ABSTRACT
The application of value added taxes (VAT) has expanded during the past couple of decades in developed and developing countries alike, including in Africa, where the number of countries with VAT increased from 2 to 30 in the 1990s (Ebrill et al. 2001). VAT is a tax on final consumption, collected by applying a tax on sales made at each stage of production, where enterprises are generally permitted to offset tax paid on inputs against that collected on sales of output. The corresponding net value represents the amount transferred to government as VAT revenue. Major advantages of VAT are that it avoids the cascade effect of turnover taxes and is viewed widely as a fair and efficient way to raise revenue. While this rise in popularity may also be related to the role of VAT as a
key element in IMF-sponsored fiscal reforms, it has also been advocated in the theoretical literature as an effective tool for ‘reaping the efficiency gains of tariff reform’ while strengthening public finances (Keen and Ligthart 2001: 491). Keen and Ligthart (2001) model a one-for-one increase in domestic consumption tax rates to offset tariff reductions and leave consumer prices unchanged, with a ceteris paribus resultant increase in welfare and public revenue for a small open economy. However, as they recognize, this result does not extend to include non-tradables, intermediate goods or the case of imperfect competition. Their approach also ignores the potential distributional effects caused by the impact of tariff reform on factor incomes. Keen and Ligthart’s (2001) theoretical analysis also disregards some
potentially important developing country constraints on effective tax policy. Burgess and Stern (1993) highlight some of these, which include a large primary sector, segmented and fragmented markets, poor education levels, weak administrative capabilities, pervasive corruption, widespread tax evasion and a vast number of small-scale enterprises, many of which operate outside the tax system. As a consequence, ‘even when broad-based taxes are used, the evidence suggests that in practice revenue is collected from only a fraction of the activity that by statute should be covered’ (Gordon and Li
2005: 2). In particular, the presence of a large informal sector may hinder the potential for offsetting border taxes with domestic taxes. Emran and Stiglitz (2005) develop a model to incorporate this additional
feature which allows for an informal sector, with imperfect substitution of formal and informal goods.1 In their analysis, a generalized uniform reduction in import tariffs reduces the distortion between tradable and non-tradable goods.However, the accompanying revenue-neutral increase in VAT creates distortions between the formal sector, which is subject to VAT, and the informal sector, which is not, with negative welfare effects. They further consider a selective tariff reduction and selective VAT reform and show that under ‘plausible’ conditions welfare is reduced. Although they recognize that VAT collected on imports may moderate this decline, they find this incompatible with the view that VAT is a more efficient tax, advocating the elimination of VAT collected at the border. Again, being a theoretical model, some aspects are ignored, such as the existence of non-tradables, intermediate goods, smuggling and the difference in administrative costs between applying border taxes and domestic taxes, although they claim that these are likely to add weight to their results. Empirically, the behaviour and impact of VAT will depend on a variety of
country-specific factors. How a VAT system is applied may vary in terms of the extent of use of exemptions, of zero-rating and the size threshold above which enterprises are required to register for and comply with VAT.2 The importance of the informal sector in economic activity and administrative capacity will clearly also impact on how effective VAT is at fairly and efficiently raising revenue, and how its incidence compares with other taxes. The above considerations highlight the importance of country studies.3
