full employment policy against harmful influence from outside. It would 
have advantages as compared with a direct manipulation of the exchange, 
if it were combined with the Dornbush proposal of not applying the tax to 
the current account. I can not judge the administrative feasibility of 
such a variable tax, but in theory it would be the ideal answer to the 
tyranny of foreign interest policy. This may become clearer in the 
following. 
The effects of the Tobin tax are much more difficult to predict once we 
turn to the movement of the exchange rate in time. 
A great deal of the speculation means holding positions only for a very 
short time within the day and the speculative gains made are small in the 
individual transaction. The proponents of the tax assume that a 
speculator in a bullish market who has paid the tax (we assume that it is 
always the buyer who pays it ) expects to recuperate it and make a gain, 
too, by selling after a short time. The tax will make it very much less 
likely that he has a chance to succeed in a short spell of time so that 
he will either hold the position for a much longer time or will not 
engage himself at all. The tax would thus even out small short term 
movements and induce speculators to hold longer positions. 
This interpretation depends on the assumption that the movements of the 
price net of tax are indpendent of the imposition of the tax. But is that 
plausible? Let us visualize the speculative market as interposed between 
a supply of exchange coming from the exporters and a demand coming from 
the importers of goods. If we start from a state of cleared markets and 
assume that an additional demand from the importers emerges then the 
speculators are supposed to come in and provide the supply on condition 
that they will make a gain; well, if there is the Tobin tax they will ask 
for a gain plus the amount of the tax so that the fluctuations of the 
exchange rate will be made steeper by the tax. To put it differently: If 
the tax would eliminate gains and speculations up to the amount of the 
tax then the demand from the importers might not always find a ready 
supply and they would drive up the price until the speculation would be 
ready to come forward. The importers will produce such gains as are 
necessary to make the speculation function. This may, however, not be 
true if the importers prefer to wait and queue up rather than pay higher 
prices. Similar considerations will apply to a bearish market where the 
buyers will deduct the amount of the tax from their offer and the decline 
of the prices will be reinforced in a cumulative way by the tax. Which of 
the two cases will materialise depends on the bargaining strength of the 
importers and exporters. If the importers can not pay more than a certain 
amount they will cut their price offer to the speculators, if they are 
able to shift the cost to the price of the goods they will bid up the 
exchange rate. We must consider that beside the importers of goods also 
exporters of capital provide an "ultimate demand” in our market. These 
capital exporters will in some cases be in a position to bid up the 
exchange rate very strongly so that they can swallow a lot of tax. 
We may perhaps discern two effects of the tax: First, it will encourage 
speculators to hold long positions, because in a chain of short positions 
the tax will pile up and make it less likely to leave a satisfactory gain 
after some time. Second, the chain of speculative transactions will be 
shorter and the price will move upward or downward more steeply and more 
quickly on account of the tax because of the cumulation which will soon 
make the importer unwilling to pay more or the exporter to accept less. 
Thus we shall get steeper and shorter price movements and less 
transactions in the short run. In this sense the short term speculation 
(say, over the day) would be diminished while the speculation over longer 
periods ( six or eight weeks which is considered by Schulmeister (1988) a 
crucial period for the formation of speculative bubbles ) may quite 
likely increase. These brief considerations can not suffice to decide how