34
MAINSTREAM THEORIES AND PRACTICES
overvalued exchange rates, large inefficient bureaucracies) without resorting to assumptions of ignorance or willful misbehavior by politicians and bureaucrats. Instead, the state is seen as a rational actor which seeks to maximize its economic and political utility by extending its influence within powerful groups in society, albeit at the cost of slower long-term growth and development (Rausser and Thomas 1990: 372-3).
However, the question remains whether this neoliberal political economy model is any more relevant to problems of Third World political development than the Keynesian pluralist model of the state that it seeks to replace. It has yet to offer any sophisticated theory of the Third World state, whether based in an examination of the internal structures and mechanisms of the state itself or in an analysis of the social composition of the state within broader societal structures.15 While public choice theory offers an analytical framework for understanding the seemingly irrational development strategies that have been followed by many Third World states, it provides no logical apparatus for political reforms that may produce more effective policies. In the end, this paradigm is limited by its narrow and cynical view of the political process. While this provides much of the ideological affinity between public choice theory and neoliberalism in general, it offers little analysis of real-world political complexities in which states may engage in a range of both productive and predatory behavior according to a variety of internal and external influences.
15      A more detailed critique of mainstream and neoclassical political theory appears in Brohman (1995a and 1995b).
2
Strategies of Growth and Industrialization
Chapter 1 described the evolution of development theories. Chapter 2 now considers how these theories were transformed into the practical strategies that dominated postwar development in the South. It first looks at the agroexport or primary export model, the prevailing strategy in most smaller, rural countries. The analysis then moves to the Keynesian strategy of import-substitution industrialization, which, for most of the postwar period, has been the dominant development approach in most larger, urban countries. The recent rise of neoliberal strategies of export-led growth is then covered. Finally, there is a detailed study of nontraditional exports, a new outward-oriented growth sector which many analysts believe offers good development prospects. Although contradictions and shortcomings can be uncovered in all these strategies, they contain positive aspects that ought not to be overlooked in the formulation of new development approaches.
The Agroexport Model
Throughout the postwar era, nearly all mainstream development strategies (including, most recently, neoliberalism) have called on the majority of Third World countries to exploit their 'comparative advantages' in cheap land and labor by expanding exports of agricultural goods and other primary commodities. This advice was particularly aimed at poorer, smaller countries in regions such as Central America and sub-Saharan Africa, which had little or no history of export-oriented industrialization. Indeed, agroexport production did expand rapidly during the early postwar period in regions like Central America, becoming a 'motor' for outward-oriented development, Bui many analysts now contend that a number of inherent tout radii lions have gradually played themselves out during recent years to
36
MAINSTREAM THEORIES AND PRACTICES
render the agroexport model, at least in its classical form, dysfunctional to future Third World development. At the same time, however, many countries possess few realistic alternatives to agroexport production in the short-to-medium term, especially to earn foreign exchange necessary for macroeconomic stability. Thus, a debate is now being waged over whether to retain agroexport production as a major axis of development and, if so, in what form. This section addresses these issues, first by examining the traditional agroexport model, and secondly by exploring possibilities for the establishment of alternative strategies which retain a significant agroexport component.
Comparative Advantage and Outward-Oriented Growth
As we saw in chapter 1, the principle of comparative advantage has been a key element of neoclassical development. This principle assures Third World countries that the road to higher growth and development lies in specialization and exchange - even with the advanced industrial countries of the First World. The classical nineteenth-century Ricardian conception of comparative advantage has since been modified and incorporated into neoclassical theory, especially through the work of Heckscher, Ohlin, and later Samuelson (1948). The resultant Hecksher-Ohlin-Samuelson (HOS) model of international trade assumes equal access to production technologies throughout the world, so that comparative advantages arise only from differential factor endowments (e.g., land, labor, capital). It follows that land- or labor-abundant Third World countries should specialize in and export land- or labor-intensive goods, while leaving the production of capital-intensive goods to those (First World) countries with greater endowments of capital.
The HOS model thus maintains that specialization and trade will increase levels of production and consumption in both developed and developing countries. The model also suggests that there is a tendency toward equalization of factor prices, including wage rates, following the development of trade. This would allow wage rates to rise gradually in many Third World countries, as specialization in labor-abundant goods eventually reduces their relative abundance of labor and drives up wages. Conversely, continuing capital-intensive production will steadily reduce relative labor scarcity in the First World, gradually decreasing its marginal productivities and wage rates. The neoclassical HOS model, therefore, holds out prospects for growth and development in the South on two fronts. During their initial stages of development, Third World countries can enjoy higher levels of growth and gradual wage increases through the production of primary com modities for the world market. During subsequent stages of development, many Third World countries will begin to gain comparative advantages in
.....

STRATEGIES OF GROWTH AND INDUSTRIALIZATION       37
semi-manufactured goods as gradual changes in their factor endowments allow for specialization in more sophisticated products.
In addition to stressing the advantages of specialization and trade, neoclassical theory also suggests that Third World countries can move to higher stages of development more quickly by maintaining an outward economic orientation. Due to the supposed lack of entrepreneurship and technological skills in most traditional societies, private foreign investment is seen as the best means of providing the capital and expertise needed to employ more sophisticated production techniques. It is implied that developing countries can only harm themselves by placing artificial limits on the free flow of trade and investment, since benefits from both the comparative advantage principle and the theory of mobility of capital transfers depend on the adoption of policies promoting outward-oriented economic growth.
Basic Elements of the Classical Agroexport Model
Neoclassical development strategies have thus focused on both stimulating growth according to comparative advantages and attracting foreign capital through infrastructure projects and other programs designed to provide a profitable and stable environment for investment. In many Third World countries, this has led to increased exports of primary commodities, particularly from the agricultural sector, to exploit comparative advantages based on relatively cheap labor and land. Especially for many of the poorer and smaller Third World economies, growth in the agroexport sector was thought to be critical for attracting (foreign) investment capital, creating a positive trade balance, and expanding job creation through the operation of 'multiplier' and 'spread' effects.
Since the Second World War, export agriculture has indeed attracted substantial foreign capital and has been the principal source of growth for many of the smaller Third World economies. 'Traditional' export crops (e.g., coffee, tea, bananas, sugar, cotton), as well as newer agroexports (e.g., animal feeds, beef, fruit and vegetables), have claimed a growing proportion of cultivable land in the South, including large tracts of tropical rainforest and grassland savanna. In many areas of Latin America, Africa, and, to a lesser extent, Asia, the deterioration of fragile ecosystems has been linked to agroexport expansion and the accompanying displacement of peasants Into environmentally sensitive areas (e.g., Lopez 1992; Amin 1993). At the same time, the production of staple foods for the domestic market has stagnated or declined in most agroexport-dominated economies; indeed, m many ol the South's main agricultural zones, it has been supplanted entirely. Moreover, society has become polarized through the exigencies of ugiorxpoit production, in which land concentration and absolute labor
38                   MAINSTREAM THEORIES AND PRACTICES
exploitation essentially generate comparative advantages. This has blocked possibilities for creating more broadly based development models stressing economic diversification for both internal and external markets.
The reorientation of agricultural production toward exports has also produced a profound structural transformation in many rural areas. The agroexport model, at least in its classical form, implies the destruction of small/medium peasant forms of production and enterprise, and of the rural village communities upon which they are based. Characteristically, the lateral expansion of large-scale agribusiness has squeezed peasants off their land in traditional food-producing areas. These former peasants have commonly been converted into a landless or near-landless floating reserve of labor which is often seasonally employed for peak periods of labor demand, such as to harvest crops for export. Peasant displacements have also contributed to growing rural-urban migration in many countries, thereby swelling the urban labor reserve, which exerts downward pressure on wages and undermines non-wage relations of production in urban areas. An enlargement and reconstruction of the overall surplus population takes place as members of rural families (notably youths, the elderly, and women) are pushed into the latent labor reserve, with many forced to eke out a bare existence in the nebulous informal sector.
For many Third World countries, particularly in Latin America, the agroexport model was associated with high rates of economic growth from the end of the Second World War until the mid-1970s. From 1950 to 1977, Central American agroexports increased by twelve times. During much of this period, some of Latin America's poorest and smallest countries (e.g., El Salvador, Guatemala, Nicaragua) sustained among the highest overall growth rates in the region, based on the dynamism of their agroexport sectors (Brohman 1989). By the end of the 1960s, however, many development theorists began to recognize that high rates of economic growth were not necessarily correlated with other basic objectives of Third World development. Deteriorating trends in employment, income distribution, and levels of poverty often accompanied impressive growth rates.
The Agroexport Model and Postwar Central America
Although much of the South has had some experience with the agroexport model, it is perhaps in Central America where the model has been most forcefully applied for the longest period of time. The Central American development performance therefore represents a good case study for assessing the model's long-term impact. Because economies that are concentrated on the export of primary commodities are extremely vulnerable to fluctuations in world market conditions, they characteristically follow a 'boom and bust' pattern. The Central American economies have traditionally been
..........-........---------------------*---------—————^—*—
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       39
concentrated on a few agroexport sectors and, during the postwar era, their overall economic performance has closely followed this classic pattern of boom and bust (FitzGerald 1991).
Although there were some variations in the performance of individual economies, Central America generally experienced a boom period of relatively high growth from the early 1950s to the mid-1970s. This growth was based on a very limited number of agroexports: the rise of cotton in the 1950s, beef in the 1960s, and sugar in the early 1970s complemented more traditional crops of coffee and bananas. However, this bpom period has given way to a pronounced recession that began in the late 1970s and has continued in most of these countries until the 1990s. Of the Central American economies, only Costa Rica has resumed a high rate of economic growth in recent years. However, its growth has essentially been based on the rise of non-traditional exports, tourism, and other forms of economic diversification, rather than continuing to depend on a reduced number of traditional agroexports.1
The agroexport model in Central America functions according to 'a logic of the minority' (Collins 1985: 108). The expansion of the model to serve the interests of a narrow economic and political elite has created the very conditions that are responsible for the marginalization and impoverishment of the majority. Investments and growth are concentrated among a few agricultural sectors controlled by foreign capital and allied fractions of the domestic bourgeoisie. Meanwhile, the remainder of the economy molders in neglect, unable to meet even the basic needs of the majority of the population (e.g., Whiteford and Ferguson 1991 ).2 Typically, the benefits of economic growth are concentrated among a few large-scale
1         A more detailed examination of new development strategies based on economic diversification into sectors such as tourism and non-traditional exports is carried out in the final section oř this chapter.
2         Within Central America, Costa Rica has represented an exception to this pattern of extreme polarization between social classes and economic sectors. Although agroexport production has traditionally formed an important part of the Costa Rican economy, the major agroexport sectors (at least until recently) have not been marked by excessive levels of concentration as in the other Central American economies. Historically, rural development in Costa Rica has included many medium-sized farms, in contrast to the more polarized latifundio-minifundio structure in much of the rest of Latin America. Many analysts contend that this has allowed a more diversified and internally articulated pattern of economic growth to evolve in Costa Rica which has, in turn, facilitated the rise of a social democratic political system. All of these features differentiate Costa Rican development from the classical agroexport model. 1 lowcvcr, as wc shall sec in the last section of this chapter, the recent rise of agroexport production in Costa Rica has been marked by many of the same characteristics (e.g., land concrntrutiona, peasant displacements and growing rural-urban migration, witlrnuiH rural iiiri|iirtlitirs) lluir huvr traditionally defined the agroexport model in the rest nl ( nil ml Antrum.
40                   MAINSTREAM THEORIES AND PRACTICES
export producers, primarily in the dominant agricultural sector and related agro-industries. The economic structure closely conforms to the classical Cepalista model of Latin American peripheral capitalism (see Prebisch 1950),3 with the dominant economic sectors oriented toward the reproduction and extension of a dependent capitalist mode of production. CEP AL (1983) characterized Central American export-led growth as 'superimposed development,' with a relatively modern agroexport sector superimposed on and independent of the remainder of the underdeveloped, internally oriented economy. Despite relatively high growth rates during the boom period of agroexport expansion, few internal 'multipliers' have been created that might generate jobs and income for the majority of the population.
Within this pattern of development, the evolution of the external sector 'determines the global behavior of the economy'; restrictions found in that sector 'mark the limit on the rate of domestic economic activity' (ibid.: 5-6). A direct relationship is established between the performance of the export sector, on the one hand, and overall rates of economic expansion, investment and capital accumulation, levels of employment, external balance of payments and import capacity, and the principal sources of government revenue, on the other. The export sector not only determines economic patterns, but also conditions the evolution of the social structure and configurations of political power. Torres-Rivas is among the many analysts who have stressed the fundamentally flawed nature of the agroexport model in Central America:
The export-oriented economy notably retarded national and social integration and contributed to the extreme rigidity of political and social relations . . . development has been determined by an externally-oriented dynamic whose essential nature has remained unchanged in spite of efforts (after World War II and especially after 1955) ... to implant a new productive base dependent on the growth of an internal market. (Torres-Rivas 1980: 25)
State Intervention to Support Agroexport Capitalism
In contrast to the rather haphazard involvement of the state within the Central American economies before the Second World War, state intervention in support of the accumulation requirements of agroexport capitalism has become much more pronounced in the postwar era. In addition to strength ening repressive labor policies and various other methods of coercion
3 During the 1950s and 1960s, the United Nations CEPAL or Comisión h'.corumimi para America Latina (Economic Commission for Latin America) based in Santiago, Chile, was a center of criticism of the agroexport model and other forms of dependem development in Latin America. A more detailed analysis of CEPAl.'s alternative strategy follows in the next section, concerning import-substitution industrialization.
....................-........
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       41
designed to expand the rural labor reserve, the state has employed a variety of mechanisms that have decisively influenced the overall pace and direction of the accumulation process in the dominant agroexport economy. Prominent among these state policies and programs have been: discriminatory credit, tariff, pricing, and exchange-rate policies; selective construction of roads and other infrastructure; and the provision of publicly subsidized irrigation, research and extension, storage, and processing/marketing facilities to favored producers (Brohman 1989).4 State economic intervention has accelerated tendencies toward concentration and centralization of capital in key agroexport sectors and related processing, commercialization, and import/export activities. Patterns of regionally uneven development have also been accentuated as state resources are directed toward major concentrations of large-scale agroexport production, while areas dominated by peasant producers of basic grains and other foodstuffs are left to stagnate in abject poverty and isolation.
This development strategy has been described as a 'repressive agroexport model' (Barraclough 1982: 15). If state economic measures prove insufficient to meet the accumulation requirements of the agroexport bourgeoisie, armed force is brutally applied to bring recalcitrant social sectors into line. Peasants, rural workers, and the urban poor have no real role in the system other than providing a steady source of cheap labor. Because the markets for Central America's principal (agroexport) production sectors are located overseas, methods of absolute exploitation can be used to hold down labor costs without adversely affecting demand for the goods produced. Rather than promoting the rise of a relatively 'progressive' modernizing bourgeoisie whose profitability might be based on technological advance, state policies have strengthened patterns of absolute exploitation and ownership concentration that have traditionally supported agroexport production by a reactionary landholding oligarchy.
Land Concentration, Semiproletarians, and Absolute Exploitation
Agroexport profitability stems not from increases in relative surplus linked with rising productivity, but is based on the extraction of absolute surplus
11 Moreover, many discriminatory state programs that have favored large-scale agroexport production over small-medium food production in Central America have also been supported by international aid and lending organizations (e.g., US Agency for International Development, lntcr-American Development Bank, World Bank). Particularly noteworthy in Central America (as well as in other areas such as the Brazilian Amazon, ihr Dominiem Republic, and Paraguay) have been greatly increased amounts ol mvcMiiii-iii capital supplied by international donors and lending institutions for i-xpunmiiii of ihr export brrl iniliKlry «nur- the ľ>60s.
42                  MAINSTREAM THEORIES AND PRACTICES
derived from maintaining very low labor costs.5 Land concentrations by the agroexport bourgeoisie have reinforced traditional precapitalist mechanisms of peasant exploitation and have provided impetus for the rise of newer, more capitalistic exploitative forms as land-poor peasants have been forced to seek seasonal wage labor in the agroexport sectors. The creation of a massive reserve of seasonal rural laborers has become a condition for meeting the demands of cheap labor upon which agroexport profitability and international competitivenesss are largely based. Low wages and poor working conditions associated with forms of absolute exploitation are related both to the seasonality of agroexport labor requirements and to the lack of alternative sources of steady income for masses of rural semiproletanans and itinerant proletarians (de Janvry 1981).6 At the same time, the ability of these groups to meet their families' needs from sources other than temporary wage labor in the agroexport sector (e.g., other types of seasonal wage labor, activities in the informal sector, partial subsistence production) allows agroexport producers to keep labor costs at levels beneath those which would be required to maintain a permanent, fully proletarianized labor force.
Accordingly, Central American rural development has created masses of semiproletarians and itinerant proletarians, which may be regarded as 'peculiar forms of the proletarianization process of the capitalist agroexport model' (Nunez 1980: 39). The expropriation and displacement of much of the rural population created a mobile labor force which had few alternatives but to respond to the requirments of cheap, seasonal labor within the
5         Profitability in most productive sectors of the advanced capitalist world during the postwar period has been based on increasing relative surplus value. Increases in relative surplus are linked with rising productivity, which also allows wages and labor costs to increase. In the agroexport economies of the South, however, profitability in the dominant agricultural sectors has traditionally been based on extraction of absolute surplus through maintenance of low labor and land costs rather than increasing productivity. For a more detailed discussion of forms of absolute surplus extraction linked with an analysis of the Central American agroexport model see Torres-Rivas (1981).
6         In his analysis of alternative roads of capitalist rural development in Latin America, de Janvry coined the term semiproletarians for land-poor peasants who are forced to seek seasonal wage labor (normally in agroexport harvests) to meet the social reproduction needs of their families. They are part peasant and part wage laborer, thus becoming rural semiproletarians. Itinerant proletarians are usually completely landless and have been forced into a cyclical migratory pattern of seasonal wage jobs. They are therefore fully proletarianized in that they normally work only as wage laborers; however, they are also itinerant in contrast to more sedentary permanent proletarians. These semiproletarians and itinerant proletarians, many of whom are former peasants displaced by postwar land concentrations, form ihc bulk of ihr (seasonal) labor requirements of the principal agroexport sectors in Central America and many other rural Third World economies.
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       43
dominant agroexport sectors around which capitalist growth revolved. Barry (1987: xiv) reports that, by the 1980s, rural landlessness had tripled since the 1960s and that about 80 percent of farmers possessed insufficient land to feed their families; at the same time, 85 percent of the best land was used for agroexports and 45 percent of total arable land was devoted to cattle grazing.
By the end of the postwar agroexport boom, all of the Central American countries (with the exception of Honduras)7 had developed immensely expanded rural labor reserves, composed primarily of part-time peasants and migrant workers who supplied the bulk of the labor power used by their agroexport sectors (Brohman 1989). Because other sources of income have been largely blocked as a result of agroexport expansion, the peasantry throughout Central America has remained dependent on seasonal wage labor during agroexport harvests as its principal source of family income. In Guatemala, for example, Burback and Flynn (1980) found that temporary wage labor in the agroexport sector accounted for almost three-quarters of total peasant family income. In Nicaragua, three export crops (coffee, cotton, and sugarcane), with particularly heavy labor requirements during the harvest season, controlled almost 54 percent of the total rural workforce in the 1970s (Baumeister 1984).
Concentrated Land Tenure and Income Inequalities
Within Third World economies dominated by agroexport production, there is a strong correlation between concentrations of land tenure and unequal income distributions (e.g., Barraclough 1982; Enge and Martinez-Enge 1991). Because land is generally the principal means of production in such economies, extreme concentrations of land ownership commonly lead to equally skewed distributions of income. In Central America the acceleration of land concentrations in the postwar era greatly widened income inequalities (Vilas 1984: 74). Of all the countries in the region, by the 1970s, only in Guatemala did the poorest 50 percent of the population receive a larger share of national income than the wealthiest 5 percent, and even there the margin was quite slight (23.5 percent versus 21.8 percent). In both Nicaragua and Honduras, the wealthiest 5 percent of the population received roughly double the percentage of national income that the poorest 50 percent earned. In all five countries, the share of national income garnered by the wealthiest 20 percent exceeded that of the poorest SO percent by a wide margin: 60.0 percent to 15.0 percent in Nicaragua, 50.7 percent to 20.8 percent in Costa Rica, 64.8 percent to 12.4% in El Salvador, 51.3 percent to 23.5 percent in Guatemala, and 58.9 percent to
In Honduras, ihr economic dominance of ihr 'banana enclave' along the Caribbean itmiil In! tu (hr i rrtilion of i inorr snlrnmry rural proletariat.
44                 MAINSTREAM THEORIES AND PRACTICES
17.3 percent in Honduras. The index of income polarization (measuring the inequality of average income between the wealthiest 5 percent and the poorest 50 percent of the population) was most extreme in Honduras (20.0) and Nicaragua (18.6), but was also substantial in the other countries of the region: 11.2 in Costa Rica, 15.0 in El Salvador, and 9.2 in Guatemala.8
Postwar inequalities became particularly pronounced among Central American countries, owing to the absolute domination of their economies by agroexport capital. Meanwhile, a marked tendency toward more regressive income distribution also marked rural development in other areas of Latin America that were subjected to a rapid, if somewhat less concentrated, form of agroexport expansion. A number of studies carried out near the end of the agroexport 'boom' in the 1970s reveal a staggering rate of rural poverty for Latin America as a whole: a 1975 World Bank study found that 42 percent of the region's rural population had per capita incomes of less that $75 per annum, while a 1978 CEP AL study reported that 62 percent of the region's rural households could not satisfy their basic needs (de Janvry 1981: 85). As a direct consequence of growing inequities and impoverishment, malnutrition and associated problems also increased, especially among the most vulnerable sections of the population such as poor children. By the 1980s, it was estimated that about three out of four children in Central America were malnourished (Barry 1987: xiv), while in Mexico some 90 percent of the rural population suffered from a severe deficiency of calories and protein (Esteva 1983: 13). Nor is this pattern unique to Latin America - links between agroexport growth and problems of malnutrition and food scarcity have been uncovered in many other Third World areas, including Africa (Bryant 1988), the Pacific Basin (Schuh and McCoy 1986), and Asia (Chisholm and Tyers 1982).
Increasing land concentrations, new and more onerous forms of rural exploitation, and widening rural inequalities have accompanied postwar agroexport booms in a succession of Third World countries. As Grindle (1986: 7, 112) notes, increasing rural inequalities in postwar Latin America are caused not by the isolation or backwardness of the peasantry, as the dualist thesis of the modernization approach contends, but by the ways in which peasants have been inserted into the expanding capitalist economy in the countryside:
The rural poor are not isolated or backward and have not simply been left behind by the modern sector, but the growth in their unemployment and underdevelopment, landlessness, wage dependence and migration is a direct result of developments in the modern capitalist sector and of state policies ... At the same time that policies for agricultural modernization
8        For more Central American data on land tenure and income distribution, see Itmlim.in (1989:518-19).
IMMMHHiMHMaHHMIHiBl
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       45
increasingly dominated markets and profits, peasants were driven into greater debt, squeezed from their land, forced into wage labor, and pushed to migrate in increasing numbers.
Systemic Limitations and Contradictions of the Agroexport Model
The contradictory nature of the agroexport model in Central America has not only blocked possibilities of development for much of the peasantry and domestic agriculture, it has also limited growth within the industrial sector to a narrow branch of activities tied through forward/backward linkages to agroexport production. The limited industrialization that took place under the Central American Common Market (CACM) in the 1960s and 1970s actually strengthened the hold of the dominant agrarian-based export structure over the remainder of the economy (Torres-Rivas 1980: 28-30). The polarization that characterized agricultural development was replicated and extended into the industrial sector with the establishment of a reduced group of large capital-intensive enterprises linked to agroexport production alongside a large number of small, technologically backward operations aimed at domestic consumption. This type of truncated industrialization di I little to alter the international position of Central America's economy as an exporter of largely unrefined agricultural products and an importer of manufactured intermediate and final consumption goods. The concentration of industrial production within sectors supplying goods to the advanced capitalist world did not permit the broadening of patterns of development to other areas of the economy. Very partial internal processing of most agroexports meant that most of the value-added and employment associated with turning primary agricultural products into final consumption goods was exported to the developed world.
With the possible exception of Costa Rica, the basic characterisitics of the postwar Central American economies are defined by: the dominance of overall production by the agricultural sector; a chronic crisis within agricultural production for the internal market; the weakness and narrowness of an industrial structure based on agro-industrial processing; the bloated nature of the informal sector and unproductive commercial activities; and severe polarization of growth in sectoral, regional, and class terms. All of these characteristics can be seen as direct consequences of the particular logic of BCCumulation upon which the agroexport model is based. The manner in which agroexport production takes place, supported by diverse forms of absolute exploitation and the concentration of land and other means of production, conditions the entire internal socioeconomic structure and the niturc ol political power. The linking of economic growth to a subordinate position within the international division of labor, based on the comparative advantage oflrrrd to agroexport sectors by maintaining low labor and land
15
46                   MAINSTREAM THEORIES AND PRACTICES
costs, retards national socioeconomic and spatial integration and inevitably leads to an extremely rigid polarization of social and political relations.
The growth and pattern of overall development in an agroexport economy depends on an externally oriented dynamic in which the demand for the goods of its principal production sectors comes not from domestic consumption but from overseas markets. Because their sources of demand are external, the key agroexport sectors operate according to an independent logic of accumulation which has little correspondence to the necessities of broader development for other economic sectors. The lack of domestic demand for consumption goods blocks the spread of internally oriented growth. It is also directly related to the accumulation logic of agroexport production, rooted in property concentrations and deepening forms of absolute exploitation. Mechanisms which might stimulate internal demand (e.g., agrarian reforms, rising wage levels and income redistribution, improvements in the social wage) do not have a functional relationship to the dynamic of the accumulation process of the dominant agroexport sectors. Indeed, they would impede agroexport production, serving to undermine the bases of its comparative advantage (i.e., cheap labor and land) in the international commodity markets.
Growing systemic tension within the agroexport model has recently been noted in most Central American countries (e.g., Bulmer-Thomas 1988; LaFeber 1983; Pelupessy 1991b; Torres-Rivas 1981; Williams 1986).9 Although the way in which systemic contradictions eventually manifest themselves is also dependent on indeterminate sociopolitical factors which may be particular to each social formation, there is widespread agreement among many analysts that the agroexport model, at least in its classical form, has recently become exhausted and offers no real future for development. As intractable societal problems have become more acute in Central America, the model is said to have entered the stage of its 'final crisis,' which increasingly calls into question the ability of the model to overcome its central contradictions in the absence of structural change. The focus of many recent accounts of Central American development has been on the economic and political difficulties involved in breaking with the agroexport model against the interests of the region's powerful landholding oligarchy and its domestic and foreign allies. Possibilities for broadening development based on industrialization and other forms of economic diversification have been blocked because the class fractions that dominate the major productive sectors and political arenas have opposed economic changes that miglu reduce their profits and power (FitzGerald 1991; Torres-Rivas 1981). At the same time, the logic by which the agroexport model operates has
9 Similar tensions have also been noted in other areas of the South in which poslwm development has been dominated by agroexports (e.g., Barraclough 1982; Collin« 1485; dejanvry 1981; Weisskoff 1992).
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       47
systematically impoverished large segments of the population and greatly accelerated environmental destruction in many rural areas (Whiteford and Ferguson 1991; Williams 1986).
From this perspective, the roots of the present crisis in Central America can be traced to the maintenance of conditions that permit an exploitative and exclusionary agroexport model to endure against the interests of the majority. Within agrarian-based societies such as those of Central America, strong links exist between patterns of land tenure, societal polarization, and the arbitrary, and often ruthless, exercise of political power. Many analysts portray the recent rise of political instability and military conflict in Central America as symptomatic of the exhaustion of the agroexport model. Williams (1986: 191), for example, argues that 'even if the wars in Central America ended today ... a [continuation of the same] development program would produce the conditions for a resurgence of the conflict within ten or fifteen years.' Accordingly, if stability and development are to return to Central America, new development strategies must be found that operate according to a different logic than the outmoded agroexport model. On the one hand, the model relies on a world commodities market that has substantially changed in recent years. Stagnant global demand has combined with increased supply, resulting from the adoption of similar export-led strategies by all of the South's major agricultural producers, to send prices of many traditional agroexports into a protracted decline. On the other hand, the degree of political repression, economic exploitation, and environmental destruction required to maintain the 'comparative advantages' of the major agroexport sectors is no longer feasible.
Possibilities for an Alternative Agroexport-led Development Model
Most analysts agree that future development for agroexport economies is dependent on structural change which responds to the needs and interests of the popular majority rather than of a narrow elite. Nevertheless, in traditional agroexport areas such as Central America in which neither regional autarchy nor widespread economic diversification are realistic possibilities for the immediate future, agroexport production must be maintained at least in the medium term. Agroexports in small, narrow dependent economies represent the equivalent of a 'capital goods sector' (FitzGerald 1985), in that agroexports have the unique ability to generate foreign exchange, which in turn determines the availability of the producer goods needed to generate overall economic growth. The deterioration of key agroexport sectors would cause unacceptably high social and economic costs in terms not only of foreign exchange earnings, but also of internal wviugs and investment capital* productive employment, income generation, m.I  sources  til  government   revenues.  Given  the  existing structures  of
48
MAINSTREAM THEORIES AND PRACTICES
production and social-class formations in the region, there are no economic sectors which could readily replace agroexports. It follows that any program of economic revitalization must include a strategy to recover the agroexport dynamic.
However, while selected agroexports should continue to play a role in the creation of any viable alternative development strategy, the central mechanisms by which the old model operated need to be replaced to allow for more broadly based and sustainable development. A dynamic agroexport sector need not necessarily be based on an exclusionary and exploitative latifundio-minifundio model. Instead, it might be based on alternative forms of rural organization such as medium-sized farms and cooperatives that, if given proper state support, could combine economic viability with social equity. Small/medium farmers and cooperatives might complement their traditional focus on domestic food production with a carefully managed entry into selected agroexport markets. Indeed, many small/medium farmers' associations in Central America have indicated that their members would welcome the opportunity to diversify into export sectors if provided with proper conditions (Rosené 1990). Development policies to promote exports ought to be designed to meet the specific needs of small/medium producers (through risk minimization, export diversification alongside food production, use of labor-intensive production techniques). In cooperation with national farmers' organizations, mechanisms such as credit, service, and marketing cooperatives and other forms of producers' associations ought to be encouraged to enhance farmers' technical and marketing expertise and help to promote market diversification, especially into overseas areas.
An alternative, more broadly based development model for areas such as Central America will require a more flexible and diversified export platform in which comparative advantage is derived from an increasingly skilled workforce, an expanded domestic wage-goods sector, and other factors beyond those rooted in land concentration and absolute labor exploitation that propelled the old agroexport model. Mechanisms such as distributive agrarian reform and fiscal modernization need to be implemented to facilitate the transfer of profits to sectors and social classes beyond the agroexport elite. In addition, public investment priorities should shift toward providing a basic social and economic infrastructure (e.g., rural health care, education, technical assistance programs, roads, irrigation) that would permit wider economic participation and allow real living standards to rise via productivity gains without sacrificing international competitiveness. Unit costs of labor can decline for agroexports and other globally competitive economic sectors (allowing wages to rise alongside profitability) without reliance on capital-intensive technologies, if labor intensity improves (with income incentives) and the labor force become, increasingly skilled (as a result of better social infrastructure). An additional
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       49
advantage is that the labor force could also become increasingly flexible in its ability to adapt to new production sectors and techniques. This means that an expanded wage-goods sector (to provide income incentives) and social services sector (to heighten labor skills) should be seen as essential components rather than alternatives for a renovated export-led development model in areas such as Central America.
However, a new development model for Central America should promote enlarged and more equitable primary agroexport structures as well as opportunities for diversification into both domestic wage-goods and export sectors based in new areas such as non-traditional agriculture, manufacturing, and services. In fact, Costa Rica's relatively successful growth performance since the early 1980s (based largely on non-traditional exports) helps to illustrate the potential for increased economic diversification in the region. While there is nothing inherently wrong with increasing production from traditional agroexport sectors under the conditions outlined above, this does not mean that promising opportunities for diversification into non-traditional export sectors ought to be pushed aside.
In regions such as Central America, in which productive structures and class formations have historically been dominated by the agricultural sector, forms of economic diversification based on forward/backward linkages with primary agroexport production probably need to play a leading role, at least in the initial stages, in any viable development strategy. Moreover, in addition to increasing industrialization linked to agroexports, the development of internally oriented agro-industries should also be encouraged to supply a broader range of wage-goods and other basic needs for domestic markets. Research in other areas of the South has demonstrated that, among industrial sectors, agriculturally based manufacturing often has strong links with the local economy. For example, in an inter-industry study of different strategies of export promotion for India, Dholakia et al. (1992: M155) conclude:
if our objective is to generate high income effects without sacrificing the linkage effects on the rest of the economy so as to achieve diversified high growth in the system, the agri-based manufacturing sectors are obvious candidates for intensive export promotion measures.
The Need for Regional Cooperation and State Assistance
It is widely accepted that economic diversification can help to reduce the instability in export earnings that has plagued many Third World economies dominated by primary commodity production. However, new Third World exporters attempting to gain a foothold in global markets often face substantial entry barriers and other marketing constraints within
50                   MAINSTREAM THEORIES AND PRACTICES
overseas trade. This is especially true for many trans-oceanic marketing channels in which scale of production is important and where transnational capitals and their Third World affiliates have historically dominated trade (van der Laan 1993). In addition to such constraints, monopolistic and oligopolistic firms based in the North control most international marketing chains, especially for agroexports and other primary commodities, and appropriate much of the surplus produced by exporters from the South. In the case of the coffee trade between Central America and West Germany, for example, Meister (1991) finds that Central American producers receive only one-quarter of the economic surplus that they generate and that their production costs represent only one-sixth of the coffee's consumer price. These figures would seem to indicate that Central American producers might better improve their position within global commodity markets via collective negotiation along international marketing chains (perhaps with the support of producer associations, the state, and/or regional trading bodies) rather than by taking measures to reduce production costs, such as by lowering wages. Moreover, as Meister (1991) suggests, much room exists for Central American countries to work together, in cooperation with producer associations, to increase their share of selected global markets. With state assistance, action could be taken in the fields of advertising, improving export services, inspecting and certifying the quality of exports, and so on. Actions on these fronts might improve the image of Central American products in selected markets and encourage Central American exporters to pay more attention to the international reputation of their products.
Problems such as excessive surplus extraction and steep entry barriers within global marketing channels underscore the need for more basic research into the functioning of international markets, as a critical initial step for any export-led development strategy. Such research might uncover methods for Third World countries to exploit new export opportunities, as well as exposing constraints and limitations that need to be overcome in sectors that may at first seem promising. In the case of Central America, for example, there is a critical need to examine the nature of the marketing channels for traditional and non-traditional agroexports and related agro industrial products. Particular attention should be paid to the distribution of surplus between producers and merchants, on the one hand, and to opportunities for access to new markets, on the other.
In addition to exploring possibilities for expanding and transforming trade with the North, export diversification should also be encouraged by seeking new ways to stimulate South-South trading links. In many cases, creating means to strengthen regional trading blocs and common markets (e.g., the CACM in Central America) might further this goal. As is the case in Central America, such trading blocs have often existed lor prolonged
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       51
periods 'on paper' only, or in quite limited form. Finding methods to facilitate free trade within the regions of the South opens up possibilities for rapid market expansion for firms previously restricted to relatively small domestic markets. Moreover, enhanced regional economic cooperation that conforms to the 'logic of the majority' should allow for the mutually beneficial exchange of products according to various countries' factor proportions and areas of technological expertise. Trade ties with local NICs may prove especially important for opening up new opportunities for growth in many rural countries. For example, regional economic cooperation could facilitate the exchange of Central American agrarian-based products for capital equipment or new technologies from more industrialized regional economies such as Mexico and Venezuela. Many development analysts and organizations (notably UNIDO and UNCTAD of the United Nations) believe that enhanced cooperation among developing countries has an enormous and, as yet, largely unexplored potential for overcoming problems of scale in areas such as production, and research and development. Bagchi (1990: 412), for instance, cites the case of growing cooperation between Cuba and Mexico in selected fields of biotechnology as offering an appropriate example of mutually beneficial exchange between Third World countries with different factor proportions and areas of expertise.
In regions such as Central America where the market 'logic' of the old agroexport model has produced severe societal polarization, it should also be apparent that any revitalized export-led development strategy will require considerable state intervention, at least in the initial stage. The market will not guarantee the maintenance, let alone the increase, of export production. Neither will it permit, in its presently polarized form, increasing economic participation by disadvantaged classes and social groups. Without policies to assist, for example, disadvantaged small/medium rural producers, any new cycle of export-led growth in Central America will inexorably reproduce the exclusionary character of the old agroexport model. As in the past, export-led growth will conform to the narrow interests of an elite minority rather than to the broader needs of the popular majority.
However, the inclusion within a renovated export-led development strategy of measures designed to raise general levels of rural productivity and economic participation (e.g., improved primary and secondary education, technical assistance programs, health care), as well as more tailored programs to meet the special needs of small/medium producers (e.g., risk minimization, export diversification while maintaining food production, adoption of labor- rather than capital-intensive techniques) might encourage broader diversification into new, potentially profitable export sectors that until now have been the exclusive domain of large-scale capital. For most ol Central America, the correct package of selective policies must include mlisirihulivr agrarian reforms that can be compatible with an overall
52                MAINSTREAM THEORIES AND PRACTICES
strategy to stimulate agroexport production. Various forms of support will need to be extended to small/medium producers arranged in many diverse (individual and cooperative) forms of production and exchange. The effects of such support, in terms of productive employment and income generation, as well as access to land and other major means of production, are important considerations for this strategy. Finally, attention should be focused on creating new conditions for increased social harmony and political stability, without which any future development strategy, whether or not it contains a significant agroexport component, cannot be sustained.
Import-Substitution Industrialization
For much of the postwar period, import-substitution industrialization (ISI) occupied a prominent place within development theories, in addition to playing an important role in the practical development experiences of many Third World countries. However, in recent years ISI has been subjected to a withering attack from critics on various sides of development studies, particularly from many neoliberal strategists. They argue that the contradictions and shortcomings of previous ISI strategies have made it as a fundamentally flawed model for future Third World development. This critique, though, has not gone unanswered: a group of development theorists, especially from the Latin American neostructuralist camp, have begun to question the neoliberal depiction of ISI and whether this forecloses all prospects for the development of new ISI approaches. Such debate has once again opened up possibilities for the inclusion of a renovated ISI approach within the development strategy of countries that seek to balance export-led growth with a complementary inward orientation. Any renewed approach to ISI, however, must take into consideration the successes and failures of past ISI strategies, which will be examined in this section.
The Rise of ISI in Postwar Development Strategies
In the 1950s many mainstream development theorists, especially those who had adopted the new Keynesian growth models, began to turn away from primary exports and to see industrialization as the key vehicle to propel rapid development for the South, particularly in some of its larger, more developed economies. Industrialization was regarded as especially impoi tant for alleviating capital constraints, which were commonly viewed as the key roadblock to growth and development in most countries. Moreover,
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       53
there was widespread pessimism over the ability of exports to generate earnings fast enough to keep pace with the rapidly increasing import requirements of a modernizing economy. Possibilities for achieving more rapid and self-sustaining growth became linked with the rise of import-substitution industrialization, which would promote needed economic diversification while attracting direct foreign investment and concessional capital (i.e., aid) to spur growth.
The ISI strategy called for increasing production of manufactured goods for domestic consumption to nurture national markets. This would decrease external dependency and heighten self-sufficiency; absorb surplus labor, especially from the traditional agricultural sector; reduce balance-of-payments problems; foster more advanced stages of industrialization; and establish linkages with related sectors to encourage economic diversification. A mix of policies including fiscal incentives (e.g., low-interest loans, tax concessions) and protection from foreign competition (e.g., tariffs, quotas, licencing, exchange controls) was frequently put in place to promote ISI. In many countries, state-owned enterprises were created, either to carry out import-substitution directly or to support private capitals involved in this enterprise. ISI was also often combined with efforts by regional development authorities to integrate lagging regions into national economies via industrialization and the creation of backward linkages with primary sectors supplying industrial inputs.
In a few larger countries, ISI was focused on the creation of heavy industries, but more often it was chiefly directed at establishing light industries to supply intermediate and final consumption goods. In addition to typically high levels of state involvement, ISI also commonly attracted investment by foreign capitals designed to improve their access to local markets and circumvent mounting trade barriers. Nevertheless, foreign capitals often proved to be highly resistant to state efforts to regulate and shape the direction of ISI (e.g., through policies intended to promote internal forward/backward linkages, local reinvestment of profits, shared ownership and control by nationals, higher levels of local employment). In some cases, disputes over such issues led to foreign capitals being threatened with nationalization of their local assets; in rare instances, such threats were actually carried out.
By the 1950s, ISI had become the dominant development strategy in much of the Third World, especially for the larger economies of Latin America ■rid South Asia (e.g., Argentina, Brazil, India, Mexico, Pakistan). The social origins of ISI can be traced back to the dramatic downturn in international Commodity markets caused by the Great Depression (and extended by the Second World War), which undercut previous development strategies based on export led growth. In addition, ISI became associated in many countries with the increasing economic strength and ideological assertiveness of a
54                MAINSTREAM THEORIES AND PRACTICES
new, modernizing, urban-based bourgeoisie which was seeking to wrest economic and political power from more reactionary elements of the traditional rural oligarchy that had been linked to the old agroexport model.
In some Latin American countries (e.g., Brazil under the Vargas and Kubitschek administrations, Argentina under Perón), ISI played a key role in facilitating the state-led transformation of society in the interests of a new power bloc composed of an urban-based class alliance (led by the industrial bourgeoisie accompanied by labor, technocrats, the military, and other urban middle-class elements) (Cardoso and Faletto 1969). ISI often provided a valuable tool for the centralization and consolidation of power by this new industrially oriented hegemonic bloc. In addition, the state frequently viewed ISI as strategically important for increasing national self-reliance, particularly in many of the larger, more assertive Third World countries. ISI also had strong symbolic value for many countries in that it represented a further step along the widely acknowledged path to modernization.
By the end of the Second World War, ISI had provided the major Latin American countries with most of their basic consumption goods (e.g., processed goods, textiles, footwear, pharmaceuticals) and construction materials (e.g., cement, lumber, paints). During the 1950s and 1960s, many of the larger countries in the region also began import substitution in heavy industrial sectors (e.g., steel, basic chemicals), while smaller countries initiated ISI in the consumer goods sectors. During these initial stages of ISI, growth rates were typically high. Investment capital flowed into a series of technologically simple and relatively cheap industries that had ready (and protected) markets for an array of products. However, once these relatively easy gains had been acheived, ISI strategies began to run into severe difficulties. Since the early 1970s, ISI has come under increasing attack from both Left and Right, especially from neoliberals who associate it with excessive state intervention and the undermining of market-led, outward-oriented growth (see Harris 1986).
Support by the US and Transnational Capital for ISI
Most analysts seek to explain the rise of ISI by examining domestic political factors in Third World countries and/or the effects of global economic-changes on those countries. For example, neoliberals commonly portray ISI as a key component of protectionist development policies promoted by populist and nationalist coalitions in the South. However, this interpretation neglects the fact that in many countries the most dynamic (consumer durables, intermediate and capital goods) ISI sectors were dominated by foreign-owned transnational corporations (TNCs), often with the support
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       55
of considerable state subsidization.10 It also neglects the important role that US and other transnational capitals played in sponsoring ISI, particularly in the postwar period. Using case studies of the Philippines, Turkey, and Argentina to support their analysis, Maxfield and Nolt (1990: 78) find that 'US internationalist businessmen and government officials worked with local proponents of industrialization to shape the formulation and implementation of ISI policies.' Ironically, ISI was promoted by those sectors of US society commonly associated with advocacy of liberal trade policies: the executive branch of the state and transnational corporations. Moreover, these authors contend that ISI was an important US initiative and not merely a concession to Third World nationalists in the context of the Cold War. The US promoted ISI even in countries such as the Philippines where domestic support for the strategy was weak. In other countries, such as Turkey and Argentina, where some version of ISI would have been implemented in any case, the US encouraged the adoption of a limited version of ISI that would secure favorable conditions for US direct foreign investment. This ensured that any protectionism associated with ISI would not entail the loss of foreign markets for large US corporations, but would actually facilitate their globalization by providing preferential access to protected markets while excluding trade by other foreign competitors.
The Neglect of Agriculture
In addition to being promoted by the US government and transnational capital, ISI also received considerable theoretical support from First World academics. The principal focus of mainstream development models in the 1950s and early 1960s was almost invariably on industrialization as an essential aspect of long-run development.
Industrial expansion would protect developing economies from worsening terms of trade for primary products; it would also supply a more secure basis for steady growth based on economies of specialization and scale, technological transformation, and associated learning and demonstration effects. Third World development was basically seen as a transformative process from a traditional, agricultural, and rural economy toward a modern, industrial, and urban one. Industrialization was correlated with various benefits of development such as high employment and per capita income, while underdevelopment was seen as the legacy of insufficiently
111 Evans and Gereffi (1982: 138) report that in Latin America, for example, TNCs (especially from the US) rapidly consolidated their positions within the most dynamic ISI sectors of the largesl countries (e.g., in Brazil in the sectors of automobiles, pharmaceuticals, rubber, nonlerrous metals, electrical machinery and goods; in Mexico in   die   seitois   ni   i heinii als, . nibbei,   niiiirlci.li n ill   and   electrical   machinery,   and
ll.linpoihlllllll lipiipilll'lll)
56                 MAINSTREAM THEORIES AND PRACTICES
developed industrial sectors. Given the prevailing conditions in most developing countries, the transition to industrialization was normally thought to require an inflow of capital and technology from abroad, leading to increased global economic integration and higher levels of dependency by peripheral countries on the capitalist core. The principal source of growth in a succession of models was increasing capital stock based on industrialization, with a bias toward allocations to capital-goods production rather than the consumer-goods sector, including agricultural production. Since industrialization strategies were basically inward-looking, they spawned a whole generation of closed-economy growth models that demonstrated optimal capital deployment among economic sectors. The push was always on industry and away from agriculture.
If agriculture was given a role within mainstream development strategies, it was usually focused on providing support for the transition to industrialization. In addition to representing an almost limitless labor reserve upon which industries could draw, the peasantry was expected to contribute to national development through taxation and the provision of cheap foodstuffs for the urban population. Meanwhile, traditional agroexport sectors were heavily taxed in many countries by a series of state-initiated 'distortions' of both internal and external price structures (e.g., tariffs and exchange-rate policies, indirect taxes, quantitative and/or price controls, discriminatory interest rates) to prop up the ISI sector (e.g., Oliveira 1986; Thorp 1992).
The industrial bias of most mainstream strategies resulted in widespread neglect of agriculture except as a steady source of resources to be exploited for the industrial modernization process. In a study of postwar African development, Stewart (1991: 416) notes that 'the single most important policy mistake . . . was the neglect of agriculture, which received inadequate investment, research and development, infrastructure, and prices in most countries.' In the Middle East, Lawless (1988: 19) states that 'the emphasis placed on urban-industrial development as the path to modernization has been at the expense of agriculture, from which resources have been effectively transferred.' For Latin America, Kuczynski (1988) finds that the development bias toward industry and neglect of agriculture aggravated inequalities, contributing to an unusually skewed distribution of income, even by Third World standards. In India, the industrial bias of a succcs sion of postwar development plans is seen by Singh (1987: 231) as 'a telling example of an attempt to develop on modern Western lines without considering the socio-cultural relevance of the model to Indian conditions.' Today, many analysts contend that the neglect of agriculture in favor ol rapid industrialization in much of the South has contributed to a number of interconnected problems that many countries are experiencing, including growing inequalities in socioeconomic and regional terms, the stagnation ol
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       57
domestic food production, high levels of malnutrition and associated health problems, balance-of-payments shortfalls, and foreign indebtedness.
Ironically, while the agroexport sector was frequently subjected to discriminatory state intervention via exchange and tariff policies, it nevertheless remained an important source of capital for industrial investments within many ISI strategies. The dynamism of ISI in much of Latin America, for example, came to depend in large part on revenues generated by the major agroexport sectors. This intensified the vulnerability of many Latin American economies (e.g., in areas such as public-sector financing, industrial investment and employment) to external trade shocks resulting from fluctuations in the international commodity markets. Contrary to the intent of ISI to reduce dependency, this type of industrialization commonly functioned as a 'multiplier' of external shocks, thereby transmitting fluctuations in global market conditions to industries, the public sector, and other closely related activities such as construction (Rosales 1988: 32).
Principal Economic Problems of Traditional ISI Strategies
While structural transformation associated with industrialization may be an important goal of development strategies in many countries, this does not mean that any type of industrialization is appropriate under all circumstances. A review of postwar ISI strategies in the South clearly reveals that in many cases there were inherent problems in the very nature of the industrialization process. First, rather than diminishing dependence on imports, ISI often proved to be highly import intensive. New demands were created, especially among the middle and upper classes, for an array of industrial products with very high import coefficients. Many of these products were produced by transnationals and replaced goods manufactured by smaller domestic capitals that had been strongly linked with local rather than foreign suppliers. Moreover, overvalued domestic currencies, as well as discriminatory tariff and exchange-rate policies, tended to favor imported inputs for ISI industries while discouraging a broad range of exports. In many cases, this meant that ISI strategies actually consumed more foreign exchange than they generated, aggravating balance-of-payments and fiscal problems.
Secondly, ISI strategies often tended to concentrate on highly capital-intensive industries (e.g., automobiles, household appliances, petrochemicals). The technologies employed were, for the most part, borrowed directly from the North and were therefore also capital-intensive. Moreover, state policies linked with ISI commonly had the effect of making capital artificially cheap and thus further promoted the use of capital-intensive technologies. In thr hue ol chronic shortages of hard currency reserves in many countries,
58                   MAINSTREAM THEORIES AND PRACTICES
demands for major capital outlays to support industrialization led to heavy foreign borrowing, compounding problems of growing inflationary pressures and indebtedness. The labor-saving bias of capital-intensive production techniques in many ISI sectors also meant that employment creation failed to keep pace with rising output. Problems of broadening inequalities and increasing levels of unemployment and underemployment appeared in many countries, as the urban-industrial sector provided relatively few jobs for rural-urban migrants who were streaming into the larger cities. Furthermore, the concentration on capital-intensive technologies was inappropriate to the factor endowments of most Third World economies. Scitovsky (1984: 953) states that 'import-substituting industrialization seems to have meant concentrating on the activities in which the LDCs [Less Developed Countries] had a comparative disadvantage - as if the doctrine of comparative advantage had been stood on its head.' This contributed to a situation in which the bulk of the ISI sector in most countries had only minimal backward linkages with the remainder of the domestic economy, while, at the same time its industries were hopelessly uncompetitive in world markets and maintained persistently high levels of import dependence.
Thirdly, ISI was inherently limited in many countries by demand restrictions resulting from extreme income inequalities. Rural demand for industrial products was limited by the destabilization of small/medium farmers and rising exploitation associated with agroexport production. Urban purchasing power was held down by the lack of permanent industrial jobs and by the maintenance of low wages, both as a condition for attracting foreign investment and as a reflection of growing labor reserves. The lack of an indigenous capital-goods industry in most countries, coupled with the vertically-integrated and capital-intensive nature of many ISI sectors (especially those dominated by transnationals), further restricted employment provision, particularly of higher paying, more skilled jobs that could have fostered an industrial middle class to spur domestic demand. This meant that ISI production of consumption goods in many countries was either geared to supplying luxury items (e.g., automoblies, stereo sets, TVs, washing machines) for the fortunate few or found its internal markets for an array of more popular goods (e.g., shoes, clothing, bicycles, building materials) severely limited. Such demand restrictions were further compounded by the failure of various integration schemes (Andean Pact, Caribbean Common Market, Central American Common Market, Latin American Free Trade Area) to develop expanded regional and subregional markets for local products. As a result, ISI markets (especially in the smaller Latin American countries) remained much more socially and geographically circumscribed than was originally envisioned by Raul Prebisch and other theorists at CEPAL and related development insti tutes.
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       59
Political Crisis and the Demise of ISI
In the end, the demise of the ISI approach in areas such as Latin America was probably based at least as much in mounting political contradictions as in economic incongruities. O'Donnell (1975) argues that the exhaustion of opportunities for ISI growth in the 1970s was closely tied to increasingly authoritarian rule and political repression by many regimes in the region. Black (1991: 82-3) adds:
The real crisis of the ISI strategy was felt when it became clear that continual growth would be dependent upon expansion of the domestic market, and that such expansion implied a far-reaching redistribution of wealth and power. This recognition served to unite and mobilize elites - both domestic and foreign - who saw their interests threatened. The upshot was the suppression - in many countries through armed force - of effective demand and the adoption of a new strategy.
As ISI gained strength in Latin America, it became apparent that a new industrial fraction of capital would vie for state power in many countries with the traditional oligarchy based in agroexports and financial capital. Most development theorists, including those associated with CEPAL, believed that new urban-based populist coalitions led by the industrial bourgeoisie would gradually supplant the traditional elites. However, research into populist coalitions in, for example, the Southern Cone suggests that they were effective only when strategies were developed that accommodated or coincided with the interests of the old oligarchy (Cypher 1990: 49-50). The rise of ISI came during unusual historical circumstances that allowed for almost costless change. Rapid postwar economic growth supported structural change toward ISI without excessively affecting the profitability of the dominant (agroexport) sectors, and ISI provided benefits for important capitalist, military, and technocratic groups in many countries (Kaufman 1990: 129). In the 1960s and 1970s, however, when the postwar economic boom began to slow and ISI policies started to seriously impinge on the accumulation needs of important elements of the traditional oligarchy and foreign capital, populist coalitions supporting ISI were faced with disintegrating pressures that they could not withstand. Perhaps because of the optimism generated by the earlier period of relatively costless social transition, analysts at CEPAL and related institutes failed to realize the precariousness of the pro-ISI populist coalitions, due to changing conditions following the mid-1960s.
Critique and Defense of ISI in Development Studies Hy the curly 1970», a growing number of theorists from both Right and
60                   MAINSTREAM THEORIES AND PRACTICES
Left began to label ISI as an inherently flawed development strategy. From the Right, neoliberals linked ISI with excessive state interference in market mechanisms, and particularly with protectionist policies that had prevented countries from exploiting opportunities for outward-oriented growth based on comparative advantages. It was claimed that ISI had promoted neither efficiency nor equity. State policies associated with ISI were held fundamentally responsible for the economic stagnation and other related macroeconomic problems that were worsening in much of the South. From the Left, dependency and world-systems theorists contended that the neglect of needed changes in patterns of consumption and ownership fundamentally constrained ISI strategies. In addition, ISI was criticized for deepening the foreign penetration of Third World economies, encouraging the use of inappropriate technologies, and accelerating the net outward flow of capital toward the capitalist core (see Harris 1986).
Since the 1970s disillusionment with ISI strategies has spread across a broad range of ideological and analytical perspectives within development studies. However, a group of (largely neostructuralist) theorists have also presented a strong case that the failures of ISI were essentially attributable not to the thrust of ISI strategy itself, but to the contortion of its policies by the peculiar politico-economic structures that prevailed in many Latin American countries (e.g., Dietz and James 1990; Gereffi and Wyman 1990). According to this line of thought, many of the excesses and misdirected policies of ISI were inconsistent with ISI theory as represented, for example, by the work of Prebisch and others at CEPAL (e.g., Kay 1993; Rosales 1988; Thorp 1992). There is considerable evidence that these early theorists, rather than arguing for an extreme type of closed, inward-looking economy, advocated a moderately protectionist model with a strong emphasis on increasing efficiency and technical progress in ISI sectors that would eventually lead to enhanced international competitiveness and external openness.11 Although the internal market was important to the CEPAL model of ISI, the model was not exclusively inward-looking. It was hoped that temporary protection would stimulate new export possibilities and that endogenous sources of productivity growth would allow for rapid development that would be compatible with goals of both increasing autonomy and enhanced competitiveness in global markets according to
»I Thorp (1992: 189) notes that the phrase 'desarrollo desde adentro' (development from within) occurs in Prebisch's early writing, in contrast to the later description of 'desarrollo bacia adentro' (inward-oriented development) that is usually attributed to him and other CEPAL theorists in the development literature. Rather than simply advocating an inward-oriented closed economy, the phrase desarrollo desde adentro expresses the idea that the Latin American countries ought to develop in a way that reinforces their internal capacities, respects their autonomy, and builds up their comparative advantages as gradual integration into the global economy ukrs place on terms favorable to Latin Amenta itself rather than to the TN< :• (mm lbe Nonli
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       61
long-run comparative advantage.
For many neostructuralists, the contortion of the original CEPAL model by Latin American countries and their consequent failure to sustain economic growth under ISI were based in the nature of political and economic power in the region. This was especially evident in the continuing dominance of ruling elites whose interests conflicted with an ISI-inspired social transition. Fajnzylber (1990: 335), for example, castigates the wasteful 'showcase modernity' of Latin America and compares it with the more progressive 'endogenous modernity' of the East Asian NICs. Greater societal polarization, less distributive equity, lower international competitiveness, and a short-term orientation toward consumption by urban-based elite groups characterize the former. By contrast, higher levels of societal integration and distributive equity, increasing international competitiveness, and a longer-term orientation toward savings and investment to meet strategic development goals characterize the latter. Calling into question the neoliberal assumption that 'excessive demand' by the subordinate classes in Latin America was chiefly responsible for the inflationary pressures and other macroeconomic problems that plagued the ISI model, Fajnzylber (1990: 325) contends that it was the ruling elite, not the popular sectors, that were mainly responsible for the wasteful consumption that distorted development in the region. For Fajnzylber and other neostructuralists, the eventual demise of ISI became a virtual certainty in the absence of a profound crisis in Latin America that would have forced change upon its reactionary elites and their overseas allies. The region's recent poor development performance 'was not caused by failing to "get policies or prices right," but rather by not "getting politics right"' (Dietz and James 1990: 203).
A Reconsideration of ISI
In retrospect, the branding of ISI as an inherently flawed strategy by many development theorists on both the Right and the Left has probably been an overreaction. Under ISI many countries achieved dramatic increases in industrial production, fostering the establishment of substantial industrial sectors to supply the local market. In many cases, ISI improved the economic prospects of an increasingly influential urban-based class alliance. This allowed state structures to be formed from a socially cohesive base that, for a time at least, generated rapid growth with relatively little class conflict. LSI was also chiefly responsible for the emergence of urban mass markets in some Third World countries, permitting more easily sustainable growth in the face of global economic downturns. In addition, substantial state outlays on social and economic infrastructure during the ISI period provided many countries wilh the foundations for future economic diversification and drvrlopmml,
62                 MAINSTREAM THEORIES AND PRACTICES
lor industrialization to proceed to more advanced stages, a base needs to be created through ISI (or some other equivalent program) of appropriate social, technological, and economic infrastructure. Modern industrial sectors are not established overnight, but are the end result of long periods of broad societal structural transformation. Considerable periods of learning and adaptation are normally required to make technology transfers and other aspects of industrialization appropriate to particular historical conditions prevailing in different countries. Import substitution may provide the means for countries to carry out this type of learning process. It may also offer domestic firms the opportunity to achieve scale economies upon which to eventually construct a viable export platform. Contrary to neoliberal opinion, much of the recent export-led development of, for example, the East Asian NICs such as South Korea and Taiwan grew from the industrial foundations originally established under ISI.12 Thus, the way in which the inward-oriented ISI model is juxtaposed to the EOI (export-oriented industrialization) strategy in much of the development literature (e.g., see Chandra 1992) may be a false dichotomy that does little to further constructive debate. Rather than being an either-or proposition, the two approaches may in fact be complementary, as the development experience of the East Asian NICs demonstrates. Import substitution does not preclude rapid growth and the development of a strong exporting capability. Finding the appropriate mix of ISI and EOI to suit particular countries under changing global conditions might be a more appropriate focus for development.
Necessary Components of a Viable ISI Strategy
Hulme and Turner (1990: 105) claim that policy-makers currently face two principal alternatives concerning the future of ISI. On the one hand, they can follow the prescriptions of the neoliberals for relaxing trade restrictions and other forms of state intervention that have traditionally propped up the ISI sector, thereby forcing its industries to become more efficient. On the other hand, they can follow the advice of the dependency theorists who call for significant changes in economic structures to accompany the deepening of ISI (e.g., through agrarian reforms, measures to redistribute income). Although each country must find its own appropriate path to development based on its unique local conditions and changing global circumstances, development for much of the South depends on the adoption of a hybrid approach composed of elements of both these alternatives.
12 The transition of the East Asian NICs and some other countries from an initial plntm-of ISI to a later period of l-'OI (export-oriented industrialization) will be covered hi more detail in the following section and especially in chapter 1, which Ionise» on ihr NIC experience.
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       63
In order to be sustainable, growth has "to be based on the efficient use of capital and other scarce resources. Infant industries may require an initial period of protection but this should be gradually reduced within the framework of a strategic economic policy to ensure long-term efficiency and global competitiveness. A policy of rational and selective protection would allow for the takeoff of new ISI projects in sectors of strategic importance for structural economic change (e.g., capital-goods sectors), while eliminating problems of excessive, permanent, or indiscriminate protection. A process of selective ISI might also be used to create links within key productive chains that could promote intra-industrial and inter-sectoral articulation. In many countries an endogenous industrial sector could be strengthened and better articulated to the rest of the local economy, which could ensure production of basic goods and widely used inputs at the same time as it acts as a springboard for expanded exports. The build-up of a viable ISI sector should be recognized as particularly important for countries with relatively large internal markets (e.g., Brazil, India, Mexico). In these countries, development strategies oriented almost exclusively toward the external market (as in Hong Kong and Singapore) are inappropriate. Production for the domestic market may permit the reduction of unit costs through the achievement of scale economies, allowing a country to establish a solid foundation from which to enter export markets. It may also foster socioeconomic articulation, permitting more efficient use of small/medium enterprises, which is vital for job creation and income distribution.
In general, measures to facilitate the economic participation of different classes and social groups, economic sectors, and geographic regions must also accompany industrialization so that it conforms to majority interests rather than to those of a narrow elite, as has so often happened within ISI strategies. For the great majority of Third World countries, either individually or collectively within regional trading blocs, ISI ought to play an important role in promoting goals of both economic efficiency and distributive equity. The establishment of a viable ISI sector to meet majority interests requires accompanying programs that simultaneously increase internal demand and raise standards of living among the popular sectors. Production ought to be particularly encouraged in sectors (e.g., basic goods, housing) that have strong multiplier effects on other industrial and primary sectors. In addition, various programs need to be implemented to open opportunities for expanded participation of small/medium firms in the industrial sector. This might involve, for example, initiatives to expand llibcontracting arrangements between such firms and larger enterprises, both public and private. Finally, however, it should be remembered that this type ol broadly based ISI program oriented toward majority interests may present ,\ fundamental challenge to the status quo in many highly polrtľi/ľd sonnies, (itven the past record of ISI, one might expect under
5240
64                  MAINSTREAM THEORIES AND PRACTICES
these circumstances to encounter strong opposition from the ruling elite, which will feel its interests threatened and will throw its support behind contrasting (at present neoliberal) development strategies.
The Neoliberal Stress on Export-led Growth
As was related at the beginning of the previous section, a strong current of 'export pessimism' pervaded mainstream development theory during the early postwar period. Many influential development theorists and policymakers contended that global trade, especially for primary commodities, was too erratic to form the principal 'engine of growth' for Third World economies. Instead, it was believed that ISI would offer a more secure and orderly basis for the generation of sustained growth. However, since the late 1960s support for ISI among mainstream theorists has gradually given way to a renewed emphasis on export-oriented industrialization (EOI) and other forms of outward-oriented growth. This shift in development thinking has paralleled the resurgence of neoclassical economics as the centerpiece of the neoliberal counterrevolution in development studies.13 It has also accompanied increasing interventionism by the IMF and World Bank into Third World policy-making via mechanisms such as structural adjustment lending. Generally, continued access to such lending, as well as to most other external sources of financing, has been made conditional on the adoption of policy reforms designed to reduce state economic intervention and generate market-oriented growth. In many countries such pressures have contributed to a decisive shift in development strategy away from export pessimism and ISI toward an optimism for the prospects of EOI and other forms of export-led growth.
The Neoclassical Theory of Export-led Growth
Rising support for EOI and export-led growth within mainstream development theory is based on seven interrelated arguments based in neoclassical theory. First, given low levels of domestic demand in many developing countries, growth in a range of (especially industrial) sectors is believed to be largely dependent on gaining access to global markets via export-oriented
13 Among the first to criticize ISI were a group of neoclassical theorists (e.g., Vinncr, Haberler, Bauer) who argued that ISI interferes with the 'natural process1 of development based on comparative advantage. Their view, which has remained populai among neoliberals, was that Third World countries, at least during their initial stages of development, should uniformly specialize in primary exports rather than attempt to develop more sophisticated industrial sectors through state intervention that would not conform to comparative advantages based on factor proportions.
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       65
trade strategies. Second, export-oriented policies are regarded as normally the least damaging in terms of microeconomic efficiency - in that they benefit total factor productivity more than any other popular policy option. Third, foreign trade multipliers associated with exports are thought to play an important part in facilitating long-term growth by expanding overall production and employment. Fourth, earnings from exports may foster macro-economic stability by contributing to a more favorable balance of trade and external accounts, which is important for attaining better ratings in international financial markets (and thus easier access to foreign loans and investment capital). Fifth, export earnings may also provide foreign exchange for imported goods, particularly capital goods needed to increase the production potential of an economy. Sixth, rising export volume and competition within global markets are believed to create economic efficiencies associated with increasing scale economies and technological diffusion. Seventh, given these theoretical arguments, rapid economic growth among (especially East Asian) export-oriented NICs, as well as a series of country studies showing strong correlations between exports and economic performance, is interpreted as empirical evidence supporting the export-led growth hypothesis.
The Rise of Industrial Exports in Some Third World Countries
In recent years, export-led growth has been especially strong among Third World countries that have managed to erect industrial export platforms. Reversing earlier trends, both the output and export of manufactures have grown more rapidly since the 1960s in the developing countries than in the industrialized countries (see tables 2.1 and 2.2). Growth in manufacturing production among developing countries was 9.0 percent in 1965-73 and 6.0 percent in 1973-85, while among the industrial market economies it was 5.3 percent and 3.0 percent in the same two periods, respectively. This allowed the developing countries to increase their share of global manufacturing production from 14.5 percent in 1965 to 18.1 percent in 1985, while the share of the industrial market economies slipped from
85.4  percent to 81.6 percent over the same period. Similarly, growth of manufacturing exports among developing countries was 11.6 percent in 1965-73 and 12.3 percent in 1973-85, while among the industrial market economies it was 10.6 percent and 4.4 percent in the same two periods, respectively. As a result, the developing countries increased their share of global manufacturing exports from 7.3 percent in 1965 to 17.4 percent in  1985, while the share of the industrial market economies fell from
92.5   percent to 82.3 percent over this period. However, such growth has been quite unevenly distributed within the Third World. A relatively «mull number o( 'middle-income' developing countries (especially the newly established NICs) have dominated growth in manufacturing exports, while
A2B
STRATEGIES OF GROWTH AND INDUSTRIALIZATION        67
a much larger number of 'low-income' developing countries have lagged behind. Growth in manufacturing exports among middle-income countries was 14.9 percent in 1965-73 and 12.9 percent in 1973-85, while among low-income countries it was 2.4 percent and 8.7 percent in the same two periods, respectively. Consequently, the middle-income countries dramatically increased their share of global manufacturing exports from 5.0 percent in 1965 to 15.3 percent in 1985, while the share of the low-income countries fell slightly from 2.3 percent to 2.1 percent over this period.
Export-led growth has been fueled in an increasing number of Third World countries by the attraction of TNCs and other capitals to export-oriented zones (EOZs),14 which permit goods to be shipped to receiving (usually First World) countries without being subjected to prevailing tariffs, duties, and other forms of trade interference. In 1970, there were about 20 EOZs in ten Third World countries; by the late 1980s, more than 260 EOZs had been created in over 50 countries (UNCTC 1988: 169-72). The growth of EOZs has been particularly dramatic among the NICs of East and Southeast Asia and Latin America. In Mexico, for example, employment by maquiladoras (TNC branch-plants located in EOZs) jumped from 123,000 in 1982 to 412,000 in 1989, representing one of the swiftest rates of growth of manufacturing employment in the world during the 1980s (Harris 1991: 121). By 1990, Mexico had approximately 1,500 maquilas producing annual export earnings of about $2 billion (Sklair 1990: 112). In the early 1980s, most of these maquilas were relatively small plants owned by smaller US firms. Production was concentrated in low-skill operations in labor-intensive sectors and employed a high proportion of unskilled (particularly female) workers. Recently, however, average plant size has grown with the increasing participation of larger TNCs (including those based not only in the US, but also in Europe, Japan, and other Asian countries). Production has also begun shifting to more capital-intensive sectors that require a more highly skilled workforce in order to enhance the quality of output and allow more sophisticated '0-error' delivery systems to be implemented (Harris 1991).
Industrial Phases and the Internationalization of Production
Although the neoliberal development literature commonly contrasts (successful) EOI strategies with the (failed) experience of ISI in the Third World, in most of the NICs ISI and EOI are better conceptualized as two phases within the same dynamic process of internationalization of
n These /.ones arc sometimes also called 'free-trade zones' (FTZs) because they allow exporters ireer access to normally restricted overseas markets. Another term for these wines i» 'export-processing zone*' (HI'Zs), which emphasizes their common orientation loward activities ol final assembly or processing ol inputs produced elsewhere.
68                 MAINSTREAM THEORIES AND PRACTICES
the circuit of productive capital (Bina and Yaghmaian 1988).15 Existing evidence indicates that ISI, rather than promoting 'self reliance,' has further integrated developing countries into the network of global production. However, because ISI was restricted to separate local markets, it eventually proved to be a constraint to the unification of the world market and the ability of capital to accumulate on a global scale. EOI has been the response of international capital to the need to develop a more unified network of capitalist production and exchange. According to this theory, ISI represents a prelude to EOI and the attainment of a higher level of global integration. In a sense, it set the stage for EOI and the internationalization of production for the world market. Many of the most successful Asian and Latin American exporters of manufactures (e.g., Brazil, Mexico, South Korea, Taiwan) developed their export platforms on the foundation of industries established during the import-substitution period. The basic difference between ISI and EOI, then, lies in the location in which their circuits of money capital are completed. For ISI the circuit is completed in the local market; for EOI it is completed externally by the realization of the value of domestically produced manufactures on the world market.
The Focus on Primary Exports in Most Third World Countries
Some highly publicized export-led growth strategies have focused on industrial exports by a small group of mostly middle-income developing countries, particularly the NICs. However, for a broader group of Third World countries, export-led growth has been concentrated in a few primary products. In many countries, the export structure remains essentially the same as it was in the (neo)colonial era under the agroexport model. This is especially true for sub-Saharan Africa, where only a few countries (notably Botswana, Mali, Mauritius, and Zimbabwe) have managed to diversify exports away from primary products during the last two decades. In Africa's low-income countries, the proportion of primary products within exports actually rose from 92 percent to 94 percent between 1965 and 1987, while it declined only slightly from 95 percent to 90 percent for the region's middle-income countries (Stewart 1991: 425).
Neoliberal development strategies associated with SAPs have encouraged export specialization in primary products for many Third World countries by expanding incentives for traditional exports. However, continuing export specialization in primary products historically has severely restricted efforts
15      These authors theorize that the progressive internationalization of capital has chartu teristically followed three stages: an early stage of primitive accumulation, a mote advanced stage of primitive accumulation and ISI, and a  Hnul staue of export led industrialization and global integration. ISI and KOI thus represent the final two wages of this process.
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       69
to broaden development in social, sectoral, and regional terms. Trade strategies concentrated on primary products have also commonly suffered from low demand and supply elasticities and a long-term decline in terms of trade in global markets. In the early postwar period, Prebisch (1950) and others linked worsening terms of trade for Third World primary commodities to an asymmetry on the demand side: exports from the South have a lower income elasticity relative to those from the North. However, more recent research by Krugman (1987) has also uncovered asymmetries on the supply side. The industrial sector in the North has increasing returns to scale (based on superior capital endowments and lower average costs in producing industrial goods), while the primary sector in the South has only constant returns to scale. Free trade between the North and South, then, will not only produce erratic growth and worsening terms of trade for the South, but will also lead to its systematic deindustrialization, as it finds itself caught within a low-level equilibrium trap that prevents entry into more sophisticated economic sectors (see Eswaran and Kotwal 1993). Countries that fail to diversify exports beyond unrefined primary products are denied access to expanded employment opportunities and value-added, as well as vital learning processes associated with the production of higher-level goods. Experience gained in the production of more sophisticated goods results in the accumulation of labor skills and human capital, as well as general improvements in production techniques and organizational methods. Moreover, the effects of this learning process tend not to be confined to the sectors in which they originate, but spread through spin-off effects to other sectors as well. In recent years, an international division of labor based on the relative sophistication of production sectors has grown not only between the North and South, but also among developing countries themselves. A few (middle-income) countries have been able to stimulate development by diversifying production, especially of their leading export sectors, into more sophisticated and profitable economic activities. The bulk of (poorer) countries, however, remain specialized in production and export of a restricted number of primary products that, for the most part, have performed poorly and offer few prospects for breaking out of an inferior position within an increasingly rigid international division of labor.
Exploitation, Polarization, and Repression Accompanying Export-led Growth
In many cast's, increasing exploitation and repression have accompanied export led development, particularly of primary commodities. In this way lite new w.ivr oi export led nrnlihrrulism closely resembles more traditional
70                MAINSTREAM THEORIES AND PRACTICES
outward-oriented development strategies such as the agroexport model: the comparative advantage of their major export sectors is commonly derived from similar processes of absolute exploitation and monopolization of land and other means of production. The social marginalization that normally accompanies such processes means that repression is commonly applied on a massive scale to enforce labor discipline and ensure overall systemic stability. Within this context, Hettne (1990: 31-2) reports that: 'The comparative advantage will be with the most aggressive and repressive nation states, at least until their own disintegration sets in.' Indeed, programs of liberalization and export-led growth in areas such as Latin America have characteristically attacked workers' rights and benefits accumulated over previous decades. Institutions and mechanisms that had been established to protect workers and other members of the popular sectors have been systematically undermined in many countries. Banuri (1991: 193) argues that these policies represent an effort to resolve the incompatibility between the logic of export-led development and a persistently high degree of worker mobilization in areas such as Latin America. Beginning in the early 1970s, attempts to spur export-led growth through liberalization measures have been marked by increasing societal polarization and confrontation in many countries (e.g., Argentina, Chile, Peru) leading to the frequent employment of brutal repression by the security apparatus.
If stess is not placed on the creation of local linkages to spread the benefits of growth in social, sectoral, and regional terms, neoliberal export-led strategies risk replicating the vicious cycles of polarization and repression so commonly associated with past export-oriented development models. Under the guise of the new export-led models of the neoliberals, countries risk creating new and more sophisticated forms of polarized and dependent development. What is missing from strategies that focus only on increasing exports is a concern for the broader development goals of raising living standards of the popular majority and promoting more balanced growth among different economic sectors and geographic regions. In the absence of well-developed linkages between the export sectors and the rest of the economy, a limited and polarized form of development takes place thai cannot act as a stimulus for overall growth and development to serve-majority interests.
Issues and Questions for Export-led Growth Strategies
A set of criteria may be developed to evaluate the effects of export led growth on overall development. These might include: the extent of linkages to the domestic economy; the creation of employment and value-added; the effect on external accounts and balance of payments; the fostering ol genuine and appropriate technology transfer rather than merely technology
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       71
relocation; the generation of jobs for skilled labor as well as for local managers, technicians, and other highly trained personnel; the establishment of favorable wages and working conditions relative to those prevailing in the country; and the rise of a relatively equitable social, sectoral, and regional distribution of the costs and benefits of growth. This would mean that maldevelopment accompanying export-led growth might be associated with some combination of: the destruction of internal linkages in the domestic economy; the failure to create satisfactory levels of local employment or value-added; the worsening of balance-of-payments problems and foreign indebtedness; the transfer of inappropriate (often capital-intensive) technologies developed for First World rather than Third World factor intensities; the loss of local skills and the failure to create skilled jobs for the local population; the intensification of labor exploitation; and the inequitable distribution of the costs and benefits of growth.
Reports of many of these problems appear with disturbing frequency in the development literature devoted to analyzing export-led growth in developing countries. At a general level, Black (1991: 85), for example, notes 'the failure of the [export-led growth] strategy to promote balanced and equitable growth in most Third World countries.' Similarly, Fröbel, Heinrichs and Kreye (1980) argue that export-led growth, especially that associated with EOZs, has produced only a truncated, severely circumscribed type of development that has excluded the majority from participating in the benefits of growth. Following an analysis of EOI in Mexico and China, Sklair (1990: 124) concludes that 'open-door strategies seem to offer a way out of the awful dilemma between dependency without development and capitalist development without social justice but. . . there is little evidence to suggest that this is anything more than a false promise in the interests of transnational capital and its partners, capitalist or otherwise, in the Third World.'
In Africa, Saha (1991: 2760) finds that liberalization and structural adjustment measures designed to promote primary exports have deepened the underdevelopment of other economic sectors and, most troublingly, have hastened a destructive process of deindustrialization in many countries. For India, Krishnaswamy (1991: 2417) reports that recent liberalization policies threaten 'the very fabric of the Indian nation' through an excessive centralization of economic decision-making, the distortion of democratic institutions, and the neglect of the bulk of the economy that lies outside of A few modern industrial sectors. Indeed, the relatively successful experience of a few (especially East Asian) NICs with export-led development is the exception rather than the rule for the developing world. The reality for the rest of the Third World is much more problematic.
Development in general and export-led development in particular can only   he  understood  within  a specific  historical  context.  Absent from
72                 MAINSTREAM THEORIES AND PRACTICES
neoliberal development studies, particularly those based on IMF/World Bank models of structural adjustment, liberalization, and export-led growth, are country-specific analyses of class and other social relations. This omission is a product of the limited scope of neoclassical theory, and of the methods of positivist science in general, which prevent a detailed consideration of questions of class, gender, ethnicity, and other social relations. This represents a serious deficiency in the neoliberal approach since various forms of development ultimately are not just about abstract policies and empirical variables but, more fundamentally, are about class and other social relations within particular historical circumstances. Not all types of export-led development (e.g., agroexports, EOI) entail the same consequences for different social sectors. Indeed, similar export products may be produced according to different social relations, some of which may foster balanced development in majority interests, while others may promote polarized development for an elite minority. The impacts of a particular development strategy may differ dramatically as change takes place within the social structure of individual countries and in the position of these countries within an evolving international division of labor. Policies which may have had a positive impact in a particular country at a certain time may produce quite different results in another country or in another era. Likewise, policies that may produce satisfactory results at a macroeconomic scale, may have adverse effects on various social sectors and non-economic aspects of development. In order to make such distinctions and understand the underlying social dynamics, we need to deepen our analysis well beyond the narrow confines of neoliberalism based in neoclassical theory and positive science.
Non-traditional Exports: A New Growth Sector
With the renewed emphasis on outward-oriented growth which has accompanied the rise of neoliberalism, increasing attention has been focused on non-traditional exports (NTEs) as an important potential growth sector for many countries.16 Indeed, recent strategies to promote growth of NTEs have proved remarkably successful in a growing number of countries, prompting much imitative behavior in many others. However, there have also been some common problems linked to NTE promotion, which call into question its usefulness as a major component of development strategies. These include foreign domination and dependency, socioeconomic and spatial polarization, environmental destruction, cultural alienation, and
16      Another important new growth sector for many countries is tourism, which is itnnlyrril in Brohman (1996).
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       73
low levels of popular participation. This section analyzes such problems and explores ways in which they may be overcome by introducing changes in development policies. The design of alternative policies that call for increased popular participation and more coordinated state involvement in various aspects of development planning is emphasized.
Basic Elements of a Non-traditional Export Strategy
The theoretical case for the promotion of NTEs is based on the contribution that they can make to the development of a large, diversified trade sector to propel outward-oriented growth (e.g., Derosa 1992; Lücke 1993). Small trade sectors limited to a narrow range of exports based on resource-intensive products have severely circumscribed development in most Third World countries. Many of these products (especially traditional agroexports) have suffered a long-term decline in international terms of trade and have experienced severe limitations with respect to both supply and demand, affording few possibilities for export expansion at the margin (Eswaran and Kotwal 1993). At the same time, export expansion is necessary in most countries to maintain macroeconomic balance and to offset the cost of necessary imports, particularly of capital goods needed to keep industries and other domestic sectors functioning. A large, diversified export sector is also critical for generating growth in developing economies with small or underdeveloped internal markets. In addition, it permits countries to adjust more easily to periodic trade shocks caused by fluctuations in global commodity markets. If a country can diversify into a broader range of exports so that the variability of earnings from one subset of exports is largely offset by that from another, then that country will tend to face less uncertainty in its ability to finance imports and other necessities for development.
In global terms the strategy focuses on enlarging and diversifying the export sector, especially by exploiting niches in international markets according to a country's particular comparative advantages based on its factor proportions. For a few NICs (especially in East and Southeast Asia) this has meant rapid industrialization via export substitution; growth has been based on a growing number of increasingly sophisticated manufactures. However, for the bulk of Third World countries this strategy has concentrated on complementing more traditional resource-based exports with non-traditional agricultural exports (e.g., off-season vegetables and tropical fruits, ornamental flowers, specialty nuts) and/or low-level manufactured goods assembled in 'final touch' industries. Common sources of comparative advantage contributing to low production costs and high profitability among these export sectors include: inexpensive and abundant land and nitlurnl resources; a relatively cheap, unorganized, and compliant
74                 MAINSTREAM THEORIES AND PRACTICES
labor-force; the lack of labor, environmental, and other state regulations concerning production; and relatively low rates of taxation.
Given these sources of comparative advantage, the strategy calls for specialization in low-wage, labor- and land-intensive export sectors. In order to exploit opportunities for growth based on the principle of comparative advantage, new neoliberal development strategies maintain that trade restrictions ought to give way to liberalization and that macroeconomic policy should provide incentives to move resources from non-tradable sectors to tradable (i.e., export) sectors. Essentially, growth is to be export led; production of manufactures, food, and other goods for the internal market should occur only when domestic producers can successfully compete against importers without subsidies, duties, or other forms of state protection.
Examples of NTE Growth
Among African countries, Zimbabwe has enjoyed especially rapid growth in agricultural NTEs during the 1980s and 1990s. The growth of Zimbabwe's NTEs has been led by flower exports to Europe, which are based on Zimbabwe's comparative advantage of a (southern hemisphere) growing season that coincides with the European winter when flower prices there are at their highest. By the end of the 1980s, horticulture production in Zimbabwe had risen from almost nothing at the start of the decade to about 9 percent of total agricultural output. The value of Zimbabwe's flower exports increased sixfold between 1985 and 1989, easily representing the most rapid rate of NTE expansion in the country (Smith 1990: 160-1). The Zimbabwean government has been quick to recognize the potential of this sector not only as a source of foreign exchange, but also as a creator of rural employment. It is estimated that horticultural crops create an average of 0.7 full-time and 2.0 seasonal jobs per hectare, one of the highest employment: land ratios in Zimbabwe's agricultural sector (ibid.:162).
In Latin America, Chile and Costa Rica have been among the most sue cessful non-traditional agricultural exporters during the 1980s and 1990s. In contrast to the historical domination of their economies by traditional resource-based exports (copper in Chile, and bananas and coffee in Costa Rica), much of these countries' export growth since the early 1980s has been based on NTEs. Like Zimbabwe, both countries began exploiting comparative advantages derived from their southern locations to ship high-value tropical vegetables and fruits, nuts, and horticultural products to North America during its winter months. Other important exports, primarily also for the North American market, have been wine from (lliilr and pharmaceuticals and seafood from Costa Rica, ľrom 1984 to ľ>8'> NTE growth was .348 percent in Costa Rica and 222 percent in Chile,
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       75
representing an annual rate of growth of 28 percent and 17 percent, respectively (Barham et al. 1992: 49). Rapid growth of NTEs was a key factor in propelling the economies of these countries to among the highest rates of growth in the region during this (recessionary) period: the annual rate of GDP growth for 1984-89 was 4.0 percent in Costa Rica and 6.4 percent in Chile.
Common Problems of NTE Strategies
Given the success that a growing number of countries have enjoyed with NTEs, many Third World governments have begun turning to non-traditional exports in order to restimulate the growth that both traditional exports and domestic market expansion have failed to generate in recent years. NTE expansion has been made a key element in the neoliberal growth strategies of many countries and, as such, has often been actively promoted (in terms of inspiration, resources, managerial direction, etc.) by bilateral (e.g., US Agency for International Development (USAID), Canadian International Development Agency) and multilateral (e.g., World Bank, Inter-American Development Bank) aid and lending agencies. Intellectual support for the inclusion of NTEs as a critical part of most development strategies has also come from the community of development scholars (e.g., Hiemenz 1989; Paus 1989; Pelupessy 1991b).
However, there is also growing evidence that a series of problems commonly associated with the more traditional agroexport model is being replicated by NTE expansion. These problems include: the progressive concentration of land and other major means of production among a narrow minority; rising inequalities and dislocations, especially in rural areas; the inability of small/medium farmers to participate in the programs without state support; and increasing dependence on First World markets that are subject to wide fluctuations.
Many of these problems seem to be appearing in the very places that are being portrayed as successful cases of NTE expansion in the neoliberal development literature. A good example is in Costa Rica, where NTE expansion has been propelled by a new program of rural development, called Agricultura de Cambio (Agriculture of Change), supported by the itate and external organizations such as USAID and the World Bank. In announcing the program, former Costa Rican President Arias said: 'We arc concerned about the situation of the small producers. We know they need more help, they are the base of our democracy' (Desanti 1988). Under the program, however, poverty has increased dramatically in Costa Rica's nnun agricultural /.ones. One report states that over 80 percent of the rural population now lives below the poverty line (Rosené 1990: 371). Land tuiuenii.iiiiins, many of which have accompanied property purchases by
76                  MAINSTREAM THEORIES AND PRACTICES
foreigners, have displaced increasing numbers of peasants from traditional production areas of basic grains and other foodstuffs for the domestic market. NTE growth has been concentrated among larger, more affluent farmers that enjoy considerable advantages in important areas such as access to capital and bank credit, technical assistance programs, and expertise in marketing and import/export activities. State provisions of rural credit and technical assistance have increasingly been shifted away from encouraging peasant diversification, toward the further expansion of large-scale NTE operations (Lowder 1990: 97-8).
Without sufficient access to credit, technical and marketing assistance, or other forms of state support, small/medium farmers who have tried the Agricultura de Cambio have had very high failure rates. In an article quite critical of NTE development and the Agricultura de Cambio program in Costa Rica, Rosené (1990: 373) states:
The Agricultura de Cambio program seems to be working directly against the small producers with the apparent intention of forcing them off the land to become cheap laborers in agribusiness farms or in the assembling and service industries . . . According to the farmers, the program will end up creating a dangerously skewed land tenure system. It will destroy Costa Rica's ability to feed itself, thus creating more dependency. It will increase social tension, threatening the stability that Costa Rica has enjoyed for so many years . . .
Evidence points to a similar process of concentration and centralization of capital accompanying the rise of NTEs in other Third World countries. In Zimbabwe, for example, differential access to capital, technological expertise, and forward/backward linkages has accentuated the concentration of horticultural exports among large-scale producers. The dominant position of the agrarian bourgeoisie in NTE expansion was solidified in 1986 with the establishment of the Horticultural Promotion Council by the Commercial Farmers Union, which represents the interests of large-scale farmers. Among its various activities, the HPC officially represents horticulture at the government level, organizes technical and marketing assistance, coordinates requests for foreign exchange to purchase inputs, and negotiates shipping arrangements with air carriers (Smith 1990: 161-2). Without a similar association, small and medium producers face considerable technical, organizational, and financial barriers that prevent entry into expanding NTE markets.
In Chile, the benefits of rapid increases in NTEs such as wine, tropical fruits, and vegetables have similarly been monopolized by the agrarian bourgeoisie to the exclusion of small/medium producers. In a process renii niscent of the agroexport boom in Central America three decades earlier, land concentrations, peasant displacements, and an increasing seasonality
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       77
to patterns of rural labor demand have accompanied the rapid growth of NTE production in Chile (Carter and Mesbah 1993). The inability of the peasantry and other small/medium producers to participate in the country's NTE boom has led a growing number of observers to describe recent Chilean agroexport growth as 'exclusionary' (e.g., Ortega 1988; Cox, Nino de Zepeda, and Rojas 1990; Gomez and Echenique 1988). At the same time, the strong bias against state intervention in Chile's neoliberal development program has prevented the adoption of policies that might target groups of small/medium producers for assistance in establishing a presence in NTE markets.
Elements of an Alternative, Broadly Based NTE Strategy
It policies are not implemented to assist disadvantaged producers, the new cycle of export diversification based on NTEs threatens to reinforce the exclusionary character of more traditional postwar development models focused on agroexport production. As has happened so often in the past, export-led growth will serve the narrow interests of an elite minority rather than the broader needs of the popular majority. Conversely, however, the inclusion within NTE strategies of policies designed to meet the specific needs of small/medium producers (e.g., risk minimization, diversification into exports while maintaining food production, adoption of labor- rather than capital-intensive techniques) might encourage them to move into potentially highly profitable NTE sectors. This might create new sources of employment and accumulation that could help to reverse tendencies toward social, sectoral, and spatial polarization that have marked previous export-oriented development strategies. It might also help to lay the social foundations for increased political stability, without which any future development strategy, whether or not it contains a significant NTE component, cannot be sustained.
While many NTE sectors have been highly profitable in recent years, they have also often been quite risky, especially for smaller producers during the initial stages of production. Some of this riskiness is due to the uncertainties of adapting new production methods and technologies to diverse and often harsh local conditions. However, other important elements of risk stem from the vagaries of producing 'luxury' goods for n severely restricted group of First World countries. In most Third World areas, NTE production is aimed at a quite limited market (e.g., at the US from Latin America and East Asia, at Europe from Africa and the Middle Fast). Like more traditional exports, NTEs often compete with similar products from many other developing countries. While these exports may dominate their production sectors within a particular Third World country, they normally represent only a small fraction of similar imports by a large
78                 MAINSTREAM THEORIES AND PRACTICES
economy such as the US. Costa Rican flowers, for example, represent only 1 percent of all US flower imports; at the same time, flower exports to the US comprise 91 percent of all Costa Rican flower production (Rosset 1989). Under these circumstances, NTEs from the South often have only minimal demand stability in the large markets of the North. Moreover, because many NTEs are discretionary or 'luxury' items subject to sudden cutbacks during economic downturns, demand instabilities for these products are further accentuated.
Faced with such fluctuations, a transnational corporation may readily switch production to other goods or shift exports to another country. However, these types of changes are often much more difficult for small/medium producers who normally lack the capital, technical expertise, marketing arrangements, and knowledge of global market conditions to make such shifts smoothly. For all of these reasons, it is critical that policies encouraging the formation of support organizations for small/medium producers be put in place to minimize the considerable risks inherent in NTE production and assist in penetrating overseas markets. These support organizations might take on different forms according to the particular historical conditions prevailing in individual countries. In some countries, they might take the form of credit, service, and marketing cooperatives operating under the auspices of a national peasant organization. In other countries, independent producer associations focused on particular NTE sectors might be created.
Nevertheless, whatever type of producer-support organizations might be created for each country, they will normally require considerable state support, at least in their initial stages. The mechanisms by which such state assistance might be made available may, once again, be variable enough to conform as much as possible to particular historical conditions. At the same time, however, there are a number of areas that, given the nature of NTE production in most countries, will almost certainly have to be addressed in any program to heighten the participation of small/medium producers and spread the benefits of growth in NTE sectors to the popular majority. These include: increasing access to land and other major means of production; extending short- and long-term credit and other forms of financial support; providing agricultural extension programs and other technical assistance targeted at specific NTE sectors; improving provisions of more general social and economic infrastructure (e.g., education, health care, transportation systems), particularly in outlying rural areas; facilitating backward linkages, both to increase articulation with other domestic economic sectors and to increase access to needed imports at fair world prices; promoting forward linkages (e.g., in processing, refining, packaging) to capture more value-added and increase local multipliers; and, finally, assisting in penetrating foreign markets.
STRATEGIES OF GROWTH AND INDUSTRIALIZATION       79
It should be emphasized again that the mechanisms by which various forms of state assistance might be provided may differ among countries according to local conditions. For example, increased access to land has successfully been provided for many small/medium rural producers via land reforms in some countries (e.g., Nicaragua, South Korea, Taiwan). However, given the present political realities of many other countries, less politically contentious methods of land redistribution might have to be employed, at least in the short term, in order to gain necessary support from powerful economic and political interests. For example, following the analysis of Carter and Mesbah (1993: 1085) in Chile, such methods might focus on the extension of 'self-financing land market reform policies - which include progressive land taxation, and the creation of land banks or land financing institutions such as mortgage banks - [that] will achieve the traditional land reform goal of linking agrarian growth with poverty reduction.'
Likewise, the state may choose different means to offer assistance in areas such as foreign market penetration or export financing. In a recent study of export promotion among East and Southeast Asian countries, the World Bank (1993b: 143-5) found that various states successfully employed different methods in both of these areas in order to meet goals for export diversification and expansion. Virtually every country in the study had some programs to ensure access to credit, often at subsidized prices. But there was also an impressive degree of variety within successful programs of export financing, including in the types of credit (long- versus short-term), the degree of subsidization (guaranteed access versus subsidized rates), the selectivity (all exports versus targeted export activities), and the means of delivery (specialized state-controlled financial institutions versus market subsidies). Similarly, nearly all of the states in the study recognized the difficulty that new exporters often face in penetrating foreign markets. But, once again, various means were chosen to encourage these exporters to overcome such problems. Some states directly subsidized export activity (direct income tax incentives), some subsidized market penetration (through exporter associations), some subsidized small/medium exporters to offset their particular difficulties in market penetration, and some promoted the creation of international trading companies.
Various means may be chosen, but active state involvement is needed in most Third World countries if NTE growth is to avoid reinforcing tendencies toward societal polarization that have accompanied previous export-led development strategies. If managed efficiently (and there are a growing number of examples of successful state-directed export promotion, particularly among the Asian NICs), the dividends that this would p«y in terms of stimulating more broadly based and socially sustainable economic growth would far outweigh any short-term costs involved in
80                  MAINSTREAM THEORIES AND PRACTICES
state intervention. There is nothing inherently wrong with development strategies that seek to increase production of NTEs, or of agroexports in general, if they are consistent with larger societal goals (e.g., promoting growth with equity, maintaining access to affordable food and other basic needs).
Indeed, many peasants and other small/medium producers are not opposed to diversifying into NTE sectors. However, they wish to do so on terms by which they can effectively compete with larger producers and which allow them to minimize risks to acceptable levels. This normally entails putting mechanisms in place (e.g., technical assistance, credit programs, forward/backward linkages) to allow for true productive diversification - involving the continuing production of basic grains and other internal consumption goods alongside NTEs, rather than a complete switchover to the more risky export sectors. It also entails the creation of institutions (e.g., different forms of cooperatives, producer associations) to enhance producers' technical and marketing expertise and to assist, when necessary, in promoting market diversification. As with traditional agroexports, import substitution, or any other recent development approach, NTE strategies must include such considerations if they are to move beyond the immediate objectives of an elite minority toward the longer-term interests of the popular majority.
3
The Asian Newly Industrializing Countries
How has a group of small, resource-poor Asian countries sustained rapid development over the last three decades, while most other Third World countries have slipped into stagnation and crisis? Neoliberal development theorists claim they have an explanation, and argue that it should form the centerpiece of a new development model for the rest of the South. However, does the neoliberal explanation of development in the Asian newly industrializing countries (NICs) stand up to serious scrutiny?1 Is it transferable as a new model of development that other Third World countries should emulate? These are some of the most pressing questions in development studies today. They are addressed in this chapter through analysis of the following aspects of NIC development: the role of the state in development, the compatibility of inward- and outward-oriented elements of development, the influence of internal historical and sociocultural conditions on development, and the impact of external geographical and historical factors on development.
Comparing the Asian NICs and Latin America
In recent years, proponents of neoliberal development strategies have often buttressed their arguments by pointing to the performance of the Asian NICs as an empirical illustration of the superiority of their outward-oriented, market-led development model. The development experience of the Asian
The original Asian NICs are the 'Four Tigers' of Hong Kong, Singapore, South Korea, and Taiwan. The recent development performance of these countries is also sometimes compared wnh lhal of Japan in the early postwar period. In addition, a number of
other Aman • ■■........--- (e.g., Indonesia, Malaysia, Sri Lanka,  Thailand, Turkey) have
leu-nlly hern Riven ihe »tain« ol NIC« in much of the development literature.