Feasibility of the Tobin Tax

By Rodney Schmidt, International Economic Analysts (Canadian Department of Finance)

" DRAFT " (Hand-Written on the Original Paper – Note of the Transcriber)
January 19, 1995
Introduction

 This paper assesses the revenue potential and incidence of a tax on short term international capital flows, applied by the Group of Seven countries (G7). This is the Tobin Tax, named after Nobel Laureate James Tobin who first made the proposal (Tobin, 1978). There are a number of other general and specialised potential sources of international finance. These include licenses and user fees for a new international automated foreign currency exchange for non-financial institutions, monetary measures such as the sale of IMF gold and new issues of Special Drawing Rights (SDRs), and charges for the use of the global commons. The revenue potential and feasibility of these other schemes will be examined in a separate note.

 The motivation for the most recent surge in interest in the Tobin Tax and other sources of international finance is a non-economic one. This is its potential for generating revenue to finance international public programmes at a time when demand is exploding and funding increasingly difficult to obtain. International public health, education, relief, and refugee programmes, provided largely by the UN group, and orderly international financial development and trading systems, provided by the Bretton Woods institutions, require ongoing financing for delivery of goods within states. In addition, there has recently been an increase in demand for financing for international peacekeeping operations, economies in the transition to a market structure, environmental cleanup and regulation, and de-militarization.

 In the face of these demands, it is apparent that current ways of funding international public goods are not adequate. The present system is based on voluntary and discretionary national contributions, and on ad hoc specialised windows and procedures within multilateral institutions. It suffers from changeable political currents, aid fatigue, and pressure on national budgets. The result is that financing of international public programmes, institutions, and initiatives is often irregular and unreliable, particularly in the UN area. For example, 43 per cent of the UN’s 1993 peacekeeping budget is in arrears (Mendez, 1993).

 The search for sources of international revenue stems from a desire to address these shortcomings. The intent is to duplicate domestic systems of public finance on the international scene. This would replace public international assistance flows with international public finance. The attractive features of such scheme are automaticity, reliability, and continuity, direct treatment of international negative externalities, and large revenue bases. This paper examines the most prominent of several proposed schemes, the Tobin Tax. Members of the United Nations group in particular officially support various formulations of the Tobin Tax, including the ILO (1994), UNDP (1994), and UNCTAD (1994). The Australian government intends to raise the issue at the UN-sponsored World Summit for Social Development to be held in March 1995. The UN secretariat summarised the Australian proposal as follows :

 In addition to augmenting the flow of resources through established channels the Economic and Social Council should be asked to consider new and innovative ideas for generating funds for social development including the possibility of raising finance through international taxation of certain types of international activities such as financial transactions, air travel and telecommunication. (Document for negotiation at the third preparatory committee for the World Summit for Social Development, New York, January 1995, para. #2)

 An earlier paper of ours (Schmidt, 1994) focused on the directly economic interest in the Tobin Tax. That analysis concluded that, in normal times, in the absence of an exchange rate crisis, an international applied Tobin tax would in principle reduce exchange rate volatility and lead to a net improvement in economic welfare. This view motivated Tobin’s original proposal and has the backing of some distinguished economist (1). Our earlier paper also concluded that, if applied unilaterally or to prevent attacks against a specific currency, the tax would be ineffective and damaging to domestic financial markets. This appears to be the reason for dismissal of the Tobin tax by the private but influential Bretton Woods Commission Report on reform of the international monetary system (1994). The IMF also remains firmly opposed to any measure to re-impose controls on cross-border movements (2). In assessing the revenue potential and incidence of the Tobin Tax, the present paper assumes the tax is implemented uniformly throughout the G7 area, as originally envisaged by Tobin. The analysis yields to following conclusions :

The revenue potential of the Tobin Tax is enormous under any realistic assumptions regarding exemptions, contraction of the foreign exchange market, and flight to offshore financial centres (OFCs). No other source would raise so much revenue on a continuing basis. Further studies of the desirability of the tax should take into account the welfare gains to be had from the use of these revenues, in addition to the direct economic impact of the tax.

The revenue base of the Tobin tax consists of very short term, two-way speculative and financial arbitrage transactions in the inter-bank market. Absent an exchange rate crisis situation, these transactions have little association, if any, to movements in underlying or fundamental economic variables. Even a low nominal Tobin tax rate may reduce these transactions, and therefore the revenue base, substantially. However, there would also be a reduction in excessive exchange rate instability, defined as that unrelated to fluctuations in fundamental economic variables.

The immediate burden of the Tobin tax would fall disproportionately on very short term, two-way foreign exchange transactions in the inter-bank market. However, one-way commercial trade and exchange rate risk-hedging transactions in the non-bank foreign exchange market would also be affected through an increase in the bid-ask spread.

If the Tobin tax reduced excessive exchange rate instability, banks and other financial institutions, who now earn large profits by actively speculating in foreign exchange markets, would be clear losers. By the same token, commercial traders and international investors would be major winners from a reduction in exchange rate instability.

Evading the Tobin tax by transferring foreign exchange transactions to OFCs outside the G7 area may be possible in substantial amounts in the long term. Were evasion to prevent a reduction in exchange rate instability, the tax may lead to a net loss in overall economic welfare (neglecting benefits from the use of the still large remaining tax revenues). However, the nature of foreign exchange markets and OFCs, and the scale of the evasion required, suggest that is not likely.

This paper is organised as follows. We begin by outlining the determinants of potential revenue from the Tobin tax, and presenting preliminary estimates. We then examine the revenue base of the Tobin tax, and presenting preliminary estimates. We then examine the revenue base of the Tobin tax, through a review of the structure of the foreign exchange market and the behaviour of exchange rates. Then we assess the likely incidence of the tax. Here we are interested in who ends up paying the tax, and who benefits from its impact on the exchange rate. Finally, we examine the potential for evasion of the tax outside the G7 area. Revenue Estimates For a given tax rate, the factors that determine revenue from the Tobin tax pertain to the tax base, the foreign exchange market in the G7 area. There are two main revenue determining factors involved : the structure of the foreign exchange market, including the potential for evading the tax, and the behaviour of the exchange rate. The latter depends both on macroeconomic policy and exchange rate management and on the rationality principle governing private trade in foreign exchange, whether macroeconomic or private. This section provides preliminary estimates of revenue from the Tobin tax (the method is adopted from Felix, 1994). These estimates are purposely conservative, based crudely on the above considerations. Later sections examine the determinants of potential revenue in more detail, as well as the likely incidence of the tax.

Let us assume a Tobin tax rate of 0.1 per cent. Let us also accept the following estimates. Before the tax, US $ 700 billion is traded on the foreign exchange markets in the G7 area every business day (BIS, 1993, Table IV) : there are 240 trading days in the year. Twenty per cent of these trades are exempted from the tax. Of the remaining trades, twenty per cent evade the tax. The resulting volume of trade falls by 50 per cent due to the impact of the tax. These estimates imply an effective tax base of about US $54 trillion per year, and tax revenues of about US $54 billion per year. This is a significant amount, especially since it accrues every year and will rise with growth in the world trade and investment. In comparison, the regular budget of the United Nations was US $2.7 billion for the 1994-95 fiscal year (General Assembly resolution 48/6/Part I), the budget for international peacekeeping in 1993 was US $3.5 billion, and for voluntary programmes was over US $3billion (Mendez, 1994).

Structure of the Foreign Exchange Market

 The advent of floating exchange rates among the major industrialised countries in the early 1970s, foreign exchange and financial market de-regulation since the late 1970s, and the introduction of new technology are behind four salient developments in international markets foreign exchange. These are :

 1) A rapid rise in the size of the foreign exchange market. In the mid-1980s daily turnover worldwide was about US $150 billion. By 1992 this had increased to US $880 billion. About 88 per cent of trades in 1992 took place in the inter-bank market for foreign exchange ; only about 12 per cent consisted of deals between banks and dealers and non-financial customers. Almost three quarters on inter-bank trade was international (cross-border), while nearly the same share of customer trade was domestic (BIS, 1993, Table III and p.12)

 2) An increase in the geographical concentration of foreign exchange activity. In 1992 the UK (29 per cent), the US (18 per cent), and Japan (12 per cent) accounted for 59 per cent of worldwide trading. This was up slightly (by one percentage point) from the figure for 1989. Moreover, Japan’s share of trading declined slightly (from 15 per cent in 1989) while the shares of the UK and the US rose from three years ago (25 per cent and 17 per cent, respectively). In 1992 Singapore accounted for a further 7 per cent of worldwide foreign exchange trading, with Switzerland and Hong Kong close behind at 6 per cent each (BIS, 1993, Table IV).

 3) A rapid rise in the use of new and sophisticated foreign exchange instruments in the inter-bank market (BIS, 1993, Table V, and PP. 1611). In 1992 the spot market, consisting of exchanges of two currencies that are settled within two business days, still accounted for the largest share of all activity, at 49 per cent. Eighty-five per sent of spot trading was between banks, and most involved the US dollar. However, growth of the spot market, at 13 per cent, was relatively low, and the share of the spot market in total activity was down from 58 per cent in 1989.

 The outright forward market, which is similar to the spot market, except that deals are settled more than two days hence, grew by 60 per cent. However, its market share was still only 6 per cent in 1992. The outright forward market is suitable for hedging commercial risk. About 50 per cent of this market is with customers. Most trades are local and many involve domestic currency.

 The swap market consisting largely of agreements to exchange two currencies at one rate spot, and to reverse the exchange to another rate forward, grew by 56 per cent and increased its market share by 18 percentage points to 40 per cent between 1989 and 1992. The swap market is suitable for hedging financial risk and engaging in covered interest arbitrage. Eighty-six per cent of this market is between banks, more than half of it is international, and most of it involves the US dollar.

 The futures market consists of standardized contracts for delivery of a currency at a date more than two days in the future. Currency options give the purchaser the right, and not the obligation, to buy or sell a set amount of currency at a set rate in the future. These two markets grew at a pace of 60 per cent and 124 per cent respectively between 1989 and 1992. However, their combined market share in 1992 was only 5 per cent. These are almost entirely inter-bank markets, involve the US dollar, and are concentrated in the US and the UK.

 4) Short average maturities of forward instruments. Two third of all outright forward and swap transactions have a maturity of no more than seven days ; only one per cent have a maturity of more than one year (BIS, 1993, p.19).

 This structural features of the global foreign exchange market provide evidence on the nature of the market, the behavior of the exchange transactions (discussed further in the next section), the potential to evade the Tobin tax and the incidence of the tax (both treated more extensively in later sections). In particular, the dynamics of foreign exchange transactions (the base of the Tobin tax) and the exchange rate are dominated by events in the inter-bank market. This market is very large, highly concentrated by geography, currency, and, to an increasing extent, instrument, and has a very short term transaction horizon.

Behavior of the Exchange Rate

 The behavior of the exchange rate is relevant to a feasibility study of the Tobin tax for two reasons. First, the exchange rate itself determines the quantity of foreign exchange transactions, and hence, the tax base. An unstable exchange rate creates ample opportunities for profitable speculation (in the spot market and options markets) and covered interest arbitrage (in the swap market). This is especially so for banks and other financial institutions, whose primary business is to deal in foreign exchange, and who therefore enjoy an international advantage over non-financial customers, whose participation in the market is derived from real trade and investment activity.

A controversial feature of the new shape of the financial system is that the bulk of its participants now have a vested interest in instability. This is because the advent of high-technology dealing rooms has raised the level of fixed costs. High fixed costs imply a high turnover is required for profitability to be achieved. High turnover tends to occur only when markets are volatile. The analysts at Salomon Brothers put it clearly : "Logically, the most destabilizing environment for an institutional house is a relatively straight rate environment" (Walmaley, 1993, p.13)

 A special case of the influence of (expected) exchange rate behavior on foreign exchange transactions is related to intervention by central banks in foreign exchange markets, and by overall exchange rate policy. In particular, attempts by central banks to defend exchange rates which are clearly out of line with fundamental variables (defined below), as occurred recently in the European Monetary System, lead to a large one-way expected movements in exchange rates. These virtually unlimited profit opportunities with little or no risk lead to speculative attacks against the exchange rate which, given the size of the market, cannot be resisted. Although revenues from the Tobin tax will peak during such crises, due to a sudden increase in the (speculative) tax base, they will be dwarfed by the transfer of resources from central banks to private speculators that occurs when the exchange rate collapses.

 Second, the exchange rate provides evidence as to the nature of the underlying demand for and supply of foreign exchange, the ultimate tax base. That is, we are interested in what determines the exchange rate. This is essential to any exercise to estimate Tobin tax revenues, since the tax itself will affect demand for and supply of foreign exchange, and thus the exchange rate.

 We have already noted the relatively small size of the non-financial market for foreign exchange (12 percent of the total market). We may safely presume that most of this markets consists of spot transactions in support of real trade and investment activity, and outright forward transactions to hedge exchange rate risk associated with the same underlying economic activity. These one-way transactions take the exchange rate as approximately given (McKinnon, 1979). Hence, the exchange rate is primarily determined in the inter-bank market. We now examine how it is thus determined.

 Theory suggests that long term equilibrium exchange rates are determined by underlying or fundamental macroeconomic variables. These variables include international differences in inflation rates, demand and supply of exports and imports, cross-border interest payments, and persistent capital flows deriving from international differences in saving and investment rates. Long term levels of these variables are determined both by private economic activity and by macroeconomic policy. Theory also suggests that short and medium term exchange rates, to the extent that they deviate from equilibrium rates because of real policy shocks, are determined by interest arbitrage (Dornbusch, 1976), and tend to return to their long term equilibrium rates.

 These theoretical determinants of exchange rates do not perform well in empirical attempts to predict exchange rates (Williamson, 1993). Only about half of the movements in the widely fluctuating US dollar in the 1980s can be explained by fundamental economic variables of interest arbitrage (Koromzay, Llewellyn, and Porter, 1987 ; Krugman, 1985 ; Marris, 1987). More generally, for prediction horizons out to two years, the random walk model of the exchange rate, which posits no influence of fundamental economic variables on the exchange rate, outperforms alternative models based on fundamental variables (Moose and Rogoff, 1983). In addition, the following features of foreign exchange markers are inconsistent with the ‘fundamentalist’ view of exchange rates (Ohno, 1994) :

the variance as well as the average value of exchange rates drifts over time. This means that there are alternating periods of a few months or more in which the exchange rate is calm and relatively stable, and in which it is turbulent and tends to move in one direction the type of news the exchange rate responds to varies over time of the three main or vehicle currencies, the leading one tends to change every several months or few years the center of trading driving exchange rates tends to shift

If the exchange rate behavior is not determined by fundamentals in the short term, what is it determined by ? The answer appears to be extensive use in the foreign exchange market of very short term mechanical trading rules, which may be rational from the point of view of the individual trader, rather than of predicted movements of fundamental economic variables to long term levels, which is rational from a macroeconomic perspective. Examples of such trading devices are automatic stop-loss rules, which, by instructing traders to sell when rates fall below a pre-set level, limit the riskiness of portfolios (Krugman and Taylor, 1992) and chartism, which bases foreign exchange trading on very short term analysis and extrapolation of past price movements (Frankel and Froot, 1986 ; Allen and Taylor, 1990). Both devices shorten trading horizons, increase the short term variability of the exchange rate, and increase the potential for a snowballing effect that can lead to extreme and prolonged exchange rate misalignment (Krugman and Miller, 1992).

 This discussion suggests that the base of the Tobin tax is very short term speculative and arbitrage foreign exchange transactions in the inter-bank market. Paradoxically, as noted in an earlier paper (Schmidt, 1994), the original motivating principle for the tax was to restore the influence of fundamentals in exchange rate determination, by discouraging precisely these kinds of transactions. If the tax were successful, as reflected in a more stable exchange rate, it would greatly reduce the tax base, and hence tax revenues. It would also substantially affect the incidence of the tax, which is the topic of the next section.

 Incidence of the Tobin Tax

 The issue of the incidence of the Tobin tax on participants in the foreign exchange market relates to its direct effect on the market for foreign exchange, its indirect impact on liquidity and transactions costs, as reflected in the bid-ask spread, and its wider impact on macroeconomic activity. We are interested in who ends up paying the tax, and who benefits from the resulting reduction in the instability of the exchange rate.

 Tobin’s original proposal (Tobin, 1978) was for a moderate proportional tax on all realized conversions of one currency into another. Such a tax would affect short term, two-way foreign exchange transactions the most, such as those associated with speculation and financial hedging (covered arbitrage) (see Table 1). Since banks account for the majority of foreign exchange deals of this nature, they would be most directly affected. Customers, whose activity consists largely of trades to support commercial activity, and therefore most of whose transactions are one-way, are least directly affected. Note that there us no maturity on a one-way deal, so, in our example, the annual tax rate is constant at 0.1 per cent.

 Table 1 : Effective Tobin Tax on a Two-Way Transaction (1)
 
. 1 day 1 week 1 month 3 months 1 year 5 years
Effective Tax Rate (annual %) 73  10.4  2.4  0.8  0.2 0.04 
Annual Tax Paid on $24 million Deal   (2)  17.5  2.5 0.6  0.2  0.1  0.0
(1) Based on a nominal tax rate of 0.1 per cent. The horizon of the transaction, motivated by speculation or financial arbitrage, is as indicated in each column.
(2) Average size of a financial swap in the inter-bank market. The figure indicates the amount of tax paid if $24 million were invested for one year and taxed at the effective annual rate.

 In any case, the Tobin tax need not apply to all participants in the foreign exchange market equally. It would seem that in order to be effective in reducing exchange rate instability, the inter-bank market for foreign exchange would have to be taxed. However, it may not be necessary or desirable to tax the non-financial customer market. This would eliminate the direct impact of the tax on commercial activity entirely (but not the indirect effect, as seen below).

 In order to avoid evasion of the tax by the use of new sophisticated instruments, the tax must cover all realized currency conversions obtaining from all foreign exchange instruments, including options when exercised. There should be flexibility to bring new instruments under the tax as they are developed. One issue concerns the treatment of netting arrangements, which, by normally offsetting amounts owed, could significantly reduce the amount of currency actually exchanged in some transactions. But the tax could be made to apply to the gross amounts involved.

 The Tobin tax is therefore remarkably selective in term of its direct impact. It can be applied only to the inter-bank market, and, within that market, has a disproportionate incidence on very short term speculative transactions, and covered interest arbitrage transactions. Note that, given the size of the inter-bank market relative to the customer market, the profitability of the later type of transaction depends to a great extent on underlying inter-bank speculative transactions rather than on commercial hedging pressure. Thus, a tax that raised the cost of interest arbitrage would also substantially reduce the source of profitable arbitrage opportunities. It would therefore not necessarily have a large direct impact on liquidity in customer markets. That is, it may not significantly hinder the ability of financial institutions to deliver foreign exchange market-making services, consisting of arbitrage of interest rates, bid-ask spreads, and maturity structures, to customers.

 The Tobin tax has, however, an indirect liquidity effect on both the inter-bank and customer markets for foreign exchange. This is captured in the response of the bid-ask spread to the tax. There are actually two spreads : a narrow one in the inter-bank market, and a wider one (typically by 200 per cent) in the customer market. The differential in the spread is an indication of the segmentation in the markets, and arises due to the efficiency, information, and credit risk advantages enjoyed by the inter-bank marker. The existing differential in inter-bank and customer bid-ask spreads means that customers already pay more to make spot transactions and hedge exchange rate risk than do banks.

 The bid-ask spread in the inter-bank market depends on two main factors : market liquidity, that is fixed costs and brokerage fees, economies of scale in institutions and in trading centers (the later deriving from the efficiency of the clearing and settlements system), and currency risk. Therefore, on the one hand the Tobin tax may tend to increase the spread if it reduces liquidity (see below for more on the Tobin tax and market liquidity). On the other hand, the Tobin tax may tend to reduce the spread it is reduces exchange rate instability. The net direct impact of the tax on the inter-bank spread is therefore likely to be small. In any case, the Tobin tax itself is analytically equivalent to a transaction cost. It is the intention of the Tobin tax to increase transaction costs (inclusive of the bid-ask spread) and reduce liquidity in the inter-bank foreign exchange market. Therefore, any targeted increase in transaction costs in the inter-bank market can be achieved through a combination of the tax itself and the resulting change in the inter-bank bid-ask spread.

 Of more consequence is the impact of the Tobin tax on the bid-ask spread offered to customers of banks. This spread, and the inter-bank/customer spread differential, is likely to rise with the tax. In particular, since customer participation in the foreign exchange market is derived from underlying real commercial activity, and since that activity is likely to grow over time, due to trade and financial liberalization, demand and supply for foreign exchange by customers is probably inelastic. In the case, banks may be able to pass on a substantial share of the tax to customers, via an increase in the spread differential. The tax will still result in a substantial net loss to banks, though, if it reduces exchange rate instability, but reducing opportunities for profitable speculation and interest arbitrage.

 The Tobin tax will, then, raise the cost to customers of ‘legitimate’ foreign exchange transactions, those that hedge against exchange rate risk and support commercial activity. Nevertheless, if the tax reduce exchange rate instability, the net impact on customers, and on commercial activity, will likely be beneficial. The strategic calculation made by proponents of the tax is that it will reduce exchange rate instability, and that the benefits accruing, from lower uncertainty, lower hedging requirements, lower resource misallocation, and lower instability in key macroeconomic variables in general, will more than compensate for the increased direct cost of participating in the market.

 At bottom, proponents of the Tobin tax predict that a modest tax rate that effectively stabilizes exchange rates will lead to a net increase in overall welfare, including that of investors and traders, whose net hedging costs may well fall. A stable exchange rate can be regarded as a publicly provided hedge against currency risk, financed from tax revenues. Banks will be the clear losers from the tax, due to reduced trading volumes and exchange rate instability.

 Evasion

 Up to this point we have ignored the problem of evasion of the Tobin tax by the transfer of the foreign exchange market to jurisdictions outside the G7 area. This issue is, however, central to an assessment of the feasibility of the tax. Were it possible to avoid the tax in large part, revenues would be substantially reduced, the tax would not stabilize exchange rates, and the cost of commercial hedging, purchased mostly by non-financial customers of banks, whose deals are much less mobile, would rise.

 Resolution of this issue rests on the incentives to evade the tax, on the technical feasibility of relocating such an enormous worldwide foreign exchange market to untaxed jurisdictions, and on the likelihood of bringing new major OFCs under the umbrellas of the tax. Before we address those aspects, we note that, given the existing concentration of the market in the major international trading centers of the G7, and the stunk costs invested in the administration of and infrastructure for the market, the prospects for significant evasion of the tax are much higher in the long term than in the short term. This would give the administrators of the tax room to respond to new developments.

 First, if the initial implementation of the tax effectively stabilizes exchange rates, the incentives facing banks to move the market to OFCs, which would involve significant up front costs, will be greatly reduced. Opportunities for profitable speculating in the inter-bank market depend on exchange rate turbulence. Similarly, opportunities for profitable covered interest arbitrage largely depend, in the current market, on underlying inter-bank speculation. Hence, there is the potential here of falling into either a vicious or a virtuous cycle : in the first case, the Tobin tax initially fails to moderate exchange rate movements, incentives to evade the tax to profit from speculative and arbitrage dealing are high, and a significant share of the foreign exchange market moves to OFCs outside the G7 area, perpetuating the volatility of the market. In the second case, the Tobin tax initially stabilizes the foreign exchanges, incentives to evade the tax are significantly reduces, and the new stability of the market is enhanced.

 Second, given the current size and structure of the foreign exchange market, the problems and costs involved in moving an important share of it to an OFC are not negligible, at least in the short and medium terms (Cassard, 1994). We have already noted the tendency in the last decade for trading to concentrate in the established international centers. This was no doubt greatly facilitated by general financial market de-regulation, but is also a result of the competitive advantage afforded by economies of scale, sophisticated technology, efficient payments and settlements systems, and low political risk. As evidence of this, the rise in use of sophisticated instruments has also been concentrated in the major centers – the US and the UK account for about 60 per cent of the increase in total reported gross forward and swap transactions (BIS, 1993, p.18). There also appears to be a tendency for banks to substitute use of derivatives, available only in the major trading centers, for traditional activity in the inter-bank market, the primary business of OFCs. Cassard (1994) concludes that the traditional comparative advantage of OFCs has been significantly reduced, and, in the current environment, major new offshore foreign exchange markets are unlikely to appear.

 The question remains, however, whether the imposition of a Tobin tax in the G7 area would restore a competitive niche for OFCs. To address this question, it is necessary to understand the nature of OFCs, of which there are three types. The first is the major established international trading center, currently located entirely in the G7 area and accounting for the bulk of foreign exchange trading activity, even during the period of significant foreign exchange and financial market regulation. The second is the regional OFC, of which Singapore is the most prominent. These centers developed in large part because of their proximity to fast growing regional economies, and serve as a conduit for foreign funds flowing in the area. In some cases, such as Singapore, the market is quite sophisticated. The third type is the booking center, which, with a very limited on site physical presence, existed to exploit acknowledgment loopholes in the domestic tax policy of nearby major financial and foreign exchange trading centers. Booking centers largely served the Eurobond market. Hence, currently only the regional OFCs, and in particular Singapore, suggest themselves as a potential destination for foreign exchange markets seeking to evade the Tobin tax.

 Typically, regional and other OFCs, including the Eurocurrency market in London, developed for a range of reasons, not all associated with the provision of a tax haven. These reasons are proximity to a fast growing regional economy, as mentioned above, the depth and liquidity of financial markets relative to the regional alternatives, efficiency of the payments system, quality of the supervisory regime, and, in some cases, secrecy. Here the experience of the Eurocurrency market in London is instructive. This market developed in response to interest rate controls (the Interest Rate Equalization Tax) and onerous reserve requirements and prudential regulation imposed in the US in the 1960s and early 1970s. However, these taxes and controls were generally much more stringent and comprehensive than the envisaged Tobin tax, governing the entire financial market rather than just the foreign exchange market. Moreover, legislation explicitly allowed for tax avoidance by conducting or registering transactions offshore. This was done to permit financial institutions to compete in international markets, while tightly controlling them at home. Use of other OFCs has declined in the past few years as these loopholes have been closed in conjunction with efforts, centered in Basle under the auspices of the Bank for International Settlements, to tighten and coordinate international banking supervision.

 Third, precedents set by the Basle Committee, which establishes international standards for bank regulation and supervision, suggests that it is possible to coordinate banking legislation internationally. Hence, given the large initial sunk costs involved in setting up a sophisticated new foreign exchange market, the possibility of bringing developing OFCs under the tax in future, as well as the heightened political risk of the expropriation outside the G7 area may well act as a deterrent to evasion.

 In sum, while some evasion of the Tobin tax is probably inevitable even in the short term, it seems that wholesale evasion, on the scale necessary to substantially reduce revenues or prevent a reduction in exchange rate instability, is unlikely to occur outside the G7 area. This is particularly true if the tax is set at a low rate. However, substantial evasion in the long run, using newly developed techniques and instruments, and by moving offshore, cannot be entirely ruled out. This is one area in particular that deserves more study, at the microeconomic and institutional level.

 Conclusion

 This paper has made a preliminary broad assessment of the revenue potential and incidence of the Tobin tax, if it were applied uniformly in the G7 area. We conclude that the tax could yield significant revenues, and could place the funding of international programmes and projects on sound principles of international finance. Aside from the benefits obtaining from the international use of Tobin tax revenues, the tax could also significantly reduce hindrances to international trade and long term investment and growth by reducing excessive exchange rate instability. Moreover, the potential to evade the tax, by moving foreign exchange trading centers outside the G7 area, may not be sufficiently large to prevent these outcomes.

 However, two important caveats attend these positive conclusions. First, we have ignored the main obstacle facing the tax : namely gearing agreement among the G7 countries to implement the tax in a coordinated and uniform fashion. To date, none of these governments officially support the Tobin tax, and some, together with the IMF, vigorously oppose any interference in international capital markets. Second, further study at the institutional level of the technical feasibility of imposing and collecting the tax, as well as of the potential to evade the tax by mis-invoicing or hiding transactions and developing new instruments, is required.