A Test for Balanced and Unbalanced Growth

Abstract

Listen

Translate article

A LTERNATIVE strategies for economic development have frequently invoked the doctrines of balanced and unbalanced growth.' Policy recommendations in favor of balanced or unbalanced growth are based on a priori notions about the relationship between (lack of) balance and the process of development. For balanced-unbalanced growth to become an empirical hypothesis, rather than a doctrine, it has to be formulated in a way that it is given the chance to be proven wrong. An operational formulation involves three steps: balance (or imbalance) is defined in an unambiguous way that renders itself to quantification; an observable relationship between balance and economic development is postulated; this observable relationship is further specified by establishing causality and by determining the flow of causation. The hypothesized observable relationship between balance and economic development is sufficiently clear in the literature. The proponents of the balanced growth theory specify positive association between balance on the one hand and overall growth rate in national income (per capita). The relationship is reversed for the unbalanced growth theorists. The discussion on causality between balance and development is less unequivocal. In general, the flow of causation is supposed to run from balance (or imbalance) to development.2 Two examples will suffice for illustrative purposes. Nurkse [8] [31 advocates balanced growth on the grounds that it increases the reinvestible surplus, it provides inducements to invest, it creates external economies in complementary industries and as a result it leads to higher economic development. On the opposite side of the field Hirschman [4] perceives the causal link between imbalance and development in terms of external economies of vertical type and in terms of decision-making which is induced by disequilibria.3 The operational definition of the concept of balance or imbalance is probably the weakest part in the formulation of the theory. In a recent article Swamy [9] formulated the criterion of balance in terms of the dispersion of sectoral growth rates and he tested the hypothesis by correlating the resulting index with the overall rate of growth for an international cross section and time periods between 1948-1960. In this article we basically employ the same operational framework, data sources and time periods with Swamy. The significant divergence in results is due to the more suitable index of imbalance and the more refined data set that we use. In section I we discuss the different variants of balanced-unbalanced growth and we introduce the appropriate for each measure of dispersion. After a brief description of the data in section II, section III presents the evidence on the relationship between growth rates and indices of imbalance. In section IV the indices of imbalance are related to the level of development, as determined by a country's national income per capita. Finally, we draw the conclusions from our analysis and we compare our results with parallel investigations.