8  The Chinese Economy

The Chinese Economy, vol. 47, no. 3, May–June 2014, pp. 8–22. © 2014 M.E. Sharpe, Inc. All rights reserved. Permissions: www.copyright.com ISSN 1097–1475 (print)/ISSN 1558–0954 (online) DOI: 10.2753/CES1097-1475470301

Wen-Jen Hsieh, Jun-Jen Huang, and Ching-Lin Wei

Is There a Bilateral Trade-Off Between Foreign Direct Investment and Trade? Case of China, Japan, and Korea Abstract: Larger inflows of foreign direct investment (FDI) induce a high volume of trade because supply chains set up by multinational enterprises intensify trade networks across nations. Several empirical studies have uncovered complementary relationships between trade and FDI among East Asia nations, but do not consider the dynamic transition of the Chinese economy and the role of the trilateral freetrade agreement (FTA). This article studies the feedback effects of FDI on trade among China, South Korea, and Japan from 1994 to 2010. Our empirical models capture the dynamic transition of trade and FDI between China and Korea or Japan and can also be used to predict the impact of the trilateral agreement on the network of trade and FDI among these countries. Our results indicate that the Trilateral Agreement could generate longterm positive reciprocal benefits from China to Japan and Korea. China has become the economic heavyweight of the global market, and its role in East Asia has never been so profound. Since 2004, China has surpassed the United States as the top trade and foreign investment partner of Japan and South Korea (Zhang, Zhang, and Fung 2007). Countries in East Asia have begun to cement a network of trade and investment with China, and in 2002, China and ASEAN (As-

Wen-Jen Hsieh is a professor in the Department of Economics, National Cheng Kung University, Taiwan; email: [email protected]. Jun-Jen Huang is an assistant professor at the Graduate Institute of Finance, National Taiwan University of Science and Technology, Taiwan. Ching-Lin Wei is a research assistant in the Department of Economics, National Cheng Kung University, Taiwan. The authors would like to thank Professor Ying-Yi Tsai of the National University of Kaohsiung for his valuable comments. 8

may–june 2014  9

sociation of Southeast Asian Nations) signed an agreement which led to the creation of the ASEAN-China Free-Trade Area (ACFTA) in 2010.1 The ASEAN-China Investment Agreement was signed in 2009. This agreement aims to build a favorable environment for investments in the ASEAN countries and China. Meanwhile, in 2004, Korea proposed a trilateral free-trade agreement between itself, China, and Japan. The agreement aimed at boosting international relationships, disaster relief, and the flow of trade and investment in East Asia. The first summit meeting of the leaders of China, Japan, and Korea (CJK) was held in 2008, and the agreement was institutionalized on September 2011 with the launching of a Trilateral Cooperation Secretariat in Seoul, Korea. In May 2012, the leaders of the CJK countries signed the Trilateral Agreement not only to establish a free-trade area, but also to promote, facilitate, and protect investment. The negotiations on the trilateral agreement, at least to some extent, highlighted the competition between China and the United States and defined the trade landscape in Asia (Wignaraja 2011). The success of the Regional Comprehensive Economic Partnership (RCEP), which includes the ASEAN countries, Australia, New Zealand, India, and the CJK countries, will hinge upon the outcome of the CJK Summit meetings. However, for Japan, Korea, and even the ASEAN countries, the agreement could further increase their dependency on China through international trade and foreign investment.2 The Chinese penetration of other economies in the region, by means of trade and foreign direct investment (FDI), is so profound and dynamic that the conventional theories on trade and FDI cannot adequately explain the phenomenon. Theoretically, relationships between FDI and trade flows hinge on different patterns of FDI. It is usually held that horizontal FDI inflows substitute for imports (Horstmann and Markusen 1992), whereas vertical FDI outflows increase exports (Helpman 1984). Nonetheless, China has witnessed a mixture of both vertical and horizontal FDI since the economic reform in 1978. Since the mid-1980s, firms in Hong Kong, Taiwan, Japan, Korea, and other Asian countries have gradually moved their labor-intensive industries to China. Beginning in 1992, inward FDI in China rapidly accelerated. As a result, China became the largest recipient of FDI among developing countries in 1993, and one of the world’s top three recipients of FDI from 2003 to 2005 (Nanto and Chanlett-Avery 2006). Foreign-invested enterprises (FIEs) play a vital role in the growth of China, with wholly owned foreign affiliates beginning to be prevalent. The model of wholly owned foreign affiliates has been adopted by most Asian firms in China, even more so than equity joint-ventures in FIEs. A vertical trade network incorporating the trade between parent foreign firms and their affiliates in China accounts for a large portion of China’s trade. In the first decade of the twenty-first century, foreign mergers and acquisitions (M&As) became a major element of FDI inflows. The rise of cross-border M&As in China can be attributed mainly to the rapidly expanding local consumer markets (Davies 2012). Nevertheless, there has been growing concern over the hollowing-out effect, the prospect of a majority of investment from many of China’s neighbors shifting to China because of its relatively low labor costs. Korea, for example, has

10  The Chinese Economy

experienced outward FDI to China with capital goods and technology-intensive industries (Gaulier, Lemoine, and Ünal-Kesenci 2005). Theoretically, heterogeneity has comparative advantages in countries that allocate their specialized production processes in a particular segmentation. Orthodox trade theory asserts that the difference in factor proportions stimulates FDI from countries with relatively abundant capital and labor shortages to countries with the opposite of that factor proportion (Helpman and Krugman 1985; Krugman and Obstfeld 2008). China can make a significant contribution to cross-border production networks, and thereby create more opportunities for trade and FDI.3 From a long-term perspective, inward FDI to China would be regarded as complementary rather than competitive for China’s neighbors. However, policymakers throughout Asia are convinced that China is absorbing so huge a share of FDI that neighboring countries are losing manufacturing industries and jobs. Foreign direct investment substituting trade between China and neighboring countries may be responsible for the hollowing-out effects in neighboring countries. If inward FDI to China accumulates at the expense of neighboring countries, the increase in FDI flows to China will generate a corresponding decline in trade with China, thereby weakening the vitality of all of the concerned economies. The research in this article investigates how China influences Korea and Japan through FDI and trade. It examines how the bilateral flows of FDI from Japan and Korea interact with China to affect bilateral trade. The impact of FDI on trade among China, Korea, and Japan from 1994 to 2010 is analyzed. The results of this research make a contribution to an understanding of how FDI networks influence trade among China, Korea, and Japan and the issue of whether China is crowding out exports from Japan and Korea by absorbing FDI. Although relationships among CJK have been deteriorating, the leaders of the three countries have decided to focus on economic issues on the eve of the establishment of the CJK Trilateral Agreement. Literature Review Both substitute and complementary relationships between FDI and trade have been inspected. Blonigen (2001) analyzed whether Japanese production of automobiles in the United States was complementary to Japanese exports of automobiles to the United States during the late 1970s to early 1990s. Results show both substitute and complementary effects. However, regarding product-level data (Japanese-produced automobiles), the only evidence found that Japanese production substituted Japanese exports. Swenson (2004) studied how inward FDI affected U.S. imports from 1974 to 1994. He found that FDI substituted for imports in the areas of product and industry, but encouraged imports in areas of manufacturing. On the other hand, research shows that vertical FDI can create trade between developing and industrialized counties. Aizenmana and Noy (2006), using an annual panel data set of 81 countries from 1982 to 1998, discovered that linear feedback from FDI to trade (FDI “Granger” causes trade) is stronger in developing than in

may–june 2014  11

industrialized countries. Although this finding is consistent with the conjecture that outward FDI to developing countries is vertical, feedback on the influence of FDI on trade is significantly positive regardless of whether the countries are industrialized or developing. Some empirical studies have uncovered a complementary relationship between trade and FDI among East Asia nations. Fukao, Ishido, and Ito (2003) tested the relationship of bilateral vertical intra-industry trade (IIT) between Japan and its 43 major trading partners in East Asia from 1988 to 2000, surmising that FDI makes a significant contribution to the rapid growth of vertical IIT. A study of Chinese bilateral IIT in 50 countries from 1992 to 2001 found evidence supporting that FDI has a positive impact on vertical intra-industry trade in China (Zhang, Witteloostuijn, and Zhou 2005). Xing (2007) showed that Japanese direct investment in China was a remarkable driver of bilateral IIT between China and Japan based on panel data from 1980 to 2004. Despite these findings, there are several research gaps in the above-mentioned empirical studies that need to be filled in. First of all, the patterns of FDI and trade between China and its major trading partners encompass both vertical and horizontal trade and FDI. Second, some existing studies show that economic integration could boost FDI among member countries (Büthe and Milner 2008; Guerin and Manzocchi 2009). China joined the WTO in 2001 after the 1997 Asian financial crisis Years later, the world witnessed the launching of the trilateral agreement between China, Japan, and Korea. However, most empirical studies give no indication of how economic integration (e.g., a bilateral free-trade agreement) would affect the relationship between trade and FDI. Therefore, future studies should consider whether the increase of FDI to China under the CJK Trilateral Agreement has reduced China’s trade with its Asian neighbors. Methodology and Data Selected Variables and Definitions The econometric models developed in the research estimate how bilateral FDI activities with China affect bilateral trade with Japan and Korea. Bilateral FDI in China is both inward and outward with Japan and Korea; similarly, bilateral trade also includes Chinese imports or exports with Japan and Korea. To be specific, the major selected variables in the research are inward FDI from China to Japan or Korea, outward FDI from Japan or Korea to China, exports from Japan or Korea to China, and imports from China to Japan or Korea. In order to study the relative influence of China, bilateral trade between Japan/Korea and China is transformed into a percentage of total trade of Japan/Korea. Bilateral FDI between Japan/Korea and China is measured as a percentage of the total flow of bilateral (inward or outward) FDI to the total stock of (inward or outward) FDI. We selected relative real GDP growth and exchange rates as our control variables. Relative real GDP growth per capita is defined as the difference of real GDP growth between China and its

12  The Chinese Economy

partner countries. Table 1 lists the definitions and abbreviations of the variables used in the econometric models. Econometric Models We employ the panel data method to study the annual statistics of bilateral FDI and trade, relative real GDP growth per capita, and exchange rates. The interaction effect between FDI and trade is also included, although it is rarely considered in the literature. The interaction terms (FDI multiplies trade) can reveal whether partial effects of FDI on trade will additionally increase or decrease along with trade. After economic integration, it is expected that trade will rise and thereby influence this partial effect. For example, if outward FDI negatively affects export, it is likely that this negative effect will be either neutralized or intensified due to the increase of exports stimulated by free-trade agreements (FTAs). In other words, export orientation policies or FTAs could have a spillover effect on FDI. Since the patterns of FDI among China, Korea, and Japan are relatively more dynamic than trade, structural changes are also relevant. We investigate whether there was a structural change of FDI on trade before or after 1998. We further postulate that FDI could have increased trade before 1998, while trade might have been substituted by FDI after 1998. We also include the J-curve effects in our models.4 The J-curve describes a situation whereby the depreciation of a country’s currency cannot cut the trade deficit initially, as business activities need time to react to the change in currency prices. When there is a J-curve effect, it is expected that current real exchange rates (CREit) will negatively affect exports to China (EXCNit), but its lag value (CREit – 1) will have a positive effect on EXCNit. Finally, in order to avoid problems arising from unobserved factors correlated with explanatory variables, panel data with fixed-effects models are adopted (Liu 2008). The empirical models are presented below. The export model is represented by Equation (1).

EXCN it = αex + αiex + δex t + β out 1OUTFDICN it-1 + βout 2 OUTFDICN it-2 + β1998 out 1 D1998 * OUTFDICN it-1 +β1998 out 2 D1998 * OUTFDICN it-2 + β ex EXCN it-1 ex REGROWTH it-1 +βout _ ex EXCN it-1*OUTFDICN it-1 + βrgdp ex ex +βex cr CRE it + β cr 1CRE it-1 + εit



(1)

ex where i = (Japan, Korea), t = (1994, 1995, …, 2010), αi is the group (country) ex effect of exports, δt is the time (year) effect of export ∑ αiex = ∑ δext = 0 , eitex is the t error term of the export model, D1998 = 1 for all the yearsi after 1998 and zero oth-

may–june 2014  13

Table 1 Definitions of Variables Variable

Abbreviation

Measurement

Exports to China

EXCN

Ratio of Japanese or Korean exports to China to total Japanese or Korean exports.

Imports from China

IMCN

Ratio of Japanese or Korean imports from China to total Japanese or Korean imports.

Inward FDI from China

INFDICN

Ratio of FDI inflows from China to Japan or Korea to inward FDI stock of Japan or Korea from China.

Outward FDI to China

OUTFDICN

Ratio of FDI outflows from Japan or Korea to China to outward FDI stock from Japan or Korea in China.

Relative growth of GDP per capita

REGROWTH

Difference between growth of real GDP per capita of China and of Japan or Korea.

Real exchange rates

CRE

Cross-rates based on 2005 prices.

Formula Japanese exports to China ÷ total Japanese exports Korean exports to China ÷ total Korean exports Japanese imports from China ÷ total Japanese imports Korean imports from China ÷ total Korean imports Chinese FDI inflows to Japan ÷ inward FDI stock from China in Japan FDI inflows from China to Korea ÷ inward FDI stock from China in Korea Japanese FDI outflows to China ÷ outward FDI stock from Japan in China FDI outflows from Korea to China ÷ outward FDI stock from Korea in China China GDP (per capita) growth rate—Japan GDP (per capita) growth rate China GDP (per capita) growth rate— Korea GDP (per capita) growth rate Japan yen/China RMB Korea won/China RMB

14  The Chinese Economy

erwise, and D1998*OUTFDICNit – 1 and β1998 are structural out 2 D1998 * OUTFDICN it-2 terms of the export model. The import model is represented by Equation (2). im IMCN it = αim + αim i + δt + β in1INFDICN it-1 998 +βin 2 INFDICN it-2 + β19 in1 D1998 * INFDICN it-1

+β1998 in 2 D1998 * INFDICN it-2 + β im IMCN it-1 +βin _ im IMCN it-1*INFDICN it-1 + βim rgdp REGROWTH it-1 im im +βim cr 1CRE it + β cr 2 CRE it-1 + εit

(2)

where aiim is the group (country) effect of imports, diim is the time (year) effect of imports, eitim is the error term of the import model ∑ αimi = ∑ δimt = 0 , and D1998 i t *INFDICNit – 1 and D1998*INFDICNit – 2 are structural terms of the import model. Through structural terms and interaction terms (i.e., EXCNit – 1*OUTFDICNit – 1), these econometric models will predict whether increasing FDI in China could reduce China’s trade with its Asian neighbors after the establishment of the CJK Trilateral FTA. Justification of Variable Selection Foreign direct investment can be either complementary to trade or can substitute trade. In most cases, if commodities made by foreign firms target the host economy’s domestic markets, then FDI will substitute trade (Vavilov 2008). In terms of market size, countries with large markets will attract FDI and trade, since large markets create ample opportunities for foreign firms. Currency prices influence both trade and FDI. A weaker currency is favorable for export and outward FDI. Lastly, as Levin, Lin, and Chu (2002) pointed out, a free-trade agreement has both positive and negative effects on FDI. Since the FTA results in a reduction of tariffs, the transaction costs of vertical FDI will drop after an FTA takes effect. This cost-driving motivation will lead to FDI growth. However, if trade merchandise substitutes for commodities produced through FDI (mainly horizontal FDI), the decrease of interregional tariffs stipulated by the FTA will cause a reduction in FDI. Therefore, the models show how FDI impacts trade and consider the effect of the FTA.5 Data Description This research utilized annual panel data for bilateral FDI and trade found on the databases of the Organization for Economic Co-operation and Development (OECD), but the sample periods of these databases span are limited (OECD International Direct Investment database). Data on Korea’s bilateral trade are available

may–june 2014  15

only after 1994 on OECD International Direct Investment database, and the 2011 bilateral FDI data from Japan and Korea have not yet been released.6 Therefore, the sample period in this research only encompasses the years from 1994 to 2010. Relative growth of gross domestic product (GDP) per capita and exchange rates are from the World Bank’s World Development Indicators. Real GDP per capita is based on purchasing power parity in constant 2005 international dollars. Exchange rates (or cross-rates) are calculated according to the consumer price index (CPI) measured in base period 2005. Table 2 shows the major economic indicators. It is evident that the economic scale of China’s GDP and trade surpasses that of Japan and Korea. Particularly, China’s real GDP is more than two-folds of Japan and ten-folds of Korea. Empirical Findings Discussion of Empirical Results Estimations of the export model and the import model are presented in Tables 3 and 4 respectively. In each of the tables, the estimations without structural changes in 1998 are in Column 1 and the estimations with structural changes in 1998 are in Column 2. Since fixed-effect models are being used, constant terms are listed in Table 3 and average values of exports or imports are in Table 4. Fixed effects of each country are also reported in Tables 3 and 4, respectively. The estimations of export models in Table 3 are interesting. Regardless of whether the structural change is included, the coefficient of OUTFDICNt – 1 is negative and statistically significant; however, the effect of the interaction term EXCNt – 1 × OUTFDICNt – 1 is positive. The estimations listed under model EXCN (2) illustrate this. A one percentage point growth in outward FDI to China will directly reduce exports from Japan and Korea to China by –0.2153 percentage points after one year. At the same time, the substitution effect of outward FDI will be gradually offset if there is a significant amount of exports that year. For example, if EXCNt is 7 percent, the total effect of an increase of one percentage point in OUTFDICN the next year is a 0.01 percentage point rise in EXCNt + 1 (3.2915*0.07—0.2153). The effect of relative real GDP growth on exports is not statistically significant in either model, suggesting that the growing market size of China has not become the major market share for multinational enterprises in Japan and Korea. The J-curve effects are statistically significant in the absence of structural change. According to model EXCN (1), exports initially go down by 0.0002 percentage points after the depreciation, but are boosted in one year. In addition, our estimations show that structural change is statistically significant at the 10 percent level, and the coefficient of D1998 × OUTFDICNt – 1 is negative and statistically significant. Shown in model EXCN (2), the direct negative effect of outward FDI on exports (0.2153 percentage point decrease in exports to China) will increase by 0.1648 percentage points (0.3801 percentage point decrease in exports to China) after 1998.

16  The Chinese Economy

Table 2 Key Economic Indicators in 2012 China Population (millions) GDP nominal (US$ million)

1,351

Japan 128

Korea 50

4,522,140.2

4,712,402.2

1,078,208.5

14,548,591.4

4,465,445.8

1,474,876.4

3,348

36,942

21,562

10,771

35,006

29,495

Exports (US$ million)

1,362,350.3

743,939.2

571,399.9

Imports (US$ million)

1,101,465.9

662,005.3

470,304.2

Real GDP (US$ million)* GDP nominal per capita (US$) Real GDP per capita (US$)*

Source: World Bank. 2013. World Databank: World Development Indicators. Available at http://data.worldbank.org/data-catalog/world-development-indicators. Notes: *denotes this currency unit is based on international dollars; see the definition on https://datahelpdesk.worldbank.org/knowledgebase/articles/114944-what-is-an-international-dollar.

Estimations of the import models in Table 4 show that the feedback effect of inward FDI on imports is relatively weak in comparison with the effect of outward FDI on exports. Without adding D1998 × INFDICNt – 1 and D1998 × INFDICNt – 2, the coefficients of inward FDI, INFDICNt – 1, and INFDICNt – 2, are statistically insignificant. Even though structural change is concerned, the negative impact of inward FDI on imports will be dominated by the interaction term IMCNt – 1 × INFDICNt – 1 if the imports from China continue to rise. According to model IMCN (2), when inward FDI of China rises by one percentage point a given year, Chinese exports to Japan and Korea will directly decrease around 0.0574 percentage points one year later, but if imports from China this year have a positive effect, the negative effect will drop by a 0.5762*IMCNt percentage point. In other words, the total effect of inward FDI from China on imports from China next year is 0.5762*IMCNt —0.0574. The coefficients of INFDICNt – 1 and IMCNt – 1 × INFDICNt – 1 are both statistically significant at the 10 percent level, but not as significant as OUTFDICNt – 1 and EXCNt – 1 × OUTFDICNt – 1. Furthermore, the coefficients of D1998 × INFDICNt – 1 and D1998 × INFDICNt – 2 are each statistically insignificant and jointly statistically insignificant as well. Our estimation thereby concludes that first, structural change is not an important factor in determining whether inward FDI tends to substitute

may–june 2014  17

imports. Second, the impact of relative real GDP growth on imports is not statistically significant in either import model, which is similar to the export models. Third, cross-rates do not significantly affect imports in the model without structural change, and the coefficient of the lagged value of cross-rates is significant at the 10 percent level, weaker than that of export models at the 5 percent level. These estimations have some policy implications. From the perspective of Japan and Korea, outward FDI to China will substitute for exports to China, but this substitution can be neutralized by an export-stimulating policy. Thus, we predict that the CJK Trilateral FTA will ameliorate the negative impact from China which is crowding out exports from Japan and Korea by absorbing FDI. A currency depreciation policy could be another option. When the market prices of the Japan yen and the Korean won are lower than the renminbi, exports to China will increase. Estimations of the import models suggest that when the Japanese and Korean governments encourage FDI inflows (investments by Chinese corporations in Japan and Korea), imports from China will drop. Besides, a depreciation policy can discourage imports from China. When an increase in outflow FDI to China reduces the exports to China, the Japanese and Korean governments can balance their domestic economies by drawing inward FDI from China and reducing their currency prices. The implications of a depreciation policy explain the growing tension in currency wars among East Asian countries. For example, Japan recently launched a depreciation policy. Aside from that, our analysis matches the trend toward FTAs. In the long term, it is expected that the CJK Trilateral Agreement will generate positive reciprocal benefits for Japan and Korea by China. The empirical result is analogous to what was found by Liu (2008). Therefore, there are still some remedies for the Chinese hollowing-out effects. Robustness of Empirical Models Table 5 presents residual tests that examine whether our empirical estimations are robust. The results from the Jarque-Bera test (1987) state that the residuals from all the empirical models are statistically insignificant against the null hypothesis of normal distribution. The unit root tests on residuals are statistically significant against the null hypothesis of the unit root process. The test methods we employ are the Levin-Lin-Chu test (2002) and the PP-Fisher chi-square test (Choi 2001; Maddala and Wu 1999; Phillips and Perron 1988). This indicates that the trend effect in each model has been eliminated successfully. We further applied the Ljung-Box test (1979) on serial correlation. Although the serial correlation tests show that most of the residuals in the empirical models are statistically insignificant against the null hypothesis of no serial correlation, the residuals of the export models with structural change have statistically significant serial correlation. As a result, the estimation of the export models understates the standard errors of every coefficient. Whether each of the coefficients in the export models is statistically significant cannot be judged by t-statistics.

18  The Chinese Economy

Table 3 Estimations of Export Models EXCN (1)

EXCN (2)

0.11*** (0.03) –0.19* (0.09) –0.01 (0.02) –0.03 (0.06)

0.09** (0.03) –0.22* (0.08) 3.e–03 (0.04) –0.01 (0.05)

CREt

–2.e–04* (1.e–04)

–1.e–04 (1.e–04)

CREt – 1

3.e–04** (1.e–04) 0.18 (0.21)

5.e–04** (2.e–04) 0.03 (0.24)

1.82* (0.78)

3.29** (0.93) –0.16* (0.08)

Constant OUTFDICNt – 1 OUTFDICNt – 2 REGROWTHt – 1

EXCNt – 1 EXCNt – 1 × OUTFDICNt – 1 D1998 × OUTFDICNt – 1 D1998 × OUTFDICNt – 2

–0.02 (0.05)

Fixed effects Japan

–0.8%

1.5%

Korea

0.8%

–1.5%

N

30

30

df

18

16

Adj. R Akaike information criterion

0.99 –8.12

0.99 –8.665

Schwarz criterion Test of a structural change χ2

–7.02

–7.5

2

4.94*

Sources: Authors’ estimation. Notes: ***, **, and * denote levels of significance at 1%, 5%, and 10%, respectively. Parentheses below the estimated coefficients indicate standard errors.

may–june 2014  19

Table 4 Estimations of Import Models IMCN (1)

IMCN (2)

Constant

0.1** (0.04)

0.24** (0.08)

INFDICNt – 1

–0.02 (0.02)

–0.06* (0.03)

INFDICNt – 2

3.30e–05 (0.01)

0.02 (0.01)

REGROWTHt – 1

0.11 (0.13)

0.06 (0.12)

CREt

–2.e–04 (4.e–04)

7.e–04 (6.e–04)

CREt – 1

–4.e–04 (3.e–04)

–9.e–04* (4.e–04)

IMCNt – 1

0.62* (0.3)

–0.53 (0.68)

IMCNt – 1 × INFDICNt – 1

0.22 (0.18)

0.58* (0.27)

D1998 × INFDICNt – 1

–0.06 (0.04)

D1998 × INFDICNt – 2

–0.04 (0.02)

Fixed effects Japan Korea

–2.7% 2.7%

2.2% –2.2%

N df

30 18

30 16

Adj. R 2

0.96

0.97

Akaike information criterion

–6.39

–6.77

Schwarz criterion

–5.31

–5.6

Test of a structural change χ2

3.35

Sources: Authors’ estimation. Notes: ** and * denote levels of significance at 5% and 10%, respectively. Parentheses below the estimated coefficients indicate standard errors.

20  The Chinese Economy

Table 5 Robustness of Estimations Residual test Normality test Jarque-Bera Unit root tests Levin, Lin, & Chu PP-Fisher chi-square Serial correlation tests Q-statistic(1 lag) Q-statistic(2 lags) Q-statistic(3 lags)

EXCN(1)

1.59 –1.93** 22.03*** 2.03 2.04 2.04

EXCN(2)

0.17 –3.17*** 33.02*** 8.19*** 8.56** 10.06**

IMCN(1)

1.6e–03 –2.38*** 16.41*** 0.04 0.41 2.38

IMCN(2)

0.03 –2.95*** 24.06*** 0.09 0.97 1.24

Note: *** and ** denote levels of significance at 1% and 5%, respectively.

Conclusion China, Japan, and Korea have signed a trilateral free-trade agreement aimed at boosting economies in Asia. The agreement has not only strengthened the trade landscape of Asia, but will increase the competition between China and the United States. Nevertheless, the Chinese economy has penetrated Japan and Korea so profoundly that their dependency on China could rise significantly after the establishment of the trilateral free-trade agreement. The dependency is due to growing concern over the hollowing-out effects from China, since a majority of investments have been shifting to China and trade among China, Japan, and Korea is gradually falling. This article studies the feedback effects of foreign direct investment on international trade and on trade among China, Korea, and Japan during the years from 1994 to 2010. We adopted bilateral flows of FDI and trade rather than the total amount of flows in a country. Our empirical findings are summarized as follows. First, outward FDI from Japan and Korea to China will substitute exports of Japan and Korea to China, but the substitution effect will be gradually dominated by the positive effect of the interaction of exports and outward FDI as exports to China rise. As a result, the CJK Trilateral FTA will neutralize the substitution effect. Moreover, structural change is only statistically significant in the export model, and the negative impact of outward FDI on exports is larger after 1998. Second, the feedback effect of inward FDI on imports is not as strong as that of outward FDI on exports. Likewise, the negative impact of inward FDI on imports will be dominated by the positive effect of the interaction of imports and inward FDI as imports from China go up. Finally, there are still some remedies for hollowing-out effects from China, such as a depreciation policy, a FTA, and policies for attract-

may–june 2014  21

ing inward FDI from China. If an increase in outflows of FDI to China lowers the exports to China, the Japanese and Korean governments can balance their domestic economies by drawing inward FDI from China and reducing their currency prices. Nonetheless, the CJK Trilateral Agreement may generate a long-term benefit for China, Japan, and Korea. After the implementation of the free-trade agreement, China may not expand its economic growth at the expense of reducing trade with its Asian neighbors. Notes 1. The Republic of Korea (South Korea) is hereafter referred to as Korea in this article. The free-trade agreement (FTA) between ASEAN and China was first proposed after the Asian financial crisis in 1998. It was expected that closer cooperation with China would help improve the regional economy (Zhang, Zhang, and Fung 2007). 2. There has been a strong linkage of trade between China and the ASEAN–5 comprising Indonesia, Malaysia, Philippines, Singapore, and Thailand (Zhang, Zhang and Fung 2007). 3. The estimated contribution of the China-ASEAN FTA is the 0.9 percent growth of GDP among ASEAN countries and 0.9 percent growth for China (Zhang, Zhang, and Fung 2007). 4. Fan (2010) elaborated the potential impact of China’s RMB exchange rate on its trade policy. 5. Liu (2008) used pooled data to study how regional trade agreements will influence China’s foreign direct investment inflows. 6. Our data was collected from OECD International Direct Investment database in late 2012. We found that bilateral FDI data between China and Japan, China and Korea was available for years 1994–2011, but bilateral FDI data between Japan and Korea was only accessible for years 1994–2010. Since regression analysis must be based on identical periods of time for each of the variables, we had to perform this empirical study with the time span 1994–2010. The latest OECD FDI data has updated to 2013. Bilateral FDI data during 2011–2013 among China, Japan, and Korea is now available online at http://stats .oecd.org/index.aspx#.

References Aizenmana, J., and I. Noy. 2006. “FDI and Trade—Two-Way Linkages?” Quarterly Review of Economics and Finance 46: 317–37. Blonigen, B.A. 2001. “In Search of Substitution Between Foreign Production and Exports.” Journal of International Economics 53, no. 1: 81–104. Büthe, T., and H. Milner. 2008. “The Politics of Foreign Direct Investment into Developing Countries: Increasing FDI Through International Trade Agreements?” American Journal of Political Science 52, no. 4: 741–62. Choi, I. 2001. “Unit Root Tests for Panel Data.” Journal of International Money and Finance 20: 249–72. Davies, K. 2012. “Inward Foreign Direct Investment in China and Its Policy Context.” China: An International Journal 10, no. 1: 62–74. Fan, M. 2010. “Quantifying the Impact of Trade Liberalization.” The Chinese Economy 43, no. 3: 68–76. Fifth Trilateral Summit Meeting among the People’s Republic of China, the Republic of Korea, and Japan. 2012. “Making Sense of the CJK FTA.” May 13, 2012. Beijing, China.

22  The Chinese Economy

Fukao, K.; H. Ishido; and K. Ito. 2003. “Vertical Intra-Industry Trade and Foreign Direct Investment in East Asia.” Journal of the Japanese and International Economies 17: 468–506. Gaulier, G.; F. Lemoine; and D. Ünal-Kesenci. 2005. “China’s Integration in East Asia: Production Sharing, FDI & High-Tech Trade.” CEPII Working Paper no. 2005-09. Available at http://cepii.com/PDF_PUB/wp/2005/wp2005-09.pdf. Guerin, S.S., and S. Manzocchi. 2009. “Political Regime and FDI from Advanced to Emerging Countries.” Review of World Economics 145, no. 1: 75–91. Helpman, E. 1984. “The Factor Content of Foreign Trade.” Economic Journal 94: 84–94. Helpman, E., and P. Krugman. 1985. Market Structure and Foreign Trade. Cambridge: MIT Press. Horstmann, I., and J. Markusen. 1992. “Endogenous Market Structures in International Trade.” Journal of International Economics 32: 109–29. International Monetary Fund. 2012. World Economic Outlook: Growth Resuming, Dangers Remain. April. Washington, DC: IMF. Available at www.imf.org/external/pubs/ft/ weo/2012/01/pdf/text.pdf. Jarque, C.M., and A.K. Bera. 1987. “A Test for Normality of Observations and Regression Residuals.” International Statistical Review 55, no. 2: 163–72. Krugman, P., and M. Obstfeld 2008. International Economics: Theory & Policy, 8th ed. Boston: Pearson Education. Levin, A.; C.F. Lin; and C.S.J. Chu. 2002. “Unit Root Tests in Panel Data: Asymptotic and Finite-Sample Properties.” Journal of Econometrics 108: 1–24. Liu, T. 2008. “Impact of Regional Trade Agreements on Chinese Foreign Direct Investment.” The Chinese Economy 41, no. 5: 68–102. Ljung, G., and G. Box. 1979. “On a Measure of Lack of Fit in Time Series Models.” Biometrika 65: 297–303. Maddala, G.S., and S. Wu. 1999. “A Comparative Study of Unit Root Tests with Panel Data and a New Simple Test.” Oxford Bulletin of Economics and Statistics 61: 631–52. Nanto, D.K., and E. Chanlett-Avery. 2006. “The Rise of China and Its Effect on Taiwan, Japan, and South Korea: U.S. Policy Choices.” Congressional Research Service Report RL32882. Organization for Economic Co-operation and Development (OECD). 2012. Investment Statistics and Analysis: International Direct Investment Database. Available at www .oecd.org/daf/inv/investmentstatisticsandanalysis/internationaldirectinvestmentstatisticsoecd.htm. Phillips, P.C.B., and P. Perron. 1988. “Testing for Unit Roots in Time Series Regression.” Biometrika 75: 335–46. Swenson, D.L. 2004. “Foreign Investment and Mediation of Trade Flows.” Review of International Economics 12, no. 4: 609–29. Vavilov, S. 2008. “Trade & FDI in Petroleum Exporting Countries: Complement or Substitute.” Université Paris 1 Panthéon-Sorbonne working paper. Wignaraja, G. 2011. “ASEAN+3/+6 or TPP? Pathways Towards East Asian FTA Consolidation.” Asian Development Bank, Tokyo. World Bank. 2013. World Databank: World Development Indicators. Available at http:// data.worldbank.org/data-catalog/world-development-indicators. Xing, Y. 2007. “Foreign Direct Investment and China’s Bilateral Intra-Industry Trade with Japan and the US.” Journal of Asian Economics 18: 685–700. Zhang, J.; A. Witteloostuijn; and C. Zhou. 2005. “Chinese Bilateral Intra-Industry Trade: A Panel Data Study for 50 Countries in the 1992–2001.” Review of World Economics 141, no. 3: 510–40. Zhang, Y.; G. Zhang; and H. Fung. 2007. “The Prospects for China’s Free Trade Agreements.” The Chinese Economy 40, no. 2: 5–35.