II Level of Taxation

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

The tax systems of the countries of the sub-Saharan region differ in many ways. One of these differences—the relationship of tax revenue to GDP—is discussed in this section. Subsequent sections deal with the structure of these tax systems. Table 3 shows that there is a considerable spread in the ratio of tax revenue to GDP (i.e., in the average tax rate). It ranges from below 10 per cent in countries such as Chad, Ghana, and Uganda to well above 20 per cent in countries such as Botswana, the Congo, Gabon, Ivory Coast, Liberia, Nigeria, Seychelles, Somalia, Swaziland, and Zambia. In a few of these countries, the ratio exceeds 25 per cent—a high level for developing countries.

While interesting, the mechanical relationship between tax revenue and GDP is not a satisfactory indicator either of the adequacy of these revenues in relation to the government’s needs (i.e., public expenditure) or to the level of taxation that the government could raise, given the particular structure and characteristics of the economy at that particular time. In other words, from the knowledge that a given country’s average tax ratio is, say, X, it cannot be stated that such a ratio is high or low in view of the country’s potential nor that it is adequate to meet the country’s expenditure need.

Two questions might then be asked. First, is there any way in which these average tax ratios can be compared among countries, taking into account their potential as reflected in their economic characteristics? Second, how do these ratios relate to the amount of expenditure in these countries? The first question relates to the issue of the potential available to a country and the extent to which this potential has been utilized. The second question relates to the need for revenue by particular countries, regardless of their potential.

Different approaches have been developed in the literature to answer the first question, but all of them have shortcomings that are well known and need no elaboration in this study. The Fiscal Affairs Department of the Fund has been interested in this question over the years and has developed a methodology to permit a comparison of tax burdens through an index for international tax comparisons. This index attempts to measure the extent to which countries are taking advantage of the revenue possibilities available to them through the various tax bases. The methodology followed in a recent paper of the Department1 has been applied in this study to 34 countries of the sub-Saharan region. This approach consists of correlating the ratio of actual tax revenue to GDP (T/Y) against per capita nonexport income (Yp – Xp), the share of mining in GDP (Ny), and an export ratio excluding mineral exports (Xy). In theory, these variables are chosen because they capture the relevant characteristics of the economies.2

The application of this method to the sub-Saharan African countries yielded interesting results. The R2 was 0.54, which is relatively high for this type of study, and the coefficients of the independent variables were highly significant, except for per capita nonexport income. The derived equation3 is as follows:

The equation indicates that the share of mining in GDP and the export ratio, excluding mineral exports, are the most significant determinants of the tax ratio. Per capita income plays no role. The results in Table 4 have been calculated by using this equation. The four columns show (1) the rate of actual tax revenue to GDP, (2) the ratio predicted from the equation, (3) the International Tax Comparison (ITC) index derived by dividing the actual tax ratio by the predicted tax ratio, and (4) the ranking. A country with an actual tax ratio equal to the one predicted by the equation has an ITC index of 1, whereas a country with an actual ratio above the predicted one has an index greater than 1. According to these columns, the five countries that were exploiting their tax potential to a greater extent were Togo, Somalia, Zambia, Benin, and Seychelles, for all of which the ITC index exceeded 1.24. In a relative sense, these countries could be considered to be heavily taxed. On the other hand, Uganda, Ghana, Chad, Malawi, and Sierra Leone could be considered to have a relatively light tax burden because their ITC index was 0.80 or lower. The relative position of other countries can easily be seen in the table.

Next, tax revenues can be compared with government expenditures. These results again must be used with caution since countries may have revenues other than those derived from taxes. For example, some of the sub-Saharan African countries receive substantial grants from other countries, allowing them to maintain public expenditures greater than their tax revenue. And others may have access to profits from marketing boards, but these could not be taken into account for lack of data. However, keeping in mind other sources of revenue, Table 5 is of interest. While in a few countries (Cameroon, Senegal, Swaziland, and Upper Volta) tax revenues were of the same order of magnitude as public expenditures, for other countries public expenditures far exceeded tax revenues. In some of these countries (Benin, Botswana, Chad up to 1976, the Congo, Gabon up to 1977, The Gambia, Ghana, Malawi, Mauritania, Sudan, Tanzania, and Zaire up to 1977), public expenditures approached or were twofold the revenue from tax receipts. This divergence has serious implications for the balance of payments and for the behavior of prices, but these effects can only be mentioned here.

Table 5 indicates that in many sub-Saharan African countries the relationship between expenditure and tax revenue has progressively worsened in recent years. For example, between 1973 and 1978, the ratio of expenditure to tax revenue, expressed as a percentage, increased in The Gambia from 122 to 233, in Ghana it increased from 157 to 253, and in Mauritania it increased from 149 to 324. Significant rises are also shown in Benin, Botswana, the Congo, Ivory Coast, Liberia, Mauritius, Nigeria, Somalia, Sudan, Tanzania, and Zaire. Also, for some of the countries shown in Table 5, the ratio of expenditure to tax revenue fell over the period—for example, Burundi, Cameroon, Gabon in 1978, Kenya, Senegal, Seychelles, Sierra Leone, and Swaziland.

This deterioration of the relationship between expenditure and tax revenue occurred in spite of efforts by the countries to increase their tax revenue. These efforts are best illustrated in Table 6, which indicates the buoyancy of tax revenue with respect to GDP. The buoyancy concept incorporates both spontaneous changes in tax revenue and changes in the economy, as well as changes brought about by discretionary measures taken by the governments. Table 6 shows that for the majority of the countries the buoyancy of tax revenue exceeded 1 and in many it exceeded 1 by substantial amounts. A buoyancy of 1 indicates an unchanged ratio of tax revenue to GDP, while a buoyancy greater than 1 indicates that the ratio has been increasing over the period. Therefore, the results in Table 6 are consistent with those in Table 3, which indicate that for the majority of these countries the ratios of tax revenue to GDP have risen in recent years.

In summary, the data in Tables 4, 5, and 6 suggest that the worsening of the fiscal situation in many sub-Saharan African countries, by and large, was not the result of inelastic tax structures but perhaps the result of ambitious expenditure programs.

Alan A. Tait, Wilfrid L. M. Grätz, and Barry J. Eichen-green, “International Comparisons of Taxation for Selected Developing Countries, 1972–76,” International Monetary Fund, Staff Papers, Vol. 26 (March 1979), pp. 123–56.


The figures in parentheses below the coefficients are t-ratios.

    Other Resources Citing This Publication