THE IMPACT OF ECONOMIC INTEGRATION ON FDI AND EXPORTS competing RIAs.He finds that at fixed tariffs,the only way to integrate between two regional blocs is through FDI,which leads to full trade diversion and investment creation. The effect of FDI on the tariff level is also analysed.Given that high tariffs would encourage FDI more than they encourage exports,MNCs exert a pressure upon the bloc's capacity to set high tariffs and prevent tariff wars internationally.Nevertheless,it looks to us that if the purpose of the bloc is to protect itself from both FDI and exports from non- members (for which a justification has also to be found),then setting lower tariffs may indeed increase the opportunity-cost of MNCs to invest in the bloc,but it also reduces the cost of exports!Moreover,as highlighted by Motta and Norman(1996),regional blocs will benefit more in terms of welfare if they concentrate on reducing internal barriers rather than coordinate on tougher external trade policy.The explanation is that since FDI is beneficial for the host country,each member of the RIA has an incentive in encouraging it, but non-bloc MNCs will flow in only under the condition of 'market-accessibility',which is ensured by minimising the level of intra-bloc trade barriers.Finally,whatever is the assumed relationship between exports and FDI(complements or substitutes)or the nature of FDI (horizontal or vertical)a tariff reduction will always benefit imports of intermediary goods in the host country or imports of the final good in the home country,hence favouring both exports and FDI. The net result on FDI in a RIA of these tariffs adjustments therefore depends on which effects will prevail,the pro-FDI or the anti-FDI one.Since it is not clear from the theoretical point of view,it becomes an empirical issue,as pointed out by Winters (1997). Of course I am not considering other tools to protect members of a bloc from non-bloc investors here,as for example the rules of origin set up by the EU.By imposing a local- content requirement',the EU can protect itself from imports of Korean goods produced in Eastern Europe,that benefit from free-trade under the Europe Agreements.The effect is to discourage such investment by non-members and it would not be picked up only by looking at the level of the tariffs. I would just like to point out here that in the theoretical literature the implicit (crucial) assumption is that barriers to trade are a good incentive for FDI.But if this were to be true,real-world data should show that most FDI arises between developed and developing countries (or blocs),where the average MFN tariffs are higher.Nevertheless in 1998 over 90%of FDI flows were within developed countries.In my empirical analysis I explicitly introduce the tariffs variable into the FDI equation. Similar arguments apply to NTBs'reduction.There is a straightforward positive effect on exports,given that their reduction diminishes trade costs,but the impact appears less clear for FDI.On the one hand,exports become more profitable compared to FDI,when a firm decides how to service the foreign market;on the other hand,and especially for non- members of the RIA,their reduction transforms investment in a member country into a good export-platform to reach all the other member countries (the market access argument)and hence we should observe an increase from non-members.The net effect is again uncertain theoretically and justifies the direct inclusion of this variable in the gravity equations. 9
THE IMPACT OF ECONOMIC INTEGRATION ON FDI AND EXPORTS 9 . competing RIAs. He finds that at fixed tariffs, the only way to integrate between two regional blocs is through FDI, which leads to full trade diversion and investment creation. The effect of FDI on the tariff level is also analysed. Given that high tariffs would encourage FDI more than they encourage exports, MNCs exert a pressure upon the bloc's capacity to set high tariffs and prevent tariff wars internationally. Nevertheless, it looks to us that if the purpose of the bloc is to protect itself from both FDI and exports from nonmembers (for which a justification has also to be found), then setting lower tariffs may indeed increase the opportunity-cost of MNCs to invest in the bloc, but it also reduces the cost of exports! Moreover, as highlighted by Motta and Norman (1996), regional blocs will benefit more in terms of welfare if they concentrate on reducing internal barriers rather than coordinate on tougher external trade policy. The explanation is that since FDI is beneficial for the host country, each member of the RIA has an incentive in encouraging it, but non-bloc MNCs will flow in only under the condition of ‘market-accessibility’, which is ensured by minimising the level of intra-bloc trade barriers. Finally, whatever is the assumed relationship between exports and FDI (complements or substitutes) or the nature of FDI (horizontal or vertical) a tariff reduction will always benefit imports of intermediary goods in the host country or imports of the final good in the home country, hence favouring both exports and FDI. The net result on FDI in a RIA of these tariffs adjustments therefore depends on which effects will prevail, the pro-FDI or the anti-FDI one. Since it is not clear from the theoretical point of view, it becomes an empirical issue, as pointed out by Winters (1997). Of course I am not considering other tools to protect members of a bloc from non-bloc investors here, as for example the rules of origin set up by the EU. By imposing a ‘localcontent requirement’, the EU can protect itself from imports of Korean goods produced in Eastern Europe, that benefit from free-trade under the Europe Agreements. The effect is to discourage such investment by non-members and it would not be picked up only by looking at the level of the tariffs. I would just like to point out here that in the theoretical literature the implicit (crucial) assumption is that barriers to trade are a good incentive for FDI. But if this were to be true, real-world data should show that most FDI arises between developed and developing countries (or blocs), where the average MFN tariffs are higher. Nevertheless in 1998 over 90% of FDI flows were within developed countries. In my empirical analysis I explicitly introduce the tariffs variable into the FDI equation. Similar arguments apply to NTBs’ reduction. There is a straightforward positive effect on exports, given that their reduction diminishes trade costs, but the impact appears less clear for FDI. On the one hand, exports become more profitable compared to FDI, when a firm decides how to service the foreign market; on the other hand, and especially for nonmembers of the RIA, their reduction transforms investment in a member country into a good export-platform to reach all the other member countries (the market access argument) and hence we should observe an increase from non-members. The net effect is again uncertain theoretically and justifies the direct inclusion of this variable in the gravity equations
FRANCESCA DI MAURO 2.2 Impact of monetary integration The issue of exchange rate variability and its effect on trade has been debated at length. Many empirical studies have been produced since the collapse of the dollar exchange standard in 1971,and the passage from a fixed to a flexible exchange rate regime.The studies either compare different exchange rates regimes(De Grauwe and Bellefroid(1986)) or model explicitly exchange rate variability and trade flows(see Peree and Steinherr(1989), Dell'Ariccia(1998)).The purpose here is not to find a better measure of it,but to look at the relative impact on exports and FDI. The economics behind a supposed(negative)relationship between exchange rate variability and trade flows stems from the fact that the exporter incurs some production costs in his own currency,while the future revenues are expressed in a foreign one8.Any change in the exchange rate during the period before payment is therefore a potential loss or benefit for the exporter.Of course,forward currency markets exist and can be used,but generally they are unavailable for periods longer than one year and are practically non-existent in developing countries.Moreover,the presence of long-term commitments to the export activity,as for example to obtain export licenses,may act as sunk costs not hedgeable'in the forward market.These factors all lead to assume a certain risk-aversion against exchange rate variability and hence that trade would be favoured by stable exchange rates. Previous empirical studies have not arrived at a clear-cut conclusion on the issue,with the results ranging from no effect of exchange rate variability (Gagnon (1993))to some findings of a negative effect(De Grauwe and Bellefroid (1986),Dell'Ariccia(1998),Peree and Steinherr(1989)) As far as FDI is concerned,it has been very much neglected by the previous empirical analysis,even though it shares many characteristics with trade and therefore one would expect economists to investigate its relationship with exchange rate variability as well.The first step of the reasoning is that if exchange rate variability is really bad for trade,we may expect that it will induce FDI,because then by directly localising the production abroad firms would eliminate part of the exchange rate risk.Secondly,if a RIA does reduce exchange rate variability among its members,then there is a strong case to observe a reduction in FDI from non-members,ceteris paribus,and hence an increase in exports as the preferred channel of integration,given the reduced exchange rate risk.But contrary arguments could also apply,if exchange rate variability is intrinsically bad for FDI too.For example repatriated profits are worth less in case of a revaluation of the domestic currency, exported inputs may be worth less in case of a devaluation or imported final goods may be more expensive in case of a devaluation.Again,even a priori the net effect on FDI is not clear from a theoretical point of view and needs to be evaluated empirically.In my analysis I will consider two medium-term measures of exchange rate variability,following De Grauwe and Peree and Steinherr.This is especially appropriate for FDI,whereby foreign investors look at the long-term stability of a currency when deciding where to invest,rather than the short one,given the greater sunk costs that they incur. 8 A similar argument can be made for the importer,whereby his buying contract is in foreign currency,while the selling price is set in domestic value and is therefore subject to unexpected variations. 10
FRANCESCA DI MAURO 10 2.2 Impact of monetary integration The issue of exchange rate variability and its effect on trade has been debated at length. Many empirical studies have been produced since the collapse of the dollar exchange standard in 1971, and the passage from a fixed to a flexible exchange rate regime. The studies either compare different exchange rates regimes (De Grauwe and Bellefroid (1986)) or model explicitly exchange rate variability and trade flows (see Perée and Steinherr (1989), Dell'Ariccia (1998)). The purpose here is not to find a better measure of it, but to look at the relative impact on exports and FDI. The economics behind a supposed (negative) relationship between exchange rate variability and trade flows stems from the fact that the exporter incurs some production costs in his own currency, while the future revenues are expressed in a foreign one8. Any change in the exchange rate during the period before payment is therefore a potential loss or benefit for the exporter. Of course, forward currency markets exist and can be used, but generally they are unavailable for periods longer than one year and are practically non-existent in developing countries. Moreover, the presence of long-term commitments to the export activity, as for example to obtain export licenses, may act as sunk costs not ‘hedgeable’ in the forward market. These factors all lead to assume a certain risk-aversion against exchange rate variability and hence that trade would be favoured by stable exchange rates. Previous empirical studies have not arrived at a clear-cut conclusion on the issue, with the results ranging from no effect of exchange rate variability (Gagnon (1993)) to some findings of a negative effect (De Grauwe and Bellefroid (1986), Dell’Ariccia (1998), Perée and Steinherr (1989)). As far as FDI is concerned, it has been very much neglected by the previous empirical analysis, even though it shares many characteristics with trade and therefore one would expect economists to investigate its relationship with exchange rate variability as well. The first step of the reasoning is that if exchange rate variability is really bad for trade, we may expect that it will induce FDI, because then by directly localising the production abroad firms would eliminate part of the exchange rate risk. Secondly, if a RIA does reduce exchange rate variability among its members, then there is a strong case to observe a reduction in FDI from non-members, ceteris paribus, and hence an increase in exports as the preferred channel of integration, given the reduced exchange rate risk. But contrary arguments could also apply, if exchange rate variability is intrinsically bad for FDI too. For example repatriated profits are worth less in case of a revaluation of the domestic currency, exported inputs may be worth less in case of a devaluation or imported final goods may be more expensive in case of a devaluation. Again, even a priori the net effect on FDI is not clear from a theoretical point of view and needs to be evaluated empirically. In my analysis I will consider two medium-term measures of exchange rate variability, following De Grauwe and Perée and Steinherr. This is especially appropriate for FDI, whereby foreign investors look at the long-term stability of a currency when deciding where to invest, rather than the short one, given the greater sunk costs that they incur. 8 A similar argument can be made for the importer, whereby his buying contract is in foreign currency, while the selling price is set in domestic value and is therefore subject to unexpected variations