Trade agreements and participation in global value chains: Empirical evidence from Latin America

Published: 22-10-2020| Version 1 | DOI: 10.17632/pvb847766y.1
Eduardo Sanguinet


This database was built from secondary and public sources (i. GVC-UNCTAD-EORA; ii. Mario Regional Trade Agreements; iii. CEPII gravity data). The do-file and matlab codes were used to build the data panel (1995 - 2015) with 180 countries with bilateral trade flows in added value. The regional classification is based on United Nations Standard Countries / Area codes.


Steps to reproduce

We adopted three main data sources: i. GVC-UNCTAD-EORA ii. Mario Regional Trade Agreements iii. CEPR gravity data Steps and estimation procedure: 1. Data on bilateral value added flows is from multiregional EORA input-output tables. Available at (“2. Country by country breakdown, 1990 - 2015”). 2. Stata Do-file “” was used to create the main data file. 3. Industry-flows calculations are based on share of interindustries trade according to EORA data. The Matlab code (“code_eora_c_final.m”) was made available by Aslam, Novta, and Rodrigues-Bastos (2017) to supporting the creation of matrix value-added trade flows ( Then, Stata Do-file was used to compile specific-pair flows. 4. The data on the “Mario Larch's Regional Trade Agreements Database” of trade agreements was cleaned up, available at, with the Stata do-file (.do file) code “”. In this step, the codes and names of the countries were standardized according to Table 1. 5. The gravitational data had the source and destination codes according to the United Nations Standard Countries/ Area codes. The data were obtained by CEPII ( 6. The EORA_c_c.dta database was consolidated. Both bases were joined in these single file. 7. The estimation results were obtained from the Stata Code “”.