibility would be to shift to what some
have called an export substitution strategy, in which the labor-intensive
manufactured goods industries developed in easy ISI begin to export rather
than continue to produce exclusively for the domestic market. Many East
Asian governments adopted this approach, as we shall see in Chapter 7.
The second alternative, and the one adopted by most governments
outside of East Asia, was secondary ISI. In secondary ISI, emphasis shifts
from the manufacture of simple consumer goods to consumer durable
goods, intermediate inputs, and the capital goods needed to produce
consumer durables. In Argentina, Brazil, and Chile, for example,
governments decided to promote domestic automobile production as a
central component of secondary ISI. Each country imported cars in pieces,
called complete knockdowns, and assembled the pieces into a car for sale
in the domestic market. Domestic auto firms were required to gradually
increase the percentage of locally produced parts used in the cars they
assembled. In Chile, for example, 27 percent of a locally produced car’s
components had to be manufactured domestically in 1964. The percentage
rose to 32 percent in 1965 and then to 45 percent in 1966 (Johnson 1967).
By increasing the percentage of local components of cars and other
goods in this manner, governments hoped to promote the development of
backward linkages throughout the economy (Hirschman 1958). Backward
linkages arise when the production of one good, such as a car, increases
demand in industries that supply components for that good. Thus,
increasing the percentage of locally produced components of cars, by
increasing the demand for individual car parts, would increase domestic
part production. The latter would in turn increase demand for inputs into
part production: steel, glass, and rubber, for example. Industrialization,
therefore, would spread backwards from final goods to intermediate inputs
to capital goods as backward linkages multiplied.
Governments promoted secondary ISI with three policy instruments:
government planning, investment policy, and trade barriers. Most
governments structured their efforts around 5-year plans (Little 1982, 35).
Planning was used to determine which industries would be targeted for
178
development and which would not, to figure out how much should be
invested in a particular industry, and to evaluate how investment in one
industry would influence the rest of the economy. India’s second Five
Year Plan (1957–1962), for example, sought to generate ambitious growth
in manufacturing by targeting the development of capital goods production
(Srinivasan and Tendulkar 2003, 8). The plan thus served as the
coordination device that governments thought necessary, given their belief
that the market itself could not coordinate investment decisions.
With a plan in place, governments used investment policies to promote
targeted industries. Most governments either nationalized or heavily
controlled the financial sector in order to direct financial resources to
targeted industries. Governments also invested directly in those economic
activities in which they thought the private sector would not invest. Much
of the infrastructure necessary for industrialization—things such as roads
and other transportation networks, electricity, and telecommunications
systems—it was argued, would not be created by the private sector. In
addition, the private sector lacked access to the large sums of financial
support needed to make huge investments in a steel or auto plant.
Moreover, it was claimed that private-sector actors lacked the technical
sophistication required for the large-scale industrial activity involved in
secondary ISI.
Governments invested in these industries by creating state-owned and
mixed-ownership enterprises. In Brazil, for example, state-owned
enterprises controlled more than 50 percent of total productive assets in the
chemical, telecommunications, electricity, and railways industries and
slightly more than one-third of all productive assets in metal fabrication
(Trebat 1983). Indian state-owned enterprises provided 27 percent of total
employment and 62 percent of all productive capital (Krueger 1993a, 24–
5). In Africa, governments in Ghana, Mozambique, Nigeria, and Tanzania
each created more than 300 state-owned enterprises, and in many African
countries, state-owned enterprises accounted for 20 percent of total wagebased
employment (World Bank 1994b, 101). Throughout developing
societies, therefore, the shift to secondary ISI was accompanied by the
emergence of the state as a principal, and in many instances the largest,
owner of productive capacity.
Finally, governments used trade barriers to control foreign exchange and
protect infant industries. Because export earnings were limited,
governments controlled foreign trade to ensure that foreign exchange
supported their development objectives (Bhagwati 1978, 20–33). After all,
many elements critical to industrialization, including intermediate inputs
179
and capital goods, had to be imported. Protection also allowed infant
industries to gain the experience needed to compete against established
producers. In Brazil and India, for instance, the state prohibited imports of
any good for which there was a domestic substitute, regardless of price and
quality differences.
The scale and the structure of protection that governments used to
promote industrialization are illustrated in Table 6.4, which focuses on
Latin America in 1960. In all but two of the listed countries, nominal
protection on nondurable consumer goods was well over 100 percent, and
for all but three countries, tariffs on consumer durables also were over 100
percent. Mexico and Uruguay stand out as clear exceptions to this pattern,
which has more to do with those countries’ extensive use of import quotas
in place of tariffs than with an unwillingness to protect domestic producers
(Bulmer-Thomas 1994, 279). It is also clear that tariffs were lower for
semi-manufactured goods, industrial raw materials, and capital goods (all
of which were items that developing countries needed to import in
connection with industrialization) than they were for consumer goods. This
pattern of tariff escalation was common in much of the developing world
(Balassa and Associates 1971).
The costs of ISI were borne by agriculture (see Krueger 1993a; Krueger,
Schiff and Valdes 1992; Binswanger and Deininger 1997). Governments
taxed agricultural exports through marketing boards that controlled the
purchase and export of agricultural commodities (Krueger et al. 1992, 16).
Often established as the sole entity with the legal right to purchase,
transport, and export agricultural products, marketing boards set the price
that farmers received for their crops. In the typical arrangement, the
marketing board would purchase crops from domestic farmers at prices
well below the world price and then would sell the commodities in the
world market at the world price. The difference between the price paid to
domestic farmers and the world price represented a tax on agricultural
incomes that the state could use to finance industrial projects (Amsden
1979; Bates 1988; Krueger 1993a). The trade barriers that protected
domestic manufacturing firms from foreign competition also taxed
agriculture. Tariffs and quantitative restrictions raised the domestic price
of manufactured goods well above the world price. People employed in the
agricultural sector, who consumed these manufactured goods, therefore
paid more for them than they would have in the absence of tariffs and
quantitative restrictions (Krueger 1993a, 9).
TABLE 6.4
180
Nominal Protection in Latin America, circa 1960
(percent)
Source: Bulmer-Thomas 1994, 280, Table 9.1.
Such government policies transferred income from rural agriculture to
the urban manufacturing and nontraded-goods sectors. The size of the
income transfers was substantial. As a World Bank study summarized,
the total impact of interventions … on relative prices [between agriculture and
manufacturing] was in some countries very large. In Ghana … farmers
received only about 40 percent of what they would have received under free
trade. Stated in another way, the real incomes of farmers would have
increased by 2.5 times had farmers been able to buy and sell under free trade
prices given the commodities they in fact produced. While Ghanaian total
discrimination against agriculture was huge, Argentina, Cote d’Ivoire, the
Dominican Republic, Egypt, Pakistan, Sri Lanka, Thailand, and Zambia also
had total discrimination against agriculture in excess of 33 percent, implying
that in all those cases, farm incomes in real terms could have been increased
by more than 50 percent by removal of these interventions.
(Krueger 1993a, 63)
Thus, ISI redistributed income. The incomes of export-oriented
producers fell while those of import-competing producers rose.
A Closer Look
Import Substitution Industrialization in Brazil
In the late nineteenth and early twentieth centuries, Brazil was the
classic case of a country that exported primary commodities. Its
181
principal crop, coffee, accounted for a large share of its production and
the overwhelming majority of its export earnings. This economic
structure was supported by a political system dominated by the
interests of coffee producers and other agricultural exporters (Bates
1997). Political authority in Brazil was decentralized, and the states
used their power in the country’s federal system to influence
government policy. As a result, Brazil pursued a liberal trade policy
throughout the late nineteeth and early twentieth centuries. World War
I and the Great Depression disrupted these arrangements. The world
price for coffee fell sharply in the late 1920s and early 1930s,
generating declining terms of trade and rising trade deficits. The
government responded to this crisis by adopting protectionist measures
to limit imports. The initial turn to protectionism was accompanied by
political change. A military coup in 1930 handed power to Getúlio
Vargas, who centralized power by shifting political authority from the
states to the federal government. Even though Vargas did not adopt an
ISI strategy, this period represented in many respects the easy stage of
ISI (Haggard 1990, 165–6). Protectionism promoted the growth of
light manufacturing industries at a rate of 6 percent per year between
1929 and 1945 (Thorp 1999, 322). Concurrently, the centralization of
power created a state that could intervene effectively in the Brazilian
economy. Although the export-oriented interests did not lose all
political influence in this new political climate, the balance of power
had clearly shifted toward new groups emerging in urban centers: the
professionals, managers, and bureaucrats who constituted the
emerging middle class and the nascent manufacturing interests. As
Brazil moved into the post-World War II period, therefore, the stage
was set for the transition to secondary ISI.
A full-blown ISI strategy emerged in the 1950s. The government
restricted imports tightly with the so-called law of similars, which
effectively prohibited the import of goods similar to those produced in
Brazil. In 1952, the Brazilian government created the National
Economic Development Bank (BNDE), an important instrument for
industrial policy through which the Brazilian state could finance
industrial projects. In the late 1950s, the government created a new
agency, the National Development Council, to coordinate and plan its
industrialization strategy. In taking up its task, the council was heavily
influenced by structuralist ideas (Haggard 1990, 174). Studies
conducted within these agencies—and, in some instances, in
collaboration with international agencies such as the United Nations
182
(UN) Economic Commission on Latin America—focused on how best
to promote industrialization (Leff 1969, 46). Most of these studies
came to similar conclusions: industrialization in Brazil would quickly
run into constraints caused by inadequate transportation networks
(road, rail, and sea), shortages of electric power, and the
underdevelopment of basic heavy industries such as steel, petroleum,
chemicals, and nonferrous metals. Building up those industries thus
became the focus of the government’s development policies. The
Brazilian government had little faith that the private sector would
create and expand these critically important industries. Instead,
policymakers determined that the state would have to play a leading
role. In the early 1950s, the state nationalized the oil and electricity
industries and began investing heavily in the expansion of capacity in
both. A similar approach was adopted in the transportation sector (in
which the government owned the railways and other infrastructure), in
the steel industry, and in telecommunications. By the end of the 1950s,
the state accounted for 37 percent of all investment made in the
Brazilian economy. As a result, the number of state-owned enterprises
grew rapidly, from fewer than 35 in 1950 to more than 600 by 1980.
Beyond creating these basic industries, the Brazilian government
also sought to create domestic capacity to produce complex consumer
goods. To achieve this objective, Brazil, in contrast to many other
developing countries, drew heavily upon foreign investment to
promote the development of certain industries. The auto industry is an
excellent example. In 1956, the Brazilian government prohibited all
imports of cars. Any foreign producer that wanted to sell cars in the
Brazilian market would have to set up production facilities in the
country. To ensure that such foreign investments were not simple
assembly operations in which the foreign company imported all parts
from its suppliers at home, the Brazilian government instituted local
rules that required the foreign automakers operating in the country to
purchase 90 percent of their parts from Brazilian firms. In order to
induce foreign automakers to invest in Brazil under these conditions,
the government offered subsidies; by one account, the subsidies offset
about 87 percent of the total investment between 1956 and 1969.
Relying on this strategy, Brazilian auto production rose from close to
zero in 1950 to almost 200,000 cars in 1962.
Brazil’s ISI strategy helped transform the country’s economy in a
remarkably short time. Imported consumer nondurable goods (the
products targeted during easy ISI) had been almost completely
183
replaced with domestic production by the early 1950s (Bergsman and
Candal 1969, 37). Imported consumer durables, the final goods
targeted in secondary ISI, fell from 60 percent of total consumption to
less than 10 percent of total consumption by 1959. Imports of capital
goods also fell, from 60 percent of total domestic consumption in
1949, to about 35 percent of consumption in 1959, and then to only 10
percent by 1964. Finally, imports of intermediate goods, the inputs
used in producing final goods, also fell continually throughout the
decade, to less than 10 percent of total consumption by 1964. Thus, as
imports were barred and domestic industries created, Brazilian
consumers and producers purchased a much larger percentage of the
goods they used from domestic producers and a much smaller
percentage from foreign producers. As a consequence, the importance
of manufacturing in the Brazilian economy increased sharply: whereas
manufacturing accounted for only 26 percent of total Brazilian
production in 1949, by 1964 it accounted for 34 percent.
The strategy of ISI promoted rapid economic growth in the 1960s and
1970s: developing countries’ economies grew at annual average rates of
between 6 percent and 7.6 percent during this period. In many countries, it
was the manufacturing sector that drove economic growth. Argentina,
Brazil, Chile, Mexico, Mozambique, Nigeria, Pakistan, and India, to select
only a few examples, all enjoyed average annual rates of manufacturing
growth between 5 percent and 10 percent during the 1960s. A glimpse
back at Table 6.1 indicates that, in Latin America, manufacturing’s share
of the total economy increased substantially between 1960 and 1980. Thus,
although the policies that governments adopted had important effects on
the distribution of income, they also appeared to be transforming
developing societies into industrialized economies.
Reforming the International Trade System
Developing countries also tried to alter the rules governing international
trade. For many developing-country governments, these efforts reflected
their experience with colonialism. India’s perspective was not unique:
international trade was “a whirlpool of economic imperialism rather than a
positive instrument for achieving economic growth” (Srinivasan and
Tendulkar 2003, 13). Consequently, as early as 1947, India, Brazil, and
Chile were arguing that the multilateral rules the United States and Great
Britain were writing failed to address the economic problems that
184
developing countries faced (Kock 1969, 38–42). Advancing the infantindustry
justification for protection, many developing countries argued that
their firms could not compete with established producers in the United
States and Europe. Yet, GATT rules not only made no provision for the
infant-industry justification for protection but indeed, explicitly prohibited
the use of quantitative restrictions and tightly restricted the use of tariffs.
Developing countries insisted that they be given a relatively free hand in
the use of trade restrictions to promote economic development, because
the GATT failed to do so.
Developing countries continued to press for GATT reforms throughout
the 1950s (see Kock 1969, 238; Finger 1991). By the early 1960s, a
coalition of developing countries dedicated to far-reaching reform had
emerged. Its first important success was achieved with the formation of the
United Nations Conference on Trade and Development (UNCTAD) in
March of 1964. The UNCTAD was established as a body dedicated to
promoting the interests of developing countries in the world trade system.
At the conclusion of this first UNCTAD conference, 77 developingcountry
governments signed a joint declaration calling for reform of the
international trade system. Thus was born the Group of 77, the leading
force in the campaign for systemic reform. During the next 20 years, trade
relations between the developing world and the advanced industrialized
countries revolved almost wholly around competing conceptions of
international trade rules embodied in the GATT and UNCTAD.
During the 1960s, the Group of 77 used UNCTAD to pursue three
international mechanisms that would increase their share of the gains from
trade (Kock 1969; UNCTAD 1964; Williams 1991). First, the Group of 77
sought commodity price stabilization schemes. Commodity price
stabilization was to be achieved by setting a floor below which commodity
prices would not be allowed to fall and by creating a finance mechanism,
funded largely by the advanced industrialized countries, to purchase
commodities when prices fell below the floor. Stabilizing commodity
prices would be an important step toward stabilizing developing countries’
terms of trade (recall the Singer-Prebisch hypothesis). The Group of 77
also sought direct financial transfers from the advanced industrialized
countries to compensate them for the purchasing power they were losing
from declining terms of trade (UNCTAD 1964, 80). Developing countries
also sought greater access to core-country markets, pressuring the
advanced industrialized countries to eliminate trade barriers on primary
commodities and to provide manufactured exports from developing
countries with preferential access to the core-countries’ markets.
185
These reform efforts yielded few concrete results. Core countries agreed
to incorporate concerns specific to developing countries into the GATT
charter. In 1964, three articles focusing on developing countries were
included in the GATT Part IV. Part IV called upon core countries to
improve market access for commodity exporters, to refrain from raising
barriers to the import of products of special interest to the developing
world, and to engage in “joint action to promote trade and development”
(Kock 1969, 242). In the absence of meaningful changes in the trade
policies pursued by the advanced industrialized countries, however, Part
IV provided few concrete gains. The advanced industrialized countries
also allowed the developing countries to opt out of strict reciprocity during
GATT tariff negotiations. The developing countries that belonged to the
GATT were therefore able to benefit from tariff reductions without having
to offer concessions in return. Benefits from this concession were more
apparent than real, however: GATT negotiations focused primarily on
manufactured goods produced by the advanced industrialized countries
and excluded agriculture, textiles, and many other labor-intensive goods.
Developing countries were therefore exporting few of the goods on which
the advanced industrialized countries were actually reducing tariffs. In the
late 1960s, the advanced industrialized countries agreed to the
Generalized System of Preferences (GSP), under which manufactured
exports from developing countries gained preferential access to advanced
industrialized countries’ markets. This concession, too, was of limited
importance, because advanced industrialized countries often limited the
quantity of goods that could enter under preferential tariff rates and
excluded some manufacturing sectors from the arrangement entirely.
Even though their efforts during the 1960s had achieved few concrete
gains, the Group of 77 escalated its demands in the early 1970s. Escalated
demands were sparked by the 1973 oil shock. The oil shock was a clear
illustration of the potential for commodity power. The world’s major oilproducing
countries, working together in the Organization of Petroleum
Exporting Countries (OPEC), used their control of oil to improve their
terms of trade. OPEC’s ability to use commodity power to extract income
from the core countries strengthened the belief within the Group of 77 that
commodity power could be exploited to force fundamental systemic
change.
Greater confidence in the possibilities that their control of commodities
offered led the Group of 77 to develop a set of radical demands dubbed the
New International Economic Order (NIEO). The NIEO represented an
attempt to create an international trade system whose operation would
186
promote development (see Krasner 1985). The NIEO, which the UN
General Assembly adopted in December 1974, embodied a set of reforms
that would have radically altered the operation of the international
economy. In addition to the three mechanisms that developing countries
had demanded during the 1960s, the NIEO included rules that would grant
developing countries greater control over multinational corporations
operating in their countries, easier and cheaper access to northern
technology, a reduction in foreign debt, increased foreign aid flows, and a
larger role in the decision-making processes of the World Bank and
International Monetary Fund (IMF).
Governments in the advanced industrialized countries refused to make
significant concessions, and by the mid-1980s the NIEO had disappeared
from the international agenda. The failure of the NIEO has been attributed
to a number of factors. First, developing countries were unable to establish
and maintain a cohesive coalition. The heterogeneity of developing
countries’ interests made it relatively easy for the advanced industrialized
countries to divide the Group of 77 by offering limited concessions to a
small number of governments in exchange for defection from the broader
group. In addition, the Group of 77 had hoped that OPEC would assist it
by linking access to oil to acceptance of the NIEO. But OPEC
governments were unwilling to use their oil power to help other
developing countries achieve broader trade and development objectives.
Most importantly, however, by the early 1980s, many developing
countries were facing serious balance-of-payments problems and turned to
the IMF and the World Bank for financial support. The need to obtain IMF
and World Bank assistance altered the balance of power in favor of the
advanced industrialized countries. This power shift sparked a reform
process that changed fundamentally development strategies throughout the
developing world.
CONCLUSION
Throughout much of the postwar period, developing countries insulated
themselves from the world trade system. The interaction between domestic
politics on the one hand, and economic shocks and decolonization on the
other, generated governments that were highly responsive to the interests
of import-competing manufacturing industries and a growing class of
urban workers. Influenced greatly by structuralism, most governments
transformed the political incentive to protect these domestic industries into
ambitious state-led development strategies. Structuralism’s critique of the
187
ability of domestic and international markets to promote industrialization
led governments to intervene in domestic markets to overcome the market
imperfections that reduced private incentives to invest in manufacturing
activities.
Policy Analysis and Debate
The Sustainable Development Goals
Question
Can the Sustainable Development Goals eradicate extreme poverty?
Overview
Members of the UN agreed in 2015 that for the next 15 years they
would focus their development policies on 17 Sustainable
Development Goals (SDGs). The SDGs constitute an ambitious
attempt to build on the gains realized through the Millenium
Development Goals, and include (among other things) end extreme
poverty everywhere (measured as living on less than $1.25 per day)
and cut the numbers living in poverty in half by 2030. In addition, the
SDGs place greater emphasis on sustainable development—and thus
have a variety of environmental goals—and they attach greater
importance to protection of human rights. Governments are to achieve
these goals through extensive planning at the domestic and
international levels. Policies based on these plans will in turn be
supported by foreign aid offered by the international community. For
that purpose, the UN has called upon rich countries to provide aid
equal to 0.7 percent of GDP to developing countries and provide
technical assistance and technology transfers where it is useful to do
so.
The logic upon which SDGs rest is similar to the thinking that at the
broad level shaped the government’s role in ISI. The SDGs rest on a
diagnosis of poverty that emphasizes structural factors. Rather than
emphasize market failure, however, contemporary thinking
emphasizes a “poverty trap”:
When poverty is extreme, the poor do not have the ability—by
themselves—to get out of the mess … When [people] are utterly
destitute, they need their entire income, or more, to survive … There is no
188
 ■
 ■
 ■
 ■
 ■
 ■
margin of income above survival that can be invested for the future.
(Sachs 2005, 56)
People can escape the poverty trap with help from the contemporary
analogue of the “big push.” The international community must provide
“a leg up” through well-funded and well-conceived government policy
initiatives. Given the logic upon which they are based, do you think
the SDGs will be successful?
Policy Options
An SDG-like strategy is necessary if the world is to eradicate
extreme poverty. Governments must embrace these goals.
The SDGs rest on faulty logic and thus cannot reduce extreme
poverty. Governments should re-evaluate their approach to the
problem of global poverty.
Policy Analysis
Do developing-country governments have incentives to
implement the policies called for by the SDG strategy? Why or
why not?
Do advanced industrialized countries have incentives to provide
the foreign aid that is required to support SDG policies? Why or
why not?
Take a Position
Which option do you prefer? Justify your choice.
What criticisms of your position should you anticipate? How
would you defend your recommendation against these criticisms?
Resources
Online: To learn more about the SDGs and current progress toward
achieving them, conduct an online search for the keywords UN and
MDGs. Look especially for the UN’s annual progress reports.
In Print: Read the alternative perspectives embodied in Jeffrey Sachs’
Ending Poverty: Economic Possibilities of Our Time (New York:
Penguin Press, 2005), and William Easterly’s The White Man’s
189
Burden: Why the West’s Efforts to Aid the Rest Have Done So Much
Ill and So Little Good (New York: Penguin Publishers, 2006).
To the extent that developing countries participated in the global trade
system, they sought to achieve far-reaching reform of the rules governing
the system. Again, the structuralist critique served an important role in this
effort, as it suggested that developing countries could not expect to gain
from trade with the advanced industrialized countries until they themselves
had industrialized. Moreover, structuralism claimed that trade based on
GATT rules would only make industrialization harder to achieve. Rather
than accept participation in the global economy on what they viewed as
vastly unequal terms, developing countries battled to change the rules
governing international trade in order to capture a larger share of the
available gains. Thus, an international struggle over the distribution of the
gains from trade arose as an important counterpart of the domestic strategy
of redistributing resources from agriculture to industry embodied in ISI.
KEY TERMS
Backward Linkages
Big Push
Complementary Demand
Easy ISI
Enclave Agriculture
Export Substitution Strategy
GATT Part IV
Generalized System of Preferences
Group of 77
Import Substitution Industrialization
Monoexporters
New International Economic Order
Pecuniary External Economies
Secondary ISI
Singer-Prebisch Theory
Structuralism
Sustainable Development Goals
Terms of Trade
United Nations Conference on Trade and Development
SUGGESTIONS FOR FURTHER READING
For a readable introduction to structuralism and development strategies more
190
generally, see Ian Little, Economic Development (New York: Basic Books,
1982). For an in-depth look at Latin America, see Victor Bulmer-Thomas, The
Economic History of Latin American since Independence, 3rd edition
(Cambridge: Cambridge University Press, 2014).
For a comparative study of the role of the state in development, see Atul Kohli,
State-Directed Development: Political Power and Industrialization in the
Global Periphery (Cambridge: Cambridge University Press, 2004).
For a detailed examination of the New International Economic Order, see the
recent special issue of Humanity (2015 6 (1), http://humanityjournal.org/issue-
6-1/), Paul Adler, 2017. “‘The Basis of a New Internationalism?’ The Institute
for Policy Studies and North-South Politics from the NIEO to Neoliberalism.”
Diplomatic History 41(4): 665–93, and the now classic, Stephen Krasner,
Structural Conflict: The Third World against Global Liberalism (Berkeley:
University of California Press, 1985).
191