Sourcing patterns of export‐platform foreign affiliates: The case of Japanese affiliates in Mexico

Abstract

This study investigates the sourcing patterns of Japanese export‐platform foreign affiliates in Mexico, which mainly export to the United States and Canada. We propose a novel approach to estimate intermediate input elasticities of exports by sourcing country. We find that, on average, Japanese export‐platform foreign affiliates in Mexico source intermediate inputs from third countries, including the United States and Canada, rather than from Japan and Mexico, suggesting that Japanese export‐platform foreign affiliates in Mexico are mainly integrated into the vertical production networks back and forth between Mexico, the United States and Canada. In turn, Japanese foreign affiliates selling domestically in Mexico source intermediate inputs not only within the North American Free Trade Agreement countries but also from Japan. In addition, we find that export‐platform foreign affiliates in Mexico use more labour‐intensive production than do foreign affiliates selling domestically in Mexico. This suggests that saving labour costs is one of the motives for export‐platform foreign direct investment in Mexico.

Spatial Clustering of Foreign Direct Investment: The Case of Japanese Automotive Suppliers in Mexico

“…Automotive suppliers seem to choose their locations abroad following the automotive assemblers with whom they have relationships in Japan (Hayashi, 2012;Martin, Swaminathan & Mitchell, 1998;Smith & Florida, 1994). Kondo (2018) analyses the international sourcing patterns of export-platform foreign affiliates from a new approach: an extension of the production function estimation in the total factor productivity literature. The author decomposes production into country-by-country transactions and decomposes intermediate inputs into country-by-source purchases.…”

Networks of foreign affiliates: Evidence from Japanese micro‐data

“…The remaining margin of error concerns activity below reporting thresholds, which is indistinguishable from true zeroes. 13 While the gravity framework has been developed in the context of trade in goods, it has been shown to be a good fit also for trade in services and foreign affiliate sales -see Ceglowski (2006[47]); Kimura and Lee (2006 [30]); Kleinert and Toubal (2010 [29]); Rouzet, Benz and Spinelli (2017 [6]). 14 As one of the main variables of interest is the restrictiveness of services regulation, captured by the OECD STRI which is defined at the country-sector level and does not yet have a sufficiently long time series to allow robust inference from time variation in policies, the regressions do not include country fixed effects.…”

“…A number of other robustness checks have been carried out to test that the results are robust to the sample composition and alternative choice of variables. 29 Table A.2 in the annex shows that most Japanese foreign affiliates are located in China and in the United States. To test that our results are not driven by these two destinations, we run the baseline specification excluding these two economies and got very similar results, indicating that our estimates are not sensitive to host markets with large numbers of Japanese affiliates.…”