Measuring the burden of taxation : an index number approach

Quantitative indicators of the tax burden used in political discussions, quite often do not prove very reliable for two reasons; either they are rather crude summary measures or they are rather selectivly chosen benchmark tax load figures. A natural way to improve the situation is to introduce an index number concept. The paper deals with index numbers for commodity taxes first. Fix weight commodity tax indices are defined and calculated numerically for Austria in close relation to official statistical price index analogues. To overcome the bias of Laspeyres indices economic tax burden indices are defined, which may be extended to include the deadweight loss of distorting consumer taxes as a burden element. Direct taxes may be included in both statistical and economic household outlay indices, which also are closely related to the well known price indices. Nevertheless many convenient properties do not hold any more due to endogenous labour supply, distortions caused by the income tax code, and progressive tax schedules. Although the economic approach tends to relate the tax index number concept to applied general equilibrium models, the empirical relevance of tax burden indices is demonstrated by calculating and interpreting monthly time series for Austria.