Editors' Summary

The Brookings Trade Forum held its sixth annual conference in Washington on May 15-16, 2003. This volume contains the papers, invited commentary, and general discussions from that conference.

The volume's first three papers focus on implications of heightened foreign competition, as developing countries increase their exposure to the global market.The first paper uses a dynamic simulation model to better understand the interrelated channels through which trade liberalization affects the efficiency of domestic firms. The second paper tackles a novel aspect of globalization's impact on labor standards, empirically examining effects of foreign pressures on compliance with minimum wage laws in Indonesia. The third undertakes a detailed case study of the controversial economic reform of the cashew sector in Mozambique, estimating gains and losses incurred and drawing broader lessons for liberalization.

The remaining three papers address macroeconomic topics. One undertakes a broad empirical analysis of when countries are able to sustain fiscal consolidation. It finds clear differences in experience across low-, middle-, and high-income country groups. Another empirically examines the linkages between external debt and economic growth, finding high-debt stocks to have particularly deleterious implications for both capital accumulation and productivity. The volume's final paper uses both theory and empirics to evaluate the efficacy of introducing collective action clauses into bond contracts, a recent approach intended to address the persistent problem of sovereign debt crises.

IN THE FIRST PAPER Erkan Erdem and James Tybout take a new look at whether trade liberalization really does tend to increase manufacturing efficiency in developing countries. Over the past two decades, a substantial body of evidence has accumulated on the firm- and plant-level effects of openness in developing countries, including influential earlier work by Tybout. This literature has generated findings, now typically accepted as stylized facts, about how increased foreign competition acts to discipline domestic firms. Import discipline is believed to work primarily through two channels. First, it squeezes price-cost margins among import-competing firms. This heightened competitive pressure induces productivity gains among these same firms. Second, additional efficiency gains are expected to come from reallocations of firms' market shares.

Erdem and Tybout argue that although these findings are useful, they leave many central issues unresolved and may be misleading because they tell only part of the story. Their paper highlights concerns that arise from problems with the way in which firm-level productivity has been measured in the literature. It also emphasizes additional dimensions of firm-level responses and recognizes that these may become increasingly important over time. Taking a more nuanced perspective, the effects of trade liberalization may be quite different in the short run than the long run.

The objective of the Erdem and Tybout paper is to begin to address these shortcomings. The authors' approach is to develop a computable model of industrial evolution that enables them to simulate the effects of import competition as they unfold over time. They are able to demonstrate what types of managerial behavior, long-term transition paths, and welfare effects are consistent with the findings of previous firm- and plant-level empirical studies. The model also provides a rigorous means to evaluate the combined effects of various forces.

Erdem and Tybout's paper begins with a brief overview of the logic behind the import discipline hypothesis. It then surveys the empirical evidence supporting this hypothesis, focusing on analyses of five natural experiments. Particularly informative is the table summarizing results from studies of liberalization in Chile, Mexico, Côte d'Ivoire, Brazil, and India. As this survey shows, these studies are consistently supportive of the hypothesis that trade liberalization increases productivity.

The authors then explain why problems in how productivity is measured may induce biases in both the cross-section and time-series variation in efficiency gains across firms. Primarily because of data limitations, empirical productivity studies measure output as deflated revenues and intermediate inputs as deflated expenditures. Neither of these reflects the considerable heterogeneity in both outputs and inputs across firms. Thus in addition to productivity, these...