達成目標

□ 未達成目標（請說明，以 100 字為限）

□ 實驗失敗

□ 因故實驗中斷

□ 其他原因 說明：

本研究建立一個賽局理論模型，探討異質性廠商進入海外市場之模式的抉擇以及不同進入 模式對投資績效之影響。本研究的主要特色在於理論模型中同時考慮了研發競爭以及產業 內廠商之異質性對進入海外市場之模式的抉擇。實證分析方面，本研究運用本研究將利用 經濟部統計處發布之1999~2005 年的「工廠校正」調查統計資料來進行經濟計量分析，以 驗證本研究之理論架構的預測能力。此外，本研究也進一步探討不同進入模式對廠商進行 海外直接投資後之研發支出以及技術效率的影響。本研究的理論分析以及實證研究內容與 原計畫構想大致相符。

2.

研究成果在學術期刊發表或申請專利等情形：

論文：□已發表 □未發表之文稿 撰寫中 □無 專利：□已獲得 □申請中 □無

技轉：□已技轉 □洽談中 □無 其他：（以 100 字為限）

附錄二

3.

請依學術成就、技術創新、社會影響等方面，評估研究成果之學術或應用價值（簡要敘述 成果所代表之意義、價值、影響或進一步發展之可能性）（以 500 字為限）

在學術成就上，本研究的理論與實證發現，進一步延伸目前國際貿易上有關異質 廠商方面的研究。相較於過去之相關研究，如 Head and Ries (2003), Helpmand et al.

(2004)等, 因為本研究在模型中考慮了研發支出，把廠商的生產技術內生，進一步釐 清了有關生產效率以對外投資間之相互因果關係，同時也進一步釐清有關廠商對外投 資模式之選擇問題。

在應用價值上，我國已是對外投資大國。對外投資對國內經濟的影響一直是產官 學界關注的焦點。本文的實證研究成果可以做為政府擬定對外投資政策之參考。特別 是，本研究發現，廠商對外直接投資對其研發支出有正面效益。此一發現有重要的政 策意涵，亦即，廠商經由對外直接投資將生產活動外移，有益於其生產規模之擴大，

因而提高了廠商從事研發活動之誘因。由此可知，產業外移與產業升級並不必然相互 衝突，產業外移甚至可能因為產業生級而有助於廠商根留台灣。

一、參加會議經過

報告人於 102 年 6 月 25 日中午由桃園機場搭機前往美國。美 國時間 6 月 25 日早上抵達舊金山。美國時間 6 月 25 日中午由舊 金山搭機前往美國科羅拉多州丹佛市，當天下午兩點多抵達丹佛 機場；傍晚時分住進丹佛市區旅館。6 月 26 日前往開會地點認識 週邊環境以及交通動線。

6 月 27 日中午前往開會地點註冊，並參加當天下午的兩場研 討會。其中一場有一篇論文探討的議題與報告人發表的文章十分 接近。該文作者為日本名古屋大學的中國籍教授。後來與該教授 有進一步的聯繫並探討未來的合作研究事宜，也算是一個意外的 收穫。 6 月 28 日早上在其中一場研討會碰到舊識大阪大學的 Hiro Lee 教授主持該場研討會，且其中有一篇論文係由台大經濟系的 黃鴻教授發表。當天中午碰巧也碰到以前的指導學生吳世傑教授。

6 月 29 日早上參加了一場專題演講，由美國 Stanford University 的 Orley Ashenfelter 教授主講，探討美國早其實施 酒禁之效果，相當具有啟發性。6 月 29 日中午參加了西方經濟學 會會長 George G Kaufman 的午餐演說，題目為公共政策與銀行 危機關系。Ashenfelter 教授碰巧就坐在我旁邊，我順便跟他聊 了一下他的演講內容。6 月 29 日下午參加了一場有關自由貿易區 的研討會，其中有大阪大學 Hiro Lee 教授一篇有關亞太地區區域 整合經濟效果之實證分析，與我手上一篇進行中的論文相關，因 此順便跟他要了該文的全文。

我發表的文章安排在 6 月 30 日下午，評論人是銘傳大學的黃

教授(Shu-Chin Huang)，她提出了一些不錯的意見。當天下午我

也參加了另一場有關 Agent-based Computation Modeling 的研討

會，對該研究方法有了進一步的認識。7 月 1 日的研討會只有半

天。我參加的一場研討會由銘傳大學的 Shu-Chin Huang 主持，其

中一篇文章係由淡江大學的陳柏儒教授發表，也是探討對外投資 進入模式的問題<，但她聚焦於財產權保護與對外投資進入模式的 關係。

7 月 2 日早上我由丹佛機場搭機回國。中午抵達舊金山；下 午由舊金山啟程，台灣時間 7 月 2 日晚上回到桃園機場。

二、與會心得

此次會議雖也吸引世界各國許多經濟學家參加，台灣來的學者也 相當多。但相較幾年前在舊金山舉辦的年會而言，此次少了許多歐美 的經濟學者。由此可見，研討會地點的選擇也攸關研討會的成敗。另 外，此次會議的一項特色就是有多場有關運動方面的研討論文。這應 該是由於丹佛市運動風氣鼎盛之故吧。

另外，值得一提的是，丹佛市為了照顧殘障人士，特別規定所有 公車必須設計無障礙設施，讓殘障人士方便進出。此一措施是多年前 丹佛市民激烈抗爭後得來的。不過，此政策在美國其他大城市並不多 見，是否值得其他城市效法值得深思。再者，丹佛市有色人種人口比 例相當高，治安上似乎有些問題。開會第二天我在旅館的房間門口要 開門時皮夾被一黑人青年搶走，還好人平安無恙且金錢的損失不大。

不過，精神上受到相當大衝擊。

三、考察參觀活動(無是項活動者略) 無。

四、建議 無。

五、攜回資料名稱及內容

無。

六、其他

無。

Firm Heterogeneity, R&D Competition, and Foreign Market Entry Strategies

Kun-Ming Chen

Department of International Business National Chengchi University

64, Section 2, Zhi-Nan Road, Taipei 116, Taiwan Tel.: 886-2-29387515. Fax: 886-2-29387699

E-mail: [email protected]

Shu-Fei Yang

Chung-Hua Institution for Economic Research 75, Chang-Hsing Street, Taipei 106, Taiwan

Tel.: 886-2-27356006 ext. 436.

E-mail: [email protected]

Abstract

This paper investigates the relationship between a multinational firm's choice of entry modes of international expansion and its R&D activity. A game-theoretical model is developed to derive testable hypotheses. Our model predicts that a firm belonging to an industry with higher R&D productivity or having a higher capability in R&D activity tends to undertake FDI to “jump the tariff”. In addition, facing relatively high R&D linkage and relatively small difference in costs between cross-border M&A and greenfield FDI, a multinational firm tends to choose greenfield investment rather than cross-border M&A. A firm-level panel dataset on Taiwanese manufacturing firms is then used to test the validity of our theoretical framework. The empirical results are generally consistent with our theoretical prediction.

Keywords: Foreign direct investment, R&D, entry modes JEL: F21, O31, F23

Firm Heterogeneity, R&D Competition, and Foreign Market Entry Strategies

Kun-Ming Chen

Department of International Business National Chengchi University

64, Section 2, Zhi-Nan Road, Taipei 116, Taiwan Tel.: 886-2-29387515. Fax: 886-2-29387699

E-mail: [email protected]

Shu-Fei Yang

Chung-Hua Institution for Economic Research 75, Chang-Hsing Street, Taipei 106, Taiwan

Tel.: 886-2-27356006 ext. 436.

E-mail: [email protected]

1. Introduction

A multinational firm can serve a foreign market mainly via three different channels:

producing at a home plant for exporting, acquiring an existing foreign plant (cross-border mergers and acquisitions), or setting up an entirely new foreign plant (greenfield investment). The latter two modes are classified as foreign direct investment (FDI). Using data from the Bureau of Economic Analysis, Klein and Rosengren (1994) show that more than 60 per cent of overall inward FDI in the USA between 1980 and 1991 were in the form of cross-border mergers and acquisitions (M&A); this share increased particularly in the late 1980s and early 1990s to far more than 80 per cent of the total. Lall (2002), however, notices that firms with lower R&D intensity are more likely to buy technological capabilities abroad by cross-border M&A, whereas those with strong technological advantages are likely to set up greenfield ventures. These facts suggest that the entry mode choice hinges on the R&D capabilities of the firms.

While FDI has received an enormous amount of attention in the literature, most of this literature has dealt exclusively with a single mode of FDI, mainly greenfield investment. Until recently, there does not appear to be much in-depth analysis in the literature of why a firm would choose one alternative or the other. In particular, the interaction between technological innovation and firm entry mode choice has still been a

neglected issue. Buckley and Casson (1998) first provide a general framework for an integrated analysis of the foreign market entry decisions, encompassing the choice between export, licensing, joint venturing, and whole-owned FDI. They don’t, however, consider R&D spillover effects. Petit and Sanna-Randaccio (2000) develop a game-theoretical model to examine the impact of a firm’s mode of foreign entry on the incentive to innovate, as well as the effects of R&D activities and technological spillovers on the firm’s international strategy. Their results indicate that there is a positive relationship between international expansion and R&D investment, but they do not consider the firm’s choice between cross-border M&A and greenfield FDI. Although

there are a few other theoretical papers dealing with mode-choice between cross-border M&A and greenfield FDI, such as Görg (2000), Hijzena et al. (2008), Nocke and Yeaple

(2007), Kim (2009), and Raff et al. (2009), none of these studies focuses on the possible interaction between a firm’s R&D activities and its choice of entry mode.

Yokota and Chen (2012) is the first paper that studies the possible interaction between a firm’s R&D activities and its choice of entry mode. They develop a three-stage game-theoretical model. In the model, they allow for imperfect appropriability (i.e., technological spillovers between firms). It is shown that if there exist relatively high R&D leakages and relatively small difference in cost between M&A and greenfield FDI,

an R&D-intensive foreign firm tends to choose greenfield investment rather than cross-border M&A, while if there exist relatively low R&D leakages, the foreign firm is more likely to choose cross-border M&A rather than greenfield FDI. However, Yokota and Chen (2012) have not examined the possible impact of a firm's capabilities on its choice of entry mode.

The relationship between firm heterogeneity and the entry mode choice has been investigated in a few recent papers, such as Helpman, et al. (2004), Head and Ries (2003), Nocke and Yeaple (2007) and Raff, et al. (2009). However, Helpman, et al. (2004), and Head and Ries (2003) focus on the choice between greenfield FDI and export without considering cross-border M&A. Whereas Nocke and Yeaple (2007) and Raff, et al. (2009) extend their models to include cross-border M&A or joint ventures, they have not considered the interaction between a firm’s R&D activity and its foreign entry mode choice.

In this paper, we extend the analysis in Raff, et al. (2009), and Yokota and Chen (2012). The novelty of this paper is that we investigate the relationship between a firm's capabilities and its entry mode choice with R&D spillover effect. In addition, a firm-level panel dataset on Taiwanese manufacturing firms is used to test the validity of the theoretical prediction of our model.

The rest of this paper proceeds as follows. In the next section, we develop a game-theoretical model to analyze the impact of a firm's capabilities on the entry mode choice. Section 3 presents our empirical models and estimation method. In Section 4, we introduce our dataset and then discuss our empirical results. Brief concluding remarks are given in the final section.

2. A Game-theoretical Analysis 2.1 Theoretical Model

In this section, we develop a simple game-theoretical model to analyze the interaction of a multinational firm’s entry mode decision and its R&D activity. The model combines the salient features of the models used in Raff et al. (2009), and Yokota and Chen (2012). Suppose that there are three firms: one foreign (multinational) firm and two identical local firms. The products of the firms are homogeneous and three firms compete in the local market.

The game proceeds in the following order: The foreign firm makes a take-it-or leave-it offer of M&A at an acquisition priceP , to one of the local firms. We will label ^A the foreign multinational firm as firm 1, and the potential local target as firm 2, leaving the other local firm, firm 3, as an independent producer in all scenarios. If firm 2 accepts

the offer, then only firm 1 survives and firm 2 becomes firm 1's division. If the M&A offer is rejected, then firm 1 chooses to enter the market by either export or greenfield investment. After the decision of entry mode, firm 1 determines its R&D spending level,

I

. After observing firm 1's R&D spending level, all independent producers choose output levels as Cournot competitors.

Suppose that the inverse demand function can be written as

p    Q

, where

p

is market price and Q is aggregate supply, and



is constant and assumed to be large enough so that the price is strictly positive. We assume, for simplicity, that only the foreign multinational firm invests in R&D. We consider the process innovations, which can reduce the marginal cost of the investing firm. The local firm might enjoy the spillover effects from the foreign firm’s R&D investment by a certain proportion when firm 1 adopts cross-border M&A strategy. For simplicity, it is assumed that there are no spillover effects under other scenarios. ¹ We also assume the cost functions as follows:

1

where c is constant, the superscript

E

indicates export,

A

indicates cross-border M&A , and

G

indicates greenfield FDI, respectively; and



is also a constant which is used to

1 In this paper, the major driving force in the determination of entry mode choice is the relative extent of R&D linkages. If we allow smaller spillover effects in the other two entry modes, the basic results will be the same.

indicate the nature of the industry. A higher value of



implies the industry is more technology-intensive so that the potential for R&D activity to reduce manufacturing cost is higher. Besides,



denotes the R&D spillover effects from firm 1 to firm 3 when the multinational firm acquires firm 2. The total cost of R&D investment is assumed to be quadratic, i.e.,

 I

²

/ 2

, with



>0, which is inversely related to the cost effectiveness of the firm’s R&D investment. A firm with a lower value of



indicates that the firm is more capable in engaging in R&D activity. The specification of the cost of R&D investment also implies that there exists decreasing returns to R&D activity.²

In the case of export strategy, suppose the foreign firm must pay an additional trade costs, t , including transport costs or import tariffs, when it exports to the local market.³ To ensure that exporting yields non-negative profits, we assume that

t  (   c ) / 3

. The profit function of firm i under scenario j is denoted by



_i^j

( =1, 2,3; i j  E A G , , )

. In addition, it is assumed that greenfield FDI requires a fixed cost,

F

, including purchasing price of land, building cost of plant, setting-up cost of machinery, and so on. Following Roff et al. (2009), we make the simplifying assumption that cross-border M&A does not involve fixed cost. Hence, one can view

F

as the differential fixed cost of greenfield investment relative to cross-border M&A.

2.2 Export

If the foreign firm chooses export as its entry mode, given its R&D level, the profit-maximizing output, and profit of each firm under this scenario will be:

2 Although Raff et al. (2009) allow the firms to undertake cost-reducing investment in the case of M&A or joint ventures in their model, they exclude the possibility under the case of greenfield FDI.

3 Greenway and Kneeter ( 2007), p. F151.

1 1

The first-order condition yields the foreign firm's optimal R&D level as follows:

₁^*

3 (

₂

3 )

implies the following equilibrium output and profit of each firm under this scenario:

₁^*

2 (

₂

3 )

If the foreign firm chooses greenfield investment as its entry mode, given its R&D level, the profit-maximizing output, and profit of each firm under this scenario will be:

1 1

₁

1

₁ ² The first-order condition yields the foreign firm's optimal R&D level as follows:

₁^*

3 (

₂

)

A necessary condition for an interior solution is

9 2

8

  

. In our model, the differences in

export and greenfield investment is that the former strategy has a higher variable cost and a lower fixed costs. As a result, it is clear the equilibrium output and R&D levels in the case of greenfield investment is a special case of export strategy in which

t  0

and level, the profit-maximizing output, and profit of each firm under this scenario will be:

1 1 3

The first-order condition yields the foreign firm's optimal R&D level as follows:

₁^*

2 ( ) ( 2

₂

)

greenfield investment to export, any M&A offer with a acquisition price greater than firm 2's profit under greenfield investment strategy, i.e., P^A

 

₂^G*, will be accepted. However, if a multinational firm prefers export to greenfield investment, an M&A offer will be accepted only if the acquisition price is greater than its profit under export strategy, i.e.,

* effectiveness of the firm in engaging in R&D, but negatively related to the trade costs. It is also negatively related to R&D spillover effect under cross-border M&A case.

Proof: Differentiating Eq. (6) with respect to

, andt , we have

* 2 1

2 2

2( )(2( 2 ) 9 ) (2( 2 ) 9 ) 0

IA c

   

   

     

  

.



The economic intuition behind the results regarding the relationship between the optimal R&D and



,



, t or



is straightforward. All of these parameters will make R&D activity more profitable. Note that the higher the productivity of R&D activity of in the industry, or the higher cost-effectiveness of a firm in engaging in R&D is, the higher the returns from R&D activity will be. The lower the trade costs are, the. higher the output is. Since R&D spending is a fixed cost, and thus the average cost of R&D will be lower. This is referred to as “cost-spreading effect” (Cohen and Klepper, 1996a, 1996b), or “sales-increasing effect” (Petit and Sanna-Randaccio, 1998) in the literature. Finally, the higher value of



is, the lower the appropriability of a firm's R&D activity will be, due to a free-rider problem, so as to reduce its incentive in undertaking R&D.

Proposition 2: A multinational firm prefers greenfield investment to export if the fixed

cost of greenfield investment is not too large, i.e.,

3 (2 2

₂

3 )

2(8 9 )

t c t

F  

 

  



^.

Proof: A comparison of 

₁^G*and



₁^E*yields the result. 

Proposition 3: A multinational firm has a higher cost-effectiveness or belongs to an

industry with a higher R&D productivity prefers greenfield investment to export.

The economic intuition behind these results is also straightforward. Given the fixed cost of greenfield investment, if a firm belong to an industry with a higher R&D productivity or is more capable in engaging in R&D activity, it could benefit more from investing abroad, because a lower variable cost from FDI could induce more R&D investment and result in higher profits. These results suggest the possible interaction between entry mode choice and a firm's R&D decision.

Proposition 4: A multinational firm prefers greenfield investment to cross-border M&A

if the R&D spillover effect is large and the fixed cost of greenfield investment is not too large; otherwise, a multinational firm prefers cross-border M&A to greenfield investment.

Proof: Differentiating the differences in the profits between cross-border M&A and

greenfield investment with respect to



, we have

Propositions 3 and 4 suggest the firms in technology-intensive industry where R&D activity plays a crucial role in market competition are more likely to "jump the tariffs"

to undertake FDI. In addition, the R&D spillover effect is crucial in the firm's choice between grrenfield investment and cross-border M&A, which is consistent with the conjecture of Zejan (1990) and Brouthers and Brouthers (2000).

3. Empirical Model and Estimation Method

To empirically estimate the determinants of a firm's decision to invest into foreign countries and its entry mode choice, we adopt a bivariate Probit with sample selection

model. We assume that a firm has a two-stage decision process: At the first stage, the

firm decides whether or not to invest abroad, and at the second stage, if the firm decides to invest abroad, then it determines its entry mode.

In the following, selection equation is used to investigate the determinants of a firm's OFDI decision, and entry mode equation is used to investigate the determinants of a OFDI firm's entry mode. The selection equation is specified as follows:

* The entry mode equation is specified as follows:

*

Here, X is the vector of explanatory variables, ₂



₂is the vector of coefficients to be estimated, and u is a random error. The random errors of these two ₂ equations are assumed to be distributed as a binomial normal distribution as follows:

each type of firms is as follows:

1 1 1

Pr(y



0)

  

( X



), (33) P r (y₁



1 ,y₂



0 )

  

X(



_{1 1}

 

) ₂X



(_{1 1}X

 

,₂ , (34) ₂ , ) P r (y₁



1 ,y₂



1 )

 

₂ X



(_{1 1}X

 

,_{2 2}. (35) , )

From the probability equations, we can derive the log likelihood function,

^L

, as follows:

1 2 2 1 1 2 2 1 1 1

In this paper, we are interested in testing the relationship between R&D capability (or spillover effects) and FDI as well as the relationship between R&D capability (or spillover effects) and entry mode choice. In order to control industry characteristics, industry dummies are used as explanatory variables in the selection equation. Besides, the following explanatory variables are used to test the driving forces of FDI:

In this paper, we are interested in testing the relationship between R&D capability (or spillover effects) and FDI as well as the relationship between R&D capability (or spillover effects) and entry mode choice. In order to control industry characteristics, industry dummies are used as explanatory variables in the selection equation. Besides, the following explanatory variables are used to test the driving forces of FDI:

在文檔中 廠商異質性、研發競爭與國外市場進入策略 (頁 65-99)