I Introduction

Vito Tanzi, M. Yücelik, Peter Griffith, and Carlos Aguirre
Published Date:
October 1981
  • ShareShare
Show Summary Details

This study provides a general statistical framework for assessing the tax systems of the sub-Saharan African countries. It does not analyze the administrative and political reasons for the revenue structures and levels in these countries. It simply presents the statistical results and attempts to link them to some relevant economic characteristics. A companion study (see Part I) provides the analytical assessment of the effective, as well as the statutory, tax systems and attempts to pinpoint the significant weaknesses and strengths of those systems.

A word of caution is in order. Because this survey covers such a large number of very diverse countries, it was impossible to treat all of them in depth in the limited time allotted. This study was limited also to statistical information derived from sources immediately available within the Fund. Fund sources are quite extensive—possibly greater than those available in any other institution—but they are not necessarily sufficient for a country-by-country analysis. Although care has been taken to be accurate in regard to details of fact and classification, it is possible that some inaccuracies remain. Also, in many countries, the most recent year for which data were available was 1978 or even 1977; thus, if some of the tax systems have experienced drastic changes since that time, these would not be reflected in the data.

This study deals with more than 30 countries ranging in size, population, and gross domestic product (GDP) from very large to very small (Tables 1 and 2). In terms of surface area, they range from a country as large as Sudan, with more than 2.5 million square kilometers, to one as small as Seychelles, with only 278 square kilometers. In population they range from 74.6 million for Nigeria to 60,000 for Seychelles. Even in terms of per capita GDP the variation is remarkable, as the group includes countries as poor as Ethiopia, with a reported annual per capita income of approximately $100, to countries such as Ivory Coast with a per capita income of over $900 a year.

Other differences are also considerable and inevitably affect the structure of the tax systems. For example, the share of agriculture in the total economy ranges from a low of about 15 per cent in Liberia and Zambia to a high of about 50 per cent in Ghana, Malawi, and Rwanda, among other countries. Manufacturing in sub-Saharan Africa also varies considerably, with a share in GDP that ranges from about 3 per cent in Lesotho to 26 per cent in Swaziland. Another aspect of the economy that inevitably affects the tax system is its openness, measured by the ratio of imports to GDP (Table 1), which varies from a low of about 5 per cent in Uganda to a high of more than 75 per cent in Botswana, The Gambia, and Seychelles.

These characteristics of an economy affect the tax system: (1) through their influence on the demand for public goods and government services in general, and (2) through their effect on tax bases (in other words, through the “tax handles” available to the respective governments). Therefore, a demand, as well as a supply, consideration relates to the economic structure of these countries. On the one hand, the structure of the economy affects the governments’ demand for tax revenue and, consequently, the tax level; on the other hand, it affects the tax structure by determining the types of taxes that are easiest to levy. An economy that is largely agricultural has less demand (or need) for public services. Such an economy would normally provide less usable tax bases (unless the agricultural output is mainly exported) than an urban economy, which relies substantially on imports and has a large manufacturing sector. Apart from these objective reasons, cultural and political differences are also likely to affect both the level and the structure of taxation. These differences, while important, are difficult to take into consideration in the type of analysis carried out in this study.

    Other Resources Citing This Publication