Use the free Adobe Acrobat Reader to view a pdf file of this article
A Trade Policy Perspective
on Capital Controls
Capital is tradable in the same way many goods and services are: it can be imported or exported at prices that reflect international demand and supply. Much trade analysis is thus valid for capital movements, and experience with trade restrictions provides insights into how different capital controls might work.
FREE TRADE—whether in goods, services, or capital—is usually the best policy. If a country is unable to free capital movements because of a weak policy environment or concerns about investor behavior, it must exercise great care in choosing among the various types of capital controls. Experience with trade restrictions shows that different instruments have different economic impacts—tariffs and other price-related restrictions do less harm than quantitative restrictions and prohibitions—and this is likely to be true for restrictions on capital flows as well as for those on trade in goods and services.
Parallels and differences
Like any good or service, capital flows have a price. But capital flows are intertemporal exchanges—that is, a price has to be paid for using foreign capital in the present and paying it back in the future. The interest rates on loans and bonds, for example, are the periodic prices for incurring foreign debt. The principle of arbitrage—that price differences induce traders to move goods, services, and capital to the markets that offer them the highest returns until international returns are approximately equalized—applies equally to international trade and international finance. Arbitrage is the source of shorter-term capital flows, as investors move their capital to different countries in an effort to benefit from exchange rate differentials between markets, as well as of longer-term capital flows, such as foreign direct investment.
The principle of comparative advantage also applies both to international trade in goods and services and to capital movements. Consider a firm that has a comparative advantage in raising capital in the form of a floating-rate loan in yen because it has accumulated little debt of this type and is able to get a new loan at a low interest rate. If the firm really wants a fixed-rate loan in U.S. dollars, it can seek out a firm that has an advantage in borrowing in U.S. dollars but wants a floating-rate loan denominated in yen, and the two firms can arrange an interest rate swap.
The fit between trade in goods and services and trade in capital is not exact, however. The capital that flows into a country (an import) can leave the country again as an outflow (an export) almost instantaneously. This rarely occurs with finished goods and services. However, imported inputs often leave a country in another form—as intermediate or finished products; the pattern is similar to putting capital to work in a foreign country and repatriating it at a later date.
Another difference is that the price and potential returns of capital can be quite uncertain, whereas the prices of goods and services are usually known. The risks associated with such uncertainty can be low for certain transactions—for example, the purchase of government bonds from countries with little public debt. But they are high for others—such as buying shares in volatile markets. Capital movements that reflect the willingness of investors to take greater risks in exchange for high potential returns are often defined as speculative. Examples of speculation in merchandise trade include some forms of commodity trading and stockpiling or forward purchases or sales (which represent "bets" on future price developments).
Export and import restrictions can be important impediments to both capital flows and trade in goods and services. However, export restrictions are rarely imposed on the latter, whereas a number of countries still have controls on capital exports. In addition, certain arguments that can be made in favor of protectionism in trade have very limited applicability to capital flows. Concepts such as infant industry protection and strategic trade policy have few parallels in policies governing capital movements, although capital controls can be used to protect underdeveloped regulatory systems much the way certain trade restrictions protect infant industries. The drawbacks are similar in both cases: protection reduces the incentive to reform regulatory systems and discourages infant industries from "growing up."
Some of the problems that have surfaced in capital markets—such as "herding" and "asset bubbles"—and that have been cited by proponents of capital controls rarely feature in trade in goods and services. One of the main arguments frequently made in favor of controls is that countries with little transparency and weak regulatory frameworks are likely to experience problems with macroeconomic management and their financial systems in the face of herding behavior (that is, when international investors seem to act en masse, blindly following each other in moving large amounts of short-term capital into and out of countries). The social costs incurred in such situations may outweigh the benefits of free capital movements. This was presumably the reason Malaysia reintroduced capital controls in September 1998.
Free trade and tariff protection
Starting with a bit of theory, consider the classical case of free trade versus import restrictions, as presented in the figure. The figure illustrates the market for capital but could represent any other market for goods and services. In free trade, domestic capital of qd and foreign capital of q* – qd are supplied at the interest rate r*. Free trade, however, does not mean the complete absence of any restrictions on international capital flows. Even countries without such restrictions impose prudential regulations on banks and other financial institutions, such as limitations on open foreign exchange positions or maturity mismatches, to protect the stability of their financial systems.
Assume now that a tariff, t, is introduced on foreign capital inflows. With this tariff, total capital supplied would decline from q* to qt, and the interest rate would rise to rt. Domestic capital supply would increase to qdt, while capital inflows would fall to qt – qdt. Proponents of controls would argue that such a tariff is desirable if it reduces the social costs of inflows as discussed above.
Abstracting from the social costs, what are the welfare implications of the tariff? Transfers to domestic capital owners are represented by the trapezoid a. This is a rent capital owners would not enjoy under a free trade regime. The government's revenues from the tariff are represented by the rectangle c. Welfare losses are represented by the two triangles b and d. However, additional welfare losses are likely to arise from rent-seeking behavior (that is, expenditures made to capture a transfer). Domestic financial institutions have an incentive to lobby for protection to gain the transfer a (for example, under the pretext of raising financial sector stability), and this process has costs as well. The costs resulting from the loss of efficiency and from rent seeking can be very high. The high costs of protectionist practices in "conventional" trade have been amply demonstrated in the literature on political economy.
Tariffs on capital flows could take various forms. They could be levied through non-interest-bearing reserve requirements, whereby interest would accrue to the central bank. (Chile had such requirements for a number of years but abolished them in September 1998.) Tariffs could also be levied as a proportionate tax on capital inflows and outflows. Transaction taxes that aim to discourage short-term flows relative to long-term flows are called "Tobin taxes," after James Tobin, the economist who first proposed them. Taxes and fees are already imposed on financial services and related capital flows in many instances, and an extension of such charges to international capital flows is conceivable. Variable levies to deal with surges in capital inflows or outflows could reduce volatility through "fine-tuned" protection. These levies would need to be applied in such a manner that they temper volatility without insulating the domestic economy from international markets and without raising financing costs during crises. Dual- or multiple-exchange-rate systems whereby different exchange rates are applied to different types of capital movements are also conceivable, but experience with such systems has been unsatisfactory.
Quantitative restrictions versus tariffs
The debate on capital controls also considers quantitative restrictions and prohibitions of capital movements. Such restrictions limit capital flows to an amount X, or they prohibit certain movements entirely. However, quantitative restrictions can also be applied more subtly through licensing systems, limits on the location and number of foreign service suppliers and on the volume of turnover, prohibitions of certain financial instruments, and various other means. Although the discussion below focuses on inflows, the same arguments apply to outflows.
Quantitative restrictions on capital flows are typically much more harmful than tariff-like protection, and import prohibitions are the strongest form of quantitative protection. When comparing the effects of quotas with those of tariffs, one should consider both the initial (static) and the longer-term (dynamic) effects. First, quotas usually stimulate more rent seeking than tariffs. Second, quantitative restrictions insulate domestic capital markets from foreign competition and are likely to discourage innovation.
With respect to rent seeking, the total size of the rent to be captured is much larger with quantitative restrictions (areas a and c in the figure) than with tariff-like protection (area a only). This is because a quota, unless it is auctioned (and it is probably difficult to auction a financial quota), does not result in government revenue but in an additional rent for the involved parties. The latter then have a stronger incentive to engage in rent seeking and are likely to waste more resources doing so.
The additional rent accrues to foreign suppliers of capital if the latter can benefit from the higher rate of return in the protected market. (However, owners of foreign capital may oppose controls despite higher returns because they may see controls on inflows as an indicator that governments would also control outflows when it was opportune to do so.) Bureaucrats in the agency administering controls can gain if part of the quota rent is transferred into their pockets—for example, through corruption. The rent accrues to domestic and foreign financial intermediaries if they can, for example, borrow from international capital markets at international interest rates and lend in the domestic market at the higher domestic rate.
The dynamic costs of quantitative restrictions are typically even more serious. With tariff-like restrictions, inflows of foreign financing can vary over time, whereas quotas limit foreign market penetration and insulate the domestic market from developments in international markets. A country with quotas would therefore not benefit from an influx of capital or lower domestic prices even if international interest rates were to drop or financial innovation were to lower the costs of intermediation. In addition, quantitative restrictions can undermine the development of financial markets and the quality of financial intermediation if protection reduces intermediaries' incentive to innovate and introduce new and better financial products. This, in turn, can undermine economic growth and development over the long term.
It is also worth comparing the likely response by traders to controls in capital markets with reactions to restrictions imposed on "conventional" trade. Protection generates a price difference across markets. If the price difference is large enough, arbitrage through circumvention of controls is likely to take place in any market. Again, this incentive is strongest when prohibitions are applied to capital flows. Controls increase circumvention and smuggling. These can take the form of hiding capital account transactions in the current account through various means, including over- or underinvoicing, exaggerating travel expenses, and exporting cash-filled suitcases.
Although quantitative restrictions may be easy to administer for clearly identifiable products—such as cars or bananas—that physically cross borders, the administration of quantitative restrictions on capital flows is likely to be much more difficult and may require a sizable bureaucracy. Corruption is likely to increase, with government officials "selling" exemptions from prohibitions or quotas. Various studies and anecdotes on foreign exchange rationing and other types of controls in a number of countries confirm that this is what happens.
The debate on capital controls should take into account the lessons learned from experience with international trade policy. It is not a coincidence that tremendous efforts were made to replace quantitative restrictions with tariffs in agriculture and many other sectors during the Uruguay Round negotiations. A similar case can probably be made for capital flows and controls more generally. In other words, if certain types of capital flows (such as short-term lending) are considered "hazardous," tariff-like protection is more desirable than quantitative restrictions or prohibitions, and more thinking may be required on how best to design such price-based restrictions.
In this context, it might be worth comparing the restrictions applied by Chile until September 1998 with those introduced by Malaysia in the same month. In 1991, Chile introduced one-year, unremunerated reserve requirements on foreign loans and subsequently extended these requirements to most types of foreign financing. The tax implicit in these requirements was determined by the duration of the investment, with higher implicit rates for shorter-term financing. This seems to have resulted—at least temporarily—in a change in the maturity composition of foreign inflows away from short-term flows. However, changes in the "errors and omissions" line in the balance of payments and in trade invoicing suggest that some substitution and circumvention took place. By contrast, Malaysia's controls consist largely of authorization requirements and quantitative restrictions. The latter include the prohibition of credit facilities between residents and nonresidents, as well as limits on imports and exports of ringgit and exports of foreign currency. All offshore ringgit had to be repatriated before October 1, 1998, and proceeds from domestic security sales could no longer be repatriated unless the securities had been held for at least one year. In February 1999, the one-year holding requirement was replaced by a system of graduated exit taxes penalizing short-term capital inflows.
Price-based restrictions are clearly preferable. Experience with Malaysia's quantitative restrictions has been very limited but, over time, administrative costs, circumvention, inefficiency, rent seeking, and the disadvantages of insulation from world financial markets are likely to grow. It is not obvious that Malaysia's restrictions were necessary in the first place, and the abolition of remaining quantitative restrictions may well be desirable. If they are not abolished, their early replacement by price-based measures should be considered. But it should not be forgotten that progressive liberalization is desirable in trade in goods, services, and capital for the reasons outlined above. Any type of protection should be no more than a temporary measure that provides time to create the proper policy framework.
Suggestions for further reading:
W. Max Corden, 1974, Trade Policy and Economic Welfare (Oxford, United Kingdom: Clarendon Press).