Journal of Policy Modeling 29 (2007) 165–180

Impacts of regional economic integration on industrial relocation through FDI in East Asia夽 Young-Han Kim ∗ Economics Department, Sungkyunkwan University, Myungnyun-dong 53, Seoul, 110-745, South Korea Received 1 July 2005; received in revised form 1 February 2006; accepted 1 April 2006

Abstract This paper examines the impacts of regional economic integration on the industrial relocation of participating countries focusing on the role of foreign direct investment. Focusing on the integrating countries’ asymmetries in technology and market size, this paper demonstrates that preferential trade agreement increases intra-bloc vertical FDI flows when the integrating countries show large differences in factor costs. Moreover, when the technology gap is relatively large between the integrating countries, inter-bloc horizontal FDI tends to inflow to a county with a higher technology level even though its factor cost is higher. These results imply that Korea–China FTA might increase the inter-bloc horizontal FDI inflows into Korea when Korea has significant technological advantage while the intra-bloc vertical FDI inflows into China might be increased with increased pressure on the Korean economy to specialize in the headquarter service sectors. © 2006 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved. JEL classification: F12; F13; F15 Keywords: Preferential trade agreement; FDI; Industrial relocation

1. Introduction The major driving force of the recent upsurge of regional economic integration in East Asia is the market access motivation. Especially, small open economies such as Korea have jumped 夽 Many invaluable comments of Horst Raff and participants of the Annual Conference of European Trade Study Group at Dublin, and anonymous referees are deeply appreciated. This work was supported by Korea Research Foundation Grant (KRF-2004-041-B00135). ∗ Tel.: +82 2 760 0615; fax: +82 2 744 5717. E-mail address: [email protected].

0161-8938/$ – see front matter © 2006 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jpolmod.2006.04.008

166

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

into the Asia-wide efforts for the formation of preferential trade agreements mainly from the concerns about being left out from the preferential market access chances. In addition to the market access motivation, the expectation for the increased foreign capital inflows in terms of foreign direct investment has worked as a second driving force for preferential trade arrangements. The backgrounds for this FDI inducing motivation for preferential trade agreement is that FDI inflows will enhance national industrial efficiency and productivity in addition to the growth of domestic production. However, these Asia-wide efforts for the formation of preferential trade agreements, which were mainly pushed forward by the concerns about alienation from the preferential market access chances, tends to produce some unintended features. The first feature observed is that FDI inflows into the PTA member countries are very asymmetric in terms of absolute volume and the increase rate. For example, after the formation of NAFTA, there has been sharp increases in the FDI inflows into Mexico, while the FDI inflows into Canada has shown quite stagnant trends comparing to the case of Mexico. That is, since 1994 when NAFTA was formed, the share of the US FDI outflows to Mexico was increased to 3.8% in 2003 from 2.8% in 1994. However, the share of US FDI outflows to Canada was reduced to 10.8% in 2003 from 18.4% in 1988 since Canada–US FTA (CUFTA) was formed in 1988. That is, from the perspective of Canada, the formation of regional integration caused negative impacts on FDI inflows from member country, the US, in terms of the ratio in the total US FDI outflows while it worked positively to Mexico. Moreover, the recent sharp decrease in the international transaction costs in terms of communication costs and transportation costs in addition to the removal institutional transaction costs such as tariff and non-tariff trade barriers has caused an upsurge of fragmentation, especially within the member countries of preferential trade agreements. Especially in the Asia, the recent efforts for FTA formation are strongly influenced by the expectation on the FDI inflows as a result of the preferential trade agreement formation, and there have been mixed expectations on the impacts of preferential trade agreement on FDI structure in the East Asian region. For example, Korea–Japan FTA is expected to produce negative impacts to Korea in terms of trade balances and industrial production, i.e., negative static impacts on Korean economies.1 Notwithstanding with these short-term negative impacts, large-scale inter-bloc FDI inflows from Japan into Korea due to FTA was expected to enhance the total factor productivity in the long-term in addition to the dynamic capital accumulation effects by the supporters of Korea–Japan FTA. However, a strong argument against Korea–Japan from the perspective of Korea goes that there will be little positive impacts on inter-bloc FDI inflows into Korea considering the little production factor cost differences between two countries. Moreover, fragmentation between two countries will be deepened pushing Korean industry to specialize in the less-value adding production process while the technology intensive headquarter service sectors are supposed to move to Japan. Another example is the case of Korea–China FTA, which is expected to accelerate the hollowing out of Korean industries to China. The major argument is that Korea–China FTA will increase the hollowing out also in the technology intensive sector through the increased horizontal FDI. A counter-argument supporting Korea–China FTA emphasizes that the increased intra-bloc vertical FDI inflows from Korea to China after the bilateral FTA will enhance the specialization of Korean

1 Refer Yamazawa and Ippei (2001), McKibbin et al. (2004), and Lee & Park (2005) for a detailed discussion on the impacts of Korea-Japan FTA.

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

167

Fig. 1. The impacts of technological asymmetry on inter-bloc FDI.

industries to the headquarter services in technology intensive, while the inter-bloc FDI inflows into Korea via horizontal FDI. The above arguments are a few examples of unresolved issues on the impacts of preferential trade arrangements on the inter- and intra-bloc FDI flows through different types of FDI modes such as horizontal FDI and vertical FDI. Trying to clarify the abovementioned unresolved issues, this paper targets to develop a model to examine the influences of preferential arrangements on the intra- and inter-bloc FDI flows considering the asymmetry in technology and market sizes (Fig. 1). This paper examines the impacts of regional economic integration on the industrial structure of participating countries focusing on the role of foreign direct investment. Traditionally, it has been believed that the FDI inflows into the preferential trade agreement (PTA) participating countries have been one of major driving force for PTA formation. This paper finds that the transaction cost reduction due to PTA creates more industrial fragmentation within the PTA participating countries while non-member countries’ FDI inflows tend to agglomerate into a country with lower production factor costs. Based on the above observations, this paper demonstrate that when PTA participating countries show technical asymmetry, a participating country with

Fig. 2. Impacts of factor cost asymmetry on the choice of inter-bloc FDI mode.

168

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

Fig. 3. Impacts of technology and factor cost on the HFDI country selection.

an intermediate level of technology gets mixed impacts on FDI inflows from PTA participation. Moreover, this paper demonstrates that preferential trade agreement increases intra-bloc vertical FDI flows when the integrating countries show large differences in factor costs (Fig. 2). When the technology gap is relatively large between the integrating countries, inter-bloc horizontal FDI tends to inflow to a county with a higher technology level even though its factor cost is higher (Fig. 3). These results imply that Korea–China FTA might increase the inter-bloc horizontal FDI inflows into Korea when Korea has significant technological advantage while the intra-bloc vertical FDI inflows into China might be increased with increased pressure on the Korean economy to specialize in the headquarter service sectors. 2. The model We assume that there are three countries, h, m and l, with asymmetric technologies and asymmetric market sizes. Country l has the lowest marginal production cost level, while the country h’s marginal cost is highest: cl < cm < ch . However, fixed cost of headquarter services, which mainly involves the R&D process and other technology intensive headquarter services of production, is assumed to be in adverse direction: Fl > Fm > Fh . In terms of market sizes, country h is largest while country m is smallest: al < am < ah .2 Each country has a representative firm and each firm produces products not only for domestic market, but for two foreign markets. Each firm can enter into foreign markets by exporting or through foreign direct investment while competing in Cournot fashion. While diverse foreign market entry modes are available, the consumer utility preferences in each market are assumed to be symmetric with the inverse demand function in country m given as: Pm = am − bm (xm + xhm + xlm ). The structure of game is as follows: the government of each country decides its trade policy including its tariff rate and arranging preferential trade agreement in the form of free trade

2 The assumptions of the model try to describe the East Asian economies such as Korea having the intermediate level of technology compared to Japan and China while the domestic market size is smallest.

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

169

agreement. In next stage, each representative firm observes the government policy decision and makes its output decision competing in each market with two other countries’ firms in Cournot fashion. The equilibrium structure when countries m and l form FTA leaving out country h as an outsider is examined first. When the firm in country h enters the markets of countries m and l, which are integrated via FTA, through exporting, the profit functions of firm m, l, and h are defined as follows: FTA m (EXP) = (pm − cm )xm + (pl − cm )xml + (ph − cm − th )xmh − Fm FTA (1) (EXP) = (pm − cl )xlm + (pl − cl )xl + (ph − cl − th )xlh − Fl l FTA h (EXP) = (pm − ch − tm )xhm + (pl − ch − tl )xhl + (ph − ch )xh − Fh where ti is the import tariff imposed by country i. There are two different paths through which preferential trade agreements influences FDI structures: (i) the structure of FDI inflows from non-member countries into the member countries, and (ii) the structure of FDI inflows between the member countries of preferential trade agreements. We examine the first case as a benchmark discussion. When firm h, which is a firm in the non-member country, tries to enter the integrated market through FDI, it decides its FDI destination between countries m and l based on the comparison of expected profits from investing in each candidate market. Moreover, firm h has to decide the specific mode of FDI in addition. The equilibrium foreign market entry mode under FTA can be determined through the analysis of equilibrium dominance among five regimes of foreign market entry. Among four types of FDI, we confine our discussion to twopresentative types of FDI, i.e., Vertical FDI and Horizontal FDI because Partial horizontal FDI is just a mixture of two types and relocating FDI is not that commonly observed mode.3 3. The impacts of FTA formation on the inter-bloc FDI inflows When countries l and m form FTA and firm h from a non-member country enters the integrated market through FDI, firm h has to decide in which country to invest and what types of FDI it has to choose. There are four types of possible equilibria, (i) firm h invests in country l through horizontal FDI (HFDIhl ), (ii) firm h invests in country l through vertical FDI (VFDIhl ), (iii) firm h invests in country m through horizontal FDI (HFDIhm ), (iv) firm h invests in country m through vertical FDI (VFDIhm ). The equilibrium FDI mode can be found through equilibrium dominance analysis. First, we check the case of HFDIhl . The equilibrium in case of HFDIhl can be determined by backward induction. When firm h invests in country l through horizontal FDI mode under FTA regime, the profit function of firm h, l, and m are defined respectively as

3 Partial horizontal FDI is a mixture of vertical and horizontal FDIs. It has the property of vertical FDI in the sense that foreign production plants operates based on the headquarter services provided from home country headquarter. At the same time, it has the property of horizontal FDI in the sense that it produces the same final products simultaneously both in home and foreign country. Full relocating FDI is not that often observed in large scale multinational corporations, but mainly observed in the small scale firms in non-technology intensive sectors such as small scale Korean firms in labor intensive industries moving to China. Therefore, we focus our discussion on two representative FDI modes: vertical and horizontal FDIs.

170

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

follows4 : FTA h

FTA l

FTA m

(HFDIhl ) = (ph − ch )xh + (pm − cl )xhm + (pl − cl )xhl − Fh − Fl (HFDIhl ) = (pl − cl )xl + (pm − cl )xlm + (ph − cl − th )xlh − Fl (HFDIhl )

(2)

= (pm − cm )xm + (pl − cm )xml + (ph − cm − th )xmh − Fm

The social welfare function is defined as summation of surplus of each sector, such as consumer surplus, producer surplus, and government surplus. Then, the social welfare of country h, whose firm has invested in country l, is defined as follows5 :  x∗ +x∗ +x∗ h mh lh SWh (FDIl ) = CSh + PSh + GSh = ph (xh , xlh , xmh ) dx 0

∗ ∗ − p∗h (xh∗ + xlh + xmh )+

FTA h

HFDIhl + th (xlh + xmh )

(3)

Each firm has three types of outputs, a product for domestic market, two types of products for two foreign markets. For example, firm h decides its optimal outputs for the domestic market h, and for the two foreign markets, l and m. The equilibrium in each market is derived from backward induction. The optimal output of each firm in each market is derived from the following first order conditions of profit maximization problem:    h h h ∂ FTA ∂ FTA ∂ FTA h (HFDIl ) h (HFDIl ) h (HFDIl ) = 0, = 0, =0 ∂xh ∂xhl ∂xhm    ∂ m (HFDIhl ) ∂ m (HFDIhl ) ∂ m (HFDIhl ) (4) = 0, = 0, =0 ∂xm ∂xml ∂xmh    ∂ l (HFDIhl ) ∂ l (HFDIhl ) ∂ l (HFDIhl ) = 0, = 0, =0 ∂xl ∂xlm ∂xlh We derive the equilibrium values by backward induction. By solving the profit maximization problem of each firm in each market, we obtain the following equilibrium outputs in market h for example: ah − 3ch + cl + cm + 2th , 4 ah + ch − 3cl + cm − 2th FTA xmh (HFDIhl ) = 4

xhFTA (HFDIhl )=

FTA xlh (HFDIhl ) =

ah + ch − 3cl + cm − 2th , 4

FTA (FDIhl ): the profit of firm h when firm h FTA h invests in country l after the FTA between country l and m is formed; h (EXP): the profit of firm h when firm h enters FTA the integrated markets by exporting after the FTA between country l and m is formed; h (FDIhm ): the profit of firm MFN h when firm h invests in country m after the FTA between country l and m is formed; h (FDIhl ): the profit of firm h MFN 4

The definition of profit function in each case is defined as follows:

when firm h invests in country l under the most-favored nation (MFN) regime, i.e., without any FTA; h (EXP): the profit of firm h when firm h enters the foreign markets by exporting under the most-favored nation (MFN) regime, i.e., MFN (FDIhm ): the profit of firm h when firm h invests in country m under the most-favored nation without any FTA; h (MFN) regime, i.e., without any FTA. 5 It is assumed that the profits of FDI affiliates are repatriated to the home country of each FDI.

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

171

Then, substituting these equilibrium outputs into social welfare functions, and solving for country h’s social welfare maximization problem, the following optimal tariff rate is derived: ∗ thFTA (HFDIhl ) = (3ah − ch − cl − cm )/10. Then, substituting the optimal tariff rate into the output functions of each firm in country h, we obtain the final equilibrium value of each firm’s outputs as follows: 2ah − 4ch + cl + cm , 5 ah + 3ch + 3cl − 7cm FTA xmh (HFDIhl ) = 10

xhFTA (HFDIhl ) =

FTA (HFDIhl ) = xlh

ah + 3ch − 7cl + 3cm , 10 (5)

In the same way, we can obtain the equilibrium outputs of each firm in each market. Then, based on these equilibrium values, we can obtain the equilibrium profits of each firm as follows: FTA

(al −2cl +cm )2 +(am − 2cl +cm )2 (2ah +cl + cm − 4ch )2 + − F h − Fl 16 25 FTA (al − 2cl + cm )2 + (am − 2cl + cm )2 (ah − 7cl + 3cm + 3ch )2 − Fl + (HFDIhl ) = l 100 16 FTA (al +2cl − 3cm )2 +(am + 2cl −3cm )2 (ah + 3cl − 7cm + 3ch )2 h + − Fm m (HFDIl ) = 16 100 (6) h

(HFDIhl ) =

Firm h’s strategy to invest in country l through horizontal FDI under FTA between countries l and m is an equilibrium strategy when the following conditions are held: FTA

 h (HFDIhl ) ≥ FTA h (VFDIl ): condition for firm h has no incentive to deviate from horizontal FDI to vertical FDI in country l.  h ) ≥ FTA (HFDIh ): condition for firm h has no incentive to deviate from hor(ii) FTA (HFDI m h h l izontal FDI in country l to horizontal FDI in country m.  h ) ≥ FTA (VFDIh ): condition for firm h has no incentive to deviate from hor(iii) FTA (HFDI m h h l izontal FDI in h to vertical FDI in country m. (i)

h

The above conditions can be checked after deriving the equilibrium profits in each mode of FDI. Now we check the case when firm h invests in country l by vertical FDI under FTA. Then, the profit function of firm h is defined as follows: FTA h h (VFDIl ) = (ph − cl − th − ml )xh + (pm − cl − ml )xhm + (pl − cl − ml )xhl − Fh FTA (7) (VFDIhl ) = (pl − cl )xl + (pm − cl )xlm + (ph − cl − th )xlh − Fl l FTA h m (VFDIl ) = (pm − cm )xm + (pl − cm )xml + (ph − cm − th )xmh − Fm Just as the earlier case of horizontal FDI, the equilibrium profit is derived by backward induction as follows: FTA 1 2 (VFDIhl ) = − 100am cl + 600cl2 + 50am cm − 680cl cm (64ah2 + 25al2 + 25am h 400 2 + 194cm − 400Fh − 150am ml + 1880cl ml − 1068cm ml + 1474m2l

− 50al (2cl − cm + 3ml ) − 64ah (5cl − 3cm + 8ml ))

172

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

Now we examine the condition under which firm h chooses horizontal FDI in country l rather than vertical FDI in country l checking the following condition: FTA FTA (HFDIhl ) ≥ (VFDIhl ) h

h

When we substitute the equilibrium values into the above condition, we obtain the following results: Firm h will choose to invest horizontally in country l under FTA if the following condition holds: (al − 2cl + cm )2 + (am − 2cl + cm )2 (2ah + cl + cm − 4ch )2 + − F h − Fl 16 25 1 2 2 − − 100am cl + 600cl2 + 50am cm − 680cl cm + 194cm (64ah2 + 25al2 + 25am 400 − 400Fh − 150am ml + 800cl ml − 1068cm ml + 1474m2l − 50al (2cl − cm + 3ml ) − 64ah (5cl − 3cm + 8ml )) > 0 The above result shows that when country l’s fixed cost is higher, i.e., when the technology level of country l is lower, firm h chooses vertical FDI rather than horizontal FDI. Now, we check the impact of marginal cost asymmetry and market size asymmetry on the firm’s choice of FDI mode. The above result shows that when firm h, which is a firm from non-member country of FTA, chooses its FDI entry mode into country l, it is mainly influenced by the fixed costs of country, that is the technology level of country, while it is less sensitive to the marginal cost difference between m and l. In most cases, it is found that firm h prefers to choose vertical FDI to country l rather than horizontal FDI with the parameter values that satisfy the model assumptions. The intuition behind this result is that with the formation of FTA, when a firm from an outsider country invests into a country with a lower technology, it will mainly choose VFDI, i.e., mainly in a production plant of less-value adding sectors. Now, we check, when firm h chooses HFDI to integrated regions, which country it will choose to invest in between country l and m. Firm h will choose to horizontally invest in country l rather than in country l only when the following condition holds6 : FTA FTA (HFDIhl ) ≥ (HFDIhm ) (8) h

h

FTA h

FTA h

(HFDIhl ) =

(HFDIhm ) =

(2ah + cl + cm − 4ch )2 (al − 2cl + cm )2 + (am − 2cl + cm )2 + 16 25 − Fh − Fl

(2ah + cl + cm − 4ch )2 (al + cl − 2cm )2 + (am + cl − 2cm )2 + 16 25 − Fh − Fm FTA

The profit of firm h when it horizontally invests in country m ( h (HFDIhm )) can be derived by backward induction in the same way as the case of HFDI into country l, and therefore, the details of derivation are skipped. 6

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

173

Therefore, the condition for firm h to horizontally invest into country l rather than country m can be expressed as follows: FTA h

(HFDIhl ) −

FTA h

(HFDIhm ) =

3 c(al + am − 2cm + c) − F ≥ 0 8

(9)

where c = cm − cl and F = Fl − Fm . The above result shows that when the technology gap between countries l and m is lower, and the marginal cost difference between countries m and l is larger, firm h from non-member country is more likely to choose to horizontally invest in country l. However, when the technology difference is larger, firm h will horizontally invest into country m with a higher technology level even though the marginal cost is higher in country m compared to country l. Now, finally, we check the case which country h will choose to invest when it vertically into the integrated market between country: FTA h

(VFDIhl ) =

1 2 (64ah2 + 25al2 + 25am − 100am cl + 600cl2 + 50am cm − 680cl cm 400 2 + 194cm − 400Fh − 150am ml + 1880cl ml − 1068cm ml + 1474m2l

− 50al (2cl − cm + 3ml ) − 64ah (5cl − 3cm + 8ml ))  h We have to check the equilibrium value of FTA h (VFDIm ) in the same as earlier case, and following the backward induction, we can obtain the following equilibrium value: FTA h

(VFDIhm ) =

1 2 (64ah2 + 25al2 + 25am + 50am cl + 194cl2 − 100am cm − 680cl cm 400 2 + 600cm − 400Fh + 64ah (3cl − 5cm − 8nm ) + 50al (cl − 2cm − 3nm )

− 150am nm − 1068cl nm + 1880cm nm + 1474n2m ) Firm h will vertically invest in country m only when the following condition hold: FTA FTA (VFDIhm ) − (VFDIhl ) ≥ 0 h

h

(10)

1 (256ah + 75(al + am ) − 203(cm + cl ) − 737(ml + mm ))(cl − cm + ml − mm ) ≥ 0 200 When we assume the management cost of operating the foreign plant is same between country m and l (ml = mm ), the above condition is simplified as follows: ⇒

1 (256ah + 75(al + am ) − 203(cm + cl ) − 1474m)(cl − cm ) ≥ 0 200 Therefore, it is found that when firm h will vertically invest into country m when the following condition holds: ⇒

203(cm + cl ) + 1474m > 256ah + 75(al + am )

(11)

What the above result shows is that when the management cost of running a foreign production plant is larger than a critical value, firm h chooses to vertically invest in country m. In other words, when the market sizes of trading countries are relatively large compared to the management cost of running the foreign plant, firm h chooses to vertically invest in country l instead. With the

174

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

Fig. 4. The impact of lower coordination costs on inter-bloc VFDI flows.

Fig. 5. The impacts of technology and coordination cost on intra-bloc FDI mode.

development of information technology, the communication and coordination cost of managing foreign plants is getting lower, leading to firm h to vertically invests in country l with a lower marginal production cost as shown in the following diagram7 (Fig. 4). From the above discussion, we find that when firm h, which is the firm from a non-member country, chooses horizontal FDI, it will invest into country m when the technology difference is large between countries m and l. However, when firm h chooses vertical FDI, it will invest into country l unless the coordination cost is higher than a critical value, which is unobservable in reality with the recent communication technology development. The above findings are summarized as follows (Fig. 5). Proposition 1. Firm h, a firm from non-member countries trying to serve markets integrated by PTA, chooses to vertically invest (VFDI) into a country with a lower production cost. When the technology gap is large with the significant management cost of foreign plants, firm h chooses to horizontally invest (HFDI) into country m even though the production factor cost is higher in country m.

7 The critical value of m, above which, firm h chooses to vertically invest in country m is: m ˆ = (1/1474)(256ah + 75(al + am ) − 203(cm + cl )).

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

175

4. The impact of FTA formation on the intra-bloc FDI structure The formation of FTA influences the FDI structures between the FTA member countries mainly thorough the reduction of transaction costs between the FTA member countries while keeping the trade barrier towards non-member countries. For example, the formation of Korea–China FTA is supposed to boost up massive FDI outflows from Korea to China. Based on a same model specification as Section 3, we examine how the FTA formation will influence firm m’s decision on FDI into country l. Firm m will choose to vertically country l rather than choosing horizontal FDI if  invest into m ) ≥ FTA (HFDIm ). the following condition holds: FTA (VFDI m m l l Firm m’s profit from vertical FDI into country l is defined as follows: FTA (VFDIm l ) = (pm − cm )xm + (pl − cl − ml )xml + (ph − cl − ml − th )xmh − Fm m

Then, the profit functions of firm l and h are defined as follows: FTA (VFDIm l ) = (pl − cl )xl + (pm − cl )xlm + (ph − cl − th )xlh − Fl , l

FTA h

(VFDIm l ) = (ph − ch )xh + (pm − ch − tm )xhm + (pl − ch − tl )xhl − Fh

(12)

The equilibrium profit of each firm is derived by backward induction as earlier case as follows8 : FTA 1 1 (VFDIm (6am + 3ch + 7cl − 16cm )2 − Fm + (7al +3ch − 10cl −21ml )2 l )= m 441 841 1 (ah + 3ch − 4cl − 7ml )2 + 100 Firm m’s profit from horizontal FDI into country l under FTA is as follows9 : FTA 1 1 (HFDIm )∗ = (ah + 3ch − 4cl )2 + (2al + ch − 3cl )2 l m 100 49 1 (6am + 3ch − 7cl − 16cm )2 − Fl − Fm + 441

(13)

Now, we examine which FDI mode firm m will choose when it invests into country l after the formation of FTA with country l. Firm m will choose to vertically invest into country l if the following condition holds: FTA FTA (VFDIm (HFDIm l )− l )>0 m

m

8 When we focus on firm m’s FDI strategy assuming that firm h from non-member country serves the integrated markets by exporting, countries m, l, and h come to decide optimal import tariffs as follows, respectively: {(1/21)(3am − 9ch + 7cl − cm )}, {(1/29)(−al − 17ch + 3(6cl − ml ))}, {(1/10)(3ah − ch + 2cl − ml )}. When the tariff rates are assumed as exogenous variables, firm m’s equilibrium profit from vertical FDI is as follows. ((ah + 3ch − 4cl − 7ml )2 /100) + ((al + ch − 2cl − 3ml + tl )2 /16) + (((6am + 3ch + 7cl − 16cm )(am + ch + cl − 3cm + tm ))/84) − Fm . 9 The optimal tariff of countries m, l, and h is as follows, respectively: {(1/21)(3a − 9c + 7c − c )}, {(1/7)(3a − m h l m l 3ch + 2cl )}, {(1/10)(3ah − ch + 2cl )}. When the tariff rates are assumed as exogenous variables, firm m’s equilibrium profit from vertical FDI is as follows. ((ah + 3ch − 4cl )2 /100) + ((al + ch − 2cl + tl )2 /16) + (((6am + 3ch + 7cl − 16cm )(am + ch + cl − 3cm + tm ))/84) − Fm − Fl .

176

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

FTA

FTA

1 (6am + 3ch + 7cl − 16cm )2 − Fm m m 441 1 1 (ah + 3ch − 4cl − 7ml )2 + (7al + 3ch − 10cl − 21ml )2 + 100 841  1 1 1 − (ah + 3ch − 4cl )2 + (2al + ch − 3cl )2 + (6am + 3ch − 7cl − 16cm )2 100 49 441  1 1 − F l − Fm = − (ah + 3ch − 4cl )2 − (2al + ch − 3cl )2 + Fl 100 49 (VFDIm l )−

(HFDIm l )=

1 1 (ah + 3ch − 4cl − 7ml )2 (7al + 3ch − 10cl − 21ml )2 + 100 841 The comparative static analysis of the above condition shows that after the formation of FTA, intra-bloc FDI structure is mainly characterized by the vertical FDI rather than horizontal FDI throughout the parameter values which satisfy the model assumptions as shown in the following diagrams. Moreover, as the technology level of the country with a lower factor price (country l) is lower, the vertical FDI becomes much more dominant strategy than a horizontal FDI for firm m, which tries to invest in country l after FTA formation. Finally, we check the formation of FTA between countries m and l might cause a hollowing out of country m’s industry. Hollowing out is a drastic feature of horizontal FDI with a total movement of all production process. First, we check the difference between firm m’s equilibrium profit from horizontal FDI before the FTA formation and after FTA formation. If the following condition holds, we might say that the probability for the industrial hollowing out might occur from country m to l: FTA MFN (HFDIm (HFDIm (14) l )− l )>0 +

m

m

The equilibrium profit of firm m when it chooses horizontal FDI into country l is derived as follows by backward induction: FTA 1 1 (ah + 3ch − 4cl )2 + (2al + ch − 3cl )2 (HFDIm l )∗ = m 100 49 1 + (6am + 3ch − 7cl − 16cm )2 − Fl − Fm 441 Firm h’s profit from horizontal FDI under MFN (most-favored nation) regime before the FTA formation is derived as follows10 : MFN m

∗

(HFDIm l ) =

(3ah + 96al + 52ch − 151cl )2 114921 (21ah − 6al + 25ch − 40cl )(81ah − 120al − 65ch + 104cl ) − 76614 1 + (2am + ch + cl − 4cm )2 − Fl − Fm 25

10 The optimal tariff of countries m, l, and h are respectively: {(3a − c − c − c )/10}, {(12a + 45a − 131c + m h l m h l h 74cl )/339}, {(29ah + 24al + 13ch − 66cl )/226}.

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

177

Then, substituting the above equilibrium profits into the condition for hollowing out, we obtain the followings: FTA m

(HFDIm l )−

MFN m

(HFDIm l )=−

(3ah + 96al + 52ch − 151cl )2 114921

1 1 + (ah + 3ch − 4cl )2 + (2al + ch − 3cl )2 49 100 (21ah − 6al + 25ch − 40cl )(81ah − 120al − 65ch + 104cl ) + 76614 1 1 + (6am + 3ch + 7cl − 16cm )2 − (2am + ch + cl − 4cm )2 441 25 When we check the impacts of cost asymmetry and market size asymmetry on the hollowing out effects from the above condition, it is found that Firm h’s HFDI into country l is reduced after the formation of FTA between countries m and l, and therefore, hollowing out is not a serious threat of FTA with a country with a lower factor costs as shown in the following diagram. The findings about the impact of FTA formation on intra-FDI flows are summarized as follows. Proposition 2. After the formation of PTA, firm m will choose VFDI when the factor cost difference is relatively large. FTA formation between two countries are not going to increase the hollowing out due to the fact that tariff reduction decreases the motivation for HFDI. 5. Policy implications and the concluding remarks Based on an oligopoly model of three countries competing in Cournot fashion, this paper examined the impacts of preferential trade arrangements on the industrial relocation between the integrating countries and the inter-bloc FDI flows between asymmetric countries in terms of technology levels and market sizes. Using the equilibrium dominance criteria, we derived the optimal foreign market entry mode considering five possible entry modes such as exporting, vertical FDI, horizontal FDI, partial horizontal FDI and full relocating FDI. The major finding from the examination is that the impacts of FTA formation between countries with asymmetric technology, factor prices, and market sizes on FDI flows and the resulting industrial relocation differs between the inter- and the intra-bloc FDI cases (Fig. 6). In case of inter-bloc FDI flows after the formation of FTA, firms from non-member country tend to choose vertical FDI into a country with lower factor costs. However, firms from non-member countries choose horizontal FDI into a country with higher technologies and higher factor costs when technology gap is big between the member countries. In case of the impacts of FTA formation on the intra-bloc FDI flows, firms from a member country, which have relatively higher technologies and factor costs, tend to choose vertical FDI into a member country with lower factor costs. Moreover, after the formation of FTA, intra-bloc horizontal FDI inflows into a member country with a lower factor costs tend to be decreased because the reduction of FTA agreement decreases the motivation for horizontal FDI among the member countries. The above results show that when FTA is formed among technically asymmetric countries, vertical FDI inflows from non-member countries into a country with a lower production factor cost will be increased. However, horizontal FDI inflows from non-member countries might be increased into a country with a higher technology when the technology difference is large. These

178

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

Fig. 6. Impacts of cost and market size asymmetry on the hollowing out effects.

results imply that in case of Korea–Japan FTA, a significant increase in vertical FDI inflows into Korea from non-member countries is unlikely when the factor cost difference between two countries is not significant, as is true in reality. Moreover, the increase in horizontal FDI inflows into Korea from non-member countries is also unlikely due to the lower Korean technological level compared to Japan. When we examine the intra-bloc FDI flows from Japan to Korea after the formation of the bilateral FTA, the probability for the increase of vertical FDI from Japan to Korea is low due to the insignificant marginal cost differences between two countries while the horizontal FDI might be reduced due to the lowered trade barriers. Therefore, Korea–Japan FTA seems to provide little positive impacts on FDI inflows into Korea in terms of both inter- and intra-bloc FDI inflows. This result is supported by the real data showing the decreased FDI inflows from Japan to Korea after the launch of the Korea–Japan Bilateral Investment Treaty (BIT) in 2003, which ensured the national treatment of Japanese firms in Korean territory, and consequently provided similar effects as FTA in reducing the transaction costs between two countries in the fields of direct investments. That is, after the launch of BIT on 1 January 2003, the FDI inflows from Japan to Korea has decreased by 61.5% in 2003 compared to the previous year, 2002. Although the FDI inflows into Korea from Japan had increased to 2258 million US dollars in 2004, still it is lower than the level in 2000 (Table 1). When the result of model analysis is applied to the case of Korea–China FTA, inter-bloc FDI inflows into Korea from non-member countries are expected to be increased when the technology gap between two countries is significantly large. In addition, intra-bloc FDI flows from Korea to China will be increased in terms of vertical FDI, while the horizontal FDI from Korea to China might be reduced when the technology gap between two countries is large. The above results imply that after the formation of Korea–China FTA, the probability for the hollowing out in the technology intensive industries from Korea is very low while the hollowing out in labor intensive sectors of industries might be increased due to the increased vertical FDI. In other words, the increased vertical FDI from Korea to China will work as a driving force for the industrial restructuring of Korean industries specializing in headquarter service sectors. Based on the above results, the following policy implications are derived. First, when a FTA participating country targets to specialize its industries in the technology intensive sectors via its vertical FDI into the member country, it is optimal to form a FTA with a country that has a lower

Investor

2000

2001

2002

2003

2004

2005

Number

Amount

Number

Amount

Number

Amount

Number

Amount

Number

Amount

Number

Japan ROW

615 3658

2451 12770

593 2828

776 10510

474 1968

1404 7689

494 2102

540 5929

552 2555

2258 10528

611 3094

1878 9684

Total

4273

15221

3421

11286

2442

9093

2596

6469

3107

12786

3705

11562

Source: Ministry of Commerce, Industry and Energy, Korea, FDI Statistics Database, 2005.

Amount

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

Table 1 FDI Inflows from Japan into Korea around BIT 2003 (Million US dollars)

179

180

Y.-H. Kim / Journal of Policy Modeling 29 (2007) 165–180

level of technology. Moreover, when the FTA participating country intends to induce horizontal FDI from the non-member countries in the technology intensive sectors, it is required to keep technology leadership within the member countries. Second, from the perspective of Korea, which has an intermediate level of technology and relatively a small domestic market size, the optimal policy to induce horizontal FDI inflows involving technology intensive headquarter services is to arrange bilateral FTA with China or ASEAN which have relatively lower technologies than Korea. The possible hollowing out in labor-intensive industries from Korea, which might be accelerated with the above FTAs, will have positive impacts on Korean industrial restructuring in long terms. This result can be generalized to the countries with intermediate level of technologies. The future works as an extension of this paper include the examination of the impacts of preferential trade agreements on the factor costs after the firms’ relocation decision making. Moreover, by introducing a general equilibrium model reflecting the oligopoly competition features, we could obtain more policy implications on the income effects and the long-term relocation strategy considering the endogenous factor cost variations. Reference Yamazawa, & Ippei. (2001). Assessing a Japan–Korea free trade agreement. Developing Economies, 39(1), 3–48.