Raising Revenue with Transaction Taxes in Latin America - Or Is It Better to Tax with the Devil You Know?

In recent years, various Latin American governments have resorted to taxes on bank debits and financial transactions as alternative ways of raising revenue. Considerable interest has developed in understanding the consequences of such reforms. The author constructs a dynamic general equilibrium model to assess the size of distortions and other quantitative implications associated with a transaction tax. The distinctive feature of the model is the non-neutrality property of the tax in the sense that it distorts the structure of relative prices of intermediate transactions, giving rise to tax "pyramidation." The effective tax rate ultimately borne by the economy is shown to depend on the complexity of the transaction structure. Calibrated for Latin America, the model finds that, contrary to existing evidence and conventional wisdom, a transaction tax is not a particularly burdensome levy in terms of economic growth and efficiency costs. The model also shows that if a government can credibly commit itself to an announced two-step reform in which it first uses a transaction tax temporarily and then replaces it with any other conventional tax, this policy will improve economic welfare relative to a tax reform where a consumption tax (or a labor income tax or a capital earnings tax) is exclusively used from the start to raise the required additional revenue.