This paper examines the income distributional impact of different VAT schemes on the urban and rural household groups in Bangladesh. It argues that, among the existing approaches to evaluating the nonmarginal tax reform proposals, the Computable General Equilibrium modeling approach is inappropriate for Bangladesh, both on analytical and empirical grounds. An alternative approach, which is based on the “equivalent variation” measure and focuses on the household and welfare side, is adopted to identify the losers and gainers of a policy change in a heterogeneous population.
This approach involves estimating parameters of the Linear Expenditure System for each of the 12 urban and rural household groups (based on their levels of per capita monthly expenditure). The pattern of the results shows that a single, uniform value-added tax (VAT) rate for all commodities, where the rate is chosen such that it has a revenue-neutral effect, is very regressive (relative to the current state) in its impact on the income of different households. In general, richer households gain while the poorer households in both urban and rural areas (the majority of the population) lose.
This paper also explores an alternative package consisting of a basic rate of VAT with exemptions for certain foodstuffs, plus additional excise taxes on tobacco, commercial energy, and sugar. The choice of commodity groups for exemptions and additional excises was facilitated by the estimation of distributional characteristics of the goods discussed in the text. This analysis permitted the identification of commodity groups for VAT exemption and for the imposition of excises.
The welfare consequences of the alternative package is found to be much less regressive relative to the uniform, proportional VAT case. It is therefore likely to be more acceptable to the general public and to policymakers.