Global Policy Forum

By Fernando J. Cardim de Carvalho

These remarks will be limited to the debate on the Tobin Tax (TT). TT proposals have been on the table for quite some time already, after the idea was revived by the wave of financial instability initiated with the Mexican crisis of 1994/1995. At this point, most of the debates on the principles and aims involved are largely exhausted. To move beyond the clarification of principles requires the debate on concrete features of such a proposal. The remarks below are presented in this spirit. They address two essential points for the operationalization of a TT: the need to establish the primary aims of the tax, among its possible (but possibly contradictory) goals; the need to asses the feasibility of such a tax under different implementation possibilities.

1. The Need to Set Clear Goals

TT proposals have been made mainly with a view to reach two different goals. The first, that was originally raised by James Tobin in the late 1970s, is to contribute to reduce international financial instability. A second goal would be to raise revenues to support development programs or poverty reduction initiatives. Although there is nothing to prevent a TT to serve both aims, it is necessary to define what the primary goal is in order to design a system capable of maximizing the probability of reaching that goal. In fact, there is an obvious contradiction between the two goals. A TT that could eliminate all kinds of speculative operations and, thus, were fully successful to reach international financial stability would not raise any revenues, because no taxable speculative operation would be made. On the other hand, if the primary aim is to raise revenues, one would design a system in which speculative operations would be realized so one could impose tax on them. Therefore, it is necessary to define what the aims of the TT are so one can set rates and design collection mechanisms that could actually reach the desired results. To examine this question a little deeper, let us quickly examine what could a TT do best.

TT as a Stabilizing Mechanism

To debate the efficiency with a TT can promote financial stabilization let us begin by recognizing two quite different types of financial instability, the first hitting developed economies and the second mainly emerging economies. A TT could perhaps be efficient to inhibit speculative financial operations in developed economies. With some exceptions, in developed economies, financial speculation does tend to concentrate on relatively small profit opportunities, as it is typically the case of the so-called convergence trades, in which hedge funds, for instance, try to explore small deviations from established relations between interest rates in different countries. The margins of profit to be reached in this kind of operation are small and this kind of transaction can be made unprofitable by the imposition of a tax at rates as low as it is proposed in most TT proposals currently under discussion. Of course, a TT would not be effective even in developed countries in the face of larger profit opportunities as those envisaged when a major event is expected to take place, as it was the case of the speculative attack on the Pound Sterling and the Italian Lira in the early 1990s. But a TT could contribute to redirect the energies of speculators like those mentioned above to other activities.

In most of the 1990s, though, international financial instability was generated in the emerging economies segment or in respect to them (with the exception of the LTCM crisis in the fall of 1998, that was, in any case, initiated by the Russian default of July 1998). The Mexican crisis of 1995, the Asian crisis of 1997, the Russian crisis of 1998, the Brazilian crisis of 1999, and, in smaller scale so far, the Turkish crisis of 2000 and the current Argentine crisis, all these episodes could hardly be avoided just by a TT. The reason for that, is, in contrast to what happens in the case of developed economies, interest rates in the case of emerging countries are dominated by risk factors (the country risk) that are frequently larger than the base rate (usually interest rates paid on risk-free US Treasury Bonds) and much more volatile. Country risk spreads are larger because these economies are structurally more vulnerable to external shocks. They are more volatile because this vulnerability, and the options an emerging country has to face it, is constantly reevaluated by financial markets. The risk spread on Argentine debt changed from about 6% over treasuries in the beginning of this year to almost 17% in April. It has been reduced to between 9% and 14% afterwards, but this gives some indication of the size of the change one is referring to in these cases. No feasible TT rate would be capable of really stabilizing such kinds of speculation. This does not mean, of course, that one should throw the arms in despair and let the markets operate freely. Far from it, the problem is not that markets know best, it is that a TT is far too little. Capital controls are the effective way to protect emerging economies from external vulnerability.

TT as a Revenue Source

An alternative aim for a TT is to raise money. In this case, one has to consider two possibilities, that are not really exclusive. A TT can be thought of as a kind of international tax, raised by some supranational organism, or as national taxes administered by the countries where the operations are made. Of course, an international tax would face many difficulties, beginning with questions of sovereignty and leading all the way down to the question of who would collect and distribute the revenues from the tax. In a passing remark during the 1999 financial crisis of Brazil, President Cardoso indicated that a TT could do some good and then added that the IMF could manage it. It is not obvious that strengthening the IMF would be in the minds of 101 out of every 100 proponents of a TT!

A much less complicated alternative would, of course, consist in thinking of a TT in national terms. Many countries could tax financial operations in national terms, including developing countries to use them to finance development programs or poverty reduction initiatives. Of course, this would mean to give up plans for an income redistribution in an international scale, unless individual richer countries decided to transfer the revenues to poorer nations, something that is unlikely to happen given the currently very low propensity of richer countries to offer aid to those nations. A national TT seems still to be the best bet in the short term.

2. The Feasibility of a National TT

One can say for sure that to impose a tax on financial transactions is feasible, at least as a national tax. This is so because some countries have actually been doing this, among them Brazil. In that country, a tax (known by the acronym CPMF, which stands for Provisional Contribution on Financial Transactions) of this kind has been collected continuously since 1997 and has been used to finance public health policies. It responds for about 5% of federal revenues, and a little under 2% of GDP. It taxes every payment that is made in the economy other than those made with cash. This means that it financial transactions like cashing a check destined to purchase goods, or using the check itself to pay for goods are taxes in the same way that payments made to buy stock or other kinds of securities. As Tobin acknowledged in his original proposal, it may not be possible to discriminate those operations. On the other hand, its practically universal incidence allows a TT rate to be set at reasonably small values so as to minimize both allocative effects and the standard deadweights effects any tax is supposed to have on the economy. As a tax, it does have some advantages, like the one of being collected through the banking system, making it unnecessary to rely on an individual's disclosure of one's financial operations. In addition, although this may introduce some further difficulties, it remains possible to single out some operations that one may wish to exempt from the tax (for instance, the payment of wages or the transfer of resources from one bank to another). It is very difficult to evade this tax, particularly in the larger operations, since the only way to avoid paying the tax is by using only cash. Finally, given the differentiated access to banking services in emerging economies, such a tax may even be progressive, since only middle classes and high income groups use intensively the banking system.

3. Conclusion

A TT is feasible, and, if approached as national tax, can be implemented without the delays that would necessarily plague any initiative that depended on an international agreement, especially given the strong opposition of the US to these measures. It can be an effective channel through which the financial sector can contribute to finance social demands for public health, universal education, etc. As a stabilization instrument, particularly in the case of emerging economies, a TT would probably be less effective than the adoption of capital controls, on the one hand, and putting pressure on the IMF to stop it from imposing destructive policies on countries undergoing balance of payments difficulties on the other.