Optimal policy for FDI incentives: An auction theory approach

Transcription

1 European Research Studes, Volume XII, Issue (3), 009 Optmal polcy for FDI ncentves: An aucton theory approach Abstract: Israel Lusk*, Mos Rosenbom** A multnatonal corporaton s (MNC) entry nto a host country brngs benefts to the economy of that country, some drect (such as ncreasng producton and employment) and some ndrect (such as productvty spn-off). overnments that vew MNCs as engnes for growth and regonal development have begun to encourage the flow of foregn nvestment nto ther country n hopes of ncreased local employment, market producton and export capacty. MNCs consder frst the maxmzaton of proft when selectng a ste to establsh ther subsdares. An MNC examnes possble nvestment stes and ndcates those that are best ftted for the nvestment. The countres that reman at ths stage are smlar n terms of ther economc characterstcs, and they compete wth each other for recevng the foregn nvestment. In ths paper we use tools from aucton theory to analyze the competton between host countres and MNCs and nvestgate the exstence of Nash equlbrum strateges. The characterstcs of ths equlbrum are consdered and assessed. We developed a general model for examnng the ncentve competton between two countres and then apply t for several subgroups accordng to the number of MNCs and the avalablty of nformaton. It turns out that the characterstcs of the equlbrum depend on the number of MNCs as well as on the structure of ther contrbuton to the host country economy. Key words: FDI, multnatonal corporatons, FDI ncentves, aucton theory. JEL Classfcaton: F, F3. *lusk@bgu.ac.l

2 6 European Research Studes, Volume XII, Issue (3), 009. Introducton A multnatonal corporaton s (MNC) entry nto a host country s accompaned by many benefts for that country s economy, some drect (such as ncreasng producton and employment) and some ndrect (such as productvty spn-off). Many governments that vew MNCs as an engne for growth and regonal development, wth the possblty of expandng local employment as well as the market s producton and export capacty, have begun to encourage the flow of foregn nvestment nto ther country. The consderatons of an MNC for choosng a ste to establsh a subsdary are mostly concerned wth the maxmzaton of proft. In the frst stage, the MNC examnes possble nvestment stes and desgnates those that are sutable for the nvestment. In the second stage, the MNC chooses the ste that maxmzes ts profts, where the level of proft s affected by the ncentves offered by the host country. The countres remanng at ths stage are smlar n terms of ther economc characterstcs, and they compete wth each other for the foregn nvestment by offerng attractve ncentves. In ths paper, we use tools from aucton theory to analyze the competton among the home country (HC) and the MNCs and nvestgate the exstence of Nash equlbrum strateges. The characterstcs of ths equlbrum are presented and assessed. We developed a general model for examnng the ncentve competton between two HCs. Then, we extend the basc model and apply t to several subgroups accordng to the number of MNCs and the avalablty of nformaton. It turns out that the characterstcs of the equlbrum depend on the number of MNCs, as well as on the structure of ther contrbuton to the host country economy. The structure of the paper s as follows: The followng secton provdes a bref revew of the relevant lterature. The thrd secton presents the consderatons of an MNC when determnng an optmal locaton for ts subsdary and the consderatons of a government that uses FDI ncentves. Secton four dscusses several models of ncentves competton. In secton fve models of ncomplete nformaton are dscussed. A bref summary of the man results s presented n the last secton.. Background The establshment of a subsdary by an MNC may ncrease the HC s DP n varous ways: ncreasng employment, transfer of new technology, access to world markets, access to the MNC s R&D, and so on. The beneft to the HC s hgher when there are productvty and technologcal spllovers. Chuang and Ln (999), Drffeld (00), and Lpsey and Soholm (00) found that technology spllovers exst n the UK, Tawan and Indonesa and contrbute to the economc growth of these economes (see also Dmels and Lour, 00). However, Konngs (000) showed that n several countres (e.g., Bulgara and Rumana), the FDI may have a negatve mpact on the economy.

3 Optmal polcy for FDI ncentves: An aucton theory approach 63 The current locaton theory (see, for example, Dunnng, 993; loberman and Shapro, 999; Shapro and loberman, 00) asserts that the MNC optmal locaton s determned n a two-stage process. In the frst stage, a short lst s prepared of potental locatons that are characterzed by a stable economy, bg markets, hgh and growng ncome per capta, modern nfrastructure and good tradng condtons wth other countres. In the second stage, the MNC assesses the fnancal condtons n each of these locatons, ncludng the followng varables: corporate tax rates, factors prces, labor costs and so on. In ths stage, countres that are on the short lst compete for the FDI by offerng grants, tax reductons, and subsdes for varous factors of producton. For example, INTEL receved a 300 mllon dollar grant from the government of Costa Rca, whch was nstrumental n persuadng the company to choose that country for ts new plant. Canon Company chose Vetnam for ts new subsdary after the Vetnamese government offered a reducton n tax rates for 0 years (see Borvatn and Eckel, 006). A stuaton where countres wth smlar characterstcs compete for FDI by offerng ncentves s called ncentves competton. Ths competton affects the allocaton of benefts between the host country and the MNC. Oman (000) asserted that the MNC s share s close to one. Blomstrom and Kokko (003) even proposed that rules of game be mposed to prevent ths outcome and to leave a larger share of the benefts n the host country. Other studes emphaszed the advantage of competton among countres. For example, Barros and Cabral (000) showed that when countres dffer n ther sze and unemployment rates, subsdy competton leads to optmal FDI locaton. Smlar results can be found n Borvatn and Eckel (006), who show that ncentves competton may cause the MNC to locate ts subsdary n the country that has the hghest benefts of FDI. In our paper, we nvestgate the ncentves competton by applyng models of aucton theory, and then we present characterstcs of the equlbrum strateges. 3. Consderatons of the MNC and the HC The problem of an MNC s to decde where to locate ts subsdary. Let denote the expected profts of an MNC n a country at tme τ. Let t denote the corporate tax rate n country. The net present value of the expected stream of proft f I s nvested n the subsdary, whch s located n country s: π τ () NPV ( t) ( + k) n = τ= π τ τ I, where k s the MNC dscount rate.

4 6 European Research Studes, Volume XII, Issue (3), 009 Under the assumpton that only one subsdary can be establshed, the MNC should choose the optmal locaton, that s, the country that maxmzes the present value of the proft per unt of nvestment: () PI = n ( t ) τ= ( + k ) I π τ τ + where s the grant offered by country to MNC. s a decson varable of country. When the HC competes for an MNC, t consders the benefts that can be derved from the MNC s actvtes n the country. We can classfy these benefts nto two knds: drect benefts (DB), such as wages pad to workers who where prevously unemployed and corporate taxes pad by the MNC to the HC s government, and ndrect benefts (IDB), such as spllovers of new technologes, and so on. The total benefts country derves from the MNC, TB, at tme τ s:, (3) TB =DB +ID B Let r represent the dscount rate of country. Then, the present value of the benefts due to establshng a subsdary by MNC, R s: () R = n TB τ τ= ( + r ) τ The grant,, s pad by the HC to the MNC, and therefore the net present value of the benefts of HC s: (5) N B =R - The HC s goal s to maxmze ths value, subect to the condton that NB > 0. As we noted before, the proftablty for the HC and for the MNC depends on future outcomes. The qualty of the estmated level of profts depends on the avalablty of nformaton on the future expected values of the cash flows stemmng from the proect. Complete nformaton enables us to predct the accurate future cash flow of the proect. In ths case, both the MNC and the HC are ndfferent to the type of ncentves (a grant or tax relef) when both types prove to have the same net present value. Under ncomplete nformaton, the mportance of an

5 Optmal polcy for FDI ncentves: An aucton theory approach 65 ncentve scheme ncreases. Usually, the host country prefers tax relef that mnmzes ts rsk, but the MNC prefers grants so that possble changes n future tax schemes can be avoded. Both cases are consdered n our paper.. Models of Incentves Competton under Complete Informaton We assume that the HC and the MNC have complete nformaton and can make an accurate predcton of the future value of cash flows from the MNC s nvestment n the HC. In addton, we assume that the tax rate s constant for the proect s lfetme and that the only way to provde FDI ncentves s by offerng the MNC a grant. We consder the ncentves competton to be an aucton where each HC offers a grant and the MNC chooses the country that provdes t wth the hghest net present value of profts. The grant s pad only f the subsdary s actually establshed. Therefore, we can apply models of frst-prce aucton. We deal wth the followng types of stuatons: Two countres compete for a sngle MNC. Two countres compete for two MNCs. There are homogenous and nonhomogenous countres.. Two Homogenous Countres Compete for One MNC Two HC (=,) compete for an nvestment of one MNC, where R represents the value of total net benefts of host country (R =R ). When a HC offers a grant ( ), ts net benefts are: NB=R-. Each country can offer a specfc amount of the grant from a lst of possble grants (=,,M), such that <...< M. When the two countres are economcally homogenous, t means that that the MNC s aftertax cash flows t are the same n both countres. The MNC that wshes to maxmze ts PI (the present value of the profts per unt of nvestment) wll choose the host country that pays the hgher grant. We look for the domnant strategy of HC n terms of the grant ( ) to be offered. Proposton. For each HC strategy x:{ =R } domnates strategy y:{ >R }. Proof: Let us assume that HC uses strategy y:{ >R }. What would be the best strategy x for country one? NB NB NB ( x, y) = ( R) = 0 f < ( x, y) = ( R) < 0 f = ( x, y) = ( ) < > R 0 f Snce we assume complete nformaton, the competton between HCs can be descrbed as ether grant competton or tax relef competton.

6 66 European Research Studes, Volume XII, Issue (3), 009 When HC s grant s smaller then that of HC, the MNC chooses to nvest n HC. In ths case, NB =0. For all other cases, NB <0: f HC wll pay = >R =R, then HC wll wn wth the probablty of ½, and NB wll be negatve. If > >R =R, then HC wll wn, but NB wll be negatve. Proposton. When R =R, strategy x:{ =R } domnates strategy y:{ <R }. Proof: For smplcty we assume that m = m for each m. Table presents the NB for each country for all possble grants: M M Table. Net benefts of each country that faces a sngle MNC (assumng homogenous countres) ( R ) R ; 0 M R ; 0 M 0 ;R ( R ) M R ; 0 M O... M 0 ; R M 0 ; R M M R ( ) M The equlbrum strateges are: =R and =R. As R =R, the MNC s ndfferent between the two HCs. Therefore, we can summarze t n the followng proposton: Proposton. 3 When R =R, strategy x:{ =R } s Nash equlbrum. Proof: For homogenous HCs n equlbrum, both HCs use the same strateges { =R } where all the HC s benefts are transferred to the MNC. Usually, an HC has budgetary constrants on the value of the grant. Note that mposng budgetary constrants on the grants s way for coordnaton between the HCs. Let assume that Y s the maxmum amount of the grant that can be offered by HC. There are several possbltes: Y R and Y R : In ths case, the budgetary constrants are not effectve, so there s no change n the equlbrum. Y R and Y < R : In ths case, the equlbrum strateges are such that HC offers Y (lower than R ) and HC offers Y + and wns the MNC. In ths case, the surplus of HC s R -(Y +) > 0. Y < R and Y < R and Y > Y : In ths case, the equlbrum strateges are such that HC offers Y, and HC offers Y + and wns the MNC. In ths case, HC s surplus s R -(Y +) > 0.

7 Optmal polcy for FDI ncentves: An aucton theory approach 67 The above model explans why some countres make ther grant polcy vague. Each HC determnes ts optmal grant by usng nformaton on the other HC s constrant. Secrecy can nduce the HC wth the hgher budget constrant to make errors and lose the MNC s nvestment.. Incentves Competton when Countres Dffer Let us consder the case where the two HCs dffer wth respect to ther FDI benefts. The dfference may stem from dfferent spllovers from the MNC to the HC, dfferent unemployment rates, and so on. We nvestgate the mpact of ths dfference on the characterstcs of the equlbrum. As n the prevous sectons, two HCs are competng for a subsdary of one MNC, such that R >R. Each HC offers a grant ( ), whch can be chosen from a lst of M possble levels such that: <...< M. We assume that f there are the same yearly after-tax cash flows t n both countres, then the MNC would prefer the country that offers the hgher grant. The domnant strategy for each HC can be consdered accordng to the followng propostons: Proposton. The hghest grant that HC offers s less then R. Proof: See Proposton s proof. Proposton 5. If HC uses strategy y:{ =R } then HC s domnant strategy s x: { =R +. Proof: For HC strategy y:{ =R }, The net beneft of country one depends on as follows: NB NB NB ( x, y) = ( R ) ( x, y) = ( R ) ( x, y) = 0 f f f = > < Note, that f HC offers =, t wll wn the MNC nvestment wth probablty of ½, and the expected added value s as descrbed above. Snce R >R, t s worth t for HC to offer a hgher grant value ( > ) and to wn the MNC nvestment. Proposton 6. The strategy of HC y:{ =R } and HC x:{ =R + are equlbrum strateges. Proof: For HC strategy x:{ =R +, the changes n HC s net beneft as a result of changng ts strategy y:{ =R } are:

8 68 European Research Studes, Volume XII, Issue (3), 009 NB NB ( x, y) = 0 f = R ( x, y) = 0 f < R When HC offers a grant such that = R +, HC does not gan by offerng the same grant (proposton ), as the expected benefts of HC are negatve. On the other hand, HC does not gan by offerng a grant smaller than R. Therefore, the equlbrum strateges are: y:{ =R } and x:{ =R +. In ths secton, we have shown that, when there are two dfferent HCs and one MNC, the HC wth the hgher beneft wns the MNC nvestment. Ths HC pays a grant amount that s equal to the benefts of the other country. Thus, part of the HC s benefts remans n the HC. Note, f budgetary constrants are relevant, then the HC wth the hgher constrant wns the MNC s nvestment not necessarly the HC wth the hgher benefts. Thus, we can conclude that budgetary constrants may cause neffcency..3 Two Countres and Two MNCs Assume that there are two HC s (=,) and two MNCs. If there s no MNC nvestng at HC, the benefts of ths HC s zero (R 0 =0). If one MNC nvests at HC, the benefts are R, and f both MNCs nvest n HC the total benefts are R. We assume decreasng margnal benefts when the number of subsdares ncreases R ( R > ), and, f the MNCs are smlar, the offered grant by HC s the same for both HCs. The net benefts of HC are: NB = R (where =0,, and count the number of MNCs that nvest n HC ). Table. Net benefts matrx for two countres and two MNCs + + ( R ) ( R ) 0 R ; 0 0 ; R ( R ) ( R ) M R M 0 ; 0 ; R M

9 Optmal polcy for FDI ncentves: An aucton theory approach 69 M M R M M ; 0 M R M ; 0 O... M ( R ) ( R ) M M Table presents the net benefts for each HC. Each HC can offer the MNCs a grant of m (m=, M). The varous types of grants are ordered by ther sze ( <...< M ) and m = m for each m. For example: f HC offers and HC offers 3, both MNCs wll nvest at HC and net benefts for HC are NB = R. If HC offers 3 and HC offers 3, then the outcome s as follows: the probablty that both MNCs nvest at HC s ¼; the probablty s ½ that one MNC nvests n HC and ¼ that there s no nvestment at HC. The expected net benefts for HC n ths case are: R 3 3. ( ) + ( R ) + 0 If HC offers and HC offers 3, both MNCs wll nvest at HC and net benefts for HC s zero. We wll show that each HC gans by ncreasng ts grant up to a specfc lmt Max. Proposton 8. Equlbrum exsts for every par of grants ( = ) that satsfes: where: Mn Max 3 Mn = R R 0 and 3 Max = R + R 0. Proof: The lower boundary: Let us assume that HC offers < Mn ; HC can offer <. In ths case, HC wns and both MNCs nvest n HC and NB (HC s net beneft) equal to 0.

10 70 European Research Studes, Volume XII, Issue (3), 009 If HC offers =, then, wth probablty ¼, both MNCs nvest n HC ; wth probablty ½, one MNC nvests n HC ; and wth probablty ¼ there s no nvestment n HC. The expected net beneft for HC, n ths case s: NB = ) = ( R ) + ( R ) + 0= R + R + 0 ( However, HC can offer = +< Mn. In ths case, both MNCs nvest n NB ( = + ε ) = R +ε. It turns out that: HC, and ( ) NB ( = ) > and that 0 NB = + ) > NB ( = ) 0. ( ε > If HC offers < Mn then HC gans by settng = +< Mn. Therefore, any grants below Mn are not n equlbrum. The mddle range: Let us assume that HC offers a grant such that Max > > Mn. Now there are three possbltes for HC :. HC can offer <. In ths case HC wns both subsdares and NB (HC s net beneft) s equal to 0.. HC can offer =. In ths case, there s a lottery (wth a probablty of ¼ that both MNCs nvest n HC ; a probablty of ½ that one MNC nvests n HC and ¼ that there s no nvestment n HC ). The expected net beneft for HC s: NB = ( R ) + ( R ) + 0= R + R HC can offer = +>. In ths case, both MNCs nvest n HC and the net beneft s: NB = R ( +ε). Smple calculatons reveal that: If < Max, then offerng = provdes HC wth a hgher NB than t can get by choosng <. If < Max, then offerng = provdes HC wth a hgher NB than t can get by choosng =( +ε). Therefore, equlbrum exsts for any par of grants ( = ) that satsfes:. Mn Max..

11 Optmal polcy for FDI ncentves: An aucton theory approach 7 The upper boundary: HC wll not offer > Max. Explanaton: If HC offers that s greater than Max and f >, then both MNCs nvest n HC, and ts expected net beneft s: NB 3 = R Max ε ( + ε) = R R + R 0+ ε = R R 0 < If = then a lottery occurs and: NB = R + R ( Max+ ε) = R + R R + R + ε < 0 Snce, n both cases, NB s negatve, HC never offers a whch s above Max. To summarze, we showed that: When an HC offers < Mn the other HC gans by ncreasng ts grant. HC wll never offer > Max. Any such that Max >> Mn s an equlbrum f both HCs offer ths grant. The lower boundary and the upper boundary of the equlbrum strateges can be derved from the followng argument: Each HC may gve up an MNC nvestment and then ts NB s 0. On the other hand, a country may ncrease ts grant offer and can attract the nvestment of both MNCs. In ths case the NB equals R -. Then = s an equlbrum f: ( R ) + ( R ) + 0 max{ 0,R }. Ths nequalty provdes us wth the two boundares. The lower boundary s: 3 = R R 0, Mn and the upper boundary s: 3 = R + R 0. Max Note that =, such that Max > > Mn, s an equlbrum, but each HC may suffer ex-post loss. The equlbrum strateges ensure that the expected net beneft s postve. But, f an HC pad a hgh grant and won both MNCs, then the total amount of the grants may be hgher than the total net benefts of the HC, whch are derved from the entrance of the two MNCs. Ths outcome may

12 7 European Research Studes, Volume XII, Issue (3), 009 explan why n most countres there s no automatc system of offerng grants for all possble MNC proects.. The Number of MNCs Is Large We have shown n the prevous sectons that the power of the HCs ncreases as the number of MNCs rses. When the number of HCs s hgh and there are only a few MNCs, the HCs compete for the FDI n order to gan ts benefts. Each HC s wllng to offer the MNC a large share of the benefts rather than reman wth no FDI. On the other hand, when the number of MNCs s large, each HC knows that one of these MNCs wll eventually nvest there even f the ncentve s low or there are no ncentves at all. Furthermore, an HC may set constrants on FDI, such as lcensng. An MNC may prefer to pay the HC the lcense rather than cancel ts nvestment plan and lose any potental profts. Such an HC gans all the benefts of the FDI as well as part of the MNC s profts. 5. Incomplete Informaton of Benefts In the prevous sectons, under the assumpton of complete nformaton, all the partcpants n the ncentves competton know the value of all the varables. In ths secton, we ntroduce a model wth ncomplete nformaton. Specfcally, each country knows ts own beneft, R, but has no precse nformaton on the compettor s R. Smlarly, the MNC knows ts own expected profts, but does not know each country s R. For smplcty s sake, we assume that each R s unformly dstrbuted on the [0,] range, and ths nformaton s avalable to all the partcpants. 5. The Case of Two HCs and One MNC We assume that R denotes the net beneft for HC (=,), and that t s unformly dstrbuted (R ~U[0,]). Each HC offers a grant that depends on ts own R : = f ( R ). We also assume symmetrc grant functon such that f R = R then =. HC consderaton HC s net beneft s: R = NB (,, R ) ( R ) 0 f f f > = <

13 Optmal polcy for FDI ncentves: An aucton theory approach 73 Note that we actually gnore the case where =, as a contnuous densty functon s assumed. The two HCs use the same grant functon, f (R) =f (R) due to the symmetry between HC and HC. The optmal grant polcy of HC s the one that maxmzes ts expected net beneft: MAX ( R ) ( f( )) Pr R, > and due to the unform dstrbuton the abect functon can be wrtten as: MAX f ( R ) f ) ( f - s defned as the nverse of = f ( R ). Equlbrum exsts when f satsfes (see Krshna, 00): = R. Let us show that the strategy: each HC suggests a grant that s equal to half of the MNC net beneft s equlbrum. Assume that = R. In ths case, HC acts as follows: MAX MAX ( R ) Pr > R = MAX( R ) Pr( > R ) ( R ) Frst-order condton for maxmum s: = R 0, and therefore, = = R.

14 7 European Research Studes, Volume XII, Issue (3), 009 When HC uses f =½, then HC uses f =½ as well. Because of symmetry, f HC uses f =½, then HC uses f =½ as well. Therefore, equlbrum exsts when each HC offers a grant that s half of ts net beneft. 5. The Case of Several HCs and One MNC Under Incomplete Informaton There are several symmetrc HC (=, N) and one MNC, assumng ncomplete nformaton such that each HC knows ts own net beneft (R ) but the other Rs are unknown except ther dstrbuton functon. An HC s consderatons are: NB (,, 3,..., Thus, the obect functon s: MAX N R =, R ) ( R ) 0 ( R ) Pr ( > max f ( R ) f f f > max = max < max and the equlbrum strateges are (see Krshna, 00): f N N ( R) = R for each HC. We can see that as the number of HCs ncreases, the offered grant ncreases as well. The reason for ths s that the probablty of wnnng the MNC s nvestment decreases as the number of HCs ncreases, and each HC compensates by rasng ts grant. Therefore, when the number of HCs s large, the approaches R, and the HC s net beneft decreases. In summary, ths secton nvestgates cases of ncentves competton under ncomplete nformaton. When one MNC faces two HCs, equlbrum exsts when each HC offers a grant that s equal to half ts benefts. In equlbrum, the subsdary wll be establshed n the HC where ts contrbuton to the economy s hgher. The HC enoys half of the benefts of the FDI. An ncrease n the number of potental HCs ncreases the grant offered by each HC as well. Therefore, more ntense ncentves competton ncreases the share of the HC s benefts. 6. Summary and Conclusons Ths paper nvestgated ncentves competton n cases where there s complete nformaton on the HC s benefts as well as the MNC s profts. Each

15 Optmal polcy for FDI ncentves: An aucton theory approach 75 country tres to attract FDI by offerng the MNC an ncentve. We appled models of aucton theory to ths problem, presented the optmal strateges for each HC, and characterzed the Nash equlbrum strateges. There are two man questons regardng the equlbrum of the ncentves competton:. What s the porton of an HC s benefts that s spent on FDI ncentves?. Is the global allocaton of nvestment optmal? That s, should FDI go frst to HCs wth hgher productvty, and only when the sze of the FDI s large should the low productvty HCs get ther share of the nvestment? The man results are as follows. When there s one MNC and two (dentcal) HCs competng for an nvestment, at equlbrum, each HC offers an ncentve that s equal to ts total beneft derved from the FDI. Ths s the only case where our results concded wth those of Blomstrom and Kokko (003), who argued that a set of rules s requred to lmt the level of ncentves and to let each MNC gan a maor part of the FDI benefts. If there s a constrant on the amount of the ncentves (due to a budget shortage or global regulaton), the ncentves competton leads to an equlbrum where a larger share of the benefts remans n the HC, but t can also lead to neffcency where the new subsdary s located n the less productve HC. Ths fndng supports Borvatn and Eckel (006), who argued that competton between HCs mproves the effcency n the allocaton of FDI n HCs. When the two HCs dffer wth respect to ther benefts, the HC wth the hgher beneft wns the race and can offer ncentves that are equal to the other HC s lower beneft. When the HCs face several MNCs, and under the assumpton of dmnshng benefts, the equlbrum of the ncentves competton conssts of a range of possble ncentves. In ths case, t s possble that the actual value of the ncentves s greater than the total benefts of the HC. Thus, we have shown that under complete nformaton the characterstcs of the ncentves competton equlbrum depend on the number of partcpants n the game, whch s the number of HCs that compete for FDI as well as the number of MNCs who are lookng for nvestment locaton. Couples of models are analyzed under the assumpton of ncomplete nformaton. When one MNC faces two HCs, equlbrum exsts when each HC offers a grant that s equal to half the benefts, and the subsdary wll be establshed n the HC where ts contrbuton to the economy s hgher. An ncrease n the number of potental HCs ncreases the grant offered by each HC as well. Therefore, more ntense ncentves competton ncreases the share of the MNC s benefts.

16 76 European Research Studes, Volume XII, Issue (3), References:. Barros, P.P. and L. Cabral (000), Competng for Foregn Drect Investment, Revew of Internatonal Economcs, Vol. 8, Borvatn, K. and C. Eckel (006), Polcy competton for foregn drect nvestment between asymmetrc countres, European Economc Revew, Vol. 50, ssue 7, Blomstrom, M. and A. Kokko (003), The Economcs of Foregn Drect Investment Incentves, CEPR Feb Dmels, S. and H. Lour (00), Foregn Ownershp and Producton Effcency: A Quantle Regresson Analyss, Oxford Economc Papers, Vol. 5, Dunnng, J. (993), Multnatonal Enterprses and the lobal Economy, Readng, PA: Addson-Wesley.. loberman, S. and D. Shapro (999), The Impact of overnment Polces on Foregn Drect Investment: The Canadan Experence, Journal of Internatonal Busness Studes, Vol. 30 (3), Krshna, V. (00), Aucton Theory, San Dego, Academc Press. 6. Oman, C. (000), Polcy Competton for Foregn Drect Investment: A Study of Competton among overnments to Attract FDI, Pars: OECD. 7. Shapro, D. and S. loberman (00), Natonal Infrastructure and Foregn Drect Investment, mmeo, Smon Fraser Unversty (February). 8. Shapley, L. S. (953), A Value for n-person ames, n Contrbutons to the Theory of ames II (Annals of Mathematcs Studes 8), H. W. Kuhn and A. W. Tucker (eds.), Prnceton Unversty Press, Thalassnos, E., Thalassnos, P., (006), Stock Markets' Integraton Analyss, European Research Studes Journal, Vol. IX, Issue Thalassnos, E., Kyrazds, Th., Thalassnos, J., (006), The reek Captal Market: Caught n Between Poor Corporate overnance and Market Ineffcency, European Research Studes Journal, Vol. IX, Issue -, pp Thalassnos, E., (008), Trends and Developments n the European Fnancal Sector, European Fnancal and Accountng Journal, vol. 3, no. 3, pp