Article

Measuring “Tax Effort” in Developing Countries

Author(s):
International Monetary Fund. Research Dept.
Published Date:
January 1967
Share
  • ShareShare
Show Summary Details

THE QUESTION often arises whether there is scope for an increase in the level of taxation in a country as part of a stabilization program, for the mobilization of resources to finance a development program, or for other purposes. A relevant consideration in answering this question is how the tax effort of the country compares with that of other countries in similar circumstances.

This paper discusses certain criteria by which tax effort might be measured and ranks selected countries on the basis of these criteria.

Ratio of tax revenue to gross national product

Tax effort is measured by relating actual tax collections to some indicator of taxable capacity.1 In past discussions, considerable attention has been focused on the ratio of tax revenue to national income,2 with some suggesting there is an upper limit to this ratio. This formulation implies that total income is a proper indicator of taxable capacity and that tax effort is measured by the ratio of tax revenue to income.

Table 1 shows for 72 countries the ratio of tax revenue to one measure of national income—gross national product (GNP). The country with the highest tax ratio is ranked 1 and the lowest, 72. For most countries, the data are three-year averages.3 The rankings should be interpreted cautiously because they do not take account of certain economic and noneconomic factors which may be especially important in evaluating the tax effort of particular countries.4 It should be emphasized that a low or high ranking in tax effort does not imply that taxes should be increased or decreased. An informed opinion on that question should take into consideration economic conditions, the character and level of government expenditures, administrative capacity, and a variety of other matters.

Table 1.Ranking of 72 Countries According to Tax Effort Measured by Ratio of Tax Revenue to Gross National Product, Recent Years’ Average
RankTax RatioRankTax Ratio
Per centPer cent
France137.7South Africa3617.2
Sweden237.2Kenya3717.2
Germany334.8Ecuador3816.7
Norway434.6Trinidad and Tobago3916.7
Austria534.5Iran4016.3
Netherlands632.9Peru4116.0
Italy729.6Malagasy Republic4215.7
United Kingdom828.9Jamaica4315.6
Belgium928.7Cameroon4415.4
Denmark1028.7Mali4515.4
Iceland1128.0Panama4615.1
Canada1227.8Turkey4715.1
Israel1326.4China, Republic of4815.0
United States1426.2Ghana4913.9
New Zealand1526.0Uganda5013.9
Australia1623.6Costa Rica5113.8
Ireland1723.4Tanzania5213.8
Congo, Democratic Rep. of1822.5Nicaragua5313.5
Sudan5413.3
Algeria1922.4Chad5513.0
Uruguay2021.8Thailand5612.6
Brazil2121.4India5712.5
Switzerland2221.0Spain5812.2
Chile2320.9Niger5912.0
Finland2420.9Malawi6011.7
Malaysia2520.5Nigeria6111.3
Greece2620.4Colombia6210.9
Argentina2720.1El Salvador6310.9
Japan2819.4Philippines6410.8
Ceylon2918.6Paraguay6510.2
United Arab Republic3018.5Mexico669.9
Honduras679.9
Burma3118.4Haiti689.6
Iraq3218.4Guatemala699.3
Dominican Republic3317.9Korea709.0
Portugal3417.9Ethiopia718.3
Liberia3517.7Afghanistan725.9
Source: Derived from Table 5, Appendix I, p. 491.

Tax revenues of all levels of government—national, provincial, and local—are included. Taxes are defined as compulsory levies that are not necessarily related to particular benefits received. They are thus distinguished from fees, prices, grants, and other sources of government revenue that are either voluntary or are more akin to proprietary income. Tax revenues are interpreted to include profits and losses of fiscal monopolies5 and of marketing boards. For example, profits of salt, matches, sugar, tobacco, and similar monopolies are included on the grounds that the profits are an alternative to taxes. The tax figures include foreign exchange profits only when they are shown as revenue in the government budget. Fees are included when receipts from them are of such a size as to indicate an element of taxation or when it has been impossible to exclude them; license revenues are included. The data were obtained from local sources, from nonconfidential reports of the International Monetary Fund, International Bank for Reconstruction and Development (IBRD), U. S. Agency for International Development, Organization for Economic Cooperation and Development (OECD), and United Nations (UN).6 Only employee and nongovernment employer contributions to social insurance funds are included. Many data are based on a study of social security financing prepared in the Fund’s Fiscal Affairs Department.7

Ratios of taxation to GNP in market prices, such as those presented here, are based on the most readily available and widely understood measure of national income. Previous studies have also used GNP for this purpose. Three alternative measures of national income—gross domestic product (GDP), net national product (NNP), and national income at market price (NI)—would have certain theoretical advantages as an indicator of a country’s potential tax base. GDP includes income produced locally which accrues to nonresidents and excludes income received from abroad by residents, whereas GNP excludes the former and includes the latter. For developing countries, GDP frequently exceeds GNP by a considerable margin, and the income accruing to foreign investors and other nonresidents can be taxed. NNP and NI exclude capital consumption allowances and may therefore be regarded as a measure of the output that can be used for private and government purposes without reducing the capital stock. However, estimates of GDP, NNP, or NI are available at present for a much smaller number of countries than are estimates of GNP.

The main problem is not the choice of national income aggregate, but the quality of the available GNP figures. The unreliability of GNP estimates is indicated by a comparison of the 1953 GNP for 31 developing countries presented in the 1958 and the 1963 editions of the UN Yearbook of National Accounts Statistics. In the 1963 edition, 15 estimates for the 1953 GNP were changed from the 1958 edition—4 by more than 10 per cent—and 5 estimates were eliminated for one reason or another. Most revisions have been in an upward direction.

Need for improving the tax ratio measure

In addition to aggregate income, the denominator in the tax ratio, other factors affect a country’s taxable capacity.8 One of the most important is the level of economic development. Economic development is usually accompanied by a higher rate of literacy, increased monetization, and stricter law enforcement—all of which can be expected to increase taxable capacity. Economic development has many dimensions and cannot be measured precisely either by a single variable or by a simple combination of variables. However, one variable frequently used by economists to give a rough idea of the development stage is per capita income. Hence, one would expect taxable capacity and per capita income to move in the same direction.

There is another reason to expect a positive relationship between per capita income and taxable capacity. For two countries with the same total income but with a per capita income of, say, $50 in the first country and $1,500 in the second, taxable capacity is greater in the second because a smaller proportion of total income is required for subsistence needs and more “surplus” is available for taxation and other purposes. It follows that, if the two countries raise the same total amount of tax revenue and thus have equal tax ratios, the first country is making the greater tax effort.9

It appears that taxable capacity also increases with the size of the foreign trade sector for two reasons: First, it is administratively easier to tax trade inflows and outflows than domestic transactions. Second, the “degree of openness” in many countries, especially in early stages of development, indicates the relative importance of cash crops and subsistence agriculture, as well as the degree of urbanization and industrialization.10 This relationship suggests that openness is associated with conditions which facilitate internal taxation.

In this paper, explicit consideration is given to only three measures of taxable capacity: aggregate GNP, per capita GNP, and the size of the foreign trade sector.11 Other important factors which are believed to affect taxable capacity but which are not considered explicitly include the size distribution of income, the industrial origin of output, and the composition of government expenditures. In considering our rankings, the reader should keep in mind important deviations from the average in any of these factors.

A comment is warranted on the relationship between the measurement of tax effort and the statistical explanation of the actual level of tax revenues. Our approach was to select a few factors for which we felt allowances should be made in measuring tax effort. We used a statistical technique (regression analysis) to show how these factors are in practice associated with tax ratios in a large number of countries. The purpose of the statistical exercise was to establish an empirical basis for the weights to be assigned to the factors and thus to provide norms for our appraisals of tax effort. If the objective had been to find the best statistical explanation of tax revenues, we would have experimented with a number of other variables and would have chosen the formulations that best accounted for prevailing tax ratios, regardless of the normative significance of the variables.12

Allowances for per capita income and openness13

Our work indicates that a small difference in the formulation of the tax effort index can have a significant impact on the country rankings. Therefore, our procedure was to present several formulations and to rank countries on the basis of each. Presumably, countries receiving the highest tax effort rankings are taxing closer to their capacity than countries with lower rankings.

In Table 1, countries were ranked on the basis of their simple tax ratios, total taxes (T) divided by GNP (Y). Allowance has to be made for differences in per capita income (Yp). It is extremely difficult to make intercountry comparisons of Yp, partly because the GNP figures include differing portions of imputed value added to total output by subsistence sectors; there also is the problem of converting the Yp figures into standardized units for comparative purposes.14

For conversion of the per capita GNP into common units, we had these alternatives: (1) to convert the national currency figures into dollars at the official exchange rates and (2) to convert the figures by using a series of “exchange rates” on a purchasing-power parity basis derived in the United Nations. These UN “exchange rates” were “estimated by adjusting the official or free market exchange rates in 1938 in each country by the relative change in the level of prices from 1938 to the year in question, between the United States and the country concerned.”15 Both methods have weaknesses. Neither set of rates accurately measures the purchasing power of currencies for the different collections of goods actually used in different countries; at best, the exchange rates reflect relative prices of internationally traded goods. Official exchange rate adjustments sometimes lag behind relative price changes in countries, and hence overvalue currencies in countries where prices are increasing rapidly. An important drawback of using the 1938 “exchange rate” series as the base is that the price indices used for adjustment are not fully reliable. Some of the results obtained by using their series are implausible. For example, when Australian and New Zealand per capita incomes were converted to U. S. dollars by this method, the 1962 per capita income for the latter was $2,060, but only $818 for Australia.16

We elected, with certain exceptions,17 to use the official exchange rates. Therefore, for some countries, our estimates of per capita income in U. S. dollars are appreciably higher or lower than estimates based on the alternative series. Generally, an overestimate of per capita income results in too low a ranking in tax effort; an underestimate has the opposite effect. Readers may wish to allow for this possibility in interpreting our rankings.

Initially, we “took account of” per capita income by measuring the deviations from a straight-line estimate of the relationship between tax ratios and per capita income for a group of countries. Specifically, we made a least-squares estimate of the relationship:

With the resulting estimates of a1 and b1 we were able to estimate an “average” tax ratio for any country, given its per capita income. The actual tax ratios and our estimates were then compared; the country whose actual tax ratio exceeded the estimated one by the largest percentage was given the highest tax effort ranking, and the country whose actual tax ratio fell short of our estimate by the largest percentage was given the lowest ranking.

It was suggested earlier that, in addition to per capita income, some allowance should be made for the importance of foreign trade in a country’s economy. Hence, we compared countries also on the basis of the following relationship:

where FY is the sum of exports (f.o.b.) and imports (c.i.f.) as a percentage of GNP.

Separate calculations were made for low-income countries and high-income countries.

Statistical results

Table 2 ranks 72 countries on the basis of the three measures discussed above. In the tax ratio column, countries are ranked in descending order on the basis of their tax ratios in Table 1. In the equation (1) column, countries are ranked on the basis of the percentage difference between their actual tax ratios and ratios estimated with allowance for per capita income differences.18 In the equation (2) column, countries are ranked with allowances for the degree of openness and for per capita income.

Table 2.Ranking of 72 Countries According to Tax Effort Measured by Different Methods, Recent Years’ Average
Tax Effort Rank Measured by
Tax ratioEquation (1)1Equation (2)2
France165
Sweden21820
Germany31410
Norway41212
Austria522
Netherlands6815
Italy754
United Kingdom83027
Belgium92036
Denmark103738
Iceland113340
Canada125147
Israel132421
United States146464
New Zealand154443
Australia165752
Ireland171930
Congo, Democratic Rep. of1813
Algeria1937
Uruguay20139
Brazil2141
Switzerland226669
Chile232314
Finland245553
Malaysia25725
Greece262111
Argentina273117
Japan283426
Ceylon291013
United Arab Republic30118
Burma3196
Iraq321631
Dominican Republic332218
Portugal342824
Liberia351748
South Africa364137
Kenya371528
Ecuador382616
Trinidad and Tobago394757
Iran403229
Peru413535
Malagasy Republic422722
Jamaica434554
Cameroon442933
Mali452519
Panama465051
Turkey474023
China, Republic of483634
Ghana494642
Uganda503845
Costa Rica515655
Tanzania523944
Nicaragua535456
Sudan544239
Chad554341
Thailand564946
India574832
Spain586158
Niger595250
Malawi605359
Nigeria615849
Colombia626360
El Salvador636068
Philippines645962
Paraguay656563
Mexico667067
Honduras676770
Haiti686261
Guatemala697171
Korea706865
Ethiopia716966
Afghanistan727272
Note: For the usual significance tests, see Appendix IV (p. 497).

To get a better idea of the major differences between the three measures, consider the following examples. On a simple tax ratio basis, Israel ranks 13. However, when account is taken of Israel’s per capita income, which is high relative to that of other countries, its tax effort rank falls to 24. In contrast, the tax effort of the Malagasy Republic, a low-income country, appears considerably greater when allowance is made for income differences.

Consider now the effect of allowing for differences in the importance of foreign trade. The fact that c2 in equation (2) equals 0.0790 means that on average, for every percentage point increase in the importance of foreign trade as measured by FY, the tax ratio increases by 0.0790 percentage point. A comparison of the last two columns of Table 2 indicates the effect of allowing for the difference in FY. The tax effort ranking of Malaysia, a country with a large foreign trade sector (FY = 80.8), falls as a result of allowing for the openness factor. In contrast, the ranking of Japan, a country less dependent on foreign trade (FY = 18.7), is improved.

It may be argued that tax effort standards for high-income and low-income countries should not be the same. Accordingly, countries were divided into two groups on the basis of per capita GNP above and below US$800 in the last year of those covered, and the intragroup tax performances were compared.

The results for the countries in the low-income group are given in Table 3. It should be noted that because of the different weights attached to YP and -y in this group compared with those in the equation for all countries, it is possible for the intragroup rankings to differ. For example, according to equation (1) in Table 2, Argentina ranked ahead of Iran, whereas in Table 3, the rankings are reversed.

Table 3.Ranking of 52 Low-Income Countries According to Tax Effort Measured by Different Methods, Recent Years’ Average
Tax Effort Rank Measured by
Tax ratioEquation (1)1Equation (2)2
Congo, Democratic Rep. of112
Algeria224
Uruguay386
Brazil431
Chile5149
Malaysia6415
Greece7128
Argentina82113
Ceylon967
United Arab Republic1075
Burma1153
Iraq121020
Portugal131816
Dominican Republic141312
Liberia151132
Kenya16917
South Africa172725
Ecuador181610
Trinidad and Tobago193239
Iran202019
Peru212324
Malagasy Republic221714
Jamaica233036
Cameroon241922
Mali251511
Panama263535
Turkey272618
China, Republic of282223
Uganda292430
Ghana303129
Tanzania312528
Costa Rica323938
Nicaragua333837
Sudan342826
Chad352927
Thailand363431
India373321
Spain384341
Niger393633
Malawi403740
Nigeria414034
El Salvador424248
Colombia434543
Philippines444144
Paraguay454645
Honduras464750
Mexico475049
Haiti484442
Guatemala495151
Korea504846
Ethiopia514947
Afghanistan525252
Note: For the usual significance test, see Appendix IV (p. 497).

Among high-income countries, an insignificant relationship was indicated between per capita GNP and the tax ratio (see Appendix IV, p. 497). The addition of the index of openness did not improve the statistical results. In short, apparently neither of our techniques of estimation is of much assistance in measuring the taxable capacity of high-income countries. The tax ratio of a high-income country is more an index of political preference for the appropriate size of the government’s role than an index of taxable capacity. For this reason the high-income countries are not considered further as a separate group in this paper.

For the low-income countries, both per capita GNP and the degree of openness were found to be significantly related to the tax ratio, although these two factors did not account for a large part of the variance among ratios. We conclude that tax revenues in low-income countries are often limited by taxable capacity and that our methods give a better indication of tax effort than do simple tax ratios.

In Table 4, low-income countries are divided into three groups—high, average, and low tax effort—on the basis of the percentage difference in tax revenue of the individual countries from the group norm with allowances for differences in per capita GNP and openness (as specified in equation (2)). To understand the meaning of Table 4, consider the following examples. Brazil, which ranks first in tax effort, collects considerably more revenue than would be expected for a country with its per capita GNP and foreign trade ratio. In contrast, Afghanistan’s tax revenues are much lower than the amount that would be predicted by the equation underlying the table, that is, solely on the basis of per capita income and openness.

Table 4.Ranking of 52 Low-Income Countries, in Groups, According to Tax Effort Measured with Allowance for Per Capita Gross National Product and Openness1
Tax Effort RankTax Effort Rank
High tax effort countriesAverage tax effort countries (concluded)
Brazil1
Congo, Democratic Rep. of2Tanzania28
Burma3Ghana29
Algeria4Uganda30
United Arab Republic5Thailand31
Uruguay6Liberia32
Ceylon7Niger33
Greece8Nigeria34
Chile9Panama35
Ecuador10
Mali11Low tax effort countries
Dominican Republic12Jamaica36
Nicaragua37
Average tax effort countriesCosta Rica38
Argentina13Trinidad and Tobago39
Malagasy Republic14Malawi40
Malaysia15Spain41
Portugal16Haiti42
Kenya17Colombia43
Turkey18Philippines44
Iran19Paraguay45
Iraq20Korea46
India21Ethiopia47
Cameroon22El Salvador48
China, Republic of23Mexico49
Peru24Honduras50
South Africa25Guatemala51
Sudan26Afghanistan52
Chad27

Table 4 gives only a rough indication of the relative tax performance of countries. Only two of the several factors that should be allowed for in a full assessment of a country’s tax performance have been taken into account. Nevertheless, we believe that Table 4 comes closer to a meaningful comparison of tax performance than is possible through an examination of simple tax ratios.

Geographic differences

Examination of Table 4 reveals wide differences among countries in the same geographic area but suggests a clustering that is worthy of comment. Of 11 Central American and Caribbean countries in the table, 9 are in the low tax effort group. The low average ranking of these countries may be due to various political, cultural, and historical factors which condition attitudes toward the role of government and taxation, as well as to economic factors which were not investigated in this study. Further details for countries grouped by geographic area appear in Table 7, Appendix III (p. 496).

APPENDICES
I. Basic Statistics
Table 5.Average Gross National Product, Tax Revenues, Per Capita Gross National Product, Population, Exchange Rates, and Openness Percentage, 72 Countries, Recent Years’ Average
PeriodAverage GNPAverage Tax Revenues Including MunicipalitiesAverage Openness1Average PopulationAverage Per Capita GNPAverage Exchange Rates for Period
Million U.S. dollarsPer centMillionsU.S. dollarsUnits per U.S. dollar
Afghanistan1963-656013623.415.63963.4
Algeria1963-642,77562251.611.52424.94
Argentina1963-6515,2633,04715.322.0692157.30
Australia1964-6621,3214,05628.411.11,9130.898
Austria21963-658,5302,94339.37.21,18125.88
Belgium21963-6513,7504,44784.09.71,41649.70
Brazil1963-6519,7473,51716.578.8194156.363
Burma1963-651,66630629.724.2694.791
Cameroon1964-656109445.15.2116246.853
Canada1963-6544,22412,27832.819.32,2931.0766
Ceylon1963-651,56229047.410.91434.758
Chad1963-652042738.83.363246.853
Chile1963-655,3161,08722.58.46322.773
China, Republic of1963-652,36635432.412.119640.10
Colombia1962-645,35357819.216.93167.8373
Congo, Democratic Rep. of1964-661,19526958.315.577150
Costa Rica1963-655547645.91.43996.62
Denmark21963-658,9512,57651.84.71,8976.91
Dominican Republic1962-6498417634.13.42911.00
Ecuador1963-651,03317227.94.920618.18
El Salvador1963–657408048.02.82642.50
Ethiopia1964–661,30910818.522.6582.50
Finland21962–646,5001,35738.24.51,4313.22
France21963–6587,93533,12821.548.41,8154.90
Germany21963–65103,58836,02730.356.11,8453.99
Ghana1963–651,89526335.37.52520.357
Greece1962–644,63494623.38.554630
Guatemala1963–651,30712128.14.33031.00
Haiti1963–653303223.64.6725
Honduras1963–654664644.62.22112
Iceland21962–643289272.30.21,72943.06
India196337,4645,43010.9460.5814.785
Iran1964–665,74593733.422.825175.75
Iraq1962–641,64530366.77.42220.357
Ireland21962–642,34854960.52.88250.357
Israel1963–653,09881937.72.51,2493.00
Italy21963–6552,75215,60125.651.11,032624
Jamaica1964–6683513058.41.74790.357
Japan21963–6569,10811,17118.795.9720359.5
Kenya1963–6577313360.89.1857.14286
Korea1962–642,94927516.626.9110172
Liberia1963–6522039100.91.02191.00
Malagasy Republic1963–6563610035.06.2103247
Malawi1964–6617620757.83.9450.358
Malaysia41961–632,04541880.88.62373.06
Mali1964-662864430.34.662246.853
Mexico1962-6415,8601,56714.439.939712.49
Netherlands21963-6516,9745,57974.312.11,3983.60
New Zealand1964-664,8031,51441.72.61,8532.7586
Nicaragua1963-655187.049.41.63247.0
Niger1964-652613022.43.283246.853
Nigeria1963-653,40741036.056.5600.357
Norway21963-656,3142,18651.83.71,7097.16
Panama1963-655828841.61.24841.00
Paraguay1963-653984021.32.0202126
Peru1963-653,59157534.711.331726.82
Philippines1964-655,04254631.331.81593.723
Portugal51962-643,11355835.49.034428.90
South Africa1963-659,2621,59434.617.05431.393
Spain51962-6415,5601,90517.631.150059.96
Sudan1963-651,321175636.113.11010.348
Sweden21963-6517,5726,53642.67.72,2925.18
Switzerland21963-6512,7922,68848.25.92,1824.32
Tanzania1963-657339653.99.8757.14286
Thailand1963-653,53944633.629.711920.83
Trinidad and Tobago1964-6664511570.31.06451.714
Turkey1963-657,6871,15713.230.02569.00
Urganda1963-656028455.97.6807.14286
United Arab Republic1962-643,89972231.728.01390.435
United Kingdom21963-6592,42026,68830.554.21,7050.357
United States1963-65605,400158,6877.1194.53,1121.000
Uruguay1962-641,38729327.72.652516.66
Sources: Tax revenue from IMF, IBRD, and local sources, except as noted. GNP, exports, imports, and population from IMF, International Financial Statistics.
II. Introducing a Higher Rate of Progressivity into the Tax Effort Formulation

To express the relationship between taxable capacity and per capita income, we selected equation (1)

implying that

Thus, we assumed that the incremental tax ratio was growing at a constant absolute rate with per capita income. The progressivity can be further increased if the incremental tax ratio grows at an increasing absolute rate with per capita income.19 Since equation (2) has openness as the independent variable, a double logarithmic formulation was not feasible to test this relationship, and the following equation was used:

The incremental tax ratio will be

When ca is positive, TY will grow at an accelerating rate with per capita income. The explanatory power of equation (3) is higher than that of equation (2). Thus, there is reason to prefer equation (3) over equation (2) on the basis of the explanatory value. Table 6 shows the results of the two calculations for the low-income countries.

Table 6.Comparison of Tax Effort of Low-Income Countries Measured on the Basis of Equation (3) and Equation (2)
Equation (3)1Equation (2)2
RankFractional residualRankFractional residual
Per centPer cent
High tax effort countries3
Brazil140.5139.1
Congo, Democratic Rep. of231.6233.3
Algeria330.5428.8
Burma427.8329.8
United Arab Republic526.4526.2
Uruguay623.8623.7
Ceylon720.5720.4
Greece818.7819.1
Ecuador918.71016.7
Dominican Republic1018.31115.6
Average tax effort countriess3
Chile1114.7917.8
Turkey1214.316-1811.3
Portugal1314.016-1811.3
Malaysia1413.714-1512.3
Mali1513.51216.3
Malagasy Republic1612.414-1512.3
Iran1712.3199.7
Iraq1810.0208.5
Kenya199.316-1811.3
Argentina208.91314.6
China, Republic of217.2235.5
Peru227.0243.9
Cameroon235.7226.4
India244.6216.6
South Africa25—1.025—0.5
Sudan26—3.926—2.6
Ghana27—3.929—6.9
Chad28—7.227—3.9
Liberia29—7.732—8.8
Thailand30—8.731—8.1
Tanzania31—9.128—6.1
Uganda32—9.530—7.0
Nigeria33—13.133-34—10.7
Panama34—13.335—14.4
Niger35—14.433-34—10.7
Low tax effort countries3
Jamaica36—17.136—18.2
Nicaragua37—18.637—22.1
Costa Rica38—19.238—22.2
Colombia39—26.043—30.8
Spain40—27.441—28.7
Philippines41—28.544—31.2
Paraguay42—28.745—31.8
Trinidad and Tobago43—29.039—23.9
Malawi44—30.540—25.4
Haiti45—33.842—30.2
Korea46—38.046—37.0
El Salvador47—41.848—45.6
Mexico48—42.749—47.5
Ethiopia49—49.947—44.8
Honduras50—50.450—53.3
Guatemala51—53.651—59.0
Afghanistan52—116.652—106.9
Note: For the usual significance test of the empirical results, see Appendix IV (p. 497).

Compared with equation (2), equation (3) makes the tax effort lower for the countries with the lowest income (e.g., Uganda, Chad, Nigeria, Tanzania, and Niger) and for Chile and Argentina, countries with the highest income, while the middle per capita income countries show a better tax effort than from the straight-line measure (e.g., Peru, Ghana, Nicaragua, Colombia, El Salvador, Mexico, Guatemala, and Paraguay).

III. Geographic Grouping
Table 7.Geographic Groupings of Low-Income Countries Ranked on Basis of Percentage Difference Between Actual tax Revenue and Revenue Obtainable with Average Effort, After Allowing for Differences in Per Capita Gross National Product and Openness
Equation (2)Equation (2)
RankFractional residualRankFractional residual
Per centPer cent
Central America and Caribbean countriesAsia, Africa, and Middle East (concluded)
Dominican Republic115.6Mali616.3
Panama2—14.4Malagasy Republic713.5
Jamaica3—18.2Malaysia812.3
Nicaragua4—22.1Kenya911.3
Costa Rica5—22.2Iran109.7
Trinidad and Tobago6—23.9Iraq118.5
India126.6
Haiti7—30.2Cameroon136.4
El Salvador8—45.6China, Republic of145.5
Mexico9—47.5South Africa15—0.5
Honduras10—53.3Sudan16—2.6
Guatemala11—59.0Chad17—3.9
Average residual for group—29.2Tanzania18—6.1
Ghana19—6.9
South AmericaUganda20—7.0
Brazil139.1Thailand21—8.1
Uruguay223.7Liberia22—8.8
Chile317.8Niger23—10.7
Ecuador416.7Nigeria24—10.7
Argentina514.6Malawi25—25.4
Peru63.9Philippines26—31.2
Colombia7—30.8Korea27—37.0
Paraguay8—31.8Ethiopia28—44.8
Afghanistan29—106.9
Average residual for group6.7Average residual—2.8
Asia, Africa, and Middle EastEurope
Greece119.1
Congo, Democratic Rep. of133.3Portugal211.3
Turkey311.3
Burma229.8Spain4—28.7
Algeria328.8
United Arab Republic426.2Average residual for group3.3
Ceylon520.4
IV. Summary of Regression Results

All countries

Equation (1)

Equation (2)

Equation (3)

High-income countries

Equation (1)

Equation (2)

Equation (3)

Low-income countries

Equation (1)

Equation (2)

Equation (3)

Evaluation de l’effort fiscal dans les pays en voie de développement

Résumé

On se demande parfois, étant donné la situation d’un pays, s’il est possible de relever le niveau des impôts dans le cadre d’un programme de stabilisation, de mobiliser des ressources pour financer un programme de développement ou pour d’autres objectifs. Pour répondre à cette question, il convient de comparer l’effort fiscal de ce pays avec ceux des autres pays se trouvant dans une situation analogue.

Le présent document examine en premier lieu les divers facteurs dont on devrait tenir compte pour juger un effort fiscal. Etant donné que le rapport qui existe entre les recettes fiscales d’un pays et son produit national brut (PNB) constitue une évaluation très grossière de son effort fiscal, on s’efforce alors de formuler, en termes quantitatifs, un indice de l’effort fiscal légèrement plus significatif.

Cette formulation tient compte des différences qui existent dans le quotient individuel de revenu et dans l’importance du secteur du commerce extérieur en établissant la régression du coefficient impôts/PNB de chaque pays sur le revenu par habitant et le degré d’ “openness” de l’économie. On utilise alors la différence entre le coefficient fiscal évalué avec cette méthode et le coefficient effectif de chaque pays pour classer ceux-ci suivant leur effort fiscal.

Les auteurs arrivent à la conclusion que l’effort fiscal mesuré de cette façon n’a guère de sens pour les pays industrialisés. C’est pourquoi la dernière partie de cette étude porte uniquement sur les pays en voie de développement.

Medición del “esfuerzo tributario” de los países en desarrollo

Resumen

Frecuentemente surge el problema de si hay posibilidades de aumentar el nivel de los impuestos de un país dado como parte de un programa de estabilización, para movilizar los recursos con que financiar un programa de desarrollo, o para otros fines. Un aspecto importante a considerar para contestar este punto es el de comparar el esfuerzo tributario del país con el de otros países que se encuentren en circunstancias similares.

Este trabajo comienza con un examen de los diversos factores que deben tenerse en cuenta al aquilatar el esfuerzo tributario. Como la relación entre los impuestos de un país y su producto nacional bruto (PNB) ofrece una medida demasiado imprecisa de su esfuerzo tributario, se trata de formular un índice en términos cuantitativos que exprese algo más acerca de ese esfuerzo.

Dicho procedimiento toma en cuenta las diferencias en el ingreso por habitante y en la magnitud del sector comercio exterior, formulando un análisis de regresión de la razón impuestos/PNB de cada país en función del ingreso por habitante y de la magnitud del aludido sector en relación con el PNB. La diferencia entre la relación impuestos/PNB calculada de este modo y la relación efectiva correspondiente a cada país se utiliza entonces para catalogar a los países según su esfuerzo tributario.

Se llega a la conclusión de que esta forma de medir el esfuerzo tributario dice poco en cuanto a los países desarrollados y, por lo tanto, la última parte del estudio no se refiere a ellos.

Mr. Lotz, economist in the Fiscal Affairs Department, is a graduate of Copenhagen University. He was formerly with the Economic Consultant of the Danish Tax Department.

Mr. Morss, also economist in the Fiscal Affairs Department, has a doctorate from Johns Hopkins University. He was formerly an assistant professor of economics at the University of Michigan.

The U.S. Advisory Commission on Intergovernmental Relations, Measures of State and Local Fiscal Capacity and Tax Effort (Washington, 1962); R. Desai, “Fiscal Capacity of Developing Economies” in Fiscal Policy for Economic Growth in Latin America, papers and proceedings of a conference held in Santiago, Chile, December 1962 and issued by the Joint Tax Program of the Organization of American States, Inter-American Development Bank, and Economic Commission for Latin America (Baltimore, 1965), pp. 43–64,

See C. F. Bastable, Public Finance, 3rd ed. (London, 1903), pp. 136–37; Colin Clark, “Public Finance and Changes in the Value of Money,” The Economic Journal, Vol. LV (1945), pp. 371–89; Colin Clark, Taxmanship: Principles and Proposals for the Reform of Taxation, Institute of Economic Affairs, Hobart Paper No. 26 (London, 1964).

Table 5, Appendix I (p. 491), indicates the period for each country.

Also, it should be noted that several countries have had increasing or declining tax ratios during the period included. For these countries the tax effort ranking will not reveal the situation in the most recent year.

The distinction between a fiscal monopoly and a public enterprise is difficult to make. Here, the local usage of the terms has been the decisive factor.

The figures in Table 5, obtained from OECD and UN publications, are national income definitions of the revenues. Hence they do not include capital levies and are not strictly on a cash basis. The differences, however, are usually small and have been disregarded. Since these sources have been used mostly for developed countries, they have only a very small impact on findings for developing countries.

Franco Reviglio, “Social Security: A Means of Savings Mobilization for Economic Development,” Staff Papers, Vol. XIV (1967), pp. 324–68.

Josiah Stamp, Wealth and Taxable Capacity (London, 1930); Lewis H. Kimmel, Taxes and Economic Incentives, The Brookings Institution (Washington, 1950), pp. 5–16; Richard Goode, “An Economic Limit on Taxes: Some Recent Discussion,” National Tax Journal, Vol. V (1952), pp. 227–33; P. D. Ojha, ‘Taxable Capacity in Underdeveloped Economies,” Journal of the University of Bombay, Vol. XXIV, Part 1, July 1955, pp. 22–41.

Richard Bird, “A Note on Tax Sacrifice’ Comparisons,” National Tax Journal, Vol. XVII (1964), pp. 303–308.

Harley H. Hinrichs, A General Theory of Tax Structure Change During Economic Development, Law School of Harvard University (Cambridge, Massachusetts, 1966), pp. 7–31.

The same choice of factors is found in “Tax Potential and Economic Growth in the Countries of the ECAFE Region,” presented at the Fourth Workshop on Problems of Budget Reclassification and Management, Economic Commission for Asia and the Far East, Bangkok, August 1966.

The usual significance tests for our regression equations are presented in Appendix IV (p. 497).

For underlying statistical data, see Table 5, Appendix I (p. 491).

Everett E. Hagen, “Some Facts About Income Levels and Economic Growth,” The Review of Economics and Statistics, Vol. XLII (1960), pp. 62–67.

United Nations, Yearbook of National Accounts Statistics, 1965. The year 1938 was chosen because exchange rates were thought to be relatively close to equilibrium values because of the absence of restrictions in that year.

Countries in which the official exchange rates set a significantly lower value on the local currency than do those using 1938 rates as a base are Burma, Italy, the Netherlands, New Zealand, Norway, Portugal, South Africa, and Turkey. Countries in which the opposite is true are Australia, Brazil, Chile, Costa Rica, Dahomey, El Salvador, Finland, Ghana, and Iraq.

Brazil, Chile, Colombia, and the Philippines. In these countries there was a difference between the official exchange rate and the average rate used for commodity transactions.

Equation (1) is a linear formulation, that is, it relates absolute changes in tax ratios to absolute changes in per capita income. Using equation (1) for T normative purposes presupposes that TY should increase by the same amount for equal absolute changes in Yp at all levels of Yr. For the application of a formulation in which TY changes at an accelerating rate as Y, increases, see Table 6, Appendix II (p. 494).

See Alan T. Peacock and Gerald Hauser, “An Agenda for Analysis of Fiscal Systems in Southern European Countries,” Government Finance and Economic Development, edited by them, Organization for Economic Cooperation and Development (Paris, 1965), pp. 260–61; Henry Aaron, “Some Criticisms of Tax Burden Indices,” National Tax Journal, Vol. XVIII (1965), pp. 313–16.

*

Coefficient is significantly different from zero at the 95 per cent level. The number in parentheses is the regression coefficient divided by its standard error.

Other Resources Citing This Publication