Entry Mode, Technology Transfer and Management Delegation of FDI. Ho-Chyuan Chen

Transcription

1 ntry Mode, Technology Transer and Management Delegation o FDI Ho-Chyuan Chen Department o conomics, National Chung Cheng University, Taiwan bstract This paper employs a our-stage game to analyze decisions o FDI Multinationals with management delegation. It inds that management delegation leads to higher optimal technology-transer o the MN than no-delegation case. It also leads the direct-entry MN to optimally transer higher technology than an acquisition one, which reverses the result o no-delegation case. Management delegation decreases the likelihood o the MN adopting direct-entry i the costs o technology transer are suiciently high, while it increases the likelihood i otherwise. When the entry cost is suiciently high, acquisition becomes the equilibrium but induces lower technology transer than o-equilibrium mode. s ar as the domestic welare is concerned, when the costs o entry are suiciently low (high), acquisition (direct-entry) mode generally achieves higher domestic welare but is not the equilibrium mode o the MN. In act, management delegation makes this disparity more signiicant. Keywords: ntry mode; technology transer; acquisition, FDI, delegation JL classiication: F5, F3, O30 This is a rough and incomplete drat. Constant revision is under working. Please no citation, no post-up, and no circulation. Comments are welcomed. June 0, 04 Ho-Chyuan Chen, Proessor at Department o conomics, National Chung Cheng University, #68 University Rd., Min-hsiung, Chiayi County 6, Taiwan. Tel: ext.340; Fax: ; -mail: runawaychen@gmail.com

2 Introduction merging markets collectively accounted or $7 billion in 000, but $ billion in 008, in outward international M&, which shows about an average annual increase o 00% over the 8-year period. The M& entry mode in this period accounted or about 65%, 8%, and 7% o FDI (oreign direct investment) lows in the world, the developed economies, and the developing economies, respectively, according to UNCTD statistics (00). Due to the enormous growth in FDI has signiicantly increased the number o multinational enterprises (MNs), the issue o technology transer and entry mode o oreign irms has gained increasing attention. mong the FDI researches, an important variable, i.e., the issue o management delegation o an MN, is let out, which is the ocus o the paper. Direct entry and acquisition entry are two alternative entry modes o the FDI lows (e.g., lango, 005; Williams, 005). cquisition may be preerred when entry costs are very high (l-kaabi et al., 00; Fatica, 00), or when enabling market development is desired in less developed markets (Teixeira and Grande, 0). On the other hand, direct entry may be preerred when direct historical and cultural ties exist between home and host countries (Demirbag et al., 00). In the literature, country size is also shown to be a matter. Large countries are more likely to attract acquisitions, the intermediate-sized countries may be served predominantly through trade, and small countries are most likely to experience either or no entry, as ound in icher and Kang (005). Dierent entry mode may induce dierent level o technology transer o the MN, which makes the technology transer be an important strategic decision or MNs (Kasuga, 003; icher and Kang, 005) and the host regulations. Doubtless, It accounted or about 74%, %, and 8% o FDI lows in U, sia and China, respectively, and China. Hong Kong and other South-ast sian countries were the main beneiciaries o the heightened FDI lows.

3 the MNs seeks the optimal entry mode to prevent the dissipation o their technological advantages (e.g., thier and Markusen, 996; Markusen, 00), while the host government regulates on the MNs entry mode or larger social welare. Mattoo et al (004) showed that technology transer o the MN is higher when acquisition mode is chosen when the number o irms is small. Muller (007) showed that direct entry becomes optimal i the technology transer cost is suiciently high and the local market is with low competition. O course, the ability and eectiveness o adopting new technology or host countries are essential or MNs. For instance, the knowledge gaps between MNs and the developing countries usually result in larger costs o technology transer or M& and makes direct entry more preerable to MNs (Bjorvatn, 00). The host governments, on the other hand, generally have incentive to attract FDI or gaining eect o technology transer (Glass and Saggi, 00) with restrictions imposed on the MNs. Due to China s oreign- ownership restrictions, Taiwan bio-tech irms generally consider acquisition as the irst-priority mode or entering mainland China (Lee et al., 0), which may lower the quality o technology transerred (Lee and Shy, 99). Instead, Kasuga (003) ound that acquisition mode becomes more diicult when the capital market o local countries is ineicient. However, MNs always have inclination to adopt management delegation because they are not amiliar with oreign markets. This becomes one o main dierences in running a business between local and oreign irms, thereby resulting in dierent choice behaviors between them. The industrial organization literature has explored the idea that altering a irm s behavior can alter equilibrium outcomes (Fershtman and Judd, 987; Sklivas, 987; Vickers, 985; Fumas, 99; Miller and Pazgal, 00 and 005). They have shown that owners can beneit rom delegating

4 decisions to managers in oligopolistic markets. The compensation schemes or managers serve as commitment devices used by owners to precommit managers to certain actions in later stage, which in turn alter the actions taken by rival managers. In this paper, we extend the literature on management delegation to a model or FDI decisions. To contrast the dierence between local and oreign irms, we assume the owner o the oreign irm oers a managerial incentive scheme to his managers or delegating decision o market competition, while the domestic irms do not. Through this way, we are able to investigate how management delegation aects the MN s decision on technology transer and entry mode. We thereore set up a our-stage game model: the owner o the oreign irm chooses mode o entry (acquisition or direct entry), sets delegation incentive scheme to his managers, decides the level o technology transer, and inally the managers choose output to Cournot-compete with domestic irms in the domestic market. The results are interesting. We ind that management delegation leads to higher optimal technology-transer o the MN than either entry mode o no-delegation case. It also leads the direct-entry MN to optimally transer higher technology than an acquisition one, which reverses the result o no-delegation case. Management delegation decreases the likelihood o the MN adopting direct-entry i the costs o technology transer are suiciently high, while it increases the likelihood i otherwise. cquisition becomes the equilibrium mode when the entry cost is high, but it induces lower technology transer than o-equilibrium mode. Furthermore, when the entry cost is suiciently low (high), acquisition (direct-entry) mode generally achieves higher domestic welare, but the equilibrium mode o the MN is direct-entry (acquisition). This disparity becomes more signiicant when management delegation is employed by the MN. On the other hand, when the entry cost is somehow 3

5 intermediate, both the MN and the host government like direct-entry mode, with a smaller likelihood than no-delegation case. This paper is organized as ollows. Section discusses the theoretical setup and describes the two entry modes (acquisition and direct entry) or the oreign irm to enter the host country. Section 3 provides a our-stage game model to analyze the extent o technology transer and the optimal entry mode when a oreign irm decides to enter the market o the host country. Section 4 ocuses on the impact o the oreign irm s entry decision on the host country s welare. Section 5 oers our conclusions. The Model There are two domestic irms (i=, ) and one oreign irm (), producing and competing or sales o homogeneous products in the home country. The oreign irm (the MN, hereater) enters the domestic market, either via acquisition () or direct entry (), and transers superior technology to lower the marginal production cost o its subsidiary. Due to unamiliarity with the domestic market, the MN usually delegates the management o production and sale to abroad managers. The our-stage model we construct incorporates the above-mentioned three issues: entry mode, technology transer, and management delegation. In the irst stage, the owner o the MN has two options or entering the domestic market: acquiring a domestic irm or setting up a wholly owned subsidiary that directly competes with domestic irms. ither one has to pay an entry payment. I it chooses to acquire a domestic irm, it makes an oer o a ixed transaction price (T ) to the target irm. To simpliy analysis, we assume that the MN has all the bargaining power to make a take-it-or- leave-it oer. I the target irm accepts the Mattoo et al (004) shows that, in equilibrium, the oreign irm does not choose partial acquisition. Throughout the paper in order to proceed to a concise mathematical analysis, we consider only the case o ull acquisition o the domestic irm. 4

6 oer, they orm a new irm owned by the oreign irm. I the target irm reuses the oer, the oreign irm can enter the market by establishing its own subsidiary or by acquiring some other domestic irm. Thereore, the reservation payo o a target domestic irm to accept an acquisition oer is equal to the proit it makes as a competitor when the other domestic irm is acquired instead. 3 On the other hand, i the oreign enter the market by establishing its own subsidiary, we assume that it pays a ixed entry cost T. Under the acquisition entry mode, the number o irms in the market is smaller and the entry cost o the MN depends on the expected reservation payo o the acquired irm, which is aected by the MN s decisions o delegation scheme and technology transer. These are the key dierence rom the mode o direct entry. ter selecting the mode o entry, the owner o the oreign irm in the second stage sets a delegation incentive scheme (), which gives a wage unction based on both irm s proit and revenue to the managers. The owner then chooses the amount o technology transer to its subsidiary in the third stage. To reduce the marginal production cost by amount x, the oreign must invest C(x) =x / in transerring technology. Parameter represents the process diiculty o technology transer. In the last stage, the manager o the oreign irm chooses quantity to compete with domestic irms in a Cournot-Nash ashion. Let a linear orm o p(q) = a Q be the inverse demand unction, where Q is total output o the market. For notation convenience, we hereater use superscript and to represent the case when the oreign irm enters by acquisition and direct entry, respectively. Thus, the total output Q{Q, Q }, where Q = q + q and Q = q +q + q. Notice that we assume irm to be the one acquired by the oreign irm. 3 This is also used in Mattoo et al (004). 5

7 Since the reservation payo o the acquired irm is equal to the proit it makes as a competitor when the other domestic irm is acquired, the speciied ee (T ) or the MN to buy-out irm must equal. lso, we assume that the ixed entry cost or the MN entering by establishing its own subsidiary T equals t(ac) or the ease o calculation. We denote net proit o the MN owner under the acquisition and direct-entry mode as ~ and ~, respectively, which are net o gross proit and. Moreover, or simplicity, we deine the overall proit margin acq m and net proit margin ac+xq M. Proit unctions are then represented as ollows: ~ Q q T, where ( m x ) q ( x ) ; {, }; q ; Q q q q ; T ; T t( a c). and () On the other hands, proit unctions o the domestic irms are represented as ollows: ;,,. m q i m qi i () Next, to ormalize managerial delegation o the MN, we ollow Lambertini (000) and assume the managerial incentive scheme to be I, and ( a ). q, where c (3) That is, the MN owner assesses the perormance o their managers according to 6

8 observable quantity o the irm s products, in addition to their proits. 4 For the ease o calculation, we deine = (a-c) throughout the paper and term the as the delegation incentive (measure). Beore closing this section, we make the ollowing assumptions to ensure non-negative equilibrium quantity: 5 ssumption ssumption Model analysis 3. Product market By backward induction, the analysis starts rom the last stage (i.e., the product market), in which the domestic irms and the manager o MN simultaneously choose their outputs to maximize their proits. By () and (3), we obtain the irst order conditions as ollows: 6 The MN: I q q ( a c) M q ( a c) 0, {, }. (4a) i The domestic irm : m q 0; m qi 0, i,. q q i (4b) 4 s shown in Fershtman and Judd (987), the real compensation to manger takes the orm W = a + bm U 0, where parameter a and b are chosen by owner and U 0 is the manager s reservation utility, which is normalized at U 0 = 0. Since he market or managers is assumed to be competitive, the participation constraint is binding. ccordingly, the proit o owner does not include the term o compensation W. For the setting that does not assume a competitive market or managers and so managers do not necessarily receive just their reservation wage, see Lambertini, L., Trombetta, M. (00). 5 See equation (9) and (3a). 6 Note that the speciied buy-out ee (v) is paid in the irst stage, and so is sunk when we consider strategy choice in third and second stages. 7

9 By solving (4a) and (4b) simultaneously under each entry mode, we obtain the ourth-stage equilibrium outputs as ollows, in which n denotes the total number o domestic irms ater the entry: q i ( )( a c) x n and q [ ( n ) )]( a c) ( n ) x n Where n = i = and n = i =., (5) It readily shows that delegation incentive scheme expands the MN s output, while it squeezes the domestic irms outputs. ccording to equations (5), we require the ollowing restricts or non-negative outputs o an individual domestic irm: x, or {, }. a c The comparative statics o individual equilibrium output with respect to incentive scheme are in Lemma : Lemma. ll other things being equal, i ( a c), 3 q i ( a c), 4 q q ( a c), and 3 q 3( a c). 4 Management delegation incentive gives the oreign irm itsel a larger output increase than the total output decreases o domestic irms, and it gives larger output increase o direct entry mode than that o acquisition entry mode. That is because the 8

10 larger total number o irms under direct entry mode leads to higher substitution between products o domestic irms and the oreign irm, which gives an aggressive output choice (i.e., higher delegation incentive) a larger impact on equilibrium. In this stage, the oreign irm aces higher domestic market competition when it adopts direct entry. 3. Technology transer In the third stage, given an entry mode and expecting q i and q in (5), the owner o the oreign irm maximizes his proit by choosing the level o technology transer x, which reduces its marginal production cost by x at the cost o C(x) = x /. Taking derivative o equation (3) and (4) with respect to x, we obtain the irst order conditions or technology transer: d dx dq d dqi n q dx qi dx Delegation ect Strategic ect where q C x x. x q Scale ect x C x MCTT 0, (6) Technology transer o the MN has three eects on the owner s proit. The cost reduction due to technology transer gives the managers much incentive to expand output, which is reinorced under the management delegation scheme and accordingly results in more decrease o owner s proit margin. It is represented by the irst term q dq dx and is termed as delegation eect (D). It is expected to be negative. On the other hand, the owner s proit margin could be increasing because that the cost reduction squeezes other irms outputs ( q dq dx ) and directly gives cost 9

11 savings or each unit o product (the q in term x ). The irst one is termed as strategic eect (S) and the second one as scale eect (), and both are expected to be positive. The sum o these three eects constructs the marginal revenue or technology transer (hereater, denoted as MRTT), while the term Cx x in x represents the marginal cost o technology transer (termed as MCTT, hereater). We rewrite equation (6) in equation (7): n MRTT ( a c)( ) n( q )( ) q n n D S ( n ) n( n ) ( n ) ( a c) x, and ( n ) ( n ) ( n ) dc dx x MCTT x. (7) s noted beore, n = or mode and n = or mode. The three eects in equation (6) are aected by the choice o delegation incentive (), which is shown is ppendix and summarized in Lemma. Lemma. The ollowing properties hold: D () 0; D 0; n D 0. n S () 0; S S 0; n n 0 only when is suiciently large. (3) 0; n 0; n 0. MRTT (4) 0; MRTT n MRTT 0; 0. n 0

12 The negative delegation eect is decreasing in delegation incentive scheme, implying that a higher delegation incentive will erode more o the owner s proit through the delegation eect. However this eect is diluted in a market with higher competition in terms o higher number o irms. Basically, management delegation entails aggressive output, which in turn largely squeezes the rivals output and help to increase the strategic eect and the scale eect o technology transer. Obviously, the scale eect will be strengthened in a market with higher competition (i.e., a market when the MN enters with direct entry), while the strategic eect can only be strengthened with higher competition when delegation incentive is large enough. When increases, both the strategic eect and the scale eect increase and sum up to dominate the decrease o delegate eect, which in turn results in the increase o the marginal revenue or technology transer (MRTT). In words, a higher delegation incentive causes higher marginal revenue o technology transer to the MN, and thus leads the MN to decide on a higher level o technology transer. Moreover, the -impact to D, S and turns out to decrease, increase and increase, respectively, in a market with increasing competition (due to increasing number o irms). lso, the increase o the last two sums up to dominate the decrease o the irst one, resulting in a smaller increase o MRTT in a market with higher number o irms. That is MRTT < MRTT. Thereore, when the technology transer is small (then MCTT = x is small), the MRTT could be less than MRTT or a suiciently small. 7 In this case, we can expect that equilibrium technology transer under direct-entry mode may be smaller (larger) than that under the acquisition mode or a suiciently small (large) delegation scheme. To see this, we can solve the optimal technology 7 Note that, when = 0, MRTT = x and MRTT = x.

13 transer or the oreign owner by setting MRTT = MCTT in (7): x ( )( a c) 9 8 and 3( )( a c) x (8) 8 9 Since MCTT is independent o and the MRTT is increasing in (according to Lemma ), we expect that, when increases, the equilibrium technology transer increases. We build Proposition as ollows: Proposition. Taking delegation incentive as given, the ollowing properties or MN exist: ()Under each speciic entry mode, the higher the delegation incentive is, the higher the technology transer o the MN. () x x could happen under management delegation scheme i the delegation incentive is suiciently high such 5 that. 6 Proo. x ( a c) > 0 and 9 8 x 3( a c) > x x = 5 6, (9 8)(8 9) which is greater than zero when 5, which is smaller than. 6 Clearly, without delegation scheme (i.e, = 0), x x always holds. However, with delegation scheme, the technology transer under direct entry mode could be larger than that under acquisition mode i the delegation incentive to the management is suiciently high. This is a sharp dierence rom the traditional view, ie., the case without management delegation scheme.

14 3.3 Optimal management delegation scheme In the second stage, the MN owner chooses the level o incentive scheme (). By (5) and (8), we derive the equilibrium outputs as ollows: ( a c) 4 ( a c) q 8 4, q 0 (9a) ( a c) 3 3 ( a c) 9 9 i, q 3 0 (9b) q Since q < q, it requires q 0 to ensure all non-negative outputs, which i i obtains what the ssumption states: ( 4 6) (4 3). ccording to ppendix, we then have the MN s net proit margins: M a c ( )(9 8) (4 ) ( ) and a c ( )(8 9) 3(3 3 ) M ( ) (0) Substituting unctions (8)~(0) into proit unction (), and taking FOC o with respect to, i.e., 0. We show in the ppendix to obtain the optimal solutions o the incentive design or each entry mode: 53 4(9 6)(9 ) i ; i (8 6)(8 8) 44 (a) (b) 3

15 ~ Where 0, 0, 0. 6 Since both - > 0 and ~ - > 0, we draw Figure and build the ollowing proposition: Figure quilibrium delegation scheme Proposition. In equilibrium, the direct entry mode leads MN to choose higher delegation incentive to the management than acquisition entry mode. s demonstrated in Lemma, a higher delegation incentive (ie., a more aggressive behavior in output) helps to increase the marginal revenue o technology, and this help turns out to be larger in a market with higher number o irms (i.e, a market with the entering directly). This drives the owner o the MN to set up a higher delegation incentive under the direct-entry mode. Plugging () into (8), we will obtain the optimal technology transers or each entry mode. For comparison, we irst look at the level o technology transer when there is no management delegation scheme, which gives us that 4( a c) 9 8 x 0 and 4

16 3( a c) 8 9 x 0 and x > x holds always. Moreover, according to Proposition, x and x with delegation scheme will be greater than that without delegation. We calculate all the our types o optimal technology transer and draw Figure. (a-c) Delegation: x x No Delegation: Figure quilibrium technology transer Proposition 3. Management delegation leads to higher optimal technology-transer o the MN than no-delegation case with either entry mode. It also leads the acquisition MN to optimally transer lower technology than direct-entry one(i.e., x < x ), which reserves the result o no-delegation case. The x > x in Mattoo et al. (004) and Chen and Chong (04), which do not include management delegation, is reversed in this paper. It s ound x < x to hold always in the equilibrium with a delegation scheme. 3.4 Optimal entry mode o the oreign irm In the irst stage, the oreign irm chooses the mode or entering the host country. We assume that the oreign irm has all the bargaining power under acquisition and it 5

17 oers the take-it-or-leave-it price T = to buy out a domestic irm, which equals m q. Net proits are the gross proit net o the entry payments. ccording to equation (), the MN s net proit unction turns out to be ~ = =. Contrarily, the MN pays a ixed entry ee T when adopting M q x m q the mode o direct entry, which is assumed to be t(a-c) or simplicity. Thereore, its net proit is ~ = t(a-c) = M q x t(a-c). By substituting (8)~() to above-mentioned proit unctions, we obtain each irm s optimal gross proit in a reduced orm o given each speciic entry mode. In this stage, the MN will decide which entry mode be chosen, depending on whichever provides higher net proit. Note that, i we substitute = 0 instead o () to proit equation (), we obtain each irm s optimal gross proit in a reduced orm o or the case without management delegation: ~ 0 and 0. With the case o no management delegation, the market with direct-entry MN has higher competition than the market with acquisition. ccordingly, the ormer has higher diminishing rate o gross proit than the latter when the cost o technology transer becomes higher (i.e., becomes higher.) Though direct-entry mode induces the MN to reduce more costs via transerring more technology, its gross proit goes less than that under acquisition mode when is enough high ( = 4.54 shown in Figure 3). The same reasoning applies in the case with delegation, except that both mode earns more gross proits or the MN and < occurs late when =

18 (a-c) Figure 3 Gross proits with reduced orm The MN considers its net proits to decide which mode to adopt. I the net-proit dierence ~ - ~ > 0, acquisition-entry becomes the equilibrium mode o the game. Since all the reduced orm o proit unctions contains the term (a-c), we can scale out the term: ( ~ ~ c )/( a ) = ~ ( )/( a c) + t. Deine ( ~ c )/( a ) t, then acquisition is the equilibrium entry-mode i t > t, otherwise direct-entry mode is the equilibrium. Likewise, deine ( ~ c 0 0)/( a ) t 0, then net-proit dierence under the no-delegation case ( ~ ~ c )/( a ) > 0 i t > t 0, implying acquisition is the equilibrium entry-mode when there exists no delegation. Figure 4 shows the choice o the optimal entry mode. When technology transer is not too ineicient (such that.75), t i > t 0, indicating the probability o direct-entry to be the optimal mode increases due to the adoption o management delegation. On the other hand, the MN s adoption o management delegation increases the probability o acquisition being chosen when technology transer is too ineicient (such that >.75) (because t i < t 0 ). 7

19 0.5 I t > t, mode is chosen under delegation I t > t 0, mode is chosen under no delegation 0. t t Figure 4 The equilibrium entry mode Proposition 4. Management delegation induces the MN to choose acquisition when costs o entry are higher than t. It decreases the likelihood o the MN adopting direct-entry i the costs o technology transer is suiciently high ( >.75); It increases the likelihood i otherwise. ccording to proposition 3, x < x always holds under management delegation. Thereore Figure 4 indicates that, when entry cost is too high such that t > t, the acquisition entry o MN becomes the equilibrium mode, which results in a lower level o technology transer than the direct-entry NN. Moreover, according to Proposition 4 and 3 we see that, when.75, management delegation increases the probability o direct-entry being an equilibrium mode, which then provides higher technology transer than o-equilibrium mode. On the contrary, when >.75, management delegation decreases the probability o direct-entry being an equilibrium mode and results in lower technology transer than o-equilibrium mode. 8

20 Proposition 5. cquisition is the equilibrium mode when the entry cost is high (t > t ), but it induces lower technology transer than o-equilibrium mode. When the cost o technology transer is high ( >.75), the likelihood o direct entry being the equilibrium decreases, but it results in higher technology transer than o-equilibrium mode.. 4 Host country welare Since the consumer surplus (CS) equals Q and the domestic proit equals i and + v =, the host country welare can be expressed as: 8 W ( Q ) and W W t( a c), where W ( Q ), ( q ) ( a c) i i i () It is obvious that Q > Q and i, giving an undetermined sign or W W. However, utilizing (9) and (), we could reduce equation () to a unction o, which can be then simulated as shown in Figure 5. s it shows, W W < 0 always holds i That is, when the technology transer is suiciently eicient, the host government always preers acquisition mode; but when > it preers acquisition only i the MN s entry ee t < t. gain or the sake o comparison, we calculate social welare o the case without delegation scheme, utilizing (9) with = 0. Figure 5 shows W W > 0 only i t < t 3, where t 3 is always greater than t. Thereore, the MN s management delegation decreases the likelihood o acquisition mode being the host government s like. 8 Our paper starts with a given demand unction, rather than the consumers utility unction. Thereore, we approximate welare by summing consumers surplus and irms proits, as in Shy (995), p.68. 9

21 t 3 I t < t, Host preers when MN adopts delegation I t < t 3, Host preers when MN adopts no delegation t chosen under no delegation Figure 5 Host government s preerence Proposition 5. Management delegation surely leads acquisition mode to achieve higher domestic welare i the cost o technology transer is low ( ); i otherwise, it leads acquisition to higher welare only on the condition that the entry cost is small (t < t ). In contrast with no-delegation case, the likelihood o the host government preerring acquisition decreases and is decreasing in the cost o technology transer. By placing Figure 4 and 5 together as in Figure 6, we see that t < t 3, t 3 < t 0, t 3 < t and t 0 < t when <.75. It s clear that the host government dislikes the MN optimal entry mode when (t, ) combination is such that t > t or t < t. In the ormer case, the MN optimally chooses acquisition mode but the host country likes the direct-entry mode; in the latter case the MN optimally chooses direct-entry mode while the host government likes the acquisition mode. However, both the MN and the host government likes the direct-entry mode when t < t < t. Figure 6 also shows that, when the MN adopts no management delegation, the case o both the MN and 0

22 the Host preerring direct entry mode occurs when t 0 < t < t 3. Since the space o t 0 < t < t 3 is smaller than t < t < t, the management delegation eectively reduces the probability o the host government liking the MN s optimal choice. We see that when the costs o entry are low such that t < t (t > t ), the domestic welare is generally higher under acquisition (direct entry) but the oreign irm preers direct entry (acquisition) instead o acquisition (direct entry) t t 0 t 3 t Figure 6 Preerence o the MN and the Host Corollary. When the costs o entry are suiciently low such that t < t (t > t ), acquisition (direct-entry) mode generally achieves higher domestic welare, but the equilibrium mode o the MN is direct-entry (acquisition). When t < t < t, both the MN and the host government like direct-entry mode, with a smaller likelihood than no-delegation case. 5 Conclusions International FDI lows have been growing rapidly, resulting in enormous growth o the number o multinational enterprises and the study o FDI eect. mong

23 others, the issue o technology transer and entry mode o oreign irms has gained increasing attention. However, the variable o management delegation which has to be employed by the FDI MNs is not included in the concerning FDI literature. This paper thus build a our-stage game model: the owner o the MN chooses entry mode (acquisition or direct entry), sets delegation incentive scheme to his/her managers, decides the level o technology transer, and inally the managers choose output to Cournot-compete with domestic irms. By the ramework, we are exploring the eect o management delegation on the MN s decision behaviors. The paper then inds that the direct-entry MN always chooses higher delegation incentive than the acquisition one. s a results, management delegation leads to higher optimal technology-transer o the MN than either mode in no-delegation. It also leads the direct-entry MN to optimally transer higher technology than an acquisition one, which reserves the result o no-delegation case. In contrast to no-delegation case, i the costs o technology transer are suiciently high, management delegation decreases the likelihood o the MN adopting direct-entry; i otherwise, it increases the likelihood. quilibrium mode depends on either costs o entry or technology transer. When the entry cost is high, acquisition becomes the equilibrium mode or the MN, but it then induces lower technology transer than the o-equilibrium mode. When the cost o technology transer is suiciently high, the likelihood o direct entry being the equilibrium o the MN decreases, but it induces higher technology transer than the o-equilibrium mode. s ar as the domestic welare is concerned, management delegation surely leads acquisition mode to achieve higher domestic welare i the cost o technology transer is suiciently low; i otherwise, it leads acquisition to higher welare only the entry

24 cost is small. Generally, when the entry cost is suiciently low (high), acquisition (direct-entry) mode achieves higher domestic welare, but the equilibrium mode o the MN is direct-entry (acquisition). This situation becomes more signiicant when management delegation is employed by the MN. When the entry cost is somehow intermediate, both the MN and the host government like direct-entry mode, with a smaller likelihood than no-delegation case. ppendix Deriving Lemma : D n ( a c) 0; n S n( n ) ( a c) 0; ( n ) n ( a c) 0; n D ( a c) 0 n ( n ) S n [(3n ) ( n )]( a c) (3n ) x 0 i is 3 ( n ) not suiciently large. q ( a c) x [( ( n ) ]( a c) ( n ) x n n ( n ) 0 Deriving optimal delegation incentive : a c For acquisition mode, Q q q y, where y. Then the proit margin is obtained as: which results in the domestic irm s proit m q q a c m ( a c Q ) y q, 3 when the MN enters by acquisition. The technology transer can be rewritten as x y ( a c) = a c 3 3y and the overall proit margin or the MN is M m x = 3

25 a c 3 y. ccordingly, the derivative o the MN s proit unction M q ( x ) v with respect to is: d m x q x a c ( ) q M x ( D d 3(9 8) whered ( ) ( ) ( ) ( ) 4(9 6)(9 ) 36 0 and B B ) 0, Thereore, the optimal solution B D is derived. ac 33 Q q q 3 y, where y, For direct-entry mode, i a c 4 which gives m ( a c Q ) y qi and i m qi qi. a c technology transer can be reduced to x y ( a c) 4 y The and the overall 4 a c proit margin o the MN becomes M m x 3y. Thus, the 4 derivative o the MN s proit M q ( x ) F with respect to is: d m x q x a c ( ) q M x ( D B ) 0, d 4(8 9) ( )( ) ( ) ( ) ( ) where D 6(8 6)(8 8) 44 and B 7 0. B It gives the optimal solution or. D 4

26 Reerences l-kaabi, M., Demirbag, M., Tatoglu,. (00). International market entry strategies o emerging market MNs: case study o Qatar telecom. Journal ast-west Business 6(), Bjorvatn, K. (00). Foreign ownership and market entry. Nordic Journal o Political conomy 7, 3-3. Demirbag, M., McGuinness, M., ltay, H. (00). Perceptions o institutional environment and entry mode: FDI rom an emerging country. Management International Review 50(), icher, T., Kang, J.W. (005). Trade, oreign direct investment or acquisition: Optimal entry modes or multinationals. Journal o Development conomics 77, lango, B. (005). The inluence o plant characteristics on the entry mode choice o overseas irms. Journal o Operations Management 3, thier, W.J., Markusen Jr., C.D. (996). Multinational irms, technology diusion and trade. Journal o International conomics 4, -8. Fatica, S. (00). Investment liberalization and cross-border acquisitions: The eect o partial oreign ownership. Review International conomics 8(), Fershtman, C. and Judd, K. (987), quilibrium incentives in oligopoly, merican conomic Review 77, Fumas, V.S., (99). Relative perormance evaluation o management: the eects o industrial competition and risk sharing, International Journal o Industrial Organization 0, Glass,.J., Saggi, K. (00). Multinational irms and technology transer. Scandinavian Journal o conomics 04(4), Kasuga, H. (003). Capital market imperections and orms o oreign operations. International Journal o Industrial Organization, Lambertini, L., (000). xtended games played by managerial irms, Japanese conomic Review 5, Lambertini, L., Trombetta, M. (00). Delegation and irms ability to collude, Journal o conomic Behavior & Organization 47, Lee, F.C., Shy, O. (99). welare evaluation o technology transer to joint ventures in the developing countries. The International Trade Journal 7, Lee, H.H., Yang, T., Chen, C.B., Chen, Y.L. (0). uzzy hierarchy integral analytic expert decision process in evaluating oreign investment entry mode selection or Taiwanese bio-tech irms. xpert Systems with pplications 38(4), Markusen Jr., C.D. (00). Contracts, intellectual property rights, and multinational investment in developing countries. Journal o International conomics 53, 5

27 Mattoo,., Olarrega, M., Saggi, K. (004). Mode o oreign entry, technology transer, and FDI policy. Journal o Development conomics 75, 95-. Miller, N. and Pazgal,.I. (00). Relative Perormance as a Strategic Commitment Mechanism, Managerial and Decision conomics 3, Miller, N. and Pazgal,.I. (005). Strategic Trade and Delegated Competition, Journal o International conomics 66, 5-3. Muller, T., (007). nalyzing modes o oreign entry: greenield investment versus acquisition. Review o International conomics 5(), 93-. Sklivas, S., (987). The strategic choice o managerial incentives, RND Journal o conomics 8, Teixeira,..C., Grande, M. (0). ntry mode choices o multinational companies (MNCs) and host countries corruption: review. rican Journal o Business Management 6(7), UNCTD (00). Foreign direct investment database/fdi On-line. Page. asp?intitemid=93&lang= Vickers, J., (985). Delegation and the Theory o the Firm, conomic Journal 95, Williams. D. (005). Supplier linkages o oreign-owned manuacturing irms in the UK: The inluence o entry mode, subsidiary autonomy and nationality. uropean Planning Studies 3,