Export led development Primary vs. industrial exports

Journal

of Development

Export

Economics

163-172.

North-Holland

led development

Primary vs. industrial Mukesh

41 (1993)

Eswaran

exports*

and Ashok

Kotwal

The University of British Columbia, Vancouver, BC, Canada Received

May 1991, final version

received

March

1992

Trade with North can make South poorer if South is a primary exporter. Manufactured imports from North can cause a decline in the real wage and consequently-lower the labour supply. The production nossibilitv frontier and hence the GNP of South can shrink while that of North can expand. If South gains comparative advantage in manufactured goods and reverses the trade pattern it can turn the tables on North. The analysis, based on a theoretical model, suggests a way to reconcile the divergent views on the desirability of North-South trade.

1. Introduction

Despite the spectacular success of the export-led growth strategy followed by countries like Taiwan and South Korea, many economists remain unpersuaded by the case for export-led growth.’ The work of Prebisch (1950), Emmanuel (1972) Frank (1979) and Bagchi (1982), among others, has emphasized that trade with the North is more likely to retard economic development in the South rather than to aid it. Most of this work has been based on the interpretation of the experience of primary exporting countries in Latin America [Prebisch (1950) Emmanuel (1972) and Frank (1979)] and of India during her colonial period [Bagchi (1982)]. Indeed, despite exporting primary goods for over a hundred years, a significant number of countries in Latin American, Africa and South Asia have remained poor. An important distinction between a success story such as South Korea and a failure such as Honduras is that the former has grown through exporting manufactured goods to the developed countries, whereas the latter has remained backward while exporting primary goods such as bananas and coffee. In this paper, we Correspondence to: Ashok Kotwal, Department of Economics, University of British Columbia, 997-1873 East Mall, Vancouver, BC, Canada V6T 1Zl. *This is a revised and retitled version of a manuscript previously entitled ‘Can Trade with North Make South Poorer?, December 1989. We would like to thank Charles Blackorby, James Brander, Brian Copeland, and the participants of the Economic Theory Workshop at the University of British Columbia for useful comments. ‘See Evans (1988) for a survey of alternative perspectives on trade and development. 03043878/93/$06.00

0

1993-Elsevier

Science Publishers

B.V. All rights reserved

164

M. Eswaran and A. Kotwal, Export led development

suggest one reason why the success of an export-led growth strategy might depend upon what kind of good is being exported. There is substantial literature which explores the asymmetries inherent in the pattern of trade between the North and the South. Different strands in this literature can be classified according to the different sources of asymmetry that they identify across the trading partners. The Prebisch-Singer hypothesis on the worsening terms of trade was based on an asymmetry on the demand side: exports from the South have a lower income elasticity than the exports from the North. The asymmetry featured in the open economy part of Lewis’ seminal work (1954), further developed by Findlay (1980) and Bardhan (1982), is the presence of a dual labour market in the South and its absence in the North. Chichilnisky (1981) incorporates both the asymmetries mentioned above. Sarkar (1989) takes a very different route and analyses the consequences of the North exporting investment goods in exchange for consumption goods from the South when both countries are in the situation of Keynesian unemployment. Krugman (1981, 1987) explores North-South trade when the asymmetry is on the supply side. The industrial sector has increasing returns to scale, while the agricultural sector has constant returns to scale. The asymmetry across the trading countries is captured by assuming that the North has superior capital endowments and hence lower average costs in producing industrial goods than the South. Free trade between the North and the South then inevitably leads to the de-industrialization of the South. Our paper is most akin in spirit to that of Krugman (1981). We model a two-country, two-sector, world in which labour is used in both industrial and agricultural sectors while land - the only other factor in the model - is a specific factor in agricultural production. The supply and productivity of land is fixed and thus agriculture is subject to diminishing returns. Industrial production, on the other hand, has constant returns to scale. The asymmetry between the North and South is simply represented by the fact that the North has a comparative advantage in industrial goods. The opening up of trade causes the industrial sector in the South to shrink. The released labour must be absorbed in the agricultural sector. Due to diminishing returns in agriculture, the wage in terms of the agricultural good declines, leading to a possible decline in the real wage under reasonable assumptions on preferences. If the labour supply is price elastic, it can thus decline in the South as a result of opening up the economy to trade with the North. In an economy with a fixed supply of land, a decline in the labour supply will shrink the production possibility frontier of the South. Thus, the result is the exact converse of the notion of trade providing a ‘vent-for-surplus’. The paper thus suggests yet another channel via which trade with the North could induce the underdevelopment of the South. When Myint (1958) suggested the vent-for-surplus idea, he pointed out a

M. Eswaran and A. Kotwal, Export led development

165

very important additional mechanism through which trade benelitted an economy. The principles of comparative advantage shows how trade expands a country’s consumption possibilities for a given production possibility frontier. Myint showed that by bringing into use previously unused resources, trade can even expand the production possibility frontier of the economy. Another way of thinking about Myint’s theory is that the opening up of an economy to trade can increase the prices of certain factors causing an increase in their supply. For example, an increased demand for Canadian wheat resulted in an increase in the land prices in Canada, making it possible to bring under cultivation previously uncultivated land. The surplus thus got vented. Our paper points out that the above process could work just as easily in the opposite direction to the one Myint has described. It all depends on which factor is price elastic. In countries such as the U.S., Canada and Australia, land, indeed, was in surplus. An opportunity to export primary goods, in this case, induced an increase in the supply of land, and hence an expansion of production possibilities, making trade doubly advantageous. In the labour abundant countries of South Asia and Latin America, however, land must be considered very much a fixed factor; labour, on the other hand, can be considered to be a price elastic factor. Since the opening up of the economy can result in an increase in the price of the inelastic factor (i.e. land) and a decrease in the price of the elastic factor, the production possibility frontier can shrink. The effect thus counters the expansion of consumption possibilities resulting from the trading opportunity. If the production possibility effect dominates the latter effect, it is possible that the GNP under free trade is lower than under autarky. Suppose, however, that the South is able to increase its industrial productivity (total factor productivity or, in this model, marginal product of labour in industry) at a rate that is faster than in the North. Trade becomes doubly advantageous to the South as it was to Canada in Myint’s model. The industrial sector in the South expands at the expense of the industrial sector in the North. The real wage in the South can thus increase, inducing an expansion of the labour supply and hence its production possibilities. This, in fact, may be what happened in countries like Taiwan and South Korea, which followed a successful export-led strategy. Our model thus suggests one possible explanation of why the kind of good that is exported determines the success or failure of an export-led strategy. We would like to emphasize here that our intent is not to generate a new result in trade theory but rather to relate the implications of existing theory to the varied experiences of developing countries. In addition, we would like to clarify why we think the possible shrinking of the production possibility frontier may be significant from the point of view of a developing country. As mentioned earlier, we will show in our analysis to follow that the

166

M. Eswaran and A. Kotwal,

Export led development

shrinking of an economy’s production possibility frontier may result in a decline in the economy’s GNP. From a welfare point of view, the above result can hardly be considered very significant since the greater consumption of leisure by workers in South associated with the reduction in labour supply is not accounted for in GNP. Yet, in the context of a developing country, the result is interesting for several reasons. Firstly, the developmental effort among many late industrializing countries has been spearheaded by the state; Germany, Japan, India, South Korea are just some of the examples. In most of these cases, the objective of development has been closer to GNP maximization than to any notion of Pareto-improving growth. Secondly, the success of exporters (like Taiwan and South Korea) of manufactured goods in promoting development invariably has been measured by the empirical yardstick of GNP growth. So it makes sense to couch the theoretical results in the same terms. The third, and most important reason, has to do with the fact that the essence of economic development consists in increasing the returns to human effort. Capital - physical or human - can be created through human effort. An increase in GNP enhances a nation’s capacity to build dams and schools and hence to expand the productive capacity in the next period. Along with these public goods, there are various classes of externalities associated with effort such as learning-by-doing. All these considerations lead to the conclusion that the shadow price of leisure is much lower than the private valuation of leisure. Explicit incorporation of these externality effects into our framework would have complicated our model. In view of the fact that there are a number of compelling models of trade [Bardhan (1970), Krugman (1987)] which bring out the importance of the process of learning-by-doing, we have opted to keep our model simple. It allows us to focus attention on the central point of the paper: the success of an export-led growth depends on whether or not the factor the exports use more intensively is price elastic. 2. The model We shall set up in this section a simple two-country general equilibrium model to investigate the effect of trade on a country’s GNP. For concreteness, we shall refer to these countries as India and England; parameters and quantities pertaining to these countries will be subscripted by i and e, respectively, Each country has an endowment of two factors of production: land and labour. Country j E {i, e} has an amount of land equal to Hj and a working population of N,. Agents in both countries have identical preferences defined over labour effort (n) and two goods: food (F) and cloth (C). We assume that these preferences are represented by the utility function U(F, C, n) = F1’2C1’2-fd,

(1)

M. Eswaran and A. Kotwal, Export led development

167

where n is the amount of labour effort supplied by the agent, Let p, w and u, respectively, denote the price of cloth, the wage rate and the land rental rate, with food as the numeraire good. Landlords are assumed to live off their rental income and supply no labour. From the above utility function the demand functions for leisure (giving labour supply), food and cloth can be easily computed. On the production side, we assume that in country j food is produced according to a production function that is Cobb-Douglas and linearly homogeneous in the amount of land (h) and labour (L):

fj(h,L)=Ajh%-y Cloth production, returns production g,(L)=CrjL,

however, function

jE{i,e}

and

O