What Can the Poor Expect From Trade Liberalization? Opening the "Black Box" of Trade Modeling

Abstract

Trade liberalization is expected to act positively on development and poverty alleviation, both of which have become a high priority of international community. This explains why numerous studies have focused on assessing the expected benefits of trade liberalization on poverty. The main empirical tool for these assessments has been the use of multi-country Computable General Equilibrium Models (CGEM). These models, however, have produced divergent results. As demonstrated by recent studies, the associated increase in world welfare from full trade liberalization ranges from 0.2% to 3.1% — results that differ by a factor of 15 to 1! The impact on poverty headcount is also very divergent as the number of people lifted out from poverty ranges from 72 million to 446 million —a ratio of 5.5 to 1! This is a rather contrasting picture of the effects of trade liberalization on poverty. It gives the impression that with global trade modeling, divergent results are the rule. Moreover, as a sophisticated and complex tool of analysis, CGEM often appears as a “Black Box”, the results of which are difficult to understand. The objective of this study is to examine the efficiency of trade modeling in capturing the benefits from trade liberalization. It will provide a survey of methodologies utilized to assess the impact of trade liberalization on poverty and will examine the extent to which such assessments diverge. The survey also demonstrates the benefits of “complementary analysis”, which utilizes different methodologies to study a specific topic. First, the paper examines the advantages and drawbacks of each method, with a particular focus on multi-country general equilibrium models. Second, the paper undertakes a global modeling under general equilibrium — MIRAGE model— the results of which are compared to those obtained in recent studies.

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 — 30 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.