Yugoslavia - Economic development (Vol. 4 of 6) : Volume 4

FEHLE COPY DOCUMENT OF INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT INTERNATIONAL DEVELOPMENT ASSOCIATION Not For Public Use Report No. 194a-YU THE ECONOMIC DEVELOPMENT OF YUGOSLAVIA (in six volumes) VOLUNE IV IX. Resource Mobilization and Allocation, 1947-71 X. Resource Mobilization and Allocation, Prospects XI. Stabilization: Institutions, Instruments and Policies XII. The Balance of Payments and Foreign Trade XIII. Capital Flows and External Debt XIV. Development Problems and Prospects November 26, 1973 Europe, Middle East and North Africa Region This report was prepared for official use only by the Bank Group. It may not be published, quoted or cited without Bank Group authorization. The Bank Group does not accept responsibility for the accuracy or completeness of the report.  CURRENCY EQUIVALENTS Before January 23, 1971 1 US dollar = 12.5 dinars 1 Dinar = 8 US cents From January 23, 1971 to December 22, 1971* 1 US dollar = 15.0 dinars 1 Dinar = 6.67 US cents From December 22, 1971 to February 23, 1972 1 US dollar = 17.0 dinars 1 Dinar = 5.89 US cents From February 23, 1973 to July 12, 1973 Maintained a central rate of 1 US dollar = 17.0 dinars 1 Dinar = 5.89 US cents Since July 12, 1973 The Dinar has been floating. The rate of July 12, was 1 US dollar = 15.50 dinars 1 Dinar = 6.45 US cents * All conversions of 1971 data into dollars in this Report have been made at this exchange rate.  - 242 - IX. RESOURCE MOBILIZATION ANT) ALLOCATION - 1947-71 Introduction 9.1 Changes in monetary and fiscal institutions and policies have con- stituted both a part of and a response to the progress of decentralization in Yugoslavia. During the era of rigid central planning, about 90 percent of domestic savings was mobilized and allocated through the National Budget. By 1971 it was the economic enterprises and households which accounted for over 90 percent of domestic saving formation. Parallel to the gradual increase in enterprise autonomv, the all-encompassing National Budget was decomposed into federal, republic and communal budgets and extrabudgetarv funds. Until the mid-1960s the State retained a major role in savings mobilization and alloca- tion through the system of social investment funds. It was not until these were abolished in 1963 that investment banking really came into its own, though the banking svstem for "short-term operations" had already developed a long way since the mono-bank days of central planning. With the abolition of social investment funds and certain taxes which financed them, and the devolution of their assets and liabilities to banks, the role of investment banking in resource mobilization/allocation was significantly increased. The reduction in the State's fiscal burden on enterprise income (roughly value added), coupled with the consolidation of enterprise autonomy in accordance with the 1963 Constitution, led to the increased importance of savings by enter- prices and households - the latter from increased personal incomes. With a much higher proportion of decentralized autonomous savings, the potential role of efficient financial intermediation has increased a potential which remains to be fullv developed. The relativelv slow development of financial mechan- isms and instruments was in part due to the inflation in the post-Reform pe- riod. The extension of market forces after the Reform was accompanied on the one hand bv strong investment demand arising from the low real interest rates, partial insulation of enterprises from financial risk and a desire to catch up technologicallv with Western Europe, while on the other, by a prolonged con- sumption boom fed by increased allocation to personal incomes from enterprise incomes and "demonstration effects" passed on by Yugoslav workers in Western Europe. These factors lent an inflationary bias to the Yugoslav economy and brought the problem of short-term demand management to the foreground. Selec- tive price control has not reduced these inflationary pressures -- if anything, it has increased the costs of inflation by distorting the relative price struc- ture. Monetary policy has had an impact but its efficiency has been limited bv a number of features including the partial insulation of enterprises from financial constraints by patron State units whether republics or communes. The fiscal manifold, composed of a multitude of budgetarv and extra-budgetary units, with independent revenue and expenditure authority, has proved to be a clumsy tool for coordinated demand management. 9.2 Thus the 1965-71 period, has been characterized by these twin foci of policy concerns:  - 2.3 _ (a) mobilization and allocation of resources accordin, to am- bitious Plan targets in a context of highly decentralized savings/investment decision; (b) the management of aggregate demand to encourage rapid eco- nomic growth without running into western-developed-economy- type "overheating" problems of inflation and/or excessive balance of payments deficits. The 1971 Constitutional Amendments have not materially altered these policy concerns. The further devolution of State economic power to republics and communes have, if anything, increased the problems of coordinating economic management. The Amendments further increased the share of social Product (gross material product) at the command of the enterprises. This, together with the fact that enterprises accounted for about half of total domestic savings in 1971, led to the institution of incomes policy agreements in each republic in an effort to stabilize, perhaps raise, the rate of enterprise savings. Thile these agreements may turn out to be powerful instruments for shoring up enterprise savings and weakening inflationary pressures, the main current concern on the resources side is with improving the mechanism for their allocation. Commercial banks, virtually the sole financial inter- mediaries, have a number of structural defects which suggest the need for reform or development of alternative modes of financial intermediation. Improvements in the financial mechanism are likely to be closely linked with the success of the present stabilization policies. Such improvements in re- source allocation appear to be most urgently needed in the less-developed republics. This chapter, and the one following, focus on questions of re- source mobilization and allocation. Chapter XI is concerned with issues in stabilization policy. Section A of this chapter briefly outlines the system of resource management during the central planning years. Section B covers the period 1952-63 describing the development of enterprise autonomy, banks and the new system of social investment funds, and the imiplications of these changes for the structure of savings in the economy. Section C describes the mid-1960's Economic Reforms and examines their im-pact in the post-Reforms period 1965-71, paying special attention to (1) trends and structure of sav- ings, (2) the new role of financial intermediation, and (3) the quality of resource allocation. A.. Central Planning 1947-51 9.3 In this period production was centrally planned and administratively controlled and enterprises merely acted as agents for carrying out the annual plans, which laid down production quotas, factor inputs to be used and rules for allocation of the product. The National Budget operated as a component of the central planning mechanism. In addition to the classical tax expenditure functions, the National Budget was the vehicle for appropriating and deploy- ing the surpluses of social sector enterprises. In this way the mobilization and allocation of savings and investment in the social sector was under full administrative control, leaving no intermediation role to the financial  - 244 - mechanism. Through taxes on incomie and turnover the fiscal systemn also chan- nelled some private sector resources into the social sector via the National Budget 1/. 9.4 The financial mechanism during this period was relatively simple. In effect, it consisted of a mono-bank system (the Niational Bank with its 400 odd branches) designed to provide money as a unit of account and means of payment. For the social sector, monetary policy was concerned mainly with meeting the transactions demand for money througlh short-term credits. 'lonetary policy did have a limited role in affecting expeniditure decisions in the private sector 2/. 9.5 Though the structure of administrative planning was subsequently dismantled, the process occurred at different speeds for different sectors. For example, administered prices continued to prevail in important segments of the economy until the end of the 1960's, with considerable impact on inter-sectoral and inter-regional resource allocation. B. Decentralization, 1952-63 9.6 The state retained considerable, though gradually diminishing, con- trol over the mobilization and allocation of resources throughout this period. Prices of goods and services in major segments of the economy renmained under state control. Until 1957, when the "Net Income System" was first introduced to enterprises, the state continued to regulate wages and salaries. By set- ting the rates of the "interest on business funds (capital) of enterprises" (a form of tax on socially owned capital) and the terms of lending from the new system of social investment funds, the state effectively decreed the cost of capital to enterprises. Though the planning mechanism was decentralized and "indicating" the state wielded enormous influence over sectoral choices thnrough the deployment of the social investment funds. Despite major changes in the system, of foreign trade and exchange, it remained the most tightly con- trolled se-ment of the economy. 9.7 However, the steps towards decentralization should not be minimized. After 1957 wages and salaries were increasingly determined by enterprises, which also took over an increasing role in voluntary savings formation. The banking mechanisn developed rapidly. Though banks were not allowed to engage in investment crediting until 1963, the foundations of the investment banking svstem were laid during, this period 3/. '/ There were further resource transfers t-rough fi.xin, relatIvely low pri- vate sector prices for products dlelivered to the social sector. 2/ Useful discussions of this role contnined in DiTPitrijevic and Macesicn 'oney and Finance inr Yu-orlavia, and lHauvonen. 3" See Chapter I.  - 245 - Social Investment Funds 9.8 The all-encompassing National Budget was decentralized vertically across the Federal, Republican and Communal governments, and horizontally (at each of these independent levels) across budgetary and extra-budgetary transactions. While budgetary transactions were concerned with fairly clas- sical public expenditure functions, the extrabudgetary operations included the new system of social investment funds. Unlike proper financial intermedia- ries, the social investment funds lent their resources but did not borrow them in the first place, thus creating no assets for primary savers. These semi- financial intermediaries relied on fiscal resources of various kinds 1/ (mainly the tax on owned capital of enterprises and contributions levied on the income of enterprises) which were then channelled to enterprises in the form of credits according to the economic plans of the relevant government unit. Some idea of the relative distribution of the resources of these funds across different levels of government may be obtained from Table 9.1. Table 9.1: ASSETS OF SOCIAL INVESTMENT FUNDS AT END 1963 (million dinars) Percent General Investment Fund 21,021 66.8 Republican Investment Funds 6,304 20.0 District and Communal Investment Funds 4,158 13.2 Total 31,483 100.0 Source: Sekulic, loc. cit. 9.9 The Social investment funds were not legally independent agencies. Their assets were managed by banks in conformity with plans of govermment units. For example, in the case of the General Investment Fund, the sectoral distribution of resources was determined by priorities in the annual Federal plans, leaving the allocation to specific enterprises up to the managing banks (mainly the Yugoslav Investment Bank in this case). The terms of cre- dits were agreed upon with the Federal Executive Council. Parallel proce- dures existed for investment loan funds at other levels of government. Thus the banks acted largely as agents of the state units, with little control over regional and sectoral allocation of funds, although they gradually took over the important role of supervising the terms and conditions of each individual credit. However, even toward the end of the period, state influence in the 1/ For a fuller description of thiese funds are Sekulic "Investment Capital of Socio-Political Communities," Yugoslav Survey, Nov. 1970.  - 246 - allocation of these investment funds often reached to individual credits, which were granted on socio-political rather than economic grounds 1/. 9.10 Two other kinds of investment funds were created at Republican and Communal levels during this period: (a) Housing Funds (1955) fed from levies on the income of enterprises and deployed to finance housing projects in the social sector; (b) Joint Reserve Funds (1962), again financed through compulsory contributions of enterprises, and used to credit enterprises in temporary difficulties. In both cases the appropriate state unit had the most say in the allocation of these resources. By 1963, all the investment loan funds numbered nearly two thousand. The Changing Role of Enterprises 9.11 Enterprise command over resources increased significantly after 1957 and by 1963 their contribution to voluntary savings formation was sizeable. Though the share of enterprises in total domestic savings was already 29 per- cent in 1953, most of it renresented compulsory savings according to rules decreed by the State. In 1957, with the adoption of the "Net Income System," the base was laid for growing worker control over enterprising accumulation decisions. In fact, the system underwent significant changes after 1957, especially in 1969. But the principle of treating the payments to labor as a part of "net. income" rather than an element of operating costs remained as the distinguishing feature of all variants of the system. 9.12 Table 9.2 presents the functional breakdovm of gross revenues earned by an enterprise according to the "Net Income System" after the definitional changes adopted in 1969. 2/ It is clear from the table that: f/ See Sukulic op.cit. and Hauvonen op.cit. In the early 1950's there was an attempt to allocate these investment credits on market criteria through "investment auctions." The attempt was abandoned largely be- cause the profitability incentives embedded in the distorted price structure indicated sectoral choices which were at variance with planned priorities, and the "irresponsible" approach of enterprise in making bids. 2 The main differences with the variants prevailing before were: (i) Some of thte "contractual obligations" in the post-1969 system were formerly charged against operating cost; (ii) After 1966, depreciation consisted of a newly-instituted legal minimum component plus a voluntary accelerated com- ponent.  - 247 - Table 9.2: THE NET INCOME SYSTEM I. Gross Revenue II. Operating Costs III. Social Product (Gross Value Added = I-II) IV. Depreciation (Legal + accelerated) V. Net Income A (Net Value Added = III-IV) 1. Obligations A. Contractual Obligations (a) Interest and service charges on bank credit (b) Insurance premia (c) Contributions to chambers of commerce, etc. B. Statutory Obligations (a) Interest on "business fund" - abolished in 1971 (b) Turnover tax (c) Legally imposed contributions for water, land, housing (d) Other legal obligations 2. Net Income B A. Gross Personal Incomes (a) Net Personal Incomnes (take-home wages) (b) Contributions from personal incomes including those for education, social insurance, etc. B. Enterprise Funds (allocations to) (a) Reserve Fund (b) "Business Fund" (c) Welfare Fund Source: Adapted from Gorupic and Paij, Workers' Self-Management in Yugoslav Undertakings. the enterprise did not have full autonomy on the allocation of Net Income A because of contractual and statutory obligations and certain mandatory deduc- tions from gross personal incomes. A better estimate of the distributional authority of the enterprise is obtained by adding together "net personal in- comes", "enterprise funds" and "depreciation" (to be called residual enter- prise income, REI) and comparing it to the social product generated in the enterprise 1/. Between 1958, when enterprises were first put on the "Net Income System", and 1963, the share of gross value added under enterprise control showed a mild upward trend (Table 9.3). The major increases in this share occurred later, with abolition of social investmnent funds in 1963 and 1964 and other tax changes associated with the 1965 Economic Reforms. 1/ The comparison is preferred in terms of gross (oi depreciation) concepts because depreciation included varying proportions of voluntary and legally fixed elements during the last two decades.  - 248 - Table 9.3: SHARE OF ENTEP3RISE REI IN GROSS VALUE ADDED (percent) 1958 1959 1960 1961 1962 1963 44 47 48 50 49 49 Source: Based on data from Statistical Yearbook of Yugoslavia) Federal Institute of Statistics. Trends in the Structure of Savings Table 9.4: STRUCTURE OF TOTAL SAVINGS, 1950-63 (Percent) 1950 /a 1953 1958 1963 State 90 54 48 45 Federal 90 47 25 18 Government (n.a.) (0) (0) Investment Loan Funds (n.a.) (25) (18) Other State Units 7 23 27 Government (n.a.) (6) (11) Investment Loan Funds (n.a.) (17) (16) Enterprises 29 42 35 Households and Private Producers 10 9 9 Other (including Unclassified) 7 1 11 Total 100 100 100 100 Decentralized Savings Share /b 10 53 75 82 Voluntary Savings Share la 10 19 33 33 N4S as Percent of CTP 24 31 30 35 la Estimates from Dimitrijevic and 1.asecic /b Decentralized - non-federal. Sources: Annual Reports of the N.*ational Bank of Yugoslavia, Dimitrijevic and Macesic op. cit. and mission estimates of national income.  - 249 - 9.13 The impact of these institutional innovations on the mobilization of savings is shown in Table 9.4. First, there was a significant increase in voluntary savings emanating from changes in enterprise control over value added. Secondly, there was a dramatic increase in the share of "decentralized savings" reflecting both the "returning of resources to the economy" by giv- ing increased autonomy to enterprises and the devolution of fiscal authority within the state sector towards republics and communes. Consequently, after the initial structural changes bewteen 1950 and 1953, the share of the state in savings declined much less than the share of the Federation because the role of non-Federal State units increased. Finally, the importance of social investment funds during the entire period is clearly brought out; they ac- counted for 42 percent of total savings in 1958 and 34 percent in 1963 1/. The Development of Banks 9.14 The banking system developed rapidly after 1955, when the mono-bank system of the central planning era was dismantled. Table 9.5 summarizes the banking institutions existing at the end of 1963. 9.15 Banks were not allowed to mobilize time-deposits or engage in invest- ment credits until 1963. But the experience gained during this transitional period in managing the investible resources of the social investment funds, provided a valuable foundation for the pivotal role of investment banking in resource allocation after 1963. 9.16 An important nonbank financial institution, the Social Accounting Service, was created in 1962 (previously it had been an arm of the National Bank) to police the conformity of all transactions with the various earmarking provisions prevailing, and to provide a comprehensive accounting and statisti- cal service. 1/ Tle sectoral breakdown of savings is based on the Money Flow accounts of the National Bank. Strict comparability with Yugoslav national accounts data (prepared independently) is not assured, but broad trends are prob- ably accurate. Both the money flow data and the Yugoslav national ac- counts on income and product include an element of capital gains on inventory, which is consistent with usual national income accounting practice. While a correction for this has been incorporated in the mission's national income estimates (see Statistical Note) and subse- quent discussion of savings, this correction, positively correlated with inflation, was small until 1963, and has been omitted from Tables 9.3 and 9.4 above.  - 250 - Table 9.5: BANKING SYSTEM AT END 1963 Year Bank Established Purpose Federal National Bank (including 6 Republican offices) Central Bank operations. Yugoslav Investment Bank 1955 Manage resources of the General Investment Fund and long-term foreign credits. Yugoslav Bank for Foreign Trade 1955 Provide credit for for- (Jugobanka) eign trade operations. Yugoslav Agricultural Bank 1959 For credits to socialized agricultural sector. Postal Savings Bank 1959 Accepts savings deposits, but places its resources at the disposal of National Bank. Republic Republic Banks (6) 1961 Nianage investment policy of Republic government. Communes Communal Banks (220) 1955 Engage in short-term com- mercial operations and manage communal social investment funds. Sources: Hauvonen op. cit. and Dimitrijevic and Macesic op. cit. C. Decentralization: 1964-71 9.17 1965 is usually singled out as the year of the Economic Reform. In fact, the changes in institutions and policy which were important in the monetary and fiscal fields occurred over 1963-66. The major changes which occurred during this period were: (a) abolition of the social investment funds, and the devolution of their assets and liabilities to the banks (mainly and to government bodies; (b) abolition of some of the taxes which financed  - 9R1 - these funds; (c) strengthening of enterprise autonomy over the disposition of value added; (d) passage of new laws regulating the formation and function of the banking system. These changes and their impact on mobilization and allocation of resources and stabilization policy are discussed below. (1) Trends and Structure of Savings 9.18 The institutional changes between 1963 and 1966 altered the distri- bution of national income in favor of enterprises and households, and away from the state. This had two fundamental consequences. First, gross nation- al savings (GNS) as a proportion of GNP declined significantly from 33 per- cent in 1960-63 to 29 percent in 1967 (Table 9.6) implying that the marginal propensity to save of enterprises and households was lower than that of state units at whose expense their incomes had increased. Second, the institution- al changes dramatically altered the sectoral structure of savings. Table 9.6: TRENDS IN AGGREGATE SAVING AND INVESTMENT (percent shares; current prices) 1960-63 1966 1967 1968 1969 1970 1971 /a 1. GNS/GNP 33.2 32.6 29.3 27.8 28.2 28.4 27.7 2. GDS/GDP 32.7 32.0 28.8 27.2 27.1 26.4 24.7 3. NFI/GNP -0.2 -0.1 0.2 0.4 1.0 2.2 3.4 4. GDI/GDP 35.5 33.0 30.3 29.0 29.4 31.6 31.0 5. (GNS-GDI)/GNP /b -2.3 0.0 -1.0 -1.0 -1.0 -2.4 -2.2 /a 1971 estimates are provisional. lb Deficit (-) on the current account of the balance of payments. Source: Mission estimates of national income, Statistical Annex Table 2.1 to 2.4. 9.19 The decline in the rate of aggregate savings between 1960-63 and 1967 should not, by itself, be viewed as a reflection of poor performance. Increasing the share of consumption in national product was a deliberate ob- jective of the Reform 1/. After this initial decline the GNS ratio remained 1/ This does not mean that a trend decline in the aggregate savings rate may be viewed with equanimity. Given ambitious investment targets in the 1971-75 Social Plan, stabilizing the rate of aggregate saving at the 1970 level is important. See Chapter X.  - 252 - remarkably stable until 1970, indicating an impressive performance in main- taining the high savings ratio of 28-29 percent. However, the ratio of gross domestic savings (GDS) to GDP declined further after 1967. As an index of savings performance, the GDS ratio is misleading in a context of large and growing net factor income from abroad (NFI) - after 1967 Yugoslavia bene- fited from a surge of worker remittances from abroad 1/. 9.20 The institutional changes also dramatically altered the sectoral structure of savings. Table 9.7 presents the sectoral composition of domes- tic savings according to two sets of data. The unadjusted ratios are based on current price data as available in Yugoslav statistical sources. The ad- justed ratios incorporate an inventory evaluation adjustment (also present in the aggregate savings data) 2/. The Yugoslav National Bank's money flow accounts (on which tables 9.7, 9.10, 9.16, 9.17, and 9.18 are based) have been periodically revised. As a result, the data prior to 1963 are not fully comparable to that for subsequent years. But conclusions based on broad trends remain sustainable. During 1973 a new revision was undertaken. The resulting changes for earlier years were very minor, and since the details of these changes were not available at the time of revision of this report, the older figures were allowed to stand. The 1972 data, compiled only on the new basis, is therefore not completely comparable to earlier years. 1/ The national income accounts for Yugoslav (and for many other countries) estimate consumption as a residual, by subtracting independently esti- mated investment from available resources (=GDP plus the deficit on goods and nonfactor services account in the balance of payments). This esti- mate of consumption is appropriate for the GNP concept of income. GNS, which is defined as GNP minus consumption, correctly measures the coun- try's savings effort from available income (GNP). However, GDS, which is defined as GDP minus consumption makes the unsupported assumption that all of net factor income from abroad is saved. The GNS figures in Table 9.6 also include net current transfers from abroad. Though much less than net factor incomes, this item, which is excluded from GNP does bias the GNS ratio upwards slightly as a measure of savings performance. However, as a ratio of GNP, there was no observable trend in this item, implying that its inclusion in GNP does not bias the trend in savings performance. 2/ Briefly, the Yugoslav data on product and income include an element of unrealized capital gain on inventory, which is inconsistent withi na- tional income accounting methodology. To achieve consistency and comparability, this element has been estimated and excluded from the income and saving figures (see Appendix I).  - 253 - Table 9.7: STRUCTURE OF SAVINGS (percent) A. Unadjusted Data 1960-63 1967 1968 1969 1970 1971 L972 State 36 23 24 20 15 3 5 Federal IT 11 10 7 (a) Government (-9) (2) (2) (1) (-3) (-4) (b) Tnvestment Loan FRnds (23) (9) (8) (6) (7) (0) Other State Units 22 12 13 13 10 7 (a) Government (8) (4) (4) (5) (4) 2 (b) Investment Loan Fands (14) (8) (9) (8) (6) (5) Economic Ebterprises 45 49 52 45 45 58 49 Other Social Sector Organizations a! -6 9 o 1 Households and Private Producers 9 26 26 31 27 3 3 Unclassified -11 3 TOTAL 100 100 100 100 100 100 lOC B. Adjusted Data b/ State 39 24 24 22 18 4 Federal T 11 IT 7 - - (a) Government (-8) (2) (3) (0) (-3) (-4) (b) Investment Loan Funds (24) (9) (8) (7) (9) (0) Other State Units 23 13 13 15 12 6 (a) Government (8) (4) (4) (5) (4) (3) (b) Investment Loan Funds (15) (9) (9) (10) (8) (5) Economric Enterprises 42 48 51 38 3 46 Other Social Sector Organizations a/ 6 I 9 12 1 10 Hiouseholds and Private Producers 10 2 27 37 33 1T Unclassified 3 -11 -7 3 TOTAL 100 100 100 100 100 100 GNS/GNP (current prices) 33 29 28 28 28 28 GDS/G,DP (current orices) 32 20 _T 77 7 27 a/ Includes social insurance organizations and banks in their capacity as wor'king organizations, not intermediaries. b/ Adjusted for inventory evaluation (see Appendix I: Statistical Note). Sources: Annual Repts, Yugoslav National Bank and mission estimnates (see Statistical Appendix, Table 6.2). The unadjusted data (A) are based on the Money Flow accounts of the National Bank. Strict comparability with Yugoslav national accounts data (prepared independently) is not assured, but broad trends are probably accurate.  - 254 - Both sets of data reveal two important features: (i) the sharp decline in the State's role in savings mobilization, associated with the abolition of the social investment funds. From an average share of 38 percent in 1960-63 (36 percent for the unadjusted data), the State's share fell to 18 percent in 1970 (15 percent for the unadjusted data) 1/. Though spread over all levels of the State, the decline was swiftest at the Federal level, due mainly to low or negative Federal budget current surpluses; (ii) corresponding increases in the share of households and private producers and enterprises. While the share of households in- creased steadily, the enterprise share declined after 1968, reflecting a fall in the rate of enterprising savings froml income at their command. (See para 9.21.) The adjusted data show lower shares for enterprises than the unadjusted data. This is because the inventory valuation adjustment (downward) is concentrated on the income and savings of this sector 2/. Enterprise Savings 9.21 Between 1963 and 1965 the system of social investment funds was largely abolished, along with some of the taxes on enterprise income, which had financed these funds. As a consequence, the proportion of gross value added under the distributional authority of the enterprise increased fron an average of 47 percent in 1960-63 to an average of 58 percent in 1967-71 3/. 1/ The 1971 figures reflect changes due to the Constitutional Amendments, and are discussed in Section V. 2/ The size of the adjustment, and hence, the discrepancy between the two sets of data is positively related to the rate of inflation during the year, the source of the unrealized capital gains on inventory. 3/ These figures from Statistical Appendix, Table 6.3 refer to data ad- justed for inventory evaluation. The corresponding percentage for unadjusted data are 51 and 61, respectively.  - 255 - Table 9.8: ENTERPRISE SAVING RATE OUT OF "RESIDUAL ENTERPRISE INCOME" /a (percent) 1960-63 1964 1965 1967 1968 1969 1970 1971 A. Unadjusted Data 42 45 47 43 42 37 37 41 B. Adjusted Data lb 38 33 34 42 41 31 28 32 /a Enterprise Savings = Depreciation + allocations to "Enterprise Funds" ?'Residual Enterprise Income" (REI) = Enterprise Savings + Net Personal Receipts of Workers. /b A = Ratios based on current price data as available in Yugoslav statis- tical sources. B = Adjusted for inventory evaluation. Sources: Statistical Yearbook of Yugoslavia and mission corrections for in- ventory valuation. (See Statistical Appendix, Table 6.5,) The increase in proportion of income under the distributional authority of the enterprise was accompanied, particularly after 1968, by a significant decline in the rate of enterprise savings out of "residual enterprise income" (REI) (Table 9.8.), suggesting a tendency of workers to vote increasing proportions of personal income distribution to themselves. This trend is accentuated in the adjusted series 1/. The apparent improvement in savings performance by enterprises in 1971 is partly illusory. As a result of the Constitutional Amendments the "interest on business funds" - tax 2/, which was an extra- budgetary revenue source to the Federation, was abolished. But enterprises were instead required to grant loans to the Federal Fund for Underdeveloped Regions. This replacement of a tax by a compulsory loan, shifted the savings to the enterprises' accounts, but such savings do not represent a voluntary increase. Savings of Households and Private Producers 9.22 The rapid post-1965 growth of savin-s by households and private pro- ducers, in both absolute and relative terms, was a direct result of the Econo- mic Reform. The Reform increased the share of net personal receipts in value 1/ Note that this observation contrasts strongly with the assertions of stable enterprise savings performance in the "Korac study." Aside from the inventory adjustment factor, the differencesaarise from choice of differing concepts of enterprise income and saving. Those adopted here seem more appropriate. 2/ The same levy is sometimes referred to as the "tax on business capital" in the Yugoslav literature.  added in both the social and private sectors of the economy. (Table 9.9.) In the social sector this increase was due to the rise in the share of REI in gross value added resulting from the reduction in the tax burden coupled with an initially unchanged, and subsequently declining, enterprise savings rate out of REI 1/. In the private sector the increased importance of per- sonal receipts in value added may be ascribed to the Reform-induced rise in relative prices of agricultural products, which increased private farm incomes significantly. Table 9.9: SHARE OF NET PERSONAL RECEIPTS IN SOCIAL PRODUCT (percent) 1960 1963 1967 1970 Social Sector 23.1 28.2 35.7 37.0 Private Sector 72.0 71.4 77.7 80.5 Total Economy 38.9 37.9 45.2 45.0 Source: Statistical Yearbook of Yugoslavia, Federal Institute of Statistics. The surge in workers remittances from abroad after 1966 and the increased activity in the private services sector also contributed to the growth of disposable incomes of "households and private producers." This increase in the share of GNP accruing to this sector was the main factor behind both the increased share of consumption (lower aggregate savings) in GNP after 1966, and the rapid increase in the share of gross savings done by "households and private producers." Depending on whether one chooses the adjusted or unad- justed data, the share of this sector in total domestic savings increased to either 44 or 35 percent respectively, by 1971, from a pre-Reform base of about 10 percent. 9.23 In judging the savings performance of this sector the ratio of sav- ings to net personal receipts is not an adequate indicator. The inclusion of private producers in the savings data requires the expansion of the corres- ponding income concept to include gross value added in the private sector, not just net personal receipts. Social insurance benefits should also be included as well as workers remittances from abroad. The most readily available and inclusive income concept is the total current receipts of this sector as defined in the National Bank's money-flow tables, from which the savings data originate 2/. This ratio (Table 9.10) rose fron average of 7 percent between 1960-63 to a range around 15-17 percent after the Reform. This performance is particularly encouraging given (a) the relatively poor financial incentives 1/ For definition of "residual enterprise income" (REI) see Table 9.8. 2/ It is subject to the criticism of being gross of production costs in- curred by private producers.  - 257 - for saving that existed during this period 1/ and (b) the wide ranging social insurance provided by the state which dilutes the precautionary motive in sav- ings. While no adequate investigation of the determinants of household savings in Yugoslavia has yet been conducted, it seems plausible that income and the rate of change in income are important determinants 2/. Table 9.10: RATIO OF SAVINGS TO "INCOME" BY HOUSEHOLDS AND PRIVATE PRODUCERS 1960-63 1967 1968 1969 1970 1971 7 16 15 17 15 17 Source: Annual Reports, National Bank of Yugoslavia. (2) Financial Intermediation 9.24 The changing pattern of savings generation was accompanied by parallel changes in the institutions for resource mobilization and alloca- tion. The reduction of the State role and the growing share of decentra- lized autonomous savings by non-State units substantially increased the po- tential role of the financial mechanism (institutions, instruments and rules) in sustaining an adequate level of savings and securing efficient intermedia- tion of this decentralized savings to appropriate investing units. With the abolition of the social investment funds, the burden fell on the newly re- formed banking structure. In 1963 the General Investment Fund was abolished and its assets (from past credits) and liabilities (from incomplete commit- ments) were transferred to the three specialized all-Yugoslav banks, chiefly the Yugoslav Investment Bank. In the next two years Republican and Communal social investment funds were abolished and a similar transfer of their assets and liabilities to the corresponding banks in their territory took place 3/. The Rise of Investment Banking 9.25 The swift replacement of the State by investment banks in channeling investment resources is revealed in Table 9.11. While the share of the State in total domestic finance of fixed investment fell from 60 percent in 1960-63 to 16 percent in 1970, that of banks increased from 3 percent to 51 percent. 1/ The record of financial saving was, as a result, less encouraging. 2! A draft study of the Capital Market, prepared by a team of Yugoslav offi- cials and economists suggested these and other hypotheses, but failed to provide empirical discrimination. This is clearly an important area for future study. 3/ See Sekulic 2. cit. for details.  - 25R - Table 9.11: DOMESTIC SOURCES OF FINANCE FOR FIXED INVESTMENT /a (percent) 1960-63 1964 1965 1966 1969 1970 1971 1972 1. Economic Organizations 30 26 29 39 28 27 27 30 2. Other Social Organi- zations 7 6 8 7 6 6 7 8 3. State Finance 60 36 27 15 16 16 15 20 (a) Federal (33) (7) (3) (6) (9) (9) (7) (2) (b) Republic (8) (8) (4) (3) (3) (2) (4) (14) (c) Communes (19) (21) (20) (6) (4) (4) (4) (4) 4. Banks 3 32 36 39 49 51 51 42 Total 100 100 100 100 100 100 100 100 /a Includes all investments financed out of social resources plus private investment financed through bank credit. Source: Statistical Bilten, various issues, 1972 and 1973. 9.26 For banks to cope with their new investment role, reform was neces- sary. The pre-1963 banks were largely creatures of the government units within whose territory they were obliged to operate. In nearly all cases the banks had been established by these state units, which also appointed their management. To adapt banks to their new expanded role, the Banking and Credit Law (March 1965) was passed. The new legislation was designed to accomplish three main goals: (a) decrease the influence of government units and increase that of enterprises in the allocative decisions of banks, and thus, hopefully expand the role of economic criteria; (b) empower banks to operate across the entire country, and to take on the new task of mobilizing fresh investment resources through savings deposits; and (c) stimulate merger of the large number of communal banks into a smaller number of more efficient units (now termed "business banks"). 9.27 Table 9.12 demonstrates rapid fulfillment of the third goal, and points up the rising dominance of "mixed" (commercial plus investment) banks.  - 259 Table 9.12: NUMBER OF BUSINESS BANKS AT THE END OF YEAR 1963 1966 1967 1968 1969 1970 Communal Banks 220 - - - - - Commercial Banks - 62 54 28 22 - "Mixed" Banks - 40 40 36 36 55 Investment Banks /a - 9 9 9 9 9 Total 220 111 103 73 67 64 /a These are the transforms of the three Federal specialized banks and the eight Republican banks which existed at the end of 1963 (see Table 9.5). Source: Dimitrijevic and Macesich 2p. cit. The number of banks is a little misleading. In 1963 the 10 largest banks ac- counted for nearly 60 percent of short-term credits, while nearly all invest- ment credits were concentrated in 10 banks. 9.28 However, the de-etatization of banks was diluted by the State's continued influence over the allocation of the resources transferred from the abolished social investment funds. In 1966, the State units, especially the Republican authorities exercised their right to withdraw most of these trans- ferred resources (usually referred to as "State Capital") from the credit funds of banks. W4hile a small part of the withdrawn resources was retained by the state units for extra-budgetary investment in the economy, the bulk was relent to the banks in the form of credits for the financing of specified projects or purposes. In this way the State retained considerable control over the allocation of "State Capital" 1/. Even after the withdrawal of Federal "State Capital" from the banks' balance sheets in 1969 about 30 per- cent of the business banks' stock of investment resources (liabilities) were accounted for by the item "credit for investment financing," which consisted mainly of funds earmarked by government units. (Table 9.13.) 1! In October 1969, there was a further change, when the Federal Government decided to allocate this "State Capital" in its own nam.e. The assets and liabilities corresponding to the earlier arrangements whereby the govern- ment credited the banks which, in turn, credited the final year, were shifted from the balance sheets of the banks to special Federal accounts (see Annual Report, 1969, National Bank of Yugoslavia). Despite the formal balance sheet changes, the banks continued to administer these resources on a commission basis, resulting in no effective change in the allocative procedure.  - 260 - Table 9.13: BALANICE SHEET OF INVESTh1bTT OPERATIONS OF BUSINESS BANTKS AT END OF YEAR a/ million dinars Percent 1969 1970 1971 1969 1970 1971 Assets (Uses) Investment Credits 66.3 83.5 99.4 70.8 70.6 70.9 Housing Credits b/ 17.3 21.9 26.1 18.5 18.5 18.6 Investment Credits in Foreign Ecchange 2.4 5.3 6.5 2.6 4.5 4.6 Unused Sources 4.1 3.7 3.0 4.4 3.1 2.1 Other 3.6 3.9 5.2 3.7 3.3 3.8 Total 93.7 118.3 140.2 100.0 100.0 100.0 Liabilities (Sources) Credit Funds (equity) 13.8 14.4 14.3 14.7 12.2 10.1 Time Deposits c/ 20.1 25.9 28.5 21.5 21.