Tunisia - Economic growth and tax elasticity : the Tunisian experience

 12: FILE ,COpy CONFIDENTIAL Report No. EC - 1 72 ~L-:; CoPY No., ~I To be returned to. GENERAL FILES immediately after use. This report is available only to those members of the staff to whose work it relates. Any further release must be authorized by the department head concerned. {"~I INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT INTERNATIONAL DEVELOPMENT ASSOCIATION ECONOMIC GROWTH AND TAX ELASTICITY -- THE TUNISIAN EXPERIENCE October 30, 1969 Economics Department o  (I CURRENCY EQUIVALENT U. S. $1.00 = D. 52 )/ I ..!  No. 61.EC INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT ICONFIDENTV.L I FORM (7.65) INTERNATIONAL DEVELOPMENT ASSOCIATION ECONOMIC COMMITTEE FILE COpy Ec/o/69 - 108 October 30, 1969 Tunisia: Economic Growth and Tax Elasticity 1. The attached draft study in green cover "Economic Growth and Tax Elasticity--The Tunisian Experience" (EC-172, dated October 30, 1969) was prepared in the Economics Depart- ment by Mr. Gabriele Sciolli. 2. It is intended to distribute the study, subject to final revisions and editing, to the ~~ecutive Directors as a grey-cover report. 3.~ Oral comments should be conveyed to Mr. Please (extension 4081). Written comments (two copies) should be sent to me, with an additional copy to Mr. Please, by the close of business on Monday, November 17,1969. C. F. Owen SeCl"etary Attachment Secretary's Department DIS T RIB UTI 0 N Messrs. Friedman Baldwin Lerdau Kamarck Bell Lipkowitz Adler Collier McDiarrnid Krishnaswamy de Vries Ross Rist de Wilde Sacchetti stevenson Gulhati Thompson King (B.B.)  ECONOMIC GROWTH AND TAX ELASTICITY -- THE TUNISIAN EXPERIENCE Table of Contents Page No. INTRODUCTION 1 CHAPTER I. . TAX ELASTICITY -- ITS ROLE IN THE DEVELOP1{ENT PROCESS 4 CHAPTER II. THE HISTORICAL EXPERIENCE 11 A. structure of tbe Revenue System 11 B. Trends in General Governmer.t Revenue 11 C. Reclassification of Tax Revenue 13 D. The Sensitivity of the Tax System to Changes in the Structure of the Economy 17 E. Elastici ty of the Individual Taxes 23 - Import Taxes 23 - Taxes on Domestic Production 29 ... Income Taxes 31 - Agricultural Taxes 39 - Export Taxes 41 - Consumption, Registration and Services Taxes 43 CHAPTER III. PROSPECTS UNDER PRES:FlJT TAX STRUCTURE 45 CHAPTER IV. SCOPE FOR DIPROVEMENTS AND POSSIBILE DIFFICULTIES 51 ,. CONCLUSIONS 55 STATISTICAL APPENDIX •  E@rO,iiIC JlHO~fl'H AND TAX ELASTICITY THE TUNIS IAN lAAPERIEN'CE Introduction 1. The elasticity of the tax system has arisen in Tunisia as a problem ar~,,~ within the discussion of the public savings perfonnance. Between 1960 and 1968, savings of the General Goverr~ent, including the Social Security System and Local Authorities, were on the average only about 3 percent of GDP at market prices, reaching a lOW' of 1.6 percent in 1968. The low level of public savings is particularly striking when one considers that Tur~sia has one of the highest ratios of revenue to GDP among developing countries of comparable per capita income and degree of openness (about 27 percent in 1967). 2. This implies that, over a number of years, savings performance in the country has been essentially an expenditure problem. In fact, between 196, and 1968 current expenditures increased by 12 parcent per annum as against a growth rate of GDP of only 5 percent. Had revenue in 1968 remained at the peak 1966 level of 27.4 percent of Gnp rather 'bhan declining to 26.3 percent, public savings would have still been less than 3 percent of GDP. 3. It remains true, however, that government revenues in certain years have failed to rise to expected levels or have not increased as fast as ODP. This attracted the attention of the Tunisian authorities as well as previous Bank missions and it was felt that a closer look should be given to the apparent lack of'. elasticity of the Tunisian Tax system, particularly at the Central Government level. The ratio of Central Government revenue to GDP did not change between 1960 and 1968 while during this time there were a number of increases in tax rate~ as well as a broadening of the basis of same important taxes (public enterprises were submitted to the corporation income tax in 1966). As we shall see late:r'.t the lack of elasticity between 1960 and 1968 is artificially magnified be cause 1960 was an abnormally high year followed in 1961 .. ,~ by a sharp decline which lasted until 1963. But after 1966, the ratio of revenue to GDP declined again with a substantial drop in 1968. GDP at market prices rose at 6.2 percent between 1966 and 1968 while Central Govemnent revenue only increased by 2., percent.  - 2 - 4. The purpose ot this study is to try and find out the causes explaining the apparent lack ot elasticfty of the Tunisian Tax system. We use the expression "apparent" because, as it will become clear late~ on, the very concept ot elasticity, as currently used, is rather ambigu- ous and the failure of total revenue to grow in line with GDP (if not taster) may have other reasons than a lack of elasticity of individual taxes. S. One should like to stress that no attempt is made here to describe in detail the Tunisian Tax system. This role is ads'luately performed by a number of documents to which reference will be made at the appropriate place. OUr attention will be rather concentrated on those taxes which are immediately relevant for our analysis and even then, our main interest will be the relation between their behavior and the change in the economic structure of the country rather than their technical aspects. 6. Anyone embarking on a study of the tax elasticity in a given country is presented wi th.a choice between concentrating on analysis of the historical experience, i.e., trying to reconstruct the built-in elasticity of the tax structure, and making an attempt to forecast the future relation between the growth in revenue and in economic activity on the basis of the present tax structure. This second aspect is in our opinion the most important one of real interest to both governments and international aid-giving institutions. The last two chapters of this study will try to deal with this problem. 1. The historical analysis should obviously serve as a basis in- sofar as we have to ask ourselves whether the factors which in the past were responsible for a possibly low built-in elasticity will still be present in the years to ccme. However, we should like to underline the great difficulties which one encounters when trying to separate the effects of changes in the tax structure fram the built-in elasticity of the tax system. This is particularly true in Tunisia where there bas been practically no year in which either the rates or the coverage of major taxes remained unchanged. B. These difficulties are well known. For one thing, there is no way of estimating by hat much an increase in tax rates has affected the propensity to tax evasion. Therefore if one simply reduces the amount of collection in one year by the percentage increase in the rates, one is not certain to obtain the collection in the absence of that increase. Secondly, collection of tax arrears distorts the natural trend of various taxes and it is often impossible to separate them. Thirdly, improved tax administration can increase revenue beyond the expected elasticity and here again there is no way of establishing the quantitative impact of thiS factor. All these difficulties point to the conclusion that it is in fact almost impossible to "reconstruct" a past trend in revenue which is un- affected by changes in the tax structure.  - 3- 9. Similar difficulties arise for the identification of the appropriate basis for each tax. Very often one is forced to accept second best solutions like GOP excluding agriculture as a proxy for individual and corporate income, or trends in private consumption for specific excise taxes. At the same time one should be aware of the fact that, when it comes to estimate the future elasticity of the tax system, the disaggregation of the sectors of origin of var-ious taxes is only justified to the extent in which it is possible to make equally disaggregated projections of the tax basis. • 10. A final difficulty is met when trying to calculate tax elasticity over a number of years. The results vary significantly according to the initial and terminal years which one chooses. They also vary according to. whether one takes two- or three-year averages for the calculations. The conclusion' 'is that one can only iden'Gify the most obvious cases of poor elasticity while in many other cases the implicit margin of error prevents one from taking at face value apparent elasticity coefficients slightly below or above unity. 11. A final observation on the approach to the study of the Tunisian experience. To the extent possible we have tried to identify the impact on the tax structure of the major changes in the structure of the Tunisian economy. In fact, as we shall see later, the elasticity of a tax systan depends to a large extent on its own structure from the point of view of the economic sectors which constitute the basis of major taxes. This approach gives good results for taxes which have their basis in a specific sector of the economy_ Unfortunately, in the case of direct taxes, due to the absence of information on trends in incomes and profits, one is forced to take, as basis for the tax, such groad magnitudes as GDP excluding agricu1ture~ 12. Tax elasticity has become a very popular and important concept to those who deal with problems of developing countries. It was felt that the study of the Tunisian experience pr-ovided a good opportunity for a broader presentation of the subject, both from the point of view of its conceptual framework and of its policy implication in the development process. To this general discussion is devoted the first chapter of the paper. The fact that the role of tax elasticity is being questioned in the course of such discussion should not give the impression of a dichotomy with the analysis of the Tunisian experience which has been motivated by and objective concern for insufficient revenue. Anticipating a recurrent theme in this sutdy, the lesson to be drawn from the Tunisian experience is that low tax elast;i,ci ty is sometimes in the nature of things and th at government should learn to live with it by following an expenditure policy attuned to revenue growth.  Chapter I TA.X EJ.•ASTICITY - ITS ROLE IN THE DEVELOPHENT PROCESS 13. A clarification of the terminology currently used when discussing the relations between growth in revenue and in economic activity is necessary for our analysis. Elasticity of the tax system is often referred to as the re~onsiveness of government revenue to the growth of the economy, usually iden- tified with GDP or GNP. More rigorously, the former relation should be defined as tax buoyancy in so far as it does not take into account possible changes in tax structure. Tax elasticity, on the contrary, is the built-in responsiveness • of a given tax structure to changes in economic activity. A tax system may have a lower than one built-in elasticity but a higher than one buoyancy as a result of increases in tax rates on introduction of netv taxes. 14. An immediate question which can be put at this point is whether it is at all important that a given growth in revenue, let us assume the desired one, is obtained through a tax system which has a high built-in elasticity or through ad hoc tax increases. The currently held view is in favor of the for- mer method. We shall deal with this question later on in some detail. But now we '\vant to focus on the general problem of the relation be"ttveen growth in reVenue and in GDP or GNP. 15. When discussing problems of developing countries, it is generally eJ(flected that public revenue increases faster than GDP. An elasticity of one is accepted as a minimum and whenever revenue fail to grow as fast as the economy, this is considered bad performance. These expectations originate from two types of considerations. One is of a technical nature, based on the supposition that any revenue mechanism must produce a higher than one elas- tici ty with respect to GDP; and even more so if changes in tax structure are introduced. The other type of consideration is, much more forcefully, of an economic nature and it simply reflects the need for increasing public savings as a condition for development. 16. The immediate, and often only, reason for an expected higher than one built-in elasticity lies in the progressivity of personal income taxes, which by definition should generate an elasticity higher than one, the exact coefficient depending on the relation between the average and marginal tax rates, given the assumption of growing personal incomes. 17. There are in fact other factors explaining the built-in higher than one elasticity of many tax systems. Business taxes, even if levied at flat rates should increase faster than GDP on the assumption that taxable profits tend to increase faster than GDP. Equally, indirect taxes levied on the gross value of sales in the industrial sector should also grow faster than GDP be- cause industrial production tends to increase faster than GDP. The only com- ponent of GDP, albeit an important one, which tends to grow less fast than GDP is agriculture but it so h~pens that agricultural taxes are usually a very small proportion of government revenue, and, therefore, their aluggishness does not really offset the more dynamic components of the tax system.  - 5- 18. But there are also other factors which in fact may tend to produce a lower than one elasticity. In many countries a large share of tax revenue originates in the import sector, and imports may increase less fast than GDP for a number of years. Furthermore, imports of goods are not themselves a simple basis against which to measure the elasticity of import duties. Import components are taxed at different rates and any change over time in the import structure will entail significant changes in the elasticity of import duties. A familiar situation, which characterizes the Tunisian experience, is the one in which the role of less heavily taxed commodities, such as capital goods, increases over time at the e~ense of finished consumer products. 19. The conclusion is that the elasticity with respect to GOP or GNP of the entire tax system varies 'tvith its own structure and that of the under-- lying economic activity. Furthermore, it is more correct to speak of the tax elasticity of each tax with respect to its own source than of the elasticity of the entire tax system. In other words, a less than one elasticity "ri th respect to GDP for the whole tax system does not imply that something is "wrong" wi th any particular tax. 20. HaVing indicated the reasons why a given tax system may not be elas- tic to the growth of GDP we shall briefly comment on the main factors e~lain­ ing the lack of elasticity of individual taxes with respect to their tax basis. These factors are mainly weaknesses in tax administration or a deliberate policy of tax exemptions. While there is no justification for the former, tax exemp- tions may be acceptable under the appropriate circumstances. This is not the place to discuss the role of tax exemptions and their favorable long run impact on the growth of the economy and the consequent broadening of the tax basis it- self. But if one accepts tax exemptions, one must be willing to live through a period during Which revenue will not grow as expected since the benefits of the exemptions policy will often come to fruition only after a number of years. This will also imply that current expenditure l'rlll have to be attuned to the e~ected growth in revenue. Otherwise, governments will have to either reduce the exemptions and possibly forgo their long~term benefits in terms of revenue, or increase the rates of other taxes assuming that this is feasible. 21. Let us now turn to the argument for continuously increased revenue as a reflection of the need to increase public savings. Such need could be also satisfied with revenue growing at about the same rate as GDP if it were not for the fact that current expenditure in most developing countries has a tendency to grow substantially faster than GIP. The need for increasing pub- lic savings, given the prevailing trends in current expenditure, has gradually pushed into the background any evaluation of the revenue level and structure from the viewpoint of its impact on the economy. Once we are confronted with an adequate public savings performance, we have a tendency to consider revenue and expenditure equally satisfactory (because the expenditure policy has also come to be judged in terms of public savings). 22. Admittedly, most governments and all international aid givers are aware of the excessive expansion of current expenditure but it seems easier  - 6 - to increase revenue than to curtail the growth in current expenditure.!! As a result, the expectations have had a tendency to concentrate on the revenue effort in terms of lev~land the GDP growth rate has become their minimum acceptable growth rate. This, we have seen, is not always justified in ter-ms of built-in elasticity. What we would like to stress here is that the failure to curb current e~Tenditure should not, in itself, become the yardstick against 'Which to measure the revenue performance. Particularly since the latter is more rigidly dictated by the growth of the e~onomy and its potential is certain- ly not as great as with cur-rent expenditure.-1 23. The prevailing emphasis on the need to increase revenue as a con- dition for expanding public savings can be challenged on two grounds. On the one side, there are the limits to taxable capacity which can be used as an argument against continuous increases in revenue. On the other- side, there is the so-calledllPlease Effect," as it is refer'red to in the 1tfo.rld Bank, which maintains that current expenditure of the government may not be exclusively a function of needs but also of availability of revenue.1/ 24. As 1-1e shall see later on, the Please effect is particularly signifi- cant for the discussion of built-in elasticity vs. ad hoc tax increases.But even outside this particular issue,it certainly represents an important limitation on the impact of additional revenue on public savings. The reasoning behind this argument is that a large amount of low priority current expenditure may be real- ized purely as a function of available revenue.~If one accepts the existence y ~e emphasis as far as the strlJ,ctv;.ra of expenditure is concerned, has been so far much more on investmemt expenditure than on current expenditure. The emphasis on investment ca n be explained on one side with the immediate link between investment and growth, and 011 the other with the fact that it is easier for e.id givers to advise a government on how to allocate its capital expenditure among the various sectors than on how to allocate its curront expenditure among various general functions like defense, health, educatiop, ..·polioe, etc. It seems to us that even leaving aside areas like defense and internal security, there are a number of important economic and social functions, the merit of which as far as structure and level are concerned could be profitably examined from the viewpoint of their contri- bution to t he growth of the economy. g( 0ur concern here is with those countries which, on the basis of international comparisons, already have a rather high tax burden. For countries with a relatively 10't'T tax: burden, the revenue potential or taxable capacity is obviously higher so that improvement in public savings may very well be a revenue problem. The ba.sis of our discussion for the need of expenditure control.stems from the Tunisi~n experience which, as mentioned in ·the . intrdduction, is character·ized by a ver'y high t?otio .of.. public revenue to GDP. 11 See Stanley Please, "Saving Through Taxation -- Reality or ~YIirage? II, Finance and Development, Vol. IV, Number 1, March 1967. ~ From there the conclusion that earmarking of revenue for development purposes is a possible answer to this problem.  - 7 - 6f the Please effect!!one goes back to the conclusio~ that the public savings problem can only find a satisfactory solution in the control of current expenditure, and l-lhenever this control is weak, one should rather aim at a lower leval of public revenue which would probably generate an equal. amount of public- savings. This in turn will lead to a higher amount of pr-ivate disposable income and of private savings assuming that the marginal propensity to consume of private individuals or corpo rations is lower than one of the government. 25. Our feeling is that the scope of the Please effect should be extended since to conduct the analysis Simply in ter-ms of current expenditure and current revenue is inadequate. The budget process is a global exercise which includes on one side total receipts, i.e., including bor-rowing, and on the other side total expenditure, current and capital. Additional current expend;i ture may be incurred not only on the basis of addi tionaJ.. current revenue but also of borrowing capacity. The results are lower public savings and higher borrowing, given a certain level of investment expenditure. By the same token, additional revenue, even if it results in increased public savings may give origin to additiona.l public investment expenditux:e the rate of return on which is lower than what could be achieved in the private sector. 1/ The emptricaJ. verification of the Please effect is a cha.llenging task for - public finance experts. Such verification can be conducted following two basic methods, insti tutionaJ.. and quantitative. An attempt has been made to find significant correlations between growth in revenue and in low priority current expenditure (see J. Lotz, "Patterns of Government Spend- ing in Developing Countries -- An Enpirical Study."). This study, examining a cr-oss-section of 37 developing countries with a GNP percapita of less than $800, shows that as the r·atio of taxes to GNP increases, the additional revenue is spent upon non-developmental expenditure. This finding, alone, cannot prove whether taxes were increased as a result of the decision to increase certain kinds of expenditure or: whether the latter were decided on the basis of expected revenue. It pr-ovides, how- ever, a necessary, but not sufficient, conditionibr the verification of the Please effect. One m:..:.gnt say that whenever in a country revenues have incl'leased as a result of changes in the rates or in the coverage of the taxes , it is more likely that the decision to spend came first. If on the contrary revenues have increased as a result of strong built-in elasticity or windfall gains, like in the case of oil revenue, it is mor,e likely that the availability of revenue caused the increase in expenditure. ~en these indicators, however, cannot be consider-ed as decisive which means that the institutional approach plays a necessary role in determining the true determinants of public expenditure, whether needs or revenue availability. By institutional approach we mean investigation of 'how the bud- get in a given country is put together and what are the forces at play. (See also Per Eklund, Taxation and Earmarking in Developing Countries, Economics Department Working Paper No. b3, April 1, 1969, p. 41.)  -8 - It is nor-mally assumed that once we have obtained addi tionaJ. public savings they will be used in the most economical manner in both the private and public sector but this is not necessarily true. 26. The investigation on the limits to taxable capacity is central to the question of tax elasticity insofar as, should there be a limit, an elasticity higher than one will have to be reconciled with the need of keeping the average tax ratio within a given level. A considerable amount of empir-ical work has been done inr-ecent years!! in order to establish the tax effort of developing countries. Factors like per capita GNP and the degree of openness have been considered for the purpose of classifying a number of developing countries between high and low tax effort ones. From a method- ological point of View, international comparisons of this nature can be described as the best substitute for our ignorance on the question of taxable capacity of a specific country. 1tJhat these comparisons tell us is which country shows the highest tax effort. But the fact that Brazil, with an average tax ratio above 30 percent, is at the top. of the list does not tell us whel1e the upper- limit for Brazil is, if it exists. 27. Ther-efore, for the purpose of canparing the performance of different countries we may go as far as saying that there are stronger grounds for increasing the tax burden, e.g"" in the Philippines than in Brazil. Similarly, the sane international comparison tells us that Tunisia is also a country wI th a substantial. tax effort even though not as high as Brazil. Does that mean that the tax burden in Tunisia could be raised several perc@ntage points above the present 27 percent without causing adverse effects on the econemy? 28. Quite honestly, if we did not have the information on a group of other comparable countries we would not be in the position of even deciding whether Tunisia has a lo'tv or high tax burden.. This is simply a reflection of the fact that one of the most important chapters of the theory of public finance has not yet :round a :a 8;·tisfactory verification in the real. world. Admittedly, it is an extremely difficult question which, among other things, cannot be solved without considering the effect of public expenditure financed through government revenues. 29. From the point of view of this stp.dy, it would be important if we lJare able to say that "conceptually" there must be a limit beyond which the average ta:x ratio cannot be pushed. Even if we cannot define it now for any given COUntFY, it would help in cautioning against a continuously increasing tax ratioYand make the empirical research of such limits a worthwhile effort. !I See J. R. Lotz and E. R. I'1orss, nI-1easuring Ta:x Effort in Developing Countries, If :mF Staff Papers, Nov. 1967. 2/ This is what is now happening, based on international comparisons. More - and more people feel that in countries like Brazil, UAR and TuniSia, public savings should be increased by curtailing the expansion of current e~~1?enditure rather than by increa:sing the average tax ratio.  - 9 - But does it exist? The fact that, historically, countries have successfully realized tax r-atios which previously had been deemed "fatal" for- the economy has convinced many people that there is really no upper limit to taxable capacity. Others maintain that, provided marginal .r:atesa..r:e such to avoid disincentive effects on labor, saving and investment the average rate can be pushed up with no darriage to the economy. We feel that the histo.r:ical experience only imicatesccf,hat permissible tax ratios can be higher than one would have thought, but does not prove that they can increase indefinitely.!! Since the tax ratio in many countries has been increasing partly under the influence of international aid givers, it is time that these very institutions give a hard look at this question. 30. So far we have discussed the emphasis' on the responsiveness of government revenue to the growth of the economy as reflecting the prevailing approach to the problem of increasing public savings. To the extent that one wonders about the soundness of this approach, one should raise the question of whether a tax s,ystem with high built-in elasticity is preferable to a system which requires a series of ad hoc tax measures in order to bring revenue to the level dictated by the expenditure needs. Tunisia is certainly a ver,y good example of a country which has succeeded in increasing revenue through a series of tax measures, thus offsetting the appa.r:ently poor built-in elasticity of its tax system. 31. The main practical argument in favor of built-in elasticity is that it tends to generate automatically the required increase in gove.r:runent revenue, on the assumption that it lv.ill be allocated to an acceptable program of government expenditure, both from the point of view of level and quality. FUrthermore it is argued that the government should not be required to go through the pro cess of passing every year significant tax measu.r:es which could encounter resistance by legislative bodies and stir up unpopularity to the point where the government might be discouraged from takrng such measures. 32. One should be careful in avoiding the inconsistency which exists between this kind of argument and the argument that government eA~enditure may become to a certain extent a function of available revenue and that, therefore, making revenue easily available to governments may lead to excessive levels of current expenditure rather than increase public savings. And as to the assumption concerning the utilization of resources, the predo- minant experience with trends in current expenditure for most developing countries is such that on balance one should not be complacent about having a very elastic tax system. ,. 3J~ If we have doubts as to the bapabality of certain governments in making the best use ·of their available resources, a tax system with poor built-in elasticity, thus requiring unpopular measures to offset it, may provide a powerful built-in mechanism,. against a poor expendi tUlle policy. --------------------------------__..i-- !I A. Peacock and J. Wiseman, "The Growth of Public Expenditure in the United Kingdom, II '!he Natio nal Bureau of Economic Research, Princeton, 1961.  - 10 - Admittedly, trus may sean a rather- unorthodo:x thesis but it may have some value in tho sa countries where one feels that the average tax ratio has reached a very high level (assuming that such thing as a high tax ratio exists), or mor-e generally in those countries in which the expenditure policy is clearlY getting out of hand and some inelasticity could be useful together- with contr-ols on expenditure. 34. One disadvantage of having inelastic tax systems is the possibility of substantial revenue shortfall with respect to budget estimates .and of their obvious disruptive effects on government programs. The problem here is that the government, in preparing the estimates, does not take into account the inelasticities of the tax s,ystem, sometimes because of ignorance, some/- times knowingly but, having to make ends meet, at least o.n ex-antebaSis. l Very often the inelasticity, as with tax exemptions, is the r-esult of the deliberate policy of the government, and if it were correctly reflected in the preparation of the budget, the expenditu.r:e policy would have to be adjusted accor-dingly. Once again, we would like to stress that it is the expenditure policy which tends to magnify deficiencies in the tax system. With a normally functioning tax system revenue could grow at most by two percentage points faster- than a GOP increasing at 5 to 7 percent per annum, while at the same time current expenditure has in effect no limit except in the willingness of gover-nments to resist all the pressures for higher wages, bills, increased defense expenditure, etc. ~ There is always an optimistic bias in the preparation of gover-nment budgets, often because these reflect targets rathe~ than forecasts of likely trends in revenue and expenditure.  CHAPTER II THE HISTORICAL EXPERIENCE A. Struoture of' the Revenue System 35. The revenue system in Tunisia is dominated by the Central Govern- ment whioh now aooounts tor almost 85 percent of total general government • revenue. Over the 1960-68 period the only signifioant change has been the increased role of sooia1 seourity oontributions which, as shown in Table 1, have:risal from 8 to 12 percent of the total. Table 1 Structure of General Government Revenue Plan Projection 1960 1964 1968 1972 Central Government 85.8 83.5 83.3 84.2 Local Authorities 5.1 6.3 4.1 4c4 Sooial Security 8.1 9.5 12.1 10.8 Others 1.0 0.7 0.3 0.6 100.0 100.0 100.0 100.0 36. At the Central Government level, in addition to the so-called ordinary revenue, under Title I, there are a number of reoeipt s aocruing to the Treasury' Aooounts and Special Funds (see Annex, Table 1).17 Between 1960 and 1968 these 'have aocounted on the average for about 15 peroent of' Central Governlllent revenue. Starting in 1968, a number of special Treasury funds have been integrated in Title I. Most of the revenues aooruing to the Treas- ury accounts are shareeot direct and indireot taxes colleoted by the Central Government. 37. Revenues of looal authorities are real estate taxes, participation in Funds finanoed through shares of major direct and indirect taxes and a number of other taxes and tees. The allocation ot the Funds, whioh account tor about 25 percent of total local authorities revenue, is based on various oriteria ooncerning population, road network, tax burden on real estate, etc. B. Trends in General Government Revenue " 38. The first step in our analysis is to identity in its broad lines the evolution of government revenue in relation to GDP and to single out the problem area from the point ot view ot tax elasticity. It appears imme- diately from Table 2 below that 1960, the base year of our series, is rather exceptional. 11 These inolude, among others, the National Defense Fund and the National Fund for improvement ot housing oonditions.  - 12 - Table g, Revenue as Percentage of GDP at Factor Cost 1960 1961 1964 1965 1966 1967 1968 Central Government 2,.2 22.6 23.6 24.2 27.0 26.4 2,.5 Local Authorities 1., 1.4 1.8 1.3 1.7 1.3 1.3 Social Security 2.4 2.1 2.8 3.0 3.5 3.6 3.7 Others I' 0.3 0.2 0.1 0.1 0.1 0.2 0.3 Total 29.4 26.3 28.3 28.6 32.3 31.5 30.6 As already mentioned in the introduction, the tax buoyancy between 1960 and 1968 would be practically one which implies, in view of the many increases in the rates incurred during' this period, a tax elasticity markedly belOlrJ' one. We believe that one should exclude 1960 from the series also because its very high level may reflect in part the change in 1960 or the fiscal year basis from March - April to January - December. 39. The 1961/68 period shows a higher than one tax buoyancy but we also notice that after 1966 tax buoyancy at the Central Government level fell below one and this affected General Government revenue as well in view of the predominance of the former. The conclusion, an obvious one, is that the problem area is to be found at the Central Government revenue. 40. Central Government revenue, however, does not provide an adequate basis for our analysis. Treasury Accounts which, from the table follovnng, show a lower than one buoyancy, cannot really be investigated in their com- ponents which are not entirely of tax origin. Furthermore J to the extent they consist of indirect and direct taxes, (for which no sufficient break- down is available) their behavior follows the general pattern of tax revenue under Title I. Table 3 Central Government Tax Revenue As Percentage of GDP at Factor Cost 1960 1961 1964 1965 1966 1967 1968 ,.. Central Government Receipts 25.7 22.5 24.6 25.1 27.6 27.1 25.6 Treasury Accounts!! 4.5 3.9 4.3 4.5 4.2 3.9 3.6 Title I 21.2 18.6 20.3 20.6 23.4 23.2 22.0 of which: tax revenue (19.6) (17.4) (18.9) (19.1) (21.6) (21.3 ) (20.1 ) a/ Adjusted by adding n, million to Treasur,y Aocounts and subtracting them - from Title I. (See footnote to Appendix, Table 1.)  - 13 - 41. Also Title I cannot be taken in its entirety f'rom the point of view of the tax elasticity ana~sis since what appear as other revenue and other taxes and duties, are the sum total. of' sundry revenues which cannot be mean- ingfully linked to any underlying economic basis. The conclusion, from the point of' view of' the coverage, is that we should only consider what appears in Appendix, Table 1, as direct and indirect taxes together with oil revenue.lI 42. A last oomment: neither local authorities revenue and Treasury Accoun'b3 have been a very dynamic souro e or revenue, between 1961 and 1968. In fact, their buoyancy has been less than 1. Between 1966 and 1968, their decline in terms of GDP has followed that of the Central Government, and this is largely explainable by their dependency on the trend in Central Government revenue. However, their combined share of general government revenue is rather small, less than 20 percent in 1961, and even a more vigorous growth rate during the period, let us say, with an elasticity of 1.2 with respeot to GDP, would ha~ only added D5 million to general govern- ment revenue in 1968 bringing it to 31.5 peroent of GDP, still below the peak 1966 level. C. Reolassifioation of Tax Revenue Aooording to their Economic Basis 43. The next step is to reolassify tax revenue with the purpose of relating them to their appropriate tax base'. In this lvay the a.nalYsis or- the relation between ohanges in tax structure will be more olear. Unfortunately, here the greatest difficulties arise and, as we shall see below, it will be necessary to resort to "second best" proxies of' the ideal tax base for eaoh tax or group of taxes. Table 4 Reclassifioation of Taxes Aocording to Seotors of Origin Taxes on Economic Basis Income ( individual and business) GDP at factor cost excluding agrioulture and crude oil Imports Imports of goods excluding cereals, - Customs Formalities Tax crude oil, oil produots - Import Duties - Production tax on imports - Consumption tax on imports Exports Exports Agricul ture Agriculture (value added) Oil Crude Oil produotion (value added) Transport Transport and Communioations (value added) Transactions - Registration Tax GDP factor cost excluding crude oil 1/ Exoluding royalties from Algerian pipeline.  - 14 - Taxes on Economio Basis Consumption - Consumption tax on alcohol ) Private consumption - Fiscal. monopolies ) - Oil and gasoline tax ) Domestic Production - Consumption tax on domestic products Manufacturing (value added) - Production tax on domestic products Service Tax Services (value added) 44. Particularly in the case of income taxes, the absence of data on personal incomes as well as on business profits makes it necessary to use GDP, excluding agrioulture and crude oil, as a proxy. The reason for ex- cluding agriou1ture is that, by general agreement, the contribution of this sector to income taxes has been quite modest. As to orude oil, whilst pro- duction became very significant only at tbe end of the period considered here (see Table , in Annex), its impact on income taxes has been very minimal (revenue from oil companies is shown separately). On the other side, it would raise the growth rate of the tax base for the 1961-1968 period from 7.2 to 7.9 percent.Y In the case of imports, the absence of data showing the amount of revenue obtained from the various groups of commodities will also make it difficult to analyze the relation between import taxes and imports with the neoessary "depth". As to the gasoline tax we have taken as economic basis the trend in private consumption, whilst the trend in the consumption of oil and gasoline would have been more appropriate. Unfortu- nately it was not possible to have a complete series of this item. Table 2 shows tax revenue reclassified acoording to the seotor of origin. Y Annual Rate of' Growth of GDP at Factor Cost (%) Current Prices Constant Prices 1961-1968 (prov.) 1961-1968 (prov.) GDP at factor oost 4.1 GDP less agriculture $.2 GDP less agriculture and oil 4.$ GDP less oil 3.0 Source: lBRD report No. EMA-12, August 2$, 1969, Tables 10, 11.  - 15 - Table 5: Percen~ge Breakdown 01- Tax Revenue, 1961, 1966, 1968 and 1969 - Reclassified According to Sector of Origin 1961 1966 1968 1969 Income Taxes 18.1 20.1 19.7 21.1 Indi~·idua1 7.7 9.6 9.6 10.2 Business 10.4 10.5 10,,1 10.9 Imports 31.1 ~3.2 24.8 26.4 Customs Formality Tax 2.6 2.4 1.9 2.1 Impor't Duties 10.6 10.6 6.8 7.6 Production Tax on Imports 15117 17.5 13.4 14.2 Consumption Tax on Imports 2.2 2.7 2.7 2.5 Exports 1.5 5.0 3.8 3.3 of which: Special Temporary Tax on Exports (4.1 ) (2.7) (2.3) Agriculture 4.8 1.9 1.5 2.1 Oil 6.1 ,.-- 7.4 Transport 0.7 2.2 1.8 1.2 of which: '~edevancesur 1es Transportes I' (1.3) , (l.0l Transactions 5.9 4.3 4.2 4.2 Consumption 29.2 22.0 25.4 21.5 Consumption Tax on Alcohol and Spices 2.4 1.9 1.4 1.7 Fiscal Monopolies 16.1 13.2 15.0 12.4 Oil and (hsoline Tax 10.2 6.4 8.6 7.0 Others 0.5 0.5 0.4 0.4 Domestic Production 6.9 8.2 9.1 9.2 Production Tax on Domestic Products 6.2 6.8 7.2 7.3 Consumption Tax on Domestic Products 0.7 1.4 1.9 1.9 Services 1.8 ~.1 3.6 ~.6 Total 100.0 100.0 100.0 100.0  - 16 - Table 6: Tunisia's Structure of Tax Revenue, 1960-1969 - Reclassified according to sector or origin ( In Millions of Dinars) Budget Estimate 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 Income Taxes 8.8 9.9 8.9 9.1 10.8 11.9 IB.3 IB.9 19.5 24.4 Individual 4. 2 3. tr 4. .8- - 5-:-~ - 6.3 8.7 . B.7 9~S 11.-B lhsiness 5.7 5.1 4.3 5.3 5.6 9.6 10.2 10.0 12.6 Imports 19.9 17.0 18.2 21.2 27.5 30.3 29.0 24.6 0.5 Customs Formality Tax 1.0 1. 1. 1. 2.1 2.2 2.3 1.9 2. linport Duties B.2 5.8 6.6 7.7 9.6 9.7 8.6 6.7 8.8 Production Tax on Imports 9.5 8.6 9.0 9.9 13.5 15.9 15.5 13.3 16.4 Consumption Tax cn Imports 1.2 1.2 1.2 2.2 2.3 2.5 2.6 2.7 2.9 Exports 0.8 0.8 0.8 4.4 4.6 3.2 (2.3 ) 2· B (2.7 ) 3.B (2.7 ) of which: Special Temporary Tax on Exports - (3. 6) (3.7 ) Agriculture __________________________________ _2_.2 ?_h 1_? ?~ ?_). 1 _0 __ 1.7 1.0 1.5 2.1! Oil 2.3 6.1 8.6 Transport 0.4 1.5 0.5 1.5 0.9 2.0 0.6 I.B 1.4 of which: '~edevance Sur les TransportesnY (-) (1.1) (-) (O.B) (-) (1.2) (-) (1.0) (-) Transactions 3.2 3.2 2.9 3.0 3.4 3.7 3.9 3.7 4.2 4.8 Consumption 16.0 15.3 16.5 17.6 18.-2---- 20.cr-- - 21.7 25.2 2~ Consumption Tax on alcohol and spices 1.3 1.2 1.4 1.6 1.4--- :L~7- 2.0 1. 4 1.-9 Fiscal MonopolieE 7.9 8.8 8.2 9.1 9.9 10.4 12.0 13.4 14.9 14.3 on and Chso1ine Tax 5.6 5.6 5.7 5.7 5.9 5.8 5.9 8.5 8.1 Others 0.3 0.3 0.3 0.4 0.5 0.5 0.4 0.4 0.5 Domest1cal~ manuf~ctured 3.6 3.8 3.B 4.2 5.1 6.1 7.5 B.4 9.0 10.6 Production Tax on Domestic Products 3.3 3.Ii --,3J,i- T.S-- -4.3 5.1 6.2 6.7 7.1 8.4 Consumption Tax on Domestic Products 0.3 0.4 0.4 0.7 0.8 1.0 1.3 1.7 1.9 2.2 Services 0.7 1.0 1.0 1.1 1.2 2.0 2.8 3.1 306 4.2 TOTAL 55.7 54.7 53.6 58.9 68.1 76.6 91.1 91.9 99.3 115.5 Tct1.l, c~~t:':'al Government RavenuJ! 73.1 71.a- 71.8 77.5 91.9107.5 119.1 123.2 126.1 --11~5.-8 . -.-,-.....-~- .--.,.-..-~------------ .. - - . - - -.~- - - 'bl Incl UdL"1.g other revenue, pipeline royalties and other Treasury accounts (see Appendix, Table 1). y Ac0V!pensa~ion paid to a special fund.  - 17 - D. The Sensitivity of the Tax System to Changes in The St_"Ucture of the Economy • 45. In 'Ghis section we shall try to identify the expected built-in e1a.sticity or the Tunisian tax system based on its structure as it appeared in 1961 and on the actual growth rate for the 1961-1968 period of the rela- tive sectors or origin of the various taxes. (See Table 7.) This attempt is far from satisfactor,y in view of the already mentioned statistical diffi- culties in relating taxes to their appropriate economic basis. Then we shall compare the actual with the expected growth rates in revenue and indicate the major changes in the tax structure which explain the dlf'f.erent behavior of the two series. A more detailed discussion of the single taxes will be conducted in the following section. There we shall examine the grolvth of each tax against its own economic basis and explain the main reasons for the apparent lack of' elasticity, whenever this will be the case. 46. The relation between the tax structure in 1961 and sectors of origin. is given in Table .• It appears from this table that seventy percent of tbe tax structure could have been expected to grow less fast than GDP at factor cost. Import and consumption taxes were the two single largest group of taxes in this category. The expected growth rate for the entire tax system shows an elasticity slightly below 1 (0.95). One should be aware of a statistical problem which affects the elasticity coefficient. The tax structure of 1961 does not include oil revenue which started accruing only in 1967. Gross Domestic Product in 1968 does include a certain amount of crude oil production. If we deduct this amount from the 1968 GDP in order to arrive at a more mean·· ingful growth rate for the entire period, the annual growth rate of GDP would decline to 5.8 percent, thus giving a slightly positive built-in elasticity (1.0,). In view of the margins or error implied by this exercise we would be inclined to conclude that the built-in elasticity of the tax structure as it stood in 1961 was t, around 1". 47. Being aware of the importance of the tennina1 yeaJ:"~ in any statis- tical analysis, we have tried the exercise on the expected built-in elasticity of the Tunisian tax structure by taking the 1961/62 and 1967/68 averages as far as structure of taxes and growth rate of the various sectors is concerned (Table 8). The data show a built-in elasticity (with respect to GDP) of 1. If we take the unadjusted growth rate of GDP at factor cost and we adjusted it for oil production the latter's growth rate for the 1961/62 - 1967/68 period would become 6 percent thus giving an elasticity of 1.09. The taxes growing less fast than GDP would now account for only 40 percent of the tota1o It may be su~ri81ng that with such a dramatic reduction in the role of the less dynamic taxes the elasticity does not differ substantially from the one shown in the 1961-1968 period. The reason is that for the 1961/62-1967/68 period agriculture does not contribute to the growth of the tax baSis, and this factor largely offsets the faster expansion of the import tax basis. 48. What eme~ges from the table is/the high sensitivity of the Tunisian tax structure to tue growth in imports.l By way of example, if we took tho 1/ We exclude imports of oereals, crude oil and oil products since they are largely exempt from both the oustoms duties and the production tax on ilnports.  - 18 - 1961-1966 period, at the end of which imports reached their peak value, the built-in elasticity of the tax s,ystem would have been 1.3 with GDP at factor cost growing at 6.2 percent annually. The absolute decline in imports after 1966 was the single largest factor behind the poor buoyancy of tax revenue between 1966 and 1968. To the extent that there is import substitution one could say that a decline in imports would have to be compensated by a faster increase in domestic production and, therefore, in domestic taxes. This was the case in Tunisia where in fact domestic production of oil increased the tax basis, starting in 1967. HOwever, as we shall see later on when discussing the specific taxes, import substitution often implies that domestic production is taxed at a lower rate than imports thus causing a reduction in the elas- ticity of the tax system.!! Furthermore, a decline in imports may not be dictated by import substitution 0lr. changes in consumption patterns, but rather by foreign exchange constraints.~ In such a case, the commodities which are likely to be squeezed are luxury consumption goods which usually pay the highest tariffs and, therefore, this will also reduce the elasticity of import taxes. This factor was probably of considerable importance in Tunisia during 1966-68. 49. The other group of taxes which influence significantly the growth of total revenue are those based on private consumption which in Tunisia, at least over the entire period, has grown less fast than GDP at factor cost. The trend in taxes, in such a case, is really a function of the development strategy of the country with respect to domeotic r@source mobilization for investment It is interesting to note that a low built-in elasticity with II respect to GDP of taxes based on private consumption is likely to be accom- panied by a low built-in elasticity in import taxes to the extent that a reduction in imports of the more heavily taxed consumer goods is a feature of the deflationary policy of the government. ,0. The trend in agriculture over the 1961-1968 period is particularly unfavorable. As we have seen, if we take the 1961/62-1967/68 period, there is even an absolute decline in value added from the sector. il'lhile we would normally expect agriou1ture to grow less fast than total GDP, the Tunisian experienoe is oertainly below the average! It is true that agrioultural taxes in 1961 did not play a very important role, but a more favorable evolu·· tion in the sector and, therefore, in tax revenue, could have added, assuming a growth rate of 4 peroent (still below the one of GDP), about 2 million dinars to the 1968 revenue. A not entirely negligible amount in view of the sharp deo1ine in other revenue for that year. !! See, on this subject, the article by Stephen R. LewiS, "Revenue Imp1ioa- tions on Changing Industrial Structure: An &npirica1 Study," National Tax Journal, Dec ember 1967. g( It is oonceivable that the deo1ine in import taxes could be offset, all or in part, by the intl~duction of a tax on foreign exchange.  - 19 - 51. The growth in the tax basis for income taxes (i.e., GDP at factor cost excluding agriculture and oil) called for an expected el~\s.tici ty with respect to GDP at factor cost of 1.12 to 1.17 depending on 1i;ne period con- sidered (1961-1968 or 1961/62-1967/68). We have already said thl:~t we feel rather uncomfortable in having to use GDP excluding oil and agri,oul ture as the appropriate tax basis. The elasticities above may seem on the low side but we have to take into account, as we shall see later on, the various tax exemptions and a certain level of tax evasion which characterize income taxes. 52. If we now compare actual collections of revenue during the 1961- 1968 period we have to proceed in two steps. Total collection in 1968, including the special temporary tax, t~e-full income tax on business (i.e., unadjusted for the large increase in 1966) and oil revenue, shows a growth rate of 8.7 percent per armum over 19~1. If we exclude these three large items which were not present in the 1961 tax structure, one obtains a growth rate of 6. 7 percent (shown in column V of Table 7 ), ba::~ely above the expected growth rate of 6.1. This growth rate, however, only takes into account the most significant changes, but does not allow for the other tax increases which have affected practically all major taxes and which will bo reviewed in detail later on. In view of this fact, it is safe to conclude that the actual built- in elasticity of the tax structure was probably below what could have been expected.V 53. Table 2 indicates a higher ratio between the expected and actual growth rate in revenue for the 1961/62-1967/68 period, but it still implies that the historical built-in elasticity was lower than that indicated by the growth rate of the various sectors of origin, if one took into considera- tion all the other tax increases not reflected in the table. One should also notice that the growth rate of import taxes is higher than the growth rate of GDP and this is because the averaging of 1967 and 1968 partially eliminates the impact of the sharp absolute drop in import taxes of 1968. ,4. Having so far concentrated on the impact of the changes in the structure of the economy over the built-in elasticity of the entire tax ~stem, we should like to indicate three main factors which have contributed to a low tax elasticity of specific taxes i.e., the compulsory loans of 1964 and 1968, the policy of tax exemptions and weakness in tax administration. 11 In other words, we have not tried to calculate directly the historical built-in elasticity of the various taxes because the margins of error are so large and the information so defective that, as said in the first chapter, we do not believe that this would be a very meaningful exercise.  - 20 - The role ot the oompulsory loans (tax deduotib1e) and of tax exemptions is partioular1y evident in the case of inoome taxes on individuals and business while tax exemptions have affeoted the behavior of the oil and gasoline tax. These aspeots will be discussed in detail when dealing with the speoific taxes. r.Tealmess in tax administration has certainly played a role in reducing the built-in elasticity of the Tunisian system. There are no indications, however, that it has been a speoial problem when oompared with most other developing oountries. 55. There is, however, a factor explaining in part the past weakness in tax administration and whioh is directly linked to the methodology fo110vred in the preparation of budget estimates. For almost all years in the past, exoept in 1968, actual oollections have had a tendency to exoeed budget estimates. From discussions with offioia1s of the Tax Department, it became clear that the conservative approach to estimating revenue was in many cases the result of applying to previous years oollection a rigid growth parameter almost regardless of the expeoted growth of the economY as determined by the planning authorities. But more generally it was the result of an under- standable attitude of not willing to be committed to targets which might have been hard to achieve and that, if not aohieved, might have reflected badly on the tax authorities. It is our view that'this oonservative approach in budget estimates has reduoed the eagerness of tax authorities in pursuing tax evaders particularly with respect to business income taxes and to taxes on self-employed individuals. The reason for this is that once conservative estimates are made which are easily realized and even surpassed, there is relatively little incentive for the tax authorities to capture the largest possible share of additional incomes and profits generated in the eoonomy. The responsibility for tax co11eotion is redistributed among a series of local offices each one responsible for a certain amount of taxes whioh they are expected to raise. Their attitude vis-a-vis the Central Direction in Tunis is the same as the latter's vis-a-vis the Minister of Finance~ it is important to realize the targets and anything beyond them is already an incU.cator of suffioiently good perfomance. 56. Things have now changed in that, starting in 1969, projections of revenues are being made in cooperation with the Planning Authorit,ies which provide the growth parameters for most of the taxes. If at all, these targets may be too ambitious and one has to expect some shortfalls in actual collection, but t.hey w:t11 also put more pressure on tt.e ta:~ authorities to strengthen their efforts in collecting taxes.  - 21 - Table 7: EXpected Built- In Elasticity of Tax structure I II III IV V Actual % Contribution Actual % 1/ TAXES Annual Growth to Increase Annual Grol·Tth·- As % of Rate of Tax of Total Rate of Tota1-l961 Tax Base Base Revenue Revenue A B Ax B linport 31.0 Imports of goods 6.0 1.86 5.4 excluding cereals & oil products Agri.culture 4.8 Value added in 0.8 0.38 agriculture Services 1 .. 9 Services 5.2 0.09 20.W (value added) Consumption 28.6 Private 4.8 1.37 6.7· Consumption ~ports 1.5 Exports 8.7 0.13 3.611 Transport 1.2 Transport 11.6 0.14 10.5~ (value added) Domestic 7.0 Manufacturing 6.9 0.48 13.1 Production Income 18.2 GOP excluding 7.2 1.31 7.321 agriculture and crude oil Tr~nsaction 5.8 GDP factor cost 6.1 0.35 4 • ,.i:; (excluding crude oil) TOT "..I. lOO.,n 6.11 6.7 AC~0~1 annual growth rate of re'f/enue unadjusted for changes under Column V ---- 8.7 A~lual growth rate of GOP at f~ctor cost ------------------------------- 6.4 1/ In addition to the three changes listed above it, does not include oil revenue which started to accrue in 1967. 2/ See discussion of services tax for the explanation of this high grovlth rate. 3/ Excluding special temporary tax introduced in 1965. "4/ Excluding "redevances sur les transport" on account of their accruing every even year (1962,1964, etc.). . 21 Adjusted for impact on business taxes of 'the 1966 reform extending the tax to public enterprises and estimated at 3.1~ million dinars.  - 22 - Table 8: Expected Built-In Elasticity of Tax structure Based on Actual Growth Rates of Sectors of Origin I II III rJ V 1961/62 7 Actual ~1, Annual Contribution Actual Annual Growth Rate of to Growth of Growt~ Rate As % of Tax Base Total of Revenue !A~~ Total Tax Base 1961/62-1967/68 Revenue 1961/62-1967/6~ A B AxB !mDurt 32.5 Import of goods 7.5 2.44 7.3 excluding cereals & oil products Ag:,:1.culture 3.5 Agriculture -6.7 (value added) Sel',rj.ces 1.8 Services 5.3 0.10 22.5 (value added) Conaumption 28.9 Private Cons. 4.9 1.42 7.0 Exports 1.5 Exports 9.2 0.14 3.B2I Transport 1.8 Transport & 11.8 0.21 9.821 Comm. Domestic 7.0 Manufacturing 7.9 0.55 14.8 Product Income Taxes 17.4 GDP excl. a gric 0' 7.6 1.32 9.rJ±I and crude oil Transaction 5.6 GOP factor cost 6.0 0.35 4.4 (excluding 100.0 crude oil) 6.53 8.0 h.~ li".l~l annual growth rate of revenue unadjusted for changes under column V --- 9.9 AtL"1u:ll grow'th rate of GDP at factor (~ost -------"':'----------------------- 6.5 1/ In addition to three changes listed above, it does not include oil revenue which started to accrue in 1967. ~/ Excluding special temporary tax introduced in 1965. 11 Excluding "redevances sur les transports" on account of their accruing every even year (1962,1964, etc.). kI Adjusted for impact on business taxes of the 1966 reform extending the· tax to public enterprises, and estimated at 3.4 million dinars.  - 23 - E. The Elasticity of Individual TaxeJ! Import Taxes ,7. Import taxes in Tunisia include cust'\Jm duties, as well as con- sumption and production taxes on imports. There is also a customs formalities tax imposed on the value c.i.f. of all imports. The rate of this tax is presently 1.8 percent. There are finally consumption fees (droits de consomnation), which apply to some imported goods, mostly food products, certain petroleum products, tires, alcohols and explosives. The most im- portant of these four, yielding about 10 of total tax revenue in 1961, is the oil and gasoline tax. Unfortunately, it has not been possible to separate the amount of the tax collected from imported quantities from the one col- lected on domestically produced oil products. We have therefore left this tax out of the import based taxes group. 58. We shall deal very briefiy with the features of the three main import taxes before moving over to the question of elasticity. Customs tariffs exempt a large number of products, the most important ones being energy products, metallurgic minerals, some unmarked textiles, machines f;;:)r textile industries and cereals. There is obviously a wide range of tariffs, the lower ones (around 6 perc ent) levied on equipment and inter- mediate and some finished products, and the higher ones (up to about 35 percent )on consumer goods of luxury nature. 59. MOst imports are also subject to the production tax.S! The rate of tbe tax is 14.3 percent for products destined to producers and 18 percent for products destined to non-producers. The tax is assessed on the value of the product plus the customs duty and the customs formality tax. The con- sumption tax is assessed on the same basis as the production tax with three basic rates of 23, 16 and 8 percent. The rates were increased recently from the 1968 level of 21, 14.7 and 7.35 percent. Luxury goods are obviously taxed at the highest rate of 23 percent. The lowest rate of 8 percent applies to goods like honey, conserves and fruit juic es, soaps, etc. As with the produotion tax, imports destined for non-producers pay a rate increased by 25 percent. 60. The historical ana~ysis of the elasticity of import taxes (with respect to GDP at factor cost) is fairly simple. If we exclude cereals, oil and oil products, imports of goods as peroentageo£ GDP at factor cost have stagnated after 1964 and even declined in 1967 ~1d 1968. (See Table 11.) This in itself provides a basis for a lower than one built-in elasticity of import taxes. ~n fact, excluding the above items the value of imports went down from 123 million dinars in 1966 to 104 million dinars in 1968. An absolute decline occurred also in the value of import tax collection and was particularly sharp in 1968 when the projections had indicated a much higher level of more than 30 million dinars against an actual collection of 23 million. 11 For a description of the Tunisian tax system, see, in addition to the various tax laws" "La Fiscalite En Tunisie" Secretariat d'Etat au Plan et a l'Economie Nationale, TuniS, 1969. g( Among exempted products we have printing materials, fishing equipment, cereals and petroleum imports imported by distributing and oil producing companies.  - 24 - 61. The sharp rise in taxable imports during the nrl.ddle sixties was basically the reflection of the very rapid expansion of public investment. Total imports between 1960 and 1965 grew at an annual rate of almost 11 percent compared with virtually no change in the ten preoeding years. After the devaluation of 1964 and the increasing financial diffioulties, imports of capital goods declined, reflecting a reduction in fixed investments, while imports of raw materials and intermediate goods still. increased to satisfy the demand of the expanding Tunisian industry. The sharp reduction of 1968 was mostly the result of unuti1ized stocks of intermediate goods from previous years, a decline in finished oonsumer goods reflecting import substitution, as in the case of textiles, and oontinuation of controls.!! 62. The next step is to examine the elasticity of import taxes with respect to their tax base which, as said, are imports of goods excluding cereals, crude oil and oil products on the assumption that these are largely exempted from import taxes. The figures in Table 11 show a buoyancy belovl one starting in 1964. It we consider the structure of imports throughout this period, we notice a fairly sharp decline in the role of finished con- sumer goods and foodstuff. These, on the whole, are the more heavily taxed commodities which, in itself, should introduce a downward bias in, the built-in elasticity of the taxes.gj 63. In applying this reasoning, however, one has to distinguish between the three types of taxes. In the case of customs tariffs, which have remained practically unchanged during the period, the impact is quite clear if we look at the data. The ratio of customs tariffs to imports of goods has declined steadily between 1963 and 1968, and also their role in the structure of import taxes has followed the same trend. The trend in the production tax has been affected by the increases in rates, particularly the ones of 1966 when the rate was raised from 11 to 12.5 percent. But in the remaining years, the tax shows basically the same tendency with customs duties, i.e., to decline as a result of the ohange in the structure of imports. Finally, the con- sumption tax on imports, \orhen related to total imports indicates a higher buoyancy than the other two taxes, but this is essentially the result of the very sharp increases in the rates of 1963, 1965, 1966 and 1968. In fact, if we relate the tax to imports of finished oonsumer good~a better approxi- mation to the appropriate tax basis, we notioe a much higher buoyancy also resulting essentially from the tax inoreases. In particular, one should notice that the tax increase of 1965 is not reflected at all if we relate the con- sumption tax to total imports. This is because in 1965 there was a ver,y marked decline in the share of imports of foodstuff. This is also true in 1966 and 1968 where the same factor explains why the inorease in the tax ratio is much smaller if total imports are used as a base. 11 See the latest economic report on TuniSia, The Eoonomy of Tunisia, EMA 12, June 20, 1969, paras. 124 to 127. g; In view of thespeoia1 treatment for imports of oil and oil products, on~ could argue that their decline oontributed in fact to increase the elas- ticity of the import tax system to the extent that the gap left open by the no longer required oil products made it possible to realize a h~gher level of imports of taxable commodities. This argument is only val~d if we assume as it seems appropriate in the case of Tunisia, that the his- torioal 1~ve1 of total imports was the maximum attainable in view of the foreign exchange constraints.  - 25 - Table 9: TAXES ON IMPORTS (Million dinars) Budget Est. 1968 1961 1962 1963 1964 1965 1966 1967 1968 1969 Customs Duties 8.2 5.8 6.6 7.7 8.8 9.6 9.7 8.6 6.7 8.8 Production Tax on Imports 9.5 8.6 9.0 9.9 12.5 13.5 15.9 15.5 13.3 16.4 Consumption Tax on Imports H 1.2 1.2 2.2 2.2 2.!.l .1.:2. 2.6 II 2.:2 Sub-Total 18.9 15.6 1608 19.8 23.5 25.4 28.1 26.7 22.7 28.1 1.0 ~ Customs Formalitjes Tax .1.J! 1.4 1.4 -1.8 2.1 2.2 ~ .l:.:.2 Total 19.9 17.0 18.2 21.2 25.3 27.5 30.3 29.0 24.6 30.5 =-:::z::: -==== --=== ==-== =-= -=== -=-: === ~ Structure (.~) CU3toms Duties u'i.O 37.0 39.0 39.0 37.0 38.0 35.0 32.0 29.0 31.0 Production Tax 50.0 55.0 54.0 50.0 53.0 53.0 57.0 58.0 59.0 58.0 Con sump tion Tax --1.:.Q B.o --1.:.Q 11.0 10.0 .....2.:Q 8.0 10.0 12.0 11.0 Total -- - --=- - - - - - 100.0 100.0 100.0 100.0 100.0 lQO.O 100.0 100.0 100.0 - - 100.0  - 26 - Table 10: STRUCTURE OF IMPORTS (Excluding cereals, oil and oil products) Ec. Budget Ests. 1960 1961 1962 1963 1964 1965 1966- 1967 1968 1969 Equipment (22.6) (24.2) (24.1) (26.9) (29.1) (35.2) (28.1) (25.1) (29.3) (28.6) (27.8) (28.8) (30.3) (33.4) ( 34 ~ 5) ( 36 · 9) (J~2. 7) (44.6) (40.8) (0.6) (0.7) (0.,) (0.6) (0.5) (0.7) (1.4) (1.7) (1.7) (34.4) (31.3) (31.1) (26.3) (21.9) (19.1) (17.6) (18.4) (17.0) (14.6) (15.0) (13.4) (12.8) (14.0) (8.1) (9.6) (10.2) (11.2) ~ .!QQ.:.Q 100. 0 .!QQ.:.Q 12£:.2 .!QQ..:.2 !QQ..:Q 1.QQ.:.Q .!QQ.:.Q 100 •0 ~ 100 •0 Total in million dinars (69.1) (69.1) (74.0) (81.8)(102.8)(11$.3)(117.9)(117.7)(103.1) (122.2)  - 27 - Table 11: IMPORTS AND IMPORT TAXES 1960 19b1 19b2 19b3 19b4 1965 19bb 19b7 19bB 19b9 A. fmports of Goods, excluding cereals and petroleum products as %of GDP at factor cost 24.3 21.6 22.6 23.3 27.5 27.0 27.4 25.8 21.1 22.8 B. Import Taxes, excluding customs formalities tax as % of A 27.3 22.6 22.1 24.2 22.8 22.1 23.8 22.7 21.9 23.0 - Customs llities 11.8 8.4 8.9 9.4 8.5 8.4 8.2 7.3 6.4 7.2 - Production Tax 13.7 12.5 12.2 12.1 12.1 11.8 13.5 13.1 12.8 13.4 - Consumption Tax 1.8 1.7 2.0 2.7 2.2 1.9 2.1 2.3 2.7 2.4 Consumption Tax as % of finished consumer goods 5.0 5.6 5.2 10.2 9.8 10.5 12.0 12.0 15.3 15.3  j . lb . ~..". ...... Table 12 Itfi'ORTS VAUJB 1~12121 I I (atllions or dinars) I I 1 --~ Pron- -'- .ional Plan ProJIICU" l2§2 ~ ~ 196) ~ 1965 ~ ~ ~ ~ 1970 !m !2l! Calli tal Goodll 1$.6 16.1 18.) 22.'b l2.:2 ~0.6 )).9 ~ 1Qdi 21:.2 ~ ~ ~ Agricultural equ1s:-nt. 1.1 0.9 0.7 1.3 2.1 2.2 2.2 0.9 1.7 D.a. n.a. n.a. ..a. Indus~ri&l equi~t lhS 15.8 17.6 20.7 27.8 36.4 31.7 28.7 26.7 n.a. D.a. n.a. • ••• Intermediate and Se.1-tlniabed Goode l2d 12.:l 22.4 ll:l 35.5 ~ 2!hl 2:i ~ aa ~ 63.5 65.0' Sultur 0.1 0.1 0.1 0.1 0.2 1.4 3.0 3.3 3.7 1&.8 5.1 6.6 7.S Soybean 011 Tobacco * 0.8 • 0.6 • 1.0 2.4 1.0 1.5 1.4 2.0 oS 3.9 0.8 4.0 0.7 2.2 0.9 3.4 0.6 4.7 0.6 5.0 0.5 5.6 0.5 Raw cotton Cot.ton thread • 0.6 * 0.8 0.2 0.9 0.2 0.9 0.4 1.2 0.8 2.7 0.8 2.8 1..) 2.5 1.5 2.2 1.5 3.3 1.7 1&.7 2.2 3.3 3.1 1.0 Artificlal riber 0.2 0.3 0.) 0.4 oS 0.7 0.5 0.6 1.0 1.1 1.6 1.7 IS Iron and,~teel tub•• 0.4 2.2 0.8 1.1 1.0 1.9 2.3 1.6 1.0 1.1 0.5 0.4 0.3 Iron and steel bara 1.4 1.5 1.6 1.6 2.6 0.8 3.2 0.8 1.0 1.1 0.5 0.4 0.3 Sheet:3 and profiles 1.3 1.5 1.6 1.9 2.0 2.8 2.1 3.1 1.0 2.1 2.5 2.8. 2.8 Wlre 0.2 0.1 0.2 0.2 0.3 0.2 0.3 0.4 0.5 0.6 0.7 0.7 0.5 Other lh.2 12.8 15.7 1705 24.4 28.8 30.6 31&.2 27.3 32.0 )5.9 38.0 laO.o Energy hl ti hl 1Jt. 5.8 ~ hl 4.0 Ad §.:! ~ hl 1:1 Crude 011 0.5 3.8 5.1 5.1 2.0 3.4 N C):) P~fined producta 5.9 5.9 5.9 6.4 1.5 Coke 0.3 1.2 1.5 1.) 1.5 1.6 1.6 1.6 Otber 0.4 0.5 0.4 0.3 0.5 OS 0.4 0.5 0.5 1.2 1.1& 1.9 1.9 Finished CoMu.er Oooa. ll:! ~ Zl.:.2 ll:i E:2 ~ ~ ~ 11. 6 19.0 ~ l1:.2 l2.:!! S;ynthetic cloth 1.9 1.9 2.0 2.3 3.0 2.6 1.6 1.3 1.6 1.5 0.5 0.5 0.7 Cotton clotb 3.3 • 2.4 2.2 1.9 2.5 3.0 I 3.3 1.3 0.8 0.3 0.) 0.1 0.1 Other 18.6 17.3 18.8 17.3 17.0 16.4 IS.9 19.0 15.2 17.2 18.2 18.2 11.0 ~ i:! 13.4 lL.!! 2:.2 £:l ~ g ~ l!M E& 11.5 10.5 L2 Wheat 4.6 11.3 8.1 4.5 2.2 6.7 6.9 13.1 9.9 10.1 9.6 8.S 6.6 Barley 0.1 1.4 2.4 0.3 0.4 1.3 0.1 2.7 0.6 1.G 1.5 1.4 1.4 Other 0.4 0.7 0.5 0.2 0.1 0.6 0.6 1.6 0.3 0.3 0.4 0.6 1.0 Other Food.tufr. .!Qd 10.4 2.:2. 10.$ !!f.:.!i 2.!l 11.3 ll:.2 ~ 12.0 12.5 ll:.i M Suear 3.1 2.7 2.8 3.0 7.4 3.2 2.8 3.1 2.9 2.7 2.7 2.8 2.1 Other 7.0 7.7 7.1 705 7.0 6.1 8.5 8.9 8.7 9.3 9.8 9.1 U.7 !2!!!: ~ ~ 22:2. 2ld !!2& lli..:2 1)1.2 Il7.1 !lli:i !l:Z&. 153.0 158.0 ~ .Less than DSO,OOO. ~: Statlstlques du Ca..arc. Ixterieur and S.E.P.E.N. ~. !  - 29 - Taxes On Domestic Production 64. Taxes on domestic production played a relatively small role in the early sixties. The combined receipts from the consumption and production tax on domestic products accounted fo r only 7 percent of tax revenue in 1961. The question then is whether between 1961 and 1968 the declining role of import taxes was accompanied by a process of import SUbstitution significant enough to generate at least a partially offsetting trend in domestic taxes. If we look at the trend in imports we have noti,;ed a decline, towards the end of the period, in crude oil and oil products; a decline, also in absolute terms, of finished consumer goods, in particular textiles products, and a practical stagnation of food imports, other than cereals. 65. Generally speaking in Tunisia one cannot link a decline in imports exclusively to the empact of import substitution because of the foreign exchange constraint, following the 1964 devaluation, which had the result of Simply reducing private consumption of a number of goods.!! However it is possible to identify some sectors where import substitution took place. Oil production increased from 1 million dinars in 1964 to 17 million in 1968. At the same time petroleum refinery products also increased in value between 1964 and 1968. Textile production between 1960 and 1964 rose at a rate of 10.5 percent. Finally, the food industry, the single largest sector in manufacturing, also increased very rapidly during the 1960-65 period while imports remained practically stagnant. After 1965, the value of domestic production, partially under the influ- ence of a series of bad harvests, ceased to grow while imports stagnated after a sharp rise in 1966 as a result of the already mentioned foreign exchange constraints. 66. The discovery of oil implied almost a net gain for the country in terms of government revenue in view of the special treatment enjoyed by crude oil and oil products under the import tax system. This impact was felt indirectly through the production tax and more directly through the taxes on oil companies which increased from nothing to 6.2 million dinars between 1966 and 1968. 67. The other two taxes on domestic production are the production tax, or turnover tax, and the consumption tax on domestic products which, as we saw before, are also levied on imported goods. !I Private consumption between 1965 and 1968 only increased at about 2 percent annually, as against a growth rate for the 1961-68 period of S percent.  - 30 - Table 13 Consumption ,and Production Tax on Domestic Products - 1960 1961 1962 1963 1964 1965 1966 1967 1968 ----uniIliOn'diIia"rs)- Production Tax 3.3 3.4 3.4 3.5 4.3 5.1 6.2 6.7 7.1 Consumption Tax 0.3 0.4 0.4 0.7 0.8 1.0 1.3 1.7 1.9 3.0 :nr 3.S" 4:2 ~ b.I 73" tr:4 9.0 As %of GDP at Factor Cost Consumption Tax 0.1 0.1 0.1 0.2 0.2 0.2 0.3 0.4 0.4 Production Tax 1.2 As W % of 11anuf acturin~ - 1.1 1.2 1.0 1.0 IT 1:2 1.1 W 1.2 1.4 1.4 1.5 1:7 l.9 1.1r 1.4 Consumption Tax 0.8 0.9 0.9 1.5 1.5 1.72.0 2.5 2.6 Production Tax 8.6 7.4 7.9 7.7 8.0 8.69.7 10.0 9.7 9.4 nJ 'lr.1r 9.2 93 10':3 11.7 I23 12.3 Manufacturins as ~ of ~DP at factor cost 13.5 14.3 13.1 13.0 14.4 13.8 14.8 14.8 14.9 68. In view of the fact that manufacturing, which ~ have taken as basis for these taxes, has increased faster than total value added, this should have insured automatically a higher than one elasticity of consumption and production taxes in relation to total value added. This has actually occurred. If we now look at the relation between the two . taxes and manufacturing we also observe a higher than one elasticity, or rather buoyancy, which is explailled by the increases in the rates which occurred throughout the period.!! 69. A factor which helps explain why production taxes, despite the increase in the rates, have not shown a higher buoyancy is that a number of goods, in particular minerals and foodstuff, obtain a tax rebate which obviously reduces the built-in elasticity of the tax. The fact that j,~ practice the calculation and payment of tax rebates have not worked very e£'ficiently has somewhat offset this down't'rard bias in the grOt,fth of the production tax. 11 The rates of the consumption tax which discriminates among different kind goods were raised as follows: con~~tion Tax A Abis A. Ter 19 2 10 T 1963 15 10 5 1965 16.5 11.0 5.5 1966 20.0 14.0 7.0 1968 21.0 14.7 7.35 1969 23.0 16.0 8.0 The rates of the production tax were also raised in 1965 (from 10 to 11 percent), j.n 1966 (to 12.5 percent) in 1968 (to 13.13 perc~nt) and in 1969 (to 14.4 percent).  - 31 - 70. A predictable effect of import substitution is automatic loss of revenue for the government insofar as domestic indirect taxes are normally levied at a lower rate than clls-r,oms tariffs. 1/ Such loss can only be offset by a series of tax increases in domestic taxes. In the case of Tunisia the increase in the rates of consumption and production taxes, which apply also to imported goods, has provided for an automatic increase in import based taxes, thus compensating in part for the reduction in the level of imports. Income Taxes 71~ The behavior of income tax proceeds between 1961 and 1968 has to be considered in relation to two separate periods: 1961-1965 and 1966-1968. In 1966, in fact, as a result of the extension of business income taxes to public enterprises, revenue from business taxes increased from 5.6 million to 9.6 million dinars. Other tax changes, as we shall see below, occurred during this period, but this was certainly the most significant one. It is, therefore, necessary to consider separately trends in taxes on individuals and business. 72. If we use GDP at factor cost excluding agriculture as the "second best" tax basis, we see that the progressive income tax shows a higher than one buoyancy until 1966 with a decline during the past two years.E! The peak level of 1966 was the result, basically, of the reduction in the basic exemption from 180,000 to 100,000 dinars. There is really no good explBJ'lation for the 1967 decline,while for 1968 the tax deductible compulsory loan issued in that year explains why the tax ratio failed to increase faster than the tax basis. There is also a significant innovation whose impact was probably felt more at the end of the period than in the initial years after its introduction in 1962. This is the provision which allows individuals an exemption up to 30 percent of their net taxable income if this portion is invested. in shares or bonds of various companies, the list of which is provided by the government. Y See the interesting article by Stephen U:!wis, "Revenue Implications of Changing Industrial Structure -- An Impirical Study", National Tax Journal, December 1967. 2/ As mentioned in Section D, we have also calculated the ratio of tax - revenue to GDP excluding agriculture and crude oil production (see Table 17). -  - 32 - 73. The purpose of this disposition was the obvious one of helping develop r~ capital market in Tunisia by channeling private savings into the industrial activities which the government was trying to initiute and expand. Since, for obvious reasons, only the middle and upper income groups were in a position to buy such shares or bonds, the ~esult has been a reduction of the effective progressive- ness of the income tax and therefore a weakening of its built-in elas- ticity.!! 74. As to withholding taxes on wages and salaries, their ratio to GIP at fe-.ctor cost has increased slightly towards the end of the period arte~ a depression during the 1963-1965 period. The decline in 1963 was the result of a reduction in the tax rate from 5.5 percent to 3.3 percent and of an increase of the exemption from 150 to 180 dinars. The increase in 1966 was due to the same kind of action, obviously in the opposite direction: an increase in the rate to 4.4% and a reduction in the exemption from 180,000 to 100,000 dinars. In 1967 there was again a 10 percent increase in the rate affecting the progressive income tax as well. up to 1967 the tax had been levied at a uniform rate. In 1968, a system of graduated rates was intro- duced with the minimum of 4.6 percent riSing to 5.75, 6.90 and 8.10. In view of the uniform rate, prevailing for most of the period, the withholding tax on wages and salaries has behaved according to expec- tatiens since short-falls are explained by changes in t,ax rates and exemptions. The relatively few difficulties in the administration of this tax, in particular the absence of tax evasion, have certainly contributed to its smooth operation. Tax evasion has certainly been more of a problem in the case of the progressive income tax, in parti- cular with regard to self-employed categories. 75. As already mentioned, business taxes show an apparent high buoyancy which is the result of the reform of 1966. The tendency, however, and allowance is made for such dramatic changes, is toward a buoyancy of one or less. If we take the one-year lagged relation, the pattern does not change. The system of business income taxes, as it appears from Table 15, is rather complex and it may be useful to describe it briefly. 76. The fixed fee (Droit Fixe) up to 1968 was a function of the type of activitY$ of the enterprise, of its size, number of employees, etc., its location, etc. It represented in fact a minimum tax on the enterprise regardless of its actual profits and is deductible from the 1/ Even if we take a one-year lag relationship between income tax - proceeds and GDP, the pattern of buoyancy is basically the same. If 1968 shows an increase in the ratio higher than the one in- dicated by the unlagged series, it is because estimates for 1969 have been used for 1968, and estimates tend to be optimistic with respect to actual collections (see Table 3).  - 33 - Droit Proportionnel (Proportionate Rate). As of 1969, it has bean renamed Droit DI EK:ercice • Its basis has also changed in that it is now constituted by the value of sales (on cash or credit basis) and it is levied at a rate of 1 percent (which becomes 0.5 percent when the value of sales is below a certain level). 77. The Droit Proportionnel (inclusive of given surtaxes) is levied on the net profits. The following table shows how the dif- ferent rates are applied. The system of differentiated rates was an illllovation introduced in 1966 with the purpose of giving incen- tives to a number of sectors, in particular to industry and tourism. Table 14 . Droit Proportiormel1l As of January 1969 Tax Payers Rates Individual and Partnerships Acti ve in Trade 46.50% Active in Industry, Handicraft, Tourism and Transport 29.00% Comp,anies General Rate 46.50% Active in Industry, Handicraft, Tburism and Transport 40.10% For joint stock companies, temporary advances (Acomptes Provisionels) are paid in the course of the year during which profits are realized. '!hey are based on taxable profits of the previous year and are pai,d in two instalments of 40 percent each of the total due (after payment of the Droit Fixe) on the preceding year' s profits. Final payment (or reimbursement) is made at the time of filing the annual tax return. 78. Various exemptions under the business tax are contemplated. In particular, companies which have been granted a ·'Lettre d'Etablisse- ment" (under the Investment Code provisions) are exempt from the tax for Q period of 5 years of business activity which may be prolonged to 1I Small business and handicrafts pay a rate of 11 and 16.5% respectively, according to whether their profits are belo't'1 360 or 600 dinars per annum. The National Handicraft Office and Handicraft Cooperatives pay a rate of 29% while Pharmacies were taxed at 31.5 and 46.5 percent based on profits below or above 9,000 dinars per annum.  - 34 - 10 years only for the portion of profits reinvested in the enterprises. The other important exemption concerns the portion of income which is reinvested either for the purchase of equipment, construction or exten- sion of buildings and purchase of stocks out of a list approved by the authorities. 79. It is not easy to draw general conclusions on the behavior of business taxes for the entire period. It is paradoxically easi~r to explain some of the major fluctuations in the tax year by year ..!! In the presence of so many changes in tax structure and coverage, the historical built-in elasticity of business taxes is certainly not iden- tifiable ex-post. The systems of exemptions, the special rates for certain sectors, the admitted existence of tax evasion indicate that such elasticity was prob~bly rather low. At the same time, one should not forget that, from the point of view of revenue collection, the government has still been able to increase collection faster than the growth rate of the economy_ Truly, without the major reform of 1966, revenue of business taxes would have probably declined further in terms of GDP. But it is also true that under the pressure at rising current expenditure, the government would have had to introduce changes apt to increase the proceeds from these taxes, and therefore offset the low built-in elastiCity, at least to some extent. 80. ~ the same token, a less generous policy of tax exemption could have probably produced faster increasing revenue, at least in the short run. This is a very important question in Tunisia and has quite some bearing on the future prospect,s, as we shall see in the concluding chapter of this paper. We believe that, with all the doubts which normally accompany the effectiveness of the various incentive systems, these were a necessar,y tool to foster development and one should give the Tunisian Government the credit of having followed a fairly elastic t,ax policy in the j.nterest of developing the industrial and tourist sector. If anything can be criticized it is rather the profitability 11 In 196), for instance, the drop in collection was mainly due to the suppression of the system of advances for individual firms (see line Others in Table 15). This also explained the sharp increase under the proportionate rate in 1964. We have already given the reason for the rise of 1966. As to 1967, there was the suppression of the special rate under the Fixed Fee for small businesses and probably the impact of the favor rate for tourism and industry introduced in 1966. 'Ibis same impact plus the tax deductible compulsor,y loan explains the decline, in absolute terms, which occurred in 1968. In this year, the increase in receipts from taxes on small business was certainly favored by a rise in the rates from 25 and 27.2 percent to 26.25 and 42.30 percent.  - 35 - of some of the new' ventures for 't"lhich tax benefits were granted. (Invest- ments in metals and textiles are good ey~mples.) The disappointment which even some Tunisian officials have expressed' in the role of fiscal incentiveo is to this extent misplaced and is perhaps more a function of inadequate evaluation of investment proposals.  - 36 - Table 15: Income Taxes (Thousand Dinars) 1961 1962 1963 1964 1965 1966 1967 1966 i9t? _ 4,153 3,839 4,85$ 5,$24 6,353 8,696 8,709 9 z 495 1.1 to7,IQ C~nera1 Progressive I::.:comeTax 2,473 2,033 3,360 3,776 4,697 5,998 5,507 5,680 7,.l~"~ "..C lrJJ.thho1ding on wages a.nd Salaries 1,680 1,806 1,495 1,748 1,656 2,698 3,202 ),815 4.620 gN ~USINESS 52 681 2z0~2 4z282 22 229 5,559 92643 10, 209 10,010 12, t~,2, Eu:iness Income Taxes 4,914 4,)56 3,569 4,490 4,720 8,829 9,237 8,844 11~2~O Fixed Fee (1,407)(1,251)(1,026) (l,075)(1,174)U,336)( 856)(1,038)(1,000) Proportionate Rate ( 717)( 662)( 864) (1,299)(1,428)3,522)(3,435)(2,675)(3,960) Advances (1,420 )(1,134)(1,272) (1,481)(1,456)(3,067 )(4,344 )(3,830 )(4.~ 20C) Others (1,365)(1,309)( 404) ( 594)( 660)( 903)( 602)(1,300)(2,j90) on Profits of '[lax Non-Commercial Professions 299 218 185 231 242 214 170 223 290 Tax on the Earnings of Stocks and Bonds 468 528 508 597 600 802 943 1,009 r1'0TAIJ 9,834 8,881 9,137 10,753 11,91218,339 18,918 19 z505 ~2. a:.--.:::,~.: - .  - 37 - Table 16: INCOME TAXES WITH ONE YEAR LAG (thousand dinars) 1961 1962 1963 1964 1965 1966 . 1967 __2 68_ 9lt INDIVIDUAlS 3,713 5,166 ~27i 6,445 7,6,4 8,205 8,882 IJ~~i General Progressive Income Tax ¥ 2,033 3,360 3,776 4,697 ,,998 5,507 5,680 7~);;O vlithholding on Wages and Salaries 1,680 1,806 1,495 1,748 1,656 2,698 ),202 3~a15 OK BUSINESS 5,042 4,280 5,228 5,558 9,6~ 10,209 10,009 12c~:~ Business Income Taxes 4,356 3,567 4,489 4,719 8,829 9,237 8,843 11~2~O Fixed Fee !I (1,251) (1,026) (1,075) (1,174) (1,336) ( 856) (1,038) (1~00J) Proportionate Rate!l( 662) ( 864) (1,299) (1,428) (3,522) (3,435) (2,675) (:,;:0) AdVances 6( (1,134) (1,272) (1,481) (1,456) (3,067) (4,344) (3,830) (4,200) ~ther - Proportion- ate Rate 6( ( 627) ( 346) ( 600) ( 621) ( 904) ( 602) (1,300) (2,05'0) Other - Advances!l ( 682) ( 59) ( 34) ( 40) Tax on Profits of N')n-Commercia1 218 185 231 242 214 170 223 290 Professions !I Tax on the ~rnings of Stock and Bonds Y 468 528 508 597 600 802 943 T(~ 8,755 9,446 10,499 12,003 17,297 18,414 18,891 ~,~,------------------~-------------------------- -------- ~: Go11ection of 1962 appears in 1961, 1963 in 1962, etc.  - )8 - Table 17: mOGlE Tm§. w",. 1961 1962 1963 1964 1965 1966 1967 1968 1969 ~.~'~ IncC&ft Taxes as % or 4.0 4.6 - GnP 2t Factor Cost 1:1 2.7 - 2.6 ~ &! 4.2 2.0 4.2 1.9 1.9 2.2 Or" Individuals 1.3 1.2 1.4 1.S 1.5 On Business 1.8 1.5 1.2 1.4 1.) 2.2 2.3 2.1 2.4 ~~s Taxes as ! or OJ'? 5't Fa ctor Cost !ic~dinS Agriculture 4.1 1:.2 l:!! 3.7 3116 5.2 U ~ 2.:i (Excl.Agric.& Crude Oil) (4.1) (3.S) ().4) (3.7) (3.6) (,.2) ('.~l) (5.0) (5.8) On Individuals 1.7 1.5 1.8 1.9 1.9 2., 2.2 2.) 2.6 (1.7) (1.5) (1.8) (1.9) (1.9) (2.5) (2.4) (2.4) (2.8) Progressive In- come Tax (1.0) (0.8) (1.2) (1.3) (1.4) (1.7) (1.4) (1.4) (1.6) (1.0) (0.8) (1 •.2) (1.3) (1.4) (1.7) (1.$) (1.4) (1.7) Wi thho1ding on wages & Salaries (0.7) (0.7) (0.6) (0.6) (0.,) (0.8) (0.8) (0.9) (1.0) (0.6) (0.5) (0.8) (0.9) (1.0) (l.l) On Business 2.4 2.0 1.6 1.8 1.7 2.7 2.7 2.5 2.9 (1.8) (1.7) (2.7) (2.7) (2.6) ().o) :illcc:me Taxes with One IEC!!:.'LlS ! 8S of GDP a1; F~ .....-.. C'" or Cost Bxclud- ., ........... ~~~icultura (:G~::~luding Agric. 3. 6 1:1 - 4.0 41'2 S.2 ,.2 U 2.& & ,;rl.lde 011 ) (4.2) (S.2) (,.2) (,.1) (6.0) o~~, J:,ldivlduals 1.S 2.0 2.0 2.3 2.3 2.) 2.3 2.7 (2.3 (2.3) (2.3) (2.4) (2.8) Progressiva In- cane Tax (0.8) (1.3) (1.4) (1.7) (1.8) (1.5) (1.,) (1.6) (1.7) (1.8) (1.,) (1.5) (1.8) Wi thholding on Wages & Salaries (0.7) (0.7) (0.6) (0.6) (0.,) (0.8) (0.8) (0.9) (0.6) (O.s) (0.8) (0.9) (1.0) On Business 2.1 1.7 2.0 1.9 2.9 2.9 2.6 ).1 (1.9) (2.9) (2.9) (2.7) (3.2)  - 39 - Agricultural Taxes 81. Agricultural taxes have never accounted for a large share of total tax revenue. They represented about 5 percent in 1961 and their shar-e decliu'Jd to less than 2 percent in 1968. Their constant decline in absolute terms between 1961 and 1968 reflects an obvious lack of elasticity with r-espect to GDP. It is true that at the same time the role of agriculture in total value added also declined rather sharply. However agricultural taxes as percentage of value added in agriculture also declined during this period. Before enter- ing a more detailed investigation of the various taxes on agricultur-e one should stress that year by year data are often grossly disto:rted by delays in collection and by the overlapping in calendar years of the production cycle fl'c.., ...~ the point of view of their paying the tax. It is practically impossible to have a good statistical sar-ies on such taxes. The:refore their ratio to ag:ri- cultural production should be taken on an average basis for- a number of years. The statistical difficulties have been magnified by wide year-to-yea~ fluc- tuations in agricultural production. Table 18: .A.~ricultural Value Added and Taxes 1961 1962 1963 1964 1965 1966 1967 1968 Agricultural Value added as % of GDP 24.8 23.0 24.2 23.2 22.9 17.3 14.5 17.0 Agricultural Taxes as %of Agricul- tural V.A. 3.3 1.6 2.9 2.8 1.9 2.3 1.4 1.8 82.1 The main change in the structure of agricultural taxes has been the introduction in 1963 of the agricultural tax which has substituted previous taxes on livestock, palm and date trees, orchard products. This tax is levied on the value of gross receipts resulting from the sale of fruits, vegetables, livestock, poult:ry farms, agriculture, etc. There are two different rates, one on truck gardening (recently :revised to 4.20 percent from 3.8 percent), and one for other cultivations (also recently r-aised to 5.50 percent from the previous level of 5 percent). Fbr animal husband~y minimum r-ates are estab- lished for the var-ious categories of animals.!! Table 19: Agr icul t ural Taxes (thousand dinars) 1957 /b . /58 1960- 1961 1962 1963 1964 1965 1966 1967 1968 1969 Ag: i;mltural m~~ -:-:- 4ffiJ 4i4 693 464 401 434 388 371 495 '1;ro~ on olives 349 •• 1115 263 484 731 714 539 153 512 1100 TaJ\: on cerea.J.s 65?. 581 103 868 1073 739 624 342 484 550 Tax on vineyards 368 417 389 484 182 13 48 127 III 275 roTAL 1825 2200 2599 1169 2529 2450 1867 1645 1010 1478 2420 L! The agricultural tax was only introduced in 1963 covering previous taxes on livestock, palm and date trees, orchard p~oducts. LP.. No breakdown a.vailab1e as a. result of the change from fiscal to calendar year which only gives a 9-month year for 1960. !I See for: details Bulletin Offici'31 de lao Direction des Impots, No .1, P 7,1969  - 40 - 83. The yield of the new tax after 1963, has been declining in absolute terms, and also in proportion to the value of production. The reason currently given for this is that since the tax is based on the declaration by taxpayers of their gross receipts there is greater scope for tax evasion. For one thing the large number of small farmers has made the administration of the tax rather difficult. Secondly, farmers tend to pack good accounts and thirdly, statistical data on the yields of various areas and crops have not been good enough for a sufficient period of time to provide the tax authorities with a good basis for tax assessment. 84. As to the tax on cereals, now levied at a rate of 7 percent on marketed output (it was 6.6 percent in 1965), its decline in absolute terms after 1964 is explained by a large amount of tax evasion (stocks being sold without going through the government marketing boards) and by increased direct consumption of the farmers. In the presence of increased population the average annual production of wheat for the 1966/68 period was 437 thousand tons as against 533 thousand tons for the 1961/63 period. 85. Collection of taxes on olives has followed the fluctuations in production from the point of view of the direction of the change. The', tax is also levied at a rate of 7 percent (616 percent in 1965) on the value of olives when they are brought for oil production or for exportation. Expressed in percentage of value of production, the tax has lost some ground but it is really difficult to reconstruct a proper trend in view of the timing of the crop which tends to overlap two fiscal years. Table 20. Value of Agricultural Production and Taxes (Million Dinars) 1961 1962 1963 1964 1965 1966 1967 1968 Cereals 21.5 25.3 31.4 26.3 32.0 21.2 19.8 23.8 Tax on cereals as % TI D:4 2:7 en 2.J 2:9 1.7 2:0 Olives 1".7 8.1 11.4 15.1 19.3 12.6 5.0 12.7 Tax on olives as % b.2 3.2 T.2 4]" 377 T.2 3I T.O Wine Gra:ees 6.7 8.7 8.3 8.5 6.2 4.1 4.5 4.1 Tax on grapes as % 6.2 4.5 ;:a- U 0:02 1:2" 2Jr n Othersl/ 33.3 33.4 33.6 36.9 40.5 36.6 40.7 43.4 Agricultural Taxes % -r:1i 1:2 2:r 1.2 l.O l.2 0.9' -o:cr 1/ Obtained by substracting the above items from value added. This is - arbitrar,y but there is no continuous series on total value of produc- tion for the sector.,  - 41 - 86. It is eliffieult to explain the behavior of the tax on wine grapes .. The large drop of 1965 could be related to the fact that in that year wine sales to France were terminated, following the nationalization of lands in 1964. Since then sales to France have been agreed on a basis of lower annual quotas than in the tirst halt of the 1960' s. Export Taxes 81. Export taxes have been a rather insignificant source of revenue up to 1964.The production tax on exports is levied at the lower rate of 3.5 percent (raised to 4.5 percent in 1966). This rate applies to a number of goods including fish, iron minerals, leather, cork, alfa, etc. Exports are also subject to a customs tormalities fee, which is now 1.8 percent (it was raised from 1.65 percent to this level in 1967 and from 1.5 to 1.65 per- cent in 1965). Table 21. Expo~t Taxes 1961 1962 1963 1964 1965 1966 1967 1968 Customs Formalities Tax 551 619 665 707 704 789 846 948 Export Tax on Lea- ther, etc. 48 39 28 30 45 45 15 30 Production Tax on Ex- ports 147 113 127 110 142 81 45 51 Special Temporary Ex- port Tax 141 113 127 110 3555 3673 2280 2741 Other 53 113 127 110 3555 3673 2280 2741 ']'OTAL tro5 .. - '1'71 820 B47 Ii4Iib 4594 ll86 3770 88. The most important change in export taxes occurred in 1965 when a special temporary tax was introduced concerning mostly e~orts of olive oil, (10 percent) and phosphates (15 percent). The tax was enacted following the devaluation of 1964 with the purpose of taking away for the goveument some of the additional profits accruing to producers as a result of the more favor- able e.xport market. After 1966, the tax on olive oil was repealed. 'Ihe in- tention ot the government of doing away entirely with the other special taxes, which would have been quite reasonable in view of the increasing difficulties in e.xporting phosphates, did not materialize, mostly because of the pressure for additional revenue resulting from the very rapid increase in current ex- penditure.  -42- Table 221 ~Patt Vt~S 1960-72 (1111 nOM iinars) Provi- sional Plan Projection 1-222 1221 ~ l2§.l ~ ~ ~ mr 1969 1970 Agricultural Products lli£ ~ ~ J2:l ~ ~ ~ ~ Cereals Hard Wheat 1.0 (S.2) 1.6 (1.6) 1.4 (1.4) 4.0 (3.7) ).) (2.7) 0.) (0.) 4.) ().6) 0.1 (*) - (-) (-) - (-) (-) - (-) Barley (1.1) (*) (*) (0.2) (0.5) (*) (0.7) (*) (-) (-) (-) ( -) (-) Others (0.1) (*) (*) (0.1) (0.1) (*) (*) (*) (-) (-) (-) (-) (-) Citrus 1.6 1.9 1.8 1.8 2.2 2.6 2.9 3.3 1.6 2.4 2.6 3.0 3.) Oranges (0.9) (1.2) (1.3) (1.3) (1.2) (1.1) (1.6) (2.3) (n.a.) (n.a.) (n.a. ) (n.n. ) (n.a.) I.emont.l (0.2) (0.2) (0.2) (0.2) (0.4) (0.2) (0.6) (0.3) (n.a. ) (n.a.) (n.a.) (n.a. ) (n.a. ) Others (OS) (oS) (0.3) (0 •.3) (0.6) (0.7) (0.7) (0.1) (n.a.) (n.a.) (n.a.) (n.a. ) (n.a.) Other fruits 1.2 0.8 1.3 0.7 1.6 IS 2.4 2.0 2.S 2.1 2.3 2.6 ~.'j Dates (0.4) (oS) (0.5) (0.2) (0.6) (0.) (0.9) (0.6) (0.6) (0.6) (0.7) (0.9) (1.1) Almonds (0.6) (0.3) (0.6) (0.3) (0.8) (1.1) (1.3) (1.1) (1.1) (1.1) (1.1) (1.1) (1.1) Others (0.2) (*) (0.2) (0.2) (0.2) (0.4) (0.2) (0.3) (0.2) (0.4) (oS) (0.6) (0.7) Anil!tal Products 1.9 1.6 1.6 1.7 1.1 2.0 2.1 3.0 2.6 1.6 1.9 2.2 2.8 Live animals am meat (1.1) (0.8) (0.9) (0.9) (Q.~) (1.4) (1.6) (2.2 ) (1.0) (1.0) (1.0) (1.0) (1.0) Fi::h and crustaceans (0.8) (0.8) (0.1) (0.8) (0.6) (0.6 ) (oS) (0.8 ) (0.6) (0.6) (0.9) (1.2) (1.8) Olive oil 5.8 9.6 12.8 9.9 n.9 13.5 13.4 0.0 n.9 12.6 12.6 12.6 Bevera.ges Wine 105 (1.3) 13.0 (1.9) 1.7 (7.6) .10.3 (10.0) _~.6) (6.6) 2.8 (2.6) ".4 (3.2) S.2 (L.2) 3.6 ().1) 305 (n.a. ) 3.5 (n.n.) 3 •.5 (n.a.) ;.:; (h.ll.) Other !}/ (0.2) (0.1) (0.1) (0.3) (0.4) (0.2) (1.2) (L1) oS (n.a.) (n.a. ) (n.n.) (n.a. ) Preoerved vegetables 0.4 0.4 0.5 1.2 1.4 1.6 1.4 1.8 1.2 1.4 1.8 2.4 <.:.8 Htu'iosa (0.3) (0.3) (0.3) (0.3) (0.) (OS) (oS) (0.3) (oS) (n.a. ) (n.a.) (n.a.) (n.a.) Othera (0.1) (0.1) (0.2) (0.9) (1.1) (1.1) (0.9) '1.$) (0.7) (n.a.) (n.a.) (n.a.) (n.n.) Preserved fruita 0.8 0.8 1.2 0.7 1.2 0.6 0.9 0.7 0.6 OS 0.7 J.8 1,,; Fl"llit juice (0.1) (0.2) (0.3) (0.1) (*) (0.2) (n.a.) tn.a.) (n.a. ) (n.a. ) (n.a.) (n.a.) (n.a. ) Otters (0.1) (0.6) (0.9) (0.6) (1.2) (0.6) (n.a.) (n. a.) (n.a. ) (n.n. ) (n.a.) (n.a.) (n.a.) ?reserved tiah 0.8 0.6 0.1 0.6 0.6 0.3 oS ..1.3 0.1 0.2 0.6 mnerals a txi Manufactures ll:l 1Z.:.2 &§. l2:l ~ JQd !J2d 21:§. !!h!L Phosphate prodUcts 8.9 10.3 9.2 12.2 21.1 18.3 2j.1 22.4 27.5 28.6 2-i.7 ;5 • .1 Rock phcaphate (S.9) (5.9) (6.6) (605) (11.4) (12,,) (12,s) (12.2) (1605) (16.9) (17.1) t1t.4) Super phosphate fertilizer (2.6). (3.9) (2.3) (2.6) (8.6) (5.2) (11.1) (10.0) (10.8) (1105) (l~ .t:) (Ii.:;) Ot.',(,1' ;fertilizer (0.4) (OS) (0.3) (0.9) (1.1) (0.6) (0.1) (0.2) (0.2) (0.2) \ "..;) \ .::) Phoophoric acid (3.d) Otter minerals 4.6 4.7 3.6 4.0 6.3 S.6 4.6 4." 4.2 4.2 ~s Iren ora (3.3) (3.0) (2.4) (2.4) (3.1) (2.9) (2.4) (2.0) (1.7) (1.5) (loll) Lead (1.4) (1.5) (1.0) (1.3) (2.4) (1.6) (1.4) (1.6) (1.8) (1.8) (1.9) Salt (0.1) (0.2) (0.2) (0.2) (OS) (0.4) (0.4) (OS) (OS) (oS) (u.S) Miscellaneoll,s (-) (-) (0.3.) (0.1) (0.3) (oS) (0.6) (0.2) (0.2) (0.4) (v.n :ntrle oil 4.2 10.5 14.1 20.9 20.0 18.4 Il'Qt. and steel producta 0.3 1.9 2.7 1.7 1.7 1.9 1.9 Iron and steel bars (-) (-) (-) (-) (0.4) (1.) (1 •. 3 ) (105) (loS) Cast Iron (-) (-) (-) f<2.7) (0.3) (1.5) (0.4) (0.4) (0.4) (.1.4) Cellulooe 1.1 1.6 1.8 1.7 2.9 3.1 3.0 c.5 Not Inc.Juded Above ...2&. §..:.2 &:.2. 1:l 2.:.Q. ~ 11& lld Fresh vegetables 0.4 0.2 0.4 0.3 0.4 0.8 0.5 .0.6 n.a. n.a. n.a. n.a. Esparto grass 1.3 1.0 1.0 0.7 0.9 0.4 o,s 0.4 n.a. n.a. n.a. n.Il.• Sugar produo ts 0.1 0.1 0.2 n.a. n.a. n.a. n.n. n.n. Cork and prorlucts * 0.4 * 0.4 0.3 0.6 0.9 * 0.8 * 0.7 C.7 n.a. n.a. n.a. n.a. rata 0·3 0.1 0.1 1~ 0.4 n.a. n.a. n.a. n.n. Pasta 0.2 0.2 0.3 0.2 0.1 * 0.1 0.1 n.a. n.a. n.a. n.a. n.a. Cement 1.0 0.4 0.6 oS 0.2 n.a. n.a. n.a. n.n. Miscellaneous 6.0 3.7 4.1 4.9 6.3 * 9.1 * 9.4 n.a. n.a. n.a. n.a. n.a. n.n. TOTAL EXPORTS SO.3 ~ ~ 21l ~ l!h!l !I In 1966 and 1';67 exports ai' wine as indemnlzation to French farmers ;for nationalized la!l):! mre included. '!if or which refined petroleum 'products ro.3 million, textiles and rugs Dl.2 million, and ceramic artic1ea 00.5 million. Source: Statiatiquell du Co!llTlerce Exterteur am S.E.P.E.N.  - 43 - Consumption, Service and Registration Taxes 89. There is a group of consUJll]ption taxes which include a number of excises, the most important being thl~ oil and gasoline tax and fiscal mono- polies, mostly tobacco. (See Table ~3.) On the whole, these taxes have shown a rather modest buoyancy throughout the period. The main reason is the very sluggish behavior of the oil and gasoline tax. This is e~lained essentially by the introduction in 1963 of detaxation for uses in agricul- ture as a feature of the incentive policy of the government towards this sec- tor. The law of December 31, 1962, provided for total or partial exemp- tion from the tax, in favor of individual farmers as well as to agricultural cooperatives. One should also notice that the sharp increase of the tax in 1968 is mostly due to liquidation of arrears tor 1967. Table 23. Consumption Taxes (million dinars) Fiscal Monopolies 8.8 8.2 9.1 9.9 10 . 4 12.0 13.4 Oil and Gasoline Tax 5.6 5.6 5.7 5.7 5.9 5.8 5.9 8 -I Spice 0.8 0.7 0.9 1.1 0.9 1.2 1.5 0.9 Alcohol 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 Others 0.3 0.3 0.3 0.4 0.5 0.5 0.4 0.4 TOTAL 16.0 15.3 16.5 17.6 18.2 20.0 ,g.~.,.? - - As% of GDP at Factor Cost 5.0 4.7 4.7 4.7 4.3 4.6 4.8 5.1 of which: Fiscal Mono- polies 2.8 2.5 2.6 2.6 2.4 2.8 3.0 3.0 Oil and Gaso- line Tax 2.0 1.8 1.7 1.6 1.4 1.3 1.5 A.s% of Priva te Con- sumption Fiscal Mono- polies 3.3 2.9 3.1 3.2 3.0 3.6 3.7 4.0 90. As to fiscal monopolies the very high increase in 1968 is e~lained by the integration of a special fund (for about 1 million dinars) which up to 1967 appeared in the Treasury Accounts. Major increases in the rates were  - 44 - introduced' in 1964 and in 1966. If we relate the tax to private consumption} rather than to GDP at factor cost, we have a slightly higher buoya.ncy. • 91. No particular comment can be made on the registration -taxes except for noticing that between 1961 and 1968 they failed to grow in line with GDP at factor cost, even if oil production in 1968 is excluded. As to the servi0ea tax, one has to stress its very high buoyancy, starting in 1965, both with respect to ODP at factor cost and to value added in the services sector. ThJ tax is levied at two basic rates (now 5.8 and 2.3 percent the preferential one for liberal professions and contractors) on services performed by profes-. siona1 people or by companies. The exceptional increases in the proceeds of the tax are largely due to the changes in the rates which in 1961 were resper.·· tively 2.5 and 1.5 percent.  - 45 - CHAPTER III PROSPECTS UIIDER PRESENT TAX Scr'RUGTURE 92. After having examined the main determinants of tax elasticity during the 1960-68 period, we can now look at the future and try to idel1tif~r likely trends in revenue given the present tax structure on one side and the expected evolution of the economy on the other. In a final section we sha.l1. try to explore the possibilities and the implications of certain adjustments in the tax structure which might improve it.s built-in elasticity. 93. In discussing the past, we had identified three main areas in which the elasticity was rather weak with respect to GDP at factor cost: agricultural import and consumption taxes. In addition, we had concluded that also in the case of direct taxes on individuals and corporations, the elasticity, though higher than one, was probably below its expected coeffi- cient as a result of the various exemptions granted and of the compulsory loans of 1964 and 1968 • At the same time, the oil sectors had becoll'E, starting in 1965, a very dynamic component of value added and also a growing source of revenue for the government. The first step, in looking at the future, is to investigate the expected growth of the economy and see whether it implies any Significant change as far as the structure and growth of revenue are concerned. 94. The Fourth Plan, 1969-1972, cons.titutes a good framework for a mediwn-term analysis of the prospects of the Tunisian economy. Gross domestic product at factor cost is projected to increase at 5.8 percent annually between 1969 and 1972. Agriculture and mining are supposed to grow at a lower rate; the oil sector in particular, after the rapid growth in recent years is expected to increase at 3.1 percent annually. At the same time, however, imports of goods, excluding cereals and oil, are projected to grow faster than GDP, at 7.5 percent, owing in large part to the reduction in imp0rts that occurred in 1968 and the abnormally low level of the base year. 95 e We have followed the sane method used fa r the analysis of the historical experience and set up a table showing the structure of tax revenue and the expected growth of the various sectors of origin. In the table both 1968 and 1969 are shown as base years. In view of the fact that 1968 was a ra~r exceptional year, both from the' point of view of tax collections (the 1968 compulsory loan, the integration of the Special Funds inw ordinary revenue) and of the growth in the econonlic basis (absolute decline in imports) we have chosen 1969 as the starting point for our detailed analysis. This approach has the obvious disadvantage of presenting estimates rather than actual data but it should be adequate from the point of view of identifying the relation between changes in the economy and in the tax structure. 96. Table 24 shows a built-in elasticity of the tax structure barely above 1. This is roughly comparable with the expected elasticity tor the  - 46 - 1961-1968 period which mdght be surpr1S~g if one considers that imports excluding cereals and oil products are projected to grow faster than the growth rate of GDP at factor cost. The main reason is that the 1969 tax structure includes oil revenue for 'Which Ie. moderate growth rate of only • 3 percent is projected after the sharp increase between 1966 and 1968. The possibility of new oil discoveries cannot be entirely ruled out at the lOOment and some uncertainty thus surrounds a significant component of the economic structure which could affect the built-in elasticity of the tax system to a very large extent. 97. The same uncertainty applies to the import projections, at least for the Plan Period. But even regardless of the possibility that because of foreign exchange constraints imports will have to be restrained, the present projections alre~ contain a dm~ward bias with respect to the elasticity of import taxes insofar as the role of finished consumer goods is expected to decline from 17 percent in 1968 to 15 percent in 1972 continuing the trend of recent years. Obviously, the decline will be much sharper if total import will have to be reduced from the expected level. 98. Even assuming that the Plan Projections will be realized, the 1972 ratio of imports (excluding cereals and oil) to GDP at factor cost will be 20 percent as against 24 percent in 1960. (The ratio reached a peak of 27 percent in 1966). The question is whether, looking at the longer run, we may expect the ratio of imports to GDP to stabilize if not decline, regardless of foreign exchange constraints. The evolution of the industriali·· zation process is a very important factor bearing on the import trend. If industrialization proceeds rap~ we may expect a rising trend in imports of capital and particularly intermedi~ goods, accompanied by a slower growth rate of finished consumer goods due to ir@ort substitution.11 The revenue implication of this trend. is a gradual loss of the built-iii elastiJ~ city of import taxes with respect to imports, but a probable gain of elasticit,r with respect to GDP at factor cost (iDe., total imports will increase faster than GDP). If the industrialization process proceeds more slowly the demand for intermediate goods will grow less fast, while the demand for finished conSUlOOr goods lo10uld be more sluggish because of lower growth of gross domestic product and, tb.erefore, of private constunpticn~ The revenue implication of this second growth pattern is a lower elasticity of import taxes with respect to gross domestic prod1.1,ct which will certainly more than offset any gain in elasticity of import t.a.xes with respect to imports resulting from the fact that the role of finished consumer goods, the more heavily taxed ones, would decline less fast than in the previous case. 11 Imports of consumer goods are projected by the Plan to remain ullchanged, in absolute terms, and this follows a period of an absolute decline since the early sixties. It is conceivable that as a result of a success- ful industrialization accompanied by an increase in foreign exchange earnings, imports of consumer goods could i~ttially increase above the longer run trend.  - 47 - Table 24: TUNISIA - EXPECTED BUILT-IN ELASTICITY OF TAX STRUCTURE BASED C!l ACTUAL GROWTH RATES OF SECTORS OF ORIGIN A B AxB Expected Cont.rI'bution As % Growth to Growth of Rate of of Total Total Tax Base Revenue Taxes 19b5 19b9 Tax Base 1908-72 19b9-72 !~b8-72 1909-72 Import 25.7 27.0 Imports of goods excl. 10.0 7.5 2.57 2.03 cereals and crude oil Agricul ture 1.6 2.1 Agriculture 5.7 3.8 0.09 0.08 Services 3.8 3.7 Services 6.4 6.4 0.24 0.24 Consumption 26.4 22.0 Private Consumption 6.2 5.2 1.64 1.14 1/ Exports- 1.1 1.0 Exports 7.7 5.7 0.08 0.06 2/ Transport- 0.8 1.3 Transport and Comma 4.8 5.2 0.04 0.07 Domestic Products 9.4 9.4 Manufacturing 8.1 7.1 0.76 0.67 Income Taxes 20.4 21.6 GDP excl. Agric. 6.7 6.2 1.37 1.34 Transaction 4.4 4.3 GDP, factor cost 6.5 5.8 0.29 0.25 Oi.l --B -1.:2. Crude Oil 10.0 3.1 0.64 0.24 Total 100.0 100.0 7.72 6.12 Elasticity with respect to GDP at factor cost 1.19 1.05 1/ Excluding special temporary tax. ~/ Excluding redevances sur les transports.  - 48 - 99.. Another element of uncertainty results from the fact that the lV'hc)le tarif"f structure is undergoing a major overhaul due to the recently signed association with the Common Harket. It is veri' unfortunate that, at the time of this lm ting, no information is available on the measures which 'tnll be taken. In particular, it is not clear whether on balance the average ta.r:i.ffs will be increased or decreased and 't'lhether the relative tariffs will be adjusted in such a way that the built-in elasti ci ty of import taxes with respect to the import structure will be significantly IOOdified. The only guideline so far advanced is that an attempt lo1ill be made to reconcile revenue needs with the developmental implications of the tariff structure. Certainly not much for ~ serious discussion. 100. Among the less dynamic components. of the tax struc ture, we have a continuation of past trends in the case of ~nrl.cultural and consumption taxes. As mentioned, value added in agriculture is projected to grow at 3.8 percent as against 5.8 percent for GDP. Assuming that agricultural taxes will be elastic with respect to growth of value added this will constitute a dowmiard bias in the overall elasticity of the tax system with respect to GDP. The Government hopes that as a result of the general cooperativization of the sector there might be a reduction in UL~ evasion with respect to recent years. Should this occur, we could have an increase in tax revenues above their "natural" elasticity for a certain period without, however, modifying the long-term elasticity which is bound to be 10lrer than one with respect to GDP. But even i~ir.oved tax administration may not be forthcoming at the expected pac~ insofar as the immediate impact of the cooperativizatior: program is likely to create a series of problel,1S in connection loti. th the evalua tion of the yields of the net'lly created units. Y 101. As to the elasticity of consumption taxes, actual performance ~ be even less buoyant with respect to GDP at factor cost in view of the fact that the oil and gasolille tax is likely to continue its past trend. Agriculture has been recognized as the leading sector in the developing strategy of the cO'Wltry, and exemptions under the tax for uses in agricul ture will certainly ramain and, i f at all, rl1a.y be increased. 102. Among export taxes, we have not included the special t~orary tax on the assumption that the outlook for it is a further decline if not total. disappearance, since economic considerations are against maintaining such tax lolhich mostly affects phosphates, a COIimlOdi ty now encountering increasing difficulties in world markets at its pres" "prices. At any rate, should the gove~nt rnaintain the tax, its contribution to public savings will be largely offset by the subsidies, which will have to be granted in order to export the commodi41 at competitive prices. 1/ There are signs that the cooperativization process which early this year '\-las. supposed to be completed at maxil1IDrll speed, may now be slowed dow. ibis will certainly del~ the eJ~ected improvement in ta"'C administration over the long run.  - 49 - 103. The built-in elasticity of income taxes (1.07) may look rather 1011 and compares with an expected elas·ticity of 1.12 for tile 1961-68 pericd.~ However, this largely reflects the output of the tax exempt,ions under both the individual progressiva income ta:Y. and the business tax. Using the grOt'lth rate of manufacturing as the base for the business income tax give an elasticity of 1.22. HOl-rever, in viell]' of the fact that tax exemptions are particularlY concentrated on business profits, the higher gro~l rate of manufacturing (7.1 percent as against 6.2 for GDP excluding agriculture) might prove excessive. These considerations are particul.arly valid for the medium-term outlook. Assuming in fact that a large number of new tax exempt ventures has been created in recent years, one could expect that a few' years hence these enterprises tall enter their taxable lif'e and therefor.~: broaden the tax basis. 104. In the case of taxes on incomes one should distinguish between the progressive income tax and the withholding tax on wages and salaries. As to the progressive income ta.."C, some lowering of' the built-in elasticit.Y' could result from the exenptions for the purchase of shares. ....!Ul additional rease.a, related to the present stage in the growth of the econollW, is that the increase in value added is likely to be redistributed over an increased number of people as a result of the employment policy followed by the government. Tr~'") will tend to keep a large number of' people belovl the exemption limit of the tax or widen the lower taxed income brackets with relatively little impact on total revenue.!! 105. As to withheld truces on 1'1ages and salaries one should mention that the recent introduction of progressive rates ~ very well increase the built-in elasticity of the tax when compared with the past. The elasticity will also depend on the wage policy, particularly in the public sector. The:r.'E; is at present strong pressure for lllcreasing the wage rates of government employees. Should this occur the ratio of the withheld taxes to GDP 11la¥ increase. '!/ At the S8m3 time, however, current expenditure of the govemrr.ent 1/ One could also mention the impact of the rate of increase in population on the tax basis through the system of family allowances. The annual rate of' natural increase is nOl-l estimated at about 2.8 percent (the birth rate is 4.5 percent). Even a successful. family planning program is unlikely to substantially reduce the rate of increase of popula't,ion over the immediate future but on the longer run it might affect the built-in elasticity of the tax. This is just one aspect of the more general question of the costs to the goVel'lu.len t of a rigid growth in population given an expansion of value added which is independent of any variation in such growth rates. Y Since value added (in current terms) will also increase the ratio it may remain unchanged depending upon the impact on the tax of the change in value added (whether more or less than proportionate) •  - 50 - ,dll increase so that from the point of view of public savings the increase in rE.~venue will not be a net gain. This is an example of what we may call '!3purious" elasticity, in that an increase in revenue is automatically accom- panied by an increase in current expenditure.!/ 106. The above discussion indicates that the elasticity of the tax structure based on the growth rates of the sectors of origin may, if at all, overest~imate what is likely to happen in view of the upward biases inplied by the coefficients for some major taxes. 11 Another general example is provided by the increase in customs duties which affect government purchases.  - 51 - CHA.Pl'ER IV SOOPE FOR Il[PID VEf-iEHTS AND POSSIBLE DIFFI CULTIES 107 • Having concluded that the Tunisian Ta.-..c System with its present structure is likely to have a built-in elasticity of about 1 with respect to the growth of GDP, the next question is whether this is acceptable in terms of public savings requirements. If the answer is negative, the next question is how to i~rove the built-in elasticity of the tax system and, alternatively, whether the need for additional rev'enue should rather be satisfied through ad hoc increases in the rates of certain taxes. 108. The Fourth Plan projections of public savings (see Appendix, Table 2) are based on a 6 percent growth rate in current expendi ture and on a growth rate of revenue of 8 percent. In particular, revenue towards the end of the period are projected to grow substantially faster than GDP and certa:inly beyond any built-in elasticity of the tax system which can now be expected. This acceleration reflects either increases in the tax rates or a broadening of the tax basis. The Plan document is no t specific on this point, and the Tunisian authorities consider this as a "target" level to be reached in order to bring public savings to the required level for the financing of the investment program. 109. There are serious doubts as to the possibility for the government to keep the expansion of current expenditure within the projected level. Therefore, it is likely that the savings target may not be reached even assuming the expected growth in revenue materializes. In other wordS, there is no question that from the point of vie'ti of public sa'vings the need might occur for revenue to grow at a faster rate than GDP. This incidentally 'WOuld seem to answer in a brutal way the question whether the average tax burden should be allowed to increase substantially, given its present already high level. If the revenue projections of the Plan are realized, the ratio of General Government revenue to GDP in 1972 would reach 30 percent - certainly a very high level. We do not know whether this ratio will depress the economy, but we can certainly say that Tunisia is a country where the public savings problem should be tackled by cutting down the growth of current expenditure rather than by increasing the average tax effort. Having said all this, from a normative point of view, we can now proceed to examine hOli, and if the built-in elasticity of the Tunisian tax system may be increased, and what costs this might involve, 110. The two main possibilities for increasing the built-in elasticit,y of the Tunisian tax system are improvement in tax administration and a greatsr role in the tax structure of the more dynamic sources of revenue. The latter result could be obtained either through a once and for all increase in the rates, or a reduction in the scope of existing exenption.!I !7 The impact on elasticity would be even greater if, at the same time, one could reduce the e~emptions which may keep the growth rate of the less dynamic (i.e., increasing less fast than GDP) taxes below the growth rate of their economic basis. We have in mind the oil and gasoline tax, wi th its exemption for uses in the agricul tural sector.  - 52 - If we take the 1969 tax structure, the choice 't'1ould fallon the income, services, domestic products and ~ort taxes, the la-liter not'tdthstanding the uncertainties mentioned in the previous section. An increase in the rates of these taxes could be acconpanied by a reduction in other taxes if' one did not wish a sudden rise in the average tax ratio; but the result in terms of a higher built-in elasticity 'Would be reached even if the other taxes were left lh'1.changed. 111. The goveImnent is presently reconsidering the whole eXe1iiltion system. under the business taxes and is naturally tom betueen ~tw need to provide incentives and the need to raise new revenues. A possibility now being considered would be to apply the incentives 011 a selective basis, either regional or sectoral. It seems, however, that the merit of this decision should be based on the assess:tilent of the impact of these incel1tive{~; If the incentives perfom their role it l-lould be narrow minded to forego their long-term effects for a S~lort-term illcrease in revenue. The views within the country on this issueara contrasting, even among the officials in charge of tax administration. I'l'o one can say whether certain ventures 'Won't be created in the absence of these exerti>tiol1sj however, the very existence of an investment code, particularly wit,h reference to foreign investors, makes their existence a factor in attracting foreign capital. 112. vie think that these exemptions should be retained and that any modification should aim at increasing their effectiveness, regardless of their impact on the revenue level. At the same time, however, the exelnptions should not act as an incentive for enterprises to procrastinate reaching the stage of self-sufficiency. \V'e have in mind in particular the practice, of extending the grace period under the "Letter of Establishment lt beyond five years. It is true that an extension is in principle regulated by precise norms, but one has the feeling that in reality this solution is used on a wider discretional" basis mld often gives place to pure mld simple salvage operations. This brings us back to the point already made that the task of the government is to avoid the creation of activi ties which are born weak and do not stand a chance of becoming self-sufficient. The current disenchantment with the exemption system. vlhich is prompted by the permanent scarcity of revenue, reflects, as said earlier, inadequate evaluation of investment proposals. 113. FUrthermore, a reduction in the system of exemptions, if not accompanied by a s"und policy in selecting new activities may very well lead to a situation in which the government is forced to give subsidies. ~fuat is gained in taxes is lost in subsidies with very little advantage from the point of 'View of the creation of public savings.!! It is true that Y This ma.v not 'be current subsidies but capital transfers to offset the capital losses suffered by the enterprises as a rusult of their operat- ing deficit. This "lill not affect public savings directJ.y but automati- callY. reduces the amount available for direct investment.  - 53 - open subsidies are better than hidden ones, but it is really a secondar,y questio~" The main point is that one should not set up enterprises which are not Justified even in a longer run perspective. 114. \ve have concentrated so far on fiscal advantages awarded to new enterprises. There are a number of othel· advantages granted to firms, including e.xemption from sales taxes, custom duties, financialparticipatiot1, • etc. The same considerations apply here. It is hard to believe that in a country like Tlll1isia, with a small size market and a still weak industrial sector, the introduction of less favorable provisions would not affect the expectedlevE~l of activity and generate a higher amount of revenue. 115. The most significant exemption under the progressive income tax is the one granted for the purchase of bonds or stocks of an approved list of enterprises. We have mentioned before that this exoneration, up to 3(1 percent of net taxable income, has the effect of reducing the progressivi ty of income tax ani therefore its elasticity ill tenus of personal income or GDP. Once again, its purpose can hardly be questioned from the point of View of developing a capital market and channelling private savings to private or public enterprises.!! 116. Some officials maintain that t.l-J.is is a way of financing private consumption of the higher classes which can rmrlntain their consumption level unchanged insofar as their savings are pro tanto increased by the tax deduction. The implication is that without the decuction the need to main- tain a certain savings level would force them to reduce their consumption. The question is whether, in case they should pay the full ta.."{, and wish to maintain the same level of savings (and reduce their consumption), would the same amount of stocks or bonds be purchased? We do not think so, since in Tunisia, as in many other developing cOtmtries, buying stocks and bonds is not the preferred wa:y of investing savings by individu/,,'~. In other words, one has to provide an incentive for this type of investzrent. The incentive is the one of being able to retain a part of Ol1e's income in this form rather than turning it over to the government. Thi$ means that the only way for channelling private savings to the capital market is to forego a certain amount of public savings. It remains true that private consumption of these categories is favorably treated with respect to people at lower income levels for whom the marginal utility of savings is too high for them to be attracted by what is considered a rather risl~ investment (mostly because non-traditional). Their preference for traditional forms of invest- ing their savings (buying apartments, jewelry, etc.) will leave them with a lower after tax income and a lower average consumption. 117. One can say that this system provides an incentive for people to change their habits and that more and more middle-class people will be pushed ~/ For ••• perspective sake, one should point out that the development of a capital market depends on many important factors in addition to tax exemptions which playa supplementary role, especially in the early stages. .  - 54 - towards investing a higher amount of their incoIll3 in stocks and bonds as a way o£ increasing their savings. A partial solution to this intergroup equity problem is to make it more favorable for people in the lower bracketa to invest in stocks and bonds. For instance, one could very well envisage, a. system in which the share of tax deductible income decreases with the level of income. .ill these considerations assume that the progressivi ty and!! therefore, elasticity of the income tax may be further reduced in the short term, the prize being a. faster expansion of these enterprises which can ., place their issues in larger quantities • 118. As mentioned before, the customs tariff structure is undergoing a major ovel"haul, the in:pact of which on the level of tariffs is still not lmown. In the longer run the determination of the struc ture and level of tariffs should reflect the objectives of industrial policy - short-term concern for revenue need should interfere as little as possible with these objectives. 119. Taxes on donestically manufactured goods are now characterized by the coexistence of the cascade and value added principle. The government is considering the general introduction of the value added tax particular~ for its merits with respect to tax deductions for exported products. Here too it would be very hard to envisage eliminating the present exemptions for exported products as a way to increase the elasticity of the tax. As to the rates, the full introduction of the value added tax should imply an increase in order to reflect the number of passages through which some pro- ducts are now taxed on a cumulative basis. However, in view of the fact that the present rates of the cascade taxes are already rather high, the tax authorities are concerned with the impact that an increase would have on the taArpayers. Admi ttedly, the total tax burden under the value added tax should not increase, but its effect on the various stages of the produc- ing process may be significmlt, particular~ in relation to those placed at the begirming of the process. On the other side, if the value added principle is introduced without an increase in the rates, this would cause a onoe-and-for-all loss of revenue which certainly would be hard to absorb if the current expenditure targets of the Plan are not realized. In addi tion, by reducing the role of these taxes, the elasticity of the entire tax system would also be reduced. 120. The CC)lJcJ:u.sion of this discussion is that the possibility of increasing the bli't..l.t-in elasticity of the t,ax system through either increases in the rates or a less favorable system of exemption are rather limited and in the long run would tend to depress the level of activit,y and ~le absolute level of revenue available to the government. 121. This leaves us with the issue of improvement in tax administration. Tax authorities are now engaged in a very rigorous effort to fight tax e:vssion. An interesting feature of this effort is the creation of tax squads with the purpose of r.eviewing the position of major taxpayers with respect to all important taxes at the same time. 1·1ore specifically, tax administra- tion is likely to be the major factor in the area of agricultural taxes, especially the agricultural income tax. As we have seen, agricultural taxes  - 55 - present a lower than one elasticity with respect to GDP at factor cost. However, also in view of the various subsidies (,exemptions on the gasoline tax, low interest loans, etc.) which the sector receives there is no justification for tax revenue to lag behind the increase in agricultural value added. Agriculture has been recognized as the leading sector in the development strategy of the country. This should call for a system of adequate price incentives to stimulate higher production, accompanied • by a tighter tax policy designed to prevent excessive private cons~tion. Conclusions 122. Three main constraints have reduced the tax elasticity of the Tunisian tax system in the past: weakness in tax administration, exemption policy and reliance of the tax system on sluggis~ inc~easing tax basis, (e.g., imports).l/ These factors have been offset by a series of increases in tax rates l'lhi'Ch, over the period have produced. an acceptable tax buoyancy . The outlook for the future is basically the same with the exception of the first factor, where inprovem.ent seems to be forthcoming. As to tax exemptions the temptation to do away with them, at least in part, should be set against their favorable impact on the economy. At any rate, it liOuld be senseless to eliminate exemptions because of the need to raise additional revenue. This issue must be decided on the basis of the utilization of tax- free money. The evolution of imports is largely dictated by the availabili T:" . of foreign exchange. There again, any change in tariffs should be decided on its economic merit rather than on its revenue yielding aspect. ~e plea of this study is for the Tunisian authorities to accept pasicall~" the present tax structure with its relatively low elasticity with respect to GDP. The true policy variable is current expenditure. The latter should be determined on the basis of the expected growth of revenue which in turn has to be determined as a function of economic conditions in the country. Even a tax elasticity substantially above one may be found in- adequate in the presence of annual growth rates of current expenditure of the order of 10-12 percent. It is absurd to judge the elasticity of the tax system on the basis of current eJ~enditure trends. However, if the authorities fail to control the grow1al of current expenditure and there is the need of raising more revenue, then it will be necessar,y to increase the rates on an ad hoc basis. 'trle believe this to be preferable to any attempt at artificially increasing a naturally low built-in elasticity, which finds its reasons not in weal, ta."'<: administration but in the structure of the economy«the need for exemptions is a structural element). First..., as said so many times by now, the ratio of government revenue to GnP is alre~ high and any addition should be del~ed as much as possible. Seco~•. ad hoc increases in the rates can be obtained on a selective basis and shvdld penalize private consumption rather than incentives to save and invest. 11 And the tax deductible compulsory loans of 1964 and 1968 in the case of income taxes.  • STATISTICAL APPENDIX Table 1: Central Government Current R~venue 11 (million dinars) Provisional Ec. Budget 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 Direct Taxes 11.0 12.5 10.1 11.6 13.2 13.8 20.0 19.9 21.0 26.8 - Income Tax B.8 9.<> 8.9 9.1 10.8 11.9 18.3 18.9 19.5 24.4 - Agricultural Taxes 2.2 2.0 1.2 2.5 2.4 1.9 1.7 1.0 1.5 2.4 Indirect Taxes uu.7 u2.3 u3.7 47.4 55.1 62.9 69.1 69.8 72.2 80.2 - Tax on Goods and Transports 7.B 7.6 8.6 7.9 9.3 8.7 B.l 8.9 12.1 11.9 - Customs Duties IJ.2 7.8 8.7 9.8 11.3 16. 16.4 14.0 12.3 15.0 o:f which: Import Duties (8.2) (5.8) (6.6) (--:.7) (8.8) (9.6) (9.7) (8.6) (6.7) (8.8) Customs Formalities Tax (lJJ) (1.4) (1.4) (1.4) (1. 8) (2.1) (2.2 ) (2.3) (1.9) (2.4) Special Temporar,r Tax on Exports (3.5 ) (3.7) (2.3) (2.7) (2.7) - Registration Tax 3.2 3.2 :;.~ 3. :' 3.4 3.7 3.9 3.7 4.2 4.8 - Turnover Tax lS.6 14.-; 15.3 17.6 21.2 24.1 28.7 29.8 28.6 34.2 o:fwhich: Production Tax en Impu~ts (9.5) (8.6) (9. J) (9.9) (12.5) (13.5 ) (15.9) (15.5 ) (13.3 ) (16.4) Consumption Tax on Imports (1.2) (1.2) (1. 2) (2.2) (2.2 ) (2.3 ) (2.5) (2.6) (2.7) (2.9) Production Tax on Domestic Products (3.3 ) 0.4) 0.4) (3.5) (4.3) (5.l) (6.2) (6.7) (7.1) (8.4) Consumption Tax on Domestic Products (O.3) (0.4) (0.4) (0.7) (0.8) (1.1) (1.3) (1.7) (1.9) (2.2 ) Service Tax (1.7) (1.0) (1.0) (1.1) (1.2) (2.0 ) (2.8) (3.1) (3.6 ) (4.2) - Government Monopolies 7.9 8.8 8.2 9.1 9.9 10.4 12.0 13.4 14.9 14.3 Other Duties and Taxes 1.1 0.9 1.1 1.2 1.5 1.6 1.6 2.6 3.1 3.4 Petroleum Revenues .').2 0.4 1.5 2.2 2.4 5.3 5.1 7.2 10.6 14.0 - Pipeline Rqyalties :::.2 0.4 1.5 2.2 2.4 5.3 5.1 4.9 4.5 5.4 - Oil Revenue 2.3 6.1 8.6 Other Revenues 3.4 3.1 1.7 3.2 3.6 4.7 6.0 6.1 6.5 6.9 Total Budget Title I 60.4 59.2 58.1 65.6 75.8 88.1 101.8 105.6 ll3.3 131.3 Other Treasur,y Accounts 12.7 12.6 13.7 11.9 16.1 19.4 18.1 17.6 12.8 1.4.5 Grand Total 73.1 71.8 71.8 77.5 91.9 107.5 119.1 123.2 126.1 145.8 11 In 1968., an amount o:f about D 5 million :for revenues which up to 1967 accrue to the Special Treas~ Funds, has been integrated in Title I. Source: S.E.P.E.N.  -. ( • Table 2: ~~."'.eral Governm~~.. ;! S!n.P&!. (million dinars) Fourth Plan - 1965 - - 1966 1961 1968 19b9 Provisional Ec. Budget - 1910 - 1911 1912 - Current Revenue 12205 -- 139.3 1h3.0 150,,4 168.8 119.9 194:J 162.6 - 212.1 A. Central Government 103.8 116.3 120.1 125.4 141.5 150.3 118.8 B. Local Authorities 5.8 1.5 5.9 6.3 6.9 B.O 8.8 9.4 c. Social Secun ty 12.1 15.0 16.5 18.3 19.5 20.1 21.8 22.9 D. Non-Budgetized 1I 0.2 0.5 0.5 0.4 0.9 0.9 0.9 0.8 2:UTent ~eniiture 101.1 111.8 121.6 141.2 ~9.1 151.1 166.9 111.1 A. Central Government 18.9 92.6 99.9 ll2.0 119.6 126.0 134.3 1h3.3 B. u>cal Authorities 1.1 8.2 8.4 9.0 10.3 11.3 12.3 13.3 C. Social Security 8.1 9.9 10.3 11.0 11.5 12.2 l2.9 13.5 D. Non-Budgetized 11 1.0 1.1 9.0 9.2 7.7 1.6 7.4 7.0 Current Surplus A. Central Government - 20.8 26.9 21.5 23.7 15.4 20.2 9.2 13 .. 4 19.1 21.9 22.8 24.3 27.2 28.3 35·2 35.6 B. Local Authorities -1.3 - 0.1 - 2.5 - 2.7 - 3.4 - 3,,3 - 3e5 - 3.9 c. Social Security 4.0 5.1 6.2 1.3 8.0 8.5 8.9 9.5 D. Non-Budgetized "J! - 6.8 - 6.6 - B.5 - 8.8 - 6.8 - 6.1 - 6.5 - 6.2 Current Surplus as %of GDP at market prices 4.2 4.2 2.9 1.6 3.1 3.5 3.9 4.1 Current Surplus as % of General Government Capital Expenditure 41.1 31.4 20.6 11.1 22.B 28.3 33.8 39.8 1( Financed through external resources. Sour~G: S~E !'.Eel~.  Table 3: 'fHE. ,FINANCE or _ 1 9 6 S.•19?~ It' (millions ot dinars) Provi- sional Plan Projections 1965 1966 1967 190B 1909 1910 1971 1972 Gross Fixed Capital 132.3 128.9 126.7 117.9 140.2 152.3 159.0 165.5 Formation ChJi'\~e in stocks 6.1 3.1 0.1 -3.2 -4.5 ,3.1 4.2 1.5 Groos Investment 138.4 132.0 126.8 114.1 135.7 155.4 163.2 167.0 D~estic Savings 67.4 79.2 11.8 90.4 105.0 112.8 122.1 133.6 Less: Net Factor Income Payments -l2.2 -14.2 -20.0 -22.1 -23.0 -22.6 -24~1 -25.3 Plus: Net Current Trans- fe~.' He ceipts !I - 0.3 0.0 1.3 0.9 1.4 1.4 1.4 1.4 National SaviESs 55.2 65.0 53.1 69 2 83.4 91.6 99.4 109.7 General Government 20.8 21 •.5 15.4 9.2 19.7 22.B 27.2 35.0 Public Enterprises 21.8 21.6 25.4 31.2 40.8 ( 48.4 49.4 53.4 Identified Private Ent. 5.0 6.0 7.3 6.9 7.1 ( Ba~s 2.4 2.3 3.4 4.0 4.0 4.0 4.0 4.0 Households and Uhiden- ti.fied 5.2 13.6 7.0 11.9 11.8 18.4 18.8 11.3 N9~1r~reign Capital 83.2 61.0 74.1 45.5 52.3 63.8 63.8 51.3 Fub1ic 46.1 31.0 48.4 31.5 46.5 61.0 60.4 65.1 Frivate !I 33.4 21.8 19.2 14.0 U.8 13.8 11.4 1.2 Decline in reserves 3.1 8.2 6.1 -6.1 -6.0 -li.O -B.O -6.0 ~/ Transfers for scholarships, child feeding and philanthropic donations, have been deducted fram nat current transfer receipts and added to private capital inflows. They are estimated at: (millions ot dinars) 2.b 2.6 3.4 4.9 5.2 4.8 4.8 4.8 Source: S.E.P.B.N.  • " • - ~t-- t ................·r~.........-:.........~.,..........·'!...fIIrt,....... ,.·,,·,"· ...•· ... ·~.""··"' .- . ~ '" ....... . t • I ~ T~e 4 ID'DDInJRE oR oross IA.noNAL PJI)DUCT 1960-1972 (lI111.iofUI ot current dinars) ~~ P1m Pro~ect.1om ~ ,!2g ~ 196) ~ 1S65 ~ 1967 . 1968 ~ 1970 1911 1972 Pr1yate CODSWIIptlon 25).) 266.6 282.9 29).3 3:lJ.O 349.1 331.1 )59.. ) )11.1 405.4 419.9 WU.5 471.2 Public CoDSWIIpu'OIl 56.1 59.1 63.0 " 64.5 61.2 79.) 91.2 100.0 llO.7 liS.8 123.8 130.9 139.11 Oross Fix.d CIp! tal roraat,l011 59.5 68.3 76.1 88.6 ]05.1 132.3 126.9 126.7 ll1.9 140,,2 151.) 159.0 165.5 GOTen.mt (29.8)' (27 .. 7) ()4.6) ()1S) (41.6) (41.5) (48.2) (46.1) (U.3) (55.8) (1l1.4) (40.7) (54.9 Public and a1Dd _terprU •• ( 5.3) (15.2) (19.0)· (26.0) (J7.4) (60.4) (55.5) (47.h ) ()5.?) ()9c:9) (67.7) (67.6) (67.) PrlYate aDd coopIrat!ft (20.4) (20.4) US .5) , (15 .1) (17.1) (15.1) (16 .9) (22.8~ (~.2) , (34 ..0) ()2.6) (.l8.2) (10.3 Housebold. ( 4.0) ( 5.0) ( 7.0) (9.• 2) ( 9.0) ( 9.3~ ( 8.3) . (10.2 (10.7) Un .,,S) ( b6) ( 4.5) ( ).0 Chan,. iD Swcb 'at !!Kp(1rta ot Goods u.d SeIT.i.ea3 -ll.l -23.6 8.2 -35.8 -6.4 -JU.8 -5.4 -39.4 0.6 -50:1 6.1 -71.0 3.1 -52.8 0.1 -55.0 -3.2 -24.3 ··4.5 ..·~Js1 ).1 -42.6 4.2 -41.1 1.5 -3).4 bparta 76.8 72.0 .' 69.7 16.4 84.4 98.7 114.) 120.4 126.5 1JJ)",;! 152.1 1.60.9 174.b lJIporta 100.4 107.8 '1ll.5 115.8 1)5.1 169.7 167.1 175.4 150.8 175.9 194.7 202.0 207.8 Oro•• DoIIIe.tic .Product at IlarJlltt Prica. .' )34.2 367.0 373.8 1.01.6 lU5.2 495.8 S07.S 531.1 572.2 626.2 656.5 696.S 744.2 .et hctor be.a P~t;a -1.1 -1.) -).2 -:6.0 -7.6 ' -12.2 -lIa.l -3) eO -:..., -22.1 -2).0 -22.iJ -211.1 -25el t 0I'0Il • •at1aDal. PracllliR )33.1 :365.7 370.6 )9S.6 1a8.0, Ja8).6 !al.) 511.1 .sSO.! 60),2 6))., 672.Ji 711.' i. I I i ! !. .te. Far t.ba period U60-l96I1. tM dUt__ c •• be. . . .... h .. allocated t.o prlYate o.....,u... om- lit ..n.\ pric•••niJk\ed traa : . t.b8 p-al1l:U• •ide IIIIl trG. 1ibI upaadl"a IIlda a..c•• S.I....... r  Ii ~ Table 5 • ") GROSS DOMESTIC PRODUCT BY SECTORAL ORIGIN (millions of current dinars) Provi- Plan Projections sional 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 Agriculture 70.8 79.2 75.5 84.7 86.8 98.0 74.5 70.0 84.0 93.8 93.8 98.6 104.8 Mining 6.3 5.9 5.3 5.9 7.5 11.4 15.7 23.3 29.2 36.1 38.1 39.5 40.2 Extractive industry 6.3 5.9 ~.3 5.9 6.9 10.7 11.5 1;1.8 12.2 13.4 13.7 14.3 15.3 Crud.e oil 0.6 ".7 4.2 12.5 17.0 22.7 24.4 25.2 24.9 Power and 1'later 5.9 5.9 6.3 6.7 6.9 U 8.2 8.8 9.3 11.3 11.8 12.6 13.2 Manufacturing 38.5 45.8 43.') 45.7 53.9 59.1 63.7 67.1 73.1 81.2 88.4 92.9 99.7 Food industry 22.1 29.8 25.5 26.1 27.3 28.5 26.3 25.5 28.8 31. 7 31.6 32.4 33.5 Petroleum refinery 2.6 2.3 2.5 2.6 3.4 3.5 4.0 4.5 4.8 Basic metallurgy 0.9 1.3 1.5 1.7 1.8 1.8 1.8 Mechanical and electrical 1.2 1.0 1.5 1.6 1.8 2.0 2.6 3.0 3.2 3.8 4.3 4.5 5.2 Construction materials 3.4 3.2 3.7 3.8 4.6 4.6 5.2 6.1 6.9 7.8 8.3 8.7 9.1 Chemicals 1.0 1.2 0.8 1.0 1.2 2.4 2.3 2.7 3.2 3.4 3.6 3.7 4.4 Textiles 6.3 5.4 6.2 7.8 9.3 10.8 13.8 15.1 13.7 15.4 19.3 19.4 21.6 Wood and furniture 2.0 2.6 2.3 2.2 3.3 4.3 5.1 5.1 5.5 6.2 6.6 7.1 7.6 Paper, printing, miscellaneous 2.5 2.6 3.0 3.2 3.8 4.2 5.0 5.7 6.9 7.7 8.9 10.8 11. 7 Transport and communications 19.1 20.9 24.2 27.6 29.7 37.0 41.3 43.0 45.0 46.7 47.7 50.6 54.3 Building and public works 21.2 27.8 32.5 34.4 33.4 40.0 41.5 42&0 38.0 42.0 _. 4.5 47.3 51.3 Services 83.5 88.9 93.tj 93.8 103.2 116.3 113d 121.5 126.5 134.4 141.9 151.2 161.9 Housing 12.2 13.0 14.2 15.') 15.7 16.9 17.8 19.0 20.0 21.1 22.4 23.8 25.3 Tourism 0.7 1.0 1.8 3.5 4.8 7.0 9.1 11.1 13.9 15.8 18.1 20.4 Commerce 46.7 49.2 51.9 5:.7 5', .. 66.5 6c.0 65.0 67.0 71.0 74.2 78.3 83.4 Other al 24.6 26.0 25.~ 26.3 27. 1) 28.1 28.4 28.4 28.4 28.4 29.5 31.0 32.8 General government 38.8 44.6 47.6 50.9 52.9 58.8 71.8 77.8 86.2 89.4 94.6 100.9 107.4 GDP at factor cost 284.1 319.1 327.5 349.6 374.3 428.0 429.9 453.5 491.3 534.9 560.8 593.7 632.8 Indirect taxes less subsidies 50.1 47.9 46.3 52.0 60.9 67.9 77.6 77.7 80.9 91.3 95.7 102.8 111.4 GDP at market prices 334.2 367.0 373.8 401.6 435.2 495.9 507.5 531.1 522.2 626.2 656.5 696.5 749.2 ~ L~cludes services of domestics and all other services Source: S.E.P.E.N.