Article

International Comparisons of Taxation for Selected Developing Countries, 1972–76

Author(s):
International Monetary Fund. Research Dept.
Published Date:
January 1979
Share
  • ShareShare
Show Summary Details

The international comparison of taxation is a controversial enterprise, and critics have not been lax in pointing out the shortcomings of previous studies.1 Yet summary measures, to the extent they successfully encapsulate large amounts of information, retain utilitarian value.2 One frequently used summary measure has come to be known as the “tax effort index.”3 The concept of tax effort reflects an attempt to overcome the limitations of the most straightforward of all comparisons—the simple ratio of taxes to output (hereinafter referred to as the “tax ratio”).4 In tax effort studies, the tax ratio is analyzed in terms of a country’s “taxable capacity,” and this analysis yields a presumption as to the feasibility of changes in the tax ratio.

This paper reviews the controversy as well as the demand for current results. In Sections I through III, problems of implementing the technique are discussed; previous studies are updated; and trends in taxation are described. In Sections IV through VI, new results are presented based upon an improved sample and a refined methodology; the sensitivity of international rankings to implicit assumptions is illustrated; and the paper’s main conclusions are summarized.

I. Theoretical Approach

Estimating taxable capacity involves regressing the tax ratio on variables that serve as proxies for a country’s “tax handles.”5 The ratio of actual to predicted tax ratios is computed and taken as an index for the purposes of international tax comparison.

where T= taxes
GDP= gross domestic product
A= a vector of tax handles

Following Hinrichs, Lotz and Morss, and Musgrave,6 ease of collection may be positively related to the following:

(1) the share of trade in production

(2) the sectoral composition of the economy

(3) the percentage of economic units exceeding a certain size limit and number of workers employed in such units

(4) the importance of large retail establishments

(5) the literacy level

Factor (1) can be represented by X + M/Y (the ratio of exports plus imports to GDP) or by Xy (the ratio of exports to GDP). Proxies for factor (2) might include Ny (the share of mining in GDP) and Ay (the share of agriculture in GDP). Since the importance of large producers, employers, and retail establishments is positively correlated with the level of economic development, Yy (per capita GDP) is taken as a proxy for factors (3) and (4). L (the literacy rate) represents factor (5). Thus, we have the functional relationship:

As in the study of Chelliah, Baas, and Kelly (CBK),7 the following five equations, which include various combinations of the explanatory variables, were estimated:

where Xp = export income per capita and Xʹy = the ratio of nonmineral exports to gross national product (GNP).

According to the proponents, ranking countries according to tax effort takes into account international differences in taxable capacity and, compared with the simple tax ratio, more accurately measures the sacrifice undertaken in order to raise tax revenue. It provides information about the capacity of countries to respond to fiscal problems by raising the level of taxation. In a country with a high tax effort, other measures probably should be employed to deal with a budgetary deficit. As CBK state, however, the indices “are not intended to be applied in a mechanistic fashion but rather to be considered useful additional information in judging the scope for more taxes.”8

The critics suggest that the level of taxation is not unambiguously related to any concept of effort. Even when the relative prices of publicly and privately provided goods do not differ across countries, tastes relevant to the public-private goods mix may do so. Two countries seeking to equate the marginal social benefits of public and private provision of goods and services will achieve very different tax ratios, which should not be seen as indices of effort but rather should be seen as the result of conventional maximizing behavior. The very term “tax effort” is misleading. A term that might be preferred, and that will be used in the rest of this paper, is “international tax comparison” (ITC). This change may appear to be merely cosmetic, but it eliminates the connotation that countries with low indices could try harder to raise their tax revenues and that they are making insufficient “efforts” to raise them. The neutral term “ITC” is a suitable replacement.9

II. Past Results Updated

In this section, new results comparable to those appearing in previous ITC studies are presented. The studies by CBK and Chelliah10 are updated using the identical sample of 47 developing economies. As in those studies, regression analysis is applied to cross-section data, and the “predicted” tax ratios from the taxable capacity equations are used to calculate indices of ITC.

The data are drawn from the Fund’s International Financial Statistics, Government Finance Statistics Yearbook, unpublished estimates available to the authors, and budget and other official country documents.11 The data are averaged over a period centering upon 1974 (See Table 7 in the Appendix.). When data are available for the entire 1972-76 span, a five-year average is computed. When only a subset of this data is available, a three-year average is used. As in previous studies, nontax revenues are excluded. In addition, social security contributions are excluded owing to problems of data availability and comparability. An attempt is made to gather data for all levels of government, although, as before, it is deemed permissible to exclude local revenues when they account for less than 10 per cent of the total.12 All data are converted to common currency units according to market exchange rates.

The equations are estimated using ordinary least squares (OLS). The possibility of a bias in the coefficient estimates owing to simultaneity was tested by running the same equations with two-stage least squares. Since the estimates did not change significantly when the estimation procedure was varied, OLS estimates are presented in order to maximize comparability with previous studies.

The new regression equations appear in Table 1 along with the earlier results. The results labeled “A” are drawn from the 1969–71 CBK study. The equations appearing under “B” are estimated using the same sample of 47 developing countries but with new data for the period 1972–76. The kkC” equations use a new sample of 63 developing countries and therefore are not directly comparable to the earlier results; they are discussed separately in Section IV. (For a list of these countries, see Table 7 in the Appendix.)

Table 1.Taxable Capacity Equations1
Lotz-Morss equation (per capita GNP and share of foreign trade in GNP)
A.196971(47)T/Y=11.65(7.77)+0.002(0.50)Yp+0.06(2.36)X+M/Y(8)R¯2=0.110
B.197276(47)T/Y=9.9683(6.02)+0.0003(0.18)Yp+0.1108(3.91)X+M/Y(9)R¯2=0.267F(2,77)=9.378
C.197276(63)T/Y=6.5775(3.75)+0.0003(1.20)Yp+0.1457(5.28)X+M/Y(10)R¯2=0.343F(2,60)=17.200
CBK 1969–71 equation (nonexport income per capita, share of mining and nonmineral exports in GNP)
A.196971(47)T/Y=11.47(7.84)+0.001(0.38)(YpXp)+0.44(5.45)Ny+0.05(1.17)Xy(11)R¯2=0.376
B.197276(47)T/Y=9.9949(6.15)0.0008(0.34)(YpXp)+0.4068(5.41)Ny+0.1938(3.12)Xy(12)R¯2=0.413F(3,43)=11.80
C.197276(63)T/Y=7.1134(4.82)0.0024(0.94)(YpXp)+0.5700(9.31)Ny+0.2218(4.17)Xy(13)R¯2=0.581F(3,59)=29.69
Nonexport income per capita and share of exports in GNP
A.196971(47)T/Y=10.36(6.31)+0.005(1.32)(YpXp)+0.15(3.35)Xy(14)R¯2=0.178
B.197276(47)T/Y=8.4022(5.54)+0.0005(0.22)(YpXp)+0.3037(6.49)Xy(15)R¯2=0.470F(2,44)=21.37
C.197276(63)T/Y=7.3663(4.41)+0.003(0.94)(YpXp)+0.3025(6.19)Xy(16)R¯2=0.375F(2,60)=19.58
Shares in GNP of mining, agriculture, and exports
A.196971(47)T/Y=14.46(8.12)+0.32(3.85)Ny+0.07(2.04)Ay+0.04(1.10)Xy(17)R¯2=0.445
B.197276(47)T/Y=8.0840(4.08)+0.2119(2.82)Ny+0.01581(0.36)Ay+0.2452(4.91)Xy(18)R¯2=0.542F(3,43)=19.16
C.197276(63)T/Y=9.1859(4.88)+0.3550(5.51)Ny+0.0240(0.61)Ay+0.1903(4.30)Xy(19)R¯2=0.593F(3,59)=31.12
Shares of mining and agriculture in GDP
A.196971(47)T/Y=15.66(11.07)+0.355(4.44)Ny0.08(2.37)Ay(20)R¯2=0.442
B.197276(47)T/Y=14.3579(7.67)+0.3555(4.15)Ny0.03018(0.57)Ay(21)R¯2=0.302F(2,44)=10.94
C.197276(63)T/Y=14.2423(8.54)+0.4521(6.59)Ny0.0571(1.30)Ay(22)R¯2=0.475F(2,60)=29.01

The Lotz-Morss equations (8)–(10) are estimated on the basis that taxable capacity is represented by per capita GNP and the ratio of foreign trade to national income. The coefficient estimates have not changed radically between sample periods. According to this specification, in equations (8) and (9) the proxies for taxable capacity account for over twice as much of the variance in tax ratios in 1972–76 as they did in 1969–71.

Equations (11)–(13) relate taxable capacity to per capita nonexport income and to the shares of mining and nonmineral exports in GDP. Only the mining and nonmineral export shares exhibit coefficients that are significantly different from zero. These equations attribute about 40 per cent of the variance in tax ratios to the capacity proxies, and the more recent sample improves the results.

When equations (14)–(19) are applied to the 1972–76 data, they attribute about half the variance in tax ratios to international differences in proxies for the capacity factors. For equations (14)—(16), which include nonexport income per capita and exports’ share in income as the determinants of capacity, this is a dramatic change from 1969–71. The increased explanatory power of the equation can be attributed entirely to the coefficient on exports’ share.

Equations (17)–(19) also suggest that the external sector mattered relatively more in the 1972–76 period than in 1969–71. For many of the countries in this sample with significant mineral production, the output of that sector is almost exclusively exported. Equations (12)—(14) avoid double counting by considering only nonmineral exports.

Finally, equations (20)–(22) take the sectoral composition of GDP as an indicator of taxable capacity. As in equations (17)–(19), the coefficient on agriculture is not significantly different from zero. In essence, this equation relates capacity to mineral production alone. It explains about 30 per cent of the variance in tax ratios.

According to statistical criteria, equations (11)—(15) (the CBK equations) perform most satisfactorily. Consequently, these equations are used throughout this paper for the construction of ITC indices. The reader’s attention is directed to Table 8 in the Appendix, where indices derived from all five equations are compared. The coefficients of rank correlation show that the rankings for all five equations are relatively stable. (Coefficients of rank correlation vary between 0.85 and 0.96.)

III. Trends in Tax Ratios and ITC Indices

The CBK 1969-71 study noted that tax ratios in developing countries increased steadily over the period 1953-71.13 This trend appears to have continued after 1971 (See Table 2.); the average tax ratio in 1972-76 was 16.1 per cent, as compared with 15.1 per cent in 1969–71 for the 47 countries included in this sample. In approximately 65 per cent of the countries considered in the 1969–71 study, the tax ratio increased in the period 1972–75. The comparable figure for the period from 1966–68 to 1969–71 is 80 per cent. However, the average level of taxation in developing economies is still well below that in developed countries.14

Table 2.Forty-Seven Developing Countries: Tax Ratios and International Tax Comparison (ITC) Indices, 1969–71 and 1972–76
Taxes as Per Cent of GNPRankingITC IndexITC IndexRanking
Country1972–761969–711969–71Country1972–761969–711969–711
Iran32.721.6(6)Iran1.7200.913(29)
Guyana31.723.4(3)Brazil1.6071.806(1)
Zambia30.831.3(1)Turkey1.4841.197(11)
Zaïre27.229.4(2)Sudan1.4651.440(5)
Venezuela23.120.4(7)Zambia1.3681.111(16)
Malaysia22.519.3(11)Guyana1.3501.059(19)
Trinidad and Tobago21.917.7(17)Tanzania1.3361.034(21)
Tunisia20.721.7(5)Zaïre1.2951.276(9)
Ivory Coast20.619.8(8)Chile1.2651.159(14)
Senegal20.218.1(14)India1.2521.093(17)
China, Republic of19.917.8(15)Kenya1.2191.090(18)
Kenya19.214.4(12)Morocco1.2141.224(10)
Jamaica19.019.4(10)Malaysia1.1911.193(12)
Tanzania18.913.9(24)Tunisia1.1841.639(2)
Sudan18.918.2(13)China, Republic of1.1251.304(8)
Morocco18.617.8(15)Ivory Coast1.1151.471(4)
Chile18.419.6(9)Argentina1.0990.973(24)
Egypt18.119.2(12)Jamaica1.0640.993(23)
Brazil18.122.9(4)Senegal1.0211.342(7)
Sri Lanka17.917.7(17)Peru0.9860.874(3D
Indonesia16.310.0(39)Sri Lanka0.9831.374(6)
Turkey16.215.6(20)Ghana0.9761.154(15)
Singapore15.713.2(28)Mali0.9681.055(20)
Ghana14.215.8(19)Thailand0.9680.925(28)
Peru14.014.2(23)Upper Volta0.9550.817(34)
Thailand13.912.4(32)Venezuela0.9200.958(26)
India13.913.4(25)Pakistan0.9150.728(39)
Costa Rica13.613.1(30)Colombia0.8990.901(30)
Korea13.615.4(21)Korea0.8581.181(13)
Argentina13.313.4(25)Costa Rica0.8580.970(25)
Mali12.913.2(28)Egypt0.8531.487(3)
Togo12.411.3(34)Ethiopia0.8030.705(40)
Ecuador12.013.4(25)Indonesia0.7970.658(42)
Bolivia11.88.2(43)Singapore0.7850.796(36)
Colombia11.612.5(31)Burundi0.7800.946(27)
Honduras11.511.3(34)Rwanda0.7730.602(44)
Pakistan11.48.8(41)Lebanon0.7620.782(37)
Upper Volta11.310.3(38)Bolivia0.7420.459(46)
Lebanon10.211.2(36)Mexico0.7330.490(45)
Philippines10.19.1(40)Trinidad and Tobago0.7230.834(33)
Ethiopia10.18.6(42)Philippines0.7180.683(41)
Rwanda10.07.9(44)Togo0.7030.739(38)
Burundi9.311.4(33)Ecuador0.6801.002(22)
Paraguay8.810.9(37)Honduras0.6690.800(35)
Mexico8.67.1(46)Paraguay0.6650.867(32)
Guatemala8.17.9(44)Guatemala0.5580.618(43)
Nepal5.44.4(47)Nepal0.4890.374(47)
Average16.115.1
Source: All indices are derived from equation (4). The indices for 1969-71 are drawn from CBK, op. cit., pp. 192-93. The 1972-76 indices are derived from Table 1, equation (12).

Changes in tax ratios from 1969–71 to 1972–76 were more pronounced than those from 1966–68 to 1969–71. Increased revenues for certain products (notably oil and sugar) and the 1974–75 world recession can be held responsible. Long-term trends in tax ratios are illustrated in Chart 1, which compares the 1966–68 and 1972–76 periods. Similarly, Chart 2 presents the shorter-term comparison between 1969–71 and 1972–76.

Chart 1.Forty-Seven Developing Countries: Comparison of Tax Ratios for 1966–68 and 1972–76

Sources: Calculated from text tables and from CBK, op. cit., pp. 192–93.

1 13.6 is the average tax ratio for 1966–68.

Chart 2.Forty-Seven Developing Countries: Comparison of Tax Ratios for 1969–71 and 1972–76

Sources: Calculated from text tables and from CBK, op. cit., pp. 192–93.

1 15.1 is the average tax ratio for 1969–71.

ITC rankings for 1969–71 and 1972–76 also appear in Table 2. In each case, the rankings are derived from equations (11) and (12) in Table 1. Countries with tax ratios that are above average tend to have tax indices above unity, and countries with below-average tax ratios tend to have lower than average ITC indices.15 There are five countries (Trinidad and Tobago, Sri Lanka, Indonesia, Venezuela, and Egypt) that have above-average tax ratios and ITC indices below one. There are only two countries (India and Argentina) with an ITC index above one and a tax ratio below the average. Among countries with tax ratios of less than 12 per cent, all except Pakistan have ITC indices below 0.9.

It is useful to consider how ITC indices evolve over a period of years. For this purpose, the tax indices based on 1972–76 data can be compared with those derived in previous studies. The 1972–76 data can be compared with 1966–68 data in order to make a “long-term” comparison and with 1969–71 data to construct a “short-term” comparison.

The short-term comparison shown in Table 2 shows a relatively stable ranking of the countries. Twelve countries16 that had been among the last 15 countries in 1969–71 are found again in the group of 15 countries with the lowest ITC indices in 1972–76. Of the countries in the first 15 places in 1972–76, 9 countries17 were in the same ranking interval in 1969–71. The two most striking changes are for Iran, which has the highest ranking in 1972–76 and was twenty-ninth in the 1969–71 period, and Egypt, which is thirty-first in 1972–76 and was third in 1969–71.

Also, a longer-term comparison of the ITC indices for 1966–68 and 1972–76 in Table 2 shows that the relative positions of various countries are stable. Placing countries into three general categories—high, medium, and low tax indices18—reveals that no one country moves from high to low or vice versa (though Egypt comes close to doing this). Of the 47 countries considered, 31 (or 65 per cent) remain in the same category.

Approximately the same number of countries shift between the high and medium and the low and medium categories. Of the 20 countries with rising ITC indices, 3 are oil producers. However, the majority of the countries that appear in the high, medium, or low categories in the 1966–68 period continue to do so in the 1972–76 period.

This is shown more clearly in Charts 3 and 4, which categorize the 47 countries according to whether their ITC indices are high and rising, high and falling, low and rising, or low and falling. The long-run (1966–68 and 1972–76) trend and the short-run (1969–71 and 1972–76) trend are similar.

Chart 3.Forty-Seven Developing Countries: Comparison of International Tax Comparison Indices for 1966–68 and 1972–76

Sources: Calculated from text tables and from CBK, op. cit., pp. 192–93.

Chart 4.Forty-Seven Developing Countries: Comparison of International Tax Comparison Indices for 1969–71 and 1972–76

Sources: Calculated from text tables and from CBK, op. cit., pp. 192–93.

If an evaluative comment were made solely on the basis of Charts 3 and 4, it would be that if the objective of tax policy is to mobilize resources for the public sector, then the countries in the “low and falling” category should be most concerned about their failure to improve their position, while those in the “high and rising” category should be seen as following successful tax policies worthy of emulation. If the objective were to minimize governmental interference with the private sector, precisely the opposite view should be taken. According to the U.S. Advisory Commission on Intergovernmental Relations, which espouses the latter viewpoint, “policymakers in these States [with relatively high and rising tax indices] are alerted to the need for taking a much more critical view of the costs and benefits of public versus private spending.”19 That is, in an international context, those countries with a high and rising tax index might prudently seek ways of curbing the growth of the public sector.

Table 3 shows the development of tax indices between 1966–68 and 1972–76. Among the “high” ITC countries, a few indices increased dramatically. Most notable are those of Sudan (33 per cent) and Guyana (31 per cent). The largest increases among the “low” ITC countries are found in Iran (77 per cent) and Nepal (63 per cent)—which softens the criticism implicit in Nepal’s bottom position in Table 2. Of the “low and falling” ITC countries, Ecuador’s ITC index declined most sharply (30 per cent).

Table 3.Forty-Seven Developing Countries: Development of ITC Indices, 1966-68 and 1972-761
High and rising

  • Sudan (77)

  • Guyana (33)

  • Turkey (27)

  • Tanzania (26)

  • India (19)

  • Malaysia (17)

  • Zambia (16)

  • China, Republic of (8)

  • Chile (8)

  • Peru (7)

  • Kenya (5)

  • Morocco (4)

  • Jamaica (3)

High and falling

  • Senegal (26)

  • Sri Lanka (23)

  • Brazil (10)

  • Zaïre (10)

  • Tunisia (9)

  • Ghana (4)

  • Ivory Coast (2)

  • Argentina (0)

Low and rising

  • Iran (77)

  • Nepal (63)

  • Bolivia (38)

  • Indonesia (29)

  • Pakistan (22)

  • Colombia (12)

  • Rwanda (10)

  • Costa Rica (5)

  • Singapore (4)

  • Ethiopia (3)

  • Trinidad and Tobago (3)

Low and falling

  • Ecuador (30)

  • Mali (29)

  • Egypt (25)

  • Upper Volta (19)

  • Paraguay (17)

  • Guatemala (14)

  • Korea (12)

  • Honduras (11)

  • Lebanon (11)

  • Burundi (10)

  • Philippines (7)

  • Venezuela (5)

  • Mexico (5)

  • Thailand (3)

  • Togo (1)

Source: Table 2 of this paper and Table 1 of CBK, op. cit., pp. 192-93.

Up to this point, all ITC rankings have been derived from equations (4) and (5) (two of the CBK equations). ITC indices corresponding to each of the principal groups of equations shown in Table 1 are shown in Table 8 in the Appendix. Although some indices vary considerably, the vast majority appear to be relatively stable, and the rankings to be broadly similar.20 The tables and charts showing the changes in tax ratios and indices over time focus attention on countries where changes have been large, suggest that the changes may reflect special problems, and invite further investigation of the circumstances particular to each country. Three main factors appear to be responsible (together or separately): (1) changes in important commodity prices; (2) changes in the level of trade, particularly of foreign trade; and (3) discretionary or automatic changes in the tax structure. For instance, the 31 per cent rise in Guyana’s tax index from 1966–68 to 1972–76 (See Table 3.) was owing to sugar levy receipts on higher sugar prices. Similarly, Iran’s 77 per cent increase was attributable to the sharp rise in its oil revenues. Although Venezuela enjoyed a similar increase in petroleum-related revenues—56 per cent during the period 1971–72 to 1976—this was insufficient to offset the fall in other revenues, and Venezuela is still shown as a country with a “low and falling” tax index (See Table 3 and Charts 3 and 4.).

A comparison of Charts 2 and 4 to contrast the pictures presented by tax ratios to those shown by the ITC indices indicates that the ITC indices tend to emphasize the negative character of performance rather more than the tax ratios. In Chart 4, 12 countries are shown with “high and falling” ITC indices, but in Chart 2, only 9 are in this category. Eleven countries have “low and falling” ITC indices in Chart 4, whereas only 8 are in this category in Chart 2.

For particular countries, some differences between the two methods of presentation are striking. From a simple tax ratio, Iran appears to be in the “high and rising” category, but the ITC puts it into the “low and rising” category. The same is true for Trinidad and Tobago. Venezuela is transferred from the “high and rising” category to the “low and falling” group.. Each of these countries has substantial revenue from oil, and the ITC index, to some extent, takes account of this in a way that the tax ratio does not.

IV. ITC Indices: Large and Small Countries, Per Capita Income, Population, and Geographical Areas

All studies that attempt to explain the variance of a sample of tax ratios start with an assumption about the comparability of countries with different economic characteristics. CBK consider only “low-income countries,” which, with the exception of Argentina, have per capita incomes below $1,000.21 Musgrave examines separately countries in the $0–300, $301–600, and $601 and above per capita income groups; and Hinrichs constructs similar categories.22 The rationale for distinguishing among countries according to per capita income is that countries in different stages of economic development will exhibit significantly different structural relationships between their tax ratios and other economic variables.

How significant are the differences, among countries with different economic characteristics, in the relationship between taxable capacity and its explanatory variables? Evidence was taken from regressions on a new sample of 63 developing economies (See Table 9 in the Appendix.). The convention employed by CBK of including countries with per capita GNP of less than $1,000 was retained while selecting this sample. Therefore, some of the countries examined in Sections I and II were excluded from this larger sample. All countries with the appropriate per capita GNP and for which recent data were available were included. Separate regressions were run for subsamples of countries, according to such features as population, national income, and per capita GNP. The results of these regressions are compared with those for the full sample of 63 countries; they are given in the tables in the Appendix. The rankings derived from the small (47 countries) and large (63 countries) samples seem to be quite stable. (The coefficient of rank correlation is 0.91.)

The full sample regressions appear as the “C” equations in Table 1. The larger, more homogeneous sample does not change in any consistent fashion the percentage of the variance in tax ratios attributable to the capacity proxies. R2 s fall in three cases and rise in two others in comparison with the “B” equations in Table 1. All coefficients that proved significant when the 47-country sample was used remain so and, in addition, 2 more coefficients can now be deemed significant at the 95 per cent level of confidence.

When the sample is divided into 12 countries with GNPs greater than $10 billion and 51 countries with GNPs less than $10 billion (See Table 10 in the Appendix.), the summary statistics suggest that the equation is acceptable only for the smaller economies.

When countries are classified by per capita income (See Table 11 in the Appendix.), in spite of respectable R2 s, equation (4) performs satisfactorily for countries in the $0–250 and $250–500 ranges but not for those with per capita incomes over $500.

In the two cases where equation (4) fits the data, its explanatory power is owing largely to the share of mining in GNP. Some support is lent to the hypothesis that mining and foreign trade matter relatively more for the countries with per capita GNPs in the lowest range. In fact, it is equation (5) that best explains taxable capacity for the countries in the $0–250 category, affirming that, at the earliest stage of development, the mix of domestic and foreign trade is the most important determinant of taxable capacity.23

Another hypothesis is that the relationship between taxable capacity and its explanatory variables might differ, depending on the population density of the country. Table 12 in the Appendix displays the five taxable capacity equations by population subgroups. It is not clear why population should matter. There may be significant economies of scale in the collection of various types of taxes. In general, however, differences between countries with high and low densities are not striking.

As always, it is difficult to quantify the social and political factors that influence taxable capacity. Not only economic parameters but such social variables as attitude toward egalitarianism, allocative neutrality, fiscal centralization, and the influence of various private interest groups upon the formation of economic policy affect the extent to which a government can exploit a potential tax base.24 For example, the capacity of two otherwise identical economies to tax the external trade sector would differ significantly, depending on whether the relatively wealthy consumers of imports had enough political clout to discourage import duties or whether the poorer classes were able to initiate income redistribution programs.

One way of testing the significance of social and political factors is to assume that countries in different parts of the world differ systematically in terms of political systems but have sufficient common characteristics in attitudes toward taxation, tax morality, and efficiency to exhibit cohesiveness. When grouped geographically, each of the five taxable capacity equations is presented for three geographical subgroups: Africa, Asia, and Latin America. (See Table 9 in the Appendix.) The coefficient estimates suggest that the continents differ only in that agriculture consistently influences taxable capacity in Africa, whereas in Asia and Latin America, external trade is a more significant determinant.25

Table 4.Comparison of ITC Indices for the New Sample of 63 Developing Countries with Per Capita GNPs Less Than $1,000 and the CBK Sample of 47 Developing Countries1
CountryNew SampleCBK Sample
Afghanistan0.52
Algeria1.48
Argentina1.09
Bangladesh0.68
Benin1.21
Bolivia0.740.74
Brazil1.601.60
Burma0.84
Burundi0.970.78
Cameroon1.34
Central African Empire0.79
Chile1.061.26
China, Republic of1.131.12
Colombia0.970.89
Congo1.51
Costa Rica0.860.85
Dominican Republic0.95
Ecuador0.660.68
Egypt0.790.85
El Salvador0.77
Ethiopia0.970.80
Gambia, The0.96
Ghana1.040.97
Guatemala0.580.55
Guinea1.34
Guyana1.291.35
Honduras0.690.67
India1.561.25
Indonesia0.750.79
Iran1.72
Iraq0.99
Ivory Coast1.141.11
Jamaica0.961.06
Jordan0.93
Kenya1.351.21
Korea0.910.85
Lebanon0.76
Liberia0.62
Malawi0.81
Malaysia1.181.19
Mali1.150.96
Mexico0.690.73
Morocco1.261.21
Nepal0.620.48
Nicaragua0.69
Pakistan1.100.91
Panama0.69
Paraguay0.730.66
Peru0.950.98
Philippines0.780.71
Rwanda0.910.77
Senegal0.971.02
Sierra Leone0.86
Singapore0.78
Sri Lanka0.%0.98
Sudan1.721.46
Swaziland1.04
Syrian Arab Republic0.56——
Tanzania1.531.33
Thailand1.060.96
Togo0.690.70
Trinidad and Tobago0.72
Tunisia1.141.18
Turkey1.531.48
Upper Volta1.190.95
Venezuela0.92
Yemen Arab
Republic0.87
Zaïre1.211.29
Zambia1.221.36

V. Commentary on “Traditional” ITCs

The apparent vulnerability of the ITC index to changes in the sample and in data availability must erode the validity of the results. At the same time, rank correlations are remarkably stable and, the larger the sample, the more acceptable should be the results. The tests using the present 63-country base yield better results than the earlier, more limited samples; second, the stability of the majority of the results is reassuring.

However, many countries’ ITC indices are sufficiently sensitive to changes in sampling procedures that the degree of precision and stability provided by the rankings may be misleading. How valuable, after all, are two or three decimal places when the ITC index varies, as it does in Colombia’s case, from 0.97 to 1.36?

One way to avoid the pitfalls of assuming that the results are more precise than they actually are is to construct a few broad categories similar to those shown in Table 5.26 The sample could be divided into countries with low, medium, and high ITC indices on the basis of either their full sample ratios or an unweighted average of the three subsample ratios in Table 6. When the countries are reclassified by the unweighted average of their three subsample tax effort indices, 8 of the 63 countries shift categories. No country moves from a low to a high ITC index or vice versa. Most of the changes occur when countries are on or near the borderline between categories. Rwanda, for instance, is only 3 percentage points away from being labeled medium on the basis of the average of its three subsample indices.

Table 5.Sixty-Three Developing Countries: Summary Measures of ITC Indices
Full Sample Tax Indices
Low ITC index

(less than 0.84)
Medium ITC index

(0.84-1.10)
High ITC index

(greater than 1.10)
AfghanistanBurundiAlgeria
BangladeshChileBenin
BoliviaColombiaBrazil
BurmaCosta RicaCameroon
Central African EmpireDominican RepublicChina, Republic of
EcuadorEthiopiaCongo
EgyptGambia, TheGuinea
El SalvadorGhanaGuyana
GuatemalaIraqIndia
HondurasJamaicaIvory Coast
IndonesiaJordanKenya
LiberiaKoreaMalaysia
MalawiPakistanMali
MexicoPeruMorocco
NepalRwandaSudan
NicaraguaSenegalTanzania
PanamaSierra LeoneTunisia
ParaguaySri LankaTurkey
PhilippinesSwazilandUpper Volta
Syrian Arab RepublicThailandZaïre
TogoYemen Arab RepublicZambia
Source: Equation (13) in Table 1.
Table 6.Sixty-Three Developing Countries: Summary Measures of ITC Indices
Mean of Subsample Tax Indices
Low ITC index

(less than 0.89)
Medium ITC index

(0.89-1.09)
High ITC index

(greater than 1.09)
AfghanistanBurma1Algeria
BangladeshBurundiBenin
BoliviaChileBrazil
Central African EmpireChina, Republic of1Cameroon
EcuadorCosta RicaColombia1
EgyptDominican RepublicCongo
El SalvadorEthiopiaGuinea
Gambia, TheGhanaGuyana
GuatemalaIraqIndia
HondurasJamaicaIvory Coast
IndonesiaJordanKenya
LiberiaKoreaMalaysia
MalawiMali1Morocco
MexicoParaguay1Pakistan.1
NepalPeruSudan
NicaraguaSenegalTanzania
PanamaSierra LeoneTunisia
PhilippinesSri LankaTurkey
Rwanda1SwazilandUpper Volta
Syrian Arab RepublicThailandZaïre
TogoYemen Arab RepublicZambia
Source: Computed from Tables 10, 11, 12, and 13 in the Appendix.

For 87 per cent of the countries sampled, ITC indices provide a stable indicator, in terms of broad categories, of relative fiscal behavior. Recognizing the limitations of ITC, it may be preferable to rely on broad indicators rather than to impute to the results a greater degree of precision than the methodology warrants.

VI. Conclusions

Tests of alternative equations for ITC using new data confirmed that the expressions—including nonexport income, the share of mining in GDP, and the share of nonmineral exports in GDP—that were used in earlier studies still were the most acceptable. (See Section II.) When results using straightforward tax ratios and the ITC equation were compared for the periods 1966-68, 1969-71, and 1972-76, it was noted that the changes were more pronounced between the two latest periods than between the earliest and latest periods. Countries were grouped into the following categories of tax ratios and ITC indices: “low and falling,” “low and rising,” “high and falling,” and “high and rising.” ITC indices appeared to reflect individual country mixes more realistically than simple tax ratios.

Regrouping ITC indices by subgroups for size of country, per capita income, population, and geographical area did not add significantly to an understanding of the relationships.

The most outstanding result of these tests of different equations and groupings, at a point in time and through time, was the apparent vulnerability of the absolute ITC indices to changes in the sample. However, rankings proved to be relatively stable.

APPENDIX
Table 7.Sixty-Three Developing Countries: Coverage1
CountryYearsLevel(s) of Government CoveredCountryYearsLevel(s) of Government Covered
Afghanistan1972–75CKenya1972–75A
Algeria1972–75CKorea1972–76C
Bangladesh1972–75CLiberia1972–75C
Benin1972–75C
Bolivia1972–76CMalawi1972–75C
Brazil1972–75AMalaysia1972–75A
Burma1972–75AMali1972–75A
Burundi1972–75CMexico1972–75C
Morocco1972–75A
Cameroon1972–75C
Central African Empire1972–75CNepal1972–75C
Chile1972–76CNicaragua1972–76C
China, Republic of1972–75A
Colombia1972–76APakistan1972–75C
Congo1972–75CPanama1972–76C
Costa Rica1972–75AParaguay1972–75C
Peru1972–75C
Dominican Republic1972–75CPhilippines1972–75C
Ecuador1972–75CRwanda1972–75A
Egypt1972–75A
El Salvador1972–75CSenegal1972–75C
Ethiopia1972–75CSierra Leone1972–74C
Sri Lanka1972–75C
Sudan1972–74C
Gambia, The1972–75CSwaziland1972–74C
Ghana1972–75C
Guatemala1972–76CSyrian Arab Republic1972–75C
Guinea1972–75C
Guyana1972–75CTanzania1972–75C
Thailand1972–75C
Honduras1972–76CTogo1972–75C
Tunisia1973–76A
India1972–74CTurkey1972–75A
Indonesia1972–75CUpper Volta1972–75A
Iraq1972–74C
Ivory Coast1972–75CYemen Arab Republic1972–75C
Jamaica1972–75CZaïre1972–75C
Jordan1972–74CZambia1972–75A
Table 8.Forty-Seven Developing Countries: ITC Indices Based on Five Regression Equations, 1972-761
CountryPer Capita GNP and Foreign Trade

(equation (10))
Nonexport Income, Mining, and Nonminerals

(equation (12))
Nonexport Income Per Capita and Exports

(equation (16))
Mining, Agriculture, ind Exports

(equation (19))
Mining and Agriculture

(equation (22))
Iran1.8191.7201.3891.4281.794
Brazil1.4531.6071.6111.7311.256
Turkey1.3481.4841.5861.6171.157
Sudan1.3301.4651.4281.5111.429
Zambia1.5241.3681.3391.1901.286
Guyana1.3061.3501.1401.1681.687
Tanzania1.1921.3361.2671.3381.407
Zaïre1.3601.2951.3681.2111.218
Chile1.3531.2651.4081.3161.013
India1.2361.2521.3981.3581.053
Kenya1.0931.2191.0561.1641.421
Morocco1.1471.2141.1251.1571.173
Malaysia1.1231.1911.0081.0771.481
Tunisia1.2001.1841.1541.1451.199
China, Republic of1.0381.1250.9701.0951.395
Ivory Coast1.0471.1150.9341.0521.528
Argentina1.0331.0991.0691.1500.912
Jamaica0.9871.0640.9600.9711.045
Senegal1.1991.0211.2331.1141.059
Peru1.0040.9861.0721.0460.846
Sri Lanka1.3020.9831.3811.1780.990
Ghana0.9690.9760.9510.9601.043
Mali0.8070.9680.9360.9900.989
Thailand0.9320.9680.9380.9851.003
Upper Volta0.7850.9550.9881.0280.845
Venezuela1.3610.9201.1750.9620.881
Pakistan0.8640.9150.9390.9740.845
Colombia0.8650.8990.8880.9320.830
Korea0.7750.8580.7790.8510.974
Costa Rica0.7350.8580.7400.8320.990
Egypt1.2050.8531.3351.0620.875
Ethiopia0.7810.8030.8120.8320.781
Indonesia1.0730.7971.0240.8460.788
Singapore0.5340.7850.6180.7281.087
Burundi0.7200.7800.8070.8110.733
Rwanda0.7630.7730.8170.7850.760
Lebanon0.5980.7620.6930.7770.716
Bolivia0.7270.7420.7060.6690.630
Mexico0.6700.7330.7350.7870.589
Trinidad and Tobago1.0880.7230.8990.7650.804
Philippines0.6800.7180.6880.7220.713
Togo0.7940.7030.7870.7050.665
Ecuador0.7380.6800.6960.6600.661
Honduras0.6390.6690.6280.6610.778
Paraguay0.6010.6650.6260.6830.642
Guatemala0.5400.5580.5390.5730.571
Nepal0.4600.4890.5310.5670.377
Source: Derived from Table 1.
Table 9.Sixty-Three Developing Countries, Grouped According to GNP, GNP Per Capita, Population, and Geographical Location
GNP1GNP Per Capita1
≤ $10 billion> $10 billion≤ $250$251-500$501-750> $750
  • Afghanistan

  • Algeria

  • Bangladesh

  • Benin

  • Bolivia

  • Burma

  • Burundi

  • Cameroon

  • Central African Empire

  • Congo

  • Costa Rica

  • Dominican Republic

  • Ecuador

  • Egypt

  • El Salvador

  • Ethiopia

  • Gambia, The

  • Ghana

  • Guatemala

  • Guinea

  • Guyana

  • Honduras

  • Iraq

  • Ivory Coast

  • Jamaica

  • Jordan

  • Kenya

  • Liberia

  • Malawi

  • Malaysia

  • Mali

  • Morocco

  • Nepal

  • Nicaragua

  • Panama

  • Pakistan

  • Paraguay

  • Rwanda

  • Senegal

  • Sierra Leone

  • Sri Lanka

  • Sudan

  • Swaziland

  • Syrian Arab Republic

  • Tanzania

  • Togo

  • Tunisia

  • Upper Volta

  • Yemen Arab Republic

  • Zaïre

  • Zambia

  • Brazil

  • Chile

  • China, Rep. of

  • Colombia

  • India

  • Indonesia

  • Korea

  • Mexico

  • Peru

  • Philippines

  • Thailand

  • Turkey

  • Afghanistan

  • Bangladesh

  • Benin

  • Burma

  • Burundi

  • Central African Empire

  • Ethiopia

  • Gambia, The

  • Guinea

  • India

  • Indonesia

  • Kenya

  • Malawi

  • Mali

  • Nepal

  • Pakistan

  • Rwanda

  • Sierra Leone

  • Sri Lanka

  • Sudan

  • Tanzania

  • Togo

  • Upper Volta

  • Yemen Arab Republic

  • Zaïre

  • Bolivia

  • Cameroon

  • Colombia

  • Congo

  • Ecuador

  • Egypt

  • El Salvador

  • Ghana

  • Guyana

  • Honduras

  • Jordan

  • Korea

  • Liberia

  • Morocco

  • Paraguay

  • Philippines

  • Senegal

  • Swaziland

  • Syrian Arab Republic

  • Thailand

  • Zambia

  • Algeria

  • China, Rep. of

  • Dominican Republic

  • Guatemala

  • Ivory Coast

  • Malaysia

  • Nicaragua

  • Peru

  • Tunisia

  • Turkey

  • Brazil

  • Chile

  • Costa Rica

  • Iraq

  • Jamaica

  • Mexico

  • Panama

Population1Geographical Grouping
≤ 10 million> 10 millionAfrican CountriesAsian CountriesLatin American Countries
  • Benin

  • Bolivia

  • Burundi

  • Cameroon

  • Central African

  • Empire

  • Congo

  • Costa Rica

  • Dominican Republic

  • Ecuador

  • El Salvador

  • Gambia, The

  • Ghana

  • Guatemala

  • Guinea

  • Guyana

  • Honduras

  • Ivory Coast

  • Jamaica

  • Jordan

  • Liberia

  • Malawi

  • Mali

  • Nicaragua

  • Panama

  • Paraguay

  • Rwanda

  • Senegal

  • Sierra Leone

  • Swaziland

  • Syrian Arab

  • Republic

  • Togo

  • Tunisia

  • Upper Volta

  • Yemen Arab

  • Republic

  • Zambia

  • Afghanistan

  • Algeria

  • Bangladesh

  • Brazil

  • Burma

  • Chile

  • China, Rep. of

  • Colombia

  • Egypt

  • Ethiopia

  • India

  • Indonesia

  • Iraq

  • Kenya

  • Korea

  • Malaysia

  • Mexico

  • Morocco

  • Nepal

  • Pakistan

  • Peru

  • Philippines

  • Sri Lanka

  • Sudan

  • Tanzania

  • Thailand

  • Turkey

  • Zaïre

  • Algeria

  • Benin

  • Burundi

  • Cameroon

  • Central African

  • Empire

  • Congo

  • Egypt

  • Ethiopia

  • Gambia, The

  • Ghana

  • Guinea

  • Ivory Coast

  • Kenya

  • Liberia

  • Malawi

  • Mali

  • Morocco

  • Rwanda

  • Senegal

  • Sierra Leone

  • Sudan

  • Swaziland

  • Tanzania

  • Togo

  • Tunisia

  • Upper Volta

  • Zaïre

  • Zambia

  • Afghanistan

  • Bangladesh

  • Burma

  • China, Rep. of

  • India

  • Indonesia

  • Iraq

  • Jordan

  • Korea

  • Malaysia

  • Nepal

  • Pakistan

  • Philippines

  • Sri Lanka

  • Syrian Arab

  • Republic

  • Thailand

  • Turkey

  • Yemen Arab

  • Republic

  • Bolivia

  • Brazil

  • Chile

  • Colombia

  • Costa Rica

  • Dominican

  • Republic

  • Ecuador

  • El Salvador

  • Guatemala

  • Guyana

  • Honduras

  • Jamaica

  • Mexico

  • Nicaragua

  • Panama

  • Paraguay

  • Peru

Table 10.Taxable Capacity Equation (4) by GNP Subgroup1
GNP ≤ $10 billion
(51countries)1T/Y=6.672(4.15)0.0041(0.90)(YpXp)+0.6403(9.23)Ny+0.2899(4.38)XyR¯2=0.634F(3,47)=29.91
GNP > $10 billion
(12countries)1T/Y=10.3317(3.19)+0.0034(0.96)(YpXp)+0.2096(1.14)Ny+0.1141(1.03)XyR¯2=0.083F(3,8)=0.72
Table 11.The Taxable Capacity Equation (4) by Per Capita GNP Subgroup1
Per capita GNP ≤ $250
(25countries)1T/Y=8.1107-0.1012(Yp-Xp)+0.5099+0.3077Xy(2.58)(-0.43)(4.55)(3.00)R¯2=0.476
Per capita GNP > $250 and ≤ $500
(21countries)1T/Y=12.1560(1.66)0.0188(0.99)(YpXp)+0.4506Ny(3.10)+0.2474(2.06)XyR¯2=0.448F(3,17)=6.41
Per capita GNP > $500 and ≤ $750
(10countries)1T/Y=30.8230(1.66)0.0557(2.08)(YpXp)+1.4246Ny(5.49)+0.6654(2.88)XyR¯2=0.835F(3,6)=16.17
Per capita GNP > $750
(7countries)1T/Y=22.2555(1.61)0.0083(0.67)(YpXp)+0.3650Ny(1.94)0.1421(0.63)XyR¯2=0.836F(3,3)=11.21
Table 12.Taxable Capacity Equations by Population Subgroup1
Population Subgroup
Lotz-Morss equation (3)
AT/Y=5.3893(2.48)0.0031(0.97)Yp+0.1720(4.97)X+M/Y(23)R¯2=0.418F(2,32)=13.20
BT/Y=4.2441(1.78)0.0058(1.76)Yp+0.2294(5.30)X+M/Y(24)R¯2=0.545F(2,25)=17.17
CBK equation (4)
AT/Y=6.7993(3.44)0.0057(1.28)(YpXp)+0.5589(5.77)Ny+0.2924(4.30)Xy(25)R¯2=0.505F(3,31)=12.57
BT/Y=6.8458(3.08)0.0046(1.38)(YpXp)+0.6010(7.64)Ny+0.2657(2.72)Xy(26)R¯2=0.678F(3,24)=20.00
Equation (5)
AT/Y=8.1758(4.26)0.0058(1.26)(YpXp)+0.2941(5.38)Xy(27)R¯2=0.442F(2,32)=14.50
BT/Y=5.3072(2.24)0.0060(1.51)(YpXp)+0.4657(5.73)Xy(28)R¯2=0.542F(2,25)=16.98
Equation (6)
AT/Y=5.4122(1.95)+0.2723(2.91)(Ny)+0.0270(0.55)Ay+0.2446(4.42)Xy(29)R¯2=0.526F(3,31)=13.57
BT/Y=10.4389(4.03)+0.3455(3.71)(Ny)+0.0537(0.90)Ay+0.2446(2.97)Xy(30)R¯2=0.695F(3,24)=21.57
Equation (7)
AT/Y=14.4239(6.10)+0.3498(3.02)Ny0.0500(0.86)Ay(31)R¯2=0.251F(2,32)=6.71
BT/Y=14.6733(5.92)+0.5006(5.66)Ny0.0715(1.03)Ay(32)R¯2=0.600F(2,25)=21.25
Table 13.Taxable Capacity Equations By Geographical Subgroup1
Geographical Subgroup
Lotz-Morss equation (3)
Africa
T/Y=8.8471(3.18)+0.0191(2.67)Yp+0.0677(1.33)X+M/Y(33)R¯2=0.373F(2,25)=9.04
Asia
T/Y=4.3711(1.92+0.0151(2.21)Yp+0.1085(1.88)X+M/Y(34)R¯2=0.593F(2,15)=13.37
Latin America
T/Y=3.9749(0.93)+0.0050(1.23)Yp+0.1163(2.57)X+M/Y(35)R¯2=0.244F(2,14)=3.58
CBK equation (4)
Africa
T/Y=8.1556(3.33)+0.0209(2.09)(YpXp)+0.4333(3.74)Ny+0.1387(1.47)Xy(36)R¯2=0.533F(3,24)=11.28
Asia
T/Y=4.5837(2.65)+0.0135(2.28)(YpXp)+0.4958(7.21)Ny+0.1932(2.30)Xy(37)R¯2=0.804F(3,14)=24.31
Latin America
T/Y=0.6428(0.14)+0.0073(1.61)(YpXp)+0.7893(3.86)Ny+0.2764(2.58)Xy(38)R¯2=0.461F(3,13)=5.57
Equation (5)
Africa
T/Y=7.9248(3.09)+0.0310(2.92)(YpXp)+0.1654(2.12)Xy(39)R¯2=0.386F(2,25)=9.49
Asia
T/Y=4.4438(2.08)+0.0090(1.17)(YpXp)+0.3838(4.92)Xy(40)R¯2=0.670F(2,15)=18.27
Latin America
T/Y=0.6236(0.13)+0.0093(1.85)(YpXp)+0.3267(3.31)Xy(41)R¯2=0.359F(2,14)=5.48
Equation (6)
Africa
T/Y=17.8356(4.30)+0.2756(2.72)Ny0.1426(2.04)Ay+0.0966(1.30)Xy(42)R¯2=0.539F(3,24)=11.55
Asia
T/Y=6.2692(2.02)+0.2744(2.79)Ny+0.0140(0.21)Ay+0.2765(3.21)Xy(43)R¯2=0.752F(3,14)=18.19
Latin America
T/Y=8.7356(2.71)+0.3806(1.54)Ny0.0888(0.70)Ay+0.1929(1.99)Xy(44)R¯2=0.373F(3,13)=4.17
Equation (7)
Africa
T/Y=21.8183(7.65)+0.2800(2.73)Ny0.1821(2.85)Ay(45)R¯2=0.526F(2,25)=16.01
Asia
T/Y=12.4371(4.00)+0.4638(4.63)Ny0.0515(0.63)Ay(46)R¯2=0.598F(2,15)=13.65
Latin America
T/Y=10.3779(3.03)+0.6137(2.56)Ny+0.0183(0.14)Ay(47)R¯2=0.240F(2,14)=3.53
Table 14.Sixty-Three Developing Countries: Ratios of Tax Categories to Total Taxes, 1972-76
CountryIncome TaxProperty TaxForeign Trade TaxSales TaxOther TaxDirect TaxIndirect TaxPoll Tax
Afghanistan15.21.775.94.42.816.880.30.0
Algeria66.20.07.326.50.066.233.80.0
Bangladesh9.80.030.840.718.79.871.50.0
Benin21.10.158.516.21.923.474.72.2
Bolivia9.92.353.831.22.812.285.00.0
Brazil19.61.65.968.14.721.374.00.0
Burma36.10.018.434.211.336.152.60.0
Burundi22.81.440.625.60.533.366.29.1
Cameroon17.11.847.329.93.819.077.20.1
Central African Empire24.92.137.124.58.230.361.63.3
Chile29.02.711.051.33.734.062.32.2
China, Republic of14.412.231.341.50.626.672.80.0
Colombia37.20.121.121.119.338.542.21.2
Congo37.30.135.024.00.041.059.03.6
Costa Rica23.30.331.344.40.623.775.70.0
Dominican Republic22.30.253.518.65.422.572.10.0
Ecuador25.50.650.021.52.426.171.50.0
Egypt25.82.937.628.31.333.265.54.5
El Salvador18.75.839.634.50.025.874.21.3
Ethiopia26.32.639.031.01.228.970.00.0
Gambia, The14.60.084.60.00.914.684.60.0
Ghana25.30.051.522.60.625.374.00.0
Guatemala14.73.530.051.10.618.281.10.0
Guinea25.90.053.411.86.927.765.41.8
Guyana31.41.547.017.13.132.964.00.0
Honduras27.21.134.637.10.028.371.70.0
India18.00.113.357.311.218.270.70.0
Indonesia66.80.815.612.31.570.727.83.1
Iraq86.10.87.64.60.986.912.20.0
Ivory Coast17.51.468.58.44.218.976.90.0
Jamaica47.53.317.525.46.250.843.00.0
Jordan9.60.230.116.243.99.846.30.0
Kenya39.80.322.736.11.140.158.80.0
Korea28.54.113.950.23.332.664.10.0
Liberia40.31.731.523.82.742.055.30.0
Malawi42.80.124.332.20.542.956.50.0
Malaysia30.00.631.822.515.130.754.20.0
Mali12.00.549.610.627.312.560.20.0
Mexico52.70.08.334.62.754.542.81.8
Morocco27.62.637.328.83.730.266.20.0
Nepal6.016.441.835.80.022.577.50.0
Nicaragua11.17.227.549.62.420.577.12.2
Pakistan13.62.735.841.36.616.477.10.0
Panama41.45.326.621.84.647.048.40.3
Paraguay12.49.233.130.213.922.963.31.2
Peru28.94.821.240.44.733.761.60.0
Philippines24.82.932.229.84.034.061.96.4
Rwanda19.42.855.215.86.822.371.00.0
Senegal21.20.043.325.110.421.268.40.0
Sierra Leone32.60.047.518.30.833.365.90.7
Sri Lanka17.02.044.936.10.019.180.90.0
Sudan12.40.055.620.911.112.476.50.0
Swaziland29.20.442.725.52.329.568.20.0
Syrian Arab Republic22.11.837.216.421.125.353.61.4
Tanzania30.90.523.844.80.031.468.60.0
Thailand17.02.130.746.80.022.577.53.4
Togo15.70.970.97.74.816.678.60.0
Tunisia21.70.027.927.123.321.755.00.0
Turkey43.30.522.233.30.044.555.50.7
Upper Volta12.60.056.121.23.619.177.36.5
Yemen Arab Republic5.60.074.711.48.35.686.10.0
Zaïre24.30.962.911.50.425.274.40.0
Zambia54.30.010.135.50.054.345.70.0
Average26.61.936.928.05.629.564.90.9
Table 15.Sixty-Three Developing Countries: Ratios of Taxes to GNP, 1972-76
CountryTotal Tax Over GNPPoll Tax Over GNPTotal Tax Plus Social Security1 Over GNPDirect Tax Over GNPIndirect Tax Over GNPIncome Tax Over GNPProperty Tax Over GNPTrade Tax Over GNPProduction Tax Over GNPOther Tax Over GNP
Afghanistan5.660.005.660.954.550.860.104.300.250.16
Algeria39.830.0039.8326.3713.4726.370.002.9010.570.00
Bangladesh5.800.005.800.574.140.570.001.782.361.08
Benin16.020.3517.523.7511.973.380.029.372.600.31
Bolivia11.840.0011.841.4410.071.170.276.383.690.33
Brazil18.100.0018.103.8513.393.550.301.0712.320.85
Burma7.560.007.562.733.982.730.001.392.590.86
Burundi9.280.849.283.096.142.110.133.772.370.05
Cameroon16.160.0116.163.0712.482.770.297.644.840.62
Central African Empire15.900.5315.904.819.793.960.335.893.891.30
Chile18.360.4126.016.2411.445.330.502.019.430.68
China, Republic of19.940.0019.945.3114.512.872.446.258.270.12
Colombia11.610.1411.614.474.914.320.012.452.452.24
Congo19.820.7119.828.1211.707.400.016.944.760.00
Costa Rica13.580.0017.633.2110.283.170.054.266.030.08
Dominican Republic15.040.0015.083.3910.843.350.038.052.790.81
Ecuador12.020.0012.023.148.593.060.076.002.590.29
Egypt18.150.8218.156.0211.894.680.526.835.130.23
El Salvador11.530.1511.532.988.552.160.674.573.980.00
Ethiopia10.110.0010.112.927.072.660.263.943.130.12
Gambia, The14.940.0014.942.1712.632.170.0012.630.000.13
Ghana14.240.0014.243.6110.543.600.007.333.210.09
Guatemala8.120.009.021.486.591.190.292.444.150.05
Guinea22.080.4023.746.1114.455.710.0011.782.591.52
Guyana31.270.0031.2710.2820.039.810.4714.695.340.96
Honduras11.520.0011.523.268.253.130.133.984.270.00
India13.870.0014.102.529.812.500.021.857.951.55
Indonesia16.260.5116.2611.504.5210.850.132.531.990.24
Iraq37.600.0037.6032.684.5932.380.302.871.720.33
Ivory Coast20.600.0020.603.9015.843.610.2814.111.730.86
Jamaica19.020.0019.669.678.179.030.643.344.841.18
Jordan19.340.0019.341.898.961.860.035.813.148.49
Kenya19.230.0019.237.7111.307.650.054.366.940.22
Korea13.570.0013.714.428.713.870.551.896.810.45
Liberia14.370.0014.376.047.945.780.254.523.420.39
Malawi10.050.0010.054.315.684.300.012.453.240.05
Malaysia22.480.0022.486.8912.196.750.147.145.063.39
Mali13.000.0013.001.637.821.560.066.441.383.55
Mexico8.630.158.634.703.704.550.000.712.980.23
Morocco18.620.0019.855.6212.325.140.486.955.370.68
Nepal5.370.005.371.214.170.320.882.241.920.00
Nicaragua11.090.2412.722.278.551.230.793.055.490.27
Pakistan11.390.0011.391.868.781.550.314.084.700.75
Panama11.580.0316.555.445.614.790.623.082.530.53
Paraguay8.760.119.962.005.541.090.812.902.651.21
Peru14.040.0016.924.728.644.050.682.985.670.67
Philippines10.130.6510.133.456.272.510.293.263.010.41
Rwanda10.020.0010.022.237.111.950.285.531.580.68
Senegal20.200.0020.204.2813.834.280.008.755.082.09
Sierra Leone17.010.1217.015.6611.205.550.008.083.120.14
Sri Lanka17.970.0018.073.4314.543.060.378.066.480.00
Sudan18.920.0018.922.3514.472.350.0010.523.%2.09
Swaziland23.090.0023.096.8215.746.740.099.865.880.53
Syrian Arab Republic11.290.1611.292.866.052.490.204.201.852.38
Tanzania18.940.0018.945.9612.995.860.104.518.480.00
Thailand13.960.4713.%3.1410.822.370.294.296.530.00
Togo12.380.0012.382.059.731.940.118.770.960.59
Tunisia20.680.0023.214.4911.384.490.005.765.614.82
Turkey16.190.1116.197.208.997.010.083.595.400.00
Upper Volta11.340.7311.342.178.771.430.006.362.410.41
Yemen Arab Republic7.670.007.670.436.600.430.005.730.870.64
Zaïre27.190.0027.196.8620.226.610.2517.103.120.11
Zambia30.800.0030.8016.7414.0616.740.003.1210.940.00
Average15.800.1216.295.159.814.770.255.554.260.84

Mr. Tait, Chief of the Fiscal Analysis Division of the Fiscal Affairs Department, is a graduate of the Universities of Edinburgh and Dublin. He was formerly a fellow of Trinity College, Dublin, and Professor and Chairman of the Economics Department at the University of Strathclyde. He is the author of The Taxation of Personal Wealth, The Value Added Tax, and of numerous chapters of books and articles in professional journals.

Mr. Grätz, economist in the Fiscal Analysis Division of the Fiscal Affairs Department when this paper was prepared, is currently in the Research Department. He is a graduate of the University of Vienna and was a scholar at the Institute for Advanced Studies and Scientific Research, Vienna. He is currently on leave from the Ministry of Finance of Austria.

Mr. Eichengreen, economist in the Fiscal Analysis Division of the Fiscal Affairs Department when this paper was prepared, is currently a graduate student in economics at Yale University. He was formerly on the staffs of the Brookings Institution and the Congressional Budget Office.

They assert, for example, that the economic, political, and institutional characteristics of individual countries are so unique that neither general theorizing nor comparative quantitative studies reveal more than they obscure. See Richard M. Bird, “Assessing Tax Performance in Developing Countries: A Critical Review of the Literature,” Finanzarchiv, Vol. 34 (No. 2, 1976), p. 256.

See, for example, Richard A. Musgrave, Fiscal Systems (Yale University Press, 1969), Chapters 4-7, pp. 91-206.

Jorgen R. Lotz and Elliott R. Morss, “Measuring ‘Tax Effort’ in Developing Countries,” Staff Papers, Vol. 14 (November 1967), pp. 478-99; Roy W. Bahl, “A Regression Approach to Tax Effort and Tax Ratio Analysis,” Staff Papers, Vol. 18 (November 1971), pp. 570-612; Raja J. Chelliah, “Trends in Taxation in Developing Countries,” Staff Papers, Vol. 18 (July 1971), pp. 254-331; Raja J. Chelliah, Hessel J. Baas, and Margaret R. Kelly, “Tax Ratios and Tax Effort in Developing Countries, 1969-71,” Staff Papers, Vol. 22 (March 1975), pp. 187-205 (Hereinafter this article is referred to as CBK.).

See a recent example of extended use in Organization for Economic Cooperation and Development, Revenue Statistics of OECD Member Countries, 1965-75 (Paris, 1977).

A convenient base to which a tax rate can be applied (e.g., wages and salaries paid in large-scale organizations) is often referred to as a “tax handle.” On tax handles, see Lotz and Morss,op.cit., passim; and Richard A. Musgrave and Peggy B. Musgrave, Public Finance in Theory and Practice (New York, 1973), p. 131.

Harley H. Hinrichs, A General Theory of Tax Structure Change During Economic Development (Harvard Law School, 1966); Lotz and Morss, op. cit.; and Musgrave, Fiscal Systems, op. cit.

CBK, op. cit., pp. 204-205.

CBK, op. cit., p. 195.

For a detailed discussion of limitations, see Chelliah, op. cit., pp. 298-300.

See Chelliah, op. cit., and CBK, op. cit.

International Monetary Fund, International Financial Statistics, Vol. 30 (July 1977); and Government Finance Statistics Yearbook, Vol. 1 (1977).

This is the same methodology employed by CBK, op. cit.

For the purpose of comparison, the same sample of countries used by CBK was employed in this section. Except for Argentina, Iran, Jamaica, Trinidad and Tobago, Singapore, and Venezuela, all countries covered in the present study have per capita incomes below $1,000.

See Organization for Economic Cooperation and Development, op. cit. In 1975, the average tax ratio in 23 developed countries in Europe and North America amounted to about 26 per cent net of social security and 34 per cent including social security.

Of the 22 countries with tax ratios above the 16.1 per cent average, 17 have tax indices greater than unity.

Indonesia, Singapore, Rwanda, Lebanon, Bolivia, Mexico, Trinidad and Tobago, the Philippines, Togo, Honduras, Guatemala, and Nepal.

Brazil, Tunisia, Sudan, the Republic of China, Zaïre, Morocco, Turkey, Malaysia, and Chile.

A “high” ITC index is greater than 1.10; a “medium” ITC index is between 0.84 and 1.10; a “low” ITC index is less than 0.84.

United States, Advisory Commission on Intergovernmental Relations, Measuring the Fiscal “Blood Pressure” of the States—1964–1975 (Washington, February 1977), p. 13.

The coefficient of rank correlation between the 1969–71 and 1972–76 rankings amounts to 0.72.

CBK, op. cit., p. 189.

Musgrave, op. cit., p. 111, and Hinrichs, op. cit.

The equation for these 25 countries is

T/Y=3.1331(1.02)+0.0294(1.22)(YpXp)+0.4073(14.73)XpR¯2=0.533F(2,22)=14.73

Musgrave, op. cit., p. 132.

For an alternative geographic treatment of tax effort, see W. Parmena, “A Method of Comparing Tax Effort in African Countries,” Eastern Africa Economic Review, Vol. 8 (December 1976).

Another way is to devise methods that cope with the two principal objections: (a) that where the proportion of the variance explained by economic variables is small, the assumption that the residual reflects noneconomic factors such as tax effort is less plausible than the suspicion that the explanatory variables are unsatisfactory proxies for taxable capacity, and (b) that each change in the sample involves a change in the standard against which all countries are measured. Two of the authors have devised measures using the level of taxation per capita instead of tax ratios and have obtained robust equations. See Alan A. Tait and Barry J. Eichengreen, “Two Alternative Approaches to the International Comparison of Taxation” (unpublished, International Monetary Fund, July 31, 1978).

Other Resources Citing This Publication