Studies examining the effects of free trade agreements on the activities of multinational enterprises have tended to focus on trade between home countries and their FTA partner countries. This column examines the effects of free trade agreements on local sales and exports to third countries by overseas affiliates of Japanese manufacturing multinationals. While the effects on local sales are ambiguous, regional trade agreements have positive effects on exports to other countries, suggesting that that Japanese multinationals have optimised their cross-border supply chain networks by utilising various forms of free trade agreements.
Over the last several decades, the offshore activities of multinational enterprises (MNEs) have increased as a result of technological progress, including new information and communication technologies (ICT), construction of large-scale transport infrastructure in developing countries, and an increase in global competition. MNEs have formed cross-border supply chain networks to enhance their efficiency and competitiveness by splitting production processes and moving these processes to other locations to give them competitive advantages. In this environment of economic globalisation, many governments have used free trade agreements (FTAs) to increase trade with their partners because policymakers believe that such agreements help to integrate foreign economies into their own country’s economy and improve economic efficiency by reducing tariffs and constructing cross-border supply chain networks based on MNEs, enhancing institutional transparency, and protecting foreign direct investment (FDI). According to the WTO, 375 FTAs have been implemented as of 1 September 2025.
Many studies have examined the effects of FTAs on MNE activities and found positive effects on trade and FDI. Those studies usually examined the effects of FTAs on the hub-and-spoke systems of MNEs, focusing on whether FTAs encourage MNEs to increase their trade between the home country and the FTA partner countries, but they typically do not examine the effects of FTAs on the majority of MNE activities. As Figure 1 indicates, the share of trade of overseas affiliates of Japanese MNEs with Japan is much smaller than the local sales/procurements and trade with third countries. Thus, MNEs have developed their supply chain networks beyond the hub-and-spoke systems between the home and the partner countries. The effects of bilateral FTAs on local transactions should also be examined. In addition, regional FTAs and FTAs between local governments and large economies have possibly encouraged MNEs to expand larger supply chain networks to those third countries. Based on these considerations, in a recent paper (Kato and Nishiyama 2025) we examine the effects of FTAs on local sales and exports to the third countries by overseas affiliates both theoretically and empirically.
Figure 1 Activities of overseas Japanese affiliates by region
Figure 1 Activities of overseas Japanese affiliates by region
Data source: METI, the Basic Survey on Oversea Business Activities
Theoretical analysis
First, using a Melitz-type firm-heterogeneity model (Melitz 2003), we theoretically examined how FTAs affect a firm’s behaviour, especially firm sales. We focused on the effect of a change in the parameter related to FTA breadth/depth. In our model, it is assumed that the fixed labour input for FTA regulatory compliance-related activities is the proxy variable for expansion of an FTA (FTA breadth), while the level of tariff is the proxy variable for promotion of the FTA (FTA depth). In addition, the number of countries in an FTA can be regarded as a proxy for FTA depth. Our model shows the following effects of an FTA on the sales of domestic and export firms:
The effects of FTA breadth and depth on the sales of domestic firms are not uniquely determined.
The effects of FTA breadth on the sales of export firms are positive while those of FTA depth are indeterminate.
Our theoretical analysis concludes that the effects of FTAs on domestic and export sales differ due to exogenous factors that trigger changes, such as tariffs, the number of export destinations, fixed labour input related to FTA compliance, and market size. With this understanding, we empirically examine the effects of FTAs on the sales of overseas affiliates by FTA type regarding an overseas affiliate as a firm in the model and discuss whether MNEs efficiently utilise different FTAs in their supply chain networks.
Empirical analysis
Based on these theoretical implications, we carefully examine whether FTAs affect the sales of overseas affiliates of Japanese manufacturing MNEs, considering the differences between the above FTA patterns and proxies for the factors that trigger changes. That is, we examine whether bilateral FTAs increase the local sales of overseas affiliates and whether regional FTAs or FTAs with large markets encourage overseas affiliates to export to other countries. We estimate the effects of FTAs based on calculations of depths and breadths as proxies in triggering changes such as tariffs and unit factor inputs, respectively. We separately estimate the effects of bilateral FTAs on local sales and of regional FTAs on exports to identify the effects of FTAs on local sales and exports discussed in the theoretical model. We select ASEAN, the EU, and NAFTA (the US) as regional FTAs because overseas affiliates of the Japanese manufacturing MNEs are heavily concentrated in those regions, in addition to the neighbouring countries where there was no regional FTA during the examined period. As a large market, we add China to these three regions due to its similar size in terms of GDP.
The index of FTA breadth/depth is calculated following Jinji et al. (2022) and Murakami (2025). These equations are estimated using Santos and Teneryo’s (2006) Poisson pseudo-maximum likelihood method due to the heteroscedastic error terms and the existence of zeros in the dependent variable in our estimation models.
Our estimation results show that the coefficients for bilateral FTAs on local sales are statistically insignificant for both WTO depth and breadth indexes. These results are consistent with our theoretical implication that the effects of both WTO depth and breadth are ambiguous. On the other hand, regional FTAs have statistically significant positive coefficients on exports to other countries. As expected, a decrease in tariff rates and an increase in the number of FTA partners both encourage exports. The positive coefficients for FTA breadth indicate that an increase in the compliance-related fixed labour input resulting from FTA participation increases the sales of export firms. These results imply that Japanese MNEs have optimised their cross-border supply chain networks by utilising various forms of FTAs. Bilateral FTAs between each local government and the US are statistically insignificant. This likely reflects the fact that the US strongly encourages MNEs to invest in the country.
Conclusion
Our findings highlight the need to examine what affects the structure of the multi-tier supply chains of MNEs. With major changes in geopolitical circumstances, such as a decoupling between China and the US, MNEs must use various FTA schemes effectively to reconstruct their supply chain networks and reshore their production processes. Our study offers important implications for industrial policies that are designed by governments to attract foreign firms and to encourage domestic firms to explore the global market amid the current environment in which anti-globalisation is advancing.
Editor’s note: The main research on which this column is based (Kato and Nishiyama 2025) first appeared as a Discussion Paper of the Research Institute of Economy, Trade and Industry (RIETI) of Japan.
References
Jinji, N, X Zhang and S Haruna (2022), "Deep Integration, Global Firms, and Technology Spillovers", Springer Open Access.
Kato, A and H Nishiyama (2025), "Supply Chains and FTAs", RIETI Discussion Paper Series 25-E-068
Melitz, M J (2003), "The Impact of Trade on Intra-industry Reallocations and Aggregate Industry Productivity", Econometrica 71(6): 1695–1725.
Santos Silvia, J M C and S Tenreyro (2006), "The Log of Gravity", The Review of Economics and Statistics 88(4): 641-658.
