Replication materials: Asymmetric Integration and Export Dependence: Mexico and Canada in the United States Market, 1990–2024 Regional integration deepened North American trade but distributed export dependence asymmetrically between Mexico and Canada.

This repository contains the data, code, and supplementary materials used to reproduce the results reported in the article “Asymmetric Integration and Export Dependence: Mexico and Canada in the United States Market, 1990–2024.” The study examines whether North American regional integration produced symmetric patterns of export diversification in Mexico and Canada. Using annual bilateral trade data for 1990–2024, it compares the geographic concentration of exports, the relative orientation of Mexican and Canadian exports toward the United States, structural changes in concentration, and sectoral differences in the redirectability of exports. The replication package combines data from: UN Comtrade, for annual bilateral merchandise exports and sectoral trade flows; CEPII Gravity and GeoDist databases, for bilateral distance, population-weighted distance, contiguity, and common-language indicators; World Development Indicators, for GDP and population variables. The aggregate bilateral panel covers Mexico and Canada as exporters, 193 destination economies, and 35 years. The rectangularized panel contains 13,510 exporter–destination–year cells, including reported zero trade flows. The principal gravity estimation sample contains 12,760 observations after the application of fixed effects and data-availability restrictions. The repository includes the materials necessary to reproduce: the Herfindahl–Hirschman Index, Theil concentration index, equivalent number of destinations, and United States export share; the unrestricted- and balanced-destination concentration panels; Poisson pseudo-maximum-likelihood gravity estimates with exporter-year and importer-year fixed effects; the Mexico–Canada relative export orientation toward the United States; robustness checks using alternative measures of bilateral distance, alternative samples, and positive trade flows only; the rank-based placebo-destination exercise used as a complementary diagnostic for the two-exporter comparison; Bai–Perron multiple structural-break tests for the Mexican and Canadian concentration series; sectoral gravity estimates for 96 HS2 chapters, classified according to product redirectability; the tables and figures reported in the article and its online econometric appendix. The central empirical findings reproduced by the package are that Mexico began the period less concentrated than Canada, overtook it in the early 1990s, and subsequently maintained a higher level of export dependence on the United States. The preferred gravity specification estimates a Mexico–Canada relative U.S. orientation of approximately exp(γ) = 3.38. Structural-break tests identify breaks in 1995 and 2007 for Mexico and in 1998 and 2007 for Canada. Sectoral results indicate that the Mexican excess is particularly pronounced in electrical equipment, machinery, and fresh horticultural products, whereas the automotive differential is close to parity. The code is organized in the order required to reconstruct the processed datasets, estimate the models, and generate the reported outputs. A README file documents the software environment, package dependencies, execution order, data sources, variable definitions, and correspondence between scripts, tables, and figures. Where redistribution of original provider files is restricted or impractical, the repository supplies source identifiers, retrieval instructions, and processing scripts rather than republishing the original source files. No API credentials or personal file paths are included. The repository is intended to support transparency, verification, and extension of the empirical analysis. It does not provide a causal estimate of the effect of NAFTA, USMCA, or any individual trade-policy intervention.