Internationalisation of Financial Service (web publication)

A Trade-Policy Perspective

Activity within the World Trade Organisation (WTO) over the past decade has created a trade-policy dimension for financial services. The WTO engagement, moreover, has engendered debate within countries about the relative virtues of openness and protection in the financial sector -- especially where financial systems are closed or largely closed to foreign participation.

Trade-policy Dimension of Financial Services

Activity within the World Trade Organisation (WTO) over the past decade has created a trade-policy dimension for financial services. The WTO engagement, moreover, has engendered debate within countries about the relative virtues of openness and protection in the financial sector -- especially where financial systems are closed or largely closed to foreign participation. Many of the issues in that debate are familiar from debates about protection versus free trade in other industries.

The infant-industry argument for protection is a good example. Frequently invoked to justify protection of domestic financial-service providers in developing countries, and widely regarded as a valid argument for such protection, it is in fact almost worthless in that role.

Another example is the contention, made by representatives of many industries, that their industry is "special", in ways that disable arguments for openness and justify protection. Financial-service industries do have special characteristics. But these characteristics do not justify protection of domestic producers against foreign competitors.

One special feature of the financial-service sector is that its product is a service. Services are not internationally traded in the same way as goods, and for that reason, domestic financial-service providers (FSPs) face a different form of foreign competition than an industry producing goods, and cannot be protected against foreign competition by border measures, as can producers of goods. Furthermore, the financial sector is heavily regulated, for both fiduciary and monetary-policy purposes. These features are discussed in more detail below. They do not, however, justify protection of domestic FSPs against competition from foreign FSPs.

Another common form of argument for protection starts from the proposition that something or other is amiss in the structure of the economy, and that therefore protection is justified as a corrective. The study of trade policy induces a focus on the "therefore" -- that is, on the alternative means of dealing with the problem asserted. Protection -- at least in the straightforward sense of barriers to foreign competitors -- is never the best alternative.

In the following sections, I elaborate on these and other propositions. I first briefly discuss the nature of trade and protection in the financial-service sector, and then examine some frequently heard arguments for protection of it. I then turn to the problem of matching policies to problems.

Trade and Protection in Financial Services

When a good is traded internationally, the good itself is usually sent from its country of manufacture to the country of the purchaser. That is not typically true of trade in services. Services, being intangible, often cannot be sent from one place to another. The outcome of a service, or the signifier of a service, such as an insurance policy, can be dispatched, but not, usually, the service itself. A newly painted automobile can be sent from one place to another, but a paint job -- the service -- cannot be. Service transactions call for the provider of the service and the receiver of the service to be in proximity with one another. International transactions in services therefore often entail cross-border movement of the means of producing the service, or local establishment of the service provider; not movement of the service itself.

Modern communications technology, though, blurs the notion of "proximity". If a surgeon is to perform an operation, surgeon and patient have to be in the same place. But a customer of a bank may communicate with it entirely by letter, telephone, fax or internet, and in that case, the bank need not be in the same country as its customer. Nor is there any technical reason why a resident of country A should not buy insurance from an insurance company in country B.

Many aspects of an FSP's business, however, demand knowledge of local conditions, and responsiveness of the FSP to changes in them. FSPs themselves seem to have a strong preference for offering their services from local establishments rather than from foreign-based establishments.

Protection in goods and protection in services

Domestic industries producing goods are protected from foreign competition in a variety of ways, but border measures -- tariffs and quotas -- are prominent among them. Border measures, though, are often an ineffective means of protecting domestic service providers from the competition of foreign providers.

This is in part a consequence of the different means by which international transactions are executed: barriers at the border cannot protect local providers of services against competition from local establishments of foreign service providers. Even when cross-border trade in a service is possible, however, protective border measure are often difficult to apply. Services are often "invisible" -- customs officers cannot easily detect a foreign insurance policy or an appendicitis operation purchased abroad as it crosses the frontier. Even when it is technically possible to use border measures to protect local providers of invisible services against imports, therefore, it may be expensive to do so. Under these circumstances, if indigenous providers are to be protected, an alternative to border measures is needed. Often, that alternative instrument is official regulation.

Regulation as Protection in Service Sectors

The case for official regulation of providers of services is widely accepted, and all governments regulate the provision of services in some way. It is widely accepted, moreover, that a proper role of regulation is to define the characteristics of those who may compete in the domestic market. That is usually taken to mean characteristics relevant to provision of the service -- the reserves of banks, for example, or the airworthiness of aircraft. Regulation bearing on the question of who can compete, however, is easily re-focused onto other characteristics: for example, nationality, as when regulation is used to block the entry of foreign service providers who are in all other respects well qualified. Sometimes regulation and/or regulators are "captured" by domestic providers of a service and become means of providing protection for domestic FSPs. Financial services, of course, are especially heavily regulated (Walter, 1987), and the problem of "capture" is picked up below.

Financial Services and the Case for Protection

Services have characteristics that distinguish them from goods, and financial services have characteristics that distinguish them from services in general. None of these differences, however, provide any basis for claims that services or financial services are special in ways that disable arguments for openness and competition (Hindley and Smith, 1984). If there are sound arguments for protecting domestic FSPs against competition from foreign FSPs, they must be found elsewhere.

Arguments for Protection

The notion that foreign domination of the financial system must be avoided at all costs may well be the most influential case in practice for protecting the domestic financial sector against foreign competition. To open the financial sector to international competition, the argument runs, is to risk elimination of local FSPs, leaving the domestic financial system at the mercy of foreigners. International competition must therefore be impeded or banned.

Statements of this case typically fail to spell out the costs or adverse consequences of "foreign domination" of the domestic financial-service sector, and as a consequence, the foreign-domination argument frequently appears as a populist appeal to national sentiment, without intellectual content. More sophisticated presentations of the case for protection of domestic FSPs, however, are buttressed by other arguments, and these provide the focus of this paper.

A number of excellent contributions discuss the case for internationalisation (for example, Levine, 1997 and 1996; Gelb and Sagari, 1990; and Walter, 1987). This paper, however, concentrates on arguments against internationalisation and for protection of local FSPs, asking what content they have. This necessarily entails a taxonomic approach: a compilation of extant arguments, as complete as possible, and discussion of each one.

Broadly speaking, there are two kinds of argument against internationalisation. The first kind, which I shall call "economic", asserts that some economic gain is to be had from protection of domestic FSPs. The other kind, which I shall call "regulatory", is centrally based on the putative needs and difficulties of regulating an internationalised financial sector.

Economic Arguments

Economic arguments for protection of the domestic financial-service sector include three pure economic arguments and a number that are economic in form, but nationalist in substance, in that they depend upon allegedly different responses to economic incentives by domestic as opposed to foreign FSPs. The three economic arguments are:

(a) a putative need to allow time for local FSPs to mature (the infant-industry argument for protection);

(b) that the stronger links with the rest of the world that internationalisation creates will themselves be harmful to the domestic economy: Boyd and Smith, 1992, provide one example of such an argument, Peek and Rosengren, 1997, another; and

(c) the presence in the banking system of large non-performing loans, which, it is said, require policies to create higher profits for existing banks, and therefore call for restrictions on the entry of new banks which would compete these profits away.

Economic-Nationalist Arguments Include:

(d) cream skimming -- foreign FSPs, it is said, will operate only in the most profitable market segments, thus reducing the profits of domestic FSPs and their ability to offer services in less profitable market segments, which they can only do, the argument continues, by cross-subsidising provision of them from profits made elsewhere; and

(e) the access to domestic savings that internationalisation will give to foreign FSPs (for example, sellers of life insurance) -- foreign FSPs, it is said, may prefer to invest these resources abroad rather than in the domestic economy; and, finally

(f) the possibility that lack of commitment to the local economy of foreign FSPs will lead to more rapid capital flight, and that the better international connections of foreign FSPs will facilitate such flight, thus making it more difficult for the local government to control the economy.

Regulatory Arguments

Another set of arguments against internationalisation is based on the putative difficulties of regulating a domestic financial system in which foreign FSP have a substantial presence. For example:

(g) domestic regulators are said to have a limited ability to monitor a more complex financial system -- internationalisation in such circumstances, it is said, will increase systemic risk;

(h) moreover, the demand that local regulators should try to monitor a more complex financial system brought about by internationalisation will loosen their control of local FSPs, and allow domestic FSPs to take excessive risks, at the final expense of the national government, when it acts as their guarantor.

It is also claimed that foreign banks have more experience in operating in sophisticated regulatory environments than local banks, so that:

(i) competition between domestic and foreign banks will be unfair to domestic banks if both are subject to the demands of a sophisticated regulatory system.

Assessment of Economic Arguments

These arguments, in one combination or another, provide the rationale for rejection by governments of internationalisation of the financial-service sectors of their countries. Each is therefore worth detailed consideration. The infant-industry argument, the first to be discussed, is especially important. While practical difficulties in the application of the argument might be conceded, the concept of infant-industry protection is widely regarded as being beyond criticism. The argument, therefore, is a major prop of the case for protection, especially in developing countries.

(a) Encouragement of infant financial-service industries

The starting point of the infant-industry argument is the common situation in which competition from established foreign firms makes local production of a good or service unprofitable. Absence of local production is consistent with a lack of local comparative advantage in the production of that good or service, and that inference is frequently drawn. The infant-industry argument, however, challenges this. It attempts to define circumstances in which local production will not start even though there is a local comparative advantage in that activity. The argument is first discussed in its standard application to industries producing goods, then as it applies to financial services.

The conventional version of the infant-industry argument depends upon learning by doing. In industries that are candidates for infant-industry protection, it is said, local production will give rise to a process of learning by doing that will reduce local costs of production. In such circumstances, the argument continues, restrictions on imports will allow local production to start, and learning by doing to proceed, ultimately revealing the country's inherent comparative advantage.

The infant-industry argument appeals to many people and it is incorporated into the GATT as a case in which tariffs may be used by developing countries. Moreover, the existence of learning-by-doing effects is well established.

Nevertheless, from an analytical standpoint, the argument faces problems. One lies in the question of whether local costs of production will fall fast enough to compensate for the costs of protecting the infant industry in the initial stages, when its costs of production are still high relative to the rest of the world. The proposition by a government that an industry should be protected on infant-industry grounds implies that the government has good grounds for believing that the industryÂ’s costs will fall fast enough to make the initial restrictions on imports socially worthwhile.

A more fundamental problem, however, lies in the information asymmetry that the argument seems to posit. Information about learning-by-doing effects is not a preserve of governments alone: the private sector can know of, and take account of, such effects. Indeed, producers plausibly know more about such effects than members of governments. In the light of that fact, the failure of private-sector producers to enter the industry implies that no member of the private sector believes that costs will fall fast enough to compensate for losses in the initial, high-cost, stages of production.

For a small country, however, the private costs and benefits of starting production are also social costs and benefits. In the absence of protection, the price facing local buyers is unaffected by whether or not there is local production. Hence, in the absence of protection, the welfare of local buyers is the same whether or not there is local production. All of the gains and losses from starting local production go to local producers, so that the private gains and losses of producers are also social gains and losses.

The failure to start production without assistance is therefore implicitly a statement about the flow of social costs and benefits. In particular, it says that in the view of members of the private sector, the discounted present value of the social costs from starting production exceeds the discounted present value of social benefits. That conclusion, though, is exactly opposite to that asserted by a government that says it has a sound case for assisting an infant industry.

A number of stories are available to make the governmentÂ’s view consistent with that of the private sector. Viewed as arguments for protecting the industry against imports, however, all of them have the same weakness: as soon as the reconciling circumstances are set out, it becomes clear that restrictions on imports are not the best way of dealing with them. At best, the infant-industry argument provides a third-best case for restrictions on imports.

Potential producers, for example, might be mistaken: the industry would be profitable without protection. In itself, though, this makes a case for no more than the release by the public sector of its superior knowledge or analysis. Alternatively, the private rate of discount, used by potential producers to assess the value to them of entry into the industry, may be higher than the social rate of discount, used by the government to assess the social value of establishing it. That certainly could create a situation in which the activity appeared to be socially, but not privately, worthwhile. But if imperfection of the capital market is the problem, other investment decisions will also be affected. The appropriate response is a general policy to correct or counteract defects in the capital market, not one like protection, that is specific to the alleged infant industry.

Two other possibilities are intellectually more interesting, but still fail to provide satisfactory grounds for protection. Both are based on rejection of the proposition that private producers know, and can take account of, the costs and benefits of setting up the industry.

The first possibility is that the industry is subject to economies of scale that are external to the firm -- that is, that the costs of firms in the industry depend upon the output of the industry as well as on their own output. If the external economies of scale appear as shifts in the production functions of firms, output of the industry will be smaller than is optimal, and there is a case for a subsidy to output -- but no case for protection against foreign competition. The second is that there is a first-mover disadvantage: the first producer to enter the industry will experience costs as a consequence of being first: later entrants will not bear them. In that event, no firm will want to be the first to enter the industry, and the industry may not appear, even though, if it did, it would be economically beneficial. The problem for the infant-industry argument, viewed as an argument justifying protection against imports, is that once the circumstances creating the first-mover disbenefit are known, a policy to correct them is always superior to a tariff. It is worth noting that each of these arguments requires a multi-firm industry -- neither works if the new industry consists of one firm only.

As an argument for a tariff on imports, the infant-industry argument lurks between two propositions which sit uneasily with one another. The first is that the government knows that the country has a comparative advantage in the production of some good or service even though that advantage does not display itself in current production. The second is that the government does not know why local production fails to appear -- for if it did know this, it could devise a better policy than a tariff on imports. This is a curious combination of knowledge and ignorance, in which the area of ignorance casts doubt on the putative knowledge and gives ground for scepticism that the government has enough information to be confident that costs will fall fast enough to justify protection.

To drive the infant-industry argument to the conclusion that tariffs should be applied to competitive imports is difficult or impossible when the industry in question produces goods. When the infant is financial services, the infant-industry argument is even more difficult to use as a support for restrictions on imports. A close analogue to the basic argument appears in the contention, in a country dominated by foreign banks, that locally-owned banks, given the chance, could thrive. But if they could thrive, some explanation is needed of why they do not thrive; and neither first-mover disbenefits nor external economies seems to provide a plausible answer.

As important, perhaps, in practical terms, is the simple fact that in many developing countries, local FSPs dominate the financial sector as a consequence of measures protecting them from foreign competition, rationalised in infant-industry terms. But if domestic FSPs cannot survive competition with foreign banks or insurers after 40 or 50 years of protection, the infant-industry argument has failed. The proposition that continued protection will have a different outcome might properly be greeted with scepticism -- as Samuel Johnson said of a second marriage, it is a triumph of hope over experience.

(b) Deleterious effects of international links

Perverse Capital Flows

Investment activity is inevitably associated with costs of information of various kinds, transaction costs, and contract-enforcement costs. Boyd and Smith, 1992, argue that if developed countries have created better means of coping with these costs than developing countries, then the net returns to investment in developed countries may be greater than the net returns to investment in developing countries, even though gross returns are higher in developing countries. Internationalisation of financial services by developing countries may therefore create an outflow of capital from poor countries to richer ones.

Viewed as a statement of possibility, this argument seems correct. Its implications for policy, however, are not self evident. Presumably banking services, for example, are part of -- maybe a large part of -- the better means of ameliorating frictions that developed countries are postulated to possess. That the argument can be driven to the conclusion that developing countries should stop banks from developed countries from operating in their territory is therefore unlikely.

Nor does the argument yield a straightforward case for capital controls. Forcing more capital into an inefficient (by assumption) developing-country allocative system yields no presumption of economic benefits sufficient to offset the lower returns received by domestic owners of capital.

An unambiguous conclusion is that the domestic financial system is in need of improvement. That is likely to entail the import of foreign knowhow in one form or another. Allowing establishment of foreign banks is one way of achieving this.

Transmission of Shocks

Peek and Rosengren, 1997, show that banking problems in Japan adversely affected the performance of Japanese branches in the United States, with consequent deleterious effects on the US economy. Generalising (which Peek and Rosengren themselves do not do), an economy that relies on branches of foreign banks for provision of financial services opens itself to the effects of shocks that originate in the countries in which are located the head office of those branches.

Two points are important in the present context. First, the Peek and Rosengren findings refer to branches not subsidiaries. They say (p.504): "... our evidence indicates that the shock to Japanese parent bank capital resulted in substantial loan shrinkage at their U.S. branches, but not at their subsidiaries ...". At most, therefore, their evidence constitutes a basis for an argument against openness of a particular kind (that is, to branches), not against internationalisation of financial services in general.

Second, the responsiveness of branches of foreign banks to the state of their parents will be stabilising under particular circumstances. Domestic banks, for example, might respond to a negative shock to the domestic economy with greater retrenchment than seems appropriate to domestic-located foreign branches.

To generalise from the specific Peek and Rosengren finding about Japanese branches in the US to an argument about internationalisation calls for an analogy with portfolio analysis. It seems likely that such analysis would yield the result that countries should "diversify" -- that they should admit branches from a number of countries, not just from one. Such an analysis might also have something to say about choices by host countries between branches and subsidiaries in terms of the correlation between domestic shocks and shocks in particular foreign countries. It is difficult to see how such an analysis could be driven to the conclusion that a closed financial-services sector will serve the economy better than an open one.

(c) Ameliorating the burden of non-performing loans

This argument is a version of the more general proposition that the state should use its powers of taxation to reduce the burden on banks of their non-performing debt. This version extends that idea by proposing that banks should in effect be allowed to use the stateÂ’s powers of taxation. By protecting domestic banks against foreign competition, the government allows them to charge a higher price for their services than they otherwise could: depositors receive less than they would under competition, and borrowers pay more. The banks are supposed to use these additional profits to avoid the effects of their bad debts.

Stated in this way, the proposal raises the question of whether, if banks are to be recapitalised via a tax-cum-subsidy scheme, this is the best such scheme. It seems unlikely that it is, for two principal reasons:

(i) the "tax" that finances the subsidy is levied on current customers of the banks; and

(ii) payment of the subsidy is not conditional on the subsidy being used to restore the balance sheet of the banks.

So far as the tax element is concerned, to try to protect or save a banking system by taxing those who use its services does not make good sense. Alternatively, stated in more formal economic language, there is some tax (and probably many of them) that costs less, from a social standpoint, than a tax on the use of bank services. Any of these taxes would be preferable to the tax implicitly proposed by this argument against internationalisation.

The subsidy element of the proposal is also inferior to available alternatives. That is because the subsidy is more likely to achieve its intended result -- and to achieve it at less cost -- if its receipt by banks is conditional on their performance and is not merely a vague hope.

Both the tax and the subsidy elements of the tax-cum-subsidy scheme implicit in this argument against internationalisation can be improved upon. It follows that the proposal cannot be defended on economic grounds alone.

(d) Cream skimming

The cream-skimming argument is heard in all service industries threatened with liberalisation. Unless the incumbents are allowed to continue to make high profits on some parts of their operations, it is said, they will be obliged to discontinue other parts, which they maintain only through a sense of patriotic duty and/or social responsibility and/or concern for the conditions of the poor, but could not afford to do without the high profits they earn in the operations facing liberalisation or internationalisation. Therefore liberalisation or internationalisation should be blocked.

A frequently heard cream-skimming argument is that foreign banks, if admitted to the domestic market, will bid lucrative corporate business away from domestic banks, but will not undertake retail banking activity in rural areas. Foreign insurers, it is said, will insure the rich and prosperous firms, but not the poor.

Such arguments propose, in effect, another tax-cum-subsidy scheme. The proposition that liberalisation should not go forward because liberalisation will result in cream skimming is an argument that incumbents should be allowed to "tax" one group of their customers by charging higher prices than would be possible with liberalisation, on the grounds that the increased profits will make it possible for them to provide (that is, subsidise the provision of) an allegedly loss-making service to another group. The power to "tax", however, is not conditional on provision of the allegedly unprofitable service.

Some observers will receive with scepticism the proposition that goodwill or patriotic duty will cause business firms to keep unprofitable operations going. Such observers may accept that different operations yield different levels of profit, but they are likely to suppose that incumbents earn at least a normal rate of profit on the operations they claim they will have to abandon if they can no longer take even higher profits from operations threatened with liberalisation. Such sceptical observers will doubt that liberalisation will lead to the abandonment of those operations.

Even if the premises of the cream-skimming argument were true, however, the conclusion that high profits should be maintained by blocking liberalisation is contestable. The policy problem lies in the contention that, first, the services cannot be provided at a profit, but, second, are socially worthwhile. These two propositions are in evident tension with one another. Neglecting that problem, however, the obvious solution is to subsidise provision of the service and to arrange matters so that payment of the subsidy is conditional upon performance of the service.

The basic reason for the superiority of an explicit tax and subsidy solution is the same as that under the last heading. First, there is no obvious moral or economic case to levy a tax on buyers of the service faced with liberalisation. Second, the subsidy is more likely to achieve its intended result -- and to achieve it at a lower cost -- if its payment is conditional on performance.

(e) Foreign investment of domestic savings

Domestic FSPs, it is said, are more likely than foreign FSPs to invest funds in the domestic economy. This alleged difference in behaviour seems to have only two possible explanations:

(i) domestic FSPs have more information about the local economy, and are therefore able to obtain higher returns on domestic investments than foreign FSPs, and higher returns than the international rate of return; or

(ii) domestic FSPs have a sense of patriotic duty that is lacking in foreign FSPs, and invest domestically even though the domestic rate of return is lower than the international rate of return.

The explanation based on asymmetry of information between domestic and foreign FSPs, however, does not provide a basis for a compelling argument against foreign FSPs. If domestic FSPs earn better-than-international rates of return on their assets because of their better information about the local economy (that is, if (i), above, is correct), then, at least on that account, they should be able to offer better terms to their customers or to enjoy higher profitability; and this runs counter to the idea that when domestic and foreign FSPs compete, foreign FSPs will inevitably win.

Moreover, when foreign FSPs become aware that their lack of local knowledge is causing them to miss profitable local investments, they have an incentive to hire sources of such knowledge. If they succeed in doing so, there is no obvious reason to suppose that even domination of the financial sector by foreign FSPs will lead to neglect of local investments with better-than-international rates of return.

Leaving aside what some will regard as the implausibility of the behaviour posited under (ii), it should be noted that even if the proposition were true, it is far from constituting an unambiguous case against internationalisation. For domestic FSPs to accept lower returns than are available elsewhere by patriotically investing in the domestic economy is in effect to tax domestic providers of capital in order to improve the situation of domestic users of capital -- which in the first instance will in many cases be the government itself, as when an insurance company invests in government bonds. From a social standpoint, this is not an obviously desirable redistribution. Even if it could be shown to be desirable, it would be better achieved by domestic FSPs striving to obtain the best available return on their assets, and the government using open taxation to make the redistribution away from users of financial services.

(f) Capital flight

The claim that foreign FSPs increase the probability of capital flight, and/or increase the volume of capital flight might be founded on a number of propositions. For example:

(i) the sense of patriotism of domestic FSPs prevents them from fully acting upon incentives to move capital out of the domestic economy;

(ii) domestic FSP have more inertia than foreign FSPs, and therefore do not respond so rapidly to incentives;

(iii) foreign FSPs have better foreign connections, and therefore find it easier to shift capital abroad.

When the issue is the maintenance of a fixed exchange rate, "patriotism" may give ambiguous instructions. At minimum, it is necessary to distinguish between the objectives of the national government, and what is best for the national economy. A government defending a fixed exchange rate for reasons of prestige, for example, may be following a course that is disastrous for the economy.

Nevertheless, when there is a risk of capital flight, it is at least conceivable that inert domestic FSPs will serve the economy better than more rapidly acting foreign FSPs. Inertia, though, is unlikely to be confined to capital flight, and it is difficult to think of other circumstances in which inertia may be desirable. An alternative statement of this argument, therefore, is that "domestic FSPs respond slowly to incentives, which is bad for the economy in general, but might be useful if there is a risk of capital flight". The alternative is to have more sprightly FSPs, and government policies that minimise the risk of capital flight.

Essentially the same is true of the claim that foreign FSPs have better foreign connections, and will therefore find it easier to move funds out of the country in times of capital flight. Better foreign connections are likely to be valuable to the economy in many ways. It is difficult to imagine a sensible argument to the effect that inferior foreign connections should be preferred because they will slow capital flight.

Conclusion on economic arguments

The economic arguments against internationalisation are weak. Pursuit of welfare-improving restrictions on international trade in financial services is likely to be a costly distraction from more worthwhile objectives.

Assessment of Regulatory Arguments

Before turning to assessment of arguments for protection based on regulatory difficulties, which to some extent depend upon a vision of an ideal regulatory set-up, it is necessary to take further the earlier discussion of regulation as an instrument of protection. Ideal regulatory set-ups can hardly be discussed without reference to this aspect of actual regulation.

Theories of Regulation

There are two competing theories of government regulation. One is the "public-interest theory" and the other is the "capture theory" (Stigler,1971; Peltzman, 1976).

The public-interest theory jumps from the observation that users of financial services are sometimes misinformed to the conclusion that competition should be restricted by governments or government-appointed regulators. That conclusion, though, relies on the implicit premise that those appointed with instructions to further the public interest will, ipso facto, do just that. But the most casual observation suggests that official regulation brings its own problems, and that these may be worse than those it purports to solve. Discussion of how a regulator trying to serve the public interest might act is a useful way to define the role of regulation. To assume that there is in fact a ready supply of regulators who will do their best to act in the public interest -- and are competent to approach that goal -- is foolish.

Market competition between producers tends to align private and public interests. No such statement can be made about the actions of government-appointed regulators. Government appointment per se provides no incentive to act in the public interest.

The rationale for giving regulators powers to restrict competition is that they will use those powers to improve the position of users of a service. But providers of a service will also see ways to improve on the outcome of competition. In particular, they are likely to prefer the higher incomes yielded by control of entry into their occupation, or restrictions on competition in it. Once created, regulatory powers can as well be used to achieve anti-competitive goals of the providers of a service as to protect buyers of the service from lazy or predatory providers.

Once regulators are given powers to restrict competition, there is likely to be conflict over how those powers are used. In this conflict, providers of a service have several clear advantages. They are likely to be better able to assess the effects of regulatory actions than users of the service. They will usually be fewer in number than users, and this, combined with their greater interest in the conditions of their occupation, makes it easier for them to organise themselves and the expression of their views, and to press for adoption of those views. Finally, effective regulation will often require a knowledge of the regulated industry so detailed as to be found only among practitioners or potential practitioners. Regulators therefore may have an inherent sympathy with those they ostensibly regulate.

The "capture theory" of regulation suggests that members of a regulated industry will often succeed in diverting the powers of regulators to their own purposes. When that happens, regulation will be good for the profits or life style of providers of a service, but bad for users of the service. Users might be better off with no regulation at all than to have captured regulators controlling provision of the service.

Simple and multiple targets for regulators

Members of a regulated industry are unlikely to be alone in seeing uses for regulatory powers other than protection of the interests of users of a service. Members of governments are also likely to see alternative uses for such powers. Regulation becomes even more problematic when regulators are directed by members of governments to use their powers to further such objectives.

An objective such as "maintaining the market share of domestic providers of a service", for example, will often call for actions that conflict with those required by the objective "obtain the best available price-quality combination for users of the service". A regulator directed to pursue both objectives, and allowed to justify his actions in terms of either, has substantial effective freedom. If the performance of a regulator is to be monitored (and if performance is not monitored, the problems discussed above are likely to be severe), multiple objectives make the job very much more difficult. Apart from the question of whether the alternative objectives are appropriate, multiple objectives remove a regulator from effective control.

Designing regulatory systems

The public interest theory of regulation that provides the rationale for much existing regulation faces major difficulties. Even so, it cannot successfully be argued on a priori grounds that economic efficiency requires a complete absence of regulation.

Nevertheless, some simple rules for regulatory systems are discernible. For example, economic theory suggests that:

• regulatory restrictions on market competition between providers of a service or barriers to the entry of new providers require extremely strong justification -- market competition is a valuable means of improving performance.

Moreover, the analysis above suggests two additional rules:

• regulators of service industries are prone to capture by existing providers of the service -- effective systems of regulation will take the possibility into account;

• in part because capture is possible, the actions of regulators must be monitored -- and their actions are easier to monitor when they are given a simple target.

Simplicity and transparency are likely to be hallmarks of an effective regulatory system. Nor does this run counter to the regulatory arguments against internationalisation noted earlier. All of these (the limited ability of local regulators to control a financial system containing foreign FSPs; the possibility that internationalisation causes local regulatory authorities to lose control of local banks; and the greater ability of foreign banks to cope with regulatory requirements) also argue for simple regulatory targets.

There is, however, some obvious tension between these different arguments. Is "the limited ability of local regulators to control a financial system containing foreign FSPs", for example, to be construed as a consequence of the inherently low abilities of regulators, or as a consequence of the design of the regulatory system they must operate? If local regulators are incompetent, the argument that internationalisation may cause them to lose control of the local banks, loses force: why would anyone want incompetent regulators to control banks? And if the regulatory system is badly designed, why not change the design?

The third argument, that local banks will have difficulty in coping with "sophisticated" regulatory requirements, may also ring warning bells in some minds, suggesting to them the question of how well local banks cope with the requirements of the "less sophisticated" regulatory system that is said to be possible in a financial system dominated by local FSPs (and that must be possible for the first and second arguments to have any bite).

These sceptical minds may wonder if what is at stake is a system in which regulators and FSPs are all old friends; in which the wishes of the regulators are conveyed orally over a meal or a cup of tea, rather than written down; and in which conformity with regulation is measured with a friendly elasticity, as is appropriate between chums. They are likely to note that groups of supportive old friends are prone to shared errors of judgement; and that the lack of transparency of such systems, and the broad discretion for regulators that typically goes with them, leave few defences against political interference.

Whatever the doubts of sceptics, such a system may work well in some circumstances. Whether it is good or bad, however, it is threatened by internationalisation. Internationalisation means that regulators must deal with foreign FSPs who are not old cronies, who probably donÂ’t like tea, and who want regulatory decisions in black and white and consistent with law.

Where a system of "informal" regulations is working well, the threat to it posed by internationalisation is an argument against internationalisation. Whether it is a compelling argument, given the absence of support from other arguments, is open to question.

MATCHING POLICIES TO PROBLEMS

By and large, arguments against the internationalisation of financial services lack force. Nevertheless, many governments restrict foreign competition in their financial sectors. Why they do so is an interesting question. But whatever the reason, another way of viewing the issue is as a problem in second-best economics. Given that the financial sector is to be less than fully open, what is the best way of restricting foreign participation?

The general answer is that foreign providers of a service should be allowed free entry into the market, and a subsidy paid to domestic providers, calculated on some measure of their output (Hindley, 1988). Domestic users of the service can then choose from the best price-quality combinations available in the world economy: domestic service providers are supported without forcing upon domestic users an unnecessary deterioration of the price-quality combinations available to them.

Alternatives to subsidies as means of supporting the local industry can be divided into two groups, broadly speaking -- first, taxes on foreign providers and/or their products and, second, quantitative restrictions on their operations. Either of these entails inferior price-quality combinations for local users.

Taxes on the products of foreign FSPs that are not applied to the products of domestic FSPs will almost certainly be passed on to domestic users of financial services. Domestic FSPs will therefore be able to earn higher profits than if foreign FSPs allowed unrestricted entry without taxes, or to hold a larger market share, or some combination of these. But, of course, domestic users of financial services will pay higher prices for financial services -- higher by roughly the amount of the tax, if quality is held constant -- whether they buy from foreign or domestic FSPs. The higher price paid to domestic FSPs may be compared with the subsidy that would be required to provide domestic FSPs with a similar output were entry of foreign FSPs free and untaxed. The tax and higher prices of financial services, however, will retard demand for financial services and development of the domestic market.

Quantitative restrictions on foreign FSPs (for example, on branches or permitted operations or volume of deposits) have little to recommend them other than providing relative certainty to domestic FSPs. Domestic FSPs may have difficulty in maintaining their market position even with a subsidy if the subsidy must be explicit and politically acceptable, and they may have difficulty in maintaining their market position even if their foreign competitors are subject to taxes while they are not. Quantitative restrictions on foreign FSPs, however, guarantee that domestic FSPs will have business even if they are very inefficient. Clearly, however, this advantage for domestic FSPs, and possibly for local politicians, is bought at a very high cost to the economy. Quantity-restricted foreign FSPs have no incentive to compete for market share. They may offer a better-quality product than domestic FSPs, but, if so, they will charge a price such that local users of financial services are roughly indifferent between local and foreign products. If foreign FSPs earn and repatriate high profits, the host economy is likely to be worse off as a consequence of the controls on entry.

Conclusion

To usefully think about policy towards international trade in a sector, it is necessary to have a notion of the objectives the policy is supposed to serve. Some national policies towards international transactions in financial services fail to provide such a notion, and the absence of clearly articulated objectives inhibits thinking about policy. Probably the most valuable thing that could happen in this area would be an effort by governments to explain -- even internally, without public exposure -- the objectives of the restrictions they maintain.

References

Boyd, J.H. and B.D.Smith, 1992 "Intermediation and the equilibrium allocation of investment capital: implications for economic development", Journal of Monetary Economics, December

Foster, Christopher, 1992, Privatization, public ownership, and the regulation of natural monopoly, (Oxford: Blackwell).

Gelb, Alan, and Silvia Sagari, 1990, "Banking" in P.A.Messerlin and K.Sauvant (eds), The Uruguay Round: services in the world economy, (Washington D.C.: the World Bank and the United Nations Centre on Transnational Corporations.

Hindley, Brian and Alasdair Smith, 1984, "Comparative advantage and trade in services", The World Economy, Vol. 7, No. 4, December, pp. 369-389.

Hindley, Brian, 1988, "Service Sector Protection: Considerations for Developing Countries" World Bank Economic Review , May, pp. 205-224.

Key, Sydney J., 1999, "Trade liberalization and prudential regulation: the international framework for financial services", International Affairs, January, pp. 61-75.

Levine, Ross, 1996, "Foreign banks, financial development, and economic growth" in Claude E. Barfield (ed.), 1996, International financial markets, (Washington D.C.: American Enterprise Institute).

1997, "Financial development and economic growth: views and agenda", Journal of Economic Literature, June, pp. 688-726 .

Peek, Joe and Eric S. Rosengren, 1997, "The international transmission of financial shocks: the case of Japan", American Economic Review, September, pp. 495-505.

Peltzman, Sam, 1976, "Towards a More General Theory of Regulation" Journal of Law and Economics, August, pp. 211-240.

Rottenburg, Simon (Ed),1980, Occupational Licensure and Regulation (American Enterprise Institute for Public Policy Research).

Stigler, George J.,1971, "The Theory of Economic Regulation", Bell Journal of Economics, Spring, pp. 359-365.

Walter, Ingo, 1987, "Global Competition in Financial Services", (Washington D.C.: American Enterprise Institute).