Introduction and Overview

Abstract

The number of Japanese exporters continued to increase in the decade after 2000. Figure 1, using data from the Basic Survey of Japanese Business Structure and Activities which covers Japanese manufacturing firms with over 50 employees and more than 30 million yen of capital stock, shows that the number of exporters in the manufacturing sector increased from 3,762 firms in 2000 to 4,518 firms in 2010. Figure 1 also depicts the type of exporters that increased, with firms categorized into deciles by export ratio, which is defined as the ratio of their exports to total sales. This figure reveals an increase of exporters classified in the export ratio categories of 10 % or more. This trend is similar to even non-manufacturing firms, including wholesalers, retailers, and firms in the service sectors.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Increase of Exporters and FDI Firms

The number of Japanese exporters continued to increase in the decade after 2000. Figure 1, using data from the Basic Survey of Japanese Business Structure and Activities which covers Japanese manufacturing firms with over 50 employees and more than 30 million yen of capital stock, shows that the number of exporters in the manufacturing sector increased from 3,762 firms in 2000 to 4,518 firms in 2010.¹ Figure 1 also depicts the type of exporters that increased, with firms categorized into deciles by export ratio, which is defined as the ratio of their exports to total sales. This figure reveals an increase of exporters classified in the export ratio categories of 10 % or more. This trend is similar to even non-manufacturing firms, including wholesalers, retailers, and firms in the service sectors.

Fig. 1

The number of exporting firms by export ratio (manufacturing firms). Data source: “The Basic Survey of Japanese Business Structure and Activities” from 1997 to 2010, the Ministry of Economy, Trade and Industry (METI)

The line in Fig. 2 shows that the number of Japanese firms making foreign direct investment (FDI) also increased remarkably over 10 years, from 2,592 firms in 2000 to 3,378 firms in 2010. The number of foreign subsidiaries of Japanese firms, depicted as the bars in the figure, increased even more rapidly from 8,872 to 16,457. The average number of subsidiaries per parent firm increased from 3.4 firms in 2000 to 4.9 firms in 2010. Note that the increases are different among regions. The subsidiaries in Asia increased rapidly from 6,082 in 2000 to 11,261 in 2010, especially in China, where these rose from 4,077 in 2005 to 5,631 in 2010. Subsidiaries in the European Union also increased, from 901 in 2000 to 2,127 in 2010, whereas those in the United States did not increase significantly.

Fig. 2

The number of FDI firms and their overseas affiliates by region (manufacturing firms). Data source: “The Basic Survey of Japanese Business Structure and Activities” from 2000 to 2010, the Ministry of Economy, Trade and Industry (METI)

The increase in Japanese firms’ subsidiaries varied among firms whose stock shares are different. Table 1 presents the proportion of subsidiaries, by classifying them into three categories according to the stock owned by the parent firm: 100 % owned, 50–100 % owned, and 20–50 % owned. The fully-owned subsidiaries, in particular those in Asia and Europe, increased, while those of 20–50 % owned, in particular in Asia, decreased. Firms in the wholesale, retail, and service sectors show a similar trend to the manufacturing firms.

Turning to the performance of internationalized firms, defined as Japanese exporters and FDI firms, it is seen that they are characterized by high premiums in employment size, value added, and total factor productivity (TFP) in comparison with domestic firms, as depicted in Fig. 3. The bars in the figure represent exporters, and the lines represent FDI firms. In comparison with a large number of Japanese manufacturing firms,² Japanese internationalized firms are also among “the happy few” as Mayer and Ottaviano (2007) described the higher productivity of European internationalized firms relative to those serving only the domestic market.

Table 1 The number and share of subsidiaries by ownership and region

Fig. 3

Premium of Japanese exporters and FDI firms (manufacturing firms, 1997–2005). Data source: Author’s calculation from the data “The Basic Survey of Japanese Business Structure and Activities” from 1997 to 2005, the Ministry of Economy, Trade and Industry (METI)

2 Research Questions and Book Scope

Since the late 1990s, trade economists have focused on analyzing the productivity heterogeneity of exporters and FDI firms at the firm level. Not only theoretical analyses but also empirical evidence from US and European firms clarified that exporters and FDI firms are heterogeneous in productivity. The seminal papers by Bernard and Jensen (1995, 1999), Melitz (2003), and Helpman et al. (2004) confirmed that firms with relatively high productivity tend to be exporters, the most productive firms engage in FDI, and the least productive firms serve only the domestic market. Many studies have been conducted on US and European firms, whereas Japanese firms have attracted insufficient research attention despite the large number of exporters and FDI firms.

This book aims to analyze the characteristics of Japanese internationalized firms by using the micro-level data of Japanese firms and thereby filling the research gap between US and European firms, and Japanese firms. The book’s research scope examines the following questions:

• 1. Internationalized firms, that is, exporters and FDI firms, are characterized by premiums in firm size, wage rate, and productivity. What are the specific features of Japanese internationalized firms compared with US and European firms?

• 2. The proportions of exporters and multinational enterprises (MNEs) vary substantially across industries, reflecting industry-specific attributes. What industry-specific factors make the modes of firms’ internationalization different among industries?

• 3. Although the productivity of firms serving foreign markets is on average higher than that of firms serving only the domestic market, the difference in productivity between exporters and FDI firms is not clear. What factors other than productivity should be included as important determinants of the export and FDI behavior of Japanese firms?

• 4. What are the features of Japanese FDI firms, and how are they different from US and European firms in terms of subsidiary sales, the number of investing countries, and the scale of operations in the home country?

• 5. The modes of internationalization, export and FDI, are determined not only by productivity, but also by market-specific factors in destination countries. Different modes of internationalization may be chosen by Japanese firms corresponding to their different destination countries, US and European countries and East Asian countries. If so, what market-specific factors make the modes of firms’ internationalization different among destination countries?

• 6. In comparison with US and European firms, Japanese firms do not present clearly that the most productive firms undertake FDI. There may be other reasons more productive firms export while less productive firms undertake FDI. Assuming that FDI is decided under an environment in which firms and managers make matches for production, the question is whether the match quality in the market affects the FDI decision.

• 7. Intra-firm trade undertaken by MNEs is increasing. How do organizational and institutional factors affect such trade?

3 Features of Internationalized Firms

The firm heterogeneity model of Melitz (2003) predicts that more productive firms engage in exports, while less productive firms serve only the domestic market since exporting requires additional costs. Helpman et al. (2004) extend the model and predict that the most productive firms engage in FDI, the less productive firms engage in exporting, and the least productive firms serve only the domestic market. A number of empirical studies have examined the relationship between firm characteristics and internationalization. They have found a positive correlation between firm performance and its internationalization in line with the standard firm heterogeneity model by Melitz (2003) and Helpman et al. (2004). In chapter “Features of Japanese Internationalized Firms: Findings Based on Firm-Level Data”, following these previous studies, Wakasugi, jointly with Todo, Sato, Matsuura, Ito, and Tanaka provides a comprehensive analysis of the internationalization of Japanese manufacturing firms. By using firm-level data on the Japanese manufacturing industry for the period 1997–2005 from “Kigyo Katsudo Kihon Chosa” (the Basic Survey of Japanese Business Structure and Activities),³ they first examine the characteristics of internationalized Japanese firms, namely firms that engage in exports and/or FDI. Second, we compare internationalization of Japanese firms with that of firms from selected European countries. The empirical results in this chapter are in line with those of previous works: the number of internationalized firms is very small and export firms are larger and more productive than domestic firms, while those that engage in both exports and FDI are even larger and more productive. The overall results show that the characteristics of internationalized firms in Japan are mostly similar to those of their European counterparts: (1) exports are dominated by a few top exporters; (2) although the export-to-sales ratios of very few firms exceed 50 %, these firms account for at least half of total exports; (3) internationalized firms perform better in terms of a number of the analyzed indicators than domestic firms; (4) the proportion of foreign-owned firms is higher among exporting than among non-exporting firms; (5) the number of FDI firms (extensive margin) has a larger influence on total sales by overseas subsidiaries than sales per firm (intensive margin). However, the analysis reveals notable differences between Japan and Europe in that productivity differences between domestic firms, exporting firms, and FDI firms are substantially smaller in Japan than in Europe. This finding suggests that variations in productivity alone cannot explain the export and FDI behavior of Japanese firms. The analysis also finds other remarkable differences: the dominance of exports by the top exporters has weakened over time and the proportion of foreign-owned firms among exporters in Japan is much lower than that in European countries.

4 Industry and Firm-Specific Factors

In chapter “Heterogeneity and the Structure of Export and FDI: A Cross-Industry Analysis of Japanese Manufacturing”, Tanaka focuses on the fact that the fractions of exporters and MNEs vary substantially across industries. For example, according to Bernard et al. (2007), the number of US firms exporting is nearly 40 % in some manufacturing industries but less than 10 % in others. As shown in this chapter, the variation in the fraction of exporters and MNEs across industries is systematic. First, the fraction of the sum of exporters and MNEs is higher in industries with a larger dispersion of sales. Second, the fraction of MNEs alone is higher in industries with a larger dispersion of sales. Third, relative to all active firms, MNEs are heavily concentrated in research and development (R&D)-intensive industries. This chapter uses a firm heterogeneity model presented by Helpman et al. (2004) and derives the theoretical relationship between firm heterogeneity and the fraction of internationalized firms. The model shows that industries with a larger degree of productivity dispersion have a larger fraction of MNEs, a larger fraction of the sum of exporters and MNEs, and a larger ratio of MNEs to non-MNE exporters, although the effect of an increase in the dispersion of productivity on the fraction of exporters can be either positive or negative. The model also shows that R&D-intensive industries have an advantage in conducting FDI. The empirical analysis in this chapter employs Japanese industry-level data for the 1997–2005 period from the Basic Survey of Japanese Business Structure and Activities. Using reduced-form specification, this chapter empirically analyzes the effect of the measure of firm-size dispersion, R&D intensity, and other variables on the following: (1) the fraction of exporters, (2) the fraction of MNEs, (3) the ratio of MNEs to non-MNE exporters, and (4) the fraction of the sum of exporters and MNEs. The results reveal that industries with a larger degree of productivity dispersion have a larger fraction of MNEs, larger ratio of MNEs to non-MNE exporters, and larger fraction of the sum of exporters and MNEs. In addition, the results reveal that MNEs are concentrated heavily in R&D-intensive industries. However, they do not confirm the positive relationship between R&D intensity and the fraction of non-MNE exporters against our model’s prediction. This suggests a need for a model that is more consistent with the data. The analysis also sheds light on the traditional source of comparative advantage, such as capital intensity and skill intensity. In particular, most of the estimation results show that capital intensity and skill intensity have no significant coefficient on the fraction of internationalized firms. This suggests that these traditional variables are less important in the structure of export and FDI than firm heterogeneity and R&D intensity. This chapter also shows that firm heterogeneity and R&D intensity play crucial roles in the structure of foreign trade and investment. Greater dispersion in productivity across firms within a single industry is associated with more FDI, as predicted in the model, and also with more exports. In addition, R&D-intensive industries have a larger fraction of MNEs.

Recent empirical studies on international trade at the firm level have found that firms engaging in export or FDI are generally more productive and larger than those firms serving only domestic markets. This finding is consistent with the theoretical predictions of heterogeneous firm trade models, most notably those of Melitz (2003) and Helpman et al. (2004), in which only productive firms are able to pay the entry costs associated with export and FDI and hence serve foreign markets. However, many empirical studies have also found that although the productivity of firms serving foreign markets is on average higher than that of firms serving only the domestic market, the productivity distributions of the two types of firm overlap significantly. This evidence implies that non-productivity factors are important determinants of the export and FDI behavior of firms. In chapter “The Role of Non-productivity Factors in the Internationalization of Firms”, in investigating the role of unobserved firm heterogeneity rather than productivity heterogeneity in the internationalization of firms, Todo applies a multinomial logit model with random intercepts and random coefficients (a mixed logit model) of export and FDI decisions to firm-level data for Japan, based on Todo (2011). The inclusion of random intercepts and random coefficients on prior firm status in the export and FDI decisions may control for unobserved firm heterogeneity and correct for the biases associated with endogeneity. He then uses the resulting estimation results to examine the quantitative effects of productivity and the unobserved firm-specific random effects. From the mixed logit estimation, this chapter finds that the effect of productivity on the internationalization of firms is statistically significant but economically negligible. The effect of other observable firm characteristics such as firm size, the degree of credit constraints, and access to information is also found to be very small in magnitude. Rather, the internationalization of firms is determined mostly by their previous experience in foreign markets and firm characteristics that are unobserved in standard firm-level data. This chapter further shows that in the case of Japanese small and medium enterprises (SMEs), unobservable non-productivity factors such as the risk and time preferences and the international experience of decision makers affect export and FDI behavior, based on Todo and Sato (2011). These results imply that entry costs represent the major barrier to firm internationalization and that unobserved firm characteristics such as the international experience of decision-makers are important determinants of entry costs. In addition, because firms are more likely to be concerned about the long-term discounted risk-averse utility from internationalization than the one-time risk-free utility that is assumed in standard heterogeneous firm models of trade, the risk and time preferences of decision-makers influence firm internationalization. These findings clearly indicate why there is a significant productivity distribution overlap between domestic and internationalized firms.

5 Internationalization in Multiple Regions

In chapter “Entry into Foreign Markets Through Foreign Direct Investment”, employing firm-level data, Matsuura and Sato examine patterns of Japanese FDI in detail. Starting with an overview of recent trends in Japanese FDI, they point out several empirical constants on Japanese FDI, among others that a small number of MNEs have foreign subsidiaries in multiple countries. MNEs with larger scales of operations in the home country tend to penetrate a greater number of overseas markets, and only such MNEs are able to enter less popular markets. Productivity for larger firms is relatively higher than that for small firms. Hence, patterns of FDI are substantially influenced by firm heterogeneity in productivity. Another important observation is that a substantial proportion of FDI subsidiaries functions as export platforms. Such foreign subsidiaries serve their local markets as well as neighboring markets via exports. This tendency is particularly prominent for foreign subsidiaries located in East Asia. Motivated by these observations, the researchers extend the standard Melitz-type firm heterogeneity model by allowing FDI subsidiaries to deliver goods not only to host countries’ markets but also to neighboring countries’ markets. Thus, the model highlights that not only market sizes for destination countries but also those for their neighboring countries may affect firms’ FDI decisions. Referring to such an extended market concept as “market potential,” the model shows that market potential may affect FDI sales in terms of both the number of foreign affiliates (extensive margin) and average FDI sales per firm (intensive margin) through changing the cut-off level of firms’ productivity. Indeed, Japanese FDI data show that a positive correlation between destination countries’ market sizes and the number of each destination’s foreign subsidiaries is quite weak, which suggests that some other factors, including market potential, might play an important role in determining extensive margins of FDI. Finally, using micro data on Japanese FDI sales, they estimate FDI intensive margins (average FDI sales per firm) and extensive margins (number of MNE subsidiaries) to confirm that the inclusion of market potential improves the fitness of the estimated gravity equation. The estimation results show that the market potential has a positive effect on both intensive margins and extensive margins of FDI sales. However, the coefficient for extensive margin is not statistically significant. The estimation coefficients for destination countries’ market sizes measured in real gross domestic product (GDP) are always significantly positive for extensive margins. Hence, a possible interpretation is that Japanese firms tend to set up foreign subsidiaries by initially targeting the markets of destination countries and only later on do they consider exporting from subsidiaries. Another interesting result of the estimation for Japanese FDI sales is that the negative effect of geographical distance on extensive margins is substantially greater than that obtained from the US FDI sales data (Yeaple 2009). This result suggests that the difference is partially attributable to the heavy concentration of Japanese FDI in East Asia.

The theoretical model by Helpman et al. (2004) reveals that various combinations between firms’ internationalization strategy and productivity levels are observed, corresponding to different market conditions. While previous empirical studies investigated the relationship between firm-productivity levels and mode of internationalization in the world market as a whole, studies that examine how the modes of Japanese firms’ internationalization may vary with different market-specific factors are hard to find. In chapter “Productivity and Modes of Internationalization: Evidence from Japanese Firms” Wakasugi and Tanaka attempt to fill the gap by empirically examining whether Japanese firms’ productivity levels relate to their mode of internationalization with regard to exports to, and/or overseas FDI production in, countries of the North (i.e., North America and Europe) and the South (i.e., East Asia). This chapter statistically answers two questions: how extensive is the variation in firms’ productivity corresponding to their destinations of internationalization between the North and the South and how significantly does firms’ productivity divide the modes of internationalization in the North and the South. The empirical analysis uses firm-level data pertaining to 12,000 Japanese firms to reveal some interesting results: (1) internationalized Japanese firms show higher productivity levels than non-internationalized Japanese firms, regardless of market destinations and the modes of internalization adopted; (2) firms engaged in FDI in the North have higher productivity levels as compared to firms that export to the North; but (3) firms engaged in FDI in the South do not apparently have higher productivity levels than firms that only export to the South. The third result is contrary to the Helpman et al. (2004) model’s prediction about an internationalization hierarchy where internationalized firms with higher productivity levels engage in FDI, while those with lower productivity levels export their products. The difference in variable and fixed costs, including transport costs and market sizes between the two regions, may be a source of the different productivity cut-off levels pertaining to FDI engagements by Japanese firms in the North and the South. The fact that wage rates in East Asian countries (the South) are lower than in the North, while they are similar across North America, Europe and Japan, supports this argument.

6 Management and Organization

In chapter “Foreign Direct Investment with Matching Frictions”, Sato discusses the FDI decisions of individual firms under an environment in which firms and managers have to make matches for production. More specifically, the author extends the standard Melitz-type firm heterogeneity model by incorporating the simple search and matching framework proposed by Rauch and Trindade (2003). This extension is motivated by an empirical fact that even though FDI firms are on average more productive than non-FDI firms, this hierarchy is not necessarily so clear, as will be discussed in chapter “Features of Japanese Internationalized Firms: Findings Based on Firm-Level Data”. This ambiguous “pecking order” indicates a limit of the standard Melitz-type model for firms’ decisions on FDI (e.g., Helpman et al. 2004). The model in this chapter expresses the following insights: (1) unfamiliarity about foreign countries is likely to make the search for efficient managers in foreign countries more difficult than in the home country, (2) consequently, matches in foreign countries tend to be associated with uncertainty about the quality of mangers, (3) when they hire low-quality managers, firms may not fully exert their intrinsic productivity level in foreign production, and (4) worse, even highly productive firms may not find appropriate mangers and may be forced to choose exports rather than FDI. Although the underlying idea of the model is rather simple, the author derives two interesting prognoses, which could shed light on some empirical findings that are not explained well by the standard Melitz-type model. First, predicted distributions of FDI firms are much more akin to real data than those suggested by the basic firm heterogeneity model, namely, there exists a range of firm productivities in which more productive firms may export while less productive firms may undertake FDI. Such a range of firm productivities becomes wider when either matching frictions increase or trade costs decline. Second, the model provides an explanation for the empirical finding of Yeaple (2009) that the ratio of an FDI firm’s foreign-affiliate operation size to its home-operation size tends to decrease with distance and increase with the usage of common language, holding other things constant. This finding implies that the unit production cost in the foreign factory would be systematically different from that in the home factory, which never occurs in the standard Melitz model. By contrast, in the model presented in this chapter, the productivity of foreign affiliates depends not only on firms’ intrinsic productivity levels but also on match quality. The average match quality is likely to decline in foreign countries in which it is difficult for firms to collect information about appropriate managers as much as they can do in the home country, which implies that the operation size becomes relatively small. Thus, the empirical finding that the ratio of the foreign-affiliate operation size to home operation size tends to decrease with distance and increase with the usage of common language can be readily understood in the model, given that the degree of matching frictions is negatively correlated to geographical proximity between the FDI host and home countries or the usage of a common language.

In the final chapter, Matsuura and Ito focus on increasing intra-firm trade undertaken by MNEs and empirically examine the determinants of intra-firm trade. This chapter sheds light not only on factor prices and trade costs, but also on organizational structure in terms of the ownership of overseas plants and the control over intermediate inputs for further processing. Regarding the relationship between the intra-firm trade and the control over intermediate inputs, Feenstra and Hanson (2005) have pointed out that the decision-making of MNEs on whether to supply intermediate goods from the home country to their foreign affiliates, or to engage in local procurement, is dependent on the value-added ratio of the affiliate firm and the contract environment in the local market. According to the property rights approach, the control over inputs should be given to local managers when their efforts measured by the value-added ratio are crucial. In this case, as the local procurement of intermediate inputs is optimal by allocating the control right over inputs, intra-firm trade consequently decreases. On the other hand, when local manager efforts are not important, the incentive system implies that both ownership and control should be allocated to the foreign firm. In addition, contractibility might affect the organizational structure. In a poor business environment, MNEs hesitate to give control rights over inputs to local managers since the cost of negotiations would be extremely high. Thus, when local manager efforts are important and the degree of legal enforcement is adequate, the control rights over inputs are given to the local managers. In contrast, when the degree of contractibility in the host country is low or the managers’ efforts are not crucial, foreign firms maintain both ownership and control over inputs by dispatching a manager from their headquarters and by increasing intra-firm trade. In line with these hypotheses, this chapter examines how allocating the control right over intermediate inputs affects intra-firm trade, using Japanese foreign affiliate-firm level data. Since organizational structure and intra-firm trade are jointly determined, instrumental variable (IV) estimations are employed. Results of the random effects probit model show that control over input decisions is positively correlated with the value-added ratio of affiliated firms, as predicted. The results of the random effects IV regression on intra-firm trade clearly indicate that granting control rights over input purchases to local managers has a large impact on the procurement of intermediate inputs from Japan, after controlling for endogeneity. This finding contributes to the literature by suggesting that control over input decisions critically affects the intra-firm trade of intermediate inputs. This chapter also considers differences in country-specific institutional qualities and shows that intra-firm procurement is increasing in countries that display poor performance of institutional factors. This result suggests that the improvement of legal institutions is a crucial factor for boosting the purchase of local inputs by MNEs. Further, it is suggested that policymakers should provide an FDI-friendly environment where MNE affiliates can operate for a long period of time. This is because the delegation of decision rights to local residents and the subsequent increase in local procurement is time consuming.

The articles in this book analyze empirically the features of Japanese internationalized firms, using micro-level data on Japanese firms. Such use of firm-level data is essential to investigate the sources of internationalization: firm-specific and industry-specific factors including productivity heterogeneity, management, and organization, and market specific factors. The use of Japanese firm-level panel data in each chapter is a unique advantage of this book.