Report No. 6374-TU Fiscal Policy arnd Tax Reform in Turkey (In Two VolIJmes) Volume 1: The Main Report July 7, 1987 Country Operations Division Country Department I Europe, Middle East and North Africa FOR OFFICIAL USE ONLY Document of the World Bank This report has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENT 1981 TL 111.22 1982 TL 162.55 1983 TL 225.46 1984 TL 366.68 1985 TL 521.98 1986 TL 674.50 1987 First Quarter TL 762.96 1987 May TL 821.05 FISCAL YEAR March 1 to February 28 - through 1981 March 1 to December 31, 1982 January 1 to December 31 - from 1983 ABBREVIATIONS CBT - Central Bank of Turkey CPI - Consumer Price Index PIF - Defense Industry Support Fund DSF - Development and Support Fund EBF - Extra Budgetary Fund EECF - European Econ mic Community Fund EEF - Export Encouragement Fund EIF - Export Improvement Fund FF - Financing Fund GNP - Gross National Product at market prices IAF - Investment Acceleration Fund IGMEF - Investment Goods Manufacturing Encouragement Fund IRDF - Interest Rate Differential Rebate Fund ISKI - Istanbul Water Supply and Sewerage General Directorate LAF - Local Administration Fund MASF - Mutual Assistance and Support Fund ("Poor People's Fund") MHF - Mass Housing Fund PCF - Petroleum Consumption Fund PEF - Petroleum Exploration Fund PPF - Public Participation Fund PPSF - Petroleum Price Stablization Fund PTT - Postal, Telephone and Telegraph Authority QR - Quantative Restrictions RUSF - Resource Utilization Support Fund SAF - Special Administration Fund SCF - Selective Credit Fund SEE - State Economic Enterprises SIS - State Institute of Statistics SPO - State Planning Organization SPSF - Support and Price Stabilization Fund TAF - Tax Administration Development Fund TEKEL - State Monopoly TFTU - Undersecretariat of Treasury and Foreign Trade TZDK - Agricultural Supplies Agency VAT - Value-Added Tax WPI - Wholesale Price Index FoR OmCAL USE ONLY This report is based on the findings of a World Bank Mission that visited Turkey in March 1986. The mission consisted of the following: Ritu Anand (Mission Chief) Faezeh Foroutan Charles E. McLure (Consultant) Anthony ?.Llechio Partbiisarathi Shome (IMF) Sweder van Wijnbergen The analysis is based on data and tax structure until January 1987. Major changes that occurred between January and April 1987 are mentioned in footnotes. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. FISCAL POLICY AND TAX REFORM IN TITRKEY: Table of Contents VOLUME I: THE MAIN REPORT Page No. EXECUTIVE SUMMARY (i) - (vii) Chapter I: INTRODUCTION 1 Chapter II: INFLATION, REAL INTEREST RATES AND THE FINANCING OF GOVERNMENT EXPENDITURE IN TURKEY A. Real Interest Rates, External Balance and Fiscal Policy 3 A.1 Policy Conflicts in an Open Economy: The Analytical Framework 3 A.2 Macroeconomic Developments in Turkey in 1980-1986 5 B. Inflation and the Financing of Government Fxp?nditure 9 B.1 Trends in Public Sector Resource Mobilization 9 Public Sector Expenditure and Revenues 9 Central Government Budget 12 Extra-budgetary Funds 14 Local Governments 14 State Econoraic Enterprises 17 B.2 Consistency of Fiscal Policy and Inflation Targets 19 B.3 Longer-term Concerns 24 C. Fiscal Deficits and Taxation: An Introduction to the Remainder of this Report 25 Chapter III: PRIVATE SAVINGS INCENTIVES AND THE PERSONAL INCOME TAX SYSTEM A. The Current Syatem 32 B. Past Reforms: An Evaluation 34 B.1 Adjustment of Personal Tax Schedules 35 B.2 Personal Exemptions 38 B.3 The Expenditure Rebate System 40 B.4 Measures to Improve Tax Collection 43 B.5 Tax Administration, Procedures and Penalties 45 C. Recommendations 46 C.1 The Expenditure Rebate System 46 C.2 Exemption Level and Rate Structure Reform 48 C.3 Inflation and the Personal Income Tax 50 Structure -ii- Page No. Chapter IV: THE CORPORATION TAX AND INVESTMENT INCENTIVES A. Description 52 A.l A Brief Overview 52 A.2 General Provisions 55 Investment Incentives 56 A.3 Tax Administration, Appeal Procedures and Penalties 58 B. Assessment 58 B.1 Depreciation Allowances & Investment Incentives 59 B.2 Marginal Effective Tax Rates 62 Equity Investment 62 Debt-Financed Investment 66 B.3 Interpretation 66 C. Suggestions for Reform: Corporate Taxes 69 C.l Incentive Rationalization & Inflation: Two Options for Corporate Tax Reform 69 Option 1: Expensing 70 OpLion 2: Economic Depreciation with Inflation Adjustment 71 Treatment of Inventories Under 2 Options 72 Adjustment of Tax Credit Carry-overs 73 Recommended Option 73 C.2 Implementation Issues 74 Appropriate Basis for Inflation Adjustment 74 The Choice of a Price Index 76 Indexing v. Forward-looking Adjustment 77 Transition Problems and the Potential of an Excessive Tax Burden 77 D. Suggestions for Reform: Investment Incentives 78 E. Other Corporate Tax Provisions 81 E.1 Corporate-Personal Tax Interaction 81 E.2 Treatment of Large Business 82 E.3 Taxation of Foreign Investment 83 E.4 Advance Payment of Tax 83 F. Summary of Recommendations 84 Chapter V: THE VALUE-ADDED TAX AND OTHER INDIRECT TAXES A. The Value-Added Tax 87 A.1 Structural Aspects of the Value-added Tax 87 Design of the Tax: Development and Characteristics 87 Analysis of Structural Aspects 88 Treatment of Agriculture and the Zero Rate 88 Multiple Rates 89 Exemptions 90 Small Taxpayers 90 Compensatory System 91 Three-Monthly Payment 91 Treatment of Capital Goods 92 -iii- Page No. A.2 Revenue Aspects of the Value-Added Tax 93 A.3 Administrative Aspects of the Value-Added Tax 97 Structure of Administration 97 Recent Administrative Measures and Effects 93 Measures to Improve VAT Assessment 100 B. Other Indirect Taxes 102 B.1 Supplementary VAT 102 l.2 Stamp Taxes 102 B.3 Fees 103 B.4 Motor Vehicles Purchase Tax 104 B.5 Motor Vehicles Tax 104 B.6 Petroleum Consumption Tax 105 B.7 Inheritance and Gift Tax 105 B.8 Banking and Insurance Tax 105 C. Summary and Recommoendations 106 Chapter VI: TAXATION OF FOREIGN TRADE A. Import Liberalizaton 107 B. Export Incentives 109 C. Evaluation 111 C-1 General Issues 111 C.2 Export Incentives 111 C.3 Reform of the Trade Intervention Structure 113 Chapter VII: EXTRA-BUDGETAPY FUNDS A. Introduction 117 B. Management of Extra-Budgetary Funds 119 C. Overview of Principal Extra-Budgetary Funds 120 D. Evaluation of the Funds 124 General Issues 124 Individual Funds 125 Export Encouragement Fund (EEF) 125 Resource Utilization Support Fund (RUSF) 125 Support and Price Stabilization Fund (SPSF) 126 E. Reform of Extra-Budgetary Funds 126 Chapter VIII: SUMMARY AND RECOMMENDATIONS A. Inflation and the Financing of Government Expenditures 129 B. Structure of Taxation 130 B.1 Minimizing the Distortionary Costs of Raising Revenue 130 B.2 Influencing the Structure of Economic Activity 132 B.3 Jncome Redistribution 133 -iv- Page No. C. Specific Recommendations 133 C.1 Personal Income Taxation 133 Expenditure Rebates 133 Rate Structure and Personal Exemption Level 135 Inflation and Bracket Creep 137 C.2 Taxation of Corporate Income 137 C.3 Investment Incentives 139 C.4 Value Added Tax 140 C.5 Trade Intervention 140 C.6 Reform of Extra Budgetary Funds 141 D. Total Revenue Estimates and Conclusions 143 t o Txt ablesadFiures VOLUME I: THE MAIN REPORT Table/Figure No. Page No. Chapter II: INFLATION, REAL INTEREST RATES AND THE 7INANCING OF GOVERNMENT !XPENDITURE IN TURKEY Table 2.1 The Size of the Public Sector 10 2.2 Public Sector Surplus 11 2.3 Share of Local Authorities in Total Tax Revenues Collected by the Central Government 15 2.4 Revenues and Expenditure of Local Authoritios 18 2.5 Required Deficit Reduction for Consistency with Various Macroeconomic Targets 21 2.6 Fiscal Consequences of Internal Debt Finance with no Fiscal Adjustment at a 15% Inflation Target and 18% Real Interest Rate on Internal Debt 23 2.7 Tax Effort by Type of Tax for Selected Developing Countries 26 2.8 Tax Revenue by Type of Tax for Selected Developing Countries 27 2.9 Total Tax Revenue as a Share in GNP 28 2.10 Central Government Budget Tax Revenues, 1985-86 30 Fig. 2.1 Private Savings, Fiscal Deficits and the Real Rate of Interest 4 2.2 Share of Central Government Revenues and Expenditures in GNP 13 2.3 Share of Central Government Expenditures in GNP 13 2.4 Central Government Funding of Local Governments: Revenue-Sharing Arrangements 16 2.5 Share of Central Government Tax Revenues in GNP 31 Chapter III: PRIVATE SAVINGS INCENTIVES AND THE PERSONAL INCOME TAX SYSTEM Table 3.1 Withholding Taxes and Re:es 33 3.2 Personal Income Tax Schedules, 1981-86 35 3.3 Estittated Revenue Effects due to Changes in the Personal Income Tax Schedule, 1981-85 36 3.4 Marginal and Average Tax Rates at Selected Income Levels in 1981-86 38 * 3.5 Comparison of the Ratio of Personal Deductions to Estimated Family Income for Selected Countries in 1984 39 -vi- Page Ns. Chapter III: PRIVATE SAVINGS INCEWTIVES AND THE PERSONAL INCOME TAX SYSTEM (Contd.) Table 3.6 Expenditure Rebate Schedules, 1984-86 41 3.7 Expenditure Rebate as a Percentage of Income at Selected Income Levels in 1984-86 42 3.8 Combined Effect of the Personal Income Tax and Expenditure Rebate at Selected Income Levels in 1984-86 43 3.9 Distributional Impact and Cost Savings of thse Suggested Reform of the Expenditure Rebate System 48 3.10 High Exemption--Simplified Rate Structure: Revenue Neutral Options 49 Fig. 3.1 Inflation and Income Tax Revenues: Fiscal Drag 51 Chapter IV: CORPORATION TAX AND INVESTMENT INCENTIVES Table 4.1 Corporate Income Tax as a Percent of Total Central Government Revenue, for Selected Countries in Recent Years 53 4.2 Investment Incentive Allowances (First Year Write-off) by Sector and Region 57 4.3 Present Value of Depreciation Allowances 60 4.4 Present Discounted Value of Depreciation Allowances and Investment Incentives by Sector, Region and Asset Type - Inflation Rate - 0 Percent 61 4.5 Marginal Effective Tax Rates (In Percent) for Equity Financed Projects Inflation Rate - 25 Percent 63 4.6 Marginal Effective Tax Rates (In Percent) by Sector, Region and Asset Type for Equity Financed Projects Inflation Rate - 0 Percent and 50 Percent 65 4.7 Marginal Effective Tax Rates (In Percent) Effect of Full Inflation-Adjustment of Depreciation Allowances 65 4.8 Marginal Effective Tax Rates (In Percent) by Sector, Region and Asset Type for 50 Percent Debt Financing and Alternative Rates of Inflation 66 4.9 Subsidies and Taxes on Capital Goods 68 -vii- Page No. Chapter V: VALUE-ADDED TAX AND OTHER INDIRECT TAXES Table 5.1 Use of Funds by Government as a Result of Treatment of Capital in VAT 93 5.2 Value-Added Tax Rates and Revenue Importance in Selected Countries 94 5.3 Revenues from Indirect Taxes 96 Chapter VI: TAXATION OF FOREIGN TRADE Table 6.1 Imports and Trade Taxes as a Share of GNP 109 6.2 Revenue Effects of Trade Reform 114 Chapter VII: EXTRA-BUDGETARY FUNDS Table 7.1 Principal Extra-Budgetary Funds, 1985-86 118 7.2 Extra-Budgetary Funds Revenues by Source, 1985 121 7.3 Sources and Uses of the Principal Extra-Budgetary Funds in 1985 122 Chapter VIII: SUMMARY AND RECOMMENDATIONS Table 8.1 Distributional Impact & Cost Savings of the Suggested Reform of the Expenditure Rebate System 135 8.2 High Exemption-Simplified Rate Structure: Revenue Neutral Options 136 8.3 Revenue Effects of the Proposed Reforms 144 EXECUTIVE SUMMARY A. INTRODUCTION 1. Since 1980, Turkey nas gone through a series of structural reforms at an unprecedented pace. Trade strategy has shifted from interventionist import substitution to a much more market-oriented outward orientation. A substantial effort is under way to rationalize the tax structure, with as a major reform the introduction of VAT in January 1985. On the external side, the crisis of the late seventies has been weathered; the current account deficit has been reduced and creditworthiness restored. Where developments seem less successful is in securing a corresponding adjustment on the internal side. Growth in real GNP in the last six years has been satisfactory by any standard, but high real interest rates and only mixed success in reducing the rate of inflation have contributed to unsatisfactory levels of private investment. There is, consequently, a justified concern about the sustainability of the current growth rate. High real interest rates in the face of an external balance constraint indicate that the public sector deficit is too large; public sector net savings are apparently not high enough for private sector net savings to make up the remaining gar with the external target at reasonable (by world standards) real rates of interest. As a consequence, private investment is cruwded out by the public sector. 2. The report therefore deals with public sector r.source mobilization, its consequences for fiscal balance, and its effects on private sector 4ncentives to save and invest. The report first discusses the relation between fiscal deficits, inflation and real interest rates in a framework of macro-economic developments over the last tew years. It then focuses on government revenue mobilization and offers a comprehensive analysis of the Turkish tax system. The report describes the current tax structure and analyzes ongoing reforms, their likely effect on public -:evenues and on private incentives to save and invest. Subsidies and transfers are covered, but no attempt is made to discuss the appropriateness of the structure and level of current and capital expenditure on goods and services by the public sector. Public sector investment has been covered in the World Bank Report entitled "Turkey: Adjusting Public Investment". Government consumption is outside the scope of the present report. B. ACHIEVEMENTS, ISSUES AND RECOMMENDATIONS 3. In the first part of the Report the consistency between inflation targets ; i current fiscal deficits is analyzed. The analysis suggests that, if 198F - taken as a benchmark, a substantial reduction in the fiscal deficit is necessary: an inflation target of 15 percent is shown to require, for consistency, a cut in the total public sector deficit to GNP ratio by two percentage points. This adjustment could be postponed, and the resulting financing gap covered t.hrough the issue of interest bearing domestic debt. However, the report shows that, if such a policy is followed at the current high level of real interest rates, the fiscal deficit would deteriorate rapidly because of rising interest expenditure. Thus, restrictive monetary policy at the cost of increased issue of government bonds might buy low inflation now at the cost of higher inflation in the future. Also, real depreciation of the exchange rate would raise the required deficit reduction because of the increased cost of servicing external dett. -(ii)- 4. Such a reduction in the deficit can be brought about by either cuts in expenditure or by increasing tax revenues. While this report does not analyze government expenditure (other than subsidies and transfers), some points can be made about the possibility of expenditure reduction. First, the overall deficit run by the State Enterprises is the largest component of the total public sector deficit. While there have been extensive reforms in public sector pricing, their financing requirements are still large. Second, while the share of Central Government expenditure in GNP fell between 1981 and 1986, overall government expenditure has actually increased, if local governments and extra-budgetary funds are included. Third, budgetary expenditures of Central Government, while down in 1985 from their 1981 share of GNP, have begun to rise again in 1986. Expenditure reduction should, therefore, receive serious attention. 5. The declining trend in the ratio of budget revenues to GNP was reversert with the introduction of VAT in 1985. However, increased revenues will still be necessary if the required deficit reduction cannot be achieved by expenditure reduction alone. The report therefore outlines a series of reforms that will go some way towards achieving this goal. However, these suggestions for reforms are not only based on the contribution they make towards increasing government revenues. Since all sources of government revenue influence private incentives, concern with the efficiency of the tax structure, for any given level of revenues, is of paramount importance. In addition, some of the reforms are of a long-term structural nature and would require careful transition measures. Others are less far-reaching and could be implemented at shorter notice. In what follows, the major tax instruments are covered in turn. This summary then ends by proviring an estimate of the effects of the reforms suggested on total government revenues: a revenue gain of 1.1-1.3 percent of GNP. Value-Added Tax 6. Tax reform in Turkey has increased the efficiency of raising revenues. Most important of all the recent changes is the replacement of producer taxes by a VAT system. Producer taxes cause inefficient use of domestic resources, while consumption taxes at least allow efficient use of the country's factors of production. However, the shift to VAT has not completely resulted in a consumption tax, because of the way non-encouraged investment is treated under VAT. VAT paid on goods bought for non-encouraged investment purposes is nct rebated immediately; instead the rebate is spread equally over three years. At current interest rates, and at the VAT rate of 12 percent, this procedure amounts to about a 3.6 percent tax on capital goods. This reduces the efficiency gains from the introduction of VAT and distorts investment decisions. 7. Thus the report suggests the introduction of immediate rebates on VAT paid on all goods for investment purposes, analogous to the treatment of exports under the VAT. This would have only moderate negative revenue effects, since encouraged investment projects already receive immediate rebates. 8. The report proposes reforms in the administration of the VAT. Making provincial head offices directly answerable to the Director General of Public Revenues may avoid some confusion in the line of authority. They now work in parallel on the national level. A second administrative reform that might improve collection efforts is to give local tax offices responsibility for inspection, rather than vesting it in the provincial offices. Under the -(iii)- current system, local tax offices have no incentive to raise revenue, in spite of their better access to information. In addition, low salaries of tax auditors hamper recruitment of qualified personnel. Corporate Tax 9. The corporate tax system has seen a number of improvements recently. Since the changes introduced in January 1986, the corporate tax system discriminates less between different sources of financing investment than it does in many other countries. Dividends are now exempt from personal income taxes, as are, effectively, capital gains. This implies that the new tax system does not distort the choice between retained earnings and new share issues as a source of finance. Most countries discriminate heavily against new share issues by taxing dividend income at a much higher effective rate than capital gains. However, the Turkish corporate tax system has not completely achieved neutrality; there still is preferential treatment of debt-fiilancing, as is the case in most other countries. 10. An important issue is the effect of inflation on the corporate tax system. Inflation causes income to be mismeasured, unless inflation adjustment is provided for in depreciation allowances, inventory evaluation, interest expense, and interest income. Through its impact on the tax burden, inflation makes the corporate tax a more arbitrary and unpredictable source of revenue. A positive feature of the Turkish system is the attempt to correct depreciation allowances for inflation. However, inflation adjustment for the asset side only (namely depreciation), and not the debt side, may actually accentuate distortions; it also causes corporate tax revenues to fall whenever inflation rises. Accounting for inflation-induced capital gains and losses in the calculation of taxable profits is therefore important. In addition to complete adjustment for inflation of depreciation allowances, this requires: (a) allowing deduction of only real interest payments from taxable income; this would also have positive revenue effects; (b) switching to the Last-In/First-Out accounting system for the evaluation of inventories in the calculation of taxable profits; (c) adjusting carry-overs of tax credits by the nominal interest rate to compensate for inflationary erosion and time delays. 11. Implementation of the change from nominal to real deductibility needs to be undertaken with care to avoid excessive administrative costs, unwarranted windfall gains to the financial sector and an excessive rise in the tax burden. Hence the proposal made in this report is to calculate taxable profits as is currently done; deduct all nominal interest payments; then add back to taxable profits inflation times the average value of NET monetary liabilities (debt minus monetary assets). The average could be calculated by simply taking the average of beginning and end of period values. This method, as all approximations, leads to some inaccuracies, but it is relatively robust to abuse and imposes no additional administrative burden on the private sector: it requires information that firms are already under obligation to collect. These implementation problems, their effect on the financial sector, and further issues such as the choice of which inflation rate to use are discussed in more detail in Chapter IV, Section C.2. Finally, this reform should be implemented gradually, or perhaps in a period of lower inflation, to avoid a sudden increase in the tax burden. If the system, once implemented, raises the tax burden too much, a lower corporate tax rate could be considered in the future. -( iv)- 12. The current investment incentive system, operating through extra first-year depreciation allowances, is effective, as long as the allowances are taken in the year investment takes place. This guarantees neutrality with respect to asset type and inflation, and so reduces the distortionary costs of the incentive system. However, it is not clear why the investment incentives given through the extra-budgetary funds are warranted. The subsidies of the Export Encouragement Fund are highly distortionary, discriminating against imported capital goods and against buildings and structures in favor of domestic machinery. As to the investment subsid4es through the Resource Utilization Support Fund and the tax deferral through the Finance Fund, their objectives can be achieved equally well througa the system of extra first-year depreciation allowances. A case can thus be made to reduce the role of these three funds. 13. A possible concern is that givirxg regional and sectoral incentives through investment tax credits introduces a bias towards capital intensity in industrial policy. Investment incentives slant the choice of technique towards more capital intensive methods, and, for a given technology, favor capital intensive sectors over labor intensive ones. The Government should consider whether the current system of additional personal exemptions for workers in favored regions or industries is sufficient to offset this bias. To the extent that investment incentives continue to be used, they could be simplified by combining them with supplementary first-year allowances, or with subsidies from the budget. 14. Furthermore, reduction (and preferably elimination) of the current discrimination against imported capital goods is advisable. Imported capital goods face an effective tax of 43 percent for non-encouraged projects and of 5 percent for encouraged projects. This works at cross purposes with the investment incentive system. Hence the recommendation to phase out tariffs on imported capital goods. The associated revenue loss could be covered by reduction in the subsidy to domestic capital goods. Differential treatment of domestic capital goods is difficult to justify; differential treatment under VAT, one of the current arguments, will become irrelevant if the recommendations on the treatment of capital goods under VAT (see para 7) are followed. 15. A final recommendation is to eliminate the current bias of the investment incentive system towards larger projects. The current minimum size requirement for certificates of encouragement biases towards large scale projects and against smaller scale partial renovation. If the investment incentive system is indeed consolidated into the system of first year investment tax allowance, eliminating the minimum size requirement should not lead to increased administrative costs. Personal Income Taxes 16. Personal income taxes, which form a significant share of tax revenues in Turkey, are generally well-designed. The personal income tax system is comprehensive and simple: it is not cluttered with numerous exemptions and deductions. A good feature is the absence of mortgage interest payment deductibility since such deductions would distort the allocation of savings. - (v) - 17. The most important problem is with the expenditure rebate system, through which taxpayers receive a rebate on their income tax proportional to their expenditure on certain categories of goods and services. The report recognizes that there are considerable benefits to be derived from this system, particularly in improving VAT compliance. But there can be little argument that the expenditure rebate system is too costly an enforcement mechanism to be acceptable as a permanent feature of the tax system. The system currently costs about 1.5 percent of GNP (no less than 40 percent of net VAT revenues in 1986) and is thus becoming a major drain on government revenues. In addition, it constitutes a disincentive to savings and thus is a factor in the deterioration of the current account, at a time when external balance is increasingly of concern. The report therefore suggests a gradual phasing out of the expenditure rebate system as a long term reform, to be implemented once the VAT is more solidly established. Measures to improve VAT complianCe, such as the computerization of tax records and use of tamper proof electronic cash registers, are being introduced. However, until such methods stucceed in establishing a high degree of VAT compliance, measures should at least be taken to improve the rebate system's efficiency, reduce its costs, and eliminate its anti-savings bias. 18. First of all, costs could be reduced without jeopardizing VAT compliance by excluding from the list of eligible commodities items that are exempt from VAT, such as rent. Also, many consumer durables are sold mostly in the organized sector, where compliance is less of a problem. Consumer durables could therefore also be excluded from the eligible list without reducing the system's effect on VAT compliance. Furthermore, the highest two brackets have rates below the VAT rate: 10 and 5 pe:.cent respectively, while the VAT rate now stands at 12 percent. Clearly, with rates below the VAT rate, consumers and shopkeepers can be better off by splitting the difference and bypassing VAT. These brackets can therefore safely be abolished without deleterious effects on VAT compliance. The reduction in rebates from all these accounts would be substantial across all income classes. On average a cost reduction of between 20 and 25 percent can be expected, which, with 1986 as a benchmark, equals between 0.3 and 0.4 percent of GNP. Moreover, the highest reduction would take place in thle upper income brackets, thus softening the negative distributional impact of cutbacks in such rebates. 19. Finally, the system's anti-savings bias could be reduced by including savings components such as share purchases in the list of eligible items. This would also encourage capital market developmint, a long standing goal of the government. Possibly, abuse could take place through sales of shares between individuals back and forth, with only the purchases reported. To avoid such schemes, eligibility should be restricted to registered shares only, where sales of shares are recorded. Before any such modification is implemented, a careful assessment of its administrative costs needs to be made. 20. Even with the expenditure rebate system in place, problems of evasion and reduced progressivity remain within the income tax system. The latter is an unintended byproduct of the erosion of the personal exemption through -(vi)- inflation. The personal exemption level was equal to the minimum wage in 1981. It now equals less than 25 percent of the current minimum wage; in real terms it has fallen by mote than 67 percent. This is one reason why marginal and average tax rates have fallen for all but the lowest income groups. While the wage rebate system alleviates this problem, there are more effective and less distortionary ways of achieving an improveme,.t on this front, which the Government may want to consider. 21. One such way to meet distributional objectives would be to restore the personal exemption level to its (real) level of 1981; this would.i require a threefold increase. But this measure would, by the very mechanism through which it would have its favorable eq;ity impact (exempting the poor disproportionally), also have a negative revenue impact. Hence the report also suggests an adjustment in the rate structure, with tax rates chosen such that the whole reform would have no negative impact on revenues. Also, since equity objectives would be met more closely through increased exemption levels, a simplified tax rate structure could be considered. Moreover, US experience has shown that a simplified rate structure also reduces evasion. Trade Taxes 22. Substantial improvements have been made in the taxation of international trade. Replacing the majority of quantitative restrictions with tariffs increased revenue with probably a decline in distortionary costs. However, this is also the area where slippage is occurring now. Although custom duties have become less important over time, trade levies to finance extra-budgetary funds have increase~d. The report shows that in 1985 trade taxes were higher by about 80 percent because of the extra-budgetary funds, while export subsidies have also taken on more importance. 23. The report puts forward two proposals in trade policy. First, it shows that reducing export taxes on agricultural commodities would lead to significant real income gains at revenue losses for the government well below the current revenue from those export taxes. Moreover, this revenue loss could easily be covered by the scheduled reduction in fertilizer subsidies, the cost of which exceeded TL 200 billion in 1985. 24. The second proposal is to reduce the widespread occurrence of import tariffs and export subsidies, which have opposits effects on resource allocation. Moreover, export subsidies and impo.t tariffs are ineffective instruments to influence the balance of trade. Such measures, especially when expected to be of longer duration, tend to affect both exports and imports, and thus leave the trade balance basically unchanged. For example, import protection diverts domestic production away from export markets towards home production, thus reducing exports as well as imports. Exchange rate adjustment and fiscal restraint are better instruments to correct external deficits. A useful rationalization would therefore be a simultaneous reduction in export subsidies and import tariffs. However, the phasing and macroeconomic impact of such reforms should be carefully considered prior to implementation. A large, across-the-board reduction in export subsidies and import tariffs would need to be accompanied by an exchange rate adjustment to -(vii)- speed up adjustment of the relative price of traded and non-traded goods. Also, compensating sources of revenue need to be found if adverse effects on fiscal and current account deficits are to be avoided. The measures suggested in this report will in fact generate irore than enough revenue to offset this loss. Finally, reducing export subsidies without lowering tariffs would introduce an anti-expori.- bias; such owe-sided reform snould therefore be avoided. Extra-Budgetary Funds 25. The proliferation of extra-budgetary funds (EBFs) over the past two years is a disturbing development. Their use of earmarked tax revenues makes it difficult to maintain overall efficiency of the tax structure; individual measures are taken without sufficient concern whether a specific tax instrument is the least distortionary source of revenue. The increasing reliance of the Funds on trade taxes clashes with the export strategy of the government. Moreover, the institutionalization of these Funds makes it more difficult over timei to fold them back into the budget system when the particular needs they were designed to meet have been satisfied. They also create problems of overall control and coordination. Often one measure is set up to offset the effects of another, or measures are working at cross-purposes. Some of the recommendations suggest more efficient ways of achieving the objectives of particular EBFs. 2t. However, rationalization of the EBFs and return to the budget of what should remain of their activities is probably only feasible over a longer time period. While such longer term reforms are underway, shorter term rationalization within the existing structure could provide immediate benefits. A careful review of the purpose and functioning of the EBFs is clearly necessary, but some inconsistencies are conspicuous. The report makes a number of specific proposals starting with the abolition of the financial transaction tax, possibly with a revenue equivalent reduction in investment incentives. Investment incentives could also be reduced if the surtax on corporate income by the Defence Industries Support Fund and now also the Mutual Assistance and Support Fund were to be replaced by recourse to general revenues. The planned elimination of fertilizer subsidies, while also reducing agricultural export taxes for the sectors receiving fertilizer subsidies, is another example of such a reform. C. FISCAL IMPACT OF THE REFORMS SUGGESTED 27. Tne tax and investment incentive reforms suggested in the report were not only designed to improve the efficiency of the tax system, but also to increase revenue. Estimates of the revenue effect of individual measures are provided in the report, and add up to a total revenue gain of between 1.1 and 1.3 percent of GNP. This covers the deficit reduction the report estimates to be necessary for consistency with a sustainable debt strategy and an inflation target of 25 percent. For a more stringent inflation target of 15 percent, the revenue effects of this reform package would cover only two-thirds of the required deficit reduction. Therefore, cuts in expenditure wuuld be necessary, if at least increases in tax rates are to be avoided. CHAPTER I - INTRODUCTION 1.1 Since 1980 Turkey has gone through a series of structural reforms at an unprecedented pace. Trade strategy has shifted from interventionist import substitution to a much more market-oriented outward orientation. A substantial effort is under way to rationalize the tax structure, with as a major reform the introduction of VAT as of January 1985. The interest and exchange rate policies of the seventies that bought high growth at the cost of unsustainable external debt accumulation and, in the end, accelerating inflation, have given way to more realistic policies. 1.2 It is important to realize that Turkey's attempt to transform itself into an efficient market economy is taking place in a more hostile external environment than countries like Korea faced during similar periods of transition in the late sixties. Protectionism in the developed countries is high add rising. External debt is more expensive and more risky, due to higher real interest rates and more volatile exchange rates. Nevertheless external adjustment in Turkey has clearly been a success since the crisis days of 1977-1980. While debt-output ratios have in fact risen over the period, creditworthiness has improved, primarily due to much better export performance. However, caution is still called for, mostly because of the recent slow-down in foreign exchange receipts. The Turkish authorities' decision to.gear macroeconomic policy towards achieving sustainable external deficits seems prudent. 1.3 Where developments seem less successful is in internal adjustment. Any external target on the current account needs to be maLched by a corresponding balance between private and public savings on the one hand and aggregate investment on the other. If this balance is struck at high levels of savings and investment, satisfactory and sustainable growth is po,sible within the constraints posed by external targets. However, if the balance is achieved at low levels of savings and investment, capacity constraints will eventually bring down output growth. 1.4 Turkey's rate of growth in GNP in the last six years has been satisfactory by any standard, but high real interest rates and only mixed success in reducing inflation have prevented satisfactory levels of private investment. There is, therefore, a justified concern about the sustainability of the current satisfactory growth rate. High real interest rates in the face of an external balance constraint are prima facie evidence of inappropriate fiscal policy; public sector net saving is apparently not high enough for private sector net savings to make up the remaining gap with the external target at reasonable (by world standards) real rates of interest. As a consequence, private investment is crowded out by the public sector. Moreover, fiscal imbalances also have required reliance on the inflation tax 1/ to such an extent that inflation rates have remained high over most of the period. As a consequence, the high real rates of interest have been at high levels of inflation and nominal rates of interest rather than at low ones. 1/ This concept is explained in para 2.49. -2- 1.5 Hence the focus of this eport. It deals with public sector resource mobilization, its consequences for fiscal balance, and its effects on private sector incentives to save and invest. The report first discusses the relation between fiscal deficits, inflation and real interest rates in a discussion of macro-economic developments over the last few years. It then focuses on government revenue mobilization and offers a comprehensive analysis of the Turkish tax system. The report describes the current tax structure and analyzes ongoing reforms, their likely effect on public revenues and on private incentives to save and invest. Subsidies and transfers are covered; no attempt is made however to discuss the appropriateness of the structure and level of current and capital expenditure on goods and services by the public sector. Public sector investment has been covered in the World Bank Report entitled "Turkey: Adjusting Public Investment". Government consumption is outside the scope of the present report. 1.6 The organization of the report is as follows. Chapter II discusses inflation, fiscal deficits and real interest rates. The remaining chapters analyze the tax system. Chapter III analyzes income taxation, while Chapter IV discusses corporate taxes and investment incentives. Chapter V covers indirect taxes, in particular the recently introduced VAT. Chapter VI discusses trade policy, export incentives and tariff structure. Chapter VII discusses extra-budgetary funds. Finally, Chapter VIII presents the recommendations of this study. -3- CHAPTER II - INFLATION, BDAL INTEREST RATES AND THE FINANCING OF GOVERNMENT EXPENDITURE IN TURKEY 2.1 Macroeconomics in Turkey is dominated by the twin problems of inflation and high real rates of interest. Both can be seen as consequences of fiscal imbalances. The relation between high real rates of interest and fiscal policy needs to be seen in the context of the external balance strategy Turkish authorities have been following. Explicit external balance targets and imperfect capital mobility cut the link between foreign and domestic interest rates; high real rates (compared with foreign real interest rates) can then be interpreted as the consequence of insufficient public net savings 1/ contribution to the external balance target. What is not met by public sector savings will have to be contributed by private net savings; to call forth sufficient private net savings, however, high real rates are needed. 2/ 2.2 In the short run, inflation is, beyond doubt, mostly influenced by such cost factors as the exchange rate, oil prices and agricultural shocks. However, supply factors like that explain price level shiftF, but not really why inflation, once shifted up, remains high after the initial shock. To understand the persistence of inflation, an analysis of inflation, taxes and the government budget is necessary. A. Real Interest Rates, External Balance and Fiscal Policy A.1 Policy Conflicts in an Open Economy: The Analytical Framework 2.3 The external debt problems of the late 1970s and the ensuing rationing of foreign exchange inflicted a very high cost on the Turkish economy. To avoid a recurrence, external balance has been the overriding goal of the Turkish authorities, and of the Central Bank in particular. Exchange rate and interest rate policies are used aggressively to avoid current account deficits judged unsustainable. If such a policy is coupled either with explicit capital controls or a lack of substitutability between Turkish and foreign assets, or both, the link between domestic and foreign interest rates is severed. 2.4 Real interest rates then are determined out of the interplay between fiscal policy and private savings and investment decisions, given the external targets under which the economy operates. The mechanism is straightforward and best understood starting from the current account identity linking public sector deficits, FDEF, private net savings, NPS(r) = Sp - Ip, and the current account, CA: CA = NPS(r) - FDEF (1) 1/ Net savings refers to the surplus of savings over investment. 2/ The interest-response of private savings is not firmly established; however, the interest-response of private investment is well-established. Therefore, net savings (savings minus investment) is widely accepted to respond positively to increases in real interest rates. -4- The private net savings surplus, NEPS, is shown as a function of the real rate of interest "r". Of course, other more structural factors, such as income distribution and so on, also influence these relationships. For any given external target CA, a particular fiscal deficit implies a required surplus of private savings over private investment, repeesented by the vertical line T(target) S(savings) in Figure 2.1. To bring forward the required private surplus, real interest rates will need to adjust until the economy is at the intersection of the NPS and TS schedules (A in Fig. 2.1). FIgure 2.1: PRIVATE SAVING, FISCAL DEFICITS, AND THE REAL RATE OF INTEREST r TS tS' NPS / / I /~~~~~~~ B , I~~~~~~~~~~~~~4p A~~~ NPS. TS CA + FDEF CA Cufrent Account Surplus RDEF Rscol Deficit NPS Net Priate Savings (Savns minus Investment) r Redl Rote d Interest TS Target SaWvngs. TS = CA + FDEF World Bank - 40039.2 -5- 2.5 Consider the impact of an increase in fiscal deficits: it will shift out the TS line, and so increase required private net savings; interest rates will tnen have to rise to restore internal and external balance, at point B. Even more perverse is the case where the measure increasing the fiscal deficit has a direct impact on private behavior, apart from its effect through fiscal revenues. An example is the expenditure rebate system instituted in Turkey in 1984, where partial rebates of income taxes are given in proportion to expenditure. This provides an explicit savings disincentive and so shifts up the NPS schedule: higher interest rates are then required to elicit the same amount of private savings. This will add further pressure on interest rates: the new equilibrium is at C rather than B. A.2 Macroeconomic Developments in Turkey in 1980-1986 2.6 To set the stage for the discussion of the 1980-1986 period, a brief overview of the preceding years is necessary. The period between the first oil crisis and the external payment problems of late 1977 was characterized by a moderate to high growth rate of an annual average 6 percent, slightly above the average for all oil importing developing countries over that period. Inflation remained fairly stable at about 20 percent until the end of the period. 1/ 2.7 However, interest rate and real exchange rate developments over the same period clearly suggest that this growth performance was at the cost of growing macroeconomic imbalances. Real interest rates averaged -11 percent over the period, and were decreasing 2/; the real exchange rate appreciated on average by 5 percent every year. 31 2.8 Such relative price developments can clearly not be sustained. The low level of real interest rates made it unduly attractive to cover expenditure through borrowing. The resulting demand pressure explains the rapid appreciation of the real exchange rate, with predictable results on external balance: excessive public dissaving, coupled with a lack of private expenditure restraint, produced unsustainable external deficits and a rapid run up of foreign debt. The trade deficit averaged approximately 8 percent of GNP during 1974-1977, until the payments crisis of late 1977 brought the process temporarily to a halt. Public savings and private net savings, at prevailing real interest rates, were clearly not compatible with acceptable external deficits; in the period before 1978, this conflict between private and public savings was resolved through external borrowing, a less expensive option at that time. As in the current (1985-86) period, public sector investment was the expenditure component that increased most rapidly. 2.9 The year 1978 saw an IMF program that brought some external adjustment through a small real exchange rate adjustment and actual foreign exchange rationing. But in the end, the program did not correct the 1/ The inflation rates are based on the wholesale price index series of the Undersecretariat of Treasury and Foreign Trade. 2/ The real interest rate is calculated as ((l+i)/(l+p))-l, where i and p refer to the nominal interest rate on one year time deposits and inflation rate respectively. 3/ The real exchange rate compares the Turkish CPI with a geometrically weighted average of the WPI in Germany and the U.S., using export weights after conversion into a common currency. -6- underlying problem of excessive aggregate demand; interest rates and fiscal policy were not adjusted with continued aggregate demand pressure as a result. Real interest rates went down to -23 percent in 1978 and -25 percent in 1979, suggesting strongly sustained excess demand pressure. As a consequence, the inflation rate shot up to 53 percent in 1978, and once again, to 64 percent in 1979. Any gain on the real exchange rate made in 1978 was subsequently lost: the real exchange rate appreciated by 12 percent in 1979 alone. 2.10 Output fell both in 1979 and in 1980, and growth rates in the subsequent period did not exceed the 1974-77 average until 1986. The downward shift in income was, therefore, not subsequently recovered. As a result the tax base eroded, and a structural deterioration in fiscal deficits took place, worsened by the rapid increase in the cost of external debt that took place over the same period. High inflation thus became ingrained in the system. 2.11 The year 1980 was a transition year, with a number of important adjustment measures taken. A major price reform for State Economic Enterprises (SEEs) affected the price level in an once-off fashion: the price level jumped by 29.3 percent in February alone. In addition, a nominal devaluation early in the year corrected a substantial part of the erosion of the real exchange rate that had taken place in the preceding years. With inflation now accelerating, not all of the devaluation translated in a real depreciation, but the real exchange rate still went down by more than 37 percent. Interest rates were not adjusted, however, and reached an (ex post) low of -37 percent. This is a somewhat misleading number, since the 107 percent inflation rate really should be seen as, at least partially, a once and for all price level shift. However, even in the last three quarters, real rates, using quarter-to-quarter inflation rates, were still -14 percent. This strongly suggests that, while some of the measures necessary for external balance and microeconomic efficiency were taken, aggregate demand was still out of line. 2.12 Balance was restored, at least temporarily, in 1981 and 1982. Higher interest rates reduced the conflict between internal and external balance. The real interest rate turned positive in the first quarter, toward 12 percent on average in 1981. Reductions in government expenditure took pressure off aggregate demand; and tax reforms increased revenues, and so reduced the fiscal deficit further. Tax reforms consisted of a shift from QRs to revenue raising tariffs as a trade intervention strategy; of the introduction of advance payments of income taxes (dropped one year later); and measures widening the base of the income tax. Moreover, the authorities continued aggressive use of nominal exchange rate policy in order to achieve external balance targets. As a result, the real exchange rate did not appreciate any further in 1981 and in fact depreciated by 11 percent in 1982. 1982 also saw a substantial increase in real interest rates, to no less than 21 percentage points. These price developments suggest that internal and external balance conflicts were resolved, as was excessive aggregate demand pressure. 2.13 However, the level of real interest rates is high by current international standards, even if the rate of real depreciation is taken into account. This strongly suggests that too much of the burden of external -7- adjustment was shifted to the private siector; for any external target, higher fiscal deficits require a larger privace surplus of savings over investment. Higher real interest rates are the mechanism that brings this about. 2.14 1983/1984 saw slippage on several fronts. Fiscal deficits started to increase under the influence of a steady erosion of central government tax collections (inclusive of the share transferred to local authorities); these declined from 19 percent of GNP in 1982 to 14 percent of GNP by 1984 (see Table 2.9). Also, by the end of the period, the reduction in the size of government started to be reversed, although mostly through off-budget items. Extra-budgetary funds (EBFs) were in excess of 20 percent of the Central Government Budget by 1985, up from only 8 percent two years earlier; this development reversed the trend towards smaller government. 2.15 Furthermore, less aggressive nominal exchange rate policy resulted in only a small depreciation in 1983 (by 4 percent), and by a further 6 percent in 1984. This was probably a factor in the slippage on the external account, in the face of declining remittances and a slow down in export earnings from the Middle East. The inconsistency between internal and external balance was reduced, but at the cost of a deteriorated external position. 2.16 The final structural change that occurred in 1984 is of great importance for the medium-run inflation outlook. The introduction of foreign exchange deposit accounts allowed Turkish residents to hold foreign exchange denominated deposits in Turkish commercial banks. This introduced a highly liquid asset, and so shifted down the demand for domestic base money. 1/ Since base money constitutes the basis over which the inflation tax is levied, a reduction in deficits would have been needed to prevent an increase in the sustainable rate of inflation. Instead, deficits increased. This increased the imbalance between inflation targets, on the one hand, and the sustainable rate of inflation implied by the fiscal deficit on the other. 2.17 The inflation rate doubled, from 25 percent in 1982 to 52 percent in 1984. There is no doubt that part of this represents a once-off shift in response to further SEE price reforms; but the change in financial structure transformed the price level shift into sustainable inflation at a higher rate. 2.18 1985 saw some reduction in fiscal deficits due to increased tax revenues. A set of producer taxes was replaced by a VAT system which, it turned out, brought in substantially more revenues than expected (about 3.3 percent of GNP). 2/ Also, central government budget and local government expenditures declined from 22.4 percent of GNP in 1984 down to 20.6 percent in 1985. However, the continuing proliferation of extra-budgetary funds has completely offset this decline (see Table 2.1). 2.19 1986 was marked by a high growth in output (of almost 8 percent) fueled by expansionary fiscal policies. Public fixed investment increased at a high rate (10 percent) for the second successive year but private fixed 1/ The concepts of base money, sustainable inflation and inflation tax, and the relation between them, are explained in paragraphs 2.48-2.49 and in more detail in Annex II. 2/ This number includes supplementary VAT but excludes VAT refunds (to exporters and purchasers of capital goods), i.e. it refers to net VAT revenue. -8- investment grew even faster (13.5 percent). Consumption expenditure too shot up, recording the highest annual growth rate (8.7 percent) since the stabilization program was introduced in 1980. Despite a sharp rise in domestic demand, inflation decelerated. The fall in the inflation rate was, however, largely due to a once-off decline in oil prices and it remains to be seen whether the government can take advantage of this situation to bring about a sustained reduction in inflation. While nominal interest rates on deposits were reduced by up to 10 percentage points, 1/ real interest rates in fact rose somewhat because of the more rapid decline in the inflation rate. 2.20 On the external side, the trade deficit did not change significantly (from 5.5 percent of GNP) in 1986 with an increase in capital goods imports offsetting favorable oil price and exchange rate developments. As a result of falling exports the real exchange rate depreciated by about 10 percent. However, because of the large fall in the value of the dollar with respect to other major currencies, the TL actually appreciated slightly against the dollar in real terms. These exchange rate developments, however, also resulted in an increase in the interest cost on foreign debt as the dollar value of DM obligations rose. The current account deficit deteriorated to 2.5 percent of GNP due to increased interest payments compounded by a continuing fall in workers' remittances and tourism receipts. Turkey's external debt increased sharply but a substantial part of the increase was due to a capital loss on account of the depreciation of the US dollar against the DM. 2/ 2.21 A final issue concerns a microeconomic set of measures, introduced in 1984, with particularly unfortunate macroeconomic effect. In an attempt to improve income tax as well as VAT compliance, wage-earners qualified for a partial rebate of their income taxes proportional to their expenditure on a wide range of goods and services. This scheme was later extended to all tax payers. The link with VAT compliance is that wage-earners have to show receipts in order to qualify, and these receipts can be checked against VAT records. The problem is that this scheme implies a disincentive to savings. This could not come at a worse time; Turkey is now trying both to achieve ambitious external balance targets and bring fiscal deficits under control; we identified the incompatibility between the two as a major force behind the current high level of real interest rates. Shifting back the aggregate supply curve of net private savings in this manner is the wrong thing to do at this time: it shifts up the aggregate demand curve for domestic goods, thus causing upward pressure on the real exchange rate. But, in addition, it leads to deterioration in external balance: a lower real exchange rate will be required to hit external targets at higher levels of private expenditure. To restore compatibility between internal and external balance, higher interest rates are needed in a period when interest rates are already high. 2.22 To sum up, both the period before 1980 and the past six years are characterized by a fiscal policy that ,s too expansionary to be compatible with external balance at acceptable (international) levels of real interest rates. Before 1980, this was resolved through a rapid run up of external debt. That possibility was foreclosed during the payments crises of the late 1970s and early 1980s. High real interest rates then became unavoidable to call forth the extra excess of private savings over investment that is now reconciling fiscal policy and external balance targets. While sustainability 1/ From January 1986 to January 1987. 2/ About 42% of the increase in total external debt between end-1985 and end-1986 can be attributed to such a capital loss. -9- i8 now not in immediate jeopardy from the external side, doubts should be raised about sustainability from the internal side: the private net savings surplus is coming, not so much out of high savings, but rather out of low private investment. Private savings in fact fell as a percentage of GNP, down from 11.7 percent in 1977 to 10.6 percent in 1980 and 9.7 percent in 1985. Private investment went down further however: from 11.8 percent of GNP in 1977 to 8.5 percent in 1980 and 8.4 percent in 1985. Alternatively, if private investment and output growth would not fall, external balance will deteriorate in the presence of large fiscal deficits. B. Inflation and the F.nancing of Government Expenditure 2.23 In the previous section, public sector net savings behavior was identified as a factor behind high real interest rates. Section B.1 below reviews the components of public sector expenditures and revenues in order to highlight the areas in which public sector dissaving is taking place. Section B.2 assesses whether current fiscal deficits are compatible with targets for inflation rates, another major macroeconomic area of concern. The final section, B.3, addresses some long term issues. 3.1 Trends in Public Sector Resource Mobilization Public Sector Expenditure and Revenues 2.24 The public sector comprises the c.6ntr*l government, local authorities and State Economic Enterprises (SEEs). Tables 2.1 and 2.2 show how the size ard deficits of the different components of the public sector evolved between 1981 and 1986. 2.25 Central government budget data give a misleading impression of recent public sector developments. The size of central government has fallen between 1981 and 1986. However, a different picture emerges once other components of the public sector are taken into account. Central government budget expenditure declined steadily from 23.1 percent in 1981 (as a share of GNP) to 18.3 percent in 1985. 1/ This downward trend was only partially reversed in 1986, when central government expenditure rose to an estimated 20.1 percent. However, when local governments and extra-budgetary funds are included, the size of government has actually increased, from 25.7 percent in 1981 to an estimated 28 percent of GNP in 1986. This is' uostly due to the rapid increase in expenditure channeled through EBFs: this increased from 1.1 percent in 1981 and only 1.3 percent in 1984 to 3.1 percent in 1985 and a projected 4.3 percent of GNP in 1986. But local governments too increased their expenditure as a share of GNP, from 1.5 percent in 1981 to 2.3 percent in 1985 and an estimated 3.6 percent in 1986. The increase is reinforced when SEEs are included: Table 2.1 shows that the share of the public sector in GNP inclusive of SEEs rises from 34.1 in 1981 to 37.4 in 1986. l/ Excluding VAT refunds paid to exporters and purchasers of capital goods, which have been netted out of VAT resenues. Turkish authorities report gross VAT revenues, treating these VAT rebates as expenditure items so the corresponding government reported expenditures are 19% of GNP in 1985 and 20.7% in 1986. -10- Table 2.1: THE SIZE OF THE PUBLIC SECTOR a/ (as percent of GNP) Est. 1981 1982 1983 1984 1985 1986 Central Govt. budget 23.1 22.0 22.6 20.6 18.3 b/ 20.1 b/ Extra-budgetary funds 1.1 1.1 1.2 1.3 3.1 4.3 Local governments c/ 1.5 1.4 1.7 1.8 2.3 3.6 Total public admin. 25.7 24.5 25.5 23.7 23.7 28.0 State Economic Enterprises d/ 8.4 8.2 8.6 10.4 10.6 9.4 TOTAL 34.1 32.7 34.1 34.1 34.3 37.4 a/ Sum of current and capital expenditure as a percentage of GNP. It is difficult to properly define the size of government. Table 2.1 measures the sum of all current and capital expenditure as a share of GNP. This exceeds the claim on actual resources by the public sector because it includes transfer payments. The row for SEEs gives value added produced by SEEs. b/ Excluding VAT refunds paid to exporters and purchasers of capital goods, which have been netted out of VAT revenues. Turkish authorities report gross VAT revenues, treating these VAT rebates as expenditure items so the corresponding government reported expenditures are 19% of GNP in 1985 and 20.72 in 1986. c/ Includes ISKI and Iller Bank. d/ Value added of non-financial enterprises. Source: SPO, Ministry of Finance and Customs, TFTU and Staff Estimates. 2.26 The central government has generally been the largest generator of public savings with local governments comprising a much smaller share and SEE self-financing of their investments at a very low level until recently. However, public saving has consistently fallen short of public investment, with substantial fiscal deficits as a result (Table 2.2). 2.27 The deficit measure used attempts to approximate the change in net liabilities of the government. It thus includes such items as capital losses on foreign debt due to exchange rate changes, although that does not require current financing and therefore does not increase the public sector borrowing requirement. Similarly it excludes asset sales and purchases, which influence the public sector borrowing requirement but do not affect the government's net worth. -11- Table 2.2: PUBLIC SECTOR SURPLUS a/ (as percent of GNP) Est. 1981 1982 1983 1984 1985 1986 Central government 1.1 1.0 0.0 -3.6 -2.2 -2.8 Extra-budgetary funds 1.2 0.8 1.1 1.8 0.8 0.4 Local governments b/ 0.4 0.4 - 0.2 0.2 -0.4 Total public admin. 2.7 2.2 1.1 -1.6 -1.2 -2.8 State economic enterprises c/ -9.5 -7.1 -8.0 -6.2 -5.7 -4.1 TOTAL d/ -6.8 -4.9 -6.9 -7.8 -6.9 -6.9 a/ The figures given here are net of intra-public sector tranfers. This does not affect the total public sector deficit but changes the composition of the deficit. For example, SEE expenditure covered by central government budget transfers counts as a component of SEE deficits; correspondingly this transfer is not included in the measure of central government expenditure and therefore does not contribute to the central government deficit. Similarly, transfers from the Support and Price Stabilization Fund (SPSF) to SEEs (see Chapter VII) are not counted as current expenditure of the extra-budgetary funds nor are they counted as part of SiLs' current revenues. Also, the issuance of revenue-sharing certificates (bonds) is counted as borrowing rather than as regular revenue, and purchases of government securities by government agencies are not counted as part of current expenditure. These procedures occasionally lead to substantial deviations from deficit numbers reported in IMF or previous World Bank documents on Turkey. b/ Includes ISKI and Iller Bank. c/ Includes only non-financial enterprises. d/ This deficit measure includes exchange rate losses on outstanding foreign debt as part of debt service expenditure; it also excludes inflationary capital gains on inventories from SEE profits and thus increases the deficit. On the other hand, it excludes capital gains on fixed assets; if these were included for the years they are available (TL 583 billion, TL 485 billion and TL 717.5 billion for 1983 to 1985 respectively), the actual deficits would be lower in each year although increasing over time. Source: SPO, Ministry of Finance and Customs, TFTU and Staff Estimates. -12- 2.28 The overall deficit was reduced in 1985 to 6.9 percent of GNP from the 1984 level of 7.8 percent (see Table 2.2). While both central government and SEE deficits declined, the latter still remains very large. 2.29 The SEEs contributed over 80 percent of the total deficit in 1985 (Table 2.2). Moreover, the financing requirements of SEEs were met by a large increase in borrowing; SEE profits and transfers from the government contributed less as a share of their requirements than in 1984. Net domestic credit to SEEs increased by 176 percent in 1985 compared with an increase of 53 percent to the central government and public administrations. The SEEs also increased foreign borrowing significantly whereas central government repayments of foreign debt actually exceeded borrowing in 1985. 2.30 Much of the increase in debt issue was used to finance a large increase in SEE fixed investment. Public fixed investment increased in real terms by about 13 percent in 1985. About 50 percent of this increase was due to SEE investments. The rapid growth in public sector investment continued in 1986. However, real fixed investment by SEEs is estimated to have declined as has their aggregate deficit. The numbers for 1986 are still provisional though, because the SEE accounts are not yet in. The continued high increase in public investment in 1986 is estimated to have arisen largely from local governments spending, but also from the central government budget and extra-budgetary funds. 2.31 Thus there are no signs that public expenditure growth slowed down; but the growth in revenue was a once-off event due to the replacement of producer taxes by VAT. Therefore, public sector deficits are likely to increase unless SEEs improve their profits substantially. Central Government Budget 2.32 Since 1980, the central government budget deficit (inclusive of transfers to SEEs) was reduced from 5.3 percent of GNP in 1980 to 2.1 percent in 1982, with an accompanying fall in the inflation rate. Thereafter, the fiscal situation deteriorated with the budget deficit climbing to 5 percent of GNP in 1984, and there was a reacceleration in iDfiation. In 1985 the budget deficit was reduced to 2.8 percent of GNP, and estimates for 1986 indicate a deficit of 3.2 percent of GNP. 2.33 A significant shortfall in budgetary revenues has been the main contributory factor in the worsening budget deficit of 1983 and 1984 (see Figure 2.2). From a high of 20.3 percent in 1981, the ratio of budgetary revenue to GNP declined steadily to an estimated 15.4 percent in 1984, with the budgetary tax revenue to rNP ratio falling from 18.2 percent to 12.9 percent over the same periLd (see Table 2.9). The downward trend in revenues was, however, reversed i. 1985 with revenues increasing to 15.5 percent of GNP and an estineced 17 percent in 1986. 1/ 1/ These figures differ from government reported revenues of 16.1% of GNP in 1985 and 17.5% in 1986, because VAT refunds (to exporters and purchasers of capital goods) have been deducted from VAT revenues in order to arrive at net VAT revenues. -13- FIGURE 2.2 SHARE OF CENTRAL GOV. BUDGETARY REV. AND EXP. IN GNP 25 -4 "s SHARE OF REVENUES 24 - 23 "s. ----- SHARE OF EXPENDITURES 22 (%) 21 19 18 17- 16- 15 , 1980 1981 1982 1983 1984 1985 1986 Source: SPO FIGURE 2.3 SHARE OF CENTRAL GOV. BUDGETARY EXPENDITURES IN GNP u12 CURRENT EXP. 11 10 ----- TRANSFERS 89- ....... INVESTMENT EXP. 6 _ . ....................... .... 3_ 4 ...................... ............. 2 1980 1981 1982 1983 1984 1985 1986 Source: SPO -14- 2.34 On the other hand, growth in expenditures was largely contained (Figure 2.2). Budgetary expenditures declined from 24.2 percent of GNP in 1980 to 20.6 percent in 1984, and further to 18.3 percent in 1985. 1/ Pressures, however, have arisen on the expenditure side and this is reflected in an increase in expenditures in 1986 to an estimated 20.1 percent of GNP. 1/ Both investment expenditures and transfers are estimated to have increased significantly. Transfer payments increased rapidly in 1983 and now are in excess of current expenditure (see Figure 2.3). The largest components of transfer payments are those for the expenditure rebate system, export tax rebates and to the social security fund; but the fastest growing components are interest payments on external and internal debt, the latter as a result of recent large issues of Treasury bills, and expenditure rebates. Extra-budgetary Funds 2.35 A large part of central government finance is now off-budget. The importance of extra-budgetary funds has grown rapidly with the establishment of some major funds in the past two years. While the total resources of the Funds are not known, it is estimated that their revenues in 1985 were close to TL 1 trillion, equivalent to about 20 percent of central government budget revenues or 3.5 percent of GNP. 2/ This is a rapid increase even from 1983 when their total revenues were about 8 percent of budget revenues or 1.4 percent of GNP. 3/ A conservative estimate of the total revenues of the Funds in 1986 is TL 1.7 trillion, equivalent to about 4.5 percent of projected GNP. 2.36 Almost 70 percent of the Funds' revenues are from taxes. Principal tax revenue sources of the Funds are earmarked taxes on trade, petroleum products, a surcharge on bank loans and various excises on tobacco, alcohol and beverages. Non-tax revenues were from various dources ranging from fees and penalty payments to the operating income from infrastructural facilities. 2.37 Overall, while the Funds are not a source of deficit, there are individual funds which are not able to cover their expenditures from their earmarked revenues. At present the deficits are largely being met by postponing commitments but borrowing, though small as yet, seems to have become another option. Local Governments 2.38 Local authorities include municipalities, provincial special administrations and villages. The central government approves the local budgets, which are required to balance; local governments do not borrow directly from the market, their deficits generally being covered by grants and 1/ Excluding VAT refunds paid to exporters and purchasers of capital goods, which have been netted out of VAT revenues. Turkish authorities report gross VAT revenues, treating these VAT rebates as expenditure items so the corresponding government reported expenditures are 19% of GNP in 1985 and 20.7% in 1986. 2/ On the basis of information on 12 major funds operational in 1985, whose total revenues (net of inter-fund transfers) were estimated to be TL965 billion (see Chapter VII). 3/ This estimate excludes revenues of the Petroleum Price Stabilization Fund (PPSF) and the Selective Credit Fund (SCF), for which data are not available. -15- loans from the central government. Ilier Bank, the Province Bank, acts as a disbursing and debt-collecting agent for the central government and assists local governments in preparing and implementing projects. 2.39 The main sources of local government revenues are a share of central government tax collections, local taxes, user charges for local services, and revenues from properties and assets of municipalities and regional governments. The revenue-sharing arrangements have become fairly complex. 1/ The flow of funds is illustrated in Fig. 2.4 and the total shares transferred to local authorities is sunmarized in Table 2.3. Over and above these revenue-sharing arrangements, a portion of the revenues of the Petroleum Consumption Fund is also earmarked for municipalities. Table 2.3: SHARE OF LOCAL AUTHORITIES IN TOTAL TAX REVENUES COLLECTED BY THE CENTRAL GOVERNMENT (in 2) Munici- Special Metropolitan palities a/ Administrations Municipalities b/ TOTAL 1981 5.00 1.00 6.0 1982 5.00 1.00 6.0 1983 5.00 1.00 - 6.0 1984 7.15 1.15 0.75 9.05 1985 8.15 1.15 2.6 11.9 1986 c/ 8.55 1.20 2.6 12.35 a/ Includes share of tax revenues accruing to Iller Bank through the Municipalities Fund. See Fig. 2.4. b/ Ankara, Istanbul, Izmir and Adana. The shares legislatively specified for metropolitan municipalities are actually a percentage share of their own collections. These have been translated into an equivalent share of total tax collections, assuming that 50X of total tax revenues is from metropolitan municipalities. c/ The shares for 1986 are applied to net tax revenues (i.e., net of expenditure tax rebates, VAT refunds and tax errors). The total share of gross tax revenues is equivalent to 11.13S. However, property taxes were transferred to local jurisdiction in 1986; once these are added in, the total share of local authorities in gross tax revenues rises to 12.62. Source: SPO 1/ For municipalities, a share of central government total tax revenues is transferred to a Municipalities Fund, to be used by Iller Bank to finance investment projects of municipalities. In addition to this, municipalities directly receive a share of total tax revenues. A supplemental share of tax revenues is earmarked for the large municipalities. Ankara, Istanbul, Izmir and Adana. Finally, a share is distributed to. a Local Administrations Fund, half of which accrues to municipalities. A similar revenue-sharing arrangement exists for the Special Administrations, which receive a directly distributed share of total tax revenues, in addition to the share accruing to the Special Administrations Fund and half of the share which is transferred to the Local Administrations Fund. These shares are specified in Figure 2.4. Fig. 2.4: CENTRAL GOVERNMENT FUNDING OF LOCAL GOVERNMENTS: Rovenue-Sharing Arrangements Municipafties' Share (directly distributed) Municipalities' Own 1984 - 6% Revenue Collection 1986 - 6% 1986-87 6% + Property Tax Metropolican Municipalities' Share a/ Total Revenue 1984 - 0.75% of Municipalities 1985-87 - 2.6% 196- 0.25% Local Admin. Fund 1987 - 0. 26% Total Tax 1984 - 85 - 0.30% Revenue 1986 - 0.456 1987 - 0.45% 1986 - 0.2% 1987 - 0.45%1987- 0.3% Municipalities' Fund Special Admin. Share Total Renveues 1984-85 - 2% (directly distributed) _of Special 1986 - 2.3% 1984 - 86 - 0.80% -Adfmpeial 1987 - 3% 1987 - 1.12% Admins. Special Admin. Fund Special Iller Bank 1984-86 - 0.2% Administrations' 1987 - 0.28% Own Revenue Collection a/ Ankara. Istanbul. Izmir and Adana. The shares legislatively specified for metrolitan municipalities are actually a percentage share of their own collections. These have been translated into an equivalent share of total tax collections, assuming that 50% of total tax revenues is from metropolitan municipalities. Source: SPO pk/w4O0039 -17- 2.40 Total revenues of local authorities were slightly less than 2 percent of GNP between 1981 and 1984 and rose to about 2.5 percent in 1985 and an estimated 3.2 percent in 1986 (see Table 2.4). This sharp increase was partly due to higher resource transfers from the central government on account of: (i) the modified revenue-sharing arrangements legislated in 1984; (ii) the transfer of the real property tax collection to local governments from 1986; and (iii) the transfer of revenues from extra-budgetary funds since 1985 to finance municipal expenditures. Local governments' own revenues, particularly tax revenues, have not been very elastic, but in 1984 there were significant increases in local taxes. 1/ 2.41 Local authorities were not a source of public sector deficits. However, local governments have initiated much larger expenditure programs as a result of increased resources and are projected to run a deficit in 1986 and in 1987. State Economic Enterprises 2.42 The net earnings of State Economic Enterprises (SEEs) comprise an additional source of public savings. However, the profitability of SEEs had been a matter of concern for some time. In the late 1970s, the aggregated losses of operational SEEs were about 4 percent of GNP. Their poor financial performance could partially be attributed to price controls, but on the other hand, they received various subsidies. Periodic conversion of outstanding debt into equity (the most recent such consolidation of debt having taken place in end-1983) as well as other indirect transfers, such as tax privileges and interest subsidies, made it difficult to assess the actual financial performance of SEEs. 2.43 Under the January 1980 program, price controls on SEE commodities and services were lifted (with a few exceptions); SEEs have since been required to cover their operating costs and generate funds for their investment programs through price changes and efficiency improvements. Simultaneously, their access to concessionary funds has been restricted. Also, budgetary transfers were reduced; the ultimate aim is to provide transfers solely for the few remaining subsidized activities. 2.44 Substantial price increases resulted in a considerable improvement in the financial performance of SEEs, with a situation of chronic deficits changing into one of marginal but growing profits. By 1984, SEE profits (after taxes) amounted to 1.9 percent of GNP. As a result, SEEs were able to finance one-third of their investment program in 1984, as compared with an average of 1.2 percent between 1980-83 of self-financing of their investments. In 1985, their profits are estimated at almost 2 percent of GNP. Nevertheless, their financing requirements are still large, and they are the major factor behind the public sector deficit. 1/ In some cases the municipal assembly fixes the amount of taxes and fees within upper and lower limits specified by law, whereas some other taxes and fees are determined by the Cabinet. (Law 2464 defines the taxes and fees of municipalities.) -18- Table 2.4: REVENUES AND EXPENDITURE OF LOCAL AUTHORITIES a/ (Municipalities and Provincial Special Administrations) (TL billion at current prices) Est. 1981 1982 1983 1984 1985 1986 REVENUES Share of Central Govt. Tax Revenues b/ 76 97 124 200 379 610 Transfer from Petroleum Consumption Fund - - - - - 7 Local Taxes 11 21 26 82 108 285 c/ Non-tax Revenues d/ 38 42 41 98 211 369 TOTAL 125 160 191 380 698 1271 (as 2 of GNP) (1.9) (1.8) (1.7) (2.1) (2.5) (3.2) EXPENDITURE Current 59 70 109 149 245 411 Transfers 13 20 33 21 62 139 Investment 24 33 53 168 343 870 TOTAL 96 123 195 338 650 1420 (as Z of GNP) (1.5) (1.4) (1.7) (1.8) (2.3) (3.6) BUDGET BALANCE 29 37 -4 42 48 -149 (as 2 of GNP) (0.4) (0.4) (-) (0.2) (0.2) (-0.4) a/ Includes ISKI and Iller Bank. Final accounts of local governments are available only upto 1983 for municipalities (excluding accounts of some municipalities) and upto 1984 for Provincial Special Administrations. The following years' figures are estimates, based on a survey carried out by the SPO. bJ Includes share of tax revenues transferred to Iller Bank and ISKI. c/ Includes TL 100 billion revenue from the real property tax which was transferred to local jurisdiction in 1986. d/ The share of local governments' own revenues and current expenditures may be somewhat underestimated because in some cases the non-tax revenve figures refer to net income from various municipal operations (water supply, bus services, slaughter houses, etc.). An analysis of five municipalities in Cukurova, for instance, indicated that only about 30-402 of total gross revenues of the municiE.lities in 1984 came from the central government. Source: SPO, Ministry of Finance and Customs. -19- B.2 Consistency of Fiscal Policy and Inflation TarRets 2.45 Consistenoy between fiscal deficits and other macroeconomic targets can be judged starting from the government budget constraint; this constraint says that the sum of the non-interest deficit plus the interest bill on foreign and domestic debt is equal to the sum of financing from all sources. Fiscal deficits can be financed in three ways: issue of external debt; issue of internal interest bearing debt; and, finally, monetary financing. Macroeconomic targets, for example targets for the inflation rate, external debt, GNP growth and so on, imply restrictions on each of these financing methods. These restrictions add up to a total financable deficit of a certain magnitude; if the actual deficit exceeds that number, one of the non-fiscal targets will need to give or, alternatively, fiscal policy requires adjustment. A formal framework is presented in Annex II. In the rest of this section we apply this framework to an assessment of compatibility of fiscal deficits and other macroeconomic targets in Turkey. 2.46 Consider external debt targets and the restrictions on financing they imply. There is nothing magic about any given debt-output ratio. In fact, the ratio of external debt to GNP at the beginning of 1985 was more than twice as high as it was in the crisis year 1978. Nevertheless Turkey's creditworthiness has improved dramatically, because of its improved export performance. However, due to the recent slowdown in Turkey's exports, further increases in the debt-output ratio seem imprudent. Such a target implies that the real value of the external debt (in terms of foreign goods) should not increase faster than the targeted rate of real output growth. This assumes, moreover, no change in the real exchange rate. A real depreciation will, ceteris paribus, increase the ratio of foreign debt to GNP and thus further restrict the room for external financing if the debt-output ratio is to be maintained at beginning of period levels. This link points to a potential conflict between fiscal retrenchment and increasing external competitiveness through exchange rate depreciation in the presence of a substantial foreign debt. At the 1985 real growth rate of 5 percent, maintaining the end of period debt-output ratio of 48 percent 1/ allows for foreign financing of the deficit at most to the extent of 2.4 percent of GNP; a 5 percent real depreciation would reduce the room for external financing to zero as the reduction in debt-output ratio by a 5 percent real output growth will be offset exactly by the capital losses on external debt associated with a 5 percent real depreciation. 2.47 A more contentious question involves the domestic debt issue. The real interest rate on auctioned one-year government paper now is approximately 18 percent. Issuing domestic debt at such a high interest cost will allow lower money growth but at the cost of future increases in debt service obligations and ensuing budget deterioration. The latter occurs because debt service will grow explosively at real interest rates so far above the real growth rate of the economy. A debt strategy that so clearly sacrifices future budget balance for current monetary restraint is likely to fuel inflationary expectatioins even if favorable external shocks allow a temporary decline in 1/ 48 percent refers to external debt by central and local government and SEEs plus foreign liabilities of the Central Bank minus the Central Bank's foreign assets. Also, a correction has been made for the fact that dividing debt by nominal GNP implies deflation of debt by an average rather than end of period price index. Central Bank foreign assets are taken from its Quarterly Bulletin. Other debt data are provided by the Treasury. -20- the rate of increase in prices; this in turn will keep nominal (and heace "ex-post" real) interest rates high, thus fueling a vicious circle of high interest rates, high public sector debt service, increasing budget deficits, high inflationary expectations and finally back to high interest rates. We demonstrate the negative impact of such a debt strategy on fiscal balance in Table 2.6 below. Relying on further domestic debt issue should, therefore, be avoided as cheaper forms of debt are still available. In what follows we will assume that the government does not allow any further real increase in domestic public sector debt. 2.48 Finally the third source of financing, the issue of money. The government can raise revenue through monetary financing in two ways. First, for any given inflation rate, people will want to hold a certain amount of real money balances in relation to GNP (see Annex II for a quantitative assessment). Thus.a positive growth rate of output implies that the government can increase the real money stock in line with real output growth without undue pressure on inflation targets. This source of monetary financing is called seignorage. Its extent depends on the rate at which the output growth takes place; in Turkey a 5 percent real growth rate at a 25 percent inflation rate implies seignorage revenue of only 0.54 percent of GNP (see Annex II). Lower inflation increases desired money holdings in relation to output, and will thus raise seignorage revenue. Growing monetization of the economy will likewise augment seignorage revenue. 2.49 The second source of monetary financing provides the link between fiscal deficits and inflation targets. We already saw that a particular inflation rate implies that people will want to maintain a certain ratio between real money balances and output. However, at positive inflation rates, it is not enough, for maintaining this money-output ratio, to keep nominal money balances constant even if there is no real output growth. Inflation would reduce the ratio of real money to output, unless nominal balances are increased enough to offset this "inflationary erosion". This is best seen as a form of taxation since money holders need to reduce expenditure below income to achieve this without any increase in real assets following from it. This source of revenue is therefore referred to as the inflation tax. At moderate inflation rates, the revenue it yields increases with the inflation rate. From the money demand estimates presented in Annex II, it follows that the government can expect 2.7 percent of GNP from the inflation tax, if inflation remains at 25 percent a year. 2.50 In the preceding paragraphs we outlined how much Turkish authorities can expect from the various sources of financing if prudent internal and external debt strategies are pursued and a real growth target of 5 percent and an inflation target of 25 percent is adhered to. Adding up the financing these sources offer under the constraints implied by the macroeconomic targets then yields the fiscal deficit that is compatible with those targets. This exercise yields an estimate of the financable deficit of 5.64 percent of GNP. This number should be compared with the actual deficit. The public sector borrowing requirement is not, however, the appropriate figure to look at for this comparison. The proper concept of deficit for this purpose should, besides incorporating all components of government, include only real interest payments on foreign and domestic debt, and treat exchange rate losses on foreign debt as part of the cost of servicing foreign debt. The deficit thus construed was 6.6 percent of GNP for 1985, a full percentage point above what is required for compatibility with a 25 percent inflation rate, the stated government target. The actual deficit was in fact consistent with a sustained inflation rate of 35 percent (see Table 2.5 below). A tighter inflation -21- target, say 15 percent, will slightly increase anticipated seignorage, as desired money holdings will rise; however it will also substantially reduce the revenue from inflation tax, down to 1.7 percent of GNP, from 4.5 percent, its 1985 value. 1/ The required deficit reduction (RDR) therefore increases to almost 2 percent of GNP (1.932) as Table 2.5 shows. Table 2.5: REQUIRED DEFICIT REDUCTION FOR CONSISTENCY WITH VARIOUS MACROECONOMIC TARGETS (AS PERCENT OF GNP) NO REAL DEPRECIATION; REAL GNP GROWTH 5S INFLATION TARGETS: p = 15 25S 35% 451 RDR 22 1S 01 -0.8S NO REAL DEPRECIATION; INFLATION 252 REAL GNP GROWTH TARGETS: n = 21 51 71 RDR 2.71 1X 0% REAL GNP GROWTH 51; INFLATION 251 REAL DEPRECIATION TARGETS: c a 01 52 101 RDR 1% 3.41 5.8% c rate of depreciation of the real exchange rate n growth rate of real GNP p inflation rate RDR required cut in 1985 fiscal deficit as a percentage of GNP. Note: A minus sign indicates that the deficit can actually be increased without any conflict with the other macroeconomic targets. 2.51 The deficit in 1985 was an improvement over 1984, largely due to expenditure cuts and higher than expected VAT revenues, and it was just compatible with a current inflation rate of 35 percent (see Table 2.5). However, even a moderate inflation target of 25 percent, and certainly a further reduction to 15 percent, requires further adjustment in fiscal policy. The conclusion is clear: to maintain consistency between fiscal policy and even a moderate inflation target of 25 percent, a substantial further reduction in fiscal deficits, of 1 percentage point of GNP, is required. Current fiscal policy and inflation targets cannot simultaneously be adhered to for a prolonged period of time. 2.52 Table 2.5 shows the fiscal deficit reduction required in order to maintain consistency with various values of other macroeconomic targets. Clearly a higher GNP growth rate relaxes the financing constraint by increasing revenue from seignorage and by allowing more foreign debt accumulation. Hence the smaller required adjustment effort as the growth rate of GNP rises. It is in fact conceivable that policy measures are required for less than the indicated adjustment if growth rates are higher than 5 percent and more if they are below. This is because growth will influence not only the financable defitit as indicated in Table 2.5, but also the actual deficit itself. Thus some of the required adjustmen3 might take place automatically as growth picks up, and conversely when growLh slows down. 1/ Reserve money was 10 percent of GNP in 1985, and inflation 45 percent; hence inflation tax revenues came to 4.5 percent of GNP. -22- 2.53 The final row in Table 2.5 shows the fiscal implications of an export strategy that relies on real exchange rate depreciation to maintain export growth. Real depreciation raises the cost of servicing foreign debt and so reduces room for fiscal policy. The numbers are substantial, because external public debt is now so high. Once again taking 1985 as a bench mark, public debt stands at 48 percent, 1/ s0 a 5 percent real depreciation will increase the required adjustment effort by 2.4 percent to bring it at 3.4 percent of GNP assuming 5 petcent real output growth and 25 percent inflation. Similarly, a 10 percent real depreciation will require a fiscal adjustment of 5.8 percent to restore policy consistency at this rate of inflation and real output growth. 2.54 Finally consider the consequences of an alternative financing method. In the previous paragraphs we argued that domestic debt carries such a high real interest rate that financing deficits through domestic bond sales will lead to explosive debt growth and an increasing pressure on future fiscal deficits because of higher real interest payments on internal debt. This is demonstrated in Table 2.6, where we assume a restrictive monetary policy aiming at 15 percent inflation, and otherwise the same basic policy stance (non-interest deficit) as in 1985. However, we assume that the fiscal adjustment required to sustain a monetary policy geared for 15 percent inflation is NOT undertaken. Instead, deficits are covered by issuing domestic debt at a real rate of 18 percent, about the 1986 real interest rate on auctioned government securities. Each year, the interest on debt so issued is added to the deficit to be financed. The numbers in Table 2.6 show the required adjustment in each of six successive years if fiscal consistency were to be attained in that year. 2.55 Table 2.6 clearly demonstrates that while bond sales make it possible to sustain restrictive monetary policy, they do so at the cost of ever increasing required fiscal cut backs if consistency is to be restored. The reason for this is that the real interest rate on internal debt substantially exceeds the growth rate of the economy. That is clear from a comparison of column one and two: in the first column we assume a zero real growth rate, in the second (and the third and fourth) a 5 percent real growth rate for GNP. In both cases debt issue has clearly compounded the problem: the required effort to restore consistency has gone up from 2 percent to 4.6 percent and 3.6 percent respectively after six years. In the low growth rate scenario, the increase is a full percentage point higher than in the 5 percent scenario. Also, as the last row of Table 2.6 indicates, the debt-output ratio increases more in the low growth rate scenario. 2/ 1/ 48 percent refers to external debt by central and local government and SEEs plus foreign liabilities of the Central Bank minus the Central Bank's foreign assets. Also, a correction has been made for the fact that dividing debt by nominal GNP implies deflation of debt by an average rather than end of period price index. Central Bank foreign assets are taken from its Quarterly Bulletin. Other debt data are provided by the Treasury. 2/ In fact, the situation is worse: a lower growth rate for given real interest rate not only influences the dynamics, but would also increase the base year Required Deficit Reduction (see Table 2.5). This is ignored in column one of Table 2.6 to bring out the dynamic complications more clearly. -23- Table 2.6: FISCAL CONSEQUENCES OF INTERNAL DEBT FINANCE WITH NO FISCAL ADJUSTMENT, AT A 152 INFLATION TARGET AND 18% REAL INTEREST RATE ON INTERNAL DEBT. Year n=O n=5 n=5 n=5 cmO c=O c=5 c=5 * * * ** REQUIRED CUT IN FISCAL DEFICIT (AS 2 OF GNP) 1 2 2 4.4 6.8 2 2.4 2.2 4.9 7.6 3 2.8 2.5 5.6 8.6 4 3.3 2.8 6.2 9.7 5 3.9 3.2 7.0 10.8 6 4.6 3.6 7.9 12.2 Increase in 18.9 16.4 36.1 55.7 domestic debt- output ratio (perc. points of GNP) * Ratio of external debt to GNP held constant ** Nominal dollar value of external debt constant c rate of depreciation of the real exchange rate n growth rate of real GNP RDR required cut in fiscal deficit as a percentage of GNP. 2.56 The third and fourth column change the assumptions on exchange rate policy and external debt management respectively. the difference between column two and three is that the latter assumes a 5 percent real depreciation per annum, which adds to the cost of external finance. This has implications for external debt accumulation, incidentally; the exercise has been set up under the assumption of a constant external debt-output ratio. At a 5 percent real rate of depreciation and 5 percent real growth rate of GNP, that implies no real (in terms of foreign goods) increase in external debt. In other words, to meet the external debt target, nominal debt can only go up at the foreign rate of inflation under this scenario. This clearly adds to the pressure on domestic finance, as Table 2.6 shows; the total of internal debt to GNP increases no less than 36 percentage points over the six years shown. Here too compound interest on internal debt makes for a rapid deterioration of the fiscal situation: the required adjustment rises from 4.4 percent to 7.9 percent in six years under the pressure of the increasing cost of servicing the growing internal debt. 2.57 The fourth column assumes a stricter target on external account, no increase in the dollar value of external debt. This implies about 5 percent real decline with an assumption of 5 percent world inflation. This adds a further 2.4 percent of GNP to domestic financing requirements if no matching expenditure cuts or tax increase takes place. Table 2.6 shows that under this scenario the fiscal situation will deteriorate very rapidly because of the large interest differential between internal and external debt. -24- 2.58 To sum up, the tables highlight a number of interactions between different instruments and targets. First, pursuing current restrictive money growth targets by relying on debt finance rather than cuts in the the non-interest deficit sacrifices future budget balance if real interest rates exceed the real growth rate of the economy. Second, there is a conflict between fiscal retrenchment and increasing external competitiveness through exchange rate depreciation in the presence of a substantial foreign debt. This is because of the impact of a real depreciation on the real cost of foreign debt service. Thus, an export strategy that relies on real depreciation to maintain export growth requires more stringent fiscal policy to an extent demonstrated in the preceding paragraphs. Finally, there is a link between stabilization policy and growth: a higher GNP growth rate relaxes the financing constraint by increasing revenue from seignorage and by allowing more foreign debt accumulation within a given debt-output ratio constraint. Hence the smaller required adjustment effort as the growth rate of GNP rises. 2.59 Further worries arise on a number of counts. We mention three. First, it should be realized that at an external public sector debt to GNP ratio of 48 percent, a 5 percent real depreciation reduces room for real expenditure by a full 2 percentage points of GNP if consistency with inflation targets is to be maintained. The increase in the current account deficit in 1986 and the likely necessity of a further real depreciation should, therefore, lead to extra pressure for restrictive fiscal policies. 2.60 Second, and more with a short-term focus, the improvement in inflation over 1986 was helped by an estimated 10 percentage points FALL in the dollar price of imports. This is, however, best seen as a once off fall in the price level, rather than the beginning of a sustained reduction in the rate of inflation: there is no reason to expect such a windfall gain to be repeated in 1987. Finally, the rapid increase in domestic credit (80 percent over the last 12 months) strongly suggests, if at least it is related to public sector credit expansion, that the fiscal policy performance as measured by public sector deficits is slipping. 2.61 The overall conclusion is therefore that the substantial decline in inflation since last year has NOT done away with the continued need for substantial adjustment in the deficit of the public sector. B.3 Longer-Term Concerns 2.62 A longer-term concern involves not so much the actual deficit numbers but more the increasingly fragmented structure of (fiscal) policymaking in Turkey. While substantial efforts were made in reducing the size of the central government, an excessive growth of other components of the public sector is taking place through the creation of extra-budgetary funds, now well in excess of 20 percent of the central government as measured by their revenues. However, the problem that control over the size of government is lost is not the only problem created by the rapid proliferation of EBFs. -25- 2.63 The EBFs are financed largely by earmarl. tax revenues, with rates and sources often adjusted to meet their revenue needs. This implies a haphazard development of the tax structure, spurred along by partial revenue needs and ad hoc earmarking rules, without concern for the overall efficiency of the tax structure that emerges over time. Mutual inconsistencies and conflicts with other policy objectives of the government then become unavoidable. The increasing reliance of the EBFs on trade taxes is such an example. 2.64 The final concern is the difficulty in implementing a cautious debt management policy with the different components of the government operating without central control over their funding policies. Examples are the delays in information, even within the government, on the build-up of external debt by the local and municipal authorities. Similar concerns arise out of the recent granting of external borrowing authority to some EBFs. The absence of adequate information collection mechanisms, let alone instruments of control, calls for concern. C. Fiscal Deficits and Taxation: An Introduction to the Remainder of this Report 2.65 This chapter dealt with the level of government revenues and expenditures, rather than with their composition over different tax instruments. However, a tax structure cannot be judged solely on the level of revenues it yields; since all practical sources of government revenues influence private incentives one way or the other, concern with the efficiency of the tax structure, for any given level of tax revenues, is of paramount importance. That will be the main focus of the remainder of this report. An overall picture of how the level and composition of tax revenues has changed over time is, however, useful before proceeding with a discussion of the individual tax instruments. 2.66 Turkey's central government budget tax ratio (tax revenues as a share of GDP) is somewhat below the average of the sample of middle income countries presented in Table 2.7. There are two qualifications, however. First, the tax ratio is conaiderably higher once the tax revenues of local authorities and of extra-budgetary funds are included (see Table 2.9; but then the numbers for other countries may need similar adjustment). Second, tax revenues increased later on due to the successful introduction of the VAT in 1985. The decline in tax revenues between 1981 and 1984 (Table 2.9) was to a large extent a consequence of the belated change in the income tax structure to offset the effects of prior inflation. 2.67 Inflation has had a major impact on the composition of central government tax revenues. Due to a progressive rate structure and a failure to adjust income tax brackets during the 1970s, income taxes increased their share in total central government tax revenues dramatically as inflation shifted more and more people into higher income tax brackets. The contribution of direct taxes to total central government tax revenues increased from 37 percent in 1970 to 62 percent in 1980, with a corresponding decline in the share of indirect taxes. 2.68 The tax reforms introduced since l981 have realigned the tax structure both by lowering taxes on income directly and, with the introduction of the value added tax, by placing a greater reliance on indirect taxation. Table 2.7: TAXC BY TYPE OF TAX FOR S91D DVELP SIRS (in percent of GP) QEi Tales On M & PnDfit Tames ax Taws m Other Total Nooax Total tunry Year at 1983 Total Indivixal G,rporation. Goods & Services lIt. TPade Taxes Taxes Revme Bewu* lulonesia 1983 560 14.9 0.5 13.5 2.1 0.9 0.3 A81 2.1 20.2 Evpt 1983 700 7.7 0.8 7.0 5.4 7.0 7.3 27.4 17.2 44.6 molocco 193 760 4.5 2.4 1.9 9.4 4.7 3.4 22.0 3.9 25.9 Pbiippippes 1983 760 2.3 L0 1.3 4.4 3.2 0.4 10.3 L5 11.8 lhalani 1983 820 3.0 L5 1.4 7.1 3.2 0.4 13.7 1.4 15.1
Группа Всемирного банка · Pre-2003 Economic or Sector Report
Turkey - Fiscal policy and tax reform : issues from the past and options for the future (Vol. 1 of 2) : Main report
Открыть оригинал документа
Полный текст размещён на сайте публикующей организации. lawenc.com индексирует метаданные и ведёт на официальный источник.
Полный текст
Основные сведения
Организация
Группа Всемирного банка
Тип документа
Pre-2003 Economic or Sector Report
Страна
Турция
Источник
Всемирный банк