Industrial evolution in developing countries : micro patterns of turnover, productivity, and market structure

Abstract

This book presents the results of a project designed to link changes in trade policy with patterns of producer entry, exit, and adjustment characterizing developing country producers, using plant-level data, which reveals micro aspects of the relation between commercial policy and industrial sector performance. The researchers used data gathered from the manufacturing sector of five countries, Chile, Colombia, Morocco, Turkey, and Mexico to pursue the question of how industrial sectors in developing countries evolve. The country studies are collected in part 2, with an overview provided in chapter 8 of the same part. Part 1 draws together evidence on a single issue from a number of countries, focusing on the employment shifts that result from the entry, growth, and exit of manufacturing plants; addressing the cross-firm differences in productivity at the micro level; and quantifying how the turnover process sorts out efficient and inefficient producers. Other chapters address the sources of heterogeneity across producers, with emphasis on the role of economies of scale and foreign direct investment. Chapter 1 summarizes the main lessons, particularly, that the degree of producer turnover in the manufacturing sectors of semi-industrialized countries is on average GREATER than that found in industrialized countries, so the competitive pressures are at least as great. Furthermore, the countries relatively open to trade examined in this research show no evidence that imports contribute additional pressure at the margin. Rather, dying plants become progressively less productive in their final years while new plants that survive improve rapidly. Since a large fraction of the manufacturing sector turns over within five to ten years, policies that inhibit this replacement process probably have substantial medium- and long-term detrimental effects on productivity.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 292 citations exist upstream.