Resource mobilization through public savings : a case study of Turkey's plan strategy

CCNFIDENTIAL 7F-O STUDIES IN DOMESTIC FINANCE NO. 1 RESOURCE MOBILIZATION THROUGH PUBLIC SAVINGS A Case Study of Turkey's Plan Strategy By D.C. Rao With an Annex by Kazuko K. Artus Public and Private Finance Division Development Economics Department I would like to acknowledge a considerable debt to Mr. Montek S. Ahluwalia who initiated this inquiry in an earlier paper and provided many helpful suggestions and comments. The responsibility for remaining errors, however, is mine alone. This study was prepared in connection with the Basic Economic Mission to Turkey in April 1973 led by Mr. Edmond Asfour. The analyses, results and judgments in this document, however, are subject to change in the Basic Economic Report, as the latter incorporates further analyses and revisions in the context of mission discussions and more recent data. April 1973 RESOURCE MOBILIZATION THROUGH PUBLIC SAVINGS A Case Study of Turkey's Plan Strategy TABLE OF CONTENTS I. INTRODUCTION ............................ 1 (i) A Public Savings Strategy ........................ 1 (ii) Scope of the Study ................................ 2 II. NATIONAL SAVINGS ............................ ......... 3 (i) Determination of Targets ......................... 3 - Plan Formulation in Turkey ..................... 3 - Savings in the Growth Model .................... 4 (ii) Plan Targets and Performance ................ 7 (iii) Public Savings .............. . ......... 9 (iv) Role of Factor Incomes from Abroad ............... 13 III. GOVERNMENT REVENUE AND EXPENDITURE .................... 15 (i) Trends in Tax Revenues 15 (ii) Structure of Tax Revenues ........................ .16 (iii) Third Plan Tax Projections ...................... 17 (iv) Public Consumption ...............*.............. 19 IV. STATE ECONOMIC ENTERPRISES ............................ 20 (i) Introduction .................................... 20 (ii) SEE Profits ...................................* . 22. (iii) Socio-Economic Goals ............................. 26 (iv) The Third Plan ................................... 28 V. PUBLIC SEC TOR BORROWING ................. .......... 28 (i) Introduction ....dcn.......... ................... 28 (ii) Components of Public Borrowing .................. 29 (iii) Savings Bonds .............. ........... ........ 30 (iv) State Investment Bank ......................... 32 (v) Government Bonds ............ 33 (vi) Central Bank Advances ...................... 34 (vii) Agricultural Credit ............................. 36 (viii) Third Plan Projections ........................... 37 VI. MONETARY POLICY ....................................... 40 (i) Introduction ..................................... 40 (ii) Money and Inflation .............................. 40 - Exogeneity of Money Supply .................... 41 - Money demand *** *** ** *** *** *............. 42 TABLE OF CONTENTS (CONT'D) (iii) Interest Rate Regulation ................. 43 - Bank lending rates ..... ................... 43 - neposit rates ......... . . . ... .. . . .. 46 - Corporate bond rates ........................... 48 - Effective cost to borrowers .................... 48 - Distortions .....0.......... . ....... 50 VII. CONCLUSIONS ....... ................. 0.......00.......6 51 LIST OF TABLES 1I.1 Investment and Savings, 1962-1972 11.2 Marginal Savings Rate II.3 Alternative Savings Ratios 11.4 Public Savings Estimates, 1968-1971 11.5 Unidentified Transfers in'Consolidated Budget 11.6 Public Savings Target 11.7 Factor Income, National and Domestic Savings, 1968-1971 111.1 Tax Revenues: Plan Targets and Performance 111.2 Structure of Tax Revenues 111.3 Consolidated Budget Tax Revenue Projection, 1972-1977 111.4 Public Consumption 111.5 Composition of Consolidated Budget Expenditure IV.1 Investment by Operational SEEs IV.2 SEE Surplus, Plan Targets and Performance IV.3 Net Profits of Operation'al SEEs IV.4 Rate of Return of Operational SEEs V.1 Second Plan Public Sector Balance, Targets, 1967-72 V.2 Elements of Public Borrowing V.3. Bank Credit to the Public Sector V.4 Central Bank Advances to Treasury V.5 Third Five-Year Plan Financing Balance, 1973-77 VI.1 Regression Results: The Demand for Money VI.2 Bank Deposit Interest Rates VI.3 Additional Costs of Bank Credit LIST OF GRAPHS II.1 Gross National Savings as Percent of GNP IV.1 SEE Surplus, 1963 to 1971 VI.1 Price Trends, 1962 to 1971 VI.2 Regression Estimate of Price Trends Annex A: Tax Pevenue Forecasting in Turkey ......... Prepared by: Mrs. Kazuko K. Artus Annex B: Savings Ratios and Factor Income from Abroad Appendix Table 1: Expenditure on Gross National Rrodtct, Plan Targets and Actuals, 1963 to 1972 endix Table 2: Savings: Plan Target and Actuals, 1963 to 1972 I. INTRODUCTION . (ij A Public Savings Strategy 1. A sustained acceleration of the rate of growth of output generally requires an increase in the ratio of investment to aggregate income. When per capita incomes are low and financial markets inadequately developed, it is difficult to mobilize enough additional private savings to finance the desired increases in fixed investment. Consequently, as more nations turn to central planning techniques to direct growth-policies, there hag been a tendency to rely on the public sector to generate the increased savings to finance the higher investment ratio. There are countries where such a policy could be optimal because the government has easy access to the sources of income genera- tion e.g. countries where extraction and refinihg .of minerals contribute a major share of national income. But in many countries, a strategy of dependence on public savings may cause problems of misallocation and may even be counter- productive. 2. In essence, a public savings strategy relies on measures which involve some degree of coercion. Some of these measures generate additional savings (as is probably the case with additional taxation, although this is by no means certain). To some extent, government borrowing may also encourage additional private savings in the financial sector. But other measures may simply preempt the use of resources 'which have.already been mobilized in forms which are available to finance productive investment. The most glaring example of such preemption is usually the restraint placed on the use of funds raised by social insurance agencies; but various requirements on the portfolios of private financial institutions also have a similar preemptive effect. The monopoly of the government in issuing currency is another major means of preempting resources. To the extent that these preemptive means are used, the case for a public savings strategy rests on the presumption that public investment is more desirable than private investment (on efficiency or welfare criteria) and not on the grounds that the public sector is more efficient in mobilizing savings, Even if such a presumption is justified, it must be realized that a public savings strategy has some pitfalls. 3. A public savings strategy is often accompanied by an acceleration in the rate of inflation resulting from the failure of the public sector to realize over-optimistic targets and an excessive reliance on borrowing from the Central Bank to fill the gap. A rapid expansion of public borrowing from the banking system may mean either that the private sector gets a smaller share of bank credit or that the supply of money is made to expand faster than its demand, resulting in inflation. Inflation tends to erode the savings ability of the public sector and can cause significant misallocation of resources. The latter is especially true when development plans fail to allow for price changes in determining the allocation of output and investment, as is the case when physical input-output and commodity balance approaches dominate planning methodology. 4. Additional distortions in the use of resources are created when the government puts ceilings on interest rates (which reduce the debt-servicing burden on the budget). The interest rate structure then ceases to play the -2- major role in the allocation of credit, which is then determined by a variety of rationing devices, including administrative regulation. This, too, is sometimes counter-productive (e.g, the attempt to provide "cheap" credits may mean even more serious shortages of credit to important sectors). 5. The use of a public savings strategy rather than.a policy of borrowing private savings to finance public expenditure (as in Mexico), also neglects the potential of the capital market in mobilizing savings. The growth of financial institutions is stifled for the lack of an adequate range of earning assets. Mobilization of deposits by the banking system and the issue of long-term debt by other financial institutions are made difficult. Private savings are invested in forms which do not increase productive capacity (e.g. real estate, gold, etc. or foreign assets if exchange convertibility is maintained). Private fixed investment then has to rely excessively on funds generated within the enterprise itself -- a serious handicap to entrepreneurs who do not also happen to be rich. 6. Thus strategy with excessive emphasis on public savings may indeed hinder, not help, the mobilization of savings to finance increased investment in productive capacity. Some of the pitfalls are illustrated tn an examination of the savings strategy in Turkey's Third Plan. (ii) Sco eof the Study 7L This study is essentially a limited attempt to assess the reason- ableness of projections regarding the financing,of investment in Turkey's Third Five Yeari Development Plan (1973-1977). The starting point for the analysis is, therefore, the projections of investment and national savings. The rationale for the projected size or composition of investment and foreign capital flows are not examined. Moreover, the projection of capital inflow from abroad expresses the goal of increasing self-reliance of the economy rather than the availability of external asistance to Turkey. The focus of this study, therefore, is on the means by which national resources are mobilized to finance investment. 8. 1 Chapter 1I discusses the genesis of the savings targets and trends in national savings in the First and Second Plan periods. But our ability to assess the savings projections is severely limited by the availability of data. Data on private disposable income or on the contribution of public and private sectors to aggregate savings are either nont-existant or open to question, primarily because of the size and complexity of the public sector in Turkey. 9. The accounts of the various ministries of the central government (the "General Budget") and the budgets of annexed agencies are consolidated (and referred to as the "Consolidated Budget"); but in addition, there are local administrations, revolving funds and a large number of State Economic Enterprises (SEEs) some of which are very large. Consolidated accounts for the public sector as a whole are not available and are difficult to cunstruct with any reasonable degree of accuracy. 10. A critical review of trends and projections of public savings is attempted in Chapter II(iii), on the basis of available data, It is our view -3- that the data base for a similar review of private savings does not exist at present. The following two chapters disduss the main determinant of public savings in greater detail. Chapter V discusses the various sources of public borrowing-and attempts to integrate the financing projections of' the Third Plan. The monetary impact of public savings shortfalls and the link with inflation is discussed in Chapter VI. The concluding chapter examines the policy directions that should be emphasized if adequate resources are to be 'mobilized to meet the investment (and growth) targets of the Plan. II. NATIONAL SAVINGS (i) Determination of Targets 11. Plan Formulation-in Turkey: Turkey has now embarked on its Third Five-Year Development Plan which has been formulated for the period 1973-1977. The principal motivations underlying the strategy of planning in Turkey (in common with many others) can be stated as the establishment of an industrial base increasing per capita GNP and reducing dependence on foreign capital inflows. 12. The target rate of growth for the economy is determined primarily in relation to the contemporary rates of growth in other countries -- both developing and developed. Over the last decade, Turkey's real GNP has been growing at the rapid rate of 6.8 percent per year. But in per capita terms, this reduces to 4.2 percent because of the relatively high rateof growth of population. An acceleration of the rate of growth of GNP is clearly necessary to prevent a growing gap in the level of development between Turkey and members of the European Economic Community of which Turkey will be a full member by 1995. Hence the target growth rate of GNP during the Third Plan period is set at 7.8 percent compared with a target of 7 percent in the First and Second Plans. 13. One can expect that an acceleration in the economy's rate of growth, especially through a policy of industrialization, will lead to a rapid in- crease in the demand for-imports. Therefore, the Plan also emphasizes a policy of import substitution to permit the growth of essential imports and to reduce dependence on foreign capital inflows. 14. Having defined these basic objectives, the details of the Plan are worked out with the help of a multi-sector input-output model which ensures consistency between the various macro-economic magnitudes and the inter-sectoral production relationships. 1/ In particular, sectoral output targets are made consistent with projected demand patterns, external balance requirements and the limits for import substitution. The rate of growth of exports and the balance of payments deficit are both specified exogenously. The model 1/ Third Five-Year Development Plan: Growth Model and its Solution, State Planning Organization, 1972. -4- achieves external balance by increasing the level of import substitution. The sectoral composition of import substitution is determined by parameters calcu- lated by studies outside the model.0 1/ 15. The macro-economic section of the consistency model has the charac-, teristics of a Harrod-Domar model and, consequently, the planning strategy has a very strong orientafion towards increasing investment to accelerate the rate of growth, Investment requirements are calculated by applying incremental capital-output Tatios to the projected output expansions in each of the 37 sectors of the model. Consistency requires that aggregate investment should equal aggregate domestic and foreign savings. Both factor incomes from abroad and the deficit on current account of the balance of-payments are specified exogenously. Hence the model places the burden of adjustment to the savings- investment identity entirely on domestic savings. 16. Savings in the Growth Model, As noted in Chapter I, the availability of resources is the major constraint in the implementation of an investment oriented growth strategy. The increase in the ratio of investment to output required to accelerate the rate of growth, combined with an exogenously specified rate of inflow of foreign capital, implies the need ior a major effort to increase national savings in the economy. The consistency models guiding the formulation of the II and III Development Plans in Turkey assume that the "savings gap" will be met entirely by increases in public saving, Indeed, as will be demonstrated below, public savings are calculated as a residual in the final solution of the model. 17. Private consumption is a function of private disposable income (which is defined as GNP at market prices less government revenue). The rate of growth of public consumption is projected exogenously, Thus public savings can be increased directly with government revenue. The model asstmes that the public sector's marginal propensity to save is unity, greater than the private marginal propensity to save (estimated to be about 0.16). The transfer of income from the private to the public sector, therefore, rapidly increases the ratio of aggregate savings to income. 18. Government revenue is composed of tax and non-tax revenues of the consolidated budget and factor incomes (profit and depreciation) of State Economic Enterprises. The Plan model includes in government revenue an additional variable ("additional savings requirement for the public sector") whose sole purpose is to ensure balance between investment requirements and savings. Although it is recognized that this additional revenue will have to be generated by fiscal policy, no adjustment is made in the tax or expend- iture assumptions to eliminate the need for this variable. 1/ These steps are obviously among the more crucial ones in the detailed planning exercise. However, it is beyond the scope of this study to offer any critique of the assumptions involved. Further, the consistency" referred to is only mathematical. Some of the economic relationships which are ignored are mentioned below. -5- 19. The implications of alternative fiscal policies to fill the savings gap are not explored. In particular, it is assumed that government revenue is increased by a transfer of income from private to public sectors (rather than by indirect taxes which affect relative prices) and it is assumed that the income elasticity of private consumption demand is the same for each sector of domestic output. The allocation effects of fulfilling the public . savings targets are not fully explored. 20. The consistency model treats all variables at constant prices. The implication that inflation does not affect any of the estimated relationships is unrealistic. Both government revenues and expenditures are obviously affected by inflation (in the same direction though not to the same extelnt)9 thereby affecting public savings. Private.savings decision.may also be in- fluenced by price movements, especially when they lead to revision of expecta- tions of future inflation. Inflation may also affect savings indirectly through its effect on.other macro-economic magnitudes such as the balance of payments, the -real rate of interest, the distribution of income and sectoral output (to the extent that relative prices change or price elasticities differ between sectors). The neglect of these factors in projected agg:regate savings could be important when the Plan is to be implemented in an inflationary environment, as is the case in Turkey. 21. In summary:. the required rate of growth of investment is derived from the target growth rate of output; public savings are assumed to expand to the extent necessary to equate aggregate investment and savings in the economy. - 6 - Graph 11.1 Graph Il1-. Gross National Savings as Percent of GNp 23.0 22.0 21.0 20.0 19.0 18.0 17.0 16.0 15.o 14.0 Target 13.0 Actual 12.0 . 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 4.a -7- (ii) Plan Targets and Performance 22. The First and Second Development Plans (1963-1967 and 1968-1972) required gross fixed investment to grow at 11 percent per annum. Through the First Plan period and the first two years of the Second Plan, the rate of growth of investmnnt exceeded Plan targets, although the level , as a percent of CNP, were lowe:: than target because of overestimation of invest- ment in the base year. Investment rose steadily to 19.6 percent of GNP in 1969 (see Table 111) but failed to increase in 1970 and declined to 18.5 percent in 1971 primarily due to the deterioration in domestic savings performance. 23. Consistent with the objective of increased self-reliance, gross national savings were projected to grow faster than fixed investment in both the first two development plans. Savings targets were well surpassed in the First Plan period, the improvement in performance being sufficiently better than planned to overcome the handicap of a high base-year estimate. In the Second Plan period, however, savings performance has been well below Plan targets. The ratio of savings to GNP is showing signs of stagnation and is probably being sustained primarily because of the increases in net factor income from abroad, This aspect is discussed in Chapter II (iv). 24. The Third Plan envisages an increase in the*gross national savings ratio up to 25-.4 percent in 1977, from a base year estimate of 19.6 percent 'in 1972. This implies a marginal savings rate of 38 percent, which is much higher than has been achieved so far (Table II.2). Table 11.2: Marginal Savings Rate (Percent) Average Average Target 1963r67 1968 1969 1970 1971 1968-71 1973-17 27.2 24.8 23.4 21.4 23.5 23.1 38.0 The marginal savings ratio achieved in the First Plan was 27.2 percent (with a growth rate of 6.8 percent in real GNP) and the ratio for most of the Second Plan has been lower, at 23.1 percent (with a growth rate of 5.3 'per- cent in real GY2)'. Even allowing for the fact that the faster rate of growth of real GNP projected in the Third Plan should justify some increase in the marginal savings ratio, the aggregate savings projection seems unrealistic. 25. Table 11.3 presents the aggregate savings ratios that will be achieved in 1977 under alternative assumptions regarding the marginal savings rateso The assumption regarding the rate of growth of real GNP is maintained at 7.8 percent 1/ as stated in the Third Plan. Alternative 1 uses the his iso of course, inconsistent. Lower marginal savings ratios will lead to lower rates of growth of output (unless higher foreign capital inflows compensate), which will tend to reduce (slightly) the terminal average savings ratio below the levels shown in Table 11.3. - Table ;[I.1:. Investr.ent and Sav-ings 1962-1972 (percent of GNP) irst Plan Second. Plan kvera e 1962 -1963 1964 1965 1966 1967 1368 1969 1970 1971 1972 1963-67 1968-71 Gross Fized Investen.t Plan 16.3 1/ 17.0 17.9 18.3 19.0 19.4 19.2 20.1 21.2 22.h 22.9 18.3 20.7 Actual 13.7 15.2 14.9 15.4 16.7 17.2 18.6 19.6 19.5 18.1 n.a. 15.9 19.0 Gross National Savings Plan 12.3 1/ 12.9 13.8 14.9 15.8 16.5 18.6 19.5 20.6 22.0 22.6 lh.8 20.2 Actual 10.6 13.1 15.0 15.9 17.7 17.8 18.6 19.1 19.4 20.3 n.a. 15.9 19.4 1/ Base year estimates used in the formulation of the First Plan. Soure: First and Second Five Year Development Plans and State Planning Organizations estinates. Table 11.3: Alternative Savings Ratios Marginal Savings GNS/GNP Rat, 973-1977Z Percent. 1977 Third Plan TarRet 38.0 .25.5 Alternative 1 . 24.6 21.2 2 30.0 22.8 marginal savings ratio th&t was, in fact, achieved since the inception of the First Plan. It includes the rather high marginal savings ratios that were achieved in the first couple of years. Alternative 2 uses a much higher marginal savings ratio which.has been chosen so as to reflect the higher GNP growth rate target in the 'Third Plan. it is much higher than the margi- nal savings rate actually achieved in any of the last 10 years except 1964. 26. Both alternatives yield aggregate savings ratios in 1977 which are well below the Third Plan projections. On the basis of historical aggrega- tive evidence one may conclude that the aggregate savings targets are over optimistic. 27. . The pessimism generated by the review of historical evidence on marginal savings ratio is somewhat reinforced if one includes the effect of inflation on the rate of saving (which, as noted above, the Plan model does not do). 28. Economic theory does not make an unambigous prediction about how inflation will affect aggregate savings. A decision to save, whether made by an individual or.by a government, necessarily implies a postponement of consumption. The interaction between inflation and savings propensities is very complex, depending on, among other things, the extent to which inflationary expectations are realized and what the precise objectives of savings behavior are. We will not enter into a discussion of the many theoretical niceties involved, but attempt to establish empirically what effect inflation might have on aggregate savings behavior. 29. It would also be preferable to estimate savings functions separately for the public and private sectors since inflation would affect their revenues and expenditures differently. Again, this has not been possible since reliable data on public savings have not yet been constructed over this period. The observed negative impact of inflation on the average sav- ings ratio is some mixture of the impact on the public and.private savings ratios. 30. The crucial point in assessing aggregate savings targets is that the rationale for the savings projections is not purely historical. Although present savings ratios are satisfactory by comparison with other developing countries, additional savings are necessary to meet the growth targets of the Plan. As explained in the discussion of the Growth Model used in the -9- formulation of the Plan, the projections of public savings absorb all the disequilibria between projected investment needs and -the forecasts of savings in the absence of special additional effort to generate savings. 31. In Turkey, the negative effects of inflation seem to dominate, as demonstrated in equation (2.1). In this regression using annual data from 1962 to 1971, per capita.income has a positive influence on the aggregate savings ratios, significant at the 99 percent level; inflation has a negative effect, significant at the 90 percent level. During this period, the rate of inflation accelerated from an average of 5 percent per year during the First Plan period to over 15 percent per year in 1971 and 1972. A simple-minded extrapolation of.equation (2.1) suggests that, assuming per capita real income grows at planned rates, the savings target will be achieved only if the rate of inflation is held below 8 percent per year. RSA -14.85 + 0.015 PCYR - 16.11 P ..... (2.1) (s.e0) (5.8) (0.004) (8.3) R = 0.88; DW = 1.79; N = 10; where RSA = the ratio of Gross National Savings to Gross National Product at market prices PCYR = per capita real GNP P = GNP deflator 32. Equation (2.1) represents only a naive attempt to examine the influence of inflation on savings. Since it is anticipated inflation that affects savings decisions, one would have liked to model for it in some way, perhaps by assuming that expectations are built on the basis of recent realized rates of inflation. But since a consistent series of national accounts are available only for 10 years, such an exercise would be of doubtful significance. (iii) Public Savings 33. The public sector has a crucial role to play in meeting aggregate savings targets ,not only because of the way in which targets are formulated (discussed above) but also because it is an important part of Turkey's economy. The inception of central economic planning has given the public sector in even more important role in fixed investment. From 1953 to 1962, an average of 14 percent of GNP was invested, of which the public sector accounted for about a third. 1/ From 1963 to 1971, the fixed investment ratio was raised to over 17 percent and the public sector undertook more than half. An increasing proportion of the expanding investment program is being implemented through the SEEs, partly financed by transfers from the consolidated budget. 1/ This is based on the SPO investment series. Data from the State Institute of Statistics tend to attribute a higher percentage to the public sector. - 10 - 34. Since consolidated accounts of the public sector are not available, it is impossible to establish reliably how public investment was financed. This chapter will attempt to examine recent developments in aggregate public savings but, perhaps more than anywhere else in this study, it is .necessary to draw attention to the fragility of the data. The discussion of the major components of public revenues and expenditures, on which more reliable data are available, is left f -r the following two chapters. 35. The separate coutributions of the public and private sectors to gross national savings have not been published for Turkey. There has been an attempt by the State Planning Organization to develop preliminary estimates for the period 1963-1971. Alternative estimates can also be developed using available data on the revenues and expenditures of the consolidated budget and projects of SEEs. This has been done for the period 1968-1971. Both estimates are presented in Table 11.4. Table 114: Public Savings Estimates, 1968-1971 (TL billion in current prices and percent of GNP) 1968 1969 1970 1971 (TL billion) Estimate (a) Consolidated Budget /1 4.3 6.1 5.9 3.3 Local Administration 72 0.7 0.6 0.8 0.5 Revolving Funds /2 0.4 0.4 0.4 0.5 SEE Surplus /3 1.3 1.3 0.5 2.4 Total 6.7 8.4 7.6 6.7 Estimate (b) (SPO estimate) 10.6 11.7 12.8 12.9 Public Savings as Percent GNP (Percent of GNP) Estimate (a) 6.1 7.0 5.4 3.7 Estimate (b) 9.7 9.7 9.1 7.2 Second Plan Target 9.4 10.0 10.9 12.4 /1 Revenues less expenditures on goods and services, interest payments and unidentified transfers assumed to be current transfers. /2 These data are based on the assumption that their investment is financed without borrowing. /3 Profit plus depreciation less tax. 36. The size of public savings is consistently higher in Estimate (b) than in Estimate (a). Both estimates show that public savings have declined as a percent of GNP in 1970 and 1971. But they differ in the sharpness of - 11 - the decline. Since the preliminary official estimates are not supported by detail or by explanatory notes, it is not possible to point out the ' reasons for these discrepancies. It is highly probable, however, that the main reason is the different treatment of unidentified transfers from the consolidated budget. Table 11.5 shows how rapidly this particular item has grown. Table 11.5: Unidentified Transfers in Consolidated Budget Fiscal Year /1 1967 1968 1969 1970 1971 TL Billion 1.7 2.5 /2 2.2 /2 5.5 9.6 Percent of Consolidated Budget Revenue 10.1 13.3 10.1 20.2 26.3 Percent of GNP at Current Market Prices 1.7 2.3 1.8 3.9 5.4 /1 Beginning March 1. /2 Budget estimate. 37. Unidentified transfers from the consolidated budget now account for more than a quarter of central government revenues and 5.4 percent of GNP. National accounting conventions require that current transfer payments be deducted in estimating sectoral savings, but lack of information prevents identification of the current component of transfers which include some special payments to certain public agencies and enterprises, grants for* municipalities, payments to the Pension Fund. The entire unidentified por- tion of transfer expenditures is, therefore, treated as current transfers in calculating estimate (a) in Table 11.4. The possible inclusion of some capital items in this category may result in an underestimation of the actual level of public savings, though it probably does not affect their trend in the last two years. 38. All the available evidence suggests that public savings have been less than anticipated in recent years and have deteriorated further in 1970 and 1971. The savings performance of the public sector was satisfactory in the First Plan period - exceeding the Plan target in each of the years as well as on average (see Table 11.6). The Second Plan raised the targets by nearly half from 7.4 to 10.7 percent of GNP for the period 1968-71 (11.1 percent for the whole Plan period). Actual performance definitely fell short of target, probably by a very large margin. - 12 - Table 11.6: Public Savings Target (percent of GNP; average for period) 1963-67 1968-71 1973-77 Plan Target 7.4 10.7 12.5 Estimate (a) /1 n.a. 5.5 Estimate (b) (SPO) 7.8 8.9 /1 See Table 11.4 for details. 39. Even the SPO estimates suggest that the performance during the II Plan was only slightly better than in the I Plan; that the percent of GNP saved by the public sector in 1971 was less than in any of the previous six years; and that the percentage has not increased since 1968, Estimate (a) suggests an even more dismal picture. The average share of G'P? saved by the public sector was little over half the Plan targets for 1968-71 and only 30 percent in 1971. Public savings have declined even in absolute amounts (at current prices) and the sharp increase in transfers suggests a continued erosion of savings potential. 40. The Third Plan requires that public savings reach 14.1 percent of GNP by 1977 -- a large increase even over the unfulfilled targets of the II Plan. It is inconceivable that this target can be achieved unless there is a sharp reversal of the trends of the past few years. Inflation is one of the factors that make such a reversal extremely difficult to achieve. 41. There are'a number of reasons why public savings in Turkey are probably adversely affected by inflation. First, government revenues as a whole do not automatically keep pace with inflation. The elasticity of the tax structure is only slightly higher than unity. As discussed in Chapter III (i) and Annex A., most of the increase in tax revenues in the last decade has been the result of legislation. Non-tax revenues ane annexed budget revenues, which accounted for 14 percent of consolidated budget revenue in 1971, are not responsive to inflation. Some of these revenues really constitute withdrawals from various special funds. Excluding these special funds, the buoyancy of non-tax and annexed budget revenues with respect to aggregate income was only 0.7. In the absence of special legisla- tion, therefore, inflation probably has a negative effect on the real value of consolidated budget revenues. Secondly, the prices charged by State Economic Enterprises have tended to lag behind the general rate of price increases for a variety of reasons, while wage rates and employment have not been similarly restrained. As a result, inflation has tended to reduce the profitability of public enterprises thus reducing public savings (see Chapter IV. Thirdly, government expenditures are inevitably affected by price increases in the economy. The response of expenditures on goods and services in immediate and gradual; wages and salaries tend to rise at intervals, with both the magnitude and the frequency of the changes being influenced - 13 - by the pace of inflation. Social insurance and pension benefits must also be revised upwards to take account of inflation. In summary, inflation would tend to reduce public savings in Turkey, especially if they are valued at constant prices. (iv) Role of Factor Incomes from Abroad 42. One of the most significant economic phenomena in recent years in Turkey has been the growth of factor income from abroad. From very low levels throughout the sixties (with the exception of 1966)0 net factor income grew to as much as 1.6 percent of GNP in 1970 and 3 percent in 1971. The reason for this phenomenal growth has been the increase in the remittances by Turkish workers abroad. There are an estimated 600000 Turkish workers in Europe and the average remittance has also grown from US$105 during 1965-69 to US$273 in 1970 and US$471 in 1971. The doubling of the per capita remittances from 1969 to 1971 was primarily because of the additional incentive provided by the devaluation of the Turkish lira in August 1970,, when the effective rate of exchange for workers' remittances was raised from TL 12 to TL 15.per U.S. dollar. 43. An increase in net factor incomes from abroad results in divergent trends between gross domestic savings and gross national savings because of the accounting co, -entions underlying their definitions. 1/ The ratio of gross domestic savings to GDP (gds) is always less than the ratio of gross national savings to GNP (gns) as long as there is a net inflow.of factor income from abroad. The difference can be quite large if net factor income is a high proportion of GNP. 2/ 44. In most countries, aggregate consumption expenditures are estimated residually in the national accounts by type of expenditure. The gross national savings ratio correctly reflects the propensity to save out of national resources. The gross domestic savings ratio, however, is built on the assumption that all factor income is saved. Since the assumption is hardly realistic, the gross domestic savings ratio underestimates the proportion of domestic income (excluding net factor. income from abroad) that is saved. The extent of underestimation obviously increases as net factor income from abroad rises. 45. It can also be shown that as factor incomes from abroad become an increasing proportion of GNP, the gross national savings ratio will itself rise, provided only that the marginal propensities to save out of factor income and domestic income are each higher than the gross national savings ratio. I/ Gross domestic savings are defined as GDP minus consumption; gross national savings as GNP minus consumption. They differ precisely by the amount of net factor income from abroad. 2/ See Annex B for the derivation of this and other results regarding savings ratios stated in this section. -14 46. Perhaps -the increase in the gross national savings ratio in 1971 is simply the result of the above factors. 1/ There is no direct evidence to support the hypothesis that the rate of savings from factor income is higher than that from domestic income in general. But it does appear to be a reasonable supposition: First, workers' remittances have proved to be remarkably sensititve to the exchange rate, suggesting that they constitute the transfer of savings from abroad rather than a transfer to finance the consumption expenditure of relatives living in Turkey.. If the latter motivation dominated, the devaluation of the Turkish lira should have had a significant "income effect", perhaps leading to a decrease in per capita remittances. Seconds there has been a phenomenal growth of privately held time deposits at commercial banks during this period (from TL 5.8 billion at end 1969 to TL 12.3 billion at end 1971). The timing of this increase appears to be related to the growth of workers' remittances, but it could also be due to the increase in deposit interest rates which took effect at about the same time. Thirdly, the evidence is that the public savings rate has been falling in 1970 and 1971 (see Chapter II (iii)) which implies that the private savings rate has increased substantially. One contributory influence might be that factor income from abroad accrues almost entirely to the private sector. 47. All these arguments tend to support the position that the savings ratio has been sustained in recent years because of the growth of factor incomes from abroad -- ironically, a development that was completely unfore- seen in the Second Plan. Table iI.7: Factor Income, National and Domestic Savings, 1968-1971 Average 1968 1969 1970 1971 1968-71 Net Factor Income from Abroad Plan (% GNP) 0.7 0.7 0.6 0.6 0.7 Actual (% GNP) 0.2 0.3 1.6 3.0 1.3 Gross National Savings Ratio (% GNP) 18.6 19.1 19.4 20.3 Gross Domestic Savings Ratio (% GDP) 18.4 18.8 18.1 16.7 48. The Third Plan is rather conservative in assuming that net factor income from abroad will grow from TL 4.9 billion in 1972 to TL 5.8 billion in 1977, at an annual growth rate of 3.3 percent. In fact, the actual in- flow in 1972 has exceeded the 1977 target and there will probably be some further growth in the next few years. The conservatism of the Plan is, of I/ If sb, we should expect another increase in the gng ratio in 1972, when factor incomes from abroad again increased substantially. - 15 - course, justified in view of the many uncertainties surrounding workers' remittances -- both as regards the demand for Turkish workers in Europe and the per capita remittance home by workers abroad. To the extent that the inflow of factor incomes is higher than forecast in the Plan, it will help . in raising the ratio of gross national savings to GNP, possibly by as much as . one percentage point. IllO GOVERNMENT REVENUE AND EXPENDITURE (i) Trends in Tax Revenues 49. Current receipts of the Consolidated Budget and surpluses of the State Economic Enterprises constitute about 90 percent of public revenues. 1i The balance consists of rev6lving funds and local administrations whose revenues are small and partly,derived as a percentage of general budget taxes. This section discusses the consolidated budget; State Economic Enterprises will be treated separately in the next Chapter. 50. The decade of planning has seen a very substantial increase in the tax effort in Turkey. The ratio of general budget taxes to'GNF has risen from 12.9 percent in 1962 to 17.5 percent in 1971. A comparison with 8 other countries which.have approximately the same per capita income shows that this ratio is definitely above average. 2/ 51. Tax performance during the First Plan fell considerably short of Plan targets, both in terms of rate of growth and share of GNP (see Table III.1), Tdbl,1Pl.1: Ta 'Pla anPe6trhaiftce, /1 Percent of GNP-- Annual Rates of Growth 1963-67 1968-71 1963-67 1968-71 Plan Targets Taxes 16.8 17.9 10.8 9.0 Taxes and Additional Revenue/2 17.9 19.1 12.6 10.00 Actuals Taxes 13.7 15.9 6.6 10.4 /1 Average of annual percentages. /2 Public savings gap presumed to be filled by fiscal means (see Chapter II (i)).. 1/ Some items are classed as revenue in official documents but must cor- rectly be regarded as borrowing. They are discussed in Chapter IV. 2/ These countries are Colombia, Fiji, Guatemala, Guyana, Honduras, Iraq, Ivory Coast and Malaysia. The average tax ratio is 14.2 percent in these countries. - 16 - The share of GNP mobilized by taxes was lower than tatget also during the Second Plan, especially if one considers "additional revenues". As pointed out in the discussion of how savings targets are determined (Chapter II (i)), the Plan assumes that the gap between the investment target and pro- jected savings will be met by fiscal measures. Tax performance must, there- fore, be judged against-this criterion. 52. Unlike the First Plan, however, the rate of growth of tax revenues (at-coist-ant prices) in the Second Plan has been higher than targets, averaging 10.4 percent per year, in the first four years. During this period, taxes absorbed as much as 24.7 percent of the increase in real output. 53. Some of the increase is the result of the automatic response of taxes to increases in ificome. The strength of this response has been care- fully studied, by individual taxes, and the conclusion is that the elasticity of the tax system in Turkey is only slightly greater than unity (see Annex A). The observed increase in the ratio of taxes to GNP is motly the result of the introduction of new taxes and changes in the structure of existing taxes. Changes in the tax structure accounted for 78 percent of the in- crease in tax revenues from 1963 to 1971, the automatic response contribut- ing only 22 percent. (ii) Structure of TAx Revenues, 54. The relative shares of direct and indirect taxes has changed little over the period 1963 to 1971 (see Table III.2). 'There has been a slight increase in the direct tax share, especially in recent years. Most of this is accounted for by the income tax, largely because of the revenue effects of inflation in a progressive tax structure. The decrease in taxes on foreign trade is to be expected when import substitution is taking place, and would have been greater but for increases in the rates of ciistoms duties on imports. Domestic production taxes do not compensate fully for the decrease in import duties. Table 111.2: Structure of Tax Revenues (Percent of General Budget Taxes) 1963 1967 1971 Direct Taxes 34 36 40 Indirect Taxes 66 64. 60 (Domestic) (37) (37) (37) (Foreign Trade) (29) (27) (23) Total 100. 100 100 -17 - 55. Property taxes contribute little to total revenue, primarily be- cause of outdated valuation. More recently, an attempt has been made to approximate current market value by introducing self-assessment - a system with obvious handicaps. Reform of the capital gains tax and corporation tax systems are also under way but are unlikely to provide much additional revenue, 56. Domestic indirect taxes consist largely of production taxes (pri- marily on petroleum), a heavy tax on banking and insurance transactions and revenues from fiscal monopolies (primarily in tobacco and alcohol). There is further scope for broadening the coverage of the production tax. This may be one of the principal sources of additionl tax revenue in the future, dspecially as import duties will probably decline with increasing integration with the European Economic Community. 57. A major new.tax introduced in 1972 is the "Fiscal Balance Tax" which replaces the compulsory savings bonds scheme introduced in 1961, levied .at 3 percent of personal and corporate income, and contributing 0.6 percent of GNP. While this is a substantial amount of revenue, the tax lacks progressivity and is levied on gross income, which make it incompatible with ability-to-pay principles. 58. The present income-tax system is deficient in its taxation of the income from agriculture and small businesses. It is estimated that only 1 percent of farm units pay income tax and agriculture's contribution to tax revenue is insignificant. The use of a lump-sum payment system and special exemptions result in reduced taxation of small businessmen. 59. These and other problems are the subjects of reforms in the tax system, some of which are already being implemented. But with few excep- tions, the revenue implications of these reforms are unlikely to be large, at least in the next few years. Indeed, the most promising of the efforts to increase the tax ratio is the attention being paid to tax administration and the effective collection of taxes. The organizational structure has been improved, additional staff has been recruited and new legislation is pending. (iii) Third Plan Tax Projections The Third Plan anticipates that tax revenues on the basis of the 1972 tax structure will rise to 24 percent of GDP in 1977. On the basis of tax revenue functions designed to project revenues from a constant tax structure, we estimate that the Plan projections are slightly over-estimated. 1/ I/ As Annex A points out, these estimates are subject to many methodological problems. However, a projection based on a simple extrapolation of past trends is definitely biased since it is based on the assumption that tax structures will be changed in the future .as they have been changed in the past. Such a simple projection yields an estimated tax ratio of 23.8 percent in 1977. - 18 - Table ITI.3: Consolidated Budget Tax Revenue Projection, 1972-1977 (in 1971 prices) /1 /2 Plan Projection Projection A- Projection D- TL Bil- Percent TL Bil- Percent TL Bil- Percent lion of GDP lion of GDP lion of GDP /3 1972 38.Y-- 20.6 35.2 18.9 38.3 20.6 1973 43.3 21.6 38.7 19.3 42.1 21.0 1974 48.4 22.3 42.5 19.6 46.2 21.3 1975 54.1 23.1 47.0 20.0 51.1 21.8 1976 60.5 23.9 52.0 20.6 56.7 22.4 1977 67.7 24.8 58.0 21.2 6321 23.1 Plan Period (1973-1977) 274.0 238.2 259.3 Growth Rate During Plan Period (percent per year) 12.1 10.5 10.5 /1 See Annex for the discussion in methodology. /2 Assumes the same growth rate as Projection A, but the revised official estimate for 1972 is used as the base. /3 Revised official estimate as of August 1972. 61. The Third Plan tax projections are based on the 1972 Uax structure, modified to include the tax changes pertaining to imports covered by the EEC Protocol and the burden of proposed incentive schemes. Projections A and B are both derived from the tax revenue forecasting functions described in the Annex; the difference lies in the choice of base year. -Projection A projects on the 197.1 base; Projection B accepts the official estimate of 1972 revenue and uses that as the base for subsequent forecasts. I/ 62. Both Projections A and B yield an annual average growth rate of tax revenues of 10.5 percent per year during the Third Plan period. This rate of growth is the same as was achieved in the first four years of the Second Plan with the aid of numerous and drastic changes in tax laws. Since Projections A and B are based on a constant tax structure, this suggests that the government has been remarkably successful in raising the elar.city of the tax system during the Second Plan. 1/ An assumption implicit in Projection B is that the additional revenues from the new 1972 tax measures will be proportional to the revenue generated by the 1971 tax structure in every subsequent year. -19- 63. An annual growth rate of taxes of 10.5 percent in real terms is a very creditable achievement. But it is less than the Plan requirement of a 12.1 percent growth rate. The resulting tax ratio for 1977, therefore, is lower than the Plan projection especially oh the basis of Projection A. 64. From the point of view of mobilizing.adequate resources to finance investment needs,. the relevant measure of tax performance is not the rate of growth of taxes but the increase in the ratio of tax revenue to projected income. To .achieve a tax ratio of 24.8 percent by 1977, the ratio must increase by 4.2-5.9 percentage points (depending on what the 1972 outturn actually is). On the basis of the constant structure tax revenue projections shown in Table 111.3, it appears that about half of this increase will have to be raised through new tax measures. This will require a revenue raising effort of the same order of magnitude as was achieved during the Second Plan. 3ut incremental .revenues become increasingly hard to raise when the tax ratio is already rather high. Raising the tax ratio from 19 percent in 1972 to 24.8 percent of GDP in 1977 will mean that one-third of incremental GDP wili have to be absorbed in additional taxes. Not only is this a difficult task, it could also lead to undesirable consequences. The effect of such an addition to the tax burden (unless imposed on sections.of the economy that are currently very lightly taxed) might be to damage incentives in the private sector and make tax administration even more difficult than it now is. (iv) Public Consumption 65. Until 1969, the rate of growth of public consumption was below the rate forecast in the First and Second Plans. The share of public consumption in GiP was fairly steady around 12 percent. But the picture changed dramatic- ally in 1970 and 1971, when "real" public consumption rose by 8.8 percent and 31.0 percent, respectively. Over the first 4 years of.the II Plan, the real rate of growth of public consumption has exceeded the Plan targets by 43 percent. Table III4: Public Consumption (Percent of GNP) Rate of Grdwth (%) 1963 1969 1970 1971 1963-67 1968-71 Plan Target 15.2 14.3 14.4 14.9 8.6 8.8 Actual /1 11.6 12.2 12.6 15.1 7.4 12.6 /1 As estimated in national income accounts, State Planning Organization. 66. A detailed analysis of public consumption is not ppasible in the absence .of consolidated public sector accounts. But some.insights can be gained by the study of the central government's consolidated budget, which accounts for most of public consumption. -20- Table 111.5: Composition of Consolidated Budget Expenditure (percent) 1963-67 1968-1971 Current 51.6 47.3 Transfers 21.4 30.8 Investment 26.9 21.9 TOTAL 100.0 100.0 67. Tabie III.5 shows that the expansion of consolidated budget expenditures has been primarily in "transfers" at the expense of both investment and expenditures on goods and services. We have already had occasion to comment on the growing importance of this item and the large proportion of it that remains unidentified (see paragraph 38). A comparison of national account data and consolidated budget data suggests that some part of consolidated-budget transfers do enter the public consumption estimates. But more detailed information is necessary to form a judgment about past trends and future prospects. 68. A comprehensive revision of civil service salaries and pensions contributed to the very rapid increase in current expenditures (and public consumption) in 1971. But that is not the whole story. The share of wages and salaries in the budget for current expenditures changed very little from 1970 to 1972, showing that expenditures on goods and services were expected to rise just as fast. 69. The Third Plan targets a rate of growth of public consumption at 8 percent per year in real terms. This is lower than the rate of increase allowed for in the First and Second Plans and corresponds to the actual rate of increase from 1968 to 1970, before the Personnel Law. The conditions for the Third Plan target to be achieved would appear to be that there is no major revision of civil servant'salaries (an unlikely situation unless present inflationary pressures are eliminated), and there is only a restrained expansion in economic and social services, IV. STATE ECONOMIC ENTERPRISES (i) Introduction 70. The evolution of the present structure of the State Economic Enterprise (SEE) sector'can easily be traced' to the enunciation of the principle of t'etatism" in the 1930's. The desire to eliminate foreign -21- domination of the economic scene and the absence of a domestic entrepreneu- rial or capitalist class led to the establishment of economic enterprises fully owned by the State. La 3460 of 1938 explicitly provided that SEEs could ultimately pass into private ownership.. That possibility now appears remote and direct State participation in economic activity has grown enormously. 71. The public enterprise sector now consists of over 100 enterprises, spanning a number of major sectors, contributing over 10 percent of value added in the economy and employing 6 percent of the non-agricultural labor force. For national planning purposes, attention is concentrated on 35 major enterprises which account for most of the activity in this sector. Eight of these are "financial SEEs," consisting of social security institu- tions, the State Investment Bank and financial institutions set up primarily to ensure an adequate supply of credit to agriculture, housing, small artisans., cooperatives and municipalities. The latter also engage in ordinary banking activities and, as "special law banks", are.treated as a part of the banking system. The non-financial "operational SEEs" operate in a variety of sectors ranging form steel and mining to textiles and fish. The public sector has a virtual monopoly in electricity, coal, petroleum refining, steel, nitrogen, paper, railways and air transport controls most shipping and communications and has a big share in textiles, cement, coal, sugar, machinery and chemicals. One SEE, the Soil Products Office, is the residual purchaser of cereals at official support prices. However, similar agencies that deal with tea aid tobacco are not classed as SEEs but as "annexed budget agencies" and are hence included in the consolidated budget. 72. It is particularly important to recognize that in Turkey, SEEs have been seen as an instrument to achieve the goals of economic development in all its aspects. A clear indication of the central role of SEEs in the planning process is the steady and substantial expansion of fixed investment in this sector, at an a'verage annual rate of about 20 percent per annum. Since 1964, fixed investment by SEEs has-grown from 28 percent to 40 percent of public fixed investment and from 13 to 18 percent of all fixed investment in the economy. Almost all the increase in expenditures on gross fixed investment represents addition to capital stock rather than compensation for depreciation (Table IV.1). Table IV.1: Investment by Operational SEEs 1963-67 1968-71 1973-77 TL million at current prices Gross fixed investment 9,310 19,737 80,137-' Depreciation allowance 3,711 5,450 Difference/2 5,599 14,287 Percent of.GNP (ross fixed investment 2.4 3.6 6.7 Depreciation allowance 0.9 1.0 Difference/2 1.5 2.6 /1 At 1971 prices. /2 Approximate measure of addition to fixed capital stock. - 22 - (ii) SEE Profits 73. The increase in investment in the SEE sector was accompanied by the hope that public enterprises would generate larger surpluses to finance future public investment. The First Plan envisaged that SEE surpluses would increase from an estimated 1.0 percent of GNP in 1962 (the base year) to 1.7 percent of GNP in 1967; the Second Plan envisaged a further expansion to 6.5 percent of GNP by 1971. As shown in Table IV.2, these hopes have not been realized, and S EE surplus had grown to only 1.3 percent of GNP in 1971 despite the large investments in the past decade. Table IV.29 .EE Surplus - Plan Targets and Performance (Percent of GNP) 1963-1967 1968-1971 Plan target 1.5 5.8 Actual 1.0 1.0 /1 "Surplus" is defined as profits plus depreciation le3s direct taxes, which corresponds.to the definition of the SEE contribution to public factor incomes in the Third Plan model. 74. The average for 1968-1971 is depressed by the very poor perform- ance in 1970, when devaluation raised costs and a compensating increase-in prices was only granted in 1971. While the recovery in 1971 is very hearten- ing, it must be noted that the increase is not that impressive when allowance is made for general price inflation (see Graph IV1). - 23 - Grah IV.1: SEE Surplus, 1963 to 197. TL rnillion 25007 2000 at current prices 1500> at bonstant 1968 prices 100 1963 .1964 1965 1966 1967 1968 1969 1970 1971 /1 SEE surplus is defined as profits plus depreciation léss direct taxes. -24- 75. The target rate of growth for the SEE surplus was 18 percent per year during the First Plan and 13 percent per year during the Second Plan (at constant prices). Consolidated sector data are not very reliable for the early years of this period, but it appears that the First Plan target was met. During the period 1968-1971, the actual rate of growth has been only 6.8 percent per year (at constant prices). 76. In general, the profits generated by SEEs have not been commen- surate with their volume of operations or the amount of capital resources used by them; and the growth in profits since 1963 does not seem commensurate with the large investment program in this sector. Most of the "surplus" is accounted for by depreciation provisions which are strictly,governed by law. They amounted to as much as 86 percent of gross profits from 1964 to 1971 (excluding 1970 when depreciation was three times as high as the surplus because of operating losses). 77. The rate of return for SEEs (net profit before tax as a percentage of the book value of net fixed assets) was only 2.7 in 1969 and only 0.9 percent during 1969-1971. The profit performance is even less satisfactory if one takes account of the cheap long-term financing available to them. In 1969, for example, te interest payments by SEEs were TL 1.01 billion on an outstanding debt of TL 25.66 billion, an average rate of only 3.9 percent. In 1971, long-term investment credits from the State Investment Bank were made available at 9.95 percent, while commercial bank credits cost around 20 percent. Another major source of finance for SEEs are contributions from the Budget. In the past three years about two-thirds of these contri- butions have been in the form of subsidies and -the rest as additional capital participation. In neither case does this place any additional burden of payment on the SEEs, although the latter adds to the net worth of the enterprise. 78. The interest rate structure in Turkey is very strictly regulated and,it is, therefore, impossible to assess precisely what the consequences would be if SEEs had to pay a "commercial" rate of interest. But the impact would clearly be very substantial. For instance, if SEEs had to pay interest at 15 percent on all outstanding long-term credits from the State Investment Bank alone, the additional burden would be of the order of TL 1 billion in 1971, leading to a net deficit of about TL 400 million instead of a net profit af.ter tax of TL 570 million. -25- /1- Table XV.3: Net Profits of Operational SEEs-- (TL million) 1966 1967 1968 1969 1970 1971 Agriculture 115 40 73 169 31 32 Mining and Power 273 453 381 460 380 517 Coal, Petroleum, Steel 34 256 296 412 455 707 Manufacturing 174 288 352 250 -26 259 Transport & Communication -147 -379 -49 -768 -1,458 -702 Total 449 658 653 523 -617 812 /1 Profits after depreciation and before direct taxes. 79. The huge operating deficits of the State Railways dominate the profit aggregates of the non-financial SEEs. During 1966-71-the railways made an average annual deficit (before tax) of TL 611 million while all other SEEs combined, made a iet profit of TL 1,024 million, Other non-financial StEs that also had substantial deficits in recent years are Maritime Bank, Coal, PTT, Nitrogen and Petrochemicals. 80. SEEs in the manufacturing sector have shown fairly stable, but low, profits. The profit performance of the agricultural sector tends to be erratic because of the inclusion of the Soil Products Office. The growth of profits in other sectors is largely the result of large, new investments. Table IV.4 attempts to take account of this factor by presenting net profits before taxes as a proportion of the estimated book value of net fixed assets. Relatively few SEEs earned rates of return in excess of 10 percent. These were sugar, Etibank, Electricity, Petroleum and Iron and St&el. All the other enterprises tended to make small losses (Tourism Bank, Milk, Meat and Fish) or small profits (Sumerbank, Machinery and Chemicals, Airlines, etc.) during this period. Low profit levels have led to substantial and rising level of subsidies. Transfers from the General Budget to SEEs which were at an annual rate of TL 600 million from 1963 to 1967 rose sharply thereafter to TL 1,476 million in 1969 and TL 3,530 million in 1971. - 26 - /1 Table IV.4: Rate of Return of Operational SEEs- (Percent) 1969 1970 1971 Agriculture 13 2 2 Mining and Power 16 8 8 Coal, Petroleum, Steel 10 9 9 Manufacturing 6 -1 5 Transport & Communication -12 -20 -9 Operational SEEs 2.7 -2.6 2.8 /1 Profits after depreciation and before direct taxes as a percentage of estimated book value of next fixed assets. This is an unsatis- factory measure but is the best available and gives some orders of magnitude. (iii) Socio-Economic Goals 81. Profits are an imperfect indicator of the performance of SEEs in Turkey since they are burdened with numerous conflicting goals. In determin- ing an appropriate pricing policy and in planning operations, there is inevitably a conflict between the need .to generate surpluses for new invest- ment and the need to further other goals of economic development -- influence the level of consumption of some goods, redistribute income, increase employment, encourage the development of backward regions, etc. In an inflationary environment, when the government is exhorting producers not to raise prices, SEEs may also be required to set a good example. The relative importance of these objectives has varied considerably in the last couple of decades in Turkey. Before 1960, the socio-political pressures to keep SEE prices low and to subsidize agricultural incomes led to considerable expan- tion in bank credit to SEEs. This is regarded as the major'reason for the rapid inflation which averaged 15 percent per annum from 1954 to 1959. Then the First Plan explicitly stated that prices ought to be determined on the basis of production costs but recognized that exceptions may be made to satisfy redistributive goals. The Second Plan deemphasized the role of costs and added the protection of price stability as one of the objectives of SEE pricing policy. This policy led to a delay of nearly a year before SEE prices responded to the substantially increased costs arising from the devaluation of August 1970, and was largely responsible for the deficit in 1970. 82. The Third Plan clearly states that the major problems facing SEEs are the need to improve efficiency and to adopt a pricing policy respon- sive to market conditions. There is an obvious correlation between the policy of price-rigidity during the Second Plan -period and the stagnation of SEE surpluses (in real terms) since 1966. Whether higher prices would - 27 - lead to higher profits is a question that can be answered satisfactorily only by examining individual enterprises, taking account of demand elasticities and other market conditions. This is beyond the scope of this study. By way of illustration, however, one may refer to the findings of an IBRD appraisal team regarding the State Railways. The team found that the rail- ways were operating under a tariff structure that bore little relation to costs. In particular, tariffs were low for the long distance transportation of ores, where road transport is not competitive and the elasticity of demand is probably low. Since ores constitute about a third of railway traffic (in ton/km.), and are expected to rise to more than half in a few years, the financial consequences of this tariff policy are considerable. While it must be recognized that the subject needs greater study, preliminary indica- tions are that tariff revisions (and*the consequent change in traffic pat- terns) could bring about a very substantial transformation in the finances of the State Railways. (The magnitude of this effect could be in the region of TL 0.5 billion a y.ear.) 83. When development objectives override narrow considerations of economic efficiency, public enterprises may enter areas of activity the private sector does not consider profitable, provide more employment and pay higher wages to unskilled labor, and undertake a variety of policies that reduce its profitability. As in the determination of an appropriate pricing, policy, the issue here is one of separating efficiency from other economic goals. Law 440 of 1964 provided that if SEEs were required to sell their output at an uneconomic low price, they should be compensated from the general budget. If comprehensively implemented, such a system would enable assessment of the operational efficiency of enterprises without sacrificing the achievement of their development goals. In fact, this has not been achieved for two reasons: (a) the compensation does not provide for any "normal profit" in addition to incurred costs; (b) the law restricts the system to "basic" commodities without being specific as to what they are. In fact, there were no instances of the compensation scheme being implemented in 1969. In 1970, payments amounting to TL 376 million were made to 10 SEEs (the principal recipients being Nitrogen, Machinery and Chemicals, Railways, Sumerbank and Meat and Fish). 84. The implementation of a compensatory scheme would improve the profitability of SEEs only at the cost of an explicit additional burden on the general budget. Similarly, raising prices of some SEEs'would raise input costs of other SEEs, and would not benefit the sector as much as they did the individual SEE. Nevertheless, there would probably be some net gain in the resource position of the public sector arising out of improved con- trol on the operational efficiency of the enterprises. 85. Apart from pricing policy, there are many other factors influencing the resource generation performance of SEEs which are widely recognized and have been acted upon. Following an intensive study in.1960, a standardized accounting system has been devised and is now being introduced; decisions on participations in equity have been centralized to enable better control - 28 - of the use of share capital. But many ills remain: the high turnover of management personnel, the shortages of skilled labor, the interference of the ministries in management decisions and the absence of an adequate system for assessing the operational efficiency of individual enterprises. (iv) The Third Plan 86. The Third Plan explicitly recognizes all the problems discussed above. The need to separate the goal of efficient operation from other socio-economic goals is clearly stated. Monopoly public enterprises are directed to set prices so as to iake reasonable profits; competitive SEEs are directed to use market prices. The Plan.promises budget subsidies to compensate for losses incurred in meeting the objective of optimum capacity utilization or other social objectives mentioned above. 87. The projections for the Third Plan period, however, do not reflect the magnitude of the problems facing SEEs. The Third Plan requires public sector factor incomes to grow at an average rate of 15.3 percent per year. Most of this growth is expected from operational SEEs, whose factor incomes- are expected to grow at 19 percent per year (at constant prices). 88. The proposed reforms of SEEs, however, tend to focus more on the administrative regrouping of enterprises and an attempt to free management from -political interference. The impact of such reform on the operational efficiency and resource generation by SEEs will be slow and uncertain. 89. On the basis of past performance, the content of prospective reforms and the extreme difficulty in bringing rapid change to the SEE sector, one must be skeptical that such a rapid increase in SEE profits will in fact be achieved. V. PUBLIC SECTOR BORROWING (i) Introduction 90. The analysis of the growth of public debt in Turkey is hampered by the fact that official documents relating to public sector accounts do not always make a clear distinction between "revenues" and borrowing. 91. This problem is illustrated by Table V.1 which shows that planned public expenditure was to be financed entirely out of revenues in the Second Plan, But many of the "revenue" items in fact represent an increase in the public sector's liability for future payment and should properly be treated as borrowing. Counterpart funds are treated as revenue instead of being 1/ SEE factor income is defined as profits including depreciation, after direct taxes. -29- treated as debt. In addition to those explicitly shown, some counterpart funds (generated by project credits) are also included in non-tax revenues. Saving bonds are ten-year bonds which are repayable at maturity. The "own resources" of SEEs are interpreted in the Plan as net profit plus depreciation; but the annual programs include all cash flows other than long-term investment credits and budget subsidies. In addition, SEE .debt is periodically "consolidated" and taken up by the Treasury, when it drops out of the usual reckoning of public . debt. The last major consolidati6n in 1960 involved a total of TL 5.2 billion. Table VJl: Second Plan Public Sector Balance, Targets, 1967-72 (percent of GNP) 1967 1968 1969 1970 1971 1972 Public Expenditure 27.4 27.9 28.3 29.3 .30.8 31.3 Tax Revenue 17.5 17.3 17.8 18.1 18.5 19.1 Non-tax Revenue 1.5 1.4 1.3 1.3 1.2 1.3 SEEs and Rev. Funds 4.9 5.5 5.4 5.9 6.5 6.4 Counterpart Funds 2.2 2.1 2.1 1.9 1.7 1.4 Savings Bonds 0.8 0.9 0.9 0.9 1.0 1.0 Total "normal" revenues 27.0 27.2 27.5. 28,1 29.0 29.2 "Additional" revenues 0.4 0.7 0.8 1.2 1.8 2.1 Total 27.4 27.9 28.3 29.3 30.8 31.3 Source: Second Five-Year Plan, State Planning Organization. 92. While these complications make it impossible to conduct an inte- grated, comprehensive discussion of public debt policies in Turkey, the dis- cussion of the main elements of public debt can still be useful. This is tne approach adopted in the following sections. (ii) Components of Public Borrowing 93. The main sources of public borrowing in the last decade have been compulsory savings bonds, social insurance and pension funds, Government bonds and banking system credits. The most striking feature is that very little of this borrowing can be said to represent a-"voluntary" increase resulting from the willingness of lenders to hold larger amounts of the financial liabilities of the government. The various sources all involve varying degrees of statutory requirement and moral suasion, which will be discussed in greater detail below. - 30 - 94. It is also remarkable that the sum of the main elements of public borrowing has increased very little as a share of GNP, despite the increase in the public investment program (see Table V.2). This highlights the fact that the public sector has not been able to mobilize private savings by offering financing instruments that are genuinely attractive to individuals and in- stitutions. 95. There has, however, been a significant change in the relative im- portance of the main elements of public borrowing. The social insurance sys- tem was a major source of funds around 1968 but then declined in importance. The public sector then turned to the banking system, primarily to the Central Bank, to finance the public investment program. (iii) Savings Bonds 96. Introduced in 1961, the compulsory savings bonds scheme required that 3 percent of all income and profits liable to Income or Corporation Tax should be used to purchase government savings bonds. The bonds had a maturity of 10 years and paid iiterest at 6 percent. In 1967, the coverage of this scheme was restricted to those earning incomes above TL 14,400 per year and in 1972, the saving .bond scheme has been replaced by an equivalent rate of tax. 97. The sale of savings bonds exceeded 0.7 percent of GNP from incep- . tion and rose to 0.9 percent of GNP until the reduction in. coverage in 1967. This measure reduced the income from bonds by a third -- from 5.6 percent of consolidated budget revenues to 3.7 percent and from 0.9 percent to 0.6 per- cent of GNP. Table V.2: l -illion and percent of GNP) (Net increase in amounts outstanding at end-year) 190 1966 1967 1968 1969 1970 1971 TL TL TL TL TL TL -- TL Compulsory Saving Bonds 637 0.9 770 0.9 910 0.9 625 0.6 731 06 882 0.6 1,18- 0.7 SIB Bonds held. by Social Security and Pension Funds 840 1.1 725 0.8 11031 1.1 1,190 1.1 13184 1.0 1,005 0.7 816 0.4 Treasury Bonds 117 0.2 295 0.3 597 0.6 384 0.4 365 0.3 44o 0.3 .420 0.2 Banking System Credit 1,110 1.4 1 55 1.2 1 1.2 638. 0.6 1780 1.5 1 0.9 42702 2.6 Total 2,704 3.6 22845 3.3 3,732 3.8 2837 2.7 4 3.4' . 2.6 7,123 3.9- -32- 98. The savings bond scheme was in fact identical to a tax for many individuals at lower income levels. Those who could not postpone their con- sumption needs chose to sell their bonds at discounts of up to 40 percent. Since the sale was made to other private individuals, the government still had the obligation to redeem these bonds. 1/ Subsequently, trading in sav- ings bonds was bannod and the raising of the,miniftum limit in 1967 must have reduced the need for distress sales. (At present, there is still a small . market in savings bond coupons.) 99. These measures reduced some of the inequities of the savings bond scheme. But, in essence, the savings bond scheme was as coercive as a tax without having the merits of a tax (from the government's point of view). The replacement of the savings bond scheme by the fiscal balance tax Vas, therefore, a logical step. (iv) State Investment Bank 100. The State Investment Bank (SIB) was established in 1964 to serve as a major means of financing the investment program of SEEs. Apart from the initial capitalization of TL 1 billion, the SIB was also authorized to sell bonds. 101L The sale of SIB bonds, however, has been excessively dependent on the Social Security Institution and the Government Pension Fund. Both these institutions undertake some real investments -on their own account but they are required to-invest the bulk,of their surplus in SIB bonds. In 1971, they held 97 percent of the outstanding bonds issued by the SIB and its predecessor, the Amortization and Credit Fund. Expansion of SIB activities, therefore, has been entirely dependent on the fortunes of these social in- surance institutions. Until 1969, the funds provided each year rose to a peak of .TL 1.4 billion. This increase was primarily the result of expand- ing coverage of social insurance and the corresponding increase in premium receipts. Subsequently, however, the contractual obligations of these funds caught up with their receipt of contributions and, furthermore, the Personnel Law of 1970 imposed additional burdens. Consequently, they were able to lend only TL 0.8 billion to the SIB in 1971. This was slightly less than what the SIB spent on amortization and interest payments on its domestic bonds and credits. The net transfer of funds from the SIB -to the SEEs went down from TL 1.5 billion in 1970 to a paltry TL 26 million in 1971. 1/ The government has refused to redeem matured savings bonds unless the bearer can prove that he was the original owner. Since bonds were not registered in the early years, this policy hat resulted in losses for some individuals and might tend to undermine public confidence in govern- ment bonds. -33- 102. If the present situation continues, the SIB will be entirely negli- gible from the point of view of resource mobilization. The SIB.4ill not be able to finance an expansion in the investments of the SEE sector and its only function will be to redirect a given volume of funds between various SEEs. Even this limited function will not be fulfilled if SEEs fail to. repay their credits when due (as is often the case). Any expansion of its credit activities will depend entirely on bigger surpluses in the social insurance funds and on ad hoc transfers from the central government. The SIB can fulfill its role of mobilization resources for public investment only if it can sell its bonds outside the present capital market whose absorptive capacity is limited. (7) Government Bonds. 103. The relatively small increase in the amount of government bonds outstanding seems surprising when one considers the incentives offered: the interest receipts are tax free and the buyer does not have to disclose the size of his holdings; and they are almost perfectly liquid, despite a ma- turity of 20 years at issue. But, as we shall see below, each of these advantages 'is subject to some qualification and, consequently, commercial banks have been forced to take up most of the new issues. 104. The tax-free feature makes the government bond yield of 9 percent more attractive than the corporate bond yield of 15 percent to individuals who pay a marginal tax rate exceeding 40 percent, and to all taxable cor- pbrations. The qualification to this, of course, is that a tax-free provi- sion is useful only to those individuals who actually pay their taxes at the required rate. Our estimate is that only about 50,000 individuals pay a marginal tax rate exceeding 40 percent and, furthermore, most individuals do not declare their interest earnings on corporate bonds. Hence they pay only the withholding tax of 20 percent, a rate which leaves corporate bonds yielding mote than government bonds. 105. The second advantage is that government bonds are almost perfectly liquid. They can be cashed in at any time, the only penalty being the loss of accrued interest since the last coupon payment. (Even this penalty was absent before 1970.). The qualification is that the transactions are routed (on a bookkeeping basis) through the Istanbul Stock Exchange, thereby incur- ring a fee of 0.65 percent and, more seriously, the government has been known to put pressure on institutions (such as banks) when they wished to cash sizeable amounts of government bonds. 106. Government bonds also enjoy a small captive market: commercial banks are required by law to invest 5 percent of their earnings each year in Government bonds; banks have to maintain a liquidity reserve requirement of 10 percent of all sight liabilities for which government bonds are an eligible asset; Treasury Placement Bonds are available in nine-month ma- turities and are, therefore, a remunerative way to hold surplus cash; 1/ 1/ This is also subject to the qualification noted -above. Commercial banks may be forc!d to roll over their holdings of Placement Bonds on maturity. -34- corporations are legally required to maintain a contingency reserve for which government bonds (and other public sector bonds) are the only eligi- ble assets; and government bonds are accepted as security in bids for public works contracts. 107. Despite these legal advantages,.the sale of Goveinment Bonds depends essentially on the government's ability to force commercial banks to hold more that they wish to. The prime consequence of this is a reduction in the ability of the commercial banks to finance other investors. This is simply another way in which the public sector can preempt the use of re- sources which have already been mobilized to finance investment. Unless the sale of government bonds is pursued vigorously outside the banking system, they cannot be.an important means of mobilizing resources for investment. (vi) Central tank Advances 108. The expansion of public sector borrowing in recent years has taken place primarily through -the Central Bank. Commercial bank credits to the public sector have expanded relatively little in recent years, most of the increase being in credits to municipalities and local administrations channeled through the public sector special law banks (see Table V.3). 109. Almost all Central Bank credit to the public sector falls into two categories: short-term advances to the Treasury (central government budget); and advances to the Soil Product Office and Monopolies Administra- tion. The latter category is related to the agricultural pricing policies of the government and will be discussed in the following section. 1/ 1/ In addition, the Central Bank also makes credits to agricultural co- operatives through the (public sector) Agriculture Bank. We follow the Central Bank's classification in treating these as credits to the private sector. -35- Table V.3: Bank Credit to the Public- Sector (TL million, outstanding as of end December) 1965 1966 1967 1968 1969 1970 1971. Central Bank/ 3,184 4,204 5,195 5,939 7,393 80410 12LO7 - Short-term Advances to Treasury 1,136 1,438 1,951 2,181 3,057 4,359 .6,088 - Soil Products Office 569 797 1,008 1,020 950 1,125 3,000 - Monopolies Administration 423 723 750 1,210 1,750 1,750 1,750 - Other /2 1,056 1,246 1,486 1,528 1,656 1,176 1,249 Commercial Banks /3 2,539 2,574 2,777 2,671 2,997 3,209 4,234 Total 5,723 6,778 7,972 8,610 10,390 11,619 16,321 Total Banking System ..Credit 19,182 24,284 28,367 33,316 40,336 45,270 54,673 Credit to the Public Sector (percent) . (29.8) (27.9) (28.1) (25.8) (25.8) (25.7) (29.9) /1 Including credits to special law banks and Bank Liquidation Fund. /2 Mainly credits to the Sugar Corporation. /3 Mainly credits to local administrations, public enterprises and semi- public institutions. 110. The rising pressures on public savings and the increasing reliance on bank borrowing to finance budget deficits are most clearly reflected in the expansion of short-term advances to the Treasury from the Central Bank. The year-to-year expansion, however, has been restrained by legal limits.' From specific limits in the early sixties, the limit was set at 10 percent of the expenditure appropriations,of the General Budget from 1965 to 1968; increased to 12 percent in-1969 and to 15 percent since 1970. An attempt to raise the limit,to 20 percent in 1972 was defeated in Parliament. 111. Until 1964, short-term advances were very small -.less than 5 percent of the general budget appropriation and a negligible proportion of GNP. From 1965, there has been a steady expansion and since 1967, the out- standing advances have tended to hit the statitory ceilings each,year. By 1970, they exceeded 3 percent of GNP (Table V.4). -36 Table V.4: Central Bank Advances to Treasury 1965 1966 1967 1968 1969 1970 1971 Outstanding (end December) TL billion 1.1 1,4 2.0 2.2 3.1 4.4 6.1 Percent of Outsthoding Central Bank Credit 7 24.4 21.6 22.2 21.5 23.7 29.9 37.4 Percent of GNP % 1.5 1.6 2.0 2.0 2.6 3.1 3.4 112. The expansion has been made possible by the growth in budgetary appropriations (from TL 14.4 billion in 1965 to TL 50.3 million for 1972) and the relaxation of the legal limits. They have been instrumental in transferring command of an increasing proportion of the economy's resources to the central government and have been one of the major causes of an infla- tionary expansion in money supply. (vii) Agricultural Credit 113. The government intervenes actively in the pri4Lng and marketing of a wide range of Turkey's agricultural products (inclv&ing tea, tobacco, wheat, cotton, sugar and hazelnuts). An evaluation of the costs and benefits of the agricultural pricing policies is entirely beyond the scope of this study; what follows is siply a brief.discussion it view of the fact that these policies absorb a considerableamount of resources that would other- wise be available for fixed investment. 114. Government intervention takes many forms, .but includes an open- ended purchase program at a specified price financed by the Central Bank, subsidized distribution of inputs and minimum export prices. The price support program for tea and tobacco is administered by the Monopolies Ad- ministration; cereals and sugar by.the Soil Products Office and the Sugar Company (both SEEs). Agricultural cooperatives have been used to intervene in hazelnuts, cotton, raisins, etc. 115. A major principle in setting the support prices.appears to be the maintenance of a steady relationship between agricultural and industrial prices. The objective of this policy is to influence the distribution of income between the agricultural and non-agricultural sections of the poli- cies. Since suppori prices are usually announced at harvest time (except for sugar), they-cannot influence the short-run production patterns of the farmer. These policies have also been criticized for not providing incen- tives for-cost-saving innovations and improved techniques to increase yields; and for not being used as "agents of change" (again excepting sugar). - 37 - 116. These policies have a "cost" because the margin between the purchase and sale price of the commodities is not sufficient to cover the cost of storage, handling, marketing and overhead costs of the agencies involved. The Soil Products office which purchases about 10 percent of the cereal crop annually, has operating costs at least twice as high as the permitted margin between purchase and sala prices. The commissions charged by the agricul- tural sales cooperatives barely cover their overhead costs. The Central Bank contributes substantially to the financing of the operation of all the agencies involved in price support, and when these agencies are unable to cover their costs, the result is an accumulation of their credits from the Central Bank. The annual increase in Central Bank credits is shown in Table V.3. When the accumulation becomes excessive, the-debts are "consolidated" and transferred to the Treasury. Such a consolidation exercise in 1960 in- volved more than TL 2 billion owing from the price support agencies. 117. In planning monetary policy, the Central Bank accords first prior- ity to meeting-the needs of agricultural finance. Ceteris paribus, expansion of agricultural credit in such magnitudes inevitably has an expansionary influence on money supply and probably means faster inflation, Both these propositions are borne out in our analysis in Chapter VI. Expansion of Cen- tral Bank credit to the agencies involved in price support policies has had a significant influence on the rate of growth of money supply in the last decade. The demand pressures generated by the increased private incomes (in nominal terms) has resulted in higher prices. 118. Without laboring the argument, one must point out the existence of a potential vicious circle. Inflation increases the pressures for lower selling prices (to protect the consumer) and higher purchase prices (to pro- tect the farmer), This leads to more Central Bank credits to these agencies, a nore rapid expansion in money supply and faster inflation. The likelihoo that such pressures will be significant is greater than desired investment exceeds desired savings and the monetary system is already under heavy strain. (viii) Third Plan Projections 119. The Third Plan goes considerably beyond the previous two Plans in providing estimates of domestic financial flows that must be engineered in order to balance savings and investment in the public and private sectors (Table V.5). 120. The finances required for public investment exceed public savings as projected in the Plan. The gap is as much as 2.3 percent of GNP in 1973,. diiinishing in each of the subsequent years until, in 1977, the gap is almost closed. The private sector balance follows the opposite trend, signalling the effects of the public savings strategy embodied in the Plan. Private savings exceed investment projections in the first three years of the Plan, but fall below it in the next two years. The net inflow of foreign capital is positive in each of the years but follows a falling trend for the public Table V.5: TlMKY: THITD FIVE-YEAR PLAN FINANCIN-, BALANCE, 1973-1977 (TI billion at 1971 prices) 1T973 -T - i9711395 .74~17 Public Private Total ulTR [rlvr.7 7rt,,- Public Private Total Public Private Total Public Privat Total 1. Investment 25.7 19.6 45.2 28.8 .2 5j.() 32.3 25.2 57-5 35.0 28.8 14.7 39.6 33. 7.* 2.Savings 21.0 21.8 42.8, P5.*[ 1.1 Ig.14 300 254 5.4-345 2 3 6 3. Baace7 2. 3 --2- rW- -3.21 .) -?2- -2.3 0.2 -21L -¶ -_l-7 71n- - --. -14/ - 4. Domestic Cap,tal Transfers -0.1 0.1 - -0.2 0.2 -0.4 0.4 - -0.3 0.3 -0. - 5. Wealth TaxL IT T1 - 1. :L - 1.7 -1.7 - 1.9 -1.9 2.1 -2.1 - 6. Nationalization and Fixed Assets -1.2 1.2 - -1.3 1.3 - -1.4 1.4 - -1.5 1.5 -1.6 1. 6 7. Other -0.3 0.3 - -0.5 0.5- - -0.7 7 - -0.7 07 -0.3 . i - 8. Foreign Debt 0.8 j3 2.1L 0.6 -1 0.3 1.4 1.7L Oo 1.5 1 - G.-2 1. 1.4L 9. Domestic Debt A 12 1-2 - -2.0 2.0 -21 .1 -24 10. "Internal Dpbt" -i.1 1.1 -1.1 1.1 --1.1 1.1 --] -11 11. Cash/BankL7- -0.1 0.1 - -0.9 0.9 - -1.0 1.0 - -1.0 1.0 - -1.P 1. - 12. Balancing Item. 5.2 -4.9 0.3 4.7 -4.4 0.3 4.5 -4.1 0.4 4.1 -3.7 0.4 3.3 -2.9 0.4 (Percent of GNP) 1. Investment 12.5 9.5 22.0 13.0 10.0 23.0 13.6 10.5 24.1 13.9 11.1 . 25.0 -14.1 11,9 26.0 2. Savings 10.2 10.6 20.8 11.6 104. 22.0 12.6 10.6 23.2 13.4 10.9 24.3 14.1 11.3 25.4 3. Balance -2.3 1.1 1.2 ITT ) T -1.0 -1.0 0.1 -0.9 .03 -0.2 -0.7 .0 4. Domestic Capital Transfers -0.1 0.1 - -0.1 0.1 - -0.1 0.1 - -0.1 0.1 - 0.1 .1 5. Wealth TaxL 0.7 -0.7 - 0.7 -0.7 - 0.7 -0.7 - 0.7 -0.7 - , -0. - 6. Nationalization and Fixed Assets -0.6 0.6 - -0.6 o.6 - -0.6 0.6 - -0.6 0.6 - -0.6 0. - 7. Other -0.1 0.1 - -0.2 0.2 - -0.3 0.3 - -0.3 0.3 - -0.1 Q. - 8. Foreign Debt 0.4 0.6 1.0 0.3 0.6 0.9 0.1 0.6 0.7 0.0 0.6 0.6 -0.1 ,. 0.5 9. Domestic Debt -o.6 0.6 - -0.9 0.9 --' -0.9 0.9 -o - -1.1 1.1 - 10. Internal ebt -0.5 0.5 - -0.5 0.5 - -0.5 0.5 - -0.5 J.5 - -1. Cash/BanL2b -0.1 0.1 - -0.4 0.4 - -0.4 0.4 - -0.4 0.4 - -0.6 . - 1P. Balancing Iter4 2.5 -2.4 0.1 2.1 -2.0 0.1 1.9 -1.7 0.2 1.6 -1.4 0.2 1.2 0.2 This deficit is filled by foreign savings (i.e. the current account deficit of the balance of payments). 12 The definition of this tax is unclear. It probably refers to existing taxation of property, in which case it should be included in current revenues and, hence, in public ao'i There is a slight discrepancy between foreign capital inflow and the sum of net foreign borrowing by the public and private sectors. There could be a compensating theatty it the domestic capital transfer entries. A "Internal debt" is shown in Table 114 of the TFYP as a net item (Income-Payment.) Table 118 has entries for "Loan to the Public Sector" from the private sector, which are assumed to be included in the net flow. /5 This presumably includer changes in the Treasury Cash position with the Central Bank. No distinction is made in Table 114 between the banking system dnd the non-bar- p;tlic. /6 Logically, row 3 should be exactly offset by the sum of rows 4, 8 and 9. T1h "balcncing item" represents the additional financial flow. required in order to achieve ac'oral savings/investment balance. Source: Third Five-Year Plan (1973-77), SPO, Turkey; Tables 10?, 105, 3111, 138. - 39 - sector and a rising trend for the private sector. Thus, in the final two years of the Plan,,both public and private sectors suffer a shortfall of. own savings and the gap is filled entirely by foreign capital flowing to the private sector, 121. The public sector will have to borrow substantial sums from the private sector if it is to meet the obligations anticipated in the Third Plan. For the Plan period as a whole, the additional borrowing needs amount to 1.8 percent of GNP (falling from 2.5 percent of GNP in 1973 to 1.2 percent in 1977). This borrowing requirement can be avoided in part (but not com- pletely) if the public sector does not reduce its domestic and foreign debt obligations at the pace anticipated in the Plan projections (rows 8 and 9 of Table V.5). However, the borrowing needs will be higher than projected to the extent that public savings projections prove to be over-optimistic. (The discussion in Chapters II-IV suggests that this is the case, especially as regards the projections of SEE profits.) Further, no explicit provision has been made in the Third Plan.projections for expansion in credit to the . Soil Products Office and Monopolies Administration which were major recipients of Bank credit in recent years (see Table V.3). 122, The magnitude of the required public borrowing effort can be as- sessed in comparison with past increases-in the main elements of public bor- rowing shown in Table V.2. Excluding the compulsory savings bonds (which have since been replaced by a tax) and the sale of bonds to the social funds (which are expected-to show a deficit in the Third Plan period), public bor- rowing averaged 1.6 percent of GNP from 1965 to 1971, of which 1.3 percent was credit from the banking system. The remaining 0.3 percent repref,ents the sale of government bonds, most of which was also taken up by the commer- cial banks. 123. The projections of the Third Plan imply a considerable expansion in borrowing by the public sector from sources other than the social funds. If the necessary restraint is observed in expanding Central Bank and short- term advances, this means the sale of government bonds to commercial banks and individuals (since other financial institutions do not yet offer a sizeable market in Turkey). The task of selling bonds.will assume even greater urgency to the extent that realized public savings do not match up to Plan targets. The recent upsurge in worker's remittance has presented the government with an excellent opportunity to expand borrowing from the non-bank public. But success in this venture requires the creation of an environment in which private savers will willingly expand their holdings of financial assets, which depends primarily on the nature of fiscal and monetary policies (especially interest rate policy). Creating such an environment is a critical factor in achieving the sectoral investment targets of the Third Plan. - 40 - VI. MONETARY POLICY (I) Introduction 124. The Central Bank is empowered to use a wide range of instruments for the purpose of short-term demand management and influencing the alloca- tion of credit. These include statutory reserve and liquidity requirements for commercial banks, import guarantee and advance deposits against letters of credit, a complex set-of quantitative controls on rediscount privileges of 'commercial banks and, perhaps most importantly, the power to vary the level and structure of all interest rates in the financial markets. 125. It is not the purpose of this study to comment on the efficiency with which these instruments have been used or the goals they are.meaut to serve. But monetary policy and the monetary implications of other public policies (such as agriculture credit) do have a bearing on the economy's ability to mobilize adequate resources to meet investment targets. It is from this point of view that monetary policy is considered in the following sections. 126. Section (ii) will demonstrate that the public sector's credit needs determine the supply of money which is closely,related to the rate of infla- tion in recent years. The following sections discuss how interest rate regu- lation has had a constraining influence on the growth of the capital market, thus inhibiting both public and private borrowing to finance fixed investments. (ii) Money and Inflation 127. There has been a striking acceleration in the rate of-price infla- tion in recent years. While the Wholesale Price Index and the GNP deflator show slightly different movements from year to year, the broad trends are the same in each: prices rose by about 4 to 5 percent per year in the First Plan period; by about 9 percent during 1967-1971; and the increase has been particularly high in 1971 (see Graph VI.1). 128. There ate, of course, numerous explanations that can be advanced for these.price trends. Further, there is the familiar difficulty in estab-. lishing the direction of causality in trends that are closely associated. 129. The analysis of price trends is especially complicated in Turkey since many prices (SEEs, agriculture) are under the direct influence or con- trol of the central government. To a large extent, this power is used to influence relative prices in the economy, as already discussed 'in the context of agricultural price policy and the performance of SEEs. It is, however, the case that price decisions can be forced on the government by price move- ments in other sectors of the economy. For example, many SEE prices had to be raised 1/ in 1971 because of the pressures of devaluation and general price and wage increases in the rest of the economy. It would, therefore, If Unless the central government would be willing'to provide enormous sub- sidies to SEEs, thereby increasing the budget deficit unreasonably. -41- be a mistake for an economic- analyst to consider administered prices as wholly under direct government control, impervious to inflationary pressures. 130. Any single explanation of the pace of inflation over the last decade will inevitably be an over-simplification. However, the data are consistent with the hypothesis that the expansion in money supply has been a very signif- icant influence on price movements. 131. Exogeneity of money supply: Monetarist explanations of inflation are essentially based on the correlation between money.supply and nominal income, and have to make the crucial assumption that causality runs from money to prices rather than vice versa. This assumption has been widely at- tacked in economic literature on the grounds that income could be one of the major determinants of the stock of money. Monetary expansion could be said to be a possible explanation of inflation only if the supply of money is exogenously determined. This is an empirical question, the answer to which may vary from one country to another. In Turkey, the supply of money does seem to be exogenous, as demonstrated in equationg (5.1) and (5.2). Money = 1873.71 + 2.65 TRE +3.93 MNP +2.06 AGCRS + aiD ... (5.1) (s.e.) (107.3) (0.28) (0.78) (0.36) i 2 = 0.948;-DW /1 = 1.47; N = 119 11 MONEY 2 = 195.53+ 0.399 TRE + 0.519 NP + 0.289 AGCRS + biD .. (5.2) (s.e.) (16.04) (0.05) (0.13) (0.06) 1 i -2 = 9 = 1.23; N = 119 R = 0.978; DW - where: MONEY = Money Supply MONEY 2 Money supply plus Quasi-money TRE = Central Bank Credit to the Treasury, consisting of short-term advances, bonds, advances against gold and Bank liquidation fund; MNP = Central Bank Credit to the Monopolies Administration AGCRS Central Bank credit to the Soil Products Office and Agricultural Sales Co-operatives D Monthly Dummy Variables /2 a., b. = Coefficients of the Dummy Variables 1 1 Sample period: January 1962 to December 1971 /1 The data have been transformed before estimation (using a first-order autoregressive transformation) to reduce serial correlation. The first order serial correlation of the residuals prior to their transformation was 0.81 for equation (3) and 0.84 for equation (4). /2 Two of the eleven Dummy Variables were not significant at the 95 percent cunfidence level in equation (3) and 4 were not significant in equation (4). But a joint test on all the dummy variables strongly rejects the hypothesis that seasonal factors did not exert a significant separate influence on the variation in Money and Money 2. Grap£h V.1: Price Trends, 1962 to 1971 (1968=100) (Log-scale) 150.0 140.0- Wholesale Price Index (1968=100) 130. 0 GNP Deflator 120., 110.0 100.0 90.0 80.0 70.9 1962 1963 1'964 1965 1966 1967 i 6 -8 1969 1950 71 -42- These equations imply that that part of the variation in the stock of monetary. liabilities in the,economy which is not explained by purely seasonal factors is almost completely explained by variables which are completely under the control of governmental policy. These "policy" variables are selected compo- nents of Central Bank credit -- to the Treasury, to the monopolies adminib- tration, and to agricultural sales cooperatives. 1,32. Central Bank credit to the Treasury consists almost entirely of short-term advances which haVe been discussed in Chapter V (vi) above. The other two variables depend directly on agricultural credit policies discussed in Chapter V (vii). The important point is that variations in these components of credit depend entirely on government policy. Since equations (3).and (4) use monthly data, some of the variation in money supply is due to seasonal factors. Eleven dummy variables (and the constant term) have been introduced in these equations to explain these seasonal variations. The significance,of the estimated coefficients of the other explanatory varia- bles cannot, therefore, be attributed to any common seasonality between them and the supply of money. 133. One may, therefore, conclude that the supply of money is exogenous to the income determination process in Turkey. By careful control of these specific components of Central Bank credit, it is possible to determine the stock of monetary liabilities in the economy. 134. Money demand: it is also possible to demonstrate that there is a very close link between th6 stock of money and the movement of prices in the economy. 135. The following equations specify the demand for real money balances as a function only of contemporaneous real income. The interest rate is an obvious omission in this specification. In an economy where the entire structure of interest rates is so rigidly controlled and unrepresentative of economic relationships, it was felt that the inclusion of any single interest rate would be inappropriate. It is found that the data support the hyp6the- sized specifications reasonably well, whether a "narrow" or "broad" defini- tion of money supply is used. 1/ Log (MR) = -3.03 + 1.31 Log YR ... (5.3) (s.e.) (0.3) (0,07) R2 = 0.98; DW = 1.79; N = 10; Log M2R = -3.48 + 1.47 YR ... (5.4) (s.e.) (0.36) (0.08) R2 &0.98; DW = 2.06; N = 10 1/ The definitions of money and quasi-money used here correspond to those used by the CeZtral Bank. where: Money Supply YR = GNP at market prices M2R = Money plus Quasi-money and all variables have been deflated using the GNP deflator. 136. The rate of growth of the demand for real money balances is very closely related to the rate of growth of real income. The equations prom- ised above use a log-linear specification of the relationship between real income and the real stock of money. Equations using a linear specification were also estimated with equal success (Table VI.1). In view of the relatively unsatisfactory nature of price data, equations were also estimated using the wholesale price index instead of the GNP deflator. As shown in Table VI.1 this had almost no effect on the estimates of the slope parameters although the intercept was shifted downwards. 137. The strong implication of these equations is that exransion of the money supply faster than a given rate (related to the rate of growth of real output) will be associated with corresponding increase in the price level. The increase in prices is required to equilibrate the demand for real balances with the supply of money., The actual rates of growth of prices in the last. decade are remarkably close to those predicted by the monetary explanation above (Graph VI. 2). 138. Since the rate of growth of money supply is determined by Central Bank credit to the public sector, it follows that the recent acceleration of inflation is due, in large part, to government policy. The failure of the public sector to generate adequate savings to finance investment led to an inflationary dependence on bank credit during the Second.Plan which, as discussed in Chapter II (ii), could have had an adverse effect on savings performance. (iii) Interest Rate Regulation 139. Bank lending rates: Control over the structure of interest rates has had..a long history in Turkey and has been perhaps the most important instrument of monetary policy in the last decade. Official pronouncements regarding the desirable structure and the degree of control to be exercised have varied greatly. But in practice, the policy has been to keep bank lending rates beloi the market equilibrium rate and use administrative reg- ulation to influence the allocation of credit. Interest ceilings are pre- scribed for as many as 25 categories of credits, deposits and rediscounts and, in practice, the actual rates are set at the prescribed ceilings. With minor exceptions, the structure of legal interest rate ceilings remained unaltered from 1961 to 1970, when there was an upward adjustment (of about 1.0 to 1.5 percent in most rates with time deposit rates going up by 3 percent). In February 1973, the interest rate structure was modified again. The new structure represents a partial return to the nominal rates prevailing before 1970. Table VI.1, Regre'ssion Results: The Demand for Money MR (a) -4.36 + 0.26 YR (a) 0,97 -17 (s.e.) (1.3) (0.02) IR (b) -5.76 + 0.26 YR (b) 0,98 1.69 (s.e.) (1.h) (0.01) M2R (a) = -9.58 + 0.39 YR (a) 0.98 1.86 (s.e.) (1.7) (0.02) M42R (b) = -12.3 +" 0.39 YR (b) 0.98 1.94 (s,e.) (1.7) (0.02) log MR (a) = -2.9 + 1.30 log YR (a) 0.97 1.77 (s.e.) (0.4) (0.08) log MR (b) = -3.03 + 1.31 log YR (b) 0.98 1..79 (s.e-.) (0.3) (0.07) log M2R (a) = -3.31 + 1.45 log YR (a) 0.97 1.93 (s.e.) (0.36) (0.08) log M2R (b) = -3.48 + 1.47 log YR (b) 0.98 2.06 (s.e.) (0.32) (0.07) .(a) : deflated, by the wholesalë price index; (b) : deflated by the GNP deflator; YR i reai GNP at market pricesj MR : real stock of money; M2R : real stock of money plus quasi-money. - 45- 44 Graph VI.2: Regression Estiiates of Price Trends /1 (L,og-scale) 150.0 1h0, o - GNP Deflator - Regression Estimates 130.0 . 120. 110. 100.0 90.0 80.0 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 /1 Estimates of price index derived from équation (5.3). -46- 140. The 1970 reform of the interest rate structure also introduced a system of separate subsidies to both banks and borrowers for credits in priority sectors4 These subsidies would have reduced the cost of bank credit to the borrower and, at the same time, raised the return on lending above the nominal interest rate ceilings. Compared with the interest rate on general credit of 11.5 percent., the banks could earn 11.8 percent on short-term pri- ority credits and up to 14 percent on medium-term credits to the priority sectors. The borrower' on the other hand, paid only 7.5 to 9 percent for short-term and a maximum of 11 percent for medium-term credit (only 8 percent for export and tourism, 9 percent for agriculture). The difference between what the borrower paid and what the bank earned was to be paid by a "Selective Credit Fund" established in the Ministry of Finance. Regrettably, this fund has not been funded and the interest subsidy scheme has not been made effective. The 1973 decree again includes this system of selective subsidies, but there ,is no great assurance that the scheme will be implemented. 141. Although these differential interest .rate schemes have not been successful, the allocation of credit has been very largely under the direct control of public policy. Agriculture, export, artisans, small traders and selected industries continue to receive bank credit at intere3t rates 1 to 3 percent per annum below the rate charged for general loans. To encourage banks to make loans available at these concessional rates, the Central Bank stood ready to rediscount bills of these priority sectors at interest rates 2.25 to 3 percent below its general rediscount rate. The unintended conse- quence of this arrangement was that banks made loans to priority sectors only to the extent that they were able to rediscount them with the Central Bank. The burden of providing credits to these sectors was, thus, transferred to the Central Bank where they competed with the Treasury and agriculture price support agencies for scarce funds. 142. The use of differential interest rate.ceilings to influence the allocation of credit has been supplemented by the establishment of special public banks whose purpose is to finance agriculture, residential construc- tion, tradesmen and artisans and the construction of municipal offices. These public banks account for more than half of the credit extended by the commercial banking system. However, they are not used effectively as a means of gathering deposits: the ratio of credits to deposits was 1.3.for the public banks as against 0.8 for the private banks at the end of 1970. A part of the burden of providing resources to meet the credit needs of these sectors has had to be taken up directly by the government, putting a further strain on budgetary resources. 143 Deposit rates: The rates of interest payable by commercial banks on their deposit obligations have also been subject to rather low ceilings since 1951. Table VI.2 compares the various deposit rate structures. - 47 - Table VI.2: Bank Deposit Interest Rates After February 1961-1970 1970-1973 1973 Sight deposits Saving (individuals) 3.0 3.0 2.5 Public and commercial 2.0 1.0 0.5 Time deposits 4 to 6 months 4.0 4.0 (4.0 /1 6 months to 1 year 5.0 6.0 ( 1 year to 18 months 6.0 9.0 7.0 /2 more than 18 months 6.5 9.0 9.0 7 /1 3 months to 1 year. 7T 1 year to 2 yegrs. T3 Over 2 years. 144. Depositors have earned a negative real rate of return on both sight and time deposits throughout the last decade. Naturally, banks try very hard to expand their deposit base. Since interest rate competition is forbidden, they turn to other methods. The use of lotteries with prizes to attract deposits became so common that they had to be regulated. Prizes offered are now limited to a maximum of 0.5 percent of outstanding deposits plus a fixed sum determined by the Central Bank. Advertising by commercial banks is noticeably heavy in urban centers and there has been an attempt to attract more deposits by expanding the number of branches. The number of branches of private commercial banks has risen from 1,035 in 1968 to 1680 in 1971 (+62 percent) while public sector bank branches rose from 1,127 to 1,425 (+25 percent) during this period. The spread between the average ef- fective,deposit rate and the loan rates appears to have been absorbed by higher administration expenses, which is very unfortunate from the point.of view of mobilizing financial savings. 145. ' Bank deposits form about two-thirds of all financial assets in the economy and poor mobilization of deposits probably means poor mobilization of private savings.. The effect cannot be quantified, however, since there have been many other factors influencing the rate of deposit accumulation in the economy. Government monetary policy has forced a rapid increase in the supply of deposits; inflationary pressures and interest ceilings have pre- vented the development of a capital market which would provide alternative' financial assets to private savers; improvement in the balance of payments has resulted in a reduction of blocked import deposits with the Central Bank, part of whicb must have been transferred to commercial banks; the surge of workers' remittances after devaluation might reflect a transfer of savings from abroad to banks in Turkey since the exchange risk has been temporarily reduced. - 48 - 146. The ratio of quasi-money to total moneta-y liabilities of the banking system was less than 0.27 for a number of years, but increased sharply to .0.30 at the end of 1970 and 0.33 at the end of 1971. The various factors mentioned in the, previous paragraph would also affect this ratio but the in- crease in the interest rate from 6.5 to 9 percent probably had a significant effect, indicating that deposits would respond to the stimulus of higher in- terest rates. 147. The only reason that has been advanced in favor of deposit rate ceilings has been the need to avoid "excessive" competition among banks, reduce bankruptcies and retain public faith in the banking system. These objectives can be achieved more effectively by supervision of bank liquidity (which is already being done by the Central.Bank) and the introduction of a deposit insurance scheme. These measures would permit a general increase in inter6st rates paid to depositors, would probably result in greater deposit mobilization and remove some of the distortions in the financial system. 148. Corporate bond rates: The corporate bond market in Turkey dates back only to 1967 and is still quite small and thin. The gross total of bond issues in five years is less than TL 1 billion (the amounat outstanding at end 1971 is obviously much less; in comparison, outstanding government bonds were TL 3.2 billion). A major reason for the lack of development of the corporate bond market has been the introduction of an interest rate ceiling of 15 percent in June 1970. Interest payments to individuals are subject to a 20 percent withholding tax, reducing their earnings to 12 per- cent and interest payments to banks and insurance companies are subject to a 25 percent transactions tax, reducing their earnings to 11.25 percent. However, as noted in Chapter V (v), corporate bonds are more attractive than government bonds for individuals who pay only the withholding tax (as is usually the case with bearer bonds) and for tax-free institutions (such as private pension funds). But the ceiling on interest rates does restrict the size of the corporate bond market to what individuals and insti- tutions are willing to provide at these rather low interest rates. 149. Effective cost to borrowers: The obvious consequence of interest rates below an "equilibrium" *rate is that there is usually an excess of demand for funds by investors. 1/ The lenders are, therefore, able to capitalize on this excess demand by passing on to the borrower as much of their administra- tive costs as is legally permissible. This appears to be an accurate descrip- tion of the situation in Turkey. In addition to the nominal interest rate applicable for the credit, the borrower is also required to pay a variety of taxes, commissions and fees. A detailed and consistent statement of all the permissible charges is hard to come by. But there is considerable unanimity that the actual cost of short-term bank credit in 1972 was in the region of 20 percent per year, as against a nominal interest rate of 11.5 percent for general credits. Table VI.3 illustrates the variety of charges passed on to the borrower which, in theory, could equally well have been borne by the lender. 1/ Except for periods when the investment climate is uncertain, as in 1971. - 49 - /1 Table VI.1: Additional Costs of Bank Credit-- Commission: 1.5 to 3.0'percent of value of credit. Stamp tax, 0.5 percent incurred twice if.a letter of credit is used. Communication costs: 0.5 percent. Insurance: 1.0 percent, if credit is against merchandise. Transactiofts tax: 25 percent of interest and all other charges except stamp tax and insurance. /1 Costs other than interest charges which are paid by borrower; the details are merely illustrative, having been derived from a review of a few ttansactions in 1972. The 1973 decree reestablishes minimum and maximum commission and service fees which banks may collect at rates lower than before. 150. Similarly, the cost of issuing a corporate bond is much higher than the interest ceiling of 15 percent (of which the individual lender only gets 12 percent after tax). Only exceptionally well-established firms can make bond issues without being underwritten by a bank (or development bank). With the addition of underwriting fees, taxes and miscellaneous changes, the cost of a bond issue rises to about 20 percent. 151. On this evidence, it appears that ceilings on interest rates have not achieved their major purpose: to reduce the cost of credit to investors. Of course, only fragmentary evidence is available on what the equilibrium rate really is. When a corporate bond market flourished briefly in Istanbul, interest rates around 15 percent per annum were prevalent in early 1970; compulsory savings bonds were traded in a secondary market.until 1967 at a discount which implied interest rates around 30 percent. The interest rate implied by transactions in the secondary bond market in Istanbul in mid-1972 was about .16.5 percent and the coupons of compulsory savings bonds are also traded at similar discounts. But these markets are so thin that one hesi- tates to draw firm conclusions on this basis. Since the government bond market is pegged (see Chapter V (v)) and the inter-bank transactions tax has prevented the development of a money market, there is no other evidence available on freely determined interest rates. It is our judgment,'however, that the equilibrium interest rate is not far.from 20 percent, i.e. not much higher than the rates already paid by borrowers from banks. - 50 - 152. Distortions: Although interest rate ceilings do not reduce the cost of credit, they do-cause distortions in the economy, and result in inefficiencies in mobilizing funds for fixed investment. The commission system does not eiable banks to distinguish adequately betweet good and bad risks, which would lead them always to prefer established borrowers. Since bank income is based to a large extent on various fees, they would tend to prefer credits against bills rather than unsecured credits (which may have more development impact). This preference is reinforced by the banks' desire to acquire bills which can be rediscounted with the Central Bank. 1/ These considerations contribute to heavy concentration of commercial bants lending at the short end of the maturity spectrum. Commercial banks have no incentive to develop the capability to make medium- or long-term credits to industry based on an appraisal of investment ideas rather than a discounting of bills presented. An attempt to force them into this field by decree has not been successful. 2/ 153. While banks have successfully raised their effective lending rates to collect the difference between legal ceilings and equilibrium rates, they have also succeeded, through a trade association (The Bankers Association of Turkey) to prevent.similar escalation of deposit rates. As.already noted above, there is little incentive for savers to accumulate deposits, unless it be the scarcity of other financial assets in the economy. 154. There are two private industrial development banks which provide long-term industrial finance in Turkey. Since their lending rates have been subject to legal ceilings, neither of them has been able to mobilize resources profitably from the domestic market. The larger of the two development banks (TSKB) has obtained almost all its resources from international lending insti- tutions and the government; only 11 percent of its capital is equity; and it has not borrowed at all f.rom private individuals or institutions. The other development bank (SYKB) is dependent on contributions from six shareholding commercial banks for which it pays less than market rates, The inability to raise resources has undoubtedly stunted the growth of these institutions. The long-term finance provided by them accounts for only about 10 percent of private fixed investment in manufacturing industries. 155. The financing needs of private industry are met primarily by in- creases in equity and the rollover of short-term commercial bank credit. Since most firms in Turkey are owned by a small number of shareholders, the 1/ The use of sector-specific rediscount policy is, in practice, the most important means by which the Central Bank influences the allocation of credit. A discussion of this is outside the scope of this paper. 2/ A decree was passed that at least 10 percent of commercial credits out- standing at the'end of 1972 should have maturities exceeding one year. No penalty was specified for lack of compliance. - 51 - need for additional equity is met either by the retention of profits oir by additional subscriptions from existing shareholders. This does not provide an environment suitable for the development of a vigorous equity market. 1/ 156. Regulation of the structure of interest rates has probably hindered the financing of investment in Turkey. Private financial savings are dis- couraged by ceilings on deposit rates and the scarcity of alternative finan- cial assets. The financing of private investment relies, to an undesirable extent, on the wealth of the entrepreneur. Public investment relies on dwin- dling public saving, captive sources of borrowing (e.g., social insurance) and the banking system. VI. CONCLUSIONS 157. The principal objective of Turkey's Third Five-Year Plan (1973-77) is to accelerate the rate of growth of per capita GNP to 5.2 percent from an annual rate of 4.2 percent in the previous decade of planning. There is no question that this is a modest target, whether considered in terms of Turkey's needs or its potential. There is some room for doubt, however, whether the Plan's strategy for mobilizing national resources to meet invest- ment needs is best suited to achieving this objective. 158. The savings strategy of the Third Plan relies crucially on the presumption that the marginal propensity to save is very much higher for the public sector than for the private sector. Therefore, a transfer of income from private to public sector is the major means by which savings targets are to be achieved. An examination of the individual components of public savings does not support this strategy. The problems that prevented realization of the public savings target in the 1I Plan are still.unsolved; and there are some additional problems that will affect public savings in the III Plan. 159. The public expenditure projections mean a reduction in the rate of growth of real expenditures. To achieve this, the central government will have to severely restrain the growth of transfer expenditures and adopt anti- inflationary policies to prevent pressures for wage increases that will be hard to resist. The past few years saw very significant increases in the ratio of taxes to GNP. The scope for further revenue generation, however, is somewhat limited to widening the coverage of production taxes and better implementation of direct taxes. Past efforts have made the tax ratio to GNP commendably high, which makes it increasingly difficult to increase the ratio further. To achieve the Third Plan target, taxes will have to absorb a third of the aggregate increment to gross national product -- a burden which falls primarily on wage-earners since large segments of the economy do not contribute much in tax payments. - 1/ The legal environment regarding the regulation of corporate equity is also unsuitable. Reform of this is the subject of a pending Capital Market Bill. - 52 - 160. Surpluses generated by the large and numerous State Economic Enterpiises ought to make a major contribution to public earnings in Turkey. But they do not. Part of the solution lies in a careful re-examination of the operations of a few large loss-makers -- railways, coal, etc. But there is also a need to emphasize operational efficiency and profitability objec- tives for the large number of enterprise who turn in a very low rate of re- turn on investment. One must not underestimate the magnitude and complexity of this task, or the political obstacles in achieving a reform of the mana- gerial structure of public enterprises. Some early gains may be achieved, however, by implementing more widely the admirable subsidy scheme legislated in 1964 (Law 440), by reassessing the social justification for pricing poli- cies that provide subsidies to private industry and even, perhaps, selling enterprises to the private sector if they do not serve any discernible so- cial purpose. 161. We have noted that agricultural price support policies have ab- sorbed resources that would otherwise have been available to finance invest- ment in productive capacity. Their main impact appears to be on the sectoral distribution of income rather than on productivity of agriculture. Since the agricultural sector is favored in the tax system as.well, their dis- tribution effects within the agricultural sector should be assessed. 162. In the past, the failure of the public sector to generate adequate savings has led to an inflationary reliance on borrowing from the banking system. It is doubly important to avoid this temptation in the next few years if the present, rather high, rate of inflation is not to be perpetuated. A period of high inflation leads to widespread expectation of its continuance. This will make inflation difficult to arrest, with adverse effects on the savings rate (at least until economic institutions and behavior adjust fully to the fact of inflation). The effect will be particularly strong on public savings -- the result of low tax revenue elasticities, rising costs of social services, pressures for upward adjustment of wages and social insurance pay- ments, and falling SEE surpluses. 163. The review of public savings potential has highlighted the impor- tance of a reform of the capital market in Turkey to enable long-term bor- rowing in ways that do not iniect excessive liquidity into the financial system. We have noted that the centerpiece of this reform is not institu- tional but a relaxation of regulations on interest rates. Commercial banks would then be able to mobilize deposits more effectively, increasing the po- tential for non-inflationary borrowing by the public sector. The ability to expand sales of long-term bonds to the public would significantly help the financing of private fixed investment and put some life into the moribund State Investment Bank. The availability of a wider range of earning assets would stimulate the growth of non-banking financial institutions, which would also assist in mobilizing private savings. These reforms would probably be more effective in mobilizing resources for increasing fixed investment than the policies outlined in the Third Plan. ANNEX A Page 1 Tax Revenue Projection The purpose of this annex is to describe briefly the procedure and assumptions used to obtain our revenue projection presented in Table III.3 of the main text. A more detailed discussion of tax revenue forecasting is given elsewhere and will not be repeated here. 1/ Problem of Tax Revenue Projection with Buoyancy Method A commonly used met'hod for tax revenue projection is to estimate the revenue buoyancy with respect to some postulated determinants, e.g., GWP, price index, from the observations in the recent past and to calculate the future revenue for the assumed levels of these determinants from the buoyancy. In Turkey, the tax revenue buoyancy with respect to GNP at mar- ket prices calculated from the data for 1962-1971 is 1.25. 2/ Hence, accord- ing to the buoyancy method, tax revenue would be projected to grow at about 9.8 percent a year in real terms during the Third Plan period. However, the buoyancy of 1.25 reflects not only the automatic response of tax revenue to income growth but also the effects of successive changes in tax structure. In 1971, about 7.6 percent of the revenue was collected on the bases-of the new taxes introduced since 1963. In this situation, it is necessary to assess the revenue-income relation under the present tax structure in order both to project the revenue which could be obtained during the Plan period without changing the tax structure further and to estimate the magnitude of additional tax efforts required for the Plan. Unfortunately, the estimation of the revenue functions under the "present structure" in this situation tends to impose a serious technical difficulty, for the revenue series pertaining only to the "present struc- ture" is short. A number of approaches to the .problem were suggested .by the edonomists faced with similar cases, but no perfect solutions have been found. 3/ 1/ For a discussion on technical problems of tax revenue forecasting and the description of our approach, see K.K. Artus, "Tax Revenue Forecasting for Turkey," draft, February 1973. 2/ Unless otherwise stated, "tax" in this annex includes the savings bonds. 3/ R.W. Bahl, "Alternative Methods for Tax Revenue Forecasting in Develop- ing Countries: A Conceptual Analysis," (International Monetary Fund DM/72/83, October 16, 1972)-presents a survey'of various approaches. K.K. Artus, op. cit., discusses these approaches from a different viewpoint. ANNEX A Page 2 Revenue Functions After having examined alternative approaches, we chose the dummy variable method for estimating most of the revenue functions. This is the orthodox way of introducing non-quantifiable determinants into regression equations. 1/ In this particular case, it amounts to assuming that the revenue from a class of taxes is determined by certain quantifiable deter- minants (for which we chose the gross value added in the related sectors) and the tax structure. Since the latter is non-quantifiable, it is intro- duced into the regression equations through dummy variables. The estimated coefficients of revenue functions are presented in Table A.1. These are obtained from regressions using the data for 1962-1971. The coefficients reflect the revenue pattern under the structure effective at the end of 1971. The tax structure was changed further in 1972, but the revenue functions under the 1972 structure could not be estimated because the 1972 revenue observations are not yet available at present. However, the revenue from the fiscal balance tax may be projected on the basis of the savings bond revenue function. Also, the revenue from the foreign travel expenditure tax may be assumed nil from 1972 onward. The revenue function from the income tax was estimated excluding the 1971 observation. In 1971, the share of the public service gross value added in 1971 in GDP shifted upwards due to the implementation of the Per- sonnel Law of,1970. This is likely to cause an upward bias in the income tax revenue elasticity with respect to GDP. Although the upward shift of this sector's share will not be reversed in the near future, it is unlikely to affect the elasticity itself. Therefore, the revenue function thus ob- tained can be used for projections with an adjustment for the shift in the revenue level. As Table A.1 shows,,the obtained revenue functions are all of the constant elasticity type. This form was chosen for each class of taxes after examining the results of preliminary regressions assuming alternative simple functional forms. 1/ J. Johnston, Econometric Methods (McGraw-Hill, 1972). Annox A ag; 3 Table A.1: Summary of Estimated Revenue Functions Under the 1971 Tax Structure Left Right Hand Hand Variable Variables Coefficients Income Tax In R Const. -7.63 In GDP 1.38 Corporation Tax In R Const. -6.53 In Y02(-1) 1.26 Savings Bonds-Fiscal In R Const. -7.21 Balance Tax In YSB 1.24 Defense Tax on In R Const. -8.17 Buildings in Y05 1.5h Motor Vehicles Tax In R Const. -9.55 In GDPR 0.80 In PGDP 1.00 Inheritance and Gift In R Const. -13.h4 Taxes and Real Estate in GNP 1.66 Purchase iax Production Tax (on In RT Const. -3.51 domestic goods and in GDP 0.99 imports, but ex- cluding petroleum and petroleum products) Petroleum Production In RR Const. -15.16 Tax (on domestic In GDPR 3.29 goods and imports) in PGDP -3.22 Foreign Travel In R Const. -11.69 Expenditure Tax In GD? 1.52 ROX -0.0657 Sugar Consumption Tax In R Const. 0.19 In GDP 0.88 In R(-1) -0.70 Bank and Insurance In R Const. -0.93 Transactions Tax In Y18 0.97 PTT Service Tax and in R Const.4 -.43 Transportation Tax In GDP 0.85 Annex A Page 4 Table A.1: Sumnary of Estimated Revenue Functions Under the 1971 Tax Structure (continued) Left Right Hand Hand Variable Variables Coefficients Stamp Duty (except . In R Const. -15.29 on imports) in GDPR 1.40 1n PGDP 1.22 Customs Duty and ln R Const. .2.63 Stamp Duty on in 14 0.56 Imports Other Taxes ln R Const. -5.50 In GDP 0.99 Symbols: GDP: GDP at current market prices (TL million). GDPR: GDP at constant 1968 market prices (TL million). PUDP: GDP deflator'with 1968 as the base year (1.968 = 100). G-P: CP c.t c4.rrent :.c.rkt pipes (TL iion) *Y02(-1): Previous yec:'s gross value added in industries, construction, trade, transportation, and financial services in current prices (TL million). YSB: Gross value added in industries, construction, trade, trans- portation, financial services, and public services in current prices (TL million). YO5: Gross value added in housing sector in current prices (TL million). Y18: Gross value added in financial service sector in current prices (TL million).- ROX: Exchange rate as TL per US dollar. M: Imports at c.i.f..prices (TL million). R: Revenue in current prices (TL million). RR: Revenue in constant 1968 prices (TL million). ANNEX A Page 5 Simulation with the 1962 Tax Structure The dummy variable method permits us to estimate the revenue func- tions under abolished tax structures as well as those under the present structure. The result of a simulation indicates that if the 1962 tax struc- ture had been maintained, then the revenue in 1971 would have been TL 20.8 billion in 1971 prices. This is about 63.8 percent of the actual 1971 rev- enue, TL 32.6 billion. The 1962 tax revenue was TL 13.5 billion in 1971 prices. It is clear that the revenue growth in 1963-1971 is largely due to the changes in the tax structure, including the introduction of new taxes. Reveriue Projection for 1972-77 Table A.2 presents the projection of tax revenue in 1972-77 under the 1971 structure. Except for income tax and the new taxes introduced by the Financing Law of 1970, the figures are the direct results of the extra- polation of the revenue functions. The simulation assumed a real annual GDP growth rate of 7 percent for 1972 and 7.8 percent for the Plau period. The annual rate of price inflation is assumed to be 10 percent for 1972-1977. As mentioned earlier, income tax revenue in 1971 was affected poi- tively by the Personnel Law of 1970. Since the share of the public service sector gross value added would not decline in the near future,,the revenue from this tax was projected by .applying the revenue growth rate implied by the elasticity to the preceding year's revenue, taking the actual 1971 rev- enue as the base. 1/ For the taxes introduced in 1970, the revenue observations are available only for 1971. Hence, the revenue functions for these taxes can- not be estimated with regression techniques. Only additional information is the Ministry of Finance revenue budget forecast for 1972, which is sup- posedly in 1972 prices. As a tentative basis for the projection for.these taxes, we approximated the real revenue elasticity with respect to real GDP by ln(RR1972 / RR1971) In(GDPR1972 / GDPR 1971) 1/ The elasticity with respect to GDP of 1.38 implies that the nominal revenue growth rate associated with a given nominal GDP growth rate, g, will be (1 + g)1.38 ANNEX A Page 6 where RR 1972 is the 1972 revenue budget forecast for these taxes after elim- inating the expected additional revenue from the 1972 tax measures, RR 1971 is the actual 1971 revenue, and GDPR 1972 / GDPR 1971 is assumed to be 1.07. The calculated value of b is 0,21. The figures in Table A.2 have been ad- justed for price inflation. As noted earlier, the projections in Table A.2 do not reflect the effects of the 1972 tax measures. Compared with the revised official esti- mate for 1972, our projection deflated to 1971 prices shows a shortfall of TL.3.1 billion.' For 1977, our projection in 1971 prices is almost TL 10 billion lower than the Plan projection. The difference may be due to the effect of the 1972 tax measures, which is included in the Plan projection. As Table A.3 shows, our projection indicates that the real revenue growth rate during the Third Plan period will be 64.5 percent as against 76.8 percent in the Plan projection. This implies that the Plan expects the 1972 tax measures to improve the revenue elasticities as well as bring about additional revenue. Estimate B in Table A.3 assumes that the effect of the 1972 tax measures is limited to increasing revenue; that the revenue elasticities do not change. It may be noted that in the Plan the revenue growth rate for 1974- 1977 is projected to remain constant at 11.8 percent, which implies a con- stant revenue elasticity of about 1.49 for this period. In our projections, the elasticity of the total revenue is not constant with respect to GDP; it is projected to follow an increasing trend. This is because our assumption is that the revenue elastitity of each class of tax is constant (not' neces- sarily with respect to GDP), and therefore the elasticity of the total rev- enue with respect to GDP is a weighted average of those of individual classes of taxes. Since the revenue from GDP-elastic taxes will increase faster than the revenue from others as GDP grows, the total revenue tends to become more elastic over time; in the long run, the total revenue elasticity will .con- verge to that of the most GDP.-elastic taxes. However, even in 1977 the implied total revenue elasticity with respect to,GDP is 1.44 according to our projection. Annex A PMe 7~ Table A.2: Re-en-e Projectir for 1972-1977Y -(iLn TL r.illion, cturrent prices) 1972 1973 1974 1975 1976 1977 Income Tax 12,001 15,233 19,335 24,542 31,51 39,5404 Corporation Tax 2,293 2,985 3,803 4,844 6,167 7,847 Fiscal E;lance Tax-2/ 1,518 1,932 2,459 3,129 '3,983 5,070 Defense Tax on Buildings h28 543 689 872 1,105 1,398 Motor Vehicles Tax 138 162 189 222 259 304 Inheritance and Gift Taxes and Real Estate Purchase Tax 953 1,274 1,702 2,276 3,042 4,066 Production Taxes 5335 6,290 7,458 8,842' -10,484 12,430 Petroleum Production Tax 3,175 3,310 3,451 3,598 3,751 3,911 Monopoly Revenue and Taxes 3,421 4,068 4,838 5,753 6,842 8,137 Sugar Consumption Tax 613 798* 777 921 948 1,081 Banking and Insurance Transactions Tax 1,973 2,h38 3,013 3,723 4,600 5,684 PTT Service Tax and Transpor'ation Tax 374 433 501 580 671 776 Stamp Tax 1,1488 1,861 2,326 2,908 3,636 4,546 Custcrs Duty and Stamp Duty on Imports 3,319 3,508 3,708 3,919 4,142 4,378 Other Taxes 79 888 1,053 1,249 12482 11758 Stel 37,71,9 h,723 55,298 67,319 82,263 100,926 New Taxes2/ 996 1,113 12h.4 392 1Ll7 1,7L1 Total 38,745 46,836 56,543 68,771 83,820 102,667 Total Revenue as Percent of GDP 18.9 19.3 19.7 20.2 20.7 21.4 1/ Revenue under the 1971 tax structure. 2/ Projected on the basis of revenue function for savings bonds scheme. 3/ Taxes collected in 1971 for the first time. Annex A Table A.3: Revenue -Projections in 1971 Prices, 1972-1977 Estimate B Plan Projection Estimate A Est. A Partial2y Revenue Under 1972 Revenue Under 1971 Adjusted for 197k Tax Structure Tax Structure/l.. Tax Measures A Growth Growth. Grcwth TL Billion Rate(%) TL Billion Rate(/) TL Billion Rate(%) 1972 38.3L3 17.6 35.2 8.1 38.3 17.6 1973 43.3 13.0 38.7 9.9 42.1 9.9 1974 48.4 11.8 42.5 9.8 46.2 9.8 1975 54.1 11.8 47.0 10.5 51.1 10.5 1976 60.5 11.8 52.0 10.8 56.7 10.8 1977 67.7 11-.8 58.0 11.4 63.1 11.4 Plan Poiod. (1973-1977) 274.0 76.8 238.2 66.5 259.3 66.5 /1 Total in Table A.2 deflated with the assumed rate of' price inflation; 10 percent per annum. /2 Estimate incorporating the effect of 1972 tax measures but assuming they will not affect the revenue growth rate. /3 Revised official estimate as of August 1972. ANNEX B Savings Ratios and Factor Income from Abroad The following symbols.will be used to describe variables as conventionaaly defined in national accounts: YN = Gross National Product YD = Gross Domestic Product FI = Net Factor Income from Abroad WN. = Gross Natibnal Savings SD = Gross Domestic Savings C = Consumption sn = ratio of gross national savings to GNP sd = ratio of gross domestic savings ,to GDP fn = ratio of factor income to GNP fd - ratio of factor income to GDP The symbol (*) will be used to distinguish variables that are not to be found in conventional national accounts but are, in principle, observable. ad* = the "true" proportion of GDP that is saved sf* = the "true" proportion of FI that is saved SD* - savings out of GDP SF* = savings out of FI The standard national accounts definitions) in terms of the above symbols,are: IN = YD + FI . . . . (Bl) S% = IN - C . . . . . . . . . . . (B2) SD = YD - C . . . . . . . . (B3) S = S= S+FI .. . (B . sd =SD .0 (B5) Throughout the following discussion, it will be assumed that there is a net inflow of factor income from abroad i.e. FI ;O. !roposition I The gross national savings ratio is always higher than the gross domestic savings ratio, the difference being greater, the higher the proportion of aNP repre- sented by net factor income from abroad. Proof of Proposition I is evident from equation (B6) which follows directly from simple manipulation of (Bl) to (B). sn sd+(1-sd)fn ... (B6) ANN EX B Pr]0oosition II . The gross domestic savings ratio underesti- mates the true proportion of dbmestic income that is saved, unless all net factor income from abroad is saved. Conventional national accounting methodology assumes that all FI is saved, as is evident from a comparison of (B4) and (BS). This assumption is the direct result of the definition of national and domestic savings, as shown in (B2) and (3). If one allows for the fact that, in reality, some part of FI is savad, the gross domestic savings ratio can be written as: ad SD* + (FI - SV*) . . . . . . . . . . . . (BT) .YD and, hence, ad = ad& - (1 - sf*)fd .............(B8) The proof of Proposition II is evident from (B8). PrptionjII The extent to which the gross domestic savings ratio underestimates the true domestic savings will increase as net factor incomes from abroad increase; however, the rate of increase will diminish as the marginal propensity to save out of factor income approaches unity. Proof of this proposition is evident from equation (B9) which follows directly from (B8). &)sd - d) . 1 - F . . . . . . . . ... . (B9) ANEX B Page3 Brpostion IV The ratio of gross national savings to GNP will increase over time if (a) factor incomes from abroad increase faster than GDP and (b) t4e marginal propensities to save from GDP and factor income are each greater than their respective average propensi- ties. This is a sufficient but not necessary condition. This proposition can be proven by differentiating equation (BB) with respect to time. The resulting expression can be simplified and -written as (BlO). de .YN =dYS (dD (-Usn)+4SF*f-nY dt ut- ddlD / -r + JdFI - dlD) d,* - an . . . . . . (B0) . L d3t /dI We note that if FI grows faster than YD, then dFI is greater than dYD; that dSD* and dSF* represent the "true" marginal propensities dt dlD dFI to save; and that sn is a weighted average of the average savings propensi- ties ad and sf. The necessary and sufficient condition for sn to increase over time is that the right hand side of equation (B1O) should be positive. Provided that condition .(a) holds, a sufficient condition for an to increase over time is that the expressions in the square brackets are positive. Condition (b) is a sufficient condition for this. Hence Proposition IV is proven. 1 .1 Appendix Table 1: TURKEY: EXPENITURE ON GROSS NATIONAL PRODUCT, PLAN TARGETS AND ACTUALS, 1963 TO 1972 (N of GNP at current market prices) First Plan Target Second Plan-Target 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 Conspt!ion 87.0 86.2 85.1 84.2 83.4 81.3 80.5 79.4 78.0 77.4 Public 15.2 3Ii9 11.7 15.2 .15.5 3T. -iT.3 -I.T -.9 15.3 Private 71.8 71.3 70.4 69.0 67.9 67.1 66.2 65.0 63.1' 62.1 Total Investment 17.0 17.9 18.3 19.0 19.4 20.6 21. 22.6 23.8 24.3 Changes in Stocks - - - - - -.1. 4 -1:V Gross Fixed Investment 17.0 17.9 18.3 19.0 - 19.4 19.2 20.1 21.2 22.4 22.9 Public (10.3) (10.9) (11,0) (11.3) (11.4) (10.9) (11.2) (11.7) (12.4) (12.3) Private (6.7) (7.0) (7.3) (7.7) (8.0) (8.3)- (8.9) (9.5) (10.0) (10.6) Current Account Surplus of Balance of Payments -4.1 -4.1 -3.4 -3.2 -2.8 -2.0 -2.0 -2.0 -1.8 -1.7 Total, 100.0 100.0 100 0 100.0 100.0 100.0. 100.0 100.0 100.0 100.0 Actuals Consumption 87.0 85.0 84.0 82.3 82.1 81.4 80.9 80.6 80.7 Public' T1.6 12.0. T - 11.9 12.2 12.0 12.2 15.1 Private 75.4 73.0 71.6 70.5 69.9 69;4 68.7 68.0 65.6 Total Investment 17.8 17.1 17.0 19.5 19.1 20.6 20.9 20.9 20.2 Changes in Stocks 2.6 2.2 1. -- 1.9 2.0 1.3 1.4 2.1 Gross Fixed Investment 15.2 14.9 15.4 16.7 17.2 18.6 19.6 19.5 18.1 Public (7.4) (7.8) (8.1) (8.9} (9.2) (10.2) (10.6) (10.3) (9.2) Private (7.8) (7.0) (7.3) (7.8) (8.0) (8.3) (9.0) (9.2)- (8.9) Current Account Surplus of Balance of Payments -4.8 -2.0 -1.1 -1.8 -1.2 . -2.1 -1.8 -1.5 ' -0.9 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Note: Items may not sum to total because of rounding. Source: First and Second Five Year Development Plans and -0.1mates by State Planning Organization. Appendix Table 2: TURKEY: SAVINGS: PLAN TARGET AND ACTUALS, 1963 TO -1972 (N of GNP at market prices) First Plan Target Second Plan Target 1964 1965 1966 1967 1968 1969 1970 1971 1972 Gross Domestic Savings n.a. n.a. n.a. n.a. n.a. 17.9 18.8 20.0 21.4 22.0 Net Factor Income from Abroad n.a. n.a. n.a. n.a. n.a. 0.7 0.7 0.6 0.6 0.6 Gross National Savings 12.9 13.8 14.9 15.8 16.5 18.6 19.5- 20.6 22.0 22.6 Actuals Gross Domestic Savings 13.3 15.2 15.5 17.0 17.5 184 18.8 1L7.8 17.3 n.a. Net Factor Income from Abroad -0.2 -0.2 o.4 0.7 0.3 0.2 0.3 1.6 3.0 n.a. Gross National Savings 13.1 15.0 15.9 17.7 17.8 18.6 19.1 19.4 20.3 n.a. Source: First and Second Five Year Development Plans and estimates by State Planning Organization.