In many countries the proliferation of tax incentives has led to opaque corporate tax structures and many unanticipated tax loopholes. The revenue loss and distortions in resource allocation resulting from the unintended interactions of the various credits and exemptions have led many countries' fiscal authorities to consider and in some case to introduce minimum corporate taxes. These are sometimes linked to profits, but more often to assets, which are less manipulable.

In Discussion Paper No. 684, Antonio Estache and Research Fellow Sweder van Wijnbergen assess the expected burden of such a `Minimum Asset Tax' (MAT), using an option pricing approach as an alternative to the standard KingFullerton methodology to consider the link between assessed asset value and asset income and the impact of rate-of-return uncertainty. This yields a measure of the value of a minimum tax to a government that faces uncertainty about its revenue due to the proliferation of tax incentives.

The authors apply their model to Brazilian data for the period 1981-8 and report four main conclusions. First, uncertainty must play an explicit role in the evaluation of MAT proposals. Where the option characteristics of the corporate tax dominate the impact of tax provisions on the Marginal Effective Rate of Tax (MERT) under full certainty, increased uncertainty significantly raises the corporate tax burden. Second, the simple tax code of the MAT and its marginal effect on the MERT provide a good short- cut to a comprehensive tax reform, whose revenue effects appear to be substantial in Brazil. This suggests that macroeconomic uncertainty may have much greater effects on rate-of-return uncertainty than the introduction of a MAT; if a MAT can reduce macroeconomic uncertainty by reducing fiscal imbalances, this indirect negative effect on the MERT may more then offset the direct positive effect for a given initial level of uncertainty. Third, the MAT does not reduce sectoral distortions because of the high variance of capital intensity across sectors; the standard deviation of the MERT with the MAT is higher than that without it. Finally, although high variance leads to a higher marginal impact of the MAT, it does not hit high-risk firms harder. These also tend to be high rate-of-return firms, which reduces the impact of a MAT, and this effect dominates in the sample. Concerns that the MAT will discriminate against the most innovative but riskiest firms therefore seem unwarranted.

Evaluating the Minimum Asset Tax on Corporations: An Option Pricing Approach
Antonio Estache and Sweder van Wijnbergen

Discussion Paper No. 684, June 1992 (AM)