Japan's outward FDI potential

Highlights

• Is Japan's outward FDI stock unusually high or low?

• We estimate counterfactual FDI using the gravity model for all OECD investor nations.

• We find Japan's ratio of actual to counterfactual FDI is the highest among OECD members.

• Therefore Japan has no unrealized potential for outward FDI.

Abstract

While Japan's outward FDI stock is historically high, it is not necessarily clear whether there is untapped growth potential, given the economic size of Japan and that of partner countries. This paper examines whether Japan's actual outward FDI stock is high or low relative to the FDI predicted by the gravity model using the outward FDI patterns of all OECD nations, which we call counterfactual FDI. The results indicate that the ratio of Japan's actual to counterfactual FDI is the highest among the OECD countries as of the year 2015. The regional distribution of Japan's actual to counterfactual FDI favors Southeast Asian nations, South Africa and the US. These results imply that Japan has no unrealized potential for outward FDI.

Introduction

Politicians sometimes rally nationalistic support around the concept of “fairness” by asserting that other countries either under-perform or over-perform relative to some standard that they determine.¹ While standards setting by policymakers is necessary for international agreements on mutual defense (e.g., the North Atlantic Treaty Organization, NATO) or climate change (e.g., the Paris Agreement) so that each participating country has a target to meet, these results-based approaches have been rejected in favor of rules-based approaches in the arenas of international trade and investment under the World Trade Organization (WTO). Nevertheless, results-based approaches often are adopted by policymakers seeking public support and negotiating leverage so economists are often tasked with determining standards for distinguishing unusual trade and investment patterns. For example, many studies have addressed the accusations that Japan exports too much, imports too little, and hosts too little inward foreign direct investment (FDI).² Most recently President Trump pressed Japanese business leaders to do more outward FDI, specifically into the US.³ This environment prompts our research question: Does Japan do enough outward FDI?

Japan's outward FDI has been expanding rapidly from the early 2000s. Fig. 1 indicates the value of Japanese outward FDI stock from 1996 to 2018. In 2018, Japanese outward FDI stock was historically high, reaching 181.7 trillion Japanese yen, which is about six times the level it was in 1996 (30.6 trillion yen). In 2014, the level of Japan's outward FDI stock was the 4th largest among the OECD countries.⁴

Is Japan's outward FDI unusually high or low? This question is important from the home as well as the host countries’ viewpoints. From the home country's perspective, whether outward FDI accelerates or not can be a major concern for policy makers because it may result in the hollowing out of domestic industries, even though it is a rational choice for firms for their survival. On the other hand, from the host countries’ perspective, whether foreign firms expand their activities or not is an essential concern for the local economy.⁵ In particular, local economies often spend large amounts of public resources to attract FDI inflows with an expectation of positive economic returns.⁶ These issues involving FDI are not limited to Japan but are commonly observed in many advanced countries.⁷

Questions regarding the appropriate size of FDI are nontrivial due to the many stakeholders involved in both home and host countries. In addition to policy incentives, FDI can be affected by both home and host countries’ factors such as economic size. Fig. 2 presents the ratio of outward FDI stock to GDP for Japan, the United States, and the average of the OECD countries from 1985 to 2015.⁸ Fig. 2 indicates that the ratio of outward FDI stock to GDP for Japan was 28.3 percent in 2015, which was comparable to the United States (27.6 percent) but lower than the OECD average of 44.5 percent. Fig. 2 also indicates that the ratios for Japan and the United States have been smaller than that for the average of the OECD countries for the last three decades between 1985 and 2015. This comparison implies that Japan's outward FDI may actually be somewhat low once one accounts for the size of the Japanese economy and the growth of other countries.

To evaluate whether Japanese outward FDI is unusually high or low, a reference value is necessary. Previous studies have established that the gravity model works well not only for international trade but also for FDI (e.g., Anderson, 2011).⁹ Accordingly, some studies such as Egger (2010) and Hoshi and Kiyota (2019) estimated counterfactual FDI, which is defined as the FDI predicted by the gravity model, and utilized it as the reference value. These studies then estimated the unexhausted FDI potential that is defined as the gap between the counterfactual and actual FDI stock. If the counterfactual FDI exceeds the actual FDI, this means that the gravity model predicts much larger FDI than the actual FDI. This in turn suggests that there is a potential for more FDI according to the gravity variables.

There are several studies such as Eaton and Tamura (1994) and Head and Ries (2005, 2008) that examined Japan's outward FDI in a gravity model framework. However, to the best of our knowledge, only Head and Ries (2005) addressed the above question directly.¹⁰ Head and Ries (2005) estimated the gravity model, using the data for 181 countries between 1980 and 2002. Their analysis found that Japan's actual outward FDI is smaller than the counterfactual FDI except for the period from the late 1980s to the early 1990s. While their study has important policy implications, their analysis did not cover the recent period when the Japanese outward FDI stock grew rapidly (Fig. 1).¹¹

Based on this background, this paper examines whether Japan's outward FDI still has untapped growth potential or not. To do so, we estimate a gravity model and compare Japan's actual outward FDI stock with the counterfactual FDI stock. In addition to covering a more recent period than the previous studies, our paper introduces a methodological improvement on the studies of Japanese outward FDI, many of which estimated a log linear form of the gravity model. A problem is that many country pairs have no FDI between them. Taking a log linear form implies that the analysis drops the country pairs with zero FDI. However, throwing away the observations with zero FDI results in inconsistent parameter estimates. To solve this problem, we employ the Pseudo-Poisson Maximum Likelihood estimator proposed by Santos Silva and Tenreyro, 2006. An additional contribution of our research is that we supplement our analysis of aggregate outward FDI stock with an examination of the regional distribution of Japan's actual versus counterfactual FDI stock.

Our results show that Japan's actual outward FDI exceeded its counterfactual FDI from 2013, and the ratio between the two is the highest among OECD countries as of 2015. The host countries with the highest actual-to-counterfactual ratios for Japanese FDI are Asian countries involved in Japanese supply chains (i.e., Indonesia, the Philippines, Thailand and Vietnam) but also include South Africa and the United States. Although President Trump recently pressed Japanese business leaders to invest more in the US, our research shows that the US hosted 1.7 times more Japanese FDI than the value predicted by the gravity model as of 2015. Our results imply that Japan has no unrealized potential for outward FDI at the aggregate level, nor at the country-level for the US in particular.

The paper is organized as follows. The next section introduces a gravity model of bilateral FDI. The section also describes the estimation method and the data that we use in this paper. Section 3 reports the estimation results and discusses their implications, while Section 4 presents robustness checks. Section 5 includes our conclusions and discussion of results.