Abstract
Several African countries have to increase their tax revenues to finance human and economic development. General consumption taxes, such as VATs, are the preferred instrument for doing so, because they are less detrimental to growth than income taxes. To enable their use, VAT design has to be improved. Currently, many VATs are so riddled with exemptions and zero rates on domestic goods that they resemble extended excise tax systems, while the standard rate is mainly confined to luxury goods. VAT base broadening would not only increase revenue but also reduce the economic distortions and administrative complexities of most taxes.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
In a cautious assessment, Rodrik (2014) expects moderate and steady growth of perhaps as much as 2 % per-capita. He believes that the traditional engines behind rapid growth, structural change, and industrialization still seem to be operating at less than full power.
Resource-rich countries (Cameroon and the Republic of the Congo) are included in the analysis, but only for their tax revenue (without revenue from oil and mining) as a percentage of non-natural resources GDP.
The high tax effort in Lesotho, Namibia, and Swaziland can partly be explained by the fact that the South African Revenue Service (SARS) collects the import and excise duties for these countries under the umbrella of the South African Customs Union (SACU).
In various countries, the VAT is called goods and services tax (GST, for short) but, as is well known, this is just another name for the same type of taxation.
Empirical evidence collected by Arnold et al. (2011), for instance suggests that income taxes reduce the rate of economic growth more than consumption taxes do.
Keen and Mansour (2010) argue that non-resource revenues were essentially stagnant in sub-Saharan countries between 1980 and 2005, although reductions in trade revenue have generally been largely offset by increased revenue from domestic sources. Further, Baunsgaard and Keen (2010) shows that past replacement has been robust with regard to high- and, to a lesser degree, middle-income countries, but that the evidence for low-income countries is flimsy. The authors (p. 573) conclude that their results suggest that perhaps only ‘something like one-sixth to one-quarter of low-income countries has succeeded in substantially replacing lost trade tax revenues’.
The countries that have not yet introduced the VAT are Angola, the Comoros, Eritrea, Liberia, Libya, Nigeria, São Tomé & Príncipe, Somalia, and South Sudan. Although Nigeria calls its consumption tax VAT, in fact the tax is a gross-product type of sales tax, which does not permit a credit for the tax on investment goods and services and provides export refunds, if at all, out of budgetary allocations rather than from gross VAT collections. Further, Egypt is a borderline case, because the taxation of services is not integrated with that of goods. Djibouti’s VAT as a broad-based consumption tax is also suspect, since the threshold is very high (US$278,000) and the indirect tax system is dominated by a multi-rate extended type of excise tax system.
There may be some confusion about the exact date on which some countries introduced the VAT. Various African countries—for instance, Côte d’Ivoire—started with a manufacturers’ sales tax using the tax credit mechanism (essential to a VAT) to eliminate tax on tax or cascading effects. They called these taxes VAT from the beginning, although the wholesale and retail stages were included at a later date.
In 1999, COMESA teamed up with ECOWAS and SADC to form the African Economic Community (AEC). The (ambitious) goals of the AEC are a continent-wide customs union by 2019 and a common market by 2023, followed by an economic and monetary union in 2028.
Instead of higher rates, many countries, particularly those with anglophone taxing traditions, try to improve the progressivity of their indirect tax systems by imposing separate excises on a large number of luxury goods. SACU and IGAD countries are examples of this practice. Cnossen (2004) argues that this is largely window dressing.
The use of ‘standard exemptions’ in EU VAT jargon (see, for instance, OECD 2012) is an utter misnomer, since ‘standard’ implies ‘something established by authority, custom, or general consent as a model or example’ (Merriam-Webster, International Dictionary). Surely, this meaning does not apply to the VAT exemptions in the EU, which are anything but a model or example to emulate, although many countries have copied them. Note that some of the exempt services in the Sixth Directive, particularly gambling, insurance and fee-based financial services, are taxed under the VATs in the SACU countries.
The seminal work on the role of tax handles in economic development is Musgrave (1969).
Ebrill et al. (2001, p. 50) report the following figures on VAT on imports relative to VAT revenues for African countries: Benin 70 %, Burkina Faso 51 %, Cameroon 43 %, Gabon 51 %, Ghana 50 %, Guinea 62 %, Mauritania 66 %, Mauritius 60 %, Togo 68 %, Uganda 58 %, and Zambia 67 %. Although these figures are rather old, there is no reason to assume that current data differ significantly. For Uganda, the International Monetary Fund (2005) concluded that, to a large extent, the VAT and the excises were border taxes since approximately two-thirds of the combined revenue was collected on imported goods and the remainder on a few domestic products manufactured by some 10–12 firms.
Ignoring countries for which revenue data are not available or are incomplete: Burundi, DR Congo, Gambia, Djibouti, and Sudan. The data for a few other countries may also be suspect, particularly if VAT (and excise tax) collections have not been properly separated out from import duty receipts.
Ebrill et al. (2001, Table 4.2) report on multivariate regressions, using a cross-section of (at most) 89 developed and developing countries. The dependent variable in their analysis is the (ln) ratio of VAT revenue to private consumption. They find that this measure is positively related to the (ln) standard rate, the country’s openness (trade as a percentage of GDP) and the literacy rate.
Preferably, final consumption should be aggregated over households, governments, and non-profit organizations. For lack of data, however, the C-efficiencies in this paper do not include government consumption expenditures in the denominator.
Unfortunately, separate data for the policy and compliance gaps in African countries are not available. For EU countries, Keen (2013, p. 423) concludes that policy gaps (exemptions and rate differentiation, including domestic zero-rating) are in almost all cases far larger than compliance gaps, but it is unlikely that this is the case in most African countries.
Similar to Ebrill et al. (2001), Leon Bettendorf (CPB Netherlands Bureau for Economic Policy Analysis), at the author’s request, has done a multivariate regression analysis for 35 of the countries shown in Table 2. He found a significant coefficient for each of the variables (standard rate, exemption, threshold) that represent the VAT structure. First, a rise in the standard rate by 1 % point is seen to decrease C-efficiency by 3.4 % points, presumably because evasion increases. Secondly, efficiency is lower in 20 countries that apply ‘many’ non-standard exemptions (see Table 2) than in other countries, while the opposite is found for six countries that limit the use of non-standard exemptions (denoted by ‘few’ and ‘some’ in Table 2). This result expands on Ebrill et al. (2001) analysis, the data set for which did not contain the extent of exemptions. Thirdly, efficiency is negatively related to the size of the threshold (for selling goods and measured relative to per-capita GDP as in Ebrill et al. (2001)). An increase in the threshold by one standard deviation (0.6 % of GDP per-capita) can be associated with an 8.1 % point lower efficiency. This result, perhaps not surprisingly in view of the relatively high thresholds in most African countries, again contrasts with Ebrill et al. ’s (2001) finding that the threshold never proved significant.
Bettendorf (footnote 19) also related VAT performance to economic development factors and administrative aspects for the countries under review. Thus, a rise in per-capita income of 1 % increases C-efficiency by 0.09 % points (although the share of agriculture in GDP hardly has an effect on C-efficiency), while a significant relationship is found for the GDP share of imports or trade openness. From the variables on administration and compliance, efficiency benefits significantly from a high literacy rate, an effective government, control of corruption, and a low administrative burden on the business sector.
It should be emphasized that the insights gained from this kind of analysis should not be overestimated. Errors in the measurement of consumption data in countries with large subsistence sectors may skew the results, as may errors in the classification of revenues as import duties instead of VAT or vice versa. Delays in, worse, the denial of VAT refunds at export, prevalent in many countries could bias the results significantly, as could presumptive assessments in lieu of proper accounting for VAT on sales and purchases. In short, a great number of large and small errors, intentional, or inadvertent could influence the results. See Martinez-Vazquez and Bird (2011) and Keen (2013) for wide-ranging discussions and analyses of these aspects.
Standard exemptions are not dealt with in this paper. Gambling, property and casualty insurance, and fee-based financial services can readily be taxed under the VAT, as shown by the SACU experience. There is also an emerging consensus that health care and education should be brought into the VAT base, but it is unlikely that this would raise any additional net revenue (although doing so would reduce distortions) if the subsidies for these services would have to be increased.
See Ebrill et al. (2001), on which this discussion draws.
See Ebrill et al. (2001). VAT burden distribution studies are rare outside South Africa but, for Tanzania, Mugoya (1998) argued that the preferential treatment of foodstuffs did not make the country’s VAT progressive, while it greatly eroded its revenue performance. By contrast, Muñoz, S and Cho (2003) found Ethiopia’s VAT to be progressive because of exemptions (especially of in-kind consumption). For the distributive effects of VAT of countries outside Africa, see Table 5.1 in Bird and (Bird and Gendron (2007), p.73).
See Section 11(1)(j) of the South African VAT Act. Originally, only brown bread and maize meal were zero rated, but in 1992 and 1993, following extended public protests, various other essential foodstuffs were added to the zero-rate list.
For an analysis of the Commission’s findings and information on the situation in other countries, see Cnossen (2004), who also points out that regressivity/progressivity should be considered for the tax system as a whole, not for VAT in isolation. Of further interest is that Botes (2001) found that zero-rating actually made the South African VAT a little more regressive.
Here and elsewhere, it is assumed that changes in the VAT’s rate or coverage do not involve behavioral changes.
Keen (2008) argues that agricultural inputs should be taxed as a way to secure some revenue from the sector, which tends not to be subject to the income tax either.
But for arguments to treat tourism leniently, see Ebrill et al. (2001, p. 73).
See Ebrill et al. (2001, p. 70), who note that the UEMOA Council of Ministers adopted a directive agreeing on a single-rate VAT (in the 15–20 % range) with limited exemptions (provided in a list of the UEMOA Commission) to be adopted by member countries by 2002.
References
Arnold, J. M., Brys, B., Heady, C., Johansson, Å., Schwellnus, C., & Vartia, L. (2011). Tax policy for economic recovery and growth. The Economic Journal, 121, F59–F80.
Baunsgaard, T., & Keen, M. (2010). Tax revenue and (or?) trade liberalization. Journal of Public Economics, 94(9–10), 563–577.
Botes, M. (2001). Regressivity of VAT—the first decade’s experience in South Africa. International VAT Monitor, 12(5), 237–244.
Bettendorf, L. J. B., & Cnossen, S. (2015). The long shadow of the European VAT, exemplified by the Dutch experience, FinanzArchiv (forthcoming).
Bird, R. M., & Gendron, P.-P. (2007). The VAT in developing and transitional countries. Cambridge: Cambridge University Press.
Cnossen, S. (2003). Is the VAT’s sixth directive becoming an anachronism? European Taxation, 43(12), 434–442.
Cnossen, S. (2004). VAT in South Africa: What kind of rate structure? International VAT Monitor, 15, 19–24.
Ebrill, L., Keen, M., Bodin, J.-P., & Summers, V. (2001). The modern VAT. Washington DC: International Monetary Fund.
Fenochietto, R., & Pessino, C. (2013). Understanding countries’ tax effort. IMF Working Paper WP/13/244.
Go, D. S., Kearney, M., Robinson, S., & Thierfelder, K. (2005). An analysis of South Africa’s value added tax. World Bank Policy Research Paper 3671.
International Monetary Fund (2005). Uganda: Fifth review under the three-year arrangement under the poverty reduction and growth facility, request for Waiver of performance criterion and extension of arrangement, IMF Country Report 05/306.
Katz Commission. (1994). Commission of inquiry into certain aspects of the tax structure of South Africa, Interim Report (1 and 3). Pretoria: Government Printer.
Keen, M. (2008). VAT, tariffs and withholding: border taxes and informality in developing countries. Journal of Public Economics, 92(10–11), 1892–1906.
Keen, M. (2009). What do (and don’t) we know about the value added tax? A review of Richard M. developing and transitional countries. Journal of Economic Literature, 47(1), 159–170.
Keen, M. (2013). The anatomy of the VAT. National Tax Journal, 66(2), 423–446.
Keen, M., & Mansour, M. (2010). Revenue mobilisation in sub-Saharan Africa: challenges from globalisation I—trade reform. Development Policy Review, 28(5), 553–571.
Martinez-Vazquez, J., & Bird, R. M. (2011). Value-added tax: Onward and upward? In E. Albi & J. Martinez-Vazquez (Eds.), The Elgar guide to tax systems. Cheltenham: Edward Elgar.
Mugoya, P. K. D. (1998). Equity and VAT in Tanzania. African Journal of Finance and Management, 7(1), 22–32.
Muñoz, S., & Cho, S. S. (2003). Social impact of a tax reform: The case of Ethiopia, IMF Working Paper 03/232 (November).
Musgrave, R. A. (1969). Fiscal systems. New Haven: Yale University Press.
National Treasury. (2011). The VAT Treatment of Merit Goods and Services. Pretoria: National Treasury of South Africa, Document T16/05.
OECD. (2012). Consumption tax trends 2011: VAT/GST and excise rates, trends and administration issues. Paris: OECD Publishing.
Rodrik, D. (2014). An African growth miracle? NBER Working Paper 20188.
WoldeMariam, A. (2010). VAT revenue productivity for selected fund member countries. Washington DC: Fiscal Affairs Department, International Monetary Fund.
Acknowledgments
Thanks are due to Leon Bettendorf, Ruud de Mooij, Pierre-Pascal Gendron, Cecil Morden, Judith Payne, and two anonymous referees for their helpful comments on the draft of this paper.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Cnossen, S. Mobilizing VAT revenues in African countries. Int Tax Public Finance 22, 1077–1108 (2015). https://doi.org/10.1007/s10797-015-9348-1
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10797-015-9348-1