Estimating revenue from an international foreign exchange transactions
tax (Tobin tax)
Key
Staff Involved
Rodney
Schmidt
Research
Period
July 2005 – December 2007
Output
Published
research report
Tobin’s
tax is conceived as a percentage levy on foreign exchange
transactions. It would be assessed on each conversion
by large foreign exchange traders of domestic currency
into any foreign currency, and collected by the domestic
government or its designated international authority.
Because of the great volume of foreign exchange transactions
in major currencies, Tobin’s tax can potentially
raise a large revenue to finance global public goods
or international development projects, such as the Millennium
Development Goals. A new international tax is attractive
in that it would broaden the tax base rather than raise
tax rates. We need reliable estimates of revenue to
find the appropriate tax rate and understand its impact
on market liquidity, and to plan for a tax administrator
and revenue allocation authority.
Earlier interest in Tobin’s tax identified evasion
as a problem. We have since found a feasible way to
collect the tax, through the globally integrated payments
system used to settle inter-bank foreign exchange transactions
in cash and other financial instruments. Tobin’s
tax is then equivalent to the user fees, subsumed in
the bid-ask spread, that administrators of payments
systems routinely collect on individual financial transactions.
Initial methods to estimate Tobin tax revenue and the way it responds to changes in the tax rate were not convincing, depending simplistically on gross foreign exchange trading volume to the neglect of other contributors to market liquidity. The equivalence of Tobin’s tax to a transaction cost such as user fees suggests a new way to estimate the revenue elasticity, by looking at the behavior of the inter-bank bid-ask spread for major currencies. The spread reflects market liquidity (the greater the liquidity the smaller the spread) and responds to transaction costs and market risks, including foreign exchange, liquidity, credit, and settlement risks. One can estimate statistically the relative importance of each of these components and their interaction for the behavior of the spread and of trading volume. Then one can simulate how the spread and trading volume would respond to Tobin’s tax as a change in foreign exchange transaction costs.
