The effects of the foreign direct investment liberalisation on Pakistan.

Guisinger, Stephen

INTRODUCTION

Pakistan for many years maintained strict controls on foreign direct investment. However, over the past decade controls on foreign investment in manufacturing have diminished sharply, though less so for the service sector. The government continues to impose restrictions on foreign trade, which adversely affect foreign direct investors in several ways. Nonetheless, Pakistan has moved a substantial distance toward liberalising direct foreign investment.

There are two obvious policy issues related to foreign investment raised by these developments. First, should Pakistan proceed further toward liberalisation and at what pace? Second, with a liberalised investment sector, should Pakistan become an active protagonist among developing countries for a multilateral agreement on investment?

This paper explores the macroeconomic effects of foreign direct investment liberalisation on developing countries that have yet to substantially and fully liberalise. The principal focus will be on relatively short term effects--those changes that will occur between one and five years after liberalisation, although long-term effects are also discussed. Unfortunately, very little is known about the repercussions of foreign direct investment liberalisation on host economies. There is a rich literature on the effects of trade policy liberalisation on macroeconomic variables. Considerable scholarly work has been done on the impact of foreign direct investment on host economies under existing investment regimes. However, for a variety of reasons discussed below the link between investment liberalisation and macroeconomic performance has received scant attention from researchers.

This study summarises a few pieces of this small body of research on foreign direct investment, but this only takes us part of the way. As Sebastian Edwards noted in a recent study, "applied economists often ask too much of their data sets, and try to extract information that is simply not there" [Edwards (1993)]. With that caveat in mind, the paper takes two approaches to achieving its stated purpose of exploring the macroeconomic effects of investment liberalisation. One is to review the literature on trade liberalisation's in an effort to extract implications that those studies may have for investment liberalisation. The second approach is to utilise a large-scale simulation model to examine the repercussions of investment changes on macroeconomic variables. Neither approach is totally satisfactory, but together they provide some insight into the investment liberalisation process.

For the purposes of this paper, an investment regime will refer to the array of policies directly affecting the profitability of investments. These policies fall into one of three categories: (1) fiscal incentives and disincentives; (2) investment restrictions; and (3) trade policies. Fiscal incentives refer to the cash grants, subsidised loans, accelerated depreciation allowances, tax holidays and other inducements that governments use to encourage investment. Fiscal disincentives include taxes on corporate profits and income remittances abroad. Investment restrictions are the limitations that governments impose on investment. Some of these are explicit and automatic: investment in certain industrial sectors is allowed or forbidden. Some are subjective and discretionary: investment proposals must be approved by government agencies whose criteria may not be transparent. Trade policies include import duties, quotas and export subsidies that affect the price of traded goods and the profitability of investments. While many other policies influence the investment climate, these policy groups have the most direct bearing on investment profitability.

FOREIGN INVESTMENT REGIMES NEED A MEASURE

One critical lacuna in the literature is a comprehensive measure of the restrictiveness of investment regimes. Edwards notes in his survey of the effects of trade liberalisation on developing countries that the emergence of the concept of the effective rate of protection in the early 1960s made possible large-scale studies of the protective effects of trade policies both across countries and over time [Edwards (1993), p. 1362]. Without a comprehensive measure of the impact that all trade policies--i.e, those affecting capital, intermediate and final goods--have on the profitability of investments, economists had to rely on partial. And very imperfect, measures of the restrictiveness of trade policy regimes. The emergence of the effective rate concept represented a mini-paradigm shift in the way economists and policy-makers viewed trade policy and, importantly, the reforms that were needed.

Investment regimes lack an analogous measure of net incentive. Of the three policy components that make up an investment regime, only one--trade policy--has a quantitative measure--the effective rate of protection. Theorists have made progress in providing an analogous measure for fiscal incentives. Eric Bond and Stephen Guisinger [Bond and Guisinger (1985)] incorporated incentives in an expanded version of the effective rate of protection to show that when Ireland joined the EEC, the government compensated investors in Ireland from incentives lost when Ireland adopted the EEC's common tariff by increasing fiscal incentives. Guisinger showed that the effective rate of protection, the user cost of capital and the ratio of the financial to the economic rate of return are all logically related [Guisinger (1989)]. Any one of these three indices could be used to measure the net incentive (fiscal incentives less fiscal disincentives) provided by a host government to investors.

The greatest need is the quantification of investment restrictions. It does not appear to be an insuperable problem from the standpoint of theory. A host government's denial of right of establishment is analogous to a prohibitive tariff or a zero import quota. A host government's non-national treatment of foreigners imposes costs on the inward flow of capital in much the same way a tariff raises the cost of imported goods. Lack of convertibility (or the risk that a currency might become inconvertible) raises costs to investors. There is no fundamental reason that researchers could not devise tariff-equivalents for restriction-induced costs. Perhaps the greatest challenge to theory is the representation of the deterrent effect of an investment approval agency using subjective criteria.

The real problems lie in empirical estimation. Data on investment restriction costs are much more difficult to collect than costs imposed by trade policies because of differences in the way restrictive policies are applied. Within the same industry, tariffs and quotas apply uniformly to all importers, so that only one piece of information is needed for each imported product. For example, all quota holders reap the same benefit from the quota-constrained supply of imports and each holder raises the domestic selling price above the c.i.f. price by the same amount. In contrast, investment restrictions are often imposed on a project-by-project basis even within the same industry or product category. No two investors in the same narrowly defined industrial category face exactly the same set of restrictions or incentives. Amassing reliable data on a project-by-project, identifying the dozen or more restrictions and incentives that apply to the typical investment in developing countries is a daunting task.

While it would be incorrect to say that we know nothing about the degree of restrictiveness, our knowledge is largely qualitative and subjective. As a result, comparisons of the degree of restrictiveness across countries and across time are difficult. It is easy to understand why so few studies have attempted to examine the effects of investment liberalisation on macroeconomic variables.

WHAT CAN WE LEARN FROM TRADE LIBERALISATIONS?

In the absence of studies of the macroeconomic repercussions of investment liberalisation, we can look to other policy liberalisation experiences for lessons they might provide. Of the liberalisations that developing countries undertake, trade liberalisation comes closest to resembling investment liberalisation, since both produce increases in imports and exports. One could argue that investment liberalisations and trade liberalisations are mirror images of one another: investment liberalisations free capital inflow which gives rise to new imports and exports, while trade liberalisations free imports and exports which give rise to new capital inflows. Imports, exports and capital investment expand following both types of liberalisations

The link between trade, capital and growth has been made in a number of other research studies. Robert E. Baldwin [Baldwin (1992)] has emphasised, for example, that capital accumulation from both domestic and foreign sources contributes significantly to the effects of trade liberalisation on output. Brad De Long and Lawrence Summers found a very strong link between investment in equipment and economic growth [De Long and Summers (1993)]. Although perhaps of smaller significance in developing countries than in developed countries because of the vast range of experience with growth rates, the link is nonetheless evident from the data and the line of causation appears to run from equipment investment to growth and not vice versa. For those developing countries producing few capital goods, equipment investment can only rise from liberalised imports. This is another way in which trade liberalisation and investment liberalisation are tied together. De Long and Summers find net social rates of return to capital investment averaging in the range of 25 percent. Although De Long and Summers do not examine foreign direct investment directly, it is evident that to the extent lower investment restrictions permit greater amounts of foreign direct investment, their data support a strongly positive link between foreign direct investment and economic growth.

The literature on trade liberalisation is vast. However, there have been three large-scale, multi-country studies of trade liberalisation. These studies involved large numbers of researchers over several years examining interrelationships between trade, protection and economic performance. The first of the three, sponsored by the OECD and directed by Ian Little and Maurice Scott, took more than four years to complete, produced seven book-length studies of developing countries and culminated in the landmark summary volume, Trade and Industry in Sonze Developing Countries [Little et al. (1970)]. Anne Krueger and Jagdish Bhagwati directed the second study under the sponsorship of the National Bureau of Economic Research. It took more than three years to complete and included more than a dozen country studies. Anne Krueger prepared the summary volume, Foreign Trade Regimes and Economic Development: Liberalisation Attempts and Consequences [Krueger (1978)]. The last of the three, and perhaps most ambitious of all, was sponsored by the World Bank and appeared in five volumes under the generic heading, Liberalising Foreign Trade. The study took more than five years to reach completion and covered eighteen developing countries. A summary was prepared by the three study directors, Demetris Papageorgiou, Armeane Choksi and Michael Michaely, under the title, Liberalising Foreign Trade in Developing Countries: The Lessons of Experience [Papageorgiou et al. (1990)]. It is interesting to note that while the principal theme of these studies was trade policy, they did examine other related reforms, including capital markets and foreign investment regulations. All three studies are surveyed by Edwards (1993).

It is unfair to try to capture the richness of these studies in just a few paragraphs. However, while the three covered much the same ground, each focused on a different aspect of the liberalisation process. The OECD study took a snapshot picture of the trade protection system and explored the consequences of observed differences in inter-country and inter-industry rates of effective protection for the growth of industry, trade and the overall economy. The NBER study studied changes in trade policy regimes over time. The authors posited a five-step evolutionary process. In Phase 1, countries impose across-the-board quantitative controls. In Phase 2, uniformity disappears as the system becomes more complex and discriminatory. In Phase 3, liberalisation begins with a few tentative measures. In Phase 4, liberalisation proceeds in earnest. Phase 5 culminates with complete liberalisation from quota restrictions (though not tariff protection).

The World Bank study concentrated on the consequences of liberalisation episodes on various macroeconomic variables and for that reason has the greatest relevance for the present study. The Bank project had several objectives within this concentration. First, were the effects of liberalisation different in the short and long term--for employment levels, growth, the balance of payments and income distribution? Second, were the costs in any time frame--short or long--so great in relation to benefits that liberalisation should not have been undertaken? Some of the other study objectives are suggestive about the appropriate types of investment reforms. For example, is there an optimal sequencing of policies in the liberalisation process--i.e, quota liberalisation first before all others? Was fiscal contraction a necessary adjunct to trade liberalisation? Were there common factors that explained liberalisation failures? Each of these is examined below after a general review of the findings of the study, which relies heavily on [Papageorgiou et al. (1990)].

The Bank study covered 19 countries, which together had 36 different episodes of liberalisation. Episodes were classified according to whether they were "weak" (e.g. small reductions in tariffs or only a small number of sectors liberalised) or "strong". The authors measured performance outcome according to whether liberalisation was sustained, partially sustained or completely reversed. Of the 36 episodes, 24 were either fully or partially sustained; 12 ended in failure. These 36 episodes were divided about evenly between weak and strong. One interesting finding was that both weak and strong liberalisations produced sustained liberalisations. Interestingly, a country's failure in one episode was not a good predictor of failure in subsequent episodes. Of thirteen countries that experienced an initial failure, eleven made a subsequent attempt, of which ten were successful to some degree (sustained fully or partially).

One of the unique contributions of the study was to examine the sequencing of trade policy reforms. Was it better, for example, to relax quantitative restrictions before reducing tariffs? The answer revealed by the study was an unambiguous "yes". The failure rate for liberalisations that did not first relax quantitative restrictions was almost 90 percent, the same as the success rate of those that did. Thus, care in planning the sequencing of policy liberalisations appears important.

Another finding was that the macroeconomic policy adopted by liberalising countries mattered. The study found that "expansionary fiscal and monetary policies are the single most important cause of a reversal of trade reforms" [Papageorgiou et al. (1990), p. 22]. In other words, governments that liberalised experienced, in many cases, short term reductions in revenues as tariffs were lowered. The rush to import worsened the current account of the balance of payments and in general added to inflationary pressures in the economy. A restrictive monetary and fiscal policy was necessary to offset these initial shocks; in the longer run, it was found that both policies could be gradually relaxed.

The study also looked at the reverse effect: what impact does fiscal policy have on trade reforms? The authors concluded that an expansionary fiscal policy made it difficult for countries to launch trade reforms. In other words, a precondition for trade reform was sound fiscal and monetary policy.

The broad results can be summed up this way:

1. There are no systematic adverse effects from trade liberalisation and the costs of any required adjustments are small.

2. If anything, trade reforms promote competition, stable prices and employment in the long term.

3. Stronger reforms implemented quickly appear to be more sustainable than weak ones.

4. One of the surest guarantors of the smooth implementation of trade liberalisation is a politically stable regime pursuing sound monetary and fiscal policies.

Table 1 summarises these points and offers conjectures on their implications for reforms of investment policy.

One small part of the study was devoted to the study of sequencing trade and capital market reforms, including liberalisation of foreign direct investment. The authors argue that trade should be liberalised before capital markets for three principal reasons. First, the turbulence created by capital market reforms might delay trade reforms. Second, if capital markets are reformed first, capital may flow to the most highly distorted and least efficient industries. Finally, if capital reforms are implemented first, then trade reforms might be endangered by a large and sudden influx of foreign capital. This influx would drive up the exchange rate, hurting exports and increasing the demand for imports. This imbalance would provide an additional woe of policy-makers who would have enough on their plate with trade reforms.

In only four of the thirty-six episodes studied were trade reforms accompanied by capital market reforms. These episodes included the reforms in Argentina, Chile, Uruguay and Israel. In all four cases, trade reforms resulted in large inflows of capital, and in the three Latin American cases, these inflows led to strong currency appreciations. Once the capital inflows stopped, sharp depreciations of the currency followed.

One final part of the study worth noting relates to income distribution. Because of the many factors that contribute to income distribution, the authors found it difficult to devise a theoretical case why trade reforms would necessarily improve income distribution. Trade reforms would unleash many forces, some of which would operate in opposite directions on wages and incomes of the lower income groups. The empirical evidence from the 36 episodes was also mixed. The authors note that since devaluations accompanying trade reforms normally raise incomes of workers in the manufacturing and traded goods sectors relative to incomes in the service sectors, some improvement in income distribution should occur. This assumes that the prior trade policies resulted in an overvalued exchange rate, depressing incomes in the tradeable goods sectors.

LIBERALISATION OF TRADE AND CAPITAL: A CASE STUDY OF TURKEY

One recent evaluation by Tosun Aricanli and Dani Rodrik [Aricanli and Rodrik (1990)] of the 1980-84 Turkish liberalisation confirmed a number of the conclusions emphasised in the World Bank study, but at the same time went further in underscoring the need for policy credibility. In the late 1970s, Turkey faced, as did many other developing countries, a debt crisis of major proportions. In January 1980 Turkey initiated a new exchange rate policy that produced a real devaluation of the Lira of over 50 percent by 1987. In line with the World Bank's findings on the characteristics of successful liberalisations, the government introduced austerity measures that dried up the home market, leaving manufacturers no alternative but to export. The government launched an export promotion campaign, backed by tax rebates and other subsidies. The government reformed public enterprises and announced a plan to privatise public holdings throughout the economy. Various financial liberalisations were put into place, in particular the deregulation of interest rates resulting in positive real interest rates for the first time in many years. In 1984, imports were liberalised by adopting a negative list, although tariffs were increased several times to provide new revenues.

At the same time as the general financial liberalisation, the government began to unwind restrictions that had severely limited foreign direct investment: the approval process was simplified and other bureaucratic impediments removed. And liberalisation efforts were intensified in 1985-86 when, according to Aricanli and Rodrik, all "conceivable disincentives" to foreign investment were eliminated. The result of these reforms was marked success in achieving greater exports and almost total failure in attracting foreign investment. One reason lay in the danger that the Bank study pointed to: simultaneous liberalisations can spell trouble. However, in the case of Turkey, investment reform did not result in exchange-rate-appreciating inflows of capital. No capital flowed in.

Aricanli and Rodrik explained the shortfall in capital inflow this way: "Foreign investors continued to doubt the durability of reforms and the stability of the financial system" [Aricanli and Rodrik (1990), p. 1348]. In other words, policy credibility was not achieved. While the reforms were sustained in the Bank's terminology, they were not viewed as stable by investors. The government tinkered constantly with tariffs, tax rates, controls and other measures trying to fine tune the reforms. The constant tinkering planted seeds of doubt in the minds of foreign investors that no amount of success in other areas, such as exports, could overcome.

INVESTMENT REFORMS AND LONG-TERM ECONOMIC GROWTH

The focus of the research reviewed above was on the liberalisation process and its impact on prices, employment and growth in the short term. All of the studies concluded that liberalisation would improve the allocation of resources and increase the prospects for sustained economic growth but none demonstrated it. It would be hard in empirical research to separate out the many factors that contribute to economic growth and show that liberalisation was linked to one or more of these factors. As noted earlier in this paper, no good empirical measures of investment regimes and investment liberalisation exist, so linking a very qualitative concept to quantitative data on growth is problematical if not chimerical.

Since a direct attack on this question is ruled out, two indirect approaches can be pursued. First, since liberalisation of foreign investment has a high probability of increasing exports, how does long term economic growth benefit from exports'? Second, what does modern growth theory have to say on the issue of capital accumulation and especially foreign investment.

The first question is surveyed by Edwards [Edwards (1993)] and he concludes that while substantial evidence exists to link export growth to economic growth, the issue is far from settled. In particular, Edwards finds that most studies have specified regression equations without the benefit of an underlying model theoretically linking exports to growth. Plainly, exports are a component of demand. Increases in this or any other component will produce higher rates of growth. But the authors surveyed by Edwards feel that exports provide an additional impetus to growth, perhaps stemming from positive externalities. Edwards finds the evidence on the externalities issue mixed and a fruitful area for further research.

The second question goes directly to modern economic growth theory. Paul Romer [Romer (1989)] has provided a theoretical basis for establishing a long-run equilibrium relationship between openness and growth [Edwards (1993), p. 1389]. Romer argues that where firms use capital, labour and a large number of specialised inputs, firms can either engage in production of final goods or devote resources to R&D. R&D efforts will result in a larger availability of intermediate products and a higher marginal product for capital. Freer trade allows firms to specialist in the intermediate products where they have comparative advantage. Other authors, such as Edwards (1993), have devised models, showing that free trade permits greater transfers of technology from developed to developing countries. Very recently, Coe et al. (1995) have shown that even though governments in developing countries may not support R&D themselves, they can still enjoy the products of R&D done in industrial countries through the medium of inter-affiliate transfers from parent to subsidiaries. They found that % developing country's total factor productivity is larger the greater is its foreign R&D capital stock, the more open it is to trade with the industrial countries, and the more educated is its labour force. The foreign R&D capital stock only affects productivity when interacted with the import share. The estimated elasticities suggest that R&D spillovers from the North to the South are significant and substantial". Hejazi and Safarian (1996) used the same approach to show that R&D spillovers flow from FDI.

Although growth models do not directly address the issue of foreign investment, it is not difficult to see how liberalisation of investment regimes can produce parallel results with liberalisation of trade. All of these models depend to some degree on indigenous production of R&D in developing countries. If one adds the assumption that foreign investment carries a greater initial capacity to generate R&D and that innovations are diffused throughout developing countries at a greater rate in the presence of foreign firms, then foreign investment is bound to accelerate the process of attaining higher long run equilibrium growth rates.

What remains to be done, of course, is to put these theories to rigorous empirical tests. However, as Edwards (1993) points out, most empirical testing of theories about growth are based on cross-section data that has inherent limitations when it comes to drawing out inferences about behaviour over time. Some analysis of the effects of foreign investment over prolonged periods of time has begun, but still is in its infancy. Bajo-Rubio et al. (1993, p. 104) found, for example, that "foreign direct investment [was] one of the main factors underlying the strong growth rates experienced by the Spanish economy for the last thirty years". Although Bajo-Rubio and Sosvilla-Rivero do not explore in detail the investment regimes under which foreign direct investment entered Spain during the period studied, it is cleat from their narrative that Spain underwent several liberalisation's that made increasing levels of foreign direct investment possible.

SIMULATING INVESTMENT LIBERALISATION

The second major alternative available to examine the macroeconomic effects of investment liberalisation is simulation through large-scale models. One such simulation model is IC95, a multi-region multi-sector model of the world economy designed to examine the short-run and long-run effects of economic policies [Dee, Geisler and Watts (19961)]. IC95 is a hybrid model based on the SALTER model developed by the Industry Commission in Australia and the Global Trade Analysis Project Model at Purdue University.

All simulation models have their strengths and weaknesses that follow from the way they are structured. For example, the IC95 model does not track the economy through time. Instead, it compares alternative states of the economy at a single point in time. IC95 is a system of non-linear equations that describe the interactions between major regions of the world. IC95 does not permit national exchange rates within each region to vary; instead the region's exchange rate (national rates within the region are fixed) moves against an international standard (such as Special Drawing Rights).

Capital accumulation and international capital mobility in IC95 is based on the treatment of capital in IC95 proposed by McDougall (1993). McDougall's contribution was to add equations that enabled simulations to be run, permitting capital to accumulate out of increased household savings and corporate profits. The essence of capital accumulation is, of course, a change in wealth over time. The McDougall (1993) extension to IC95 involves an artificial step of solving the capital accumulation and mobility equations outside of the model, using certain assumptions about the time paths of the explanatory variables inside the model. The model assumes a representative international financial intermediary. Foreign income recipients are not taxed. To incorporate withholding taxes would require an additional equation that explained the timing and rate of income repatriation. The IC95 model includes tax rates for land, domestic capital and households.

Bora and Guisinger (1996) have used the IC95 model to explore the implications of foreign direct liberalisation in Asia. Bora and Guisinger examine examples of three runs of the IC95 model. Investment liberalisation produces an initial "shock" to the system, in this case represented by foreign direct investment raising the stock of host country capital. This increased capital stock produces higher real wages and higher returns to land. In order to capture the strength of linkages between investment and trade, the model was run using two different data sets, both benchmarked on 1992 data. First, a 1992 trade data set was used without updates reflecting Uruguay Round or the North American Free Trade Area effects. The second data set updates the 1992 data with these new developments. The purpose of the two runs is to give us an insight into the complementarity of trade and investment liberalisation.

The results tend to parallel the findings of studies of trade liberalisation. The real values of exports and imports increase, though for exports at a faster rate, sometimes significantly so. Because of the closed nature of the model, there are no balance of payments effects because any trade surplus is offset by changes in net foreign income and net capital inflows. Consumer prices fall or remain unchanged ill most countries, though in a few countries inflation increases. The IC95 model does not allow for fiscal effects because government expenditures are always met by revenues. There is a terms of trade effect associated with the capital transfer, as prices in the capital-importing countries fall, but rise in capital-exporting countries. When the size of the capital transfers increases, these features become more pronounced. In sum, foreign investment liberalisation in the IC95 model is not a painless procedure but the pain seems limited and low in relation to benefits.

SHOULD PAKISTAN LIBERALISE?

Nothing in the preceding review supports the notion that investment liberalisation is either destabilising or anti-growth. Pakistan should experience few costs and considerable benefits from continued liberalisation. This would entail removing the remaining performance requirements, such as the deletion requirements in the automobile industry, and freeing investment in the service sector. The government has removed the major hurdles placed in the way of foreign investors, such as negative lists. Removing the few remaining barriers would put Pakistan in the elite club of developing countries that have adopted the free movement of direct investment.

Now is the time for completing the liberalisation of foreign direct investment. A poll of Asian executives conducted by the Far Eastern Economic Review in August, 1997 found that 92 percent of respondents rated India a better location for investment than Pakistan. The Board of Investment has set a target of a US$3 billion inflow of foreign direct investment in the year 2000, a substantial increase over the roughly US$1 billion received in 1995-96. To achieve this goal, Pakistan must step up its liberalisation process. Pakistan is currently negotiating a bilateral investment treaty with the United States. Now is certainly the time to move toward complete liberalisation.

Complete liberalisation would also permit Pakistan to take a leadership position in forging a multilateral agreement on investment. A leadership position on such an agreement would signal to investors around the world the government's commitment to the free movement of capital and help Pakistan achieve the type of policy credibility that Aricanli and Rodrik argued that Turkey lacked.

The OECD's purpose in proposing the MAI outside of the WTO framework is to ensure that the investment disciplines negotiated would be of the same high standard that OECD member countries enjoy. Negotiations among OECD members have reached a standstill, for a variety of reasons, not the least of which are uncertainties on how to integrate environmental and labour concerns into the agreement. The OECD believes that, once negotiated, the agreement will appeal to developing countries, such as Pakistan, because of the large increases in foreign direct investment that would follow its adoption.

Foreign direct investment is no panacea and will not solve Pakistan's resource gap by itself. Even if additional FDI inflows were to add only a small fraction to the growth rate, this is no reason to ignore or belittle its contribution. If Pakistan were to find a dozen such small measures, each raising the growth rate by. 1 of one percent, a full percentage point could be added to per capita growth.

If Pakistan takes the next and final step to complete its liberalisation, why not sign the MAI or, if not the MAI, become a proponent in the WTO for a similar multilateral agreement? I see only benefits and few costs for Pakistan.

Comments

What are the macroeconomic effects of foreign direct investment (FDI) liberalisation on developing countries that have yet to substantially and fully liberalise? Dr Guisinger's response is that there is lack of empirical evidence to answer this question. He goes on to survey the substantial evidence accumulated from OECD, NBER, and World Bank studies on trade liberalisation which generally show favourable effects on growth. He also looks at a case study of Turkey's attempts at liberalisation of trade and investment. Finally, he informs us about the favourable effects of foreign direct investment liberalisation in simulations that he did for Asia using IC95, a multi-region multi-sector model. Based on the favourable growth experience of trade liberalisation for most countries and his simulation experiments with the IC95 model, Dr Guisinger concludes that Pakistan should experience few costs and considerable benefits from continued investment liberalisation.

Dr Guisinger asserts that one of the major reasons for the lack of empirical studies of foreign direct liberalisation has been the lack of a comprehensive quantitative measure of the restrictiveness of investment regimes. He is emphatic in making this point. However, a look at the international corporate finance literature reveals that there are several indexes of country risk available. These are comprehensive indexes which incorporate economic, financial, and political risks in various countries. Given that these measures of country risk are employed routinely by multinational companies in their capital budgeting exercises involving foreign direct investment implies their immediate relevance. International banks have their own indexes of country risk based on political and financial risk factors. One of the assignments that I gave my graduate students regularly in my international corporate finance courses was to come up with comprehensive quantitative measures of country risk.

The case study on Turkey is interesting and perhaps instructive for Pakistan. In spite of an intensification of foreign investment liberalisation by Turkey, in 1985-86, which involved the elimination of all "conceivable disincentives" to foreign investment, no capital flowed in. This reminds me of the saying, "suppose you gave a party and no one came". The reason attributed to this lack of FDI in Turkey was that the policy measures lacked credibility, given that there was constant tinkering with tariffs, tax rates, controls, and other measures. At present, there is a great deal of talk in Pakistan about FDI liberalisation and what a great place Pakistan is for FDI and how FDI will solve all our economic problems. However, a poll of Asian Executives conducted by the Far Eastern Economic Review in August 1997 found that 92 percent of the respondents rated India as a better location for investment than Pakistan.

I think efforts are misplaced in looking at the effects of FDI liberalisation on growth in Pakistan. It is not as if MNCs are waiting at the border to rush in with FDI once we decide that the effects will be positive. The questions that we should be answering objectively are: Why will foreign capital flow into Pakistan? Where will it come from? When will this party really begin? I agree with Parvez Hasan and others who stress that mobilisation of domestic savings is more important than relying on FDI to solve all our economic problems.

It is useful to remind ourselves of some elementary principles of international corporate finance. FDI moves to countries which offer the highest country risk- and exchange risk-adjusted rates of returns. It flows to high growth economies. There are other less risky alternatives open to MNCs than setting up shop in the host country. These are exporting, licensing a host country enterprise, joint ventures, etc. There is no consensus on why FDI happens but we can learn from case studies of MNCs and other countries. It is high time that we development economists looked at the literature on FDI in international corporate finance.

Talking again about indicators, it would be silly to expect FDI in Pakistan when domestic investment languishes, when property rights are not secure, and when capital flight occurs. There are many Pakistani expatriates in the Middle East and the U.S.A. who are willing and able to invest in Pakistan. They even have a comparative advantage being familiar with the local language, customs, and modes of doing business. It would be instructive to know why they hesitate.

The paper talks about the favourable factors that helped in trade liberalisation. These are political stability, sustainability of liberalisation, economic fundamentals being correct, competition, stronger reforms, and policy sequencing and credibility. If all these factors were in place, I believe that Pakistan would be on a course of self-sustaining high economic growth. In these circumstances, FDI would take care of itself and we would not have to worry about why FDI is or is not happening or whether its effects are positive.

Nasir M. Khilji

U.S. Bureau of the Census & Economic Adviser, U.S./Saudi Arabian Joint Commission on Economic Cooperation

(JECOR),Riyadh.

REFERENCES

Aricanli, Tosun, and Dani Rodrik (1990) An Overview of Turkey's Experience with Economic Liberalisation and Structural Adjustment. World Development 18:10 1343-1350.

Bajo-Rubio, Oscar, and Simon Sosvilla-Rivero (1993) An Econometric Analysis of Foreign Direct Investment in Spain, 1964-89. 104-120.

Baldwin, Robert (1992) Measurable Dynamic Gains from Trade. Journal o/ Political Economy 100:1 162-174.

Bond, Eric, and Stephen Guisinger (1985) Investment Incentives as Tariff Substitutes: A Comprehensive Measure of Protection. Review of Economics and Statistics February.

Bora, Bijit, and Stephen Guisinger (1996) Simulating Investment Liberalisation: The Case of Southeast Asia. (Mimeographed.)

Coe, David, Elhanan Helpman, and Alexander Hoffmaister (1995) North-South R&D Spillovers. (Discussion paper of the Centre for Economic Policy Research, 1133.)

Dee, P., C. Geisler, and G. Watts (1996) The Impact of APEC's Free Trade Commitment. Canberra: Industry Commission.

De Long, Brad, and Lawrence H. Summers (1993) How Strongly Do Developing Economies Benefit from Equipment Investment? (Mimeographed.)

Edwards, Sebastian (1993) Openness, Trade Liberalisation, and Growth in Developing Countries. Journal of Economic Literature 31: 1358-1393.

Guisinger, Stephen (1989) Total Protection: A New Measure of the Impact of Government Interventions. Journal of International Business Studies Summer: 280-295.

Hejazi, W., and Safarian A. Edward (1996) Trade, Investment and U.S. R & D Spillovers. Canadian Institute for Advanced Research. Toronto: Ontario, Canada. (Working Paper 65.)

Krueger, Anne O. (1978) Foreign Trade Regime and Economic Development: Liberalisation Attempts and Consequences. Cambridge, MA: Balinger Pub. Co.

Little, Ian, Tibor Scitovsky, and Maurice Scott (1970) Industry and Trade in Some Developing Countries. London and NY: Oxford University Press.

McDougall, Robert (1993) Incorporating International Capital Mobility in SALTER. Industry Commission, Canberra. (SALTER Working Paper No. 21.)

Papageorgiou, Demetrious, Armeane Choksi, and Michael Michaely (1990) Liberalising Foreign Trade in Developing Countries: The Lessons of Experience. Washington, D.C.: World Bank.

Romer, Paul (1989) Models of Growth. American Economic Review. Vol. 79.

Stephen Guisinger is Professor of International Management at the University of Texas at Dallas, Richardson, Texas, USA.

Table 1
Characteristics of Trade
Reforms and Implications for Investment Reforms

 Trade Reforms

Does the Strength of the Positive Effect on
 Reform Matter? Outcome

Does an Initial Failure No
 Produce More Failures?

Is the Proper Sequencing of Yes
 Policies Important?

Does the Accompanying Yes
 Macroeconomic Policy
 Matter?

What Affect does Reform Ambiguous
 have on the Budgetary
 Position?

Is a Sound Initial Yes
 Macroeconomic Climate
 Important before Launch?

Did Agreements with No Effect one Way or
 Outsiders Play a Role? Other

Political Stability Yes

 Possible Implications for
 Investment Reforms

Does the Strength of the Perhaps Stronger Trade
 Reform Matter? Reforms Better for
 Investment Reforms

Does an Initial Failure No
 Produce More Failures?

Is the Proper Sequencing of Investment Restrictions
 Policies Important? before Investment
 Incentives; Trade
 Policies before
 Investment Regimes

Does the Accompanying Yes
 Macroeconomic Policy
 Matter?

What Affect does Reform Positive
 have on the Budgetary
 Position?

Is a Sound Initial Yes
 Macroeconomic Climate
 Important before Launch?

Did Agreements with Yes, to Initiate and Sustain
 Outsiders Play a Role?

Political Stability Yes