Introduction

Recently, multinational enterprises (MNEs) actively use regional trade agreements (RTAs) all over the world. In particular, there is a growing number of MNEs of which the overseas affiliates trade with third countries (i.e., neither home country nor host country) under RTA preferential schemes. For example, the White Paper on International Economy and Trade 2014 by the Japanese government introduces several case studies on Japanese affiliates in Southeast Asia using preferential schemes when trading with China, India, or Korea. Their use in trading with China or Korea is especially important since Japan has not yet concluded any RTA schemes with China and Korea. Namely, through trading via their overseas affiliates, MNEs enjoy preferential trade with countries with which their home country does not have preferential access.

In the academic literature, such export to third countries by MNEs is called the export platform foreign direct investment (FDI). This branch of research is motivated by the growing importance of platform-FDI activities. For instance, export sales accounted for over 40 per cent of total sales by foreign affiliates of Japanese MNEs in 2009, and export sales to third markets represented over 30 per cent of the export sales. Despite its empirical importance, there have been only a few studies on export platform FDI. Ekholm et al. (2007) is one of the pioneering studies and considers various types of export platform FDI according to the destination market of the overseas plant. The above-mentioned feature indicates that overseas affiliates by such export platform FDI enjoy preferential trade in their export.

In this study, we empirically identify significant firm characteristics in the use of preferences in export platform FDI. Specifically, we investigate the correlation of preference use with the share of “originating inputs”, intensity of exports to the third country, and experience of preference use in exporting to other third countries. For example, if the export intensity is highly and positively correlated with preference use, the absolute magnitude of preferential exports will play a significant role in that use. We can also decompose “originating inputs” into labour inputs, originating material inputs, and other originating inputs. As a result, in order to directly find the bottleneck in preference use in export platform FDI, we just examine the correlation of preference use with firm performance indicators rather than its determinants, such as more fundamental firm characteristics including firm productivity.