The Leff Conjecture: Some Contrary Evidence

Abstract

Listen

Translate article

In a recent issue of this Journal, Leff (1973) attributes the failure of export-led development to occur in nineteenth-century Brazil to the slow growth of exports, especially in per capita terms, and to the small share of exports in Brazilian national product. Leff speculates that during the same period other less developed countries (LDCs) failed to garner dynamic gains from trade for the same reasons. This conjecture contrasts with most of the existing literature, which treats imperfections in the linkages between internal development and the external trade stimulus as a main reason for persistent poverty in the Third World. In Section I, I present estimates of per capita exports in a large sample of LDCs for the years 1860, 1880, 1900, and 1913 (table 1) and of growth rates of per capita exports in these countries during two intervals of time, 1860-1900 and 1900-1913 (table 2). In Section II, I use the results to test Leff's conjecture. The periodization used here does not correspond exactly to the 1850-96 and 1897-1913 periods on which Leff focuses, in part because a sufficient quantity of data going back to 1850 is not available. Also, Leff's periodization of the nineteenth century is based on residuals from trend equations estimated for a particular country, Brazil, and may not be applicable to other LDCs. It is sufficient that the estimates span a long period of time, be widely spaced chronologically, and permit subdivision of the whole period into reasonably homogeneous segments. For purposes of brevity and in keeping with the spirit of Leff's article, the analytical discussion focuses on the period 1860-1900.