Report No. 7208-MA Malaysia: Matching Risks and Rewards in a Mixed Economy (In Three Volumes) Volume ii: Selected Topics Odober 7, 1988 Country Operations Division Country Department II Asia Region FOR OFFICIAL USE ONLY Document of the World Bank T;his report has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS Currency Unit - Ringgit (M$) M$1.0 = US$0.39 US$1.0 = M$2.59 (as of May 1988) FISCAL YEAR January 1 - December 31 ACRONYMS AND ABBREVIATIONS CIC - Capital Issues Committee CICU = Central Information Collection Unit DOS - Department of Statistics EPF P Employee Provident Fund EPU = Economic PlAnning Unit FELDA = Federal Land Development Authority 5MP = Fifth Malaysia Plan FIMA = Food Industries of Malaysia FTZ = Free Trade Zone GDP Gross Domestic Product GNP Gross National Product HICM = Heavy Industrial Corporation of Malaysia ICU = Implementation Coordination Unit IMP = Industrial Master Plan KLSE Kuala Lumpur Stock Exchange KTM = Malayan Railway LPN = National Paddy and Rice Authority MARA - Majlis Amanah Rakyat MDB = Manpower Development Board MIDA = Malaysian Industrial Development Authority MIPS 5 Malaysian Industrial Policy Studies MOF = Ministry of Finance MPE - Ministry of Public NEP - New Economic Policies NWPE - Non-Financial Public Enterprises PERNAS = Perbadanan Nasional Berhad PETRONAS = Petroliam Nasional Berhad PNB = Permodalan Nasional Berhad RISDA = Rubber Industry Smallholders Development Authority SADC = State Agriculture Development Corporation SEDC = State Economic Development Corporation SLCHP - Special Low Cost Housing Program VAT = Value Added Tax FOR OMCLIL USE ONLY MALAYSIA COUNTRY ECONOMIC MEMORANDUM Volume II Table of Contents Page No. Is NONFINANCIAL PUBLIC ENTERPRISES ........ o......oo.ooo. o.. 1 A. Size and Quantitative Dimensions.....e.s.o..o.oo.o...ooso 1 Public Enterprises Definedo....................0.0..00 1 Sector Sizeios.oo..oo.ooo....oeoo.ooo..ooeo. ee..ooo 4 Growth of the NFPE Sector.e...oooo.ooo.oo..oooo.o..e.. 14 B Perfot..ceooooooooo.oeo.oo. o.oe..eoe....ooooooo oo.oooooe 18 Introductionooooe*oooooooeee*oooo 18 Trends in Number of Profitable Firmso..........&ooo. 19 Trends in Profitabilityofooooooitaoboilitey..o....o... 20 Analysis of Loss-Making Firms..0...0....000.00..0.0.0. 26 Estimate of Welfare Loss Due to NFPEs.... ............ 31 Reasons for Poor NFPE Perfor mance.o.o...oooe..oeooooo. 31 C. Organizational Environment.o....oooo.o......*oooo***o.s 36 Breakdown of NFPEs by Ministry........e.o..o...o.o.. 36 Parent Corporationor..................... o n0..0.0.00.. 39 Board of Directors..oiooor oooe e too.oo.....o..oooorso.. 40 Privatization.... eg ....... g........g0.00000g.. .... 40 D. Improving the Efficiency of NFPEs..................E...... 45 Reform of the Control Structure..g..g..ee......e..g.e 45 Development of Information System stem................. 46 Performance Evaluation Systems. .oo .. ........ g.e... 47 Incentive Sytm. 49 Board of Directors..... ......e. ge...... ... ........ 50 [ ~~~ ~~Divestitureeeooo**e**oooeooooeoos 50 Chapter 1 Appendices Appendix 1 - Companies Identified for Data Collection by CICU... 52 Appendix 2 - List of Statutory Bodies ........................... 53 Appendix 3 - Regression Analysis of IFPE Profitabilityo.... o .. 55 Appendix 4 - Projects Being Considered for Privatizatione.o..... 57 Appendix 5 - List of NFPEs that Have Been Sold.... .............. 59 Appendix 6 - CICU's Z-score Model of Performance Evaluation..... 61 This document has a restricted d stribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. - ii - Page No. II. FISCAL DEFICITS IN MALAYSIA ......... ........................... 64 A. Introduction5.............. . 64 B. The Fiscal Deficit: Definition and Quantification........ 65 Public Sector Revenues and Expenditurees............... 65 Three Measures of Fiscal Deficits ...................... 65 Fiscal Imbalance: 1970-87 ..................... ..... 67 The Primary Deficit............... .........**...... 70 C. Fiscal Policy in Malaysia: 1972-87.7 2 - 87................. 70 Measurement of the Fiscal Impul pe.......s e.......... 70 Disaggregating the Net Fiscal Impulse...*060......-.. 72 Adequacy of Fiscal Policy.oe l i cy..............o..... 78 Subsidies and the Fiscal Deficit...................... 80 D. Public Investments and Govcrnment Subsidies ..............*. 83 Statutory Authorities and Fiscal Deficits0.... ..0** 89 E. The Financing of a Sustainable Deficit: Tax Reform ...... 91 Financing the Deficit: 1971-87 ....................... 91 Fiscal Policy Goals.................................. . 95 Public Revenues .................................... 95 Proposed Tax System Reforms........................... 96 Dependence of Tax Revenues on Trade and Natural Resources................................... 96 Narrow Domestic Tax Base.............................. 99 Taxation of Companies and Fiscal Incentives........... 106 User Charges and Self-Financing for Statutory Authorities and NFPE.......0.0.00.0....... ***00000** 113 Cha.pter II Appendices Appendix 1 - Technical Notes.................................... 116 Calculation of Net Fiscal Impulse ................ 116 Derivation of Implicit Subsidies................. 121 Financing the Real Deficit and Macroeconomically Sustainable Deficits........... ................ 119 Tax Bnxoyancy Estimates........................... 123 Appendix 2 - List of Federal Statutory Bodies and Government 124 -iii - Page No. III. TEE LABOR MARKET: ISSUES IN THE DEVELOPMENT AND EMPLOYMENT OF HUKA. RESOURCES........................................ 129 Ao Introduction ......***........... ................. 129 S. Macroeconomic Perspectiveo................................. 130 Labor 8upy..................................132 Profile of the Unemployed............................. 133 Changes in the Patterns of Unemployment............... 141 Hours of Work and Employment Structure................ 142 C. Labor Absorption by Industry.............................. 145 D. The Functioning of the Labor Market....................... 149 Labor Mobility and Market Segmentation................ 149 E. Wage and Employment Adjustments............................ 152 Wage Policies in the Public Sector.................... 152 Productivity and Wage Adjustments..................... 153 Quantity Adjustments.................................o0 157 F. The Average Wage in a Two-Tier Labor Market ........ ....... 157 Go Industrial Rltos................................. 161 The Employee Provident Fund and Other Labor Compensation Laws............................ ....e.. 162 H. Manufacturing Labor Costs and International Competitiveness ..............o......o.....o......e..ee.....o...o... 164 1. Government Measures Against Unemployment.................. 166 J. Education and Training. ...................................o 168 The Social Productivity of Higher Education........... 168 Training and Public Sector Programo................... 171 K. Policy Recommendationso.................................. . . 173 Policy Framework .. ......................... .... 173 Recommen. ations...eo..... o... .......o............... 176 Tables 1.1 Distribution by Status 1.2 Number of NFPEs in Malaysia 1987 1.3 Government Equity and Number of NFPEs 1.4 Sales of NFPEs in CICU Data Base: 1986 1.5 Value Added of Public Enterprises in Malaysia: 1986 1.6 NFPE Shares in GDP at Factor Cost 1.7 State-Owned Enterprises' Share of Value Added in Manufacturing - iv - 1.8 Taiwan's Public Enterprise Sector 1.9 NFPE Shares in Gross Fixed Capital Formation 1.10 NFPE External Debt 1.11 Growth of Public Ente.s-ises 1.12 Percentage of Profitaule Companies: Total NFPE Sector 1.13 Percentage of Profitable Companies by Industrial Sectors for 1981-86 1.14 Average Net Profitability by Sector 1.15 Summary of the Performance of Government-Owned Companies 1.16 Summary of the Performance of Government-Owned Companies with Paid- up Capital of at Least M$5 Million 1.17 Summary of the Performance of Government-Owned Companies with Paid- up Capital Below M$5 Million 1.18 Relative Profitability Under Private/Public Ownership 1.19 Relative Costs Under Private/Public Ownership 1.20 Distribution of Firms by Net Profits: 1986 1.21 Major Noncorporate Loss-Making NFPEs 1.22 Distribution of Loss-Making Firms 1.23 Investment Returns in Public and Private Sectors 1.24 NFPE Projects with Socio-Economic Objectives 1.25 List of Companies Under the Ministry of Public Enterprises Transferred to Permodalan Nasional BHD 1.26 Ministries Controlling NFPEs 1.27 Companies Under the Ministry of Finance, Incorporated 1.28 NFPE Distribution by Agency 1.29 Projects which have been Privatized 1.30 Projects which have been Approved for Privatization 2.1 Transparent Subsidies in Malaysia 2.2 Public Investment in Malaysia, 1978-87 2.3 Budgetary Position of Statutory Authorities, 1985-87 by Ministry of Supervision 2.4 Fiscal Deficits of Statutory Authorities, 1985-87 2.5 Tax Revenue in Malaysia, 1982-87 2.6 Regression Estimates for Categories of Taxes, 1970-87 2.7 Regression Estimates for Sales Taxes, 1977-87 2.8 Regression Estimates for Buoyancies of Excise Taxes Logarithmic Equations, 1977-87 2.9 Marginal Effective Tax Rates for a Standard Investment Project in East Asian Countries 2.10 Estimates of Subsidy of Major Tax Incentives Over Various Recipients of Tax Holidays, ITA, or Non-Recipients 2.11 Revenue Foregone Through Duty Exemption on Machinery and Equipment: 1983-86 3.1 Total Labor Force, Employment, Unemployment Rate and Output 3.2 Labor Force Growth, 1980-90 3.3 Ratio of Age-Specific Unemployment Rates Relative to the Average of all Age Groups 3.4 Percentage Distribution of Unemployed Persons by Education Level: Malaysia 3.5(a) Education Specific Unemployment Rates Relative to the National Average 3.5(b) Annual Growth Rate in Education Specific Unemployment Rates 3.6 Percentage Distribution of Unemployed Persons by Duration of Unemployment 3.7 Percentage of Distribution of Unemployment by Monthly Household Income Class 3.8 Relative Unemployment by State and Level of Development 3.9 Comparative Unemployment Rates: Male/Female and Urban/Rural 3.10 Percentage Distribution of Employed Persons by Industry and Hours Worked per Week 3.11 Percentage Distribution of Employed Persons by Employment Status and Industry 3.12 Level and Growth of Employment by Industry 3.13 The Relative Contribution of Industry to Employment Generation 3.14 Consumer Price Indexes in Malaysia, 1975-82 3.15 Compensation per Worker in Manufacturing and Government Relative to the National Average 3.16 Distribution of New EPF Members Registered in 1986 and 1987 by Age and Salary 3.17 Minimum Wage Offers for Workers with Less Than Three Years Experience 3.18 A Synopsis of Alternative Efficiency Wage Theories 3.19 Percentage Growth Rate of Wages, Employment and Labor Productivity in Manufacturing 3.20 Employment Value Added Elasticity 3.21 Wage and Output Elasticities of Employment Using Normalized Regressors, 1968-85 3.22 Manufacturing Unit Labor Cost 3.23 Private and Social Internal Rates of Return on Human Capital Formation in Malaysian University System by Race and Type of Graduates 3.24 Budget for Recurrent Expenditures: Higher Education Vocational/ Technical, and MOE Total 3.25 Percentage Distribution of Respondents (Industrial Training Institutes) by Status of Employment, 1981-84 3.26 Figures 1.1 NFPE Investment 2.1 Consolidated Revenues and Expenditures of the Public Sector 2.2 Current and Federal Deficits Compared 2.3 Net Fiscal Position of Monitored NFPEs as a Percent of Overall Federal Deficits 2.4 Primary and Overall Deficits 2.5 Net Fiscal Impulse, 1972-87 2.6 The Tax Impulse, 1972-87 2.7 The Expenditure Impulse, 1972-87 2.8 The Net Fiscal Impulse and Cyclically Neutral Deficits 2.9 Federal Subsidies, 1979-87 2.10a Total Federal Expenditures and Estimates of Subsidies and Transfers, 1970-87 - vi - 2.10b Subsidies and Transfers as Ratio to Total Public Expenditure and to cGP, 1970-87 2.11 Public Investment by Level of Government 2.12 Composition of Public Investment 2.13 Financing the Real Fiscal Deficit, 1911-87 (Percent Distribution from Source of Financing) 2.14 Distribution of Revenues, 1982-1987 I. NONFINANCIAL PUBLIC ENTERPRISES 1.1 Malaysia's nonfinancial public enterprises (NFPEs) account for almost a third of all economic activity within the country. This figure is higher than that of any other country in the region and three times higher than the average for most countries. In Malaysia, public enterprises operate in almost all sectors of the economy. They extract petroleum, gas and other oils; manufacture automobiles, steel, cement, bicycles and textiles; operate plantations, restaurants, hotels, golf courses and travel agencies; run air- lines, shipping lines and bus services; construct buildings and boats; culti- vate cocoa, palm oil, pineapples, cashews, asparagus and mushrooms; act as car dealers, insurance agents, property developers, investment brokers, bankers and landscapers. And do much more. 1.2 Public enterprises have been used as a response to market failure; for achieving income distribution goals and restructuring of society; for the pursuit of socioeconomic objectives; and for the creation of a heavy indus- trial base. Their growth has been rapid, and, once set up, they have acquired a dynamic of their own. Relatively autonomous corporations and holding companies have for a variety of reasons expanded and diversified. Often this expansion has been in areas not envisioned at the time of their establishment and with little comparative advantage for public ownership. As a consequence, they have probably displaced private investment and have certainly added to the external debt and the fiscal burdens of the Government. 1.3 As long as the economy was doing well, resources for the prolif- eration of NFPEs were provided ungrudgingly--even enthusiastically. But as overall macroeconomic growth slows, the burdens of the massive public enter- prise sector are beginning to acquire new transparency. There is an emerging consensus on the need for curtailing the growth of the sector, for improving its efficiency and for evolving a coherent set of policies toward the achieve- ment of these objectives. 1.4 To assist the Government in designing appropriate policies for the sector, this study of the NFPEs provides an understanding of the true size of the sector, its impact on the macroeconomy, and its oragnizational environ- ment. Based on issues identified in the foregoing sections, the study con- cludes with some suggestions for improving the efficiency of the NFPEs, which, it is hoped, will be useful. The following section lays the groundwork for the later analysis by defining what an NFPE is, determining the magnitude of the sector and how it compares to the public enterprise sectors of other countries, and identifying the factors which have caused the sector to grow. A. Size and Quantitative Dimensions Public Enterprises Defined 1.5 There seems to be little coasensus on the definition of a public enterprise (PE). Different countries have adopted their own schemes for clas- sifying PEs, thus complicating any attempt at international comparison and generally introducing ambiguity in estimates of the size of the sector. - 2 - 1.6 In perhaps the first systematic attempt at measuring the size of the PE sector, Jones defined a FE as follows: A Public Enterprise is a productive entity that is owned g d/or controlled by public authorities and whose output is marketed._ For the purpose of this study, the underlined terms are defined as follows:2/ (a) A "productive entity" is an identifiable decision-making unit with an explicit or extractable budget which produces goods and/or services. (b) "Ownership" refers to share capital being held by the government either directly or through other public entities. A minimum of 10% ownership of shares by the government would 3ualify the enterprise for inclusion in the public enterprise set.- 1/ Leroy Jones, Public Enterprise and Economic Development: The Korea Case, Korea Development Institute, 1975. For a fuller discussion of the issues involved in defining public enterprises, see the various articles in Praxy Fernandes, ed., Seeking the Personality of Public Enterprise. 2/ These definitional criteria conform with earlier studies on the public enterprise sectors of various countries. In addition to Jones's study on Korea, see A.H. Shaikh, "Public Enterprise Sector of Pakistan: Evolution, Performance and Policies: 1971-85," World Bank Report, 1987. 3/ Sometimes a 51% ownership of shares by the government is considered a criterion for defining a public enterprise. The logic for including firms with less than 50% government ownership in the public enterprise set is that if the government controls the management of the enterprises, then their behavior is likely to be similar to other public enterprises irrespective of the percentage of shares owned by the government. There- fore, in this report, a PE has at least 10% government ownership and effective government control of enterprise management. (c) "Control" is defined as involvement in the management of the enter- prise. The primary criterion for 4e exercise of control is the ability to appoint top management.- (d) Output is defined as "marketed" if the value of sales coyf;r at least 50X of the current costs of operation of the enterprise._ 1.7 This definition has two advantages: first, it is likely to become the standard for international comparisons, and thus future analysis can be more meaningfully undertaken. Second, and perhaps more important, the criterion of 102 government equity is being rsed by the Malysian Government in its own efforts to collect information on the PE sector._' This informa- tion is being collected and computerized by the Central Information and Collection Unit (CICU), recently established within the Permodalan Nasional Berhad (PNB), to gather data on all corporate public enterprises (NFPEs and financial PER) in the country. Corporate public enterprises are those regis- tered under the Corporation Act. CICU collects general, financial and socio- economic information, based upon the size of the enterprise. It currently has information on 867 PEs. 41 There are many ways iu which governments control enterprises, both within the public and the private sectors. The instruments of control include tax/subsidy policies, grant of licenses for scarce goods such as foreign exchange, frequent and continuous discussion, and in some cases the threat of nationalization. It is, however, the ability to hire and fire the top management that is the primary and most readily observable source of effective control by the government. In the case of Malaysia, the departmental and statutory bodies which function as NFPEs are directly under government control while in the case of corporate public enter- prises the government exercises influence through its ability to appoint the Board of Directors. 5/ The "marketedness" test divides public enterprises from public institu- tions which are productive entities but rely mainly upon government sub- sidy for survival. In the context of Malaysia, some statutory bodies like Bank Negara and various universities are unlikely to meet the market test and would be counted as public institutions. However, corporate entities with public ownership and commercial statutory bodies like the Electricity Board and Port Commissions are included as public enter- prises. 6/ Sensitivity analysis is carried out in para. 1.10 to determine the impact that alternative definitions of equity would have on the size of the sector. -4- Sector Size 1.8 The following discussion considers the size of the NFPE sector in terms of the number of firms, share of value added in GNP, investment, external debt, etc. Where relevant, international and intertemporal compari- sons are provided. 1.9 Number of NFPEs. According to CICU, there were 867 corporate public enterprises in Malaysia as of September 30, 1987. However, this number needs to be adjusted for five factors as follows: (a) Stat s of Firms. Many of the 867 firms listed by CICU are not active. As shown in Table 1.1, 679 firms are in active production and 8 are in active preproduction. Thus the minimum estimate of public enterprises (including both financial PEs and NFPEs) active in 1987 would be 687. (b) Undercoverage. Twenty-one firms have been identified that meet the criteria for inclusion in the public enterprise set, yet are not currently in the CICU data base (Appendix 1). While other firms may also need to be included, this provides the lower bound. (c) Noncorporate NFPEs. Statutory bodies of the federal and provincial governm-nts are excluded from the CICU data set because they are not corporate entities. Of the 76 statutory bodies identified (Appendix 2), about 30 meet the requirements for inclusion in the PE set. These include the Malaysian Timber Industry Board, the National Electricity Authority and the various Port Commissions. In fact, the Government includes 11 statutory bodies in the list of 56 large NFPEs chosen for close monitoring due to their size and importance. (d) Firms with Less than lOZ Government Equity. CICU data include 43 firms with government equity of less than 10%, although paid-up capital is greater than M$250,000. These 43 have been subtracted from the total (they have thus been a-sumed to be active). (e) Financial Public Enterprises. Since the focus here is on the nonfinancial public corporate sector, 78 financial enterprises included in the CICU data base have also been deleted from the set of PEs meeting the definitional criteria. The above adjustments are presented in Table 1.2. The total number of NFPEs in Malaysia after all the adjustments is 617. -5- Table 1.1: DISTRIBUTION BY STATUS Status Number of firms Active preproduct ion 8 Active production 679 Ceased operation 19 Closed 7 Dormant 30 Liquidation 32 Preoperation 28 Receivership 12 Shell 12 Suspended 40 Total 867 Table 1.2: NUMBER OF NFPEs IN MALAYSIA 1987 Total CICU 867 Moribund status (-) 180 Undercoverage (+) 21 Statutory bodies (+) 30 Equity less than 10% (-) 43 Total Public Enterprises 695 Active financial public enterprises (-) /a 78 Total NFPEs 617 /a There are actually 85 active financial public enterprises, but the 7 with government equity of less than 10Z have already been deducted. Note: This excludes the 13 State Economic Development Corporations and 3 State Agriculture Development Corporations. 1.10 How sensitive are these results to a definition of NFPEs that is based upon a 10% government equity ownership? Table 1.3 gives a breakdown of firms by percentage of equity. About 44% of the 560 NFPEs for which data on equity are availabl, iave a public ownership share of more than 50%. Howe &, the sales value of t .se NFPEs is much greater than 50% due to the large siz. of many of the NFPEs with majority government ownership. -6- Table 1.3: GOVERNMENT EQUITY AND NUMBER OF NFPEs Z equity No. of NFPEs Cumulative no. 90 - 1OOX 96 96 80 - 902 21 117 70 - 80Z 42 159 60 - 70Z 47 206 50 - 60% 109 315 40 - 50Z 51 366 30 - 40Z 59 425 20 - 30Z 70 495 10 - 201 29 524 0- 101 36 560 Source: CICU. 1.11 Size: Value Added of NFPEs. The second dimension of the size of the NFPE sector is the share of value added generated by the NFPEs as a proportion of total GDP. This indicates the overall importance of the NFPEs in the economy, and has been estimated as described below. 1.12 Step One: Calculation of Sales. Sales figures were obtained for active firms in the CICU data base. The aggregate figures are based upon a total of 587 NFPEs for the year 1986. Figures for the remaining 30 firms are not available, and the results therefore have a downward bias. The magnitude of the bias is, however, likely to be small due to the skewed distribution of firms, the availability of information on the large firms, and the inclusion of some firms with less than 101 equity in the CICU data set. Table 1.4 gives the sectoral composition and the total value of sales (M$26.22 billion) for the 587 NFPEs. - 7 - Table 1.4: SALES OF NFPEs IN CICU DATA BASE: 1986 No. of firms Value Sector No. of firms with data Sales added -- (M$ million) -- Agriculture 71 64 739 429 Construction 76 65 678 373 Extractive 27 25 8,502 5,116 Manufacturing 236 204 9,057 3,246 Services 216 183 4,023 3,031 Transportation 49 46 3,327 1,877 Total NFPEs 676 587 26,227 14,072 Financial 86 72 13,388 Total PE 762 659 39,614 Source: Staff estimates. 1.13 Step Two: Adjusting Sales to Calculate Value Added. The sales figures were adjusted to determine the value added for the sector as follows: an updated version of the 1978 input-output table gives the gross output and value added for different sectors. Sectoral value added by NFPEs was calcula- ted by multiplying the NFPE sales by the ratio of value added to gross output of that sector given in the input-output table. The value added for all NFPEs in the CICU data set was estimated to be M$14.07 billion. 1.14 Step Three: Adding Value Added of Non-CICU NFPEs. The value added of NFPEs not included in the CICU data base, i.e., the statutory bodies, and for which data are available from other sources, was then added to the esti- mate of corporate sales. Information on profits, wages and depreciation is available for only 11 of the NFPE statutory bodies. The total value added for the 11 NFPEs is M$1.81 billion. Since there are other statutory bodies that meet the requirements for inclusion in the NFPE set but for which the needed data are not available, this constitutes a source of downward bias in the estimates. However, this bias is probably not significant since some statu- tory bodies are holding companies (MARA, RISDA) and the corporate entities operating under them are included in other data. 1.15 The value added of Telekom (M$1.16 billion), which was recently converted from a government department NFPE to a corporate NFPE (see para. 1.81), was also added to the total to get the overall value added for the NFPE sector. 1.16 Step Four: Estimates of Value Added. Based on the above adjust- ments, the total NFPE value added in 1986 is estimated at M$16.98 billion. - 8 - Given a GDP of M$71.14 billion, NFPEs contributed about 241 to the econo B (Table 1.5). Assuming that PEs comprise 60-80Z of thfinancial sector,- the share of all public enterprises in the GDP is 29-31Z.-/ Table 1.5: VALUE ADDED OF PUBLIC ENTERPRISES IN MALAYSIA: 1986 (M$ million) Sales Value added Corporate NFPEs (587) 26,227 14,072 Statutory bodies (11) 3,296 1,806 Telekom (1) 1,559 1,157 All NFPEs (599) 31,082 17,035 CDP 71,144 NFPE/GDP 23.9Z 1.17 Further evidence of the domination of the economy by the public sector is indicated by its share of the nonagricultural GDP. Since the share of NFPEs in agriculture is negligible and agriculture accounts for about 211 of GDP, the share of the public enterprise sector in the nonagricultural GDP, assuming 80Z public ownership in the financial sector, approaches almost 401. 1.18 Comparison with Other Countries. How does the size of Malaysia's NFPE sector compare with that of other countries? According to a World Bank Report, the average share of the NFPE sector in the GDP for a sample of 45 countries is 10Z.7' (The average for developing countries tends to be slightly higher than that of the developed countries.) Since the propensity for public ownership is higher in the mining industries of developing 7/ These estimates are based on the proportion of PEs in the financial sec- tors of three mixed economies, Pakistan (68Z) South Korea (83Z) and India (84Z). See Leroy Jones, "Operational Efficiency of Manufacturing Public Enterprises in Pakistan," World Bank Report, 1981. 8/ Financial sector value added in 1986 measured in constant 1978 prices is M$5.16 billion. Multiplying by the inflator of 1.23, the estimated cur- rent price value added is M$6.35 billion. Assuming 60% public ownership, the PE value added is M$3.81 billion. This, added to the NFPE value added of M$16.98 billion and divided by total GDP, indicates that PEs contribute 29.2% of output. However, if PEs are assumed to comprise 80Z of the financial sector, then the total share of PEs in GDP is 31%. 9/ World Development Report, 1983. - 9 - countries, those countries dependent on mineral-based industries tend to have even larger NFPE sectors. This holds true of Malaysia which has one of the largest NFPE sectors in the world. 1.19 Table 1.6 gives a breakdown of countries by the contribution of the NFPE sector to GDP. For most countries, the figures reflect the late seventies but are unlikely to have changed (except perhaps for Bangladesh, which already appears low, and Chile where the figure is for 1982 and already reflects the impact of a major denationalization). Given the differences in definitions and coverage, comparisons are qualified, but the large size of the Malaysian NFPE sector is apparent nonetheless. With a 24Z share of GDP, Malaysia's NFPE sector is in the upper middle range, clcser to Nicaragua, Egypt and Guinea and much higher than the other East Asian countries which tend to have NFPE shares below 10. 1.20 Table 1.7 compares Malaysia's NFPE share in the manufacturing sector with the corresponding share of 24 other developing countries. The 24X share of the NFPEs WMalaysian manufacturing puts the country in the middle of the distribution .- However, most of the countries with higher shares than Malaysia (Romania, Hungary, Burma, Ethiopia, Tanzania, Syria) are socialist countries which have extensive public ownership for ideological reasons. 1.21 A comparison of Malaysia with another East Asian economy, Taiwan, China, highlights the large share of the Malaysian public sector relative to some of its neighbors (Table 1.8). In Taiwan, the share of public ownership in total industrial production declined from 571 in 1952 to 162 in 1985. During the same period, the share of public ownership in manufacturing declined from 56Z to 12%. Thus, while Taiwan was cutting its NFPE sector dramatically, Malaysia was increasing its sector equally dramatically. 1.22 NFPE Investment. The trend in NFPE investment is summarized in Figure 1. Starting from only M$156 million in 1971, investment by NFPEs peaked at M$7.8 billion in 1984, then declined to about M$4.0 billion by 1986. Given the macroeconomic environment, investment is likely to decline further. 1.23 The rise in NFPE investment has probably crowded out at least some private investment. The analysis of investment in Volume 1 reveals that pub- lic investment in infrastructure may be complementary to private investment, but that noninfrastructure public investment displaces private investment. NFPEs have been the principal recipients of noninfrastructure investment in manufacturing, services and other sectors. (In a dynamic context, if the efficiency of NFPE investment is lower than private investment, the long-term growth of the economy will be slowed down.) 10/ NFPE value added (VA) in manufactauring was divided by total manufac- turing VA. The total manufacturing VA for 1986 was estimated by inflat- ing the 1978 constant price VA with the general price index. - 10 - Table 1.6: NFPE SHARES IN GDP AT FACTOR COST Country X Country % Industrialized Countries France 17.0 Italy 7.5 Austria 14.5 Ireland 8.0 United Kingdom 10.9 Denmark 6.3 Federal Republic of Germany 1.0 Spain 4.1 A_stralia 9.4 Netherlands 3.6 Devo C tr Algeria 66.0 Portugal 9.7 Hungary 63.5 Kenya 8.7 Zambia 37.8 Argentina 8.6 Guyana 37.2 Philippines 8.0 Nicaragua 36.0 Benin 7.6 Egypt 31.4 Sierra Leone 7.6 Venezuela 27.5 Mexico 7.4 Tunisia 25.4 Botswana 7.3 Guinea 25.0 Liberia 6.8 Malaysia 23.9 Greece 6.1 Senegal 19.9 Korea 6.4 Chile 17.7 Pakistan 6.0 Tanzania 12.3 Bangladesh 5.7 Bolivia 12.1 Turkey 5.0 Togo 11.8 Thailand 3.6 Ivory Coast 10.5 Paraguay 3.1 India 10.3 Nepal 3.0 Sri Lanka 9.9 Mali 9.4 Sources: Peter Short, UNIDO, IIAS, and Bank files. - 11 - Table 1.7: STATE-OWNED ENTERPRISES' SHARE OF VALUE ADDED IN MANUFACTURING Country Percent Greece 2 Argentina 6 Portugal 11 Romania 92 Mexico 30 Hungary 83 Panama 4 Korea, Rep. of 14 Turkey 20 Syria 56 Tunisia 57 Ivory Coast 22 Nicaragua 20 Thailand 8 Egypt 60 Bolivia 7 Senegal 18 Pakistan 9 Sierra Leone 12 Tanzania 34 India 14 Burma 58 Ethiopia 60 Bangladesh 41 Malaysia 24 Sources: UNIDO; World Bank; Peter Short (1983). - 12 - Table 1.8: TAIWAN'S PUBLIC ENTERPRISE SECTOR Industral rduction Manufacturing Period Total Private Public Total Private Public 1952 100.0 43.4 56.6 100.0 43.8 56.2 1953 100.0 A4.1 55.9 100.0 44.1 55.9 1954 100.0 47.3 52.7 100.0 50.3 49.7 1955 100.0 48.9 51.1 100.0 51.3 48.7 1956 100.0 49.0 51.0 100.0 51.7 48.3 1957 100.0 48.7 51.3 100.0 51.3 48.7 1958 100.0 50.0 50.0 100.0 52.8 47.2 1959 100.0 51.3 48.7 100.0 54.8 45.2 1960 100.0 52.1 47.9 100.0 56.2 43.8 1961 100.0 51.8 48.2 100.0 54.7 45.3 1962 100.0 53.8 46.2 100.0 57.7 42.3 1963 100.0 55.2 44.8 100.0 59.4 40.6 1964 100.0 56.3 43.7 100.0 61.1 38.9 1965 100.0 55.2 41.3 100.0 63.2 36.8 1966 100.0 56.3 38.2 100.0 66.7 33.3 1967 100.0 58.7 34.7 100.0 71.2 28.8 1968 100.0 61.8 31.1 100.0 75.3 24.7 1969 100.0 65.3 29.4 100.0 77.3 22.7 1970 100.0 68.9 27.7 100.0 79.4 20.6 1971 100.0 79.5 20.5 100.0 84.7 15.3 1972 100.0 80.9 19.1 100.0 86.0 14.0 1973 100.0 81.1 18.9 100.0 86.2 13.8 1974 100.0 80.4 19.6 100.0 85.9 14.1 1975 100.0 81.2 18.8 100.0 85.8 14.2 1976 100.0 79.8 20.2 100.0 84.8 15.2 1977 100.0 79.5 20.5 100.0 84.2 15.8 1978 100.0 80.7 19.3 100.0 85.1 14.9 1979 100.0 81.0 19.0 100.0 85.3 14.7 1980 100.0 81.3 18.7 100.0 85.5 14.5 1981 100.0 82.1 17.9 100.0 86.0 14.0 1982 100.0 82.0 18.0 100.0 86.1 13.9 1983 100.0 82.6 17.4 100.0 86.7 13.3 1984 100.0 83.6 16.4 100.0 87.6 12.4 1985 100.0 83.9 16.1 100.0 88.0 12.0 Source: Statistics Yearbook, Taiwan, Republic of China. FIGURE 1: NFPE INVESTMENT 8- 7 r - 5 - 4 3 w 2- 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 0 NFPE Investment - 14 - 1.24 Table 1.9 shows NFPE investment as a proportion of gross capital formation for a set of countries including Malaysia. Gross investment by Malaysian NFPEs in 1986 was an estimated M$6.0 billion. This is calculated by adding the net increase of fixed assets during the year of the largest 56 NFPEs (M$tif billion) plus the depreciation on their fixed assets (M$1.9 billion).- Given a gross capital formatioa of M$18.8 billion for the Malaysian economy, the NFPEs' share was 32%. This is about three times the size of the developed country average and is greater than the 27% average for developing countries. The zhree neighboring countries included in the sample, Korea, Philippines and Thailand, all have NFPE shares in gross capital forma- tion well below that of Malaysia. 1.25 External Debt. NFPEs have emerged as majnr borrowers in the exter- nal markets. This is likely to have an important impact upon the overall external debt position of the economy in the future. Table 1.10 shows gross borrowing, net borrowing and outstanding debt of the Government, the private sector and the NFPEs during 1981-86. Over this period, the NFPEs' share in gross borrowing doubled from 16Z to 33%, and their share in total net borrow- ing rose from 15% to 53%. The stock of outstanding debt of the NFPEs in 1986 stood at M$14.86 billion, or 28% of all outstanding debt. 1.26 The phenomenal growth in the external debt of the NFPEs has increased the overall dependence of the economy on external resources. Per- haps even more alarmingly, many of the projects that borrowed heavily in international markets (e.g., Sabah Gas, Perwaja) have not been performing well and are unlikely to improve their capacity for repayment in the near future. dlore likely, since many of these loans are guaranteed by the Treasury, the latter will have to intervene, and the burden of low economic efficiency of investment, changes in market conditions, weak debt service capacity and lack of managerial skills will ultimately be borne by the Government. Growth of the NFPE Sector 1.27 Since no time series on the value added of the NFPE sector is available, the growth of the PE sector in Malaysia was estimated on the basis of the date of commencement of the PEs in five-year intervals (Table 1.11). The proliferation of PEs began in the early seventies and continued strongly during 1975-85. At least 453 (more than 50%) of all PEs were created during this period. While the largest number of firms were initiated in the manufac- turing and services sector, the pattern of government intervention appears to be uniform and widespread throughout the economy. In addition to the factors that typically result in the establishment of PEs, like market failures and natural monopolies, several factors unique to Malaysia prompted the explosive expansion of the country's public corporate sector. These factors are discussed below. 1.28 The New Economic Policy (NEP). One of the NEP's objectives is to achieve a more equitable distribution of income and to restructure the owner- ll/ This estimate shows a downward bias since it excludes many of the smaller NFPEs. - 15 - Table 1.9: NFPE SHARES IN GROSS FIXED CAPITAL FORMATION Country x Country 2 Developed Countries Norway 22.2 France 12.1 Austria 19.2 Ireland 11.8 United Kingdom 17.0 Sweden 11.& Spain 15.6 Japan 11.4 Italy 15.2 Germany 10.8 Belgium 13.1 United States 4.4 Netherlands 12.6 Canada 2.7 Developing Countries Algeria 67.6 Sierra Leone 19.6 Zambia 61.2 Argentina 19.6 Burma 60.6 Costa Rica 19.6 Egypt 47.8 Uruguay 18.3 Pakistan 44.6 Senegal 17-9 Bolivia 40.9 Kenya 17.3 Ivory Coast 39.5 Tanzania 16.3 Gambia 37.9 Peru 14.8 Mauritania 37.2 Honduras 14.6 Ethiopia 36.5 Mauritius 14.4 Venezuela 36.3 Liberia 14.1 Tunisia 35.8 Guatemala 13.3 Guyana 35.1 Chile 12.9 India 33.7 Thailand 12.8 Portugal 33.2 Haiti 12.4 Malaysia 32.0 Philippines 10.9 Bahamas 31.6 Greece 8.7 Bangladesh 31.0 Dominican Republic 8.4 Mexico 29.4 Dominica 7.8 Sri Lanka 28.4 Botswana 7.7 Panama 27.7 Mali 7.6 Turkey 27.5 St. Lucia 6.7 Jamaica 24.8 Paraguay 6.5 Brazil 22.8 Malta 6.3 Korea 22.8 Colombia 5.6 Malawi 21.2 Sources: See Table 1.6. - 16 - Table 1.10: NFPE EXTERNAL DEBT (M$ million) 1981 1982 1983 1984 1985 1986 Gross Borrowing 6,116 10,256 9,349 8,420 11,043 7,772 Federal government 3,618 5,380 5,067 4,155 7,343 2,892 NFPE 981 1,076 2,225 2,661 2,092 2,565 Private sector 1,517 3,800 2,057 1,604 1,608 2,315 Net Borrowing 5,035 8,748 7,819 5,685 2,145 3,412 Federal government 3,419 4,893 4,569 3,093 956 1,347 NFPE 769 839 2,028 2,334 962 1,815 Private sector 847 3,016 1,222 258 227 250 Outstanding Debt 15,368 24,28 31,763 37,245 42,249 50,639 Federal government 8,278 13,158 17,728 20,848 23,070 28,310 NFPE 3,071 3,715 7,657 9,656 12,027 14,862 Private sector 4,019 7,410 6,378 6,741 7,152 7,467 Source: EPU - 17 - Table 1.11: GROWTH OF PUBLIC ENTERPRISES Before Industry 1960 1960-64 1965-69 1970-74 1975-79 1980-84 1985-87 Unclear Agriculture 2 1 3 18 21 23 12 8 Construction 2 4 0 9 26 25 5 14 Extraction 0 1 0 2 16 6 0 7 Finance 0 6 3 28 22 20 4 12 Manufacturing 3 3 18 59 68 62 29 23 Services 2 1 7 43 51 85 19 30 Transportation 6 6 6 8 14 14 5 3 NFPE sector 15 16 34 139 196 215 70 87/a PE sector 15 22 37 167 218 235 74 99/a /a Includes two companies in unclassified sectors. Source: Staff Estimates - 18 - ship of assets within the society. PEs were one of the instruments chosen to bring this about. The first wave of government control of existing enter- prises and establishment of new ones can be attributed to the pursuit of NEP objectives. PERNAS, for example, was set up explicitly to increase the participation of the Bumiputera community in the nation's commercial and industrial sectors, and was one of the main reasons for the growth in the number and activities of NFPEs. Although it has been transferring some of its lucrative investments to PNB, PERNAS still has over 95 firms/subsidiaries operating in trade, retail, hotel, freight, plantations and other indus- tries. MARA has also been in YO ved in setting up subsidiaries, mostly small, in a broad range of sectors.- 1.29 Regional Economic Development. The sector also grew as a result of regional economic development objectives. This is illustrated by the large number of NFPEs set up by the 13 State Economic Development Corporations (SEDCs) which were established to undertake development projects. The three State Agriculture Development Corporations (SADCs) have also promoted public ownership in agriculture. Similarly, the Food Industries of Malaysia (PIMA) was set up in 1975 to develop food-related industries in rural areas, to promote rural industrialization and increase the welfare of the rural commu- nity. But as often happens, FIMA soon expanded into other related activities such as packaging, marketing, and retailing. PIMA's associated firms and subsidiaries now undertake activities ranging from manufactu:i of paper cartons and tin plating to the operation of restaurants and bulking term- inals. This kind of diversification points to the fact that public enter- prises, once created, acquire a dynamic of their own. If unchecked, they will continue to expand, often in areas unrelated to their original purpose. They grow not as instruments of a conscious government policy but rather as a reflection of the management's desire for expansion. 1.30 Emhasis on Heavy Industry. During the 1980s, an important contri- butor to the growth of the NFPE sector has been Malaysia's desire to establish a heavy industrial base. Although the expansion of heavy industry can be encouraged in the private sector through a system of industrial incentives, in MalQ7sia, public ownership was the favored option-perhaps even at the cost of crowding out private investment, as seems to be the case in the automobile, steel and cement sectors, for example. The creation of the Heavy Industrial Corporation of Malaysia (HICOM) in 1980 set the stage for the Government's involvement in large industrial projects. Since its creation, HICOM has invested in 15 companies, including Proton and Perwaja, with a capital invest- ment of over M$2 billion. B. Performance Introduction 1.31 An evaluation of the performance of PEs should ideally be based upon criteria that reflect their original objectives and reveal their contribution to the economy. However, due to the unavailability of data for such an analy- 12/ See Bruce Gale, Politics and Public Enterprise in Malaysia, Eastern Uni- versity Press, 1981. - 19 - sis, the focus here is on financial indicators. These indicators, related to accounting profits and profitability, are being utilized by proponents and opponents of the NFPEs alike in the debate related to the sector's efficiency. The profitability indicator is of added importance because it is invariably mentioned-by both government officials and managers--as the single most important objective of the NFPEs, though this has been relatively implicit until recently. 1.32 The performance of the PEs is assessed below through: (a) an analy- sis of trends in the number of profitable firms by sector; (b) net profitabil- ity trends; (c) equations and variables used for regression analysis, and the regression results; (d) an evaluation of loss-making firms--their magnitude, distribution and impact; and (e) a comparison of NFPE performance with that of the private sector and explanations for inferior p X -mance by the NFPEs. Trends in the Number of Profitable Firms 1.33 The number of firms operating profitably within the various sectors and the aggregate NFPE sector is a valuable indicator since it is not influ- enced by the performance of one or two very large firms and is more represent- ative of overall sector performance. However, it does not reflect the true impact of the performance of the sector upon society since, if the profitable firms are small in size but large in number, the overall impact on the sector may be minimal. The performance evaluation provided here covers the six-year period 1981-86 and is based primarily on information and documents supplied by CICU. The analysis excludes the statutory bodies due to a lack of data. 1.34 Tables 1.12 and 1.13 summarize the performance of the various sectors and the NFPE sector as a whole. No real trend is apparent for the overall NFPE sector, which seems to have experienced no major changes in prof- itability. However, the absolute number of firms making losses (roughly half) is alarming and points to the impossibility of sustaining so many loss-making firms for very long. Table 1.12: PERCENTAGE OF PROFITABLE COMPANIES: TOTAL NFPE SECTOR Year 1981 1982 1983 1984 1985 1986 X with positive 61.6 55.5 56.7 58.0 54.8 55.2 No. of firms 498 576 631 676 702 574 Source: CICU. 1.35 Sectoral Performance. What are the relatively more profitable sectors and how has the number of profitable firms changed over time in various sectors? As indicated in Table 1.13, the largest number of profitable - 20 - firms as a proportion of total firms over the six-year period was in finance, followed by transport and extraction. The least number of profitable firms was in the agricultural sector, although a dramatic turnaround occurred in 1985 when many firms began to show profits. The percentage of profitable NFPEs in the construction industry declined after 1984, and again in 1986, due to lower demand for construction activity within the economy. Similarly, the property development sector shows a high percentage of profitable firms during the property boom years of 1983 and 1984, but then begins to decline. In the plantation sector, commodity price fluctuations affected the number of profi- table firms, which peaked in 1984 at the height of Lhe palm oil price, then dropped sharply as palm oil prices plummeted in 1986. The manufacturing and services sectors, the largest in terms of the total number of firms, both show declines after 1984 due to the declining macroeconomic performance of the economy. Table 1.13: PERCENTAGE OF PROFITABLE COMPANIES BY INDUSTRIAL SECTORS FOR 1981-86 /a 1981 1982 1983 1984 1985 1986 Manufacturing 63.4 (131) 57.6 (158) 55.3 (168) 57.7 (182) 50.2 (183) 51.9 (156) Services 64.8 (125) 57.1 (147) 53.3 (167) 54.9 (184) 48.1 (189) 50.8 (140) Transportation 78.8 (33) 60.5 (38) 66.7 (45) 68.8 (48) 60.8 (51) 68.3 (41) Extractive 57.9 (19) 66.7 (21) 65.2 (23) 72.0 (25) 78.3 (23) 63.1 (22) Building and construct- tion 63.6 (11) 81.8 (11) 64.3 (14) 57.1 (14) 56.2 (16) 40.0 (15) Agriculture 29.0 (31) 28.1 (32) 38.9 (36) 27.8 (36) 83.3 (36) 85.2 (27) Finance 85.7 (49) 79.6 (54) 78.6 (56) 72.1 (61) 61.7 (60) 70.9 (55) /a Figures in parentheses are the total number of firms for which data on net profit are available. Source: CICU. Trends in Profitability 1.36 The performance of the sector in terms of profitability is measured as profits after taxes and interest, divided by total assets. In effect, this is a measure of the rate of return to assets. While a measure of profitabil- ity--private profitability--is valuable in assessing the performance of a private sector enterprise, it does not always indicate the efficiency of the management of public enterprises. Profitability can increase due to improve- ments in efficiency or due to changes in nonefficiency-related factors affecting profits. The major nonefficiency-related determinants of changes in accounting profitability are: - 21 - (a) changes in prices; for example, profits can rise due to favorable price increases; (b) purely accounting adjustments and transactions; for example, changes in the rate of depreciation can affect the calculation of the denominator and the numerator; and (c) the pursuit of nonco=_ercial objectives. Ideally, one would like to adjust for these factors, but if this is not possible the flaws of the private profitability indicator should be borne in mind in analyzing the performance of the various sectors. The results of the performance evaluation exercise are presented in Table 1.14. Table 1.14: AVERAGE NET PROFITABILITY BY SECTOR (unweighted average, percent) 1981 1982 1983 1984 1985 1986 Manufacturing /a 0.1 0.8 0.2 1.4 2.4 1.8 Services 1.7 1.3 1.6 0.9 -0.1 0.7 Transport 0.7 -1.7 -0.2 3.4 5.5 5.6 Extractive 22.9 19.0 1.7 16.1 16.6 15.0 Building and construction /b 4.7 4.2 3.9 3.4 2.2 1.5 Agriculture /c -5.4 0.6 -4.6 -1.0 -0.2 -1.2 Finance 1.2 1.1 -1.4 1.0 0.9 0.2 /a Excluding timber processing. Th Excluding property development. 7& Excluding plantations. Source: CICU. 1.37 The data shown in the table indicate that: (a) The extractive industry has had the highest rates of return to assets over the six-year period, with profits peaking in 1981 and declining since then. The trend of the extractive industry is dominated by one firm-PETRONAS--and thus reflects the movement of prices in the petroleum sector. - 22 - (b) Agriculture has the lowest rates of return, which are largely negative. It is interesting that in 1985 and 1986 when a large number of small firms were beginning to show profits, the industry's profitability worsened. This was probably due to heavy losses by some large firms. (c) The construction industry shows a steady and continuous decline in its profitability over the six year period, as the construction boom ended. Similarly, property development, which has traditionally been relatively profitable, shows a 501 decline in profitability between 1985 and 1986 as a consequence of the bust in the property development sector. (d) The transportation sector, aided by recent increases in the fares of MAS and MISC (airlines and international shipping), begins to show higher profitability after 1985. Since the sector's trend is also dominated by the large firms' performance, it is not easy to conclude anything about the performance of the smaller bus companies. However, since the number of profitable firms is declin- ing in the sector, it can be concluded that these gains are not reflective of industry-wide improvements. (e) A comparison of 1985 and 1986 reveals an important but disconcerting fact: in seven out of the nine sectors there is a decline in profitability. Only services has improved over a rather poor performance in 1985 and transportation has remained essentially unchanged. The figures for 1986 should reinforce the tendency towards reform. 1.38 Classification of Firms by Performance Criteria. On the basis of their performance, the firms can be categorized as sick, weak, satisfactory or good. The criteria for these categories is as follows: (a) SICK - companies with a negative shareholders' fund and incurring current losses. (b) WEAK - companies with positive shareholders' fund and incurring current losses. (c) SATISFACTORY - companies in which the difference between the shareholders' fund and the paid-up capital was negative but the companies were making current profits. (d) GOOD - companies in which the difference between the shareholders' fund and the paid-up capital was positive and the companies were making current profits. - 23 - A summary of the number of companies within the four categories based on the performance of the companies between 1980 and 1986 is shown in the following table. Tab,le 1.15: SUMMARY OF THE PERFORMANCE OF GOVERNMENT-OWNED COMPANIES Sick Weak Satisfactory Good No. of X of No. of Z of No. of X of No. of Z of No. of Year Companies Total Companies Total Companies Total Companies Total Companies 1980 49 12.28 102 25.56 45 11.28 203 50.88 399 1981 63 12.65 124 24.90 52 10.44 259 52.01 498 1982 87 15.10 171 29.69 50 8.68 268 46.53 576 1983 82 13.02 189 30.00 68 10.79 291 46.19 630 1984 94 13.99 186 27.68 80 11.90 312 46.43 672 1985 107 15.40 209 30.07 80 11.51 299 43.02 695 1986 63 13.94 140 30.97 48 10.62 201 44.47 452 1.39 The performance classification was then further broken down into companies with paid-up capital of at least M$5 million and companies with paid-up capital of less than M$5 million. As shown in Tables 1.16 and 1.17, more than two thirds of the companies in the "sick" and "weak" categories have paid-up capital of less than M$5 million, while a majority of companies with paid-up capital exceeding M$5 million can be considered as "good" companies. The trends over time have, however, remained stable, with roughly half the companies remaining in the categories of sick or weak. Table 1.16: SUMMARY OF THE PERFORMANCE OF GOVERNMENT-OWNED COMPANIES WITH PAID-UP CAPITAL OF AT LEAST M$5 MILLION Sick Weak Satisfactory Good No. of Z of No. of Z of No. of Z of No. of X of No. of Year Companies Total Companies Total Companies Total Companies Total Companies 1980 8 5.30 41 27.15 12 7.95 90 59.60 151 1981 12 5.91 55 27.10 16 7.88 120 59.11 203 1982 15 6.36 80 33.90 12 5.08 129 54.66 236 1983 17 6.56 81 31.27 19 7.34 142 54.83 259 1984 14 5.11 80 29.20 27 9.85 153 55.84 274 1985 19 6.79 90 32.14 17 6.07 154 55.00 280 1986 17 8.06 64 30.33 14 6.63 116 54.98 211 - 24 - Table 1.17: SUMMARY OF TEE PERFORMANCE OF GOVERNMENT-OWNED COMPANIES WITH PAID-UP CAPITAL BELOW M$5 MILLION Sick Weak Satisfactory Good No. of Z of No. of Z of No. of Z of No. of Xof No. of Year Companies Total Companies Total Companies Total Companies Total Companies 1980 41 16.53 61 24.60 33 13.31 113 45.56 248 1981 51 17.29 69 23.39 36 12.20 139 47.12 295 1982 72 21.18 91 26.76 38 11.18 139 40.88 340 1983 65 17.52 108 29.11 49 13.21 149 40.16 371 1984 80 20.10 106 26.63 53 13.32 159 39.95 398 1985 88 21.21 119 28.67 63 15.18 145 34.94 415 1986 46 19.09 76 41.53 34 14.11 85 35.27 241 1.40 Comparison with the Private Sector: Profitability. Table 1.18 shows a sectoral comparison of the profitability of NFPEs and private sector firms. Over a wide range of industries, the profitability of the private sector is higher than that of the public sector. - 25 - Table 1.18: RELATIVE PROFITABILITY UNDER PRIVATE/PUBLIC OWNERSHIP /a Sector 1981 1983 1985 Extractive: Public 0.190 0.170 0.166 Private /b 0.290 0.280 0.230 Manufacturing: Public 0.011 0.002 0.024 Private 0.086 0.065 0.049 Construction: Public 0.047 0.039 0.022 Private 0.016 0.057 0.008 Finance: Public 0.012 -0.014 0.009 Private 0.024 0.013 0.008 /a Private is really a sample of all companies, including public corpora- tions. This services to bias the figures towards those reported in the line for the public sector. The true difference between the sectors is likely to be greater than that shown here. Unweighted averages are used. /b 0.75 (other mining) + 0.25 (tin mining). Source: Private sector figures are staff estimates based on data from the Department of Statistics (DOS) Financial Survey. Public sector figures reflect profits after interest and taxes, divided by total assets. 1.41 Comparison with the Private Sector: Cost per Unit. Data are available that allow a comparison of private and public firms on the basis of cost per unit in two important industries: cement and steel. Table 1.19 summarizes the cost per metric ton of steel and cement produced by private and public enterprises during the period 1981-86. Except for cement in 1981 the average costs are lower under private ownership for both products and in all years. Moreover, the gap between costs under private and public ownership increased after 1984/85. This suggests that the relative efficiency of input use has declined under public ownership. Alternatively, the marginal invest- ments in capacity (and production) are less efficient and reflect wrong investment decisions. In addition, government-owned firms may enjoy access to credit at relatively lower rates and other advantages which are not reflected in this comparison. - 26 - Table 1.19: RELATIVE COSTS UNDER PRIVATE/PUBLIC OWNERSHIP /a 1981 1982 1983 1984 1985 1986 Cement: Public 121 134 136 132 196 192 Private 135 131 129 124 136 157 Steel: Public 1,020 861 784 785 989 1,012 Private 916 804 708 784 964 912 /a Unit cost per metric ton, measured in current prices. Source: Ministry of Trade and Industry. 1.42 Regression Analysis. Regression analysis was carried out to deter- mine the statistical significance of the effect of different factors on profitability (Appendix 3). The results of this analysis confirms some of the findings already discussed: (a) the size of the enterprise is positively related to net profitabi- lity. This is true whether either paid-up capital or total assets is used as a measure of size; (b) government equity negatively and significantly influences profitabi- lity; (c) performance during 1986 was worse than 1985; and (d) there is no significant difference in profitability across sectors once an exception is made for PETRONAS and its subsidiaries. Analysis of Loss-Making Firms 1.43 The major loss-making firms are analyzed below in terms of their characteristics, impact, and the potential gains from their divestiture or liquidation. 1.44 A breakdown of data on corporate public enterprises by net profit for 1986 indicates the following: Net Profit < 0 = 311 Net Profit > 0 = 298 Net Profit = 0 = 151 -1. 13/ Mostly not available. - 27 - Thus, 52% of the firms for which data exist had losses. The total breakdown of firms by their net profits is given in Table 1.20. The net profits of the corporate NFPE sector as a whole uould be negative but for one firm-PETRONAS. Table 1.20: DISTRIBUTION OF FIRMS BY NET PROFITS: 1986 Percentage (Z) Net Profit Assets Loans 10 -48.6 28.9 40.9 20 -51.2 30.3 42.3 30 -52.0 30.9 42.8 40 -52.1 31.2 43.0 50 -52.1 31.3 43.0 60 -52.1 31.3 43.0 70 -52.1 31.9 44.0 80 -51.8 33.4 44.8 90 -49.6 37.0 46.3 100 100 100 100 Absolute Value (M$ billion) 3.5 108.4 22.7 TOP 1Z 126.0X/a 38Z 28.2Z TOP 1 Firm 100.5Z 19% 18.0X TOP 3 Firms 118.6Z 25Z 38.4Z TOP 10 Firms 132.8Z 47Z /a Percentages of the absolute value. 1.45 Three important points emerge: (a) the absolute values of total losses by all loss-making firms is about M$1.81 billion, not including noncorporate NFPEs and some other corporate NFPEs. To partly adjust for this, the total losses of the noncorporate NFPEs were estimated where data were avail- able. This is summarized in Table 1.21; (b) the loss-making firms account for almost one third of total assets and 43% of all loans; and (c) the losses are concentrated heavily in the first 10% of the firms. - 28 - Table 1.21: MAJOR NONCORPORATE LOSS-MAKING NFPEs Company Net Loss- (S$ million) Sabah Gas /b -104 Lembaga Letric Sabah -18 KTM -8 UDA -5 L.P. Johor -2 L.P. Sabah -1 Sabah Energy Corp. -1 Subtotal -139 Kedah Cement Ic 113 Total 252 /a Measured as income after tax. 7i A corporate NFPE but data for 1986 were not included in CICU data base. 7i The Kedah Cement figure is M$182 million. In CICU, this is M$69 mil- lion. If this figure is used, the adjustment would be by 182 - 69 = 113. 1.46 To analyze the loss-making firms in greater detail, a table similar to Table 1.20 was constructed to show losses, assets and loans of the loss- making firms. According to Table 1.22, 10 of the loss-making firms (about 31) are responsible for over 85% of all losses. These same 10% account for 87Z of the assets and 88Z of total loans of all loss-making firms, as well as 93% of the foreign loans. The top 10 loss-makers account for almost 67% of all losses. - 29 - Table 1.22: DISTRIBUTION OF LOSS-MARING FIRMS (Z) Losses Assets Loans 101 85.1 86.9 88.1 201 91.6 91.6 92.1 30Z 96.1 93.5 93.4 40Z 96.8 94.5 93.8 501 97.5 95.4 94.6 602 98.2 96.3 95.3 702 98.7 97.0 95.7 801 99.0 98.0 96.1 90X 99.0 99.00 96.2 1001 100.0 100.00 100.0 Total (M$ billion) 1.81 33.9 9.8 1.47 Considerations in Sectoral Reform. The skewed distribution of loss- making firms has important implications for the design of any reform package to reduce government deficits and increase sector efficiency. It is thus recommended that the Government focus on the highest 101 of the loss-makers because the pay-off is likely to be greatest for improvements in their efficiency. A small increase in the efficiency of these firms will release greater resources than a major improvement in the efficiency of the other 90% of the firms. 1.48 In dealing with the loss-making firms, the Government should deter- mine the sources of the financial losses and whether the losses are due primarily to inefficiency or to financial reasons unrelated to efficiency, such as high taxes, interest expenses or depreciation costs, etc. The former is being done through the recently established Monitoring Unit in the Ministry of Finance while the latter was done under this study by identifying the operating profit of the loss-making firms. Operating profit is a good indi- cator of efficiency from the societal point of view because it is largely unaffected by transfers such as taxes and interest. 1.49 The analysis points to some disconcerting facts: only 7 of the 70 major loss-makers had positive operating profics and only 39 of the 310 loss- making firms had positive operating profits. Thus, the argument of some that most NFPE losses are due to nonoperating reasons is disproved by evidence from the corporate NFPE sector. This also implies that remedies which rely solely upon financial restructuring and tax breaks are unlikely to provide a cure. At the same time, policies for divestiture and liquidation become more attrac- tive as a means of decreasing the financial burden of NFPEs on the Government. - 30 - 1.50 On the basis of the forego . analysis, it is recommended that the Government focus upon the following:-_ (a) Privatization of many of the small loss-making firms if private sector investors can be found. Elaborate mechanisms for determining the sale price is not important from an economic point of view. If a particular loss-making firm cannot be made profitable with public sector management, then, theoretically, a price of zero (ir even negative) is socially efficient. However, political considerations and horizontal equity among potential private sector buyers requires some mechanism for price determination. For 90Z of the loss-making firms, these rules should be simple so that they do not divert the Government's resources from the reform of those NFPEs where returns to such effort are high. If private sector parties are not easy to find, serious consideration should be given to shutting down many of the small loss-making firms--unless there are overwhelning and legitimate noneconomic reasons for their continued operations. The key consideration for recommending the shutting down of these small- -but numerous--loss-making firms is not that the magnitude of their losses are great, but rather that focusing upon their reform leads to a dissipation of energies that can be better utilized elsewhere. (b) For the top 10Z of loss-making firms, performance evaluation exercises should be conducted, and based on these results and other legitimate considerations, it should be determined which firms are strong candidates for liquidation, divestiture or continued public operation. 1.51 Giv'en the large size of many of the NFPEs, detailed study for privatization would be needed--if that is the chosen option. The determina- tion of pricing is also complicated for large firms. What should, however, be resisted is "package deals" that involve favorable treatment to private sector buyers of the NFPEs after their purchase. Often, these concessions involve higher tariffs, selective tax breaks, subsidies, etc., which create distor- tions in the environment and may replace inefficient NFPEs with inefficient private firms. It may be better to shut down a firm than to offer distortion- ary "nonprice-related incentives" to lure the private sector to purchase some chronically sick NFPEs. 1.52 If the firms are retained under public ownership, it is important to implement a series of steps designed to introduce transparency in their performance, incentives for increasing efficiency and institutional reforms for sustaining those gains. These measures are discussed in greater detail in paras. 1.85-1.97. 14/ Government has already taken the decision to sell loss-making firms under the anti-recession measures of 1986. For companies with equity less than M$2.5 million, the decision to sell is irrespective of whether profits or losses are being made. - 31 - Estimate of Welfare Loss Due to NFPEs 1.53 To answer the question--What is the welfare loss to society of investing resources in the public enterprise sector?-this section of the report discusses the impact of inferior performance of public enterprises in terms of foregone output (and surplus). This is done by comparing rates of return in the public enterprise sector with both rates of return in the private sector (estimated at between 5.1Z and 8.12) and an estimate of the opportunity cost of capital, measured by the prevailing interest rate of 82 (Table 1.23). Table 1.23: INVESTMENT RETURNS IN PUBLIC AND PRIVATE SECTORS Rate of return in public sector (1986) /a 3.3Z Rate of return in private sector (1985)7b 5.1% Rate of return in private sector (1979-851 /c 8.1% Opportunity cost of capital (1986) /d 8.0% /a Measured as net profits (M$3.56 billion) divided by total assets (M$104.56 billion) for all firms studied by CICU. /b Net profits divided by total assets for the Malaysian economy. Staff estimate for 1985 based upon DOS Financial Survey data. /c Same as /b but average for four years 1979 = 12.32, 1981 - 8.0X, 1983 = 6.8% and 1985 = 5.1%. /d Estimated interest rate. 1.54 The difference between the rates of return for public and private enterprises ranges rom 1.8Z to 4.8%. Given a total asset value of M$104.5 billion for 1986, the annual foregone surplus would be M$2-5 billion. This result is qualified by the following important consideration: the rates of return to some NFPE activities may be low in terms of financial profits but may be higher in terms of shadow prices, i.e., after accounting for the distortions in different markets. Reasons for Poor NFPE Performance 1.55 The relatively poor performance of public enterprises in Malaysia is caused by a combination of various factors, the more important of which are discussed below. 1.56 Lack of Adequate Evaluation. The NFPEs in Malaysia have enjoyed a great deal of autonomy. However, this autonomy has not been matched by adequate accountability. Neither of the major monitoring agencies--the Implementation Coordination Unit (ICU) or the Ministry of Public Enterprises (MPE)-has developed a system for evaluating management performance or holds the management accountable for results (para. 1.89). Other central agencies, like EPU and Treasury, have more controlling power with respect to budget allocations, procurement approvals, loan guarantees, project approvals and - 32 - payments of bonuses to employees, but this power is not exercised in any systematic fashion. The absence of a meaningful evaluation system has created three problems which have interacted to create inefficiency. These are: (a) lack of clarity in objectives; (b) plurality of principals; and (c) lack of transparency in evaluation. 1.57 NFPE managers have faced multiple and conflicting objectives. While there is nothing wrong with the pursuit of multiple objectives, a vague adherence to several objectivei without any mechanism for setting priorities allows the management considerable scope for discretionary behavior. If, on top of that, there are several agencies and a Board of Directors responsible for oversight and none of them has a system for disentangling legitimate reasons for lower performance from managerial inefficiency, then there will be little incentive for the management to improve performance. While even in this environment there will be individual managers who will distinguish themselves by their competence and effort, the system as a whole will continue to provide little incentive for improving efficiency. 1.58 Pursuit of Socioeconomic Objectives. Many NFPEs in Malaysia were created to meet specific socioeconomic objectives, e.g., industrialization of backward regions, improvement of income distribution, or provision of low-cost housing. This is particularly true of the NFPEs set up by various SEDCs. In principle, some socioeconomic objectives may be well worth pursuing. However, while their pursuit benefits society as a whole, their costs are borne only by the enterprise. As a result, NFPEs which incur higher costs for socioeconomic projects look worse in terms of profit and profitability. An indirect--and perhaps more important--consequence is that managements are able to conceal inefficiencies, which would otherwise clearly show up in poor profitability, under the excuse that profitability is being negatively affected through pursuing socioeconomic objectives. The BODs and the controlling agencies therefore need to net out costs to socioeconomic objectives from the financial data. Table 1.24 presents the costs of socioeconomic projects implemented by some NFPEs under MPE. The NFPEs have petitioned the Treasury for reimburse- ment of these costs on the basis of previous approval of compensation for uneconomic services provided by the railway. 1.59 Flawed Investment Decisions. Many NFPEs are performing poorly because the original investment decision was flawed owing to the following factors: (a) Optimistic Forecasts of Demand and Prices. Sabah Gas, for example, was expected to break even if the price of methanol was US$180 per ton; however, the current price of methanol is US$70 per ton. In the case of cement, domestic demand in 1983 was 3.8 million tons, while capacity was 4.0 million tons. Nevertheless, two new firms were added, raising the capacity to 7 million tons, while domestic demand fell to 2.6 million tons. (b) Technology-Related Problems. The design of projects was often flawed. For example, Perwaja steel was established on the east coast at a cost of M$1.1 billion. Very new technology was brought in from Nippon Steel but it failed. The operational losses are - 33 - Table 1.24.: NYPE PROJECTS WITH SOCIO-ECONOMIC OBJECTIVES /a Loans million New township Sri ?nnjoong 1.80 Shop lots/houses for State of Perak 1.60 Shop lots/houses for Kampong Gajah 1.44 New township Bangi 4.66 infrastructure Tioman Islan 14.60 Shop lots/houses Jaya Gading 1.61 Shop lots/houses Kuala Krau 1.00 Shop lots/houses Slnpang Songeang 1.00 Shop lots/houses Merapoh 1.00 Shop lots/houses Padang Tengku 1.00 Shop lots/houses Rompin 1.00 Shop lots/houses Peramu 1.41 Shop lots/houses Batu Talam 0.20 Shop lots/houses Lancang 1.25 Shop lots/houses Kuala Tembeling 1.05 Urbanisation Section 9 1.41 Shop lots/houses Buluh Kubu I & II 6.34 Batik Village 2.50 Shop lots/houses MAcang 2.00 New township Baling 4.59 New township Sik 3.00 New township Lalang 1.97 Shop lots/houses PT 14 1.40 Shop lots/houses Jalan Asasad 1.30 Shop lots/houses Padang Serai 0.80 Shop lots/houses Sungai Petani 0.75 Shop lots/houses Kuala Ketil 0.65 Infrastructure Pulau Besar 6.09 Shop lots Jalan Rasah 6.79 Shop lots/houses Mata Air 0.59 Shop lots/houses Bukit Keteri 0.50 Total 75.21 /a Carried out under the Fifth Malaysia Plan by the SEDCs, financed by federal loans. - 34 - covered by compensational clauses and the loss due to closing down the DR plant (M$480 million) will be paid by Nippon. Billets are being made at a capacity of 30-40X. While the original plant was to utilize 20X scrap metal, the closing down of the DR plant means that the project has to operate on 1002 scrap metal. However, there is no scrap on the east coast and no demand for billets. Transport costs alone, therefore, currently account for M$60 per ton. Sabah Forest is another example of flawed design. Created at a cost of M$1.4 billion, it is owned entirely by the State of Sabah. However, the Treasury is the guarantor of the loan. Designed to produce pulp and paper from tropical hardwood, it has been plagued by problems with the newly installed machinery, which was not tested due to a dispute over its appropriateness between the management and the consortium. In the meantime, the Treasury has to pay the interest and principal on the loans it guaranteed. (c) Faulty Debt Structure. For many projects, especially those initiated in the late seventies, the debt-equity structure was 70:30, with the debt mostly in foreign currencies. The lopsided debt structure, combined with the appreciation of the yen (sometimes by 802), has created severe financial problems for the NFPEs. Thus, an excessive reliance on foreign credit and an inability to correctly forecast trends in exchange rates led to poor financial performance.. 1.60 Role of PNB Takeovers. A factor with possible efficiency conse- quences is the drive for transfer of NFPEs to PNB. Although the total number of firms acquired by PNB is not known, 23 firms 15ve been transferred from MPE alone, at a cost of M$162 million (Table 1.25).1)" In most instances PNB purchases the shares in NFPEs at prices far below their true value. While all trust companies are candidates for PNB takeovers, the more profitable firms have been particularly targeted. Given the higher probability of transfer of the more profitable firms, there is an inbuilt disincentive to exclusive focus on profitability. The parent organizations of multiple firms may also resort to cross-subsidization within firms to prevent the relatively profitable firms from showing excessive profits. 1.61 It can be argued that PUB is simply carrying out its functions in meeting the objectives of NEP. However, the purchase of shares at below their true value creates distortions in behavioral responses that lead to ineffi- ciency and low morale. A forum should be established to resolve issues related to PNB transfers from other agencies such as MPE, PERNAS, MARA, etc., and to minimize negative efficiency effects while accomplishing NEP objec- tives. Approval for such a forum was granted by Government in February 1988 in the context of its review of Transfer Guidelines. 15/ More recent information indicates that 39 PEs had been transferred to PNB from MPE by July 1988. - 35 - Table 1.Z5t LIST OF COMPANIES UNDER THE MINISTR'J OF PUBLIC ENTERPRISES TRANSPERRED TO PERMODALAN NASIONAL BHD (as at December 31, 1966) nuantity Total cost Companies of shares Agency Cost of transfer of transfer M$) ($) Phase I 1. Rapema Corporation 5,600,000 Bank Pembangunan 1.00 5,600,000 2. UDA Sea-Park 2,500,000 Peremba 1.00 2,500,000 3. D & C Bank 2,439,482 PKEN Selangor 1.50 3,659,223 4. Kempas Edible Oil 480,000 PREN N. Sembilan 1.00 480,000 5. Senawang Edible Oil 480,000 PREN N. Sembilan 1.00 480,000 6. Sebor Sabah 1,606,500 PKEN Sabah 1.00 1,606,500 7. Malaysian Industrlal Development Faiajoe 26,666,700 MFI 1.00 26,666,700 8. General Aluminium Works 575,332 Bank Pembangunan 1.00 575,332 9. Central Elastic Corp. 1,205,000 Bank Pembangunan 1.00 1,205,000 10. Kompleks Rewangan Malaysia Bhd. 91,150,000 MARA 1.17 108,215,519 Phase II 11. Castrol (M) Sdn. Bhd. 120,000 PMN N. Sembilan 10.00 1,200,000 12. Pltachi Semi-conductor Sdn. Bhd. 400,000 PREN P. Pinang 1.16 464,000 13. Panelex Sdn. Bhd. 630,000 PREN P. Pinang 0.95 601,328 14. Carter Semi-Conductor 900,000 PREN Perak 1.38 1,242,000 15. Timab Eangat Bhd. 4,735,749 PREN Selangor 1.00 4,735,749 16. Negara Properties Sdn. Bhd. 5,100,000 PREN Selangor 1.00 5,100,000 17. Syarikat Malaysia Sheet Sdn. Bhd. 2,400,000 PREN Selangor 1.00 2,400,000 Phase III 18. United Estates Projects Bhd. 9,625,000 PEREM8 BUD. 1.00 9,625,000 19. Perusahaan Pelangl Bhd. 27,792,000 PREN Johor 0.50 13,896,000 20. Kotak Malaysia 820,625 PUEN Melaka 1.00 820,625 21. Syarikat Metrojaya Sdn. Bhd. 417,000 UDA 1.00 417,000 22. Rompleks Revangan Malaysia Berhad 1,479,311 MARA 1.07 1,582,863 23. Rumpulan Peran8sang 18,200,000 PrEN Selangor 1.21 22,022,000 Total 206.042,699 215.814.839 - 36 - 1.62 Other Reasons. The performance of Malaysian NFPEs has also suffered from: (a) Pricing Policies. In many industries (e.g., sugar, cement, steel), prices are set by the Government on the basis of a mark-up on costs. If the costs for mark-up are based on those of individual firms or of firms with high per unit costs, this can dampen--and in some cases create perverse--incentives for efficiency. (b) Absence of Incentives. Few NFPEs have incentive systems for manage- ment that link performance with rewards and penalties. The result is a focus upon processes rather than performance in many NFPEs, especially statutory bodies. (c) Barriers to Exit. Malaysian NFPEs have been cushioned from closures and liquidation when things go wrong. For many large firms, the Treasury acts as a guarantor of loans and therefore acts as a reservoir of loss-making firms. For social- and employment-related reasons, the Government has been reluctant to shut down chronically loss-mak'ag firms. This may change in the future as the Government allows the private sector to buy firms and perhaps--under increasing financial strain--becomes less willing to bail out chronically sick NFPEs. C. Organizational Environment 1.63 The Malaysian NFPEs operate in different markets, under alternative control structures and within the supervision of numerous ministries. They interact with the Government in a complex, multidimensional fashion. It is, therefore, not possible to capture the entire NFPE sector in terms of an organizational chart that shows the hierarchy of decision making within the sector. But some feeling for the situation can be gained from an understand- ing of the dimensions of the control structure of the sector and the role of the important government ministries and the Boards of Directors of the indivi- dual enterprises. 1.64 Formal control of the NFPEs is exercised at three different levels: (a) the relevant ministry; (b) the parent corporation; and (c) the Board of Directors (in case of corporate NFPEs). Other government ministries, e.g., Finance, are also involved in indirectly influencing the NFPEs. Breakdown of NFPEs by Ministry 1.65 A breakdown by ministry of NFPEs in the CICU data base indicates the predominance of ICU and MPE in the sector (Table 1.26). - 37 - Table 1.26: MINISTRIES CONTROLLING NFPEs Ministry Number Total 862 Implementation Coordination Unit 259 Ministry of Agriculture 7 Ministry of Energy 2 Ministry of Finance 7 Ministry of Land 75 Ministry of Natural Resources 60 Ministry of Primary Industries 13 Ministry of Public Enterprises 365 Ministry of Trade 7 Ministry of Transport 18 Penang State Government 30 Sabah State Government 8 Sarawak State Government 9 Yayasan Pelabura 2 1.66 Ministry of Finance. While the Ministry of Finance has only seven NFPEs formally under its administrative supervision, the Treasury is a major shareholder in 20 PEs listed by CICU (Table 1.27). Nine of the 20 are among the 56 major NFPEs. As major shareholder in some of the firms, MOF appoints the Board of Directors (BOD) and the latter chooses the Chief Executive (CE) of the firm. In practice, however, the CE may be directly appointed by the Minister and later approved by the BOD. Depending upon the importance of the firm, e.g., PETRONAS, the decision may be made even by the Prime Minister. - 38 - Table 1.27: COMPANIES UNDER THE MINISTRY OF FINANCE, INCORPORATED 1. Heavy Industries Corporation of Malaysia Berhad 2. Syarikat Jengka Sdn. Berhad 3. Bank Islam Malaysia Berhad 4. Bank Kemajuan Perusahaan Malaysia Berhad 5. Bank Pembangunan Malaysia Berhad 6. Kumpulan Fima Berhad 7. Malaysian Rubber Development Corporation Berhad 8. Aerospace Industries of Malaysia Sdn. Berhad 9. Malaysian International Shipping Berhad 10. Malaysian Airline System Berhad 11. Malaysia Shipyard & Engineering Sdn. Berhad 12. Gula Padang Terap berhad 13. Malaysia Export Credit Insurance Berhad 14. Sports Toto Malaysia Berhad 15. Perbadanan Nasional Berhad 16. Bank Bumiputra Malaysia Berhad 17. Petroliam Nasional Berhad 18. Syarikat Perumahan Pegavi Kerajaan sdn. Berhad 19. Syarikat Malaysia Explosives Sdn. Berhad 20. Syarikat Telekom Malaysia Berhad 1.67 The MOF is also involved in NFPEs as a provider of subsidized loans or the guarantor of foreign loans. The loans may be provided directly to the NFPE or through the State Government owning the NFPE. Treasury loan rates are generally about 2-4Z lower than market rates and may be even lower for socio- economic projects. When macroeconomic conditions were better, the Treasury was content to let the borrowing firms operate with a great deal of managerial autonomy. However, due to tightening economic conditions, it has adopted an increasingly assertive role vis-a-vis the NFPEs. The Treasury's role as the guarantor of foreign loans and the subequent failure of some major NFPEs to show profits has also led to more scrutiny of NFPEs by the Treasury, which set up a Monitoring Unit and an Advisory Committee for this purpose. The Monitoring Unit is currently examining the performance of ten large and financially troubled firms with a view to identifying how performance can be improved and what possibilities, if any, exist for privatization. 1.68 Implementation and Coordination Unit. The ICU in the Prime Minister's Secretariat is responsible for the control of some of the most important NFPEs, including PETRONAS, MISC, HICOM and PERNAS, which is a trust agency. The ICU is considered a "Watchdog" for all government projects and has a unit responsible for the NFPEs. The ICU exercises its control through the review of annual reports, ex-post evaluation and approval of the financing of new projects. The firms under ICU have traditionally been powerful and relatively autonomous. Their personnel policies have been shaped more by private sector practices than those of government departments or statutory bodies. The ICU has not, however, been able to develop a systematic perfor- mance evaluation system that takes into account different objectives of the - 39 - NFPEs and holds management responsible for results. It has also been hampered in its control functions due to insufficient resources and professional staff. 1.69 Ministry of Public Enterprises. The MPE's principal role is to monitor the programs, activities and achievements of the 13 SEDCs, 3 SADCs, and 6 Federal agencies, namely, UDA, BPMB, FIMA, MSE, MIDF and KKMB. In addition, it appoints three representatives each to the BOD of SEdCs and SADCs at the holding level. The State SEDCs also appoint their respective represen- tatives to the BOD. Some of the members are representatives of the private sector and/or political appointees. 1.70 The role of the M2E is to ensure that the operation of each agency under its supervision is in line with the stated objectives of that agency and with the NEP, especially with regard to the development of Bumiputera entre- preneurs and the involvement of Bumiputera in the management and ownership of the corporate sector. The MPE's ability to carry out this role, however, is limited as it has monitoring responsibilities without any effective control over the management of the enterprises. Its focus therefore has been on balance sheets and tabulating and aggregating results. The real control is exercised by the Boards and the respective State governments. However, NFPEs requiring government funds for initiating a project need the approval of the MPE before loan approval by the Government. The loan repayment to the Federal Government by the NFPEs takes place through the State Governments which often siphon off some of the repaid loan. Parent Corporation 1.71 Table 1.28 gives a distribution of NFPEs by their parent corpora- tion/agency. As the table indicates, some of the large Government holding bodies have gained significant power in the sector. Table 1.28: NFPE DISTRIBUTION BY AGENCY Federal agencies State agencies Regional PERNAS 95 PERAK SEDC 60 DARA 17 PNB 62 SARAWAK SEDC 45 KEJORA 12 MARA 59 JOHOR SEDC 42 OTHER 23 FIMA 34 KEDAH SEDC 34 BPMB 34 ASPA 22 MISC 19 TERANG. SEDC 21 PETRONAS 19 KPSB 21 FELDA 18 PENANG SEDC 20 HICOM 16 NEGERI SEDC 18 OTHER 89 OTHER 116 FEDERAL 445 PROVINCIAL 399 REGIONAL 52 - 40 - Board of Directors 1.72 The BODs link the ownership role of the Government with the operat- ing role of the Chief Executive Officer and senior management of the enter- prise. The Government, as the principal, appoints the BOD, and defines its objectives and degree of autonomy in fulfilling those objectives. The manage- ment's responsibility is to execute the agenda set by the Government through the BOD. In theory the BODs should perform several functions, the most important being: long-range planning, performance evaluation, and appointment and dismissal of top management. They are also expected to act as a buffer between the firm and excessive political interference in the firm's day-to-day operations. While a detailed analysis of the performance of individual BODs is beyond the scope of this report, certain general features of the workings of the BODs in Malaysia deserve comment. 1.73 Historically, NFPEs in Malaysia have enjoyed considerable autonomy and been free of the kind of political intervention often found in other developing countries. This is particularly true of the corporate NFPEs. The BODs have therefore been largely successful in acting as buffers against excessive intervention. It can be argued that this management model via decentralized decision making and autonomy is preferable to the opposite model of excessive centralization and limited autonomy. There is concern, however, that in the case of Malaysia, the autonomy is perhaps not matched by the same degree of accountability. As a consequence, the scope for discretionary behavior on the part of management is quite high. Moreover, given the impor- tance of the BODs as an organizational pillar in Malaysia, more attention should be given to improving their functioning. 1.74 A prerequisite for the successful performance of the evaluation and planning function by the BODs is the nomination of people with the right blend of expertise, experience and independence. These conditions are often not met, however. Many of the appointments are motivated by political considera- tions rather than qualifications for the assignment. With increasing pres- sures on PEs to improve their performance, the traditional approach of select- ing political appointees and ministry representatives for the BODs should be reappraised. Privatization 1.75 Privatization of some NFPEs has been recently attempted due to government budgetary problems as well as the realization that the public sector has become too large and its perceived inefficiency may constitute an obstacle to higher productivity. 1.76 Objectives. The stated objectives of the Government's privatization policies are: (a) relieving the financial and administrative burden of the Government; (b) promoting competition, raising efficiency and productivity; (c) accelerating growth by stimulation of private entrepreneurship; - 41 - (d) reducing the size and presence of the public sector in the economy; and (e) support to the objectives of the NEP by ;ranferring assets to the private sector. An unstated objective is to raise revenues for the Government by the sale of i public assets. 1.77 Institutional Structure. An interagency (IA) committee on privati- zation, chaired by the Director General of the Economic Planning Unit (EPU), has been established to oversee the privatization process. Committee members include representatives of the Treasury, Public Services Department, the Attorney General and Secretary Generals of key ministries. The IA committee scrutinizes projects for privatization and forwards them to the Cabinet for approval. The committee is aided in its work by technical committees and working groups that deal with specific issues such as manpower, land, legisla- tion, etc. The sale of Government shares/companies to individuals, or their transfer to PNB, is supervised by the Investment Coordination Committee (JPP) under the chairmanship of the Deputy Prime Minister. 1.78 The Government's commitment to privatization is indicated by its intention to develop a master plan for privatization to address the issues involved in an integrated fashion, to provide guidelines for privatization, identify potential bottlenecks and suggest ways for overcoming them, and prepare detailed operational plans for some individual units. Consultants to assist in preparing the master plan are being chosen and the plan is expected to be completed in six months. 1.79 Scope of Privatization. As of November 1987, the Government had privatized ten projects (Table f.29); another 15 projects had been approved for privatization (Table 1.30) and 48 others were being considered (Appendix 4). In addition, 71 smaller firms, belonging primarily to the various SEDCs, had been sold (Appendix 5). The total value of the privatized firms is difficult to determine. The value of the ten projects privatized by the Federal Government, excluding MISC and Teman Negara, is M$1,715 million. However, MAS accounts for 74X of this. While the reduction in the size of the total PE sector has so far been small, if the Government continues to pursue the privatization policy as seriously as its public pronouncements imply, then these numbers can change. For example, when Telecom is privatized after three years of consolidation in its new corporate structure as promised, the extent of privatization will rise sharply. 1.80 Mechanisms of Privatization. The following approaches to privatiza- tion have been utilized by the Government thus far: (a) Partial divestment. In this method of privatization, used for MAS and MISC, shares in the PE are sold in the stock market, but the Government retains the majority of shares and the ability to influence the Board of Directors. Table 1.29: PROJECTS WHICH HAVE BEEN PRIVATIZED Name of Origin of Project Mode of project Sector proposal value privatization (in mln) 1. Maintenance and Manufacturing (main- Private sector 89.5 Leasing repair of RMAF tenance & repairs) aircraft 2. North Klang Construction, trans- Private sector 20 New project developed by the Straits by-pass portation private sector 3. Viaduct-Jalan Construction, trans- Private sector 80 New project developed by the Kuching/Kepong portation private sector 4. Sports, Toto Business services Private sector 40 Sale of shares Co., Ltd. 5. TV3 Entertainment Private sector n.a. New project developed by the services private sector 6. Malaysian Air- Transportation & Government 1,262 Sale of shares line System storage 7. Container Termi- Transportation & Government 112 Sale of movable properties and nal, Port Klang storage leasing of immovable proper- ties 8. Malaysian Inter- Transportation & Government n.a. Sale of shares national Shipping storage Corporation 9. Labuan Water Water supply Private sector 112 New project to be developed by Supply the private sector 10. Tourist Facili- Recreational services Government n.a. Leasing ties at Taman Negara n.a. a not available. - 43 - Table 1.30 : PROJECTS VOICH HAVE BERN APPROVED FOR PRIVATIZATION Name of Origin of Project Mode of project Sector proposal value privatization (M$ min) 1. Metrolink Transportation Private sector 134 Development of a new project by the private sector 2. Sabah-Labuan Transmission Private sector 50-60 Development of a new project Inter Connection by the private sector 3. Melaka Port Transportation Private sector 8.4 Leasing and storage 4. Security Printing Manufacturing Private sector n.a. Leasing Div. (Cov. Print- Ing Department) 5. Langkawi Island Hotel industry Government 28 Outright sale Resort 6. Motel Desa Hotel industry Government 3.28 Outright sale/lease Kuala Trengganu 7. Tanjung Jara Rotel Industry Government 7.587 Outright sale/lease leach Hotel 8. Rantau Abang Hotel industry Government 1.22 Outright sale/lease Visitor Center 9. Inspection of Business services Government - Contract Commercial Vehicles 10. North/South Construction, Government 3.5b Development of a new project (toll) Highway transportation by the private sector 11. Telecoms Dept. Communications Government 7.805 Sales of shares of Syarikat Telecom Malaysia 12. Marketing of Business services Private sector n.a. Management commercial adverti- sing time over Radio Malaysia 13. Garbage disposal, Sewerage and garbage Government 60 Development of a new project City Hall, disposal by the private sector Kuala Lumpur 14. Kuala Lumpur Construction Government 200 Development of a new project Interchanges transportation by the private sector 15. Marketing Division Wholesale Government n.a. Management of RISDA trade - 44 - (b) Conversion of a government department into an incorporated entity that is wholly owned by the Government. This approach, which was used for the Telekom Departmeat, represents a change in the organi- zational form of the NFPE, and is really not privatization. However, a change to the corporate model is likely to improve efficiency since the departmental form of organization tends to be overly bureaucratic and its culture is not conducive to operations of an essentially commercial nature. Further restructuring and privatization are also more probable for a corporate entity than for a department. This may therefore be viewed as an intermediate step towards eventual privatization. Despite the possible improvements to efficiency and behavior, this form of conversion of legal status does not effect the private/public mix. (c) Sale of shares in companies through private placements, such as Sports Toto. (d) Allowance of private investment in areas traditionally reserved for the public sector, e.g., licensing of a privately owned TV channel, TV 3, and the privatization of the water supply project at Labuan. (e) Leasing to private sector management. The Port Kelang Container Terminal, for example, was leased to the private sector although Government retained shares in the ownership. This is not privatiza- tion in the sense of transferring public sector assets into private hands. It is really a liberalization measure that will, at the margin, affect the private/public mix and if sustained will, over time, reduce the share of the public sector. (f) Outright sale of public firms to private sector individuals and firms. This has mainly affected small NFPEs belonging to the SEDCs, and the overall magnitude has been very small in terms of the value of assets transferred from the public to the private sector. However, the approach is now being encouraged for a large number of firms being considered for sale by Government, and private sector entrepreneurs are being allowed to bid on government firms or themselves initiate the acquisition of NFPEs. 1.81 Impact of Privatization. It is still rather early to judge the impact of these initial privatization efforts. In the case of partial divest- ment, the fact that only a portion of the shares are sold allows the Government to retain effective control of the firms, and it is not clear what behavioral impact this will have on the corporation's management i . .erms of increased effort and efficiency. In the case of KAS and MISC, the objectives were to restructure the balance sheets through an infusion of new equity and to generate funds for the Treasury. In both cases, it is unlikely that competition and efficiency will increase appreciably. as a direct result of the privatization measures. 1.82 The main impact of divestiture thus far appears to be an increase in government revenues through the sale of its shares. The purchasers of shares of the privatized firms have also benefited since the shares were sold at far - 45 - below theii,market value and their prices rose dramatically soon after priva- tization. Thus, those who were able to buy the shares received a wind- fall. In other arrangements for sale of shares, besides public listings, there are also problems in maximizing government revenue. Initially, private offers were considered on a "first-come-first-served" policy, .&ich was characterized by a negotiated price between Government and private parties who were granted a letter of exclusivity on the basis of their bid. This has now been weakened to give initial private bidders the first right to negotiate. There is still, however, a tension between this policy, which has the advan- tage of letting the private sector take the lead and of forcing Government departments to let go, and the objective of revenue maximization. An alterna- tive, in cases where privatization is initiated by the public sector itself, is an open tender, as for the Port Kelang Container Terminal and the Tanjing Aru Beach Hotel. D. Improving the Efficiency of NFPEs Reform of the Control Structure 1.83 Two different aspects of the control structure of NFPEs need to be distinguished. First, there is the formal control structure, represented by an organizational chart that shows the official hierarchical relationship between the different layers of the organization. Second, and perhaps more important, is the informal autonomy structure, according to which decision- making really takes place. A typical response to the problems of NFPEs is to "reorganize." Often this reorganization entails nothing more than a redrawing of the formal organizational chart. A major change in the manner in which NFPEs are organized in Malaysia is not needed at this stage to achieve the objective of increasing efficiency since: (a) the transaction cost involved in moving from the existing organiza- tion to a new one will be high. The main argument for reorganiza- tion would be to consolidate on the basis of industries to maximize economies of scale from specialization. However, given the current system of large holding firms such as PERNAS and MARA, which are operating in many industries, this kind of restructuring is likely to cause a great deal of disruption with high short-term costs; (b) a focus on reorganization of the formal control structure may act as a palliative and divert attention away from the more pressing problem of accomplishing efficiency gains that are possible within the current organizational structure. The possibilities of increas- ing efficiency within the current system of control are high, provided the steps outlined below are implemented; and 16/ This is generally true for all new issues on the KLSE and comes about partly as a result of the price-fixing guidelines followed by the Capital Issues Committee. - 46 - (e) a major reorganization may lead to an entrenchment of the NFPEs, thus reducing the momentum of privatization efforts. 1.84 While no major changes are recommended, some comparatively modest organizational changes, with high expected pay off, may be considered. These includft. (a) clarifying the role of the Ministry of Public Enterprise. The MPE is responsible primarily for NFPEs under the various SEDCs. In theory, it monitors performance, but without any effective control, it finds this task difficult. MPE is also unable to prevent inter- ference in the management of firms from the various state govern- ments. Given the dispersion of the NFPEs under SEDCs, MPE should be given increased powers in supervising the MPEs under its control and for developing impartial systems for evaluation; and (b) since the statutory bodies which are also NFPEs (Telecom, LLN) are widely dispersed under different ministerial control, there is a case for setting up a body to provide general guidelines on objec- tives, evaluate performance and resolve inter-ministerial issues wher they affect the NFPEs as a group. Such a body could perhaps be located in the Treasury, the most powerful controlling agency over the PEs, perhaps through expansion of the new UPSAK unit. Development of Information System 1.85 Good quality information is a prerequisite for correct decision- making. The creation of CICU is an important step in this direction since it will allow policymakers to evaluate individual enterprise performance and the macroeconomic impact of the NFPEs as a group. CICU has the potential to be extremely effective. Its staff is professional and enthusiastic, and it is doing a good job of collecting, storing and managing information. CICU has recently also started to develop analytical tools for evaluating the informa- tion it collects, such as PNG-Score system. This system sorts companies into problem and non-problem groupings through discriminant analysis (see Appendix 6). Its focus, however, is almost exclusively on financial results of the company. While financial performance is the best indicator of managerial efficiency in a commercially-run operation, it is not always a good indicator where socioeconomic objectives are important. Thus, the CICU system is no substitute for a broader performance evaluation system. 1.86 As CICU's data base is enlarged to generate more direct information on productivity, through computerization of operational data on volume of production, sales, raw materials consumption, employment, prices, etc., it will become increasingly useful to determine management efficiency. In this, however, the development of CICU is simply an extension, on a quantitative and centralized basis, of the role played by other supervisory agencies such as MPE and ICU. In effect, both CICU and these agencies now provide monitoring functions, albeit with different methodologies--the former favoring sectoral and intersectoral comparative statistical performance and the latter favoring personal in-depth knowledge of specific companies. The next step must be to identify actions to be taken once problems have been identified and to ensure - 47 - that Ministerial recommendations are implemented. As yet, this function has been performed by UPSAK for only a selected few of the largest NFPEs. An institutional structure that can perform these functions more systematically, and that is well coordinated with the monitoring functions is required. Performance Evaluation Systems 1.87 A necessary condition for ensuring accountability--and thus improv- ing efficiency--is the introduction of systems designed to evaluate managerial performance. At present, neither MPE nor ICU has a developed system for evaluation of the performance of the NFPEs, although they recognize that, they should do this. An important condition for meaningful evaluation--an informa- tion system-has already been met with the development of the CICU data base. As discussed above, however, this data must be supplemented by a series of targets in order to be used for an evaluation. While a more detailed set of recommendations would be necessary for the design of a comprehensive performance evaluation system, the following suggestions are offered. 1.88 Civen the complexity of the technical issues related to performance evaluation, the first priority should be the establishment of an institutional framework to resolve these issues. As a start, performance evaluation units need to be set up within the MPE and the ICU--the two agencies with the most NFPEs under operational control. As is already the case in these agencies, monitoring and evaluation should be done at the holding company level. This will reduce bureaucratic controls and is likely to be more effective in instituting internal accountability down to the lowest levels. The existing staff within the two ministries dealing with evaluation issues should form the basis of these units. Additional technical expertise should be acquired to staff the units with competent professionals, perhaps drawn from the NFPE managerial ranks, experienced in evaluation issues, financial and economic performance and industry specialists. The payoff to some selective recruiting in this area will be high and well worth the cost. Outside help, when needed, could be acquired from academia, the private sector and international agencies. 1.89 The tasks of the performance evaluation units would essentially consist of the following: (a) choosing performance indicators, e.g., profits and productivity; (b) assigning relative weights to different indicators, e.g., 0.8 to profits, 0.2 to productivity; (c) negotiating targets for the management at the beginning of the year; (d) calculating the scores at the end of the year; (e) assigning grades to the management, which would ideally determine management reward; and (f) reviewing performance with managements to identify genuine constraints, main strengths and weaknesses, and to draw appropriate lessons for the future. - 48 - None of the steps outlined above is easy to implement in practice. There is, however, an emerging consensus that systematic performance evaluation of management is necessary if NFPEs are to increase their efficiency in any major way. Several other countries, most notably Pakistan and Korea, have imple- mented performance evaluation systems. The World Bank is helping several other countries to evolve similar systems which are suited to local condi- tions. Malaysia can and should benefit from the experience of 07 Bank and other countries in developing a system geared to its own needs._ 1.90 Although the details of a performance evaluation system will have to be worked out by the relevant ICU and MPE officials and consultants recruited for this assignment, the following very general guidelines are offered, based upon the Bank's understanding of Malaysia's institutional environment: (a) The indicators, at least in the initial stages of implementing the system, should be simple. They should also be fair, both to the management and the country. Fairness to the management implies that they should be judged only on the basis of what is within their control and not penalized if their performance is affected by exogenous circumstances. Fairness to the nation implies that the indicators must reflect the contribution of the NFPE to the economic welfare of the country. The current focus on the financial viabil- ity of the NFPEs is a healthy one. At the same time, consideration must be given to improvement of etficiency, which might be mea gyed via total factor productivity or public profitability or both. (b) The annual targets should be reached following negotiation with NFPE management at the beginning of the year. The ICU and MPE should deal only with the parent NFPE where there are subsidiaries and associate bodies. For example, ICU should be concerned with eval- uating the performance of HICOM, PETRONAS or their important subsi- diaries, e.g., PERWAJA, PROTON, etc. The management of the parent organizations should be given the autonomy to determine policies vis-a-vis their subsidiaries although they should be encouraged to adopt principles of management by objectives and designate cost and profit centers within their subsidiaries to meet their targets. 17/ For a discussion of these systems see, for example, Young C. Park, "A System for Evaluating the Performance of Government-Invested Enterprises in the Republic of Korea," World Bank Discussion Paper Series, No. 02529- 210X:3, and A.H. Shaikh, "Performance Evaluation of Public Enterprises: Lessons From the Pakistan Experience" in Annals of Public and Cooperative Economy, Oct-Dec. 1987. 18/ For a discussion of the public profitability concept and its relevance to the performance evaluation of NFPEs, see Leroy Jones, "Towards a Performance Evaluation Methodology for Public Enterprises," paper presented at an international conference on performance evaluation, Islamabad, November 1981. - 49 - They should also be encouraged to develop their corporate planning capabilities and evolve uniform criteria for the preparation of accounts and budgets. (c) The education of a broad range of managers, evaluators and govern- ment officials will also be necessary. Thus, training and dissemi- nation of information through seminars, workshops, development of manuals and other means should receive priority. (d) The evaluation system should have both an evaluation component and a monitoring component. The evaluation component essentially performs ex-post analysis for determining the scores and bonus levels of the management. The monitoring component should be designed for periodic reviews and detecting early warning signals during the course of the year. Feedback from this source could be used by management to correct potential problems before they become serious. (e) There are limits to the usefulness of quantified performance indi- cators. These limits should be recognized, and allowance should be made for judgments based upon qualitative assessments, especially in service-based industries. It is important therefore that the professional judgment, impartiality, technical competence and personal integrity of the evaluators be respected by the parties involved. Incentive System 1.91 A system of performance evaluation is a necessary but not a suffi- cient condition for achieving higher efficiency within NFPEs. It is equally important to link performance with a set of rewards and penalties to get the desired behavioral response from the NFPE management. 1.92 In the design of the incentive system, several considerations are important: (a) incentives should be geared to the right set of people, i.e., senior management; (b) incentives should be sufficiently high to induce the desired behavior; (c) while financial bonuses are the single most important form of bonus, other incentives can also be effectively used, especially for senior managers. These include public recognition--for example, giving out medals--granting greater autonomy, accelerated promotions, etc.; and (d) the incentive systems can be approved relatively easily through the Boards of Directors for the corporate NFPEs. However, they may be harder to implement in the statutory bodies. A potential source of resistance is from the civil service which may view this as an attempt to increase the salaries of some government officials, and thus a violation of horizontal equity. This opposition can be - 50 - overcome by increasing the awareness of the commercial nature of the NFPEs and making sure that the bonuses are strictly linked to increased efficiency and do not degenerate into automatic increases in salary. Board of Directors 1.93 Given the large number of corporate NFPEs in Malaysia, proper functioning of the BODs is critical to achieving sector-wide gains in effi- ciency. In this regard, the most important area in need of reform is the selection of BOD members. Given the very large number of NPPE Boards and the ability of some ministries, e.g., Treasury and MPE, to put their representa- tives on the BODs, the Boards are often filled with members unable to devote the time and energy necessary for successful management of the enterprise. In some instances, individuals have been on the Bourds of 20 companies, and in one instance a person was Chairman of the Board of ten NFPEs. All this was, of course, in addition to carrying out a normal workload within the ministry. 1.94 Monetary gain is an important reason for the desire of many officials to be on the Boards of several firms at the same time. Compensation of BOD members ranges between one and two thousand ringgit and the Chairman can receive up to five thousand ringit. The Treasury is currently trying to restrict the salaries of board members, although this may not be well- advised. International experience su;ggests that to attract better candidates for BOD membership, especially from the private sector, compensation should be comparable to that of private firms. A more promising approach may be to restrict the number of Boards to which mny one person can be appointed. 1.95 Although not a widely prevalent practice, there are also instances of the Board Chairman and the CEO being the same person, e.g., in PERNAS. This obviously muddles the ownership and operational functions of the Govern- ment and ought to be avoided as much as possible. Divestiture 1.96 The case for privatization of NFPEs rests primarily on the argument of the higher relative efficiency expected under private ownership. The comparative analysis discussed earlier (para. 1.40) does in fact confirm that the efficiency of the private sector tends to be higher for a wide range of industries in Malaysia. And since the need for privatization seems to have been accepted by the Government, it appears that the issue is not whether to privatize, but what and how best to privatize to achieve maximum gains in efficiency. 1.97 The expected gains from increased efficiency will differ, depending on the method of privatization. Only if privatization is effective in the sense of transferring operational control to the private sector will the benefits of private initiative be maximized. It is for this reason that the modest pace of privatization so far is likely to have had only a marginal effect on efficiency since the corporations involved remain largely under government control. These privatizations had other important non-efficiency- related effects, such as sending correct signals to the private sector and - 51 - revenue generation for the Government and the corporation. They have also provided a symbolic initiation by the Government of its program to allow the private sector to take the lead in the development process. However, if the aim is to improve the efficiency of production, then it will be better to think in terms of outright sale or transfer of management to the private sector. 1.98 A related issue is the continued operation of loss-making NFPEs. Not all loss-making firms are inefficient. Some NFPEs make losses in spite of being efficient since the source of the losses may be non-efficiency related like the government's macroeconomic policies, high tariffs or other pricing decisions, excessive wage costs due to political or institutional reasons, disadvantageous location, incorrect forecasts in the original project evalua- tion or the pursuit of non-commercial objectives (paras. 1.55-1.60). However, the fiscal burden of loss-making NFPEs (almost 50X of all NFPEs) cannot be sustained for long. The performance of loss-making firms should therefore be analyzed to determine the source of their losses; whether their problems can be remedied; and whether the costs involved in remedying the situation are worth bearing. While government's announced policy of dealing with loss- making NFPEs is to privatize them, this should be considered as a long-run goal. In the mediumo-term, it is unlikely that the scale of privatization can be accelerated to deal with the large numbers of firms involved. The initia- tion of the performance evaluation system should be thought of as a parallel track to divestiture; one which seeks to upgrade managerial effectiveness and, hience, asset value, before privatization occurs. 1.99 The formidable barriers to the dissolution of NFPEs are not unique to Malaysia. But improvement in efficiency of production requires that resources are allocated to the most profitdble areas in the national economic sense. If it is determined that under continued public operation an VFPE cannot increase its efficiency, and no private owners can be found, serious consideration should be given to closing down that class of NFPEs. Chapter I Appendix 1 - 52 - MALAYSIA Companies Identified for Data Collection by CICU Agri-Wise Realty Sdn. Berhad Amanah Saham Sabah Berhad Likas Bay Hotel Sdn. Berhad Perkina Industries Sdn. Berhad PBS Securities Sdn. Berhad PBS Labuan Sdn. Berhad Perkina Motors (Sarawak) Sdn. Berhad Syarikat Yakima Sdn. Berhad Timbar Properties Sdn. Berhad Great Union Properties Sdn. Berhad Asean Development Corporation Sdn. Berhad Sri Pandan Services Sdn. Berhad Modal Lombong Indah Sdn. Berhad Modal Mining Sdn. Berhad Syarikat Modal Timor Sdn. Berhad Modal Bina (Pahang) Sdn. Berhad Permodalan Properties Sdn. Berhad Pahang Electronics Sdn. Berhad Syarikat Perubatan Utara Sdn. Berhad Great Eastern Mill Sdn. Berhad Perbadanan Khidmat Komputer FELDA Chapter I - 53 - Appendix 2 Page 1 of 2 MALAYSIA List of Statutory Bodies Prime Minister's Department Muslim Pilgrim Saving and Management Authority National Development and Family Board Ministry of Trade and Industry Tourist Development Corporation, Malaysia (TDC) Malaysian Industrial Development Authority (MIDA) National Productivity Center National Film Development Corporation Ministry of Finance Central Bank of Malaysia National Savings Bank Employees Provident Fund (EPF) Ministry of Transport Light Dues Board, Peninsular Malaysia Kelang Port Authority Johore Port Authority Kuantan Port Authority Bintulu Port Authority Penang Port Commission Shipping Industry Fund Center Ministry of Education -4-ra Inotittute ozf T-chi-' gyZ1 University of Malaya University Hospital National University Malaysia University of Agriculture Malaysia University of Science Malaysia University of Technology Malaysia Examination Council Education Central Board Tunku Abdul Rahman College Language and Literacy Hall Teachers Provident Fund Board The Northern University of Malaysia Ministry of Primary Industries Malaysian Rubber Exchange and Licensing Board Malaysian Rubber Research and Development Board Palm Oil Registration and Licensing Authority Malaysian Timber Industries Board Palm Oil Research Institute of Malaysia (PORIM) Rubber Research Institute of Malaysia (RRI) National Tobacco Board Malayan Pineapple Industry Board Tin Industry Research and Development Board Ministry of Land and Regional Development Federal Land Development Authority (FELDA) Rubber Industry Smallholders Development Authority (RISDA) Federal Land Consolidated and Rehabilitation Authority (FELCR..) Southern Johore Development Authority Central Trengganu Development Authority Southern Kelantan Development Authority Southern Pahang Development Authority Kedah Regional Development Authority Penang Regional Development Authority Jengka Development Corporation Chapter I - 54 - Appendix 2 Page 2 of 2 Ministry of Agriculture Agricultural Bank of Malaysian Fisheries Development Authority Farmers Organization Authority Federal Agricultural Marketing Authority (FAMA) Kemubu Agricultural Development Authority Muda Agricultural Development Authority Malaysian Agricultural Research and Development Institute (MARDI) Ministry of Labor Malaysian Migration Fund Board South India Labor Fund Board Penang Port Labor Board Manpower Development Board Social Security Organization (SOCSO) Ministry of Science, Technology and Environment Standard and Industrial Research Institute of Malaysia (SIRIM) National Council for Scientific Research and Development Ministry of Culture and Tourism Board of Trustees of National Gallery of Art Merdeka Stadium Corporation National Sport Council Board of Trustees of National Museum of Art Ministry of Defense Armed Forced Provident Fund Ministry of Public Enterprises National Padi and Rice Board Urban Development Authority (UDA) Ministry of National and Rural Development Council of Trust for the Indigenous People Cooperative College Malaysian Handicraft Development Corporation Ministry of Works Malaysian Highway Authority Ministry of Information Malaysian National News Agency Ministry of Welfa e Social and Welfare Services Lotteries Board Ministry of Energy, Telecommunications and Post National Electricity Board _ 55 _ Chapter I Appendix 3 Page 1 KALAYSIA Regression Analysis of NFPE Profitability 1. Regression analysis was carried out to determine the statistical significance of the effect of different factors on profitability. The dependent variable is defined as indicators of profitability were: Met profitability (NTTA) = Met Profits/Total Assets (TA) Other measures of profitability, such as operating profits/total assets, were also used. The results were very similar and, hence, are not reported here. The independent variables included: (a) Total Assets (TASS) (a measure of size, included in log form) (b) Total loans/paid-up capital (TLTC) (c) Foreign loans/total loans (FTTC) (4) Share of Government Equity in Total Equity - GEQ (e) A dummy variable for separating 1985 and 1986 = YEAR (f) Sectoral dummy for each of the sectors, e.g., manufacturing, services, atc. a D2, D3 - Dg 2. The results of the regression analysis are summarized in the following table. - 56 - Chapter I Appendix 3 Page 2 8EGRESSION RESULTS Dependent variable Net profits/total assets /a Model Model (1) (2) Intercept -0.282 -0.295 (-6.55) (-6.48) Total assets 0.030 0.031 (7.65) (7.47) Government equity -0.069 -0.063 (2.86) (-2.57) Loans/capital ratio -2.8 E-7 1.8 E-07 (-0.41) (-0.25) Foreign loans/total loans -0.035 -0.049 (-0.94) (-1.26) YEAR -0.030 -0.029 (-2.09) (-2.01) D2 -0.011 (-0.51) D3 0.064 (1.77) DA -0.042 (-0 .98) D5 0.020 (0.67) D6 -0.032 (-1.29) D7 0.006 (0.21) D9 0.019 (0.80) F-Value 14.29 6.76 Adjusted R-Squared 0.106 0.110 /a After adjusting for missing variables, and excluding Petronas and subsidiaries, there were 559 observations covering 1985 and 1986. Data are cross-section over a number of firms. Gapgter I MALAYSIA Projects Being Considered for Privatisation lre of Origin of Project Mode of project Sector proposal value privattistion ($ min) 1. Coal terminal rrade, storage Private sector 241 Outright sale or lease Port Klang 2. Network of sir- Transportetion Private sector n.a. n.a. ports (Subang, Pulau Pinsng. Kuching, tots Kinabalu, Lsngkawli Senai) 3. Malaysian navy Manufacturing Government 600 - Dockyard, Lumut 4. Penang Port Transportation & Government n.e. - storage 5. Bus services, Transportation 6 Government n.a. - Penang munict- storage pality 6. National Electri- Supply of electricity Government n.a. - city Board 7. Water supply, Water supply Government 750 Development of a new project Selangor by the private sector 8. Tourist complex, Hotel industry Government n.a. Outright sale/lease Desaru 9. Karkttng Of Com- B.asineg se.vic.- P-iate sector n.a. Manatement mercial advertis- ing time over TVI and TV2 10. Food catering for Food catering Private sector u.a. Contracting out hospitals 11. Sewerage, City Sewerage & waste Private sector 370 Development of a new project Rall. Kuala disposal by the private sector Lumpur 12. Viaduct/flyover Construction, trans- Government - a.a. construction over portation railway lines 13. Poet office Communication & post Private sector n.a. n.a. 14. PM stereo station Communication Private sector n.a. Development of a new project by the private sector 15. Port Klang (ex- Transportation & Government n.a. n.a. cluding container storage terminal) 16. Printing 6 sales Manufacturing Government n.s. Sale/lease dept., Dewan Bahasa 6 Pustaka 17. On-line proces- Business services Private sector u.a. Contract sing of national registration 18. Development of Construction Government n.a. Development of a new project land owned by by the private sector Urban Development Authority 19. PERDA Aquaculture Agriculture Government n.a. Development of a new project Project, Sg. by the private sector Pinang, Penang 20. Land development Construction Private sector n.a. New project to be developed by for Welfare Ser- the private sector vices Department 21. Government medi- Manufacturing, stor- Private sector n.a. - cal store age & trading Cho ter I - 58 - Atni page 2 of 2 Nam of Ortgin of Project Mode of project Sector proposl value privetiration (5 nln) 22. tllayan Railvay Transportation & Government n.a. Leasing storage 23. Happing services Business services Private sector n.a. Contracting out 24. Sate Rio Food services Private sector n.. Mnaegement 25. Herdeka Stadium Construction/services Government n.&. Development of a nev project car park by the private sector 26. Aerospace ground Vehicle maintenance Private sector n.a. Leastng support division services (afg) 27. PWD quarrtes llining & quarrying Private sector n.e. Leasing 28. Penang bridge transportation Government 800 n.a. 29. Advertising at Business services Private sector a.e. Management post offices 30. Water supply, Water supply Private sector n.a. n.a. Pulau Pinang 31. LLN new power Electricity supply Private sector n.a. Development of a new project generating plant by the private sector 32. Malaysian trana- Manufacturing Private sector - Sales of shares former 33. Socso Insurance Private sector - Management 34. Tanjong Malim Construction Government - Development of a new project -- station by the private sector 35. General Printing Privting (manufac- Private sector - n.a. Division of the turing) National Printing Department 36. Gov't quarters Services Private sector - Management 37. Gov't chalets Recreational services Private sector - Management 38. Registrar of con- Business services Private sector - Management pantes & business 39. Post-production Recreattonal services Private sector - Contract on processing of films 40. Totaltsator Board Business services Private sector n.a. n.a. 41. Cleaning of bore- Water supply Private sector n.a. Contract holds, Labuan 42. Johore Port Transportation & Government storage 43. Mini-hydro Supply of electricity Private sector - Development of a new project project by the private sector 44. Gov't pool - car Services Private sector n.a. Management services 45. Social welfare Business services Private sector n.a. n.a. lottery 46. Paddy processing Construction Private sector n.a. Development of a new project complex at by the private sector Seberang Perak 47. Latex processing Construction Private sector n.e. Development of a new project factory at Macang by the private sector Ielantan, Kelantan 48. Vehicle registra- Services Private sector n.a. n.a. tion _ 59 _ Chapter I Appendix 5 Page 1 of 2 MALAYSIA List of NFPEs That Have Been Sold qGency Companies 1) Negeiri Sembilan 1) Far East Processing SEDC 2) Perusahaan Azan Sdn. Bhd. 3) Cempake Negeri Laksmi Textile 4) Emas Negeri Sembilan Bhd. 5) Tenaga Baru Sdn. Bhd. 6) Senavang Jaya Sdn. Bhd. 7) Perusahaan Udang Negeri Sdn. Bhd. 8) Malaysia Poultry Sdn. Bhd. 9) Permas Negeri Sdn. Bhd. 2) Malacca SEDC 10) Malaka Elektronik Sdn. Bhd. 11) Syarikat Mona Industries Sdn. Bhd. 12) Pembangunan Mahamurni 13) United Strait Fuso Sdn. Bhd. 14) International Footwear Sdn. Bhd. 15) Syarikat Air Keroh Height Sdn. Bhd. JJ JXrU55@UD5 OVW~ 16) tSyarikat PA,pa T.. g Sdn. lid 4) Kedah SEDC 17) I.T. International Sdn. Bhid. 18) Kelga (M) Sdn. Bhd. 19) International Latex Incorporation 5) Johor SEDC 20) Bovis Ene 21) Syarikat Pengangkutan Senai 22) Pepper Trades (M) Sdn. Bhd. 23) Johor Estate Agencies Sdn. Bhd. 24) Pernas Jaya 25) Kahang Palm Oil Mills Sdn. Bhd. 6) Sabah SEDC 26) Angkasa Sabah 27) Kinabalu Boat Industries Sdn. Bhd. 28) Kumpulan Sabah Sdn. Bhd. 29) Sabah Shipyard Sdn. Bhd. 30) Kudat Edible Oil Sdn. Bhd. 31) Kilang Batu Bata Paper Sdn. 7) Selangor SADC 32) Ocean Land Sdn. Bhd. Chapter I - 60 - Appendix 5 Page 2 of 2 8) Perak SEDC 33) Galian Saham Sdn. Bhd. 34) Galian Keledang Sdn. Bhd. 35) Gula Perak Sdn. Bhd. 36) Timah Bersama Sdn. Bhd. 37) Ekoba Sdn. Bhd. 38) Maju Rapat Sdn. Bhd. 39) Jelutong Utara Sdn. Bhd. 40) Perwaras Sdn. Bhd. 41) Sri Lancang Sdn. Bhd. 42) Sri Bunga Sdn. Bhd. 43) Balak Kasteria Sdn. Bhd. 44) Kuala Lerek Sdn. Bhd. 45) Celok Sdn. Bhd. 46) Kayu Balak Ranan Sdn. Bhd. 47) Sri Pembalak Sdn. Bhd. 48) Sayong Budiman Sdn. Bhd. 49) Perkembaran Wangsa Sdn. Bhd. 50) Usaha Perba Sdn. shd. 51) Kaswani Sdn. Bhd. 52) Tiaran Sdn. Bhd. 53) Cahaya Bersatu Sdn. Bhd. 54) Melantai Sdn. Bhd. 55) Resi Sdn. Bhd. 56) Sinar Mersawa Sdn. Bhd. 57) Desa Sepakat Sdn. Bhd. c) T,iil4fna Ti9A Mdn. Bhd. 9) Sarawak SEDC 59) Sarawak Motor Industries Sdn. Bhd. 60) Industry Cold Storage Enterprises Sdn. Bhd. 61) General Carrier Sdn. Bhd. 62) Sarawak New Rubber Processing Sdn. Bhd. 63) Syarikat Tingan Lumber Sdn. Bhd. 64) Associated Metal Work 65) Perina Sdn. Bhd. 66) Process Utama Sdn. Bhd. 67) Smallholders Edible Oil Sdn. Bhd. (SHEO) 11) DARA 68) Batadara Sdn. Bhd. 12) KEJORA 69) Ladang Kejora-Luth Sdn. Bhd. 13) PERNAS 70) Berjaya Corporation Sdn. Bhd. 14) MISC 71) Gaya Shipping Sdn. Bhd. Note: SEDC - State Economic Development Corporation SADC - State Agricultural Development Corporation 1/ As of October, 1987. Chapter I 61 - Appendix 6 Page 1 MALAYSIA CICU's Z-score Model of Performance Evaluation 1. CICU has developed a technique of performance evaluation called the Z-score model, which it is applying to public enterprises within the corporate sector. Based upon a technique originally developed by Altman, the objective of the exercise is to choose those indicators of performance which are most significant in explaing (and predicting) the difference between problem and no problem NFPEs._ 2. The method involves the following steps: (a) selection of a sample of problem and no problem companies; (b) selection of performance indicators; (c) step-wise discriminant analysis to determine the statistical significance of various indicators and the selection of the most significant indicators for inclusion in the model; and (d) multiple discriminant analysis for determining relative weights to be assigned to the performance indicators chosen in step (c). 3. Utilizing a sample of 42 firms (21 problem firms and 21 no-problem firms), and starting with 41 performance indicators, CICU was able to identify the following seven financial ratios: A. Operating Profit/Total Liabilities B. Current Assets/Current Liabilities C. Profit after Interest and Taxes/Paid-up Capital D. Sales/Working Capital E. Current Assets - stocks - Current Liabilities/Operating Profit F. Total Shareholders Fund/Total Liabilities G. Ordinary Shareholders Fund/Employment of Capital. 4. Multiple discriminant analysis was utilized to estimate the equations with weights to the chosen indicators (ratios). These weights were 1/ For Altman's original formulation, see Altman, "Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy," Journal of Finance, September 1968. - 62 - Chapter I Appendix 6 Page 2 then used to calculate the Z scores for the no problem (ZNP) and the problem (ZP) firms. lo A B C D E F C ZNP -9.68 17.4 5.0 1.95 0.2 0.1 0.1 1.9 ZP -1.38 -5.8 -0.2 -1.1 -0.1 +0.1 -0.1 +0.3 If for any firm, the difference between ZNP and ZP : [DIFF = ZNP - ZP] is negative, it is interpreted to mean that the firm is a problem company, and vice versa. 5. The development of the Z-score model is the outcome of a growing realization of the need for performance evaluation of public enterprises. That concern is both legitimate and timely. However, the Z-score model suffers from several conceptual and practical problems which limit its usefulness as a device for improving managerial efficiency in the Malaysian context. These are: (a) it is overly complicated and difficult to comprehend; (b) the financial ratios chosen for inclusion in the Z-score are problematic: (i) they introduce double counting: thus profits are counted as the numerator in A and in C, and total liabilities are included as the denominator in A and F. (ii) the same indicator is used to increase and decrease the Z-score: thus an increase in operating profit increases A, but--because it is a denominator-decreases E. There is no logic why an increase in operating profit should be used to decrease the Z-score and thus penalize the management by making them look bad; (c) it is based upon a mechanical exercise that does not reflect the objectives of a particular enterprise; and (d) it does not provide a direct signal to the management, i.e., the Z-score is not readily convertible into meaningful targets which the management should strive to maximize, such as net or operating profit, productivity, etc. 6. Thus, the Z-score model may be a useful diagnostic device for ex- post evaluation of private firms but is of little value as a performance - 63 - Chapter I Appendix 6 Page 3 evaluation mechanism for management by objectives and for increasing the efficiency of public enterprises. - 64 - II. FISCAL DEFICITS IN MALAYSIA 1' A. Introduction 2.1 Government debt in Malaysia increased rapidly from the late 1970s to the early 1980s. The twin causes of the mounting debt were ambitious public expenditure programs promoting the objective of the New Economic Policy to restructure Malaysian society and the expectation of 'ontinued growth of revenues from oil and other exports. 2.2 To promote domestic economic growth, the Government had established statutory agencies and other nonfinancial public enterprises (NFPEs) to pursue social, industrial and agricultural development programs. Budgetary transfers to the NFPEs and statutory agencies became principal instruments of fiscal stimulus, in addition to the combined operations of the federal agencies. Many of these institutions were assigned the task of undertaking directly productive investments in industry, finance, and agriculture and some were charged with supporting programs of social restructuring. 2.3 The rise of public spending accelerated from 1978 to 1981. In the face of external price shocks (initially positive, with the 1979/80 oil price rise but later negative as other commodity prices fell), the Government maintained an expansionary fiscal stance as a countercyclical measure. Thus, while Government revenues were falling behind targets, public expenditures were maintained. The result of this policy was large fiscal deficits, which increased until 1982. Even today, these budget deficits continue to create a pattern of fiscal imbalance, given their debt service implications and the Government's continuing need to fund its high level of long-term expenditure commitments. 2.4 Fiscal deficits thus became a major macroeconomic problem by 1983. Continued fiscal expansion could not be sustained by domestic and foreign borrowing, so the danger of monetizing the debt was real. The Government, realizing the financial unsustainability of the level of imbalance, undertook the needed fiscal adjustment by instituting contractionary fiscal policies- reducing operating and development expenditures through cuts in development plan targets and through a variety of tax increases. The contraction cut deep; domestic economic growth fell. 2.5 The question for the future then is how to manage public finances so that a stable environment for _cnnomic growth is possible, in the context of sustainable fiscal deficits. To answer that question, it is essential to focus on the Government's fiscal policy stance and how it relates to the build 1/ Note the graphs and analysis in this section are based on preliminary data for 1987. New revised estimates indicate a substantially lower deficit than that presented here. Updates are reflected in Tables and text as appropriate. - 65 - up of fiscal deficits. In the process, the factors which contributed to the deficits can be quantified. This paper deals with the problem of deficits in Malaysia by, first, defining and quantifying the scope of the deficit, then by tracing the Government's fiscal stance over the period 1970-87 through an analysis of the disaggregated fiscal impulse, i.e., the tax, expenditure, cyclically neutral and discretionary impulses. The appropriateness of the policies in force at the time is then assessed, and details of public investments and subsidies are provided in the following section. Finally, the financing of the fiscal deficits and their sustainability given likely resources is assessed and priority tax reform measures are suggested in the context of the Government's twin goals of reducing the deficit and reforming the tax system. B. The Fiscal Deficit: Definition and Quantification Public Sector Revenues and Expenditures 2.6 In assessing the overall impact of fiscal activity in Malaysia, this report considers the fiscal operations of all levels of government, including the self-governing (statutory) authorities and NFPEs which sell goods and services to the general public. The NPPEs are considered in this comprehensive measure of the public sector since they all perform functions related to government objectives and are net users of government funds, through grants, loans or subsidies received either from federal agencies or other statutory bodies which have direct supervisory ovetsight of their operations. 2.7 Figure 2.1 shows consolidated public sector fiscal operations, including overall revenues and operational and developmeni,expenditures, of the federal government, the state governments, and NFPEs._/ As indicated in the figure, public expenditure has been outstripping public revenues by a substantial amount since 1979. This is true both for the federal government and for the consolidated public sector accounts. Three Measures of Fiscal Deficits 2.8 Two conventional measures of budget deficits are widely in use. The current deficit, also known as the operational deficit, represents the difference between current revenues and current expenditures; it gives a measure of public savings. The overall deficit indicates current revenues minus both current and capital expenditures, in other words, the current deficit minus capital expenditures. 2/ Data for NFPEs in this and subsequent charts are not comprehensive prior to 1981. Thereafter, data reflect consolidated information collected for some 56 major NFPEs. This is not considered a significant omission since the NFPEs were not larger users of government funds in the earlier period and many were not even operating then. - 66 - FiSre 2.1: Consolidated Revenues and Expenditures of the Public Sector 45 - 40- 36.' 30 25 - 20 1971 1972 1973 1974 197 1976 177 1978 19" 1960 161 196129319641966196 1967 0 FR + FCE O F£E A CR X CmE V CCE Note: PR a federal total revenues FCI - federal current (operating) expenditure FTM * federal total expenditure CR - consolidated public sector revenues CTE * consolidated public sector total expenditure (operating plus development) CC8 a consolidated current (operating) expenditure. - 67 - 2.9 The overall deficit is the conventional measure of the fiscal deficit. It can take two forms: the federal fiscal deficit, calculated on federal revenues and federal expenditures (both current and capital); and the more complete consolidated fiscal deficit which includes the fiscal accounts of the federal, state, and NFPE sectors. 2.10 A conceptually different measure of the deficit is the primary deficit. This is equivalent to the overall deficit net of interest payments and other nondiscretionary transfers like permanent subsidies and pensions and gratuities paid to government employees. The primary deficit is useful in providing information on how the current fiscal actions of the government relate to the improvement or worsening of the fiscal deficit. Under normal circumstances, the government has to pay all current obligations for which no discretionary element exists. The primary deficit meets this measure since the nondiscretionary components of expenditure are already removed. Essentially, it represents the deficit arising from the discretionary elements of government spending. Thus, if the primary deficit is falling, then discretionary public spending is falling, or at least tax revenues are covering an increasing share of it. The primary deficit thus provides critical information which the overall fiscal deficit fails to convey. Fiscal Imbalance: 1970-87 2.11 Figure 2.2 shows current and overall fiscal deficits from 1970 to 1987, based on federal, state, and NFPE accounts. As indicated in the figure, the overall federal deficit ranged from 61 to 101 of GNP during 1971-79, substantially lower than the consolidated public sector deficit of 101 to 161 of CUP during the same period. By 1980, the pattern of deficits shows a structural break in revenues as revenue growth begins to trail off and government investment programs grow. The deficit consequently increased to 16 of CNP in 1980, reaching almost 201 in 1981. The growth of public investment and expenditure was propelled by optimism about expected revenues from the country's main exports--oil, gas and other primary commodities--as well as a countercyclical expenditure policy to bolster the economy during a period of slack foreign demand. The major source of the deficits was related to capital spending (see paras. 3.51-3.55). In 1982-85, with the introduction of fiscal adjustment measures, the deficit began to fall, but rose again thereafter, indicating the need for further adjustment. 2.12 Throughout this period, with the exception of 1972 and 1987, the current budget showed a surplus. This ranged from 11 to 3% of GNP until 1980, and 51 to 9Z from 1981 to 1985. The current surplus was more modest in 1987. 2.13 The NFPEs' contribution to the overall federal deficit (including federal government and NFPE accounts) is shown in Figure 2.3. During 1973-80, the NFPE deficits ranged from 61 to 151 of total federal deficits, and accounted for 281 in 1984. In 1981 and 1985, howver, they contributed a surplus to the federal accounts due largely to the payment of dividends and royalties from oil enterprises. The NFPEs' dependence on government funds and the volatility of this dependence are discussed further below (paras. 2.56- 2.62). - 68 - Figure 2.2 Current and Federal Deficits Compared Current vs. Overall Deficits Fodeal v. CorwaIdefd 0.00 0. - 0.04 0 gL -0.02- i -0.04 - -0.08 -.0. 1 -0.12- -0.14- -0.16 -.0.18 -0.2- -022 | - 1-wW 197019711972197319741975197619771978197919'OW1g1 198219831984196519661987 O Cur* + 01 V Consafldabd owal Note: The budgetary positions of the NFPEs are integrated into the federal account. The deficits are shown as negative ratios since revenues vere less than expenditures. - 69 - Figure 2.3 Net Fiscal Position of Monitored NFPEs As a Percent of Overall Federal Deficits NFPE Fiscal Position An Rato X@ N O efcit 0.3 - 0.256 0.2- 0.15- b.t 0.1 0 0.06 -0.05 -0.1- I I I I I I I I I 1971 1972 1973 1974 1976 1976 1977 1978 1979 1980 1981 1962 1983 1984 196 11967 Note: The ratios are in the positive quadrant because the deficits of the NFPEs are divided by the deficits of the overall federal government. - 70 - The Primary Deficit 2.14 Figure 2.4 shows the primary deficit, defined first as net of combined debt service (composed mainly of interest payments), and second, as net of debt service and pensions and gratuities for government employees. During 1970-78, the primary deficit ranged from 31 to 7Z of GNP, with pensions and gratuities accounting for about 1Z of the total. During the period of expansionary fiscal policy in 1979-81, the primary deficit increased from 61 to 15Z of GNP, then declined considerably to register a slight surplus in 1985. As a result, overall deficits also declined. Thereafter, the rise of the primary deficit to 6Z of GNP in 1986 caused the overall deficit to grow from 8X in 1985 to 15Z in 1986. 2.15 The debt service burden during this period increased from 21 of GNP in 1971 to nearly 9X in 1987. The cost of pensions and gratuities remained relatively stable as a proportion of GNP; however, this obscures the fact that the GNP rose at a high rate over time and that maintaining the same proportion of pensions and gratuities could become an inordinate fiscal burden in times of slackened economic growth. 2.16 If the Government of Malaysia is to stabilize the fiscal front, it has to bring the primary fiscal deficit under control. This means reducing expenditures and financing any additional spending with tax revenues, thus ensuring that the level of public debt and subsequent debt service requirements do not rise. C. Fiscal Policy in Malaysia: 1972-87 Measurement of the Fiscal Impulse 2.17 The Government's fiscal policies during the period under review are best analyzed in terms of the fiscal impulse. The net fiscal impulse is a measure of the combined stance arising from fiscal activities of taxing and spending by government. It estimates the net stimulative or restrictive effects of the fiscal operations of the government in the current period and is therefore relevant to the assessment of impact on aggregate demand. The net impulse is policy-determined, resulting either from the introduction of new fiscal measures (changes in tax rates or the tax base, expansion or contraction of expenditure) or in the operation of previously existing measures which may have some automatic impact on the economy. 2.18 Measuring the net fiscal impulse over time provides some estimate of the effects of fiscal policy on the macroeconomy during the business cycle. The measure depends on potential (targeted) GNP at any given time. Calculation of the potential GNP is explained in Appendix 1 to this chapter. Given the performance of the Malaysini economy, an annual growth in CNP of 8Z to 10 in current terms (since inflation has been relatively low in the country) is reasonable, and this is the range of potential growth used as the basis for deriving the fiscal impulse. While lower rates would be reasonable in the context of the more restricted resources of the current period and the near future, the higher rates for potential GNP are appropriate when considering the prosperous 1970s. - 71 - Figure 2.4 Primary and Overall Deficits oai 001 Me 0- -0.01 -0.03 -0044 -0.06 -0.06 -0.07 t b-0.06- 5-0.09 -0.11 a-0.12- -0.13 -0.14 -0.16 -0.16 -0.17 -0.18 -0.1 9 -0.2 -0.21 - I I I I I I I I I I I I I I I I 197019711 9721973197419751976197719781979198019S1 182198319641S519861587 O P, inef DS + P, in DS+P& O 01wMI Note: Pt net of DS means primary deficit net of debt service (composed of interest payments); P, net of DS+P&G is primary deficit net of combined debt service and pensions & gratuities; FD/GNP is the ratio of the federal overall deficit to GNP; CD/CNP is the ratio of the consolidated public sector deficit to ONP. - 72 - 2.19 Figure 2.5 shows the net fiscal impulse as a ratio of GNP from 1972 to 1987, based on consolidated public sector accounts. The overall impact of government on the economy is best measured at this level of accounting. The potential GNP, whether growing at 10 or 81, makes little difference in the direction of the fiscal impulse, but affects the size of the net fiscal impulse. 2.20 The figure also shows the consolidated fiscal deficit (CD) of the public sector as a proportion of GNP (=CD/GNP). This demonstrates how the fiscal deficit moves with regard to the net fiscal impulse. The relationship is straightforward. A fiscal contraction reduces the deficit. A fiscal expansion raises the deficit. 2.21 The contractionary net fiscal impulse of 1973 equivalent to about 5% of GNP led to a reduction of the public sector deficit from 12% to 6Z of GNP. The relatively balanced net fiscal impulses from 1974 to 1979 caused little movement in the public sector deficit during that period as a percent of GNP. The large expansionary fiscal actions of 1979 to 1981 (rising from a contractionary net impulse of 21 of GNP in 1979, to 61 in 1980 and 1OX by 1981) caused the massive escalation of fiscal deficits from 8% in 1979 to 20% by 1981. Fiscal policy from 1983 to 1985 was contractionary again, and the net fiscal impulse was quite severe in impact (-11% of GNP in 1985). The effect of this contractionary fiscal policy reduced the public sector deficit to 7Z of GNP by 1985 and brought the federal primary deficit into a slight surplus. However, the decline in oil prices and general government revenue in 1986 caused the fiscal deficit to rise again. Thus, the 1980s have exhibited major fluctuations largely because of the stop-and-go fiscal policy. Disaggregating the Net Fiscal Impulse 2.22 The net fiscal impulse essentially has two components: the tax impulse and the expenditure impulse. Normally, these two effects work in opposite directions, and their combined effect produces the magnitude of the net fiscal impulse. 2.23 The Tax Impulse. An increase in tax rates usually has a contractionary effect, while a decrease in tax rates has the opposite effect. However, some adjustments in tax rates (either through their automatic impact on activities or their sliding scale effects on various tax payers) may produce different effects on the fiscal impulse. Figure 2.6 shows the tax impulse in Malaysia, as computed from two sets of average tax ratios for 1971-80 and 1976-86. As indicated by the figure, the tax ratios used in determining the tax impulse affect the magnitude of the impulse in terms of revenue gains or losses, but do not affect the direction of the impulse. Throughout 1972-87, the tax impulse was fairly stable and largely negative, and thus contractionary. The tax impulse was expansionary only in 1973, 1976 and 1984. - 73 - Figure 2.5 Net Fiscal Impulse, 1972-87 Net Fiscal Impulse ~~~~~~~~~~~wsiv Anas, a. 0.1 0.08 0.02 -0.02 -0.02 -0.06 -0.1 -0.12- -0.14 -0.16 -0.18 1972 1973 1974 1975 1976 1977 1978 1979 1980 1961 1962 193 1984 196 1966 1967 0 Al + A2 * A3 V CD/ONP Note: The legends in the chart represent assumptions about the potential for which the corresponding net fiscal impulse is calculated. Al and A2 are variations on the 0l potential GNP growth assumption; and A3 assumes 82 growth but is otherwise based on the same assumptions as Al. - 74 - Figure 2.6 The Ta Impulse, 1972-87 Tax Impulse 0.04 0.02 0.04 a.. 1-0. I -0.12- -0.14 -0.16 -0.18 -0.22 1972 1973 1974 1975 1976.1977 1978 1979 1960 1961 1962 1963 1984 1966 196 1967 0 1976-66 + 1971 -0 Y CD/OW - 75 - 2.24 Changes in tax policy in recent years have responded to the problem of the fiscal deficits. Given the difficulty of defining a single-direction for taxes (in view of the possibility of a negative response from citizens), they are often put together as a collection of tax changes with opposite effects: some are designed to provide relief, while others are designed to produce revenues. On a net basis, they are designed to raise revenues, and thus, the likely impact is a contractionary fiscal impulse. 2.25 Between 1980 and 1983, the tax impulse was negative but in 1984 it moved to a more expansionary stance. The major tax changes during 1980-81 were the increase in excise taxes for tobacco and liquor. Sates taxes were effectively reduced up to 1982: the threshold for imposition of the sales tax on small manufactures was raised from H$20,000 to M$1u0,000, and sales taxes on luxury items (designed for touristic promotion--cameras, w'atches, radios, etc.) were abolished. Personal income tax reliefs for individiduals were raised and other deductions for children were also increased. The road tax was raised steeply during this period, but the main revenue-increasing tax reforms were the doubling of the sales tax from 5 to 10X in 1983 and the increase of the service tax by the same rate (although this action was reversed in 1986 to 52). 2.26 The large contractionary tax impulse in 1985 was due to several measures which had compensating effects. Income taxes from non-petroleum sources were reduced but were more than compensated by increases in consumption-based commodity taxes. The reform of the individual income tax reduced the top marginal tax rate from 55Z to 402 to align it with the company marginal tax rate. This was accompanied by higher levels of deductions for dependent children. On the revenue raising side, excise tax rates on tobacco and liquor were converted to an ad valorem basis and excise taxes on locally assembled cars were increased. Import duties on a number of items were raised, including those on liquor and tobacco (already subjected to excise and sales taxes). A 50% flat rate on anncommercial goods brought into Malaysia was also imposed. In addition, decreases in the road tax for privately-owned and company-owned vehicles were matched by increases in car registration fees. 2.27 Tax policies instituted in 1986 were largely expansionary since they liberalized the fiscal incentives system, which was already quite generous. Liberalization of depreciation allowances, extension of the investment tax credit, and special incentives for special industries like shipping all provided a reduction of the effective tax rate and therefore created a basis for tax erosion. 2.28 An increase in the sales tax rate and the removal of some exemptions from it were introduced in 1987. While this has raised some revenues, the full impact has still to be realized although revenues during this year have risen already. 2.29 The lesson emerging from this is that while the tax impulse can be manipulated by policy, it is essentially stable. Thus, if fiscal policy is linked with taxation much more directly, the volatility of economic performance will also likely disappear. The Government has demonstrated that it can raise revenues when needed, but tax changes have taken place in an - 76 - environment in which no sharp tax increases have been undertaken without some balancing with revenue-reducing actions. The sharp tax reductions that have been introduced were undertaken often to provide incentives. Such results can come about from increases in allowances and in deductions for income taxes (both company and individiAal) and from raising threshold levels for coverage under the sales and service taxes. 2.30 Malaysian export taxes have some sliding tax rates that are predetermined by administrative ceilings on the effective threshold price for the rates. In this way, they capture the price premium during boom times, but, when prices fall, the taxes slide back to lower rates. By adjusting threshold prices, the effective tax can be influenced. Thus, when rubber prices fell sharply to the detriment of -ibber producers, an increase in threshold prices for rubber reduced the *ctive tax rate. Such measures have also been applied to tin and palm t and the outright abolition of some export taxes, such as those on processe..- igricultural items like processed palm oil, has also been tried recently o give a measure of tax relief to adversely affected sectors. 2.31 Tax rates on petroleum products for domestic consumption and for exports are also a major tax variable. The imposition of a 25Z export tax on petroleum in the late 1970s was a major element in the revenue picture for Malaysia during an earlier period. But the increases in administrative reference prices for the taxes on petroleum are a matter for periodic revision. In addition, dividends from the profits of PETRONAS are non-tax revenues to the Treasury, and these can be set as a matter of government policy. Although as a company PETRONAS can determine its dividend rate, the revenue situation of the federal government determines its dividend policy. 2.32 The Expenditure Impulse. Figure 2.7 shows the expenditure impulse. While the tax impulse reflects a series of offsetting tax policy actions, the expenditure impulse is largely a series of curges reflecting expansion or contraction of the public expenditure level in order to achieve a targeted GNP growth rate. The expenditure impulse is almost perfectly correlated with the net fiscal impulse and appears as a reflection of the net fiscal impulse shown in Figure 2.5. 2.33 The surges in the expenditure level come from increases in government operating costs, increases in investment expenditure, and the growth in the number and levels of budgetary transfers to :he statutory authorities and state governments. The volume of these tr.,nsfers rose from 16% of total federal expenditure in the early 1970s to 30% by 1980. Although this proportion fell in 1982, it has risen again to about 28% of federal expenditure or about 11% of GNP. In 1986-87, as a result of Government's countercyclical expenditure policy to reflate the economy, the expenditure impulse grew, as did the positive tax impulse, owing to the liberalization of tax incentives. - 77 - Figure 2.7 The Expenditure Impulse, 1972-87 Expenditure Impulse Seniteiy Ana 0.12 - 0.1 0.06 0.04 0.02 *9 -02 .02t o9 -0.06 a -0.08 -01M -0.12 -0.14- -0.16 -0.18 -0.2 -0-22 1972 1973 1974 1975 1976 1977 1978 1979 I9O 1981 1902 1983 1984 1960 1986 1987 0 1976-66 * + 1971-0 v CD/GN - 78 - 2.34 Because of the relative magnitude of the expenditure impulse, it often swamps butt as indicated above, may sometimes reinforce the tax impulse. During 1979-81, the large expansionary impulse from public expenditures swamped the milder contractionary tax policy; thereafter, with the need for fiscal adjustment and deficit reduction, both the expenditure impulse and tax policy were contractionary. Generally, however, because of its magnitude, expenditure policy has played a destabilizing role in the performance of the economy by creating wide fluctuations in macroeconomic behavior. Adequacy of Fiscal Policy 2.35 The net fiscal impulse can be disaggregated into two components: a cyclically neutral fiscal impulse and a discretionary impulse. The cyclically neutral deficit would be the deficit (or surplus) level required to maintain a specific level of potential (targeted) GNP, given the structure of tax and expenditure ratios during the base year. The discretionary fiscal impulse should match this cyclically neutral impulse, so that it is adequate. Whether or not the net fiscal impulse is adequate, excessive, or inadequate depends, of course, on the relative objective of economic policy. 2.36 Cyclically Neutral Deficits Malaysian fiscal policy during the 1980s has not been matched by the appropriate net fiscal impulse, which either overshot the required deficit or fell short of it. Figure 2.8 depicts the country's net fiscal impulse, the cyclically neutral deficits and the consolidated deficits. The line representing the cyclically neutral deficit as a percentage of GNP indicates gradual changes, very similar to the fairly stable character of the tax impulse. However, the line showing the net fiscal impulse is volatile, reflecting the volatile nature of the Government's expenditure impulse. When the neutral deficit required a mildly contrac- tionary deficit (in the range of -2% of GNP) during 1979 to 1982, the net fiscal impulse was expansionary, at 7% and 9Z of GNP. During the period of contractionary fiscal policy between 1982 and 1985, the net fiscal impulse was almost twice the level required to maintain the potential GNP. Thus, the tight fiscal policy contracted the economy much more than was required. The danger for fiscal policy is that a countercyclical expenditure policy can go off-target and produce more instability than desired. 2.37 The level of the potential (or targeted) GNP can play a significant role. A higher level of potential GNP requires a higher net fiscal impulse to make it cyclically neutral. In 1985, the cyclically neutral deficit given the potential GNP level associated with a 10% growth in GNP (assumption Al) was 6.2% of GNP. Under ,he alternative assumption that potential GNP grows at 8% (A3), the required cyclically neutral deficit was -5.2X of GNP. However, the actual net fiscal impulse in that year was more procyclical at about -10% of GNP, more contractionary than required for the potential GNP. - 79 - Figure 2.8 The Net Fiscal Impulse and Cyclically Neutral Deficits NF' & Neutral Deficits 1971 -0 1. 976-96 0.1 0.06 -=_' A o -°2 \ ' 0.04 0.022 -0. ~-0.04- -0.02 0.1 -0.16 -0.18 -0.2 -0.22 - ,. 1972 1973 1974 1975 1976 1977 1978 1979 1960 1961 1962 1963 1984 1966 1986 1967 4+ CND71-60 , NFIL76-86 A Nf-.71-60 Note: CUD:76-86 - cyclical neutral deficit, computed on 1976-86 taz and expenditure efforts. NFI:71-80 net fiscal impulse, computed on 1971-80 tax and expenditure efforts. - 80 - 2.38 In 1987 (based on the expenditure and revenue estimates by midyear), the net fiscal impulse is much more expansionary than required. The cyclically neutral deficit is about 4.7X of GNP, but the actual net fiscal impulse is set at 7.7Z of CNP, certainly much more expansionary. Under the lower level of potential CNP, the required cyclically neutral fiscal impulse is about 4.1Z of GNP. This involves a substantial amount of fiscal resources in any given year. Clearly, a lower level of targeted CNP (the potential GNP) is consistent with a lower level of deficits. 2.39 Prospects for Countercyclical Fiscal Policy. A fiscal policy aimed at sustaining a high level of economic growth requires a high level of public expenditure. This cannot be accomplished by the more gentle growth of tax revenues. However, since revenues are dependent on volatile factors such as the terms of trade for a country's exports, public expenditure should be set at lower target levels to ensure a more stable fiscal policy over the medium term and avoid the need for countercyclical spending. 2.40 Malaysia's countercyclical fiscal action to stimulate demand caused expenditure levels to exceed the levels required to maintain the potential GNP. These expenditures, however, failed to produce the anticipated results since they involved long-gestation activities. To be effective, countercyclical expenditures should be used for short-duration activities so that expenditures can be decreased when no longer required. They should essentially be a temporary relief activity during times of slack demand. 2B.41 Countercyclical spending should not be based on development expenditures requiring longer-term committments of public funds. Such development projects should be conceived on the basis of development criteria, not countercyclical policy demand. Countercyclical expenditure policy is therefore more suitable for expenditures on public maintenance of established infrastructure, which can begin at short notice or be postponed. These expenditures are not lumpy and have no large impact on future commitments of funds. Since they are related only to the level of existing public infrastructure, their total volume is also within control and the possibility of overshooting the required fiscal impulse during times of poor aggregate demand in the economy is less. Subsidies and the Fiscal Deficit 2.42 As already stated, statutory authorities were created with specific social and economic goals, many of them linked with the NEP. To operate, the authorities required federal budgetary transfer which resulted in a build-up of subsidies. During times of abundant budgetary resources, the cost of these subsidies was not noticed, but during times of extreme budgetary constraint, they became a source of fiscal inflexibility, since they were not necessarily discretionary public expenditures. 2.43 Problems of Estimating Subsidies. Direct subsidies, such as those listed in budget documents, are easy to tabulate, but most subsidies are indirect and cannot be easily monitored. Budgetary transfers to agencies for purposes of equity capitalization or loans (that in the end do not get repaid or loans at reduced interest rates) appear as claims on assets with an - 81 - expected return in the future. In some cases, the implied current subsidies for some operations are only found out years later, when the concerned agency or enterprise seeks additional cash infusion or when it is reorganized or liquidated. Only at these points is the cost of the budgetary transfers transparent. 2.44 Most subsidies in Malaysia are implicit or ind rect, and thus almost impossible to estimate directly (this would involve estimating the difference between market-determined costs of government-provided services and user prices). Most of these are associated with program expenditures of the statutory authorities and NFPEs involved in agriculture, housing, stock purchase, transportation, credit, health care, education, and industry. The Malayan Railways, for example, receives indirect subsidies through its exemption from sales taxes and duties since 1982, the cancellation in 1982 of M$12.5 million in accumulated pension liabilities owed to the Government, and its release from subsequent pension liabilities in 1983-87. Other types of indirect subsidies include housing obtained at below-market cost and subsidized through low-interest loans, stock equity ownership programs (sales of stock at low values) and agricultural credit subsidies. 2.45 Estimated Cost of Subsidies. Transparent subsidies provided through the government budget (see Table 2.1) are negligible as a proportion of total government expenditure. As shown in Figure 2.9, transparent subsidies have been drastically reduced since 1983, indicating the Government's ability to deal with them effectively when necessary. The subsidy components shown in Table 2.1 have been treated like discretionary expenditure budgets, which were pruned by fiscal decisions in the current year. 2.46 The less transparent the subsidy, the more it will affect the fiscal burden since the government cannot deal with it effectively. Based on an indirect estimate of implicit subsidies (by subtracting total government expenditure as recorded in the budget from demand for real goods and servic.s as recorded in the national income accounts), total government transfers and subsidies in Malaysia have been estimated for the period 1970-87. As shown in Figure 2.10(b), all subsidies, both direct and indirect, represent a significant share of total public expenditures and, as a consequence, of GNP. At their peak in 1980, subsidies and transfers accounted for 57Z of consolidated government expenditures, but declined to 37% by 1982, then rose again to nearly 401 by 1987. As a proportion of GNP, subsidies dipped from 31% in 1981 to between 18Z and 21% in 1983-87. Even though these measures are only indirectly derived, and even assuming that they may be overestimated, their magnitudes in terms of their fiscal burden are quite sizable. Table 2.1: "TRANSPARENT" SUBSIDIES IN MALAYSIA (M$ thousand) 1976-gO 1979 1980 1981 1982 1983 1984 1985 1986 1987 Operating Budget 1,027.8 165.9 713.9 944.4 987 1,360.3 437.7 289.6 284.7 304.8 Petroleum products 783.3 129.2 575.5 721.3 750 1090.9 172 9.2 8 3 Textbooks 120.3 36.7 14.2 22.4 30 34.4 26.8 41.5 41.5 41.5 National Electricity Board 13.6 - 13.6 19.5 21 52.6 60.2 68.9 31.8 39.7 Paddy price support 110.6 - 110.6 181.2 186 182.4 178.7 170 185 205 Others /a - - - - - - - - 18.4 15.6 Development Budget 595.2 132.3 258.2 262.3 306.8 313.2 295.9 138.7 137.9 190.2 Rubber replanting 215.9 48.4 87.9 54.7 100.9 95.3 74.9 33.6 20.5 89.3 Pineapple replanting 10.9 1.8 2.2 2 2.9 4.3 2.2 3.7 1.6 2.4 Agriculture/Fisheries asst. 52.6 14.4 21.6 22.2 15.9 31.1 36.7 - - - Vegetable price support - - - - - 20 6.4 - - - X Agricultural inputs 126.7 26.8 99.9 118.2 97 90 75.8 74.8 98.5 80.5 Farm mechanization 15.2 5.3 3.6 4.9 27.2 20 16 - 1.4 1.t Coconut replanting 32.2 7.1 9.3 9.7 10.9 13 13.6 13.5 6.5 6.9 Crop diversification 71.6 18.4 17.9 13.7 15.6 17 0.8 - - - Animal husbandry 5.4 1.7 1.6 15.7 17.5 9.8 11.8 13.1 9.4 9.4 Other 64.7 8.4 14.2 21.2 18.9 12.7 57.7 - - - Total 1,623 298.2 972.1 1,206.7 1,293.8 1,673.5 733.6 428.3 422.6 495 /a Others includes subsidies not elsewhere reported earlier, including nutritional subsidy and public license subsidy. Note: This table refers to identifiable subsidy schemes only; it excludes a number of other schemes which appear in the development budget but are difficult to disaggregate. It also excludes interest rate subsidies to state and local governments, public and quasipublic entities and cooperatives. Ia MWoa M I. oooo PH _ I 0 9 \~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~3 - 84 - Figure 2.10(a). Total Federal Expenditures and Estimates of Subsidies and Transfers, 1970-1987 Total Expenditures & Subsidies 40 56 30 26- 20- 16 . 10 19701971 1972197319741976i97619771978197919tS0t1 1g218S1lg4t19e5 1seEg7 0 [KP- + 4. k Figure 2.10(b). Subsidies and Transfers as Ratio to Total Public Expenditure and to GNP, 1970-87 Subsidies as Ratio To P Ce Efdhre h to GNP 0.0 0.44 I 0.2- 0.1 19701971 1972197319741975197619771978197919t061 1198219851984 1 Mg I 96 1967 a Re; to tEp. - Rtio to GN - 85 - D. Public Investments and Government Subsidies 2.47 Public Investments. Encouraged by booming exports, the Government launched a major public investment effort in the 1970s. This program carried on into the 1980s, utilizing the various statutory authorities and NFPEs as agents for accomplishing the task. As a result of the program, development spending in 1979 was 3.3 times the level of spending in 1971, while 1982 development spending at M$11.4 billion was 2.7 times that of 1979 (see Table 2.2). In 1982, a quarter of all development spending was attributable to the NFPEs. The fiscal crunch of 1981 led to drastic reductions in these expenditures, mainly through postponement of programmed capital spending and of the more discretionary components of current expenditure. Figure 2.11 shows the composition of investment by level of government activity--federal, state and NFPEs. As shown in the figure, the rise of federal public investment tapered off after 1981 and continued to grow slowly until 1983, which marked the peak in these expenditures. Thereafter, spending plummeted and has remained at low levels ever since. Public investment by the state governments remained at about the same level over this period, indicating that the engine of growth in the public sector is the federal government. 2.48 While the Government was undertaking a fiscal retrenchment, public investment by NFPEs continued to rise. The delay in controlling the fiscal operations and investment of NFPEs indicates the extent to which they were not brought into line with the fiscal stance. The eventual fall in NFPE invest- ment, which is evident after 1984, occurred two years after the major fiscal ! cuts had been ordered by the budget authorities. The decline of investment (in absolute value) by the NFPEs after 1984, however, indicates the general effect of the recessionary conditions in the economy as well as the delayed effect of the overall budgetary situation of the federal government. The relative share of NFPEs in overall public investment has become substantial, accounting for close to 45% of total public investment during 1984-86 and one third of the total in 1987. Thus, although the level of total public investment has fallen, NFPE expenditures have been affected to a lesser extent. 2.49 Figure 2.12 shows public investment by type of investment. During 1978-87, physical infrastructure accounted for at least 50% of total investment, with energy, communications and transport having the largest shares. Investment in social infrastructure (education, housing, and health) accounted for 10% to 17% of total public investment since 1979. 2.50 Public investment in directly productive activities (Table 2.2) is dominated by investments in commerce, industry and land development, which in peak years (1978 and 1987) accounted for about one fourth of total public investment. Because investment in energy and communication also includes investment in the petroleum industry, public investment in productive indus- tties is really much larger than the levels shown in the chart. Activities in land development have largely involved FELDA and urban development schemes. Public investment in industry was concentrated in heavy industry. 2.51 These estimates of investment in the public sector are based on budgetary expenditures and may not indicate the full extent of public - 86 - Table 2.2: PUBLIC INVESTMENT IN MALAYSIA, 1978-1987 (In million M$) Sector of investment 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 Commerce and industry 320 120 431 659 874 1,126 1,479 1,434 167 1,545 Energy and communication 521 774 2,011 2,148 3,169 4,039 4,986 4,624 3,938 2,455 Housing 140 173 260 593 371 528 252 296 271 320 Drainage and irrigation 114 140 220 339 411 308 261 225 161 96 Agriculture 170 265 446 503 493 299 302 376 503 648 Research and development 24 19 62 47 68 47 20 41 47 53 Education 73 324 530 747 1,024 943 951 851 1,037 1,066 Transport 594 672 971 1,215 2,012 1,650 1,664 1,380 1,404 606 Tourism - - - - - - - - 1 2 Water supply 231 257 351 492 553 700 633 730 485 687 Land development 379 428 263 331 462 335 381 980 233 120 Health 37 46 61 96 126 150 118 108 113 88 Others 816 917 597 2,123 1,815 2,396 999 1,202 2,865 918 Total Public Investment 3,419 4,135 6,203 9,293 11,378 12,521 12,046 12,247 11,225 8,6'4 COMPOSITION OF PUBLIC INVESTMENT, BY TYPE OF INVESTMENT (in percent) Public investment by sector 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 Physical Infrastructure 53.79 54.92 61.52 48.69 58.07 56.16 65.79 64.82 55.42 44.68 Energy & communication 15.24 18.72 32.42 23.11 27.85 32.26 41.39 37.76 35.08 28.53 Drainage & irrigation 3.33 3.39 3.55 3.65 3.61 2.46 2.17 1.84 1.43 1.12 Transport 17.37 16.25 15.65 13.07 17.68 13.18 13.81 11.27 12.51 7.04 Water supply 6.76 6.22 5.66 5.29 4.86 5.59 5.25 5.96 4.32 7.98 Land development 11.09 10.35 4.24 3.56 4.06 2.68 3.16 8.00 2.08 1.39 Social Infrastructure 3.22 8.95 9.53 9.07 10.11 8.73 8.87 7.83 10.24 13.41 Education 2.14 7.84 8.54 8.04 9.00 7.53 7.89 6.95 9.24 12.39 Health 1.08 1.11 0.98 1.03 1.11 1.20 0.98 0.88 1.01 1.02 Productive Activity 18.43 13.49 18.33 18.89 15.28 15.60 16.88 17.20 8.39 29.23 Agriculture 4.97 6.41 7.19 5.41 4.33 2.39 2.51 3.07 4.48 7.53 Commerce & industry 9.36 2.90 6.95 7.09 7.68 8.99 12.28 11.71 1.49 17.96 Housing 4.09 4.18 4.19 6.38 3.26 4.22 2.09 2.42 2.41 3.72 Tourism 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.02 Others 24.57 22.64 10.62 23.35 16.55 19.51 8.46 10.15 25.94 11.29 Research & development 0.70 0.46 1.00 0.51 0.60 0.38 0.17 0.33 0.42 0.62 Others 23.87 22.18 9.62 22.85 15.95 19.14 8.29 9.81 25.52 10.67 COMPOSITION OF PUBLIC INVESTMENT, BY LEVEL OF GOVERNMENT ACTIVITY Level of Government 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 Federal 69.00 70.45 54.25 58.84 58.37 56.53 40.46 42.70 50.29Z 60.412 State 18.89 21.35 20.23 13.97 11.60 7.13 9.99 10.20 7.842 7.442 NFPEs 12.11 8.20 25.52 27.19 30.03 36.34 49.55 47.11 41.872 32.15X - 87 - Figure 2. 11 Public Inavestmnt by Level of Gove t In Billion NS 13* 12 - 11 10 0. E~~981" 1g 8110 ~ 941W U~ ~~osktWF -88- Figure 2.12: COMPOSITION OF PUB3LIC INVESTMEN Composition of Public Investment Cw - Ratiao 1.5-- - -- 0.2 I - 0.6 1 a~~~// 198 97 98 181 182 193 194 98 9- 1F E '7r7 Xi PoutveOhr - 89 - investments that are maturing because of the gestation lag. The numbers in themeelves also cannot indicate the productivity of Lhe investment financed, i.e., whether or not the expenditure was justified. In fact, the construction boom in Kuala Lumpur and other cities, which was led in part by the government sector, has resulted in overbuilding and a large stock of unsold properties. Moreover, the relatively high proportion of investments in directly productive activities raises the question of the overall role of the Government in productive enterprise. Indeed, some of the major government investments in heavy industry have yet to yield the projected returns. This has led to heavy losses and the requirement of budgetary support. One particularly disturbing consequence of these large investments in heavy industry has been a crowding out of the private sector. Statutory Authorities and Fiscal Deficits 2.52 At present, some 72 statutory authorities operate under Government authority. Statutory agencies are different from non-financial public enterprises only in that they are created by an enabling law of Parliamen' with specific authorization and objectives, rather than being organized under the Company Act. In performing their functions as agents of development and social policy, the statutory authorities finance their operations from government budgetary transfers and from their own revenues. While the authorities and NFPEs are in theory supervised by the responsible ministries, their operations are basically autonomous. 2.53 The financial operations of these entities are very difficult to assess. This arises in part from the power of the statutory authorities to create new companies under the Companies Act, 1865, and to acquire companies in fields related to their own operations. With this, the fiscal operations of the government off-budget agencies become tangled, making it difficult to assess the overall fiscal situation of the authorities and of the Government in general. An additional difficulty is the ambiguity of control for fiscal purposes. The statutory authorities, being autonomous, are not accountable directly to the fiscal authorities. While the agencies receive periodic budgetary transfers to support specific or supplementary programs, they may be financially independent for a long period of time, operating on the basis of an earlier capitalization and deriving revenues for their operations. In fact, many statutory authorities have liquid financial positions and place their funds in the market. 2.54 One consequence of this arrangement--financial dependence on budgetary transfers for some programs and the abl.lity to carry on independent financial operations for current activities--is to foster lack of financial accountability to the fiscal authorities and to encourage less reliance on self-financing of programs through pricing policies. This occurs particularly during periods of fiscal prosperity. The agencies in surplus also tend to accumulate cash reserves without ever providing the Treasury with a "reverse" transfer (in the form of dividends and repayments of loans). 2.55 These pxoblems became apparent during the period of fiscal tightness, especially after 1982. Table 2.3 shows the budgetary positions of statutory authorities for the three years, 1985 to 1987 (data for 1987 are - 90 - Table 2.3: BUDGETARY POSITION OF STATUTORY AITHORITIES, 1985-87 BY MINISTRY OF SUPERVISION (in million NS) Min.of Nin.of HUn.of Hin.of Hin.of Min.of Nin.of Hin.of Other Winistry of Supervision-> T61 Educ. Labour LbRD Pr.Ind. R4RD Agr. C6Sport Nina. Total 1985 (Actual) Total revenue snd grants 80.77 939.81 1.18 784.02 115.92 398.86 1,077.86 8.35 196.42 3,603.19 Current rev. 6.78 120.78 1.18 209.34 105.22 41.12 527.15 6.59 175.57 1,193.73 Grants 73.99 819.03 0.00 574.68 10.69 357.74 550.71 1.77 20.85 2,409.46 Current expenditure 63.69 607.41 1.05 369.90 101.42 210.11 944.90 5.71 230.38 2,534.57 Capital expenditure 3.04 236.51 0.00 742.54 27.98 41.67 62.49 1.06 2.80 1,118.10 1986 (Eeetnatd atual) Totar renu c m grants 63.91 951.60 1.04 789.11 121.60 432.19 1,084.83 5.23 576.48 4,025.98 Current rev. 7.56 143.63 1.04 199.97 110.37 48.87 540.32 3.73 219.10 1,274.58 Granta 56.35 807.96 0.00 589.14 11.23 383.33 544.51 1.50 357.38 2,751.39 Current expenditure 61.67 678.95 1.06 421.72 104.34 225.53 1,001.51 6.28 268.03 2,769.09 Capital expenditure 2.38 190.93 0.00 745.69 12.91 28.61 56.86 0.19 15.87 1,053.44 18(Latest estiuts Total revenu and grants 78.63 977.37 0.96 778.64 118.27 417.53 1,155.47 30.10 260.24 3,817.22 Current rev. 7.08 127.01 0.96 172.23 104.52 44.81 630.92 28.60 232.96 1,349.08 Grants 71.56 850.37 0.00 606.41 13.76 372.72 524.55 1.50 27.28 2,468.14 Current expenditure 73.12 789.60 1.05 425.44 109.14 259.31 1,105.09 7.05 404.89 3,174.69 Capital expenditure 7.46 214.37 0.00 800.83 1.95 37.98 66.52 0.90 1.99 1,132.01 1985-1987 (totals) Total revenue and grants 223.31 2,868.78 3.18 2,351.77 355.79 1,248.58 3,318.16 43.68 1,033.14 11,446.39 Current rev. 21.42 391.42 3.18 581.54 320.11 134.80 1,698.39 38.91 627.63 3,817.40 Grants 201.89 2,477.36 0.00 1,770.23 35.67 1,113.79 1,619.77 4.77 405.51 7,628.99 Current expenditure 198.49 2,075.96 3.15 1,217.06 314.91 694.95 3,051.50 19.04 903.30 8,478.36 Capital expenditure 12.89 641.82 0.00 2,289.06 42.84 108.27 185.87 2.1S 20.66 3,303.54 Percentage distribution 1985-1987 (totals) Revenues 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 Current rev. 9.59 13.64 100.00 24.73 89.97 10.80 51.18 89.09 60.75 33.35 Grants 90.41 86.36 0.00 75.27 10.03 89.20 48.82 10.91 39.25 66.65 Expenditures 100.00 100.00 100.00 100.00 100.00 iOO.0O 100.00 100.00 100.00 100.00 Current expenditure 93.90 76.38 99.94 34.71 88.03 86.52 94.26 89.87 97.76 71.96 Capital expenditure 6.10 23.62 0.06 65.29 11.97 13.48 5.74 10.13 2.24 28.04 Source: Federal Treasury, Malaysia. - 91 - latest estimates by the Treasury). The statutory authorities with the largest fiscal programs and level of reliance on budgetary transfers from the federal government were those under the Ministry of Education (with 86.3X from grants), the Ministry of National and Rural Development (89Z), the Ministry of Land and Regional Development (75%) and the Ministry of Agriculture (48.82). Average grants as a percentage of revenues was 66.6%. 2.56 Table 2.4 shows the budgetary positions of these authorities in terms of their overall fiscal operations. Three types of deficits are shown in the table: first, current revenues versus total current expenditures; second, total revenue income (own current revenues plus grants) compared to current expenditure, to show total current deficits; and third, total revenues minus all expenditures (current and capital), which is analogous to the overall budgetary deficit. 2.57 In terms of current expenditures compared to current own-revenues, the authorities had a deficit of M$4,661 million for the three years 1985-87 or an average M$1,554 million deficit per year. When grants are added to their current revenues, they had a surplus of M$2,968 million or just under M$1 million per year. When the overall deficit is taken into account, the statutory authorities had a deficit of M$336 million or M$112 million per year. The authorities with the highest budget deficits were under the Minis- tries of Education, Land and Regional Development, and Agriculture. Together, these authorities accounted for M$3,673 million or 79% of current deficits for the three years. 2.58 The federal statutory authorities may not be able to attain budgetary surpluses in their operations, but they could at least reduce their budgetary dependence. This will be critical to the Government's efforts to reduce costs and raise public savings. In connection with this objective, the Government will have to clarify its policy on reducing budgetary transfers to the authorities, strengthen financial controls, devise a mechanism whereby surplus revenues are returned to the Treasury (as dividends or repayment of principal from the initial capitalization of the agencies) and decide on the disposition of agencies that have not met the objectives for which they were created. Regarding the latter proposai, Government has already decided to privatize some public enterprises, most notably the Malaysian Airlines System and the Malaysian International Shipping Company, and these entities have been put up for sale. E. The Financing of a Sustainable Deficit: Tax Reform Financing the Deficit: 1971-87 2.59 Figure 2.13 shows the proportion of the real deficit financing from 1971 to 1987. As fiscal deficits accumulated in the 1970s, the high level of domestic savings in Malaysia and the strong balance of payments performance (due to oil and other exports) created a huge margin of comfort. Consequently, the moderate fiscal deficits incurred before 1980 could be financed mainly from domestic borrowing. In particular, the savings generated by the Employee Provident Fund and similar compulsory savings schemes (e.g., the Armed Forces and the Teachers' Provident Funds) was a captive source of Table 2.4: FISCAL DEFICITS OF STATUTORY AUTHORITIES, 1985-1987 (in million M$) Min.of Min.of Min.of Min.of Min.of Min.of Min.of Min.of Other Ministry of Supervision-> T&I Educ. Labour L&RD Pr.Ind. N&RD Agr. C&Sport Mins. Total Deficit concept: 1985 Actual Revenue-expenditure -56.92 -486.63 0.14 -160.56 3.80 -168.99 -417.76 0.88 -54.81 -1,340.8 Revenue+grants-expenditure 17.08 332.40 0.14 414.12 14.49 188.75 132.95 2.65 -33.96 1,068.6 Revenue+grants-total expenditure 14.04 95.90 0.13 -328.42 -13.48 147.08 70.4b 1.59 -36.76 -49.5 1986 estimated actual Revenue-expenditure -54.11 -535.32 -0.02 -221.76 6.03 -176.66 -461.18 -2.55 -48.93 -1,494.5 Revenue+grants-expenditure 2.23 272.64 -0.02 367.39 17.26 206.66 83.33 -1.05 308.45 1,256.9 1 Revenue+grants-total % expenditure -0.15 81.79 -0.02 -378.30 4.34 178.05 26.47 -1.24 292.58 203.5 W 1987 (latest estimates) Revenue-expenditure -66.05 -662.60 -0.09 -253.21 -4.63 -214.50 -474.17 21.55 -171.93 -1,825.6 Revenue+grants-expenditure 5.51 187.77 -0.91 353.20 9.13 158.22 50.38 23.05 -144.65 642.5 Revenue+grants-total expenditure -1.95 -26.60 -0.91 -447.637 7.18 120.24 -16.14 22.15 -146.64 -489.51 Overall Budgetary Position for 1985 to 1987 Revenue-expenditure -177.08 -1,684.55 0.03 -635.53 5.20 -560.15 -1,353.11 19.88 -275.67 -4,660.90 Revenue+grants-expenditure 24.82 792.82 0.03 1,134.71 40.88 553.64 266.66 24.64 129.84 2,968.0 Revenue+grants-total expenditure 11.93 151.00 0.03 -1,154.35 -1.96 445.36 80.79 22.49 109.13 -335.51 Note: Revenue refers to current revenue; expenditure to current expenditure; grants to federal grants; and total expenditure to current and capital expenditures. Source: Treasury, Malaysia. - 93 - Figure 2.13 Financing the Real isal Deficit, 1971-1987 (Percent Distribution fr Soure of Ficlaing) Financing of the Real Deficit 0.S8 0.4 0.3- 0.2 1971197219731974197519761977 19781979 1 noIgal I102 II96" I Ml19S17 Dow Co 12FlON Ing ~ E f Ts DlnCn = Domestic borrowin; MM = Forign Borrowing; Sdpe = Beigorag; IS Ti = Inflation Tax. - 94 - financing. Surpluses generated by PETRONAS vere also available to finance the U public debt, and public savings during this period was likewise relatively high because of the surplus on the current budget (see Figure 2.2 for the current surplus). 2.60 Because of relative price stability, the rewards from the use of seignorage, or the profits of the monetary authorities from the high demand for holding currency, were also marked at this time, especially during 1972-73 and 1976-80. The demand for money (as measured by the ratio of base money to real GNP) rose from 17.8% of GNP in 1970 to 22% between 1977 and 1983 then dropped to slightly above 201 from 1984 to the present. This willingness of people and institutions to hold relatively more money was used by the monetary authorities to help finance the deficit through the issuance of currency without creating an excess supply of money. 2.61 Except for a few years when Malaysia experienced domestic price inflation of mild proportions (the highest annual rate was 17X in 1974, and this was largely due to the commodity price booms for exports), the inflation tax was insignificant. Thus, unlike the current situation in some Latin American countries with severe deficits, the monetary sources of financing the real deficit in Malaysia came largely from the seignorage component and not from inflation. 2.62 The country's economic strength continued to minimize the impact of the deficit on domestic prices into the early 1980s. Despite a deficit of 21% of GNP in the consolidated public sector account in 1981, the consumer price index rose by only about 10X, partly as a result of the deficit and partly from adverse terms of trade. But the signs of monetary danger were evident. The trade balance had turned negative in 1981 and 1982 for the first time in years. The deficit on the balance of payments reached 20.9% and 29.8Z of total export earnings during the two years. As a proportion of GNP, the deficits in 1981 and 1982 were 10% and 13.92, respectively. The savings rate fell from 30.7% of GNP in 1980 to 252 of GNP by 1982. Faced with the real threat of having to monetize the debt, the Government introduced fiscal adjustment measures in 1982. 2.63 From 1981 to 1986, the Government increasingly resorted to financing the deficit with foreign borrowings, with a consequent rise in its debt service obligations. Debt service charges on the public debt also climbed as a result of the 19% depreciation of the Malaysian dollar to the US dol i7r since 1980 (the depreciation with respect to the SDR is even steeper).- Because of the decision to prepay some foreign debts in 1987 instead of using those funds as a source of domestic finance, foreign borrowing was negligible 3/ The Malaysian ringgit exchange rate was M$2.1769 per US dollar in 1980 and M$2.6 in 1987. This represents a 19% depreciation over this period. Of the total foreign debt, 42.1% is denominated in US dollars; 35.1% are loans from Japan and the Federal Republic of Germany; and the rest is from multilateral sources and other European countries, Australia and Canada. - 95 - in that year. The experience of Malaysia indicates that, to achieve positive results, foreign borrowing has to be accompanied by a prudent fiscal policy and stable exchange rate. Fiscal Policy Goals 2.64 The fiscal deficits of the 1980s were spawned by an imbalance in fiscal resources, by the failure to foresee the extent and prolonged nature of external price shocks, and, most of all, by ambitious growth targets supported by a high level of public investment. The social restructuring program under the NEP also required a corresponding build-up of fiscal subsidies. Now, the Government faces the need to improve the revenue base of the economy and to reduce public expenditures that still drain the country's fiscal resources. The rest of this chapter is devoted to the revenue side of this equation. On the expenditure side, public outlays should be reviewed, particularly invest- ment in directly productive enterprises, which are best left to the private sector. Fiscal subsidies, also a drain, should likewise be reconsidered. The reduction of these expenditures and the reallocation of some of the released resources could be critical to Malaysia's economic recovery. 2.65 The fiscal deficit should be brought down to a range of 6-7Z of GNP and maintained at that level in the medium term. At this level of fiscal deficit, the economy has the capacity to attain a real growth of 4.5-2,55 a year. This deficit is attainable with the debt-to-GNP ratio of 0.725_' for domestic debt and 0.41 for foreign debt which were achieved in Malaysia in 1987. It is assumed that this is the desired proportion of debt to output, if that is the strategy used in apportioning the financing of new debts. This level of growth is compatible with a fairly stable exchange rate and domestic price regimes. It allows for a 22 rate of exchange depreciation and a 2X annual inflation rate. It further assumes that the demand for base money as a ratio of output is 0.2, meaning that 20 cents of currency are in circulation for every dollar of output. This is the current range observed in the country. 2.66 A higher growth rate---for instance, 71 annual growth in GNP--would cause a build-up of deficits to 92. Given the recent experience in Malaysia and the volatility of factors that indirectly determine the growth of govern- ment revenues, such a high growth target is likely to cause a recurrence of the adverse and unstable situation of 1981. Public Revenues 2.67 Aside from the expenditure cuts referred to above, the fiscal deficit can be reduced by improving government revenues, the main revenues being those derived from taxes. While Malaysia's tax system has shown a high degree of buoyancy, some of its features would benefit from major improvements in order to increase both revenues and efficiency. 4/ Federal government debt only. - 96 - Proposed Tax System Reforms 2.68 Four features of the tax system particularly merit reform: (a) the high dependence on petroleum and trade taxes; (b) the narrow domestic tax base; (c) the highly complex system of tax incentives; and (d) the inadequate availment of user-charges by self-financing statutory agencies. Dependence of Tax Revenues on Trade and Natural Resources 2.69 The revenue structure of Malaysia is highly dependent on taxes on international trade and, since the 1980s, on petroleum taxes (see Table 2.5). The primacy of petroleum as a source of revenues is evident from the following: during 1982-87, petroleum production companies contributed about 32% of the taxes on net income and profits of companies; furthermore, taxes and levies on petroleum products contributed about 39% of domestic excise taxes, some 17% of total import duties collected and 83% of total export duties. In addition, dividends from PETRONAS (the government oil company) and petroleum royalties from all other petroleum companies operating in the country contribuce 31% of total nontax revenues collected. All petroleum taxes accounted for 30% of all tax revenue in 1982 and 25% in 1987. Were it not for the relative decline of petroleum prices, this contribution would be even higher. It is estimated that an additional US$1 increase in the per barrel price yields an additional M$100 million in revenues from royalties and taxes on imports and exports. Conversely, price declines have costly implica- tions on revenue levels. 2.70 Taxation of primary exports has declined in relative terms, partly due to the adverse commodity price movements in recent years, but largely because of the dominating role of petroleum taxation. Of total revenues from exports, Malaysia's traditional primary exports accounted for only 16% of export taxes. As a percentage of total tax revenues, export taxes have declined in importance, from 18X in the early 1970s to only 10.5% in 1982-87, including petroleum export taxes, but only 1.7X without petroleum taxes. In fact, export duties have now been surpassed by import duties and domestic taxation of goods and services. 2.71 In spite of this, international trade taxes (including those from petroleum) dominate Malaysia's tax system. Taxes on petroleum and inter- national trade (including taxes on exports, imports, and imported goods for domestic sale) together have accounted for close to one half of total taxes in the 1980s. Of the tax and nontax revenues earned from trade and petroleum, the petroleum component (comprising taxes and royalties from production companies and dividends from PETRONAS) accounted for about 70% during 1982- 87. Not included in this figure are revenues from the taxation of goods that spin off from petroleum usage (e.g., motor vehicles) and taxes from companies whose main business is related to the export of the major primary exports. 2.72 At first glance, Malaysia's tax system appears atypical for a developing country in that direct taxes figure largely in the overall tax structure. However, the majority of the direct taxes are dependent on income taxes from companies producing petroleum. International trade taxes comprise only 21% of total revenues, while domestic taxation of goods and services represents about 19% (Figure 2.14). - 97 - Table 2.5: TAX REVENUE IN MALAYSIA, i982-87 Est. Average 1982 '983 1984 1985 1986 1987 82-87 Tax Revenue 13,164 15,917 17,208 17,478 15,438 13,552 15,459.5 Taxes on net income & profits 6,0'6 7.269 7,962 8,811 8,290 6,135 7,420.5 Oil production companies 2,075 1,999 2.570 3,130 3,072 1,282 - Petronas 1,610 1,678 1,974 2,304 - 1,021 - Other oil companies 465 321 596 826 - 261 - Other companies 2,617 3,451 3,432 3,920 3,446 3,179 - Individuals 1,362 1,815 1,975 1,749 1,761 1,666 - Other 2 4 5 12 11 8 - Taxes on property 95 100 88 87 88 86 90.6 Taxes on goods and services 2,730 3,727 3,946 3,853 3,574 3,695 3,587.5 Sales tax 788 1,284 1,320 1,233 992 1,028 - Imported goods 343 528 531 470 357 383 - Domestic goods 445 756 789 763 635 645 - Selective excises on goods 1,024 1,361 1,459 1,376 1,410 1,380 - Tobacco & alcoholic bev. 247 335 355 358 363 358 - Petroleum products 305 436 487 537 717 613 - Motor vehicle tax 406 530 563 426 269 326 - Other 66 60 54 55 61 83 - Selective excises on services 918 1,082 1,167 1,244 1,172 1,287 - Of wc: motor vehicles 538 629 696 753 726 772 - Taxes on international trade 4,035 4,483 4,785 4,357 3,204 3,379 4,040.5 Import duties 2,315 2,591 2,697 2,518 2,064 2,260 - Tobacco & alcholic bev. 394 463 357 364 321 268 - Petroleum products 233 297 337 420 576 556 - Other import duties 1,233 1,447 1,889 1,634 1,090 1,351 - Surtax on imports 455 384 114 100 77 85 - Export duties 1,720 1,892 2,088 1,839 1,140 1,119 - Rubber 110 273 161 3 1 0 - Petroleum 1,351 1,477 1,629 1,639 1,076 1,025 - Tin 159 56 35 38 0 0 - Palm Oil 75 49 193 93 18 28 - Other 25 37 70 66 45 66 - Other tax revenue 248 338 407 370 276 257 316 Nontax Reyenue 3,359 2,549 3,342 3,364 3,817 4,631 3,510.3 Property income 2,690 2,033 2,789 2,733 3,130 3,813 - Nonfinancial public enterprises 207 249 405 114 97 127 - Public financial institutions 290 253 234 263 518 476 - Rent and interest 318 540 589 807 986 773 - Dividends paid by Petronas 1,450 500 980 930 1,000 3,000 - Petroleum royalties 425 491 581 619 549 437 - Administrative Fees and Charges: Sale of goods, fines & forfeits 314 367 410 432 453 570 - Contribution to Pension Fund 330 137 134 186 219 289 - Other nontax revenue 25 12 9 13 15 19 - Capital revenue 32 29 38 105 75 80 - Total revenue 16.555 18.495 20,588 20.947 19,324 18.263 19,028.6 Foreign grants 12 8 59 2 0 2 13.8 Total revenue & grants 16,567 18,503 20,647 20.949 19,324 18,265 19,042.5 Source: Malaysian authorities. - 98 - Figure 2. 14 Distribution of Revenues, 1982-1987 Distribution of AMI R.v.nu. 0 w, (1M00 Oil Tm_ (t.714 <~~~~~~~wp% (afim *wn. OT, (10.*) Relative Composition of Petroleum and Other Trade Taxes Sales Tax on Import Gds Petreb.um (47.7 (5.0%) ENw Oub (3.1%)C kqw Dai I21 - 99 - 2.73 Thus, from the perspective of sources of tax revenues, Malaysia's tax system has much in common with those of many developing countries. How- ever, Malaysia's taxes have been more buoyant than those of other countries, partly reflecting the large revenue gains from petroleum and international trade taxes. During 1970-87, Malaysia's total. revenues had a high buoyancy level of 1.3, indicating that revenues rose about 1.3% for every percent increase in the GNP. 2.74 Table 2.6 shows estimates of tax buoyancies for different types of taxes 1970-87 and 1977-87. These tax buoyancy estimates are measured with respect to GNP. Earlier estimates of buoyancies are measured with respect to GDP. These are consistently lower than the present estimates reported in the Table. The different estimates are consistent with each other because GDP has risen faster than GNP in Malaysia. The buoyancies of direct income taxes are higher than those of indirect taxes. This phenomenon is striking for total direct income taxes (which include petroleum income taxes and royalties) because they are higher than the components (which exclude petroleum) for company and individual income taxes. The 1977-87 buoyancy estimates for direct income taxes are more than twice those for the company and the individual income tax. The individual and company income taxes (excluding petroleum) have reasonably high buoyancy estimates: 1.1 for ^ompanies and between 1.4 and 1.2 for individuals. 2.75 Clearly, the tax system is still highly tied to international trade, in particular, to export trade based on natural resources. Moreover, a large part of domestic income is induced by export income. In view of the volatility of trade and commodity prices, particularly for petroleum, Malaysia would benefit from tax changes that reduce the current trade and petroleum link. The measures described in the following discussion could help to achieve this reduction. Narrow Domestic Tax Base 2.76 The deficit problem has called attention to the country's need for increased tax resources. As mentioned above, commodity taxation has been dominated by the taxation of petroleum products. Close to 40% of excise taxes derive from petroleum products, and the average sales tax collection from 1982 to 1987 was 64Z of what PETRONAS paid as a company in income taxes for the same period. Together, all sales ant excise taxes (excluding petroleum excises) brought in tax revenues equal to only about 42% of all petroleum- related taxes collected. 2.77 The taxation of goods and services has grown along two tracks: an excise tax system covering a wide range of commodities; and a sales tax system designed to cover all other commodities but which, in fact, is narrowly applied. There is much merit in combining the two taxes into a more productive single sales tax system, with a minimal but special role playei by excise taxation and a move towards a value-added tax (VAT). 2.78 The Sales Tax. The structure of the sales tax has remained basically unchanged since its adoption. It is an ad valorem single-stage tax imposed at the manufacturing and import levels. The sales tax rate of 5% was - 100 - Table 2.6: REGRESSION ESTIMATES FOR CATEGORIES OF TAXES, 1970-87 1970-1987 1977-1987 Logarithmic Regressions L Constant GNP DCNPD R-SQ8AR D.W. Constant GNP DGNPD R-SQBAR D.W. Total revenue -3.9266 1.2906 0.008 0.992 2.24 -4.974 1.338 - 0.961 1.858 (-8.668) (27.615) (1.492) - - (-5.407) (15.890) - - - Direct taxes -32.913 4.1 - 0.87 1.066 -55.34 6.275 0.189 0.454 1.88 (-8.221) (10.759) - - - (-2.046) (2.460) (-1.270) - - Incoe tax -8.0222 1.844 - 0.561 1.549 -15.539 2.585 -0.068 0.481 1.28 (-1.974) (4.769) - - - (-1.354) (2.390) (-1.080) - - Companies -4.685 1.146 - 0.958 2.027 -4.882 1.164 - 0.839 1.977 (-7.707) (19.801) - - - (-2.807) (7.313) - - - Individuals -8.228 1.409 - o.978 1.327 -6.632 1.263 - 0.904 1.686 (-15.626) (28.113) - - - (-4.710) (9.799) - - - Indirect taxes -17.584 2.641 - 0.686 0.963 -40.841 4.865 -0.143 0.47 1.77 (-3.916) (6.179) - - - (-1.966) (2.484) (-1.247) - - Export duties -17.605 2.493 - 0.389 0.836 -41.734 4.879 -0.246 -0.119 0.915 (-2.288) (3.443) - - - (-0.778) (0.964) (-0.833) - - *-best eqn not significant. Rubber -16.728 2.214 -0.131 0.39 1.058 60.144 -5.533 - 0.331 0.926 (-1.649) (2.207) (-3.461) - - (2.629) (-2.439) - - - Palm oil -13.359 1.787 -0.251 0.428 1.325 30.385 -2.362 - 0.361 1.269 (-2.588) (3.501) (-3.754) - - (3.039) (-2.581) - - - Import duties 6 surtax -2.142 0.89 - 0.965 0.57 -2.516 0.924 - 0.811 0.564 (-1.652) (6.633) - - - Excise -3.056 0.92 - 0.976 1.318 -2.074 0.831 - 0.875 1.32 (-8.402) (26.577) - - - (-1.927) (8.437) - - - Road tax -0.255 0.572 0.026 0.953 0.66 -5.56 1.031 - 0.827 0.469 (-0.378) (8.571) (2.994) - - (-3.170) (6.988) - - - Total non-tax revenue 1.475 0.776 0.12 0.963 2.304 -5.957 1.201 0.095 0.937 1.928 (-1.485) (.611) (7.216) - - (-1.123) (2.403) (3.275) - - Note: The coefficients of GNP are interpreted as "buoyancy" estimates. The coefficients of "DGNPD," when statistically significant, either add or subtract to the buoyancy estimate for the period after 1980. - 101 - doubled to 10Z in 1983, and in 1987 a 15 tax was applied to liquor and cigarettes. Over time, however, the list of exemptions to the sales tax has widened and the threshold level for firms subject to this tax has been raised to those with sales of M$100,000 or more, a decision which narrowed the coverage to a small grcup of manufacturers and importers. This means that the effective base has narrowed, when it should have been growing over time. It is estimated that at least 75X of domestic manufacturing, production and imports are exempt from the sales tax and that the average effective sales tax rate on domestic manufacturing and imports is less than 21 of the value of sales. At present, all state enterprises are exempt from both the sales tax and import duties. Besides depriving the Government of needed revenues, this situation encourages overconsumption of materials and capital-intensive operations. 2.79 At 7.11 of total tax revenues, the sales tax in Malaysia is low compared to the 20% of tax revenues received from the sales tax in Thailand, 13X in the Philippines, and 251 in Korea. 2.80 Significantly, the 1988 budget has reimposed a sales tax of 51 on many items that used to be exempted. Previously, foodstuffs, machinery, raw materials, agricultural goods and implements, building materials, pharmaceuti- cals, tourist goods, books, and magazines were exempted, but under the recent measure a wide list of food items, building materials and other raw materials are taxed at SZ. A shortet list of items is now taxed at 101. Nevertheless, a large number of sales transactions are still excluded from the sales tax due to exemptions from the sales tax of companies enjoying tax incentives and the high threrhold level for wholesaler taxation. 2.81 A second significant development is the commitment of the Govern- ment, announced in the recent budget, to undertake a study for the purpose of revising the sales tax. As a consequence, recent sales tax reforms in neighboring countries, mostly of a value-added type, are being studied for their relevance to Malaysia's sales tax reform. The possible direction of this reform is discussed below (paras. 3.93-3.100). 2.82 The Excise Tax. The distinguishing feature of the excise tax is that it is based on specific rates while the sales tax is ad valorem. The desirability of assessing a higher, special rate of tax on some commodities is one justification for having an excise tax. But while the excise tax rate structure has to be revised periodically to adjust revenues to inflation, this need not be done with the ad valorem sales tax structure. Excise duties were originally the principal instrument for domestic taxation of commodities in Malaysia. Before the country's independence in 1957, only a few items including beer, cigarettes, petroleum, and matches were subject to tax. Over time, the list of items under excise taxation grew, and despite some reduction in the list in order to accommodate the sales tax, many commodities fall under both excise and sales taxes. 2.83 The principal revenues from the excise tax are from tobacco and alcoholic beverages, petroleum products, and motor vehicles. In the recent revisions of the sales tax rates, these items were assessed at the new and higher sales tax rate of 151. Most other items under the excise tax have small revenue incomes and could easily be integrated with the sales tax. - 102 - 2.84 The Service Tax. The tax on services was enacted in 1975 as a logical companion to the sales tax which covers commodities. At present, the service tax has a limited coverage, but it provides a beginning for a broader- based tax. The tax of 5% of value of sales service applies to hotels with more than six bedrooms and to service establishments within the hotel premises (resturants, bars, coffee houses, premises for meetings and shows) and to establishments located outside hotels with an annual turnover of less than M$500,000. 2.85 The coverage of the service tax is too limited and should be extended to more significant service transactions that affect a larger segment of the domestic trade in services, for example, the sale of electricity, telecommunications, building construction, warehousing, engineering, lease of real property (other than residences), building construction and repair, gas and water. It would also be convenient to integrate the service tax with the sales tax. 2.86 Tax Buoyancies of Sales and Excise Taxes. Tax buoyancies have been estimated for taxes on commodities for which time series data are available. These buoyancies are shown in Table 2.7 for sales taxes and Table 2.8 for excise taxes. As already explained, the buoyancy estimates reflect the response of tax revenues to movements in the GNP. The excise tax buoyancies are computed for the period 1977-86 and the sales tax for 1972-86. 2.87 The buoyancies of excise taxes are consistently lower than those for the sales tax. This is caused by the failure of specific rates to catch up with price changes, and secondly, by the phase-out of tax coverage of some items under each grouping over time. Tax buoyancy estimates greater than one (meaning that they have risen more than the rate of growth of GNP) are found for tobacco, cigars, and cigarettes, petroleum products, motor vehicles, and air conditioners. Buoyancy estimates for other products are below one. 2.88 Items under sales taxation have high buoyancy estimates. Of the 19 tax groups for which buoyancy was estimated, 16 had buoyancy greater than one. These estimates range from the low of 1.12 (for garment-making) to 1.85 (for electrical goods). The buoyancies were higher after 1981 for a number of commodity sales taxes. Buoyancy estimates may also reflect the increase in the basic sales tax rate to 10X in 1983. In comparison, the buoyancies for excise taxes are largely the same throughout the whole period. These findings would support the recommendation to reduce the items covered under excise taxation and increase sales tax coverage. 2.89 Towards a VAT. One possible way of modifying the present sales tax would be to move the base of the sales tax from the manufacturer's and importer's level to the wholesale stage, which is just one step removed from the retail stage. Because of the difficulties of the current "ring method" of isolating the tax element in sales transactions, there is a high degree of tax pyramiding in later sales, hence causing economic distortions in the costs of goods. If a "credit" method of taxations which credits previous taxes paid on purchases of materials from other suppliers, replaced the ring method, the wholesale-stage sales tax would have the features of a VAT at the wholesale stage. But whether or not the wholesale-stage tax would produce a significant jump in revenues would depend on its coverage and rate. - 103 - Table 2.7: REGRESSION ESTIMATES FOR SALES TAXES, 1977-87 Industry Acronym Constant GNP DGNPO R-BARSQ D.W. Confectionary CAND -15.265 1.534 - 0.958 1.200 (16.859) (17.953) Spirit, malt and liquor LIQ -16.285 1.849 -0.044 0.906 2.118 (-7.053) (8.210) (-1.768) Aerated waters SODAS -13.090 1.409 0.367 0.979 1.826 (-11.523) (12.713) (2.977) Tobacco, cigar and cigarettes CIGRETS -10.017 1.323 - 0.92 1.7Q6 (-9.097) (12.730) Textiles TXTILES -8.332 0.955 - 0.861 2.115 (-7.737) (9.402) Footwear FTWR -7.205 0.817 - 0.097 2.075 (-1.320) (1.586) Garmentmaking, incl. tailoring GRMNTS -9.997 1.128 0.040 0.957 1.558 (-7.077) (8.189) (2.632) Furniture other than woodprods FURNTR -12.176 1.333 0.033 0.967 1.905 (-8.990) (10.069) (2.276) Paper and paper products PAPER 12.279 1.345 0.039 0.961 1.180 (-8.021) (8.988) 2.46) Rubber products other than footwear RUBR -14.860 1.603 0.032 0.977 1.859 (-11.370) (12.576) (2.297) Paints, varnishes, lacquer PAINTS -13.319 1.379 0.063 0.963 2.173 (-7.877) (8.622) (3.447) Soap, washing and cleaning compound SOAPS -6.160 0.687 0.128 0.904 1.934 (-2.319) (2.652) (4.456) Perfumes, cojsmetic, toiletries CSMETICS -17.447 1.799 0.078 0.907 2.026 (-9.080) (9.595) (-3.743) Base metal BMETL -11.010 1.279 - t.938 1.911 (-11.952) (14.712) Motor vehicles AUTOS -13.274 1.626 0.027 0.937 1.698 (-6.030) (?7571) (1.157) Machinery MACHNRY -17.189 1.748 -0.020 0.951 1.502 (-10.127) (10.559) (-1.,01) Plastic goods PLSTCS -11.489 1.271 0.020 0.948 1.722 (-7.422) (8.418) (1.213) Electrical goods ELCTRC -16.669 1.851 - 0.983 2.175 (-24.221) (28.496) Miscellaneous MISC -9.606 1.221 - 0.838 1.006 (-6.373) (8.584) Total TOTAL -10.511 1.524 - 0.968 1.466 (-13.386) (20.570) - 104 - Table 2.8: REGRESSION ESTIMATES FOR BUOYANCIES OF EXCISE TAXES LOGARITHMIC EQUATIONS, 1977-87 Items under excise taxation Acronym Constant GNP DGNPD R-BARSQ D.W. Intoxicating liquors LIQR -6.463 0.850 - 0.808 2.531 (-4.569) (6.566) Beer from malt BEER -5.837 0.986 - 0.914 1.820 (-5.616) (10.367) Tobacco. cigars, cigarettes CCRTS -12.037 1.510 - 0.780 1.025 (-4.410) (6.042) Monosodium glutamate MNGLT -3.977 0.498 - 0.515 1.19: (-2.494) (3.410) Petroleum & petroleum products PTROL -7.122 1.218 - 0.462 0.410 (-1.659) (3.100) Rubber tires TIRES -1.403 0.293 - 0.116 0.817 (-0.403) (1.524) Inner tubes for vehicles AUTBES -10.786 0.985 - 0.879 2.018 (-8.634) (8.620) Primary cells and batteries BTTRY -2.393 0.305 - 0.037 0.750 (-0.846) (1.179) Aerated water SODAS -8.203 0.926 -0.040 0.310 2.219 (-2.028) (2.429) (-1.822) Vehicles AUTOS -5.518 1.028 - 0.341 0.542 (-1.221) (2.485) Television TV 11.867 -0.888 - 0.138 0.443 (-1.973) (-1.613) Refrigerator PRIDGE -6.702 0.810 - 0.594 0.634 (-2.995) (3.954) Airconditioning machines AIRCN -10.134 1.083 - 0.395 0.626 (-2.354) (2.747) Playing cards PLCRDS -12.021 0.969 0.022 0.892 2.000 (-3.536) (3.024) (.018) *ftt below is very poor. Hatches HTCHS 3.601 -0.214 - -0.064 1 (.693) (-.452) OTHRSS 26.405 -2.133 -0.262 0.524 1.410 (-0.622) (-0.533) (-1.122) Notes: t-values are reported under the coefficients in parentheses; r-barsq is the r-sq corrected for degrees of freedom; D.W. is the Durbin-Watson statistic. - 105 - 2.90 It is more logical, however, to move more directly to a full VAT at the retail level. Experience in some countries has shown that while a VAT at the wholesale stage could be an intermediate step, it is possible to leapfrog the process. The administrative steps needed to bring the system to a full credit method w,ith a retail-stage VAT is likely to be challenging, but should not be so daunting as to preclude making the required changes. 2.91 A VAT at the retail stage is superior to the present sales tax and even to a sales tax at the wholesale stage. The VAT is a broader-based tax: it can cover activit£es from early stages of production to the final sale. Depending on the extent to which exemptions and the number of tax rates are minimized, it can be neutral with respect to the use of productive factors by removing tax pyramiding through the application of tax crediting. It therefore has the advantage of being less distorting than other forms of sales taxation. For Malaysia, the VAT could serve as a vehicle for integrating the present sales, service and excise taxes, and derive economies in single tax administration. 2.92 The VAT has many variants. The most common form of VAT found in many countries has the following characteristics: it is a tax on consumption (in contrast to a tax on income); it is imposed on the principle of destina- tion (not the origin of supply); it uses the tax credit method of computation (rather than other methods2i and it has two or more rates in addition to a zero rate for necessities.- Each country selects the most suitable VAT depending on its objectives. 2.93 The simplest VAT rate structure is a single uniform rate. In view of the additional administrative costs of running a credit system and (initially) of upgrading the administrative machinery to oversee the tax, the minimum tax rate worthwhile is at least 7%. Given that the present sales tax now has a three-rate structure of 5%, 10%, and 15%, this minimum rate has already been reached in Malaysia. It is therefore recommended that the number of exemptions from the tax should be limited and the basic rate, except for a few items, should be set at 10%. As experience improves in running the tax, the cost of administering the VAT would fall per ringgit of revenue. In the beginning, there may be difficulties in bringing in a wide network of business enterprises. The treatment of small farmers and traders can be handled by the use of a threshold level, usually related to business turnover per year. The current threshold of M$100,000 per year seems too high and should be brought down so that the impact of the tax on revenues can be raised. 2.94 The possibility of having more than one tax rate is often predicated on the equity objective: to reduce taxes on the consumption of the poor and to tax items consumed by the rich at a higher rate. This complicates the administrative costs oz the tax and inevitably raises the question of identifying the most b sic items that qualify for lower rates and those for 5/ See M. Gillis, C. Shoup, and G. Sicat, "Lessons from Value Added Taxation in Developing Countries," World Bank, IRD Discussion Paper No. 238, February 1987. - 106 - higher rates. A possible compromise in handling this issue is to limit items pertaining to the lowest-income groups to the most basic--for instance, only the most essential food items. (The definition of "basic necessities" is always a trap that causes fairly good taxes to suffer revenue erosion.) To provide a zero-rate for these items rather than to exempt them has often been proposed, but the administrative expedient in some countries is outright exemption. Although a tess preferred alternative, exemption is normally adopted when a VAT is initially introduced and the authorities have no experience in implementing the tax. As the administrative machinery becomes more capable, the zero-rating option, which is the more appropriate, should be adopted. 2.95 If the sales tax is to serve primarily as a revenue measure, there should be no more than three rates. As in the case of tariffs, many different rates of taxation not only create complications in tax administration but also lead to more economic distortions. The VAT is weli suited as a revenue measure and efforts to make it accomplish other tasks would only overload it. 2.96 The issue of integrating the excise tax with the VAT is related to the number of rates to be applied. If the excise tax is designed as an additional tax on luxuries, it could become the vehicle for much higher sumptuary taxes. This could apply to liquor, tobacco, petroleum products, and motor vehicles. The rate structure could also be a specific levy, that is, a per unit tax. However, most of the remaining items for taxation should be integrated into the VAT, as the latter is more suited as an instrument of taxation and its ad valorem structure automatically responds to price changes. Taxation of Companies and Fiscal Incentives 2.97 Marginal Effective Tax Rate for Companies. In Malaysia, the statutory (or nominal) tax rate for companies is 40% of taxable income, which is relatively high compared to rates in other countries in the region. For competitive reasons, it is understandable that the company tax in Malaysia is similar in structure to that of Singapore. But in addition to the company tax, Malaysia has additional taxes that further bite into corporate income. These are the 5% development tax and the 3Z excess profits tax on taxable income greater than M$200,000 (abolition of the excess profits tax was announced in the 1988 budget). 2.9ts Yowever, a study of Malaysia's marginal effective tax rates on capital investment iv;licates major deviations from the statuto'y tax rate permitted under the _urrent company tax law. Taking a standard project composed of investments in buildings, machinery, equipment, and vehicles, an all-equity- financed project has an effective tax rate of 32%, or 8% less than the nominal tax of 40%. Given a more realistic situation in which project financing involves 50Z debt financing, the marginal effective tax on the investment is cut by almost half, at 20.5%. The favorable treatment of debt financing, i.e., the deduction of interest from income, helps to reduce the effective tax rate. Table 2.9 shows a comparison of effective tax rates on standard investment projects in various East Asian countries. - 107 - Table 2.9: MARGINAL EFFECTIVE TAX RATES FOR A STANDARD INVESTMENT PROJECT IN EAST ASIAN COUNTRIES Type of Capital Asset Country Build- Mac .& Vent- statutory ings equip. cles Project rate Case of All Equity Financing and Full Loss Offset Hong Kong 19.8% 13.5% 10.9% 17.3% 18.5% Indonesia 45.1% 38.7% 34.5Z 41.6% 35.0% Japan 42.0X 37.3% 32.6% 39.2% 33.3% Korea 30.2% 34.8% 31.1Z 33.1% 30.0X Malaysia 37.6% 24.3Z 15.1% 32.0% 40.0% Philippines 41.8% 38.1% 37.6% 40.4% 35.0% Singapore 26.1% 23.5! 32.6Z 28.4% 40.0% Taiwan 31.8% 32.6% 28.1% 31.9% 25.0% Thailand 28.7% 18.9% 28.1% 24.9% 35.0% Case of 50Z Debt Financing with Full Loss Offset Hong Kong 14.2Z 2.8% -1.8% 9.6% 18.5% Indonesia 40.2Z 28.8% 21.0% 34.1% 35.0Z Japan 34.1% 26.0% 17.8Z 29.4% 33.3% Korea 19.5% 27.7% 21.0% 24.6% 30.0% Malaysia 30.2% 10.8 -9.7% 20.5% 40.0% Philippines 34.4% 27.8% 26.8% 31.9% 35.0% Singapore 15.0% 2.7% 25.5% 15.2% 40.0% Taiwan 28.0% 29.6% 21.6% 28.2% 25.0% Thailand 25.1% 7.3Z 23.4% 18.6% 35.0% "Project" is a weighted average of an investment in buildings (40%), machinery and equipment (40%); and vehicles (20%). Source: A. J. Pellechio, G. P. Sicat, D. G. Dunn, "Taxation of Investment in East Asian Countries," DRD Discussion Paper No. 261, World Bank, March 1987. 2.99 The effective tax rate is also different for different types of investment, with investment in vehicles enjoying the most favorable treatment. With 50% debt financing, investment in vehicles receives an indirect tax subsidy of -9.7%. 2.100 These results do not take into account fiscal incentives offered by the Government. They simply compare the effective marginal tax rate on a standard project, given the various practices allowed by tax law in connection with the following: depreciation methods for capital investments, the carry-over of losses; and the treatment of capital gains. In Malaysia, the carry forward of losses is possible without time limit. This effectively allows the offset of losses. Dividends are not taxed but are imputed to the stockholder through the individual income tax; and a generous capital gains law is in effect, unlike the practice of some countries where capital gains are treated as ordinary income. - 108 - 2.101 Tax Incentives for Investment and Industry. The current incentives for industrial development take the form of tas holidays, exemptions, favorable tax treatment of capital acquisition, and tax allowances. As revised in 1986, the incentive system is as follows: (a) Pioneer Status. This is given to companies in the manufacturing, agricultural, hotel, and tourist industries upon approval of their application for it. The pioneer status provides relief from income and development taxes for a period of five years, regardless of company size. The tax holiday may be extended for another five years. (b) Investment Tax Allowance (ITA). This is an alternative to the pioneer status and is awarded, as in the case of pioneer enterprises, to promoted activities and products. The ITA is up to a maximum of 100l of qualifying capital expenditure incurred within five years of the ITA approval. (c) Abatement of Adjusted Income. This is a post-tax holiday incentive after expiry of pioneer or ITA benefits. The company is allowed to deduct the abatement against adjusted income, thereby reducing the chargeable income and tax payable by the company. The different types of abatements are as follows: (i) Based on export performance. This is up to a maximum of 50,Z depending on the company's export performance. (ii) Based on location. This is equal to 51 of the adjusted income if the company is located in a designated industrial area. (iii) Based on small-scale size. This is for small-scale enterprises and is equal to 5X of adjusted income. (iv) Based on compliance with the government policy on capital participation or employment in industry. This is for 5% of adjusted income. (d) Export Allowance. The export allowance is granted to trading companies which export products manufactured in Malaysia (including certain agricultural products). It is equal to 5% of the f.o.b. value of export sales. (e) Reinvestment Allowance. A 40Z allowance (raised from 25% in the 1988 budget) is provided to manufacturing and processing industries for qualifying capital expenditures for the purpose of expansion. This allowance, which was about to expire, was extended to 1990. (f) Accelerated depreciation allowance (ADA). The ADA allows a qualifying plant to fully depreciate approved investments made prior to 1988 within a period of two years. - 109 - (g) Double deduction of special expenses. Expenses incurred in the promotion of exports and those for approved training and research can be deducted at twice the cost. 2.102 Some Estimates of Tax Subsidies to Industr . The granting of tax incentives represents a discretionary power of government to provide relief from the application of some tax provisions. This is considered a routine part of industrial promotion, although the fiscal loss due to the sought incentives is rarely estimated due to sketchy data on the value of the tax incentives awarded. 2.103 The Inland Revenue Department of the Treasury and the Malaysian Industrial Policy Studies (MIPS) conducted a survey in 1984 of 800 corporate income tax returns covering the period 1958-82. The survey included companies that enjoyed incentives such as pioneer status and investment tax allowance and those that did not, but which could still benefit from provisions under the company tax law. While caution is essential in drawing conclusions from these data in view of their incomplete coverage, some general trends may be seen. 2.104 Table 2.10 summarizes the information: recipients of tax holiday, of ITA, and those not enjoying either incentive are classed according to the value of the incentive (i.e., the indirect subsidy) that they received. These incentives include the export performance incentive, export promotion, tax credit, reinvestment allowance and accelerated depreciation. Taken together, the tax subsidies do not appear large, but this can be misleading since they are only for the surveyed companies and fail to convey the tax susbsidies for the whole population of companies enjoying them. The data also refer only to tax breaks from the company tax due to the investment project. They do not reflect the tax incentives provided throughout the production process, such as those on raw materials inputs. 2.105 Further data on the cost of income incentives are available for 1987 on incentives introduced in 1986. The value of the 5% incentive for the use of indigenous materials in the production of export commodities in six categories of products exported by 20 companies amounted to M$14 million. The bulk of these incentives were received by palm oil products (M$9.8 million) and petroleum products (M$2.8 million). 2.106 Another 52 companies benefitting from the export allowance in 1986 received M$5.9 million. The bulk of this was received for food/agricultural products (24 companies receiving M$4.0 million) and palm oil (six companies receiving M$0.88 million). 2.107 Another aspect of the revenue losses from tax incentives is exemptions from duty on machinery and equipment imports. Table 2.11 shows the revenues foregone from exemptions from the import duty, the surtax, and the domestic sales tax on machiuery and equipment from 1983 to the first semester of 1986. The total revenues forgone for two and a half years amount to M$193.0 million. Of this amount, M$80.8 million represented an erosion of the sales tax. - 110 - Table 2.10: ESTIMATES OF SUBSIDY OF MAJOR TAX INCENTIVES OVER VARIOUS RECIPIENTS OF TAX HOLIDAYS, ITA, OR NON-RECIPIENTS (in M$ million) Company Enjoying Tax Investment No Summed Up Holiday Tax Credit Incentive Overall Period Relevant 1958-82 1970-82 1978-82 1958-82 No. of Firms 314 101 333 748 Specific Fiscal Incentive Export Performance 24.46 40.28 0.96 65.70 Export Promotion Expense 2.12 5.04 0.01 7.17 Pioneer Status 893.55 - - 893.55 Investment Tax Credit 0.90 138.48 0.63 140.01 Reinvestment Allowance 18.59 3.46 0.81 22.85 Accelerated Depreciation 34.33 3.29 1.45 39.07 Total Value 973.95 190.55 3.86 1,168.35 Source: Inland Revenue Department/MIPS Survey of Corporate Income Tax Returns, 1984, cited by Barjoyai Bardai and Tan Hui Gek, "Evaluation of Malaysian Corporate Investment Incentives," paper submitted to the National Conference on Tax Reform, December 1987, Malaysian Institute of Economic Research. 2.108 These figures reflect only the tax exemptions applicable to imports of machinery and equipment. Figures on exemptions from taxes on the import of raw materials for use in domestic production by privileged enterprises are not available. Also unavailable are data on exemptions from the sales tax of companies producing for domestic production. In any neutral sales tax regime which does not penalize exports, raw material imports of exporting enterprises would be tax-exempt, but those for domestic production would not be. As indicated earlier, 75% of the manufacturing industry is effectively not covered by sales taxes largely due to the prevalence of exemptions. 2.109 These numbers are also limited by reflecting only those companies receiving industrial incentives under the Ministry of Trade and Industry and the Treasury. Companies located in the Free Trade Zones and also those operating as licensed manufacturing warehouses, which receive tax incentives, are excluded. 2.110 Furthermore, while the magnitude of the incentives may seem small when compared to total tax collections for the period, it should be borne in mind that when all incentives relating to the income tax, the sales tax and import duties are considered, the overall cost of the incentives is high, especially in the context of the Government's serious fiscal deficits and the need for revenues to reduce the fiscal imbalance. Table 2.11: REVENUE FOREGONE THROUGH DUTY EXEMPTION ON MACHINERY AND EQUIPMENT: 1983-86 (M$ 1,000) 1983 1984 1985 Jan-Jun, 1986 Import Sales Import Sales Import Sales Import Sales Type of industry duty Surtax tax Total duty Surtax ta- Total duty Surtax tax Total duty Surtax tax Total 1. Textiles and garments I 1 41 43 48 20 47 115 162 3 81 246 252 34 162 447 2. Wood-based industries 0 4 8 12 647 647 622 1,916 6,045 643 1,509 8,197 2,067 4 4,988 7.059 3. Food products from edible oils 40 23 69 132 12 - 12 24 195 - 16 211 394 - 18 412 4. Metal products 21,891 10,975 18,478 51,344 1,774 54 461 2,289 3,041 176 17,531 20,748 12,389 135 5,779 18,303 5. Food and beveragea 198 111 748 12057 39 39 123 201 227 26 1,016 1,269 2,150 16 620 2,786 6. Chemicals 895 147 402 1,444 - - - - 95 - 6 101 8,408 138 4,328 12,874 7. Transportation products 7 54 107 168 6,530 5,543 6,919 18,992 203 8 344 555 483 - 109 592 8. Rubber-based products 127 47 104 278 - - - - 71 1 80 152 502 25 890 1,417 9. Engineering products 61 86 235 382 33 92 714 839 226 35 501 762 22 3 31 56 10. Electrical/electronics products 3 23 48 74 215 - - 215 - - - - 96 1 156 253 11. Plastics products 223 224 288 735 7 1 9 17 544 59 373 976 - - - - 12. Printing and paper products - - - - - - - - 206 - 240 446 117 - 54 171 13. Construction/bldg. materials and cement - - - - - - - - 637 11 19 667 254 2 75 331 14. Others 506 345 472 1,323 17,382 1,459 11,325 30,166 - - - - 188 5 18 211 Total 23,952 12.040 2100 56,992 26,687 7.855 20,232 54,774 12,882 1,091 22,366 36,339 27,321 363 17,228 44,912 Note: The figures above do not include companies in tlhe Free Trade Zones and licensed minufacturing wareh uses. "Others" category may include or not items 12 and 13 before 1985. No data available before or after first semester-1986. It seems data collecting functions are not well-integrated in work. Claim is that some computerization is being done at the moment. Source: Behagian Cukai, Perbendaharaan Malaysia Treasury. - 112 - 2.111 It is sometimes argued that tax incentives are essential to attract foreign investment and to keep the country competitive with other countries that also offer generous tax incentives. However, it is widely known that the critical factors attracting foreign investment are: the policy climate for equity investment, the stability of economic policy and economic growth prospects, and relative labor costs. Moreover, some fiscal incentives may even be irrelevant to some foreign investors. Given that American and Australian firms, for instance, are taxed on their investment income from their overseas investment, the fiscal incentives lose power. To some extent, they only provide taxes for the capital-exporting country's treasury. Hence, the incentives represent an indirect subsidy by the developing country to the treasury of the developed country. This is quite an unintended result. 2.112 It is fortunate that Malaysia, unlike many developing countries, has avoided the pitfalls of overvalued exchange rates and an excessively high tariff regime for imports, although recently tariff protection has been rising. In general, the economic penalties from the high costs of fiscal incentives are not compounded by additional subsidies due to the trade and payments regime. The results are manifested by the fairly robust performance of industry in Malaysia in recent years, especially in the export manufacturing sector. 2.113 Reform of Tax Incentives and Income Taxation. Even without tax incentives, the tax rules for companies in Malaysia reduce the marginal effective income tax that companies have to pay. Since some recent developments in neighboring countries may cause Malaysia to lower its own tax rates, the country will have to decide whether it can continue to afford the costly investment tax incentives in a regime of lower tax revenues. Singapore has announced its intention to reduce its 40% statutory tax rate for companies to 30% and eventually to 25Z when the Government's fiscal position improves. Regagling the personal income tax, the top marginal tax rate would move to 30X._ This is likely to change the nominal tax rate level throughout the region, given the competition there. The Malaysian Government, in fact, has already announced that a study will be carried out of the downward adjustment of company income tax rates. 2.114 If Malaysia moves to lower tax regimes for its income taxes on companies and individuals, it should move towards a system of taxes requiring less administrative discretion and more uniformity in application. Any reduction of the nominal tax rates from their present levels inevitably necessitates removal of the costly fiscal incentives. Two consequent issues then arise: first, will revenues be lost due to withdrawal of incentives and the reduction of tax rates, and second, how should companies now enjoying fiscal incentives be treated under the new regime. 6/ See Report of the Economic Committee, The Singapore Economy: New Directions, Ministry of Trade and Industry, Republic of Singapore, 1986, pp. 89-96. - 113 - 2.115 Income tax reform could cause a temporary loss of revenues if other parts of the tax system are not changed. One way to avoid this is to link any substantial reform of the sales tax to that of the company tax. The second alternative is to undertake a phased program of tax rate reduction, say, over a three to five year period. The first alternative would probably be preferable, since it has the ingredients of overall trade-offs in designing a better tax system. The revenue losses from the company tax could be recouped by an improved overall tax regime which would contribute both to improved economic efficiency and to longer-term improvements in the tax structure. From an administrative viewpoint, the new tax regime would cause the redeploy- ment of admnistrative resources to improving tax collection and streamlining the tax machinery and away from reviewing, monitoring, and policing tax incentives. 2.116 Firms currently enjoying tax incentives are likely to resist removal of the incentives as it is not "sweetened" by gains in tax cuts. From the viewpoint of the investor, the real issue is the level of the effective tax rate facing him: is it to be a low effective tax rate achieved through tax incentives from a regime of high statutory tax rates or is it to be from a low effective tax rate from a simpler company tax without direct fiscal incentives? If the achieved effective tax is the same, the preferred route is likely to be the uniform lower tax rate regime because the company's other activities would also benefit from the new tax regime. Like the Government, the companies would gain from administering a simpler company tax, rather than a complicated system of tax incentives. 2.117 A more critical need is to assure companies now enjoying incentives that they will not lose what has been granted to them already. This can be accomplished by giving them a once-for-all choice either to retain their existing tax privileges under the older, unreformed tax rate structure until those incentives expire, or to shift to the reformed system of lower tax rates without the incentives. The choice is theirs. But future investors would face the more simple company tax structure. User Charges and Self-Financing for Statutory Authorities and NFPEs 2.118 The statutory authorities and NFPEs largely represent a drain on the Government's fiscal position. The Government should therefore identify ways of both reducing the budgetary burden of these agencies, most of which are in deficit, and of improving the regularity with which the government derives revenues from the handful of surplus companies. 2.119 One way of reducing their budgetary impact is to privatize them. This is a possibility for some NFPEs but is not necessarily the best solution for those public institutions whose functions are best performed from within the public sector. - 114 - 2.120 An alternative would be to develop ways by which the agencies could finance part of their operations through more effective pricing policies, such as user charges. The government railway company, for example, might increase its tariffs, in addition to improving operational efficiency. Other possible areas where user charges and public enterprise pricing policy may prove helpful are discussed below. 2.121 (a) Education. Great value is placed on education in Malaysia and the demand for quality education there is high. This indicates the existence of a market for quality higher education, even at market-determined prices. Until now, university education has been inexpensive in Malaysia due to the large and varied government subsidies granted. While the current policy of heavily subsidizing education is commendable, the need to find an appropriate balance between cost recovery and subsidies should be confronted by the authorities involved. Two possible means of affecting this balance might be (a) to increase university fees, which have not been adjusted in 20 years and (b) to expand the practice of providing students with loans, repayable after graduation, rather than outright scholarship grants. By using a revolving fund for these loans, it would be possible to reach a greater number of students with the funds available. With these and other measures, the Government may be able to shift some of the costs of investment in human capital to the beneficiaries. 2.122 (b) Health. The Government plays a dominant role in the provision of health services for Malaysians. At present, the health insurance and health security funds are inadequate to finance the existing structure of demand, given the low charges for medical care. Some rationalization of treatment costs in public hospitals and other clinics could reduce this dependence. The reduction of government expenditure for health delivery should be addressed based on the findings of an ongoing Asian Development Bank study of the level of health services that should be delivered by the Government and the pricing of those services. 2.123 (c) Housing and Urban Development. The main subsidy element in housing and urban development is low-interest rates on loans. Government housing loans, for example, are offered at a rate of 6% p.a. and government loans to public agencies engaged in housing and land development, especially for projects undertaken at government behest, range from 4% to 10%. In addition, the many urban development schemes undertaken by statutory authorities are funded by low-interest government loans. In the latter case, the subsidized interest rates partly account for the oversupply of high-rise buildings constructed under those schemes. The interest rate policy related to projects undertaken by the statutory authorities is therefore a critical issue and merits further study. 2.124 (d) Agriculture. Many activities in the agricultural sector are partly financed from benefit taxes. Research institutes for rubber and palm oil are financed in part from the proceeds of export taxes, a research tax and replanting fees. Projects in waterworks and irrigation are partially recovered from the land tax, and some projects in smallholder development are subsidy-intensive with a large component of operational expenditures supported by federal grants. Cost-effective project planning and better cost recovery - 115 - is thus important to reduce the fiscal burden of these projects on the federal government. 2.125 Explicit subsidies are also reflected in various operating subsidies, the largest being the paddy support price which has made Malaysia a high-cost rice producer. Other major subsidies directed at the plantation subsector are for rubber and pineapple replanting. While the total value of these subsidies has fallen since the early 1980s, they still represent a significant budgetary problem. At the root of these subsidies is the pricing policy for goods and services sold by the government sector. 2.126 Statutory Authoritites and NFPEs with Surpluses. The few statutory authorities and NFPEs with surpluses have to some extent assisted the Government in dealing with its deficits. To date, however, PETRONAS is the only NFPE to declare dividends, although this occurred only when the Govern- ment called for dividends to be paid. PETRONAS has also provided the Governments with sizeable taxes through the profits tax and the petroleum tax. Other agencies with a capacity to contribute operational surpluses include the Electricity Board, the port authorities, and other corporations engaged in transportation (shipping and airlines). (Some of these NFPEs are being privatized, in which case, the issue under discussion becomes moot.) These agencies could contribute more to the improvement of government finances in at least two ways: (a) by more effectively screening their commitments for new projects in order to weed out uneconomic projects and reduce overly large projects; and (b) by ensuring that more regular dividends are declared by profitable enterprises. 2.127 (a) Monitoring of Large Projects and Subsidiary Activities. The Treasury, which is currently monitoring NFPE activities, should encourage these enterprises to invest prudently. When they are financially liquid, the statutory agencies and NFPEs are likely to look for projects that expand their operations. Eventually, these expansions (especially if they are long-term or prove uneconomic) affect the agencies' cash flows adversely, and in some cases they become dependent on the federal budget for operational support. Even when the NFPEs are not dependent on the federal budget, the large-scale activities they undertake can effectively reduce their potential contribution towards the Government's finances. 2.128 (b) Dividend Policy. Some mechanism should be introduced to ensure regular declaration of dividends by public enterprises. This would be helpful in assuring their financial responsibility to their stockholder, the Govern- ment. In turn, government revenues from dividends may be plowed back to the NFPEs in the form of financing for worthwhile projects. The NFPEs tend to retain earnings for re-investment in operations due to the lack of an active stockholder presence pressuring them to declare dividend income. The Govern- ment (as the owner of capital is this case) can and should exercise that pressure if only to help to regularize its dividend income as a source of nontax revenues. More important, however, is the possible effect of this pressure in improving state enterprise performance and productivity. Chapter II - 116 - Appendix 1 Page 1 of 8 TECHNICAL NOTES Calculation of Net Fiscal Impulse 1. The calculation of the net fiscal impulse in terms of changes in the fiscal variables is as follows: NFI = - (DTAX - to*DGNP) + (DGEXP - go*DPGNP) where NFI is the net fiscal impulse DTAX is the change in the tax revenue DGEXP is the change in government spending DPGNP is the change in potential GNP in current prices to is the tax-to-GNP ratio in the base year and go is the government spending ratio to CNP in the base year. 2. The above formula breaks down the net fiscal impulse into the tax impulse, which if often contractionary, and the expenditure impulse, which is normally expansionary. If actual taxes grow faster than the expected tax changes as a result of increases in the GNP in current prices, the tax impulse is contractionary. Otherwise, it is expansionary. If the change in actual government spending is faster than expected government expenditure out of additional potential GNP, then the expenditure impulse is expansionary. The expenditure impulse is contractionary when actual government spending is slower than that expected out of net potential GNP. 3. Rearranging the same formula yields the following: NFI = - (DTAX - DGEXP) + (t0*DGNP - go*DPGNP). This formula shows a different breakdown of the net fiscal impulse. The second element of the formula indicates a cyclically neutral deficit, which measures the difference between the proportional tax revenues taken out of the change in actual GNP and the proportional government expenditures expected out of the change in potential CNP. The first element, which is associated with the discretionary component of the deficit, is the difference in the changes of taxes and expenditure in the current period. 4. This measurement requires some assumptions about the potential GNP. There is, of course, some difficulty in defining the potential GNP. In one definition, it is full employment GNP. But given typical conditions in developing countries like Malaysia, there is usually structural unemployment in the labor force and many bottlenecks. As a result, some subjective Chapter II - 117 - Appendix 1 Page 2 of 8 elements come into play in the definition of potential GNP. An operational definition tracking the potential long-term performance of the economy is therefore used. In analyzing a more current period, the potential GNP could be likened to a targeted GNP, if the latter is feasible and within the bounds of historical performance. 5. In that spirit, two basic assumptions about the potential GNP are used: one in which the potential GNP grows at 10% per annum and the second at 8%. Two variations are used for the 10% growth scenario. Assumptions for the potential GNP are therefore as follows: (a) Potential GNP grows at 10% per year: if the actual historical growth of the economy exceeds 10X in the current year, the potential GNP is the same as the actual rate of growth. If the economy grows at less than 10%, the potential GNP is one that grows at 10%. (b) This is the same as (a), but if the economy experiences negative growth in any given period (a recent phenomenon in Malaysia), the highest level of GYP in the last three years is chosen as the potential GNP in that current period. The rationale for this assumption is that an economy which declines because of cyclical factors will retain its productive capacity for the period of highest GNP. Ideally, an adjustment for depreciation would be needed, but this is just a refinement which is not followed any more, especially if, in later years, positive growth is experienced again. (c) This is the same as (a), but the rate of growth is 8% a year rather than 10X. 6. Given that the assumptions about potential GNP are in the vicinity of actual GNP levels, there is never a case for a significant deviation between actual GNP and potential GNP. Therefore, the assumed potential GNP is very realistic and reflects the historical performance of the economy, except for cyclical factors. 7. Calculatio;ns of the net fiscal impulse utilize two alternative sets of values for the tax and government expenditure ratios: the average tax and expenditure ratios for 1971 to 1980 and for 1976 to 1986. Figure 2.Al shows the time path of three potential GNP assumptions. For further information see, International Monetary Fund, World Economic Outlook, April 1985, Supplementary Note 1, pp. 99-110. Derivation of Implicit Subsidies 8. The implicit subsidies are calculated by combining government budget accounting and national income accounting. Under government budget ascounting, government spending is divided into operational outlays and development outlays. Operational outlays constitute largely public consumption (payment for consumed services such as wages for government employees, contracted current services, supplies and materials) and transfers (such as transfers to statutory authorities, to state governments, or payments - 118 - Chapter II Appendix 1 Page 3 of 8 of interest for the public debt and pensions and gratuities). Development outlays are dominated by public investment expenditure. 9. Under national income accounting, the expenditure and product accounts are estimated net of interest for the public debt, pensions and gratuities for worker., and subsidies By cnvention, these are transfer payments and do not represent payment for services during the current period. Of course, from the income side of the recipient sectors, these payments are part of the computation of disposable income. The outlays of the recipient sectors are affected and, depending on the magnitude of the transfers, their own consumption or saving (and investment) accounts will change. For instance, while subsidies are netted out of the government account, private sector consumption or investment may rise by virtue of new transfers. In most cases, these are likely to affect the consumption side of the private sector account. 10. The national income accounts include estimates of public consumption and public investment. The sum of these two expenditure items of government represents the total demand of government for current output. The total actual government expenditure is known from the federal budget accounts. By deducting the aggregate in current prices of public consumption and public investment from total (consolidated public) expenditures (operating and development), an estimate of the implicit subsidies is derived. 11. The change in inventory in the public sector is not added to the estimate of public investment. If done, this would raise the total demand for government spending for current services. By the nature of its functions, the government does not carry inventories of goods as large as the private sector. By attributing all changes in inventory to the private sector, total investment by the public aector is only for fixed capital formation. Thus, the total expenditure of government for productive goods and services is represented by the combined public consumption and public fixed investment accounts. This assumption does not lead to an overestimation of public subsidies. As explained below, the subsidies are much underestimated. 12. The explicit subsidies shown in the text (see Figure 2.9) &re considered apart of these estimates of implicit subsidies. They represent a very minor percentage of the total subsidies estimated; in fact, they are not visible to the naked eye in the scale of the Figure; they are shown on the horizontal axis with government expenditures and public subsidies and transfers. The total payment for service of the public debt is taken as a ratio to the estimate of implicit subsidies. The ratio has exceeded unit value in some years, hence indicating (notably since 1983) that the value of the debt service exceeded the estimated subsidies. This should not happen since debt service is only one component of subsidies in the national income accounting sense. When the estimates of pensions and gratuities are added to debt service, the discrepancy is even larger; pensions and gratuities are about 1OZ of the value of the estimated subsidies. 13. It is also believed that there is overemployment in some sectors of the Government. From an efficiency viewpoint, this implies some redundancy in public employment. To some extent, such redundancies reflect trancfers and subsidies in the guise of productive employment. Chapter II - 119 - Appendix 1 Page 4 of 8 Financing the Real Deficit and Macroeconomically Sustainable Deficits 14. The estimates of the sources of financing of the real deficits begin with the deficit finance identity, as follows: Deficit = Domestic Net Borrowing + Foreign Net Borrowing + Changes in Money Balances or, in symbols, and using appropriate price deflators: DIP = B /P + 'B*e/P + M/P d f f where Bd refers to the nominal level of the domestic debt, Bf to the foreign debt denominated in foreign currency, e the currency exchange rate, M the base money in circulation, the dot "." above the letter corresponds to discrete yearly change, P refers to the domestic price level and Pf to the foreign price level. The star symbol "*" is a multiplier operator used throughout these technical notes. 15. All the items in the identity are in flow terms. In fact, the fiscal deficit D/P on the left-hand side can also be written as B for simplicity, where B refers to the total stock of public debt. lo. For Malaysia's purposes, the domestic price deflator P is the CPI index (=CPI*.Ol, in ratio form). Malaysian debt data are reported, after 1973, at the exchange rate prevailing at the time of receipt of loan proceeds and the data prior to that year have been revalued at the exchange rate prevailing at the end of each year. This means that data are already in the form Bf*e. The foreign price level should reflect the price level of the country's major trading partners. The foreign price level is set equal to 1 so that the yearly foreign borrowing data expressed in Malaysian ringgit which is reported by Bank Negara are used without any further adjustment. 17. Tlh changes in monetary balances can be further simplified into two components - (1) the change in real money balances, m (where m- M/P), and (2) the rate of the domestic inflation, P*, times the real money balances, m, or P**m. The first item is often referred to as the seignorage revenue and the second as the inflation tax revenue. 18. Therefore, the identity is now 2/ 1/ This follows from the following: m m = d(M/P) = dM/P - (dP/P)(M/P) = dM/P - (dP/P)*m. From the main identity therefore, dM/P + P**m, where P*(=dP/P) is the annual inflation rate. 2/ In terms of the acronyms used in the figures, the identity can be written as follows (the suffix "KP" stands for constant prices, for instance, GNPKP is real GNP): DEFKP/GNPKP = DBORKP/GNPKP +FBOR/GNPKP + dm/GNPKP + P**m/GNPKP where small d means "change from last year". - 120 - Chapter II Appendix 1 Page 5 of 8 d bd/y + be m ; + P^*m where each corresponding small letter now stands for the real or price deflated values of the variables (e.g., d-D/P, m=M/P, etc.). For ease of use, the above identity is divided through by the real GNP, y, or d/y = bd/y + bfe/y + m/y + P^*m/y (1) where the corresponding small letters refer to real values (e.g., d=D/P). The items on the left-hand side of (1) are computed frce yearly data on sources of financing, money supply, and prices. 19. In brief, the formulation (1) gives the rrtio of the real deficit as being financed from the following ratios to real GNP: (a) real domestic borrowing; (b) foreign borrowing; (c) seignorage; and (d) the inflation tax. 20. Discrepancy in Items on the Two Sides of the Equation. A discrepancy will occur between the real deficit and the sum of the sources of financing since there are errors in valuation on either side of the identity. In usual accounting, discrepancies such as these are lumped under the one or the other of these names: "errors and omissions", "residuals" or "balancing items". 21. Figure 2.A2 shows the sum of the financing sources compared with the estimate of the real deficit as ratios to real GNP. The left-hand side (real deficit) of the relation (1) is deflated simply by CPI. But this is not adequate. A breakdown would be more appropriate. From a composition standpoint, that would be equivalent to disaggregating the three components: (a) the primary deficit (i.e., net of interest service charges); (b) the real cost of domestic debt service (which is total domestic debt times the real interest rate); and (c) the real cost of debt service of the foreign debt (which is total foreign debt denominated in the currency of borrowing multiplied by the real interest rate and adjusted further for the exchange rate depreciation of the Melaysian ringgit). These details are not measured, for lack of data on the second a.d third items. Deflating the deficit by the CPI to obtain the real deficit is therefore used only as an expedient method. 22. The discrepancies are large for the early 1970s. Figure 2.14 in the text simply takes the component as the sum of the sources of financing. They therefore reflect orders of relative magnitudes. If understood in this way, the absolute total is not as significant as the relative share of each source of financing, which is the magnitude being sought in this exercise. 23. Financially Sustainable Fiscal Deficit Level. The financing identity is in effect a budget equation for the government in financing the fiscal deficit. One definition of a sustainable fiscal deficit depends on the accepted limits of inflation and other real phenomena, such as the trade imbalance, that policy makers are willing to accept. Under this definition, a level of fiscal deficits is not sustainable, even if the desired domestic and foreign borrowings are attained, if the tolerance for domestic inflation is exceeded and signs of large trade deficits are threatening the stability of the exchange rate. In Malaysia, given the record of limited inflation, it Chapter II - 121 - Appendix 1 Page 6 of 8 would seem that the definition of sustainability of the fiscal deficit would be linked to this phenomenon. A more objective technical definition of fiscal sustainability would relate to long-term maintenance of macroeconomic equilibrium. 24. Assuming that the borrowing strategy in financing the budget deficit is to maintain a specific target debt-to-output proportion from year to year. Let the target ratios be: bd =btd/y and bf = bf*e/y where b is the real level of domestic debt and bf is the real foreign debt and y tge real GNP. Maintaining this fixed strategy is the same as satisfying the condition that the borrowing during the year, denoted by b and bf, will not grow faster than the growth of the domestic and foreign degt, respectively. This implies that bd/y = g*bd and bf/y 2 (g-C)*bf where g is the real growth of GNP and c is the rate of real exchange rate depreciation. 25. This is the same as saying that the new domestic borrowing as a ratio to real GNP is equal to the rate of growth of GNP multiplied by the domestic-debt-to-GNP ratio and that the new foreign borrowing is equal to the real rate of growth of output corrected for real exchange rate depreciation multiplicd by the foreign-debt-to-GNP ratio. 26. The remaining component, the monetary financing, as stated, is divided into the seignorage component and the inflation tax. Seignorage is demand-for-money determined. It is, however, related to the level of inflation in the sense that, it has been established by empirical studies, that the proportion of money held by citizens and institutions is relatively higher when prices are stable than when there is a high inflation rate. Thus, as the inflation tax rises, the seignorage component of financing falls. In Malaysia's case, as explained in the text, seignorage is more important than inflation, which, as a rule, has been negligible. 27. Using the budget equation as a controlling framework for sustainability, it can be shown that the sustainable fiscal deficit can be stated, as follows: d = g*bd + (g-c)*bf + g*m + P^*m (2) where g is the real growth of GNP, c is the rate of real foreign exchange rate depreciation of the domestic currency, P* is the annual inflation rate, m is the ratio of real base money to real GNP, or m/y. 28. Table 3.Al shows a matrix of growth rates of the real GNP and the corresponding fiscal deficit, given some values of the target domestic debt to 7 122 - Chapter II - 122 - ~~~~Appendix I Oage 7 of 8 output and foreign debt to output ratios, the level of demand for real money balances as a ratio to GNP, and an annual inflation rate. Table 2.A1: FINANCIALLY SUSTAINABLE FISCAL DEFICITS A B C D E F C H I Growth rate CNPKP g 0.040 0.045 0.050 0.055 0.060 0.065 0.070 0.075 0.080 Domestic debt/ CNPKP ratio b/y 0.725 0.725 0.725 0.725 0.725 0.725 0.725 0.725 0.725 Foreign debt/ GNPKP ratio b*y 0.410 0.410 0.410 0.410 0.410 0.410 0.410 0.410 0.410 Foreign exchange depreciation rate c 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 Domestic inflation rate p 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020 MI/CNPKP m/y 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 0.200 Sustainable deficit as ratio to CNPKP d/y 0.051 0.058 0.065 0.071 0.078 0.085 0.091 0.098 0.105 29. The discussion in the text centered on the rates of growth of GNP between 4.51 and 5.5% per year. The discussion can be reversed by asking the following question. Given a target real fiscal deficit ratio to GNP, and assuming some planned debt to output ratios, knowledge of MI/GNP, and some tolerance limits for price inflation, what is the rate of growth of real GNP which is consistent with the fiscal objective? 30. The borrowing policy can be affected by many factors especially the presence of favorable factors. If trade surpluses occur, leading to accumulation of foreign assets, or when the domestic saving rate increases, the government may feel that it can finance larger fiscal deficits. On the other hand, it may even decide to prepay principal for some foreign loans. What is recognized in this formulation is that the economy is subject to reasonably defined constraints about financing the fiscal deficit. The consi3tency of deficit targets with the growth rates for the economy is a major consideration, among others, for avoiding larger than necessary fiscal deficits. 31. For further details, see the analysis developed in the World Bank report, Fiscal Policy and Tax Reform in Turkey, Report No. 6374-TU, Vol. II, July 1987 on which this methodology is based. Chapter II - 123 - Appendix 1 Page 8 of 8 Tax Buoyancy Estimates 32. The estimates of tax buoyancy were derived from best fit equations from among competing linear regression models. These were regressions of the time series of tax collections for a particular tax on GNP and on other variables, like time trend, other economic variables like prices, etc. After some experimentation, it was found out that linear regressions on GNP and on changes of GNP produced the most interesting results. 33. The following model was used for estimaing each taxi: Taxi = aO+ &,GNP + a2(GNP-GNPO)*DUMHY where the base year is chosen as 1980, the year when the economy experienced the terms of trade break referred to earlier. The dummy variable is equal to zero until 1980 and is equal to 1 thereafter. A significant estimate of the coefficient implies a change in the slope of the overall buoyancy coefficient. A positive estimate means that the buoyancy estimate is higher in the latter period while a negative estimate implies that the buoyancy has fallen. The increase in the sales tax in 1983, although happening two years later than the base period, will be captured by the slope of the dummy variable, if it is significant. 34. Many good statistical fits were obtained for the simple linear regressions. But the best fits were derived from the logarithmic versions. The coefficients in the logarithmic versions are interpreted as estimates of tax buoyancy, while the coefficients of the untransformed data are estimates of slopes. In the few cases where the results were better for these forms of linear regressions, the buoyancies were estimated by multiplying the slopes with the reciprocal of the average ratios of tax collections to the GNP. 35. In the case of the estimates of buoyancies for the aggregative taxes in Malaysia, two distinct periods were computed: for 1970-1987 and for 1977- 1987. In both cases utilizing the dummy for the structural break in the economy was included. It turns out that there is some difference in the estimates of the coefficients for the two periods. In fact, an overall finding is that while the tax buoyancies are higher for the later and shorter period, it is also clear that the quality of the estimates is poorer than the longer period estimates. One reason for this is the wider variance of the data for the latter period as shown by the higher standard errors (or equivalently, lower values of the t-values of the coefficients). In most of the estimates, however, the coefficient, a2, associated with the structural break was not significant for the regression using the longer time series. The results are reported in Tables 2.6, 2.7, and 2.8 of the text. Chapter II - 124 - Appendix 2 Page 1 of 5 LIST OF FEDERAL STATUTORY BODIES AND GOVERNMENT COMPANIES I. Federal Statutory Bodies Statutory Body Incorporating Legislation A. Economic Sector (a) Agriculture 1. Pihak Berkuasa Kemajuan Pekebun Rubber Industry Shareholders Kecil Perusahaan Getah (RISDA) Development Authority Act 1972 2. Lembaga Kemajuan Ikan Malaysia Lembaga Kemajuan Ikan Malaysia Act (HAJUIKAN) 1971 3. Lembaga Pertubuhan Peladang Farmers Organization Authority Act 1973 4. Lembaga Kemajuan Ternakan Negara Lembaga Kemajuan Perusahaan Haiwan (MAJUTERNAK) Negara Act 1972 5. Lembaga Kemajuan Pertanian Muda Muda Agricultural Development (HADA) Authority Act 1972 6. Lembaga Kemajuan Pertanian Kemubu Kemubu Agricultural Development CKADA) Authority Act 1972 7. Lembaga Padi dan Beras Negara Lembaga Padi dan Beras Negara Act (LPN) 1971 (b) Land and Regional Development 8. Lembaga Kemajuan Tanah Land Development Ordinance 1956 Persekutuan (FELDA) 9. Lembaga Pemulihan dan Penyatuan National Land Rehabilitation and Tanah Negara (FELCRA) Consolidation Authority (Incorporation) Act 1966 10. Lembaga Kemajuan Trengganu Tengah Lembaga Kemajuan Trengganu Tengah Act (KETENGAH) 1973 11. Lembaga Kemaj .n Johor Tenggara Lembaga Kemajuan Johor Tenggara Act (KEJORA) 1972 12. Lembaga Kemajuan Pahang Tenggara Lembaga Kemajuan Pahang Tenggara Act (DARA) 1972 13. Lembaga Kemajuan Kelantan Selatan Lembaga Kemajuan Kelantan Selatan Act (KESEDAR) 1978 14. Lembaga Rema uan Wilayah Kedah Lembaga Kemajuan Wilayah Kedah Act (KEDA" 1981 ChptrII - 125 - Catr1 Appendix 2 Page 2 of 5 Statutory Body Incorporating Legislation (c) Commerce and Industry 15. Majlis Amanah Rakyat (MARA) Majlis Amanah Rakyat Act 1966 16. Lembaga Pasaran dan Pelesenan Malaysian Rubber Exchange and Getah Malaysia Licensing Board Act 1972 17. Lembaga Pendaftaran dan Pelesenan Palm Oil Registration and Licensing Minyak Kelapa Sawit (PORLA) Authority (Incorporation) Act 1976 18. Lembaga Perindustrian Kayu Malaysian Timber Industry Board Malaysia (Incorporation) Act 1973 19. Lembaga Tembakau Negara National Tobacco Board (Incorporation) Act 1973 20. Lembaga Perusahaan Nenas Tanah The Pineapple Industry Ordinance 1957 Melayu 21. Lembaga Pemasaran Pertanian Federal Agricultural Marketing Persekutuan (FAMA) Authority Act 1965 22. Jemaah Pamasaran Lada Hitam Federal Agricultural Marketing Authority Act 1965 Fl. Lembaga Kemajuan Perindustrian Federal Industrial Development Malaysia (MIDA) Authority (Incorporation) Act 1965 24. Perbadanan Kemajuan Filem Perbadanan Kemajuan Filem Nasional Nasional Malaysia Malaysia 1979 25. Perbadanan Kemajuan Pelancungan Touri-t Development Corporation of Malaysia (TDC) Malaysia Act 1979 (d) Transport 26. Pertadbiran Keretapi Tanah Melayu Malayan Railway Ordinance 1948 (KTM) 27. Lembaga Pelabuhan Kelang Port Authorities Act 1963 28. Suruhanjaya Pelabuhan Pulau Penang Port Commission Act 1955 Pinang 29. Lembaga Pelabuhan Kuantan Port Authorities Act 1963 30. Lembaga Pelabuhan Johor Port Authorities Act 1963 31. Lembaga Pelabuhan Bintulu Bintulu Port Authority Act 1981 32. Lembaga Cukai Api Federation Light Dues Ordinance 1953 33. Pusat Kumpulanwang Perusahaan Merchant Shipping Ordinance 1952 Perkapalan Chapter II - 126 - Appendix 2 Page 3 of 5 Statutory Bodv Incorporating Legislation (e) Finance 34. Bank Negara Malaysia Central Bank of Malaysia Ordinance 1958 35. Kumpulanvang Simpanan Pekeria Employees Provident Fund Ordinance (KWSP) 1951 36. Bank Simpanan Nasional Bank Simpanan Nasional Act 1974 37. Lembaga Urusan dan Tabung Haji Lembaga Urusan Tabung Haji Act 1969 (LUTH) 38. Pertubuhan Keselamatan Sosial Employees Social Security Act 1969 Pekerja (SOCSO) 39. Bank Pertanian Malaysia Bank Pertanian Malaysia Act 1969 40. Totalisator Board Racing (Totalisator Board) Act 1961 41. Lembaga Tabune Angkatan Tentera Electricity Act 1949 (Revised) (f) Public Utilities 42. Lembaga Letrik Negara Electricity Act 1949 (Revised) 43. Lambaga Lebuhraya Malaysia Highway Authority Malaysia (Incorporation) Act 1980 44. Perbadanan Pembangunan Bandar Perbadanan Pembangunan Bandar Act (UDA) 1971 (g) Research and DevelopmenL 45. Lembaga Penyelidikan dan K_-majuan Malaysian Rubber Fund (Research and Getah, Malaysia (MRRDB) Development) Ordinance 1958 46. Lembaga (Penyelidikan dan Tin Industry (Research and Kemajuan) Perusahaan Timah Development) Fund Ordinance 1953 47. Institiut Penyelidikan Minyak Palm Oil Research and Development Act Kelapa Sawit Malaysia (PORIM) 1979 48. Institiut Penyelidikan dan Malaysian Agricultural Research and Kemajuan Pertanian Malaysia Development Institute Act 1969 (MARDI) 49. Institiut Piawaian dan Standards and Industrial Research Penyelidikan Perindustrian Institute of Malaysia (Incorporation) Malaysia (SIRIM) Act 1975 50. Pusat Penyelidikan Getah Yalaysia Rubber Research Institute of Malaysia (RRIM) Ac: 1966 Chapter II - 127 - Appendix 2 Page 4 of 5 Statutory Body Incorporating Legislation Bo Social Sector (a) Education and Training 51. Inhtitiut Teknologi MARA Institiut Teknologi MARA Act 1976 52. Dewan Bahasa dan Pustaka Dewan Bahasa dan Pustaka Act 1959 53. Lembaga Pusat Education Act 1961 54. Universiti Kebangsaan Universities and University Colleges Act 1971 55. Universiti Teknoloji Malaysia Universities and University Colleges Act 1971 56. Universiti Malaya Universities and University Colleges Act 1971 57. Universiti Sains Malaysia Universities and University Colleges Act 1971 58. Universiti Pertanian Malaysia Universities and University Colleges Act 1971 59. Majlis Peperiksaan Malaysia Malaysian Examinations Council Act 1979 60. Hospital Universiti Universities and University Colleges Act 1971 (Established in 1966) 61. Maktab Kerjasama Malaysia Co-operative College (Incorporation) Act 1968 62. Pusat Daya Pengeluaran Negara National Productivity Council (Incorporation) Act 1966 C. Miscellaneous 63. Lembaga Loteri Perkhidmatan Social and Welfare Services Lotteries Masyarakat dan Kebajikan Board Ordinance 1950 64. Lembaga Perancang Keluarga Negara Family Planning Act 1966 65. Lembaga Kumpulanwang Buruh India South Indian Labour Fund Ordinance Selatan 1956 66. Majlis Sukan Negara National Sports Council of Malaysia Act 1971 67. Lembaga Stadium Merdeka Merdeka Stadium Corporation Act 1963 Chapter II - 128 - Appendix 2 Page 5 of S Statutory Body Incorporating Legislation 68. Pertubuhan Berita Nasional Pertubuhan Berita Nasional Malaysia Malaysia (BERNAMA) Act 1967 69. Yayaaan Tun Razak Tun Razak Foundation Act 1976 70. Lembaga Amanah Muzium Seni Negara National Art Gallery Ordinance 1959 71. Majlis Agama Islam Wilayah Federal Territory (Modification of Persekutuan Administration of Muslim law Enactment) Order 1974 72. Kumpulanwang Simpanan Guru-guru Teachers Provident Fund Rulea 1962 - 129 - III. THE LABOR MARKET: ISSUES IN THE DEVELOPMENT AND EMPLOYMENT OF HUMAN RESOURCES A. Introduction 3.1 With an open unemployment rate in 1987 almost double the rate in 1981, the fear that unemployment will reach double dipit levels in 1988 is generating serious concern. This problem, now regarded as top priority by the political leadership, has become a focal point in the ongoing public debate on the urgency of stimulating the economy and the appropriate macroeconomic policies to solve the current recession. The heightened concern about the unemployment problem has raised questions about its nature, the functioning of labor markets, and their ability to efficiently a4locate human resources. It has also called into question the effectiveness and efficiency of government programs and policies for employment and human resource development. 3.2 Much of the present macroeconomic policy debate centers on the issue of reflation. Certain sectors of society are clamoring for increased government expenditures to directly stimulate aggregate demand and employment. Others like the Bank Negara, however, strongly oppose such a policy suggestion and have opted for maintaining the present policy of austerity to consolidate the gains from economic adjustments. Central to the issue of the macroeconomic policy debate are price and wage adjustment questions. 3.3 Looking at the problem from a microeconomic perspective, some analysts contend that standard macroeconomic (monetary and fiscal) policy measures cannot be relied upon to bring down unemployment. The problem, they contend, should be looked at in the context of the larger issue of the ability of labor markets and institutions to promote or hinder efficient allocation and development of human resources. This view is based on several assump- tions, the first being that a market failure may have occurred because of an alleged "balkanization" or segmentation of the labor markets that prevents movement of labor from low to high productivity work or from labor-surplus to labor-shortage areas. Another element is the supposed inability of the system to match '-he supply and demand of particular types of labor, as evidenced by high rates of graduate unemployment and skills mismatches. It is also argued that the prevailing policy regime distorts the structure of incentives and promotes capital-intensive industrialization. Finally, wages are perceived to be rising faster than productivity and, hence, there is fear that the interna- tional competitiveness of Malaysian manufactured exports may be eroding. 3.4 This chapter addresses these various concerns and attempts to resolve them by examining the Malaysian labor market in relation to the unemployment problem and selected human resource development issues. It begins with a macroeconomic perspective on the problem, followed by a descrip- tion of the labor supply, a profile of the unemployed and trends in unemploy- ment patterns. Thereafter, the hours of work and structure of employment are discussed, followed by an assessment of labor absorption by industry. The functioning of the labor market and the issues of labor mobility and wage/employment adjustments are examined thereafter, followed by a review of - 130 - industrial relations, compensation-related labor laws, including the Employee Provident Fund, overtime regulations and termination benefits. The discussion is rounded out with a comparison of labor cost trends in Malaysian manufacturing and those of some competitors, and a brief review of efficiency problems in higher education and public training institutions. Finally, in response to the issues identified in the course of the review, a set of policy suggestions are provided. B. Macroeconomic Perspective 3.5 At an earlier stage of its development, Malaysia was characterized as a labor-surplus economy. This was true in the sense that by being employed predominantly in agriculture, labor was not fully utilized during the year due to the seasonality of padi cultivation and other types of agricultural activities. By about the mid-seventies it became apparent that Malaysia was no longer a labor-surplus economy, and the country began the decade of the eighties with an impressive economic growth record and full employment. In 198Z, the unemployment rate, which may be regarded as a full employment (natural unemployment) rate, was estimated to be 4.61. 3.6 In 1983, however, unemployment increased to 5.2Z, rising to 5.81 in 1984 and 8.3Z in 1986 (see Table 3.1). As the economic recovery gathered strength, as measured by the acceleration in the growth of gross domestic cutput from 1.2% in 1986 to 5.1Z in 1987, the rate of increase of the 1, unemployment rate decelerated to 7.061 from 19.98% during the previous year.- 3.7 The analysis of the macroeconomic trends raises an intriguing question. Why should the unemployment rate rise from 1982 to 1984 when the rate of growth of rDP was accelerating at 5.91, 6.31, and 7.8% in 1982, 1983 and 1984, respectively? 3.8 One hypothesis, which will be explored below, is that much of the observed unemployment is voluntary, reflecting greater willingness to search and wait for better and more appropriate jobs. With a fully employed economy, there emerged in the eighties a stronger tendency for the work force to be more selective in choosing jobs. Households are richer and can afford to finance the cost of extended unemployment until an appropriate and more rewarding job is offered. The young, more educated and urbanized, appear to have developed a strong preference to work in the modern sector rather than in traditional activities. This development is evident in the study of the so- called second generation problem (Ogawa and Chan 1985). A study of attitudes 1/ The official unemployment rate in Table 3.1 is higher than estimates directly obtained from Department of Statistics (DOS) Labor Force Sur- veys. It is argued that a change in the definition of unemployment in the 1980s caused unrealistically low estimates of the unemployment rate. Consequently, using administrative records and the Survey of Establish- ments, the Labor Force Survey figures were adjusted upwards in an attempt to make them consistent with estimates for earlier years. The official estimates are therefore much higher than the Survey figures. Table 3.1: TOTAL LABOR FORCE, EMPLOYMENT, UNEMPLOYMENT RATE AND OUrPUT 1970 1975 1980 1981 1982 1983 1984 1985 1986 1987 Labor force 3,682 4,539 5,122 5,313 5,500 5,727 5,907 6,039 6,222 6,409 (3.59) (3.35) (3.73) (3.52) (4.13) (3.14) (2.23) (3.03) (3.00) Employment 3,396 4,247 4,835 5,061 5,247 5,429 5,565 5,624 5,707 5,851 (3.71) (3.61) (4.67) (3.68) (3.46) (2.51) (1.06) (1.47) (2.52) Unemployment 286 292 287 251 253 298 342 415 516 558 (1.90) (-.41) (-12.54) (0.8) (17.79) (14.77) (21.35) (24.34) (8.14) Unemployment rate 7.8 6.4 5.6 4.7 4.6 5.2 5.8 6.9 8.3 8.7 GOP growth rate (constant prices) - 0.5 7.4 6.9 5.9 6.3 7.8 -1.0 1.2 5.1 GNP growth rate (constant prices) - 1.9 8.8 7.5 4.6 3.9 6.8 -1.5 2.1 5.8 a Growth of GDP per worker (constant prices) - - - 2.2 2.2 2.7 5.1 -2.1 2.5 4.4 Source: EPU. - 132 - of the young towards estate jobs by Nayagam and Sepien (1981) also reports similar findings. However, while the fall in per capita incomes after 1984 should ceteris paribus attenuate voluntary search unemployment, the recession and the slow recovery of the economy complicated the unemployment situation and resulted in the acceleration of the unemployment rate in 1985 and 1986. Labor Supply 3.9 The population of Malaysia was estimated to be about 16.6 million in 1987, up from 15.8 million in 1985 and 13.9 million in 1980, for an average annual growth rate of 2.6X during the first half of the eighties. In the second half of the decade, the Fifth Malaysian Plan projects population growth at 2.5% p.a. and a population of some 17.8 million by 1990. 3.10 The working-age population defined as persons in the age bracket 15 to 64 constituted about 58.2% of the total population in 1985. This ratio, which was only 56.5% in 1980, is projected to increase to 59.3% in 1990. The working-age population has therefore been growing rapidly, at an average annual 3% in 1981-85, but is expected to decelerate to 2.8% in the second half of the eighties, rising from 9.7 million in 1985 to 11.3 million in 1990 (Table 3.2). Table 3.2: LABOR FORCE GROWTH, 1980-90 Average annual 1980 1985 1990 growth rate (Z) Age group '000 x '000 x '000 o 1981-85 1986-90 15 - 19 667.5 13.1 658.0 11.1 669.8 9.9 -0.3 0.4 20 - 34 2,455.7 48.1 2,925.9 49.4 3,365.3 49.5 3.6 2.8 35 - 44 991.6 19.4 1,187.0 20.1 1,440.4 21.2 3.7 3.9 45 - 54 666.5 13.0 780.1 13.2 904.7 13.3 3.2 3.0 55 - 64 327.6 6.4 366.1 6.2 417.7 6.1 2.2 2.7 Total 5,108.9 100.0 5,917.1 100.0 6,797.9 100.0 3.0 2.8 Source: Fifth Malaysia Plan. 3.11 About 65% of those in the working-age group are in the labor force. The labor force participation rate, which appears to be slightly declining over all, varies by age and sex. As expected, the labor force participation rate is declining among the school age group 15-19 and is rising among women. - 133 - 3.12 Given these various trends, the total labor force grew from 5.11 million in 1980 to 5.92 in 1985 at an average growth rate of 3X per year. In 1986-1990 the Fifth Malaysian Plan expects the labor force to increase at a less rapid rate of 2.8%. In absolute terms, the average annual increment in the number of labor force participants was around 168,000 between 1983 and 1987. 3.13 This number of new entrants or re-entrants into the labor force, is much larger than the number of retrenched (laid-off) workers. (The reported number of retrenched workers at its peak in 1985 was less than 50,000. Even if the number is doubled to allow for under-reporting, it would still be less than the increment in the total labor force.) It would therefore be misleading to assess the pattern of unemployment on the basis of the characteristics of the registered retrenched workers alone. Hence, in the next section the analysis of unemployment will be based on the more wide- ranging data compiled by DOS in its Labor Force Survey. Profile of the Unemployed 3.14 Who are the unemployed? And what is the incidence rate of unemployment among various socioeconomic, geographic and demographic groups? What do microeconomic data on the profile of the unemployed tell about the nature of the unemployment problem? 3.15 Measured unemployment, as defined by DOS, consists of active and passive unemployed persons. The former are those actively seeking jobs during the reference week, while the latter are not actively seeking work but would be willing to work if offered a job. According to surve, findings, 461 of the unemployed were actively seeking work in 1986. 3.16 Unemployment in Malaysia is predominantly short-term and is concentrated among the young, especially those with secondary education. Table 3.3 shows that close to 70% of the unemployed in 1986 was under 25 years old. The unemployment rate among the age groups 15-19 and 20-24 was 3.1 and 1.9 times the national average, respectively, in 1986. Table 3.4, on the other hand, shows that persons with secondary education accounted for 67.2X of total unemployment in 1985. They also exhibit the highest rate of unemploy- ment, particularly among those with upper secondary education. In 1986, the education-specific unemployment rate relative to the labor survey national average stood at 0.4 and 0.66 for persons without a formal education and for those with a primary education, respectively (Table 3.5). The ratio rapidly increases with the level of secondary education, peaking at 2.24 among labor force participants with upper secondary education. It then dramatically drops to 0.59 among tertiary education graduates. - 134 - Table 3.3: RATIO OF AGE-SPECIFIC UNEMPLOYMENT RATES RELATIVE TO THE AVERAGE OF ALL AGE GROUPS* (Percentage distribution of unemployed persons by age in parentheses) 1982 1985 1986 All Strata 15-19 3.49 (42.0) 3.28 (33.4) 3.08 (31.8) 20-24 1.74 (34.6) 1.91 (37.4) 1.87 (36.6) 25-29 0.54 (8.7) 0.58 (9.6) 0.70 (11.5) over 29 0.22-0.46 (14.7) 0.60-0.32 (19.6) 0.46-0.31 (20.1) Urban 15-19 3.16 (36.9) 3.12 (29.7) 2.96 (26.7) 20-24 1.52 (33.6) 1.63 (36.6) 1.68 (36.7) 25-29 0.62 (10.9) 0.61 (11.24) 0.71 (13.3) over 29 0.71-0.32 (18.6) 0.73-0.33 (22.46) 0.70-0.39 (23.3) Rural 15-19 3.73 (44.9) 3.25 (35.5) 3.12 (34.8) 20-24 1.97 (35.2) 2.07 (37.8) 1.99 (37.2) 25-29 0.51 (7.5) .56 (8.8) 0.68 (10.4) over 29 0.40-0.16 (12.4) 0.53-0.25 (17.9) 0.42-0.18 (17.6) Source: Department of Statistics, Labor Force Survey. *Peninsula Malaysia Table 3.4: PERCENTAGE DISTRIBUTION OF UNEMPLOYED PERSONS BY EDUCATION LEVEL: MALAYSIA No formal Total education Primary Secondary Tertiary 1980 100 11.0 26.2 61.8 1.0 1985 100 5.8 23.0 67.2 3.4 Source: Department of Statistics, Labor Force Survey. - 135 - Table 3.5(a): EDUCATION SPECIFIC UNEMPLOYMENT RATES RELATIVE TO THE NATIONAL AVERAGE* 1982 1985 1986 No formal 0.314 0.453 0.394 Primary 0.486 0.585 0.662 Lower secondary 1.343 1.358 1.296 Middle secondary 2.143 1.623 1.535 Upper secondary 2.80 2.358 2.239 University/college 0.514 0.774 0.592 Table 3.5(b): ANNUAL GROWTH RATE IN EDUCATION SPECIFIC UNEMPLOYMENT RATES* (percent) 1982-1985 1985-1986 No formal 39.4 16.7 Primary 27.4 51.6 Lower secondary 17.7 27.8 Middle secondary 4.9 26.7 Upper secondary 9.7 27.2 University/college 42.6 2.4 Source: Department of Statistics, Labor Force Survey. *Peninsular Malaysia - 136 - 3.17 By duration, it appears that 81S of the unemployed have been without jobs for less than a year (Table 3.6). Although short-term unemployment is predominant, long-term unemployment, defined as persons unemployed for a year or more, remains an important component at u=der 201 in 1986. Table 3.6: PERCENTACE DISTRIBUTION OF UNEMPLOYED PERSONS BY DURATION OF UNEMPLOYMENT* 1981 1985 1986 Less than 3 months 41.5 (46.4) /a 43.3 (43.3) 41.5 (41.5) 3 months to < 6 months 14.9 (61.0) 17.9 (61.2) 20.3 (61.8) 6 months to < 1 year 17.3 (75.2) 20.2 (81.4) 19.0 (80.8) 1 year to < 3 years 18.8 (93.9) 14.4 (95.8) 15.7 (96.5) 3 years & over 7.5 (100.0) 4.2 (100.0) 3.5 (100.0) /a Cumulative distribution. Source: Department of Statistics, Labor Force Survey. *Peninsular Malaysia - 137 - 3.18 Given the concentration of unemployment among the young, especially those with secondary education, search unemployment is expected to constitute the bulk of the unemployed. A survey of the unemployed not seeking work bears out this assumption. Of the passive unemployed, about 21Z in 1986 gave "none or no suitable job available" as their reason for not actively seeking a job, up from 7.4Z in 1982. A substantial proportion, some 581 in 1986, said they were waiting for answers to their job applications. And in the case of unemployed graduates, 20Z were not looking for work because they had gotten a job that was about to start. Only a small proportion of the unemployed gave lack of qualification as a reason for not seeking work. This is not very surprising considering that the unemployed are better educated and a substantial number of them (641) have work experience. 3.19 The unemployment rate by family income in Table 3.7 is revealing by showing that 411 of the unemployed in 1984 belonged to a household with gross monthly income below M$700 (median income at the time was M$723). Of equal importance, it reveals that households with higher per capita income tend to have a mach greater incidence rate of unemployment, indicating that many are unemployed because their families can afford it. This provides further evi- dence that voluntary search/waiting unemployment is a substantial factor in the Malaysian labor market. 3.20 The unemployment rates for the various States relative to the average for Peninsular Malaysia are given in Table 3.8, with the States arranged according to average monthly household income per capita as of 1984. The States can be split into two groups according to their 1984 household incomes: the first six (lower) income states have relative unemployment rates over one in 1985; the other (higher income) states have relative unemployment rates equal to one or less. This pattern is persistent, with only two exceptions between 1980 and 1984. Negeri Sembilan in 1984 had a relative unemployment rate slightly greater than one and Perlis had a relative unemployment rate of slightly less than one. 3.21 Of the six disadvantaged states, three are in the north and two are in the east, with Perlis, Kedah and Terengganu having the highest relative unemployment rates in 1985 (1.43, 1.40, and 1.36, respectively). Kelantan, the poorest state, ranks fifth in 1985. 3.22 Not unexpectedly, the poor, less developed states generally have higher unemployment rates. Hence, the data suggest that richer regions may be enjoying a more developed social infrastructure that facilitates job search and/or that workers in those areas can access opportunities for better jobs more easily than those in more depressed locations. Consequently, more developed states and regions tend to exhibit lower unemployment rates despite the positive income effect on voluntary unemployment. 3.23 Finally, the urban-rural incidence of open unemployment (Table 3.9) indicates that standard dualistic models of migration and unemployment are not applicable. The standard theory asserts that because wages are higher in urban than in rural areas and, given that urban wages are not flexible downwards, the onus of equilibriating the economic system falls on the unemployment rate. At equilibrium, expected wage rates in urban and rural - 138 - labor market should be equal. Hence, given the higher urban wage, the urban unemployment rate should be higher. In reality, the urban unemployment rate in Malaysia was slightly lover in 1986 than in previous years when the difference was much more substantial. Consequently, because of the higher rural unemployment rate and the fact that the greater majority of the people continue to live in rural areas, the bulk of the unemployed are to be found in those communities. Table 3.7: PERCENTAGE DISTRIBUTION OF UNEMPLOYMENT BY MONTHLY HOUSEHOLD INCOME CLASS (MALAYSIA) Relative Percent of Income class Unemployed Labor Force unemployment Unemployed (Ringgit) (percent) (percent) rate /a passive (Z) < 149 1.89 19.72 .10 66.7 150-199 2.02 4.94 .41 58.9 200-299 5.64 11.31 .50 59.0 300-399 6.78 12.z7 .54 57.4 400-499 8.29 10.77 .77 61.6 500-599 7.24 8.43 .86 50.8 600-699 8.76 6.58 1.33 50.9 700-799 6.21 4.91 1.26 54.6 800-899 5.95 3.73 1.60 49.1 900-999 5.83 2.66 2.19 51.8 1000-1249 11.04 4.79 2.30 51.6 1250-1499 8.41 2.81 2.99 53.2 1500-1749 5.61 1.76 3.19 52.6 1750-1999 4.14 1.14 3.63 39.0 2000-2499 4.45 1.38 3.22 45.9 2500-2999 2.92 .77 3.79 44.4 3000-3499 1.64 .45 3.64 49.9 3500-3999 1.04 .29 3.59 46.8 4000-4999 1.06 .42 2.52 44.6 5000 and over 1.04 .58 1.79 45.3 Total 100 100 1.00 52.6 /a Ratio of income-specific unemployment rate relative to the total average. Source: Department of Statistics, Household Income Survey, 1984. - 139 - Table 3.8: RELATIVE UNEMPLOYMENT BY STATE AND LEVEL OF DEVELOPMENT Per capital 1984 mean monthly CDP per capita Unemployment rate /b (1985) household income 1980 1984 1985 Low Income Change /d 1. Kelantan 1740 (E)/a 150 1.328 (2)/c 1.170 (3) 1.132 (5) 0.5 2. Kedah 2358 (N) 161 1.180 (4) 1.191 (2) 1.40 (3) 1.5 3. Perlis 2604 (N) 170 1.279 (3) 0.979 (7) 1.434 (1) 3.0 4. Terengganu 4719 (E) 183 1.426 (1) 1.447 (1) 1.358 (2) 0.4 5. Perak 3194 (N) 227 1.132 (6) 1.149 (4&5) 1.151 (4) 0.7 6. Melaka 2765 (C) 241 1.180 (5) 1.149 (4&5) 1.038 (6) 0.1 High Income 7. Negeri 3846 (C) 258 1.0 1.043 (6) 0.887 Sembilan -0.2 8. Johor 3324 (S) 266 0.951 0.957 1.0 0.8 9. Pahang 3495 (E) 266 0.852 0.979 0.811 -0.3 10. Penang 4120 (N) 292 0.902 0.766 0.943 1.4 11. Selangar 4963 (C) 393 0.787 0.851 0.887 0.7 12. Wilayah 7783 (C) 491 0.623 0.745 0.660 Persekutuan -0.3 Peninsular 274 1.0 1.0 1.0 Malaysia /a E = east, N = north, C = central, S = south. Ratio of state-specific unemployment rate relative to the average for Peninsular Malaysia. /c Number in parentheses refers to rank order of the six states with the highest relative unemployment rate. /d 1984/85 change in unemployment rate. Source: Government of Malaysia, Fifth Malaysian Plan; Department of Statistics. Labor Force Survey, 1984; General Report of the Population Census, 1983. - 140 - Table 3.9: COMPARATIVE UNEMPLOYMENT RATES: HALE/FEMALE AND URBAN/RURAL* 1982 1985 1986 Ratio Urban/rural 0.838 0.860 0.972 Male/female 0.596 0.778 0.908 Annual percentage growth rate of unemployment rate Both sexes Urban 19.4 42.9 Rural 18.0 26.3 All strata Male 25.0 40.8 Female 11.35 20.6 Urban Male 26.9 53.2 Female 8.9 30.8 Rural Male 24.1 34.0 Female 11.5 15.7 Source: Department of Statistics, Labor Force Survey. *Peninsular Malaysia - 141 - Changes in the Patterns of Unemployment 3.24 A comparison of data for 1982, 1985 and 1986 unemployment indicates that the increase in unemployment pervades all groups but the rate of change varies. The most noteworthy change in the unemployment rate is among grad- uates of tertiary education. Between 1982 and 1985, this rate increased by 42.61 annually (Table 3.5(b)). The rapidity with which it has risen generated serious concern about graduate unemployment, although clearly it is the secondary school leavers who are predominantly the unemployed, with tertiary education graduates constituting only 3.4Z of the total unemployed. In 1986, a dramatic deceleration in graduate unemployment occurred. The rate of change fell to 2.4%. Consequently, the graduate unemployment rate relative to the labor survey national average, which rose from 0.514 to 0.774 between 1982 and 1985, dropped to 0.592. 3.25 Except for those without formal education, all other educational groups exhibit accelerating unemployment growth rates, particularly the middle and upper secondary education groups. While group-specific rates for 1987 are not available yet, it is reasonable to expect that they have decelerated following the overall trend in the total unemployment rate. 3.26 In terms of demographic groups, Table 3.3 shows a consistent decline among 15-19 year olds in both the share of total unemployment and the un- employment rate. The trend holds true for both urban and rural areas. Among 20-24 year oldst unemployment first rose between 1982 and 1985 and then fell in 1986. Unemployment for males has worsened faster than for females. Hence, in Table 3.9, the kale to female unemployment rate ratio increased from 0.596 to 0.908. 3.27 Locadionally, unemployment in urban areas, which was 16Z less than that in rural areas, rose only slightly faster during 1982-85. From 1985 to 1986, however, urban/rural differential growth rates were much more pronounced, with the ratio of the urban to rural unemployment rate rising from 0.875 to 0.97. It would appear, then, that in the eighties the unemployment rate has moved towards greater equality not only between the sexes but also between urban and rural areas. 3.28 The pattern of unemployment by state has changed favorably for the poorest state (Kelantan). From the relative unemployment rate of 1.32 in 1980, it declined to 1.17 in 1984 and 1.13 in 1985. Perlis, on the hand, experienced the greatest increase in the relative unemployment rate in 1985, when it rose to 1.43 after declining from 1.28 in 1980 to 0.97 in 1984 3.29 What have been the changes in the duration of unemployment? It would appear that hard core term unemployment has declined perceptibly. In 1981 the proportion of persons without f'ployment for three years or more was 6X (Table 3.6), but declined to 4.2% in 1985 and to 3.5% in 1986. In addi- tion, the share of those unemployed for a year or more declined from 18.7% in 1981 to 14.4% in 1985, then increased slightly to 15.7% in 1986, a smaller decline than that for longer-term unemployment. Both hard core and very short-term unemployment came down even during recessionary times. While those unemployed for less than six months remained remarkably stable at about 61%, - 142 - the share of those under three months decreased from 46.41 in 1981 to 43.3Z in 1985 and 41.5Z in 1986. The implication is that people are engaged in a little more search or waiting for better jobs. And jobs as such do not appear to be more difficult to find as suggested by the decline in hard core unemployment. Hours of Work and Employment Structure 3.30 What are the trends in hours of work? And how does the structure of employment look in the changed economic environment of the eighties? Table 3.10 shows that the employed generally continued to work long hours as has been observed in the past. Although there are perceptible changes in recent years towards shorter work hours, they are slight and do not appear to be serious. In 198f, 43.6Z of the employed worked 48 hours or more, down from 46.5Z in 1985 and 4'i.3Z in 1982. On the other hand, those working less than 25 hours rose from 8.8Z in 1982 to 10.2Z in 1985, inching up to 10.6Z in 1986. It is interesting that in 1986 the percentage of employed labor not at work (during the survey reference week) was substantial in agriculture and increased in almost all sectors. Data here seem to indicate that some reduc- tion in overtime work and perhaps increased use of part-time workers may be happening, allowing overworked labor to take extended rest ar.d vacation. - 143 - Table 3.10: PERCENTAGE DISTRIBUTION OF EMPLOYED PERSONS BY INDUSTRY AND HOURS WORKED PER WEEK* Total c 25 25-34 35-47 48 and over Not at work All 1982 100 8.8 10.8 30.2 45.3 4.9 1985 100 10.2 9.3 29.1 46.5 4.9 1986 100 10.6 10.2 29.4 43.6 6.2 Agriculture 1982 100 18.9 19.6 26.8 26.0 8.8 1985 100 22.8 16.2 27.8 25.9 7.3 1986 100 23.3 15.0 29.1 22.5 10.0 Industry 1982 100 4.6 5.7 24.3 62.7 2.6 1985 100 5.4 4.9 24.1 61.8 3.7 1986 100 5.9 6.4 24.3 59.8 3.6 Services 1982 100 5.3 10.7 45.8 33.3 4.9 1985 100 6.3 10.5 40.3 38.2 4.7 1986 100 6.1 12.3 39.3 35.2 7.1 Source: Department of Statistics, Labor Force Survey. *Peninsula Malaysia 3.31 Table 3.11 presents the distribution of employment by status (i.e., whether the worker is an employer, employee, own account/self-employed, or unpaid family worker). After steadily rising from 57% to 66.8% between 1967 to 1985, the percentage of employees dropped to 64.6% in 1986. On the other hand, the share of self-employed and unpaid family workers shows the opposite trend. The rank order of industries in terms of their share of wage employment remained the same over time. In 1986, the service sector ranked first at 91.5%, followed by agriculture at 83% and industry at 68.5%. Manu- facturing registered at 83.3% compared to 47.3% for commerce. - 144 - Table 3.11: PERCENTAGE DISTRIBUTION OF EMPLOYED PERSONS BY EMPLOYMENT STATUS AND INDUSTRY* Unpaid Own account family Employer Employee worker worker Total All 1967 2.4 57.0 25.6 15.0 100 1979 2.8 62.5 21.7 13.0 100 1982 3.3 65.4 21.3 10.0 100 1985 3.2 66.8 20.5 9.6 100 1986 3.3 64.6 22.1 10.0 100 Agriculture 1967 0.92 43.7 32.5 22.9 100 1982 1.1 35.1 42.9 20.9 100 1985 1.4 35.1 41.3 22.1 100 1986 1.2 32.1 44.7 22.0 100 Industry 1967 4.9 64.2 23.1 7.7 100 1982 5.2 69.4 17.1 8.3 100 1985 4.6 71.0 17.0 7.3 100 1986 5.0 68.5 18.0 8.2 100 Manufacturing 1967 4.7 64.6 12.5 8.2 100 1982 3.1 83.9 10.0 3.0 100 1985 2.4 84.3 10.8 2.5 100 1986 2.5 83.3 11.2 3.0 100 Commerce 1967 7.5 41.9 36.3 14.4 100 1982 7.7 45.0 28.0 19.3 100 1985 6.7 49.2 27.6 16.5 100 1986 7.0 47.3 28.1 17.5 100 Services 1967 2.5 82.9 10.0 4.6 100 1982 2.0 92.4 4.9 0.8 100 1985 2.1 92.4 4.9 0.5 100 1986 2.3 91.5 5.5 0.7 100 Source: Department of Statistics, Labor Force Survey. . , . .. - 145 - 3.32 With one notable exception, the employment status trends observed for the economy as a whole are generally reflected in all industries. Agri- culture and manufacturing stand out, however. The latter shows the biggest increase (20 percentage points) in employees between 1967 and 1985. In agri- culture the employee share, unlike the growing share in other sectors, fell from 43.7% in 1967 to 35.1% in 1982, then falling to 32.1% in 1986. The percentage of unpaid workers remained stable, while the proportion of own- account workers rose from 32.5% in 1967 to 42.3% in 1982 and 44.7% in 1986. The trend in the agricultural sector is due to the land resettlement program, specifically FELDA, while the recent movement towards self-employment is a consequence of the slowdown or cutback in employment by the Government and the industrial sectors. 3.33 According to the DOS surveys, agriculture is the single biggest employer (35%), followed by commerce and other services (22%), and then by Government and manufacturing (15% each). With wage employment accounting for around 65% of total employment, the 15% share of government in the work force constitutes a substantial fraction of wage and salary workers. In relation to the nonagricultural wage employment markets, government probably accounts for as much as 35%. Consequently, wage and employment decisions of government have a strong influence on the labor market. C. Labor Absorption by Industry 3.34 The nature of the employment problem in the 19809 is different from that in previous decades. Previously, the problem was the lack of productive work opportunities for rural labor during seasons of low agricultural activity. When Malaysia reached full employment and completed its agro- industrial transition, it went through a process that changed the structure of the economy and the characteristics of the work force. Labor became more urbanized, educated and richer. The exceptional growth of the economy, in particular the rapid development of the modern sector in the seventies, created expectations about career paths and encouraged human capital investments geared towards modern sector wage employment. 3.35 The failure to sustain industrial growth and generate jobs in the modern sector, after an increase in workers' desires to get more rewarding and modern jobs, resulted in higher unemployment. Hence, beyond the issue of the employment effect of the general economic slowdown, it is important to under- stand the sectoral sources and the nature of the jobs being generated in the eighties. 3.36 Table 3.12 presents the level and rate of growth of employment by industry. Several things are notable in the data. First, after 1984 the most rapidly growing industry in terms of employment is the "other services" sector, of which the biggest component is commerce and trade. At its peak in 1985, this sector was growing at 8.6% per year compared to the average of 1.4%. Second, growth of agricultural employment rose after 1985 from 0.5% or less to 2.6% and over. In comparison, construction, which had the fastest growing labor demand from 1980 to 1982, registered a 2.6% decline in 1985 and another -3.6% in 1987. Manufacturing employment declined by -1.9% in 1985 and by another -2.2% in 1986. It grew again in 1987 by 1.3%, but this was much Table 3.12: LEVEL AND GROWTH OF EMPLOYMENT BY INDUSTRY /a 1980 1982 1983 1984 1985 1986 1987 Agriculture 1,911.0 1,929.3 1,924.7 1,932.4 1,903.3 1,954.7 2,005.5 (0.5) (-0.3) (0.4) (-1.5) (2.7) (2.6) Mining & 80.1 69.3 65.8 66.8 60.5 52.5 53.8 quarrying (-7.2) (-5.1) (1.5) (-9.4) (-13.2) (2.5) Manufacturing 755.1 799.4 814.6 843.9 828.0 810.0 820.5 (2.8) (1.9) (3.6) (-1.9) (-2.2) (1.3) Construction 270.2 340.4 360.9 370.9 378.7 368.7 355.8 (11.6) (6.0) (2.8) (2.1) (-2.6) (-3.5) Finance, business 78.3 89.6 94.8 98.2 101.6 104.6 108.4 services and (6.7) (5.8) (3.6) (3.5) (3.0) (3.6) real estate Transport, 209.5 234.0 239.9 254.1 264.9 269.4 271.0 storage and (5.5) (2.5) (5.9) (4.2) (1.7) (0.6) communication Government 658.2 765.1 785.8 803.2 819.5 835.9 843.9 (7.5) (2.7) (2.2) (2.0) (2.0) (1.0) Other services 855.0 938.0 984.8 1,024.3 1,112.0 1,172.7 1,228.9 (4.6) (5.0) (4.0) (8.6) (5.5) (4.8) Total 4,817.0 5,165.1 5,271.4 5,393.8 5,468.5 5,568.1 5,687.8 (3.5) (2.1) (2.3) (1.4) (1.8) (2.1) /a Number in parentheses is annual growth rate. Source: Ministry of Finance, Economic Report 1987/1988. Fifth Malaysian Plan. - 147 - lower than the growth rate in the seventies. Mining had negative growth rates in the eighties except in 1984 and 1987, when it registered a 2.5X growth. While government employment continued to grow in 1987, its 1.0% growth rate is down from 2.01 in 1985 and 7.5% around the beginning of the decade. 3.37 The Economic Report for 1987/88 of the Ministry of Finance gives data on the industry-specific increment in employment and sectoral performance in labor absorption (Table 3.13). After 1985, agriculture took much of the burden of labor absorption. From less than 0.05 or even a negative rate, the ratio of incremental agricultural employment relative to total incremental employment dramatically rose to 0.52 and 0.424 in 1985/86 and 1986/87, respectively. In contrast, manufacturing with a marginal share of over 0.127 prior to 1984 made a negative contribution with a ratio of -0.21 and -0.18 between 1984 and 1986, respectively. Its ratio of 0.088 in 1986/87 is still below its historical experience, however. - 148 - Table 3.13: THE RELATIVE CONTRIBUTION OF INDUSTRY TO EMPLOYMENT GENERATION /a 1980-1982 1982/1983 1984/1985 1985/1986 1986/1987 Total 348.1 106.3 74.7 99.6 119.7 (increment) Agriculture 0.523 -0.043 -0.390 0.516 0.424 Mining & quarrying -0.031 -0.033 -0.084 -0.080 0.011 Manufacturing 0.127 0.143 -0.213 -0.181 0.088 Construction 0.202 0.193 0.104 -0.100 -0.108 Finance, business 0.032 0.049 0.046 0.030 0.032 services & real estate Transport, storage 0.070 0.056 0.145 0.045 0.013 & communication Government 0.307 0.195 0.218 0.165 0.067 Other services 0.238 0.440 1.174 0.609 0.470 Total 100 100 100 100 100 /a The ratio of the annual increment in number of persons employed by an industry relative to the increment in total employment. Source: Ministry of Finance, Economic Report, 1987/1988. 3.38 The marginal share of government declined from 0.307 during the first two years of the eighties to 0.165 in 1985/86, plunging to 0.067 in 1987. From about 0.20 prior to 1984, the relative contribution of construc- tion turned -0.10 after 1985. Mining and quarrying was negative throughout the eighties except in 1987 when it posted a substantially positive contribution. "Other services" (commerce and other distributive industries) increased in importance during the eighties. Clearly, this sector took most of the burden of absorbing labor in the past five years. Its relative contribution of 0.238 in 1980/82, which was smaller than that of government services (0.307), rose to 1.174 in 1984/85. As agriculture increased its share of labor absorption at the margin after 1984 and manufacturing partially - 149 - recovered in 1987, the relative contribution of the other services sector declined but remained the largest. D. The Functioning of the Labor Market 3.39 The apparent failure of the labor market to clear has raised questions regarding its nature and functioning. Discussions have highlighted concerns about wage flexibility and labor mobility. The consensus is that Malaysia is losing its competitive edge in the export market due to accelerating wage increases that are out of step with labor productivity. From a macroeconomic perspective, the question of the downward flexibility of the nominal wage remains to be answered. The Malaysian labor market is also supposedly segmented and, hence, wages do not provide appropriate signals about relative scarcities. This hypothesis provides a seemingly natural explanation for the problem of "skills mismatch" and the coexistence of unemployment and labor shortages in certain areas of the economy. 3.40 The labor segmentation hypothesis, if correct, could have important implications on what the general approach should be regarding human resource development and the unemployment problem. In a world where labor mobility is restricted and relative wages give distorted signals about relative scarcities, the basic strategy will call for very highly targetted, direct and, hence, detailed government interventions. It downgrades approaches that rely on adjustments in the incentive structures facing economic agents at the sectoral or macroeconomic levels. In a relatively well-functioning labor market, it is reasonable to develop a strategy that is based primarily on such --has with direct and targetted proiects playing an important but subsidiary role. It is, therefore, important to look at the validity of the labor market segmentation hypothesis, which appears to have influenced the policy discussions of human resource and employment policy in Malaysia. Labor Mobility and Market Segmentation 3.41 The notion of labor market segmentation in Malaysia was introduced in a study showing large differentials in earnings among various categories of labor. The persistence of these wage differentials even after controlling for education and for length of work experience seemed to indicate labor markets that were not effectively linked and workers whose mobility was highly restricted. A well-functioning labor market with mobile labor should be characterized by its ability to efficiently sort and match workers and jobs, with compensating wage differentials as the signalling mechanism. The labor market segmentation hypothesis, however, relates to the notion that wage differentials persist because workers' labor mobility is highly restricted. It is this particular interpretation of labor segmentation, which led to concern about the allocative efficiency of wage signals in Malaysia, that is discussed here. 3.42 The labor market segmentation hypothesis seems doubtful. Labor mobility in Malaysia is, in fact, quite high and there are good indications of a narrowing of wage differentials between skilled and unskilled occupations. Furthermore, the evidence used to advance the idea of labor segmentation is of highly doubtful validity. - 150 - 3.43 Regarding the third point, questions have been raised about the validity of the study's assumption that, by holding age (a proxy for experience) and years of education constant, one is able to control for productivity-related differences among workers. In reality, the quality of education and experience differs among individuals. Furthermore, productivity- related cognitive and non-cognitive characteristics exist which are often not directly observable by researchers, but are known to the individual laborer and the employer. 3.44 Tha2the Malaysian labor market is highly fluid has been shown by Smith (1983),-' using the Malaysian Family Life Survey (MFLS). The data reveal that after 35 years of work experience after entry into the labor market, a typical Malaysian male held more than four distinct jobs. Furthermore, the average inter-job turnover is not simply due to a few very highly mobile workers; rather, most workers are involved in the process. Seven out of 10 male workers had switched jobs at least once after 10 years of market experience. And, although job mobility is lower in Malaysia than in the highly mobile U.S. market, where a male worker typically has 7.6 jobs after age 20, the life-cycle patterns are remarkably similar. Turnover in both countries peaks early in the career (between five and ten years of exper- ience). Finally, it is reported that 60X of job separations reflect movement between broad occupational categories, while one third are changes within detailed occupational codes. 3.45 Since formal education is easily accessible in Malaysia, except at the tertiary level, high intergenerational educational mobility has been found in the MFLS data. 3.46 Regarding residential mobility, however, it would appear that, despite the near universal participation in job change, haif of the male workers remain attached to the village in which they started work. The other half, according to Smith's (1983) study, are frequent movers, with 40% of them migrating six time or more. 3.47 This paradox can be resolved as follows. Job changes among workers are nearly universal and very frequent, partly because the industrial structure is such that costly long-term migration and transfer of residence are often not required to take advantage of developments in the labor markets. The agro-industrial transition in Malaysia achieved a balanced urban/rural distribution of industries and job opportunities. Thus, 51.3% of industrial employment is located in rural communities. In manufacturing, rural employment accounts for 54.5%. 3.48 Given the locational structure of industries, a development characteristic of the Taiwan experience, the pattern of residential migrations found in the MFLS data is not surprising. Smith found that, partly reflecting return migration or very short-term moves, 60% of all rural migrants moved to 2/ "Income and Growth in Malaysia," James P. Smith, Rand Monograph R2941- AID, May 1983. - 151 - another rural district, while 40X of urban migrants relocated to another urban community. Among migrants who changed urban/rural status, there was a two-way flow, with net migration favoring the rural areas. Remarkably, however, 60X of job changes according to the MFLS data involved transitions between agricultural and nonagricultural employments. 3.49 The labor segmentation hypothesis is further weakened by indications of a narrowing of certain wage differentials. In manufacturing, a census and survey of manufacturing industries undertaken by Soon Lee Ying in 1987 shows that, in general, the rate of wage increase is faster in the low-wage than in the high-wage industries (see the statistical annex). Survey estimates of the coefficient of variation of average wages of major industries reveal that the coefficient dramatically fell from 0.86 in 1968 to 0.41 in 1983. Some narrowing of wage differentials are also indicated for some occupational groups. The same survey shows that the wage rate of skilled and unskilled labor between 1973 and 1983 rose faster than those of managerial, clerical, and professional workers. 3.50 Further indications that the wages of unskilled and low-skilled labor are rising faster than other wages can also be deduced from Table 3.14. It has been suggested that a good substitute for wage data is the differential rates of increase in components of the consumer price indices produced by workers of varying skills (Oshima 1986). Table 3.14, which shows the rates of growth-of prices paid for services produced by unskilled and low- skilled workers, reveals that they are rising faster than the overall consumer price index and the CDP deflator for 1980-83. Table 3.14: CONSUMER PRICE INDEXES IN MALAYSIA, 1975-1982 1980 to 1983 Annual growth rates in percent of CPI in services (1980-1983) Payment for sewing, knitting 14.6 Laundering, cleaning, dyeing, etc. 12.9 Domestic services 11.8 Services of barbers and beauty shops, etc. 11.4 Expenditure in restaurants, cafes, and hotels 13.2 Other services 4.8 All Services 7.6 Sources: Upper panel data from Monthly Statistical Bulletin, Kuala Lumpur: June 1984; lower panel data from Department of Statistics. Oshima (1987). - 152 - 3.51 An often-cited paradox in Malaysia is the coexistence of pockets of labor shortages and the problem of unemployment. These pockets of unemploy- ment are often reported in the agricultural sector, especially by the estates. However, information about the magnitude of the labor shortage problem may be exaggerated since it comes mostly from sources that want to use low-wage foreign workers. Furthermore, applications of plantation employers through the Manpower Department to recruit local youths to work in the estates are usually unsuccessful, indicating lack of interest among the young to work on the plantations. Finally, a study of the labor shortage problem in planta- tions reveals that a major reason for the lack of interest is low wages. Clearly, the labor shortage complaints by agricultural employers reflect a labor market condition that requires them to pay a larger premium or compen- sating wage differential to attract local workers, whose attitudes are changing unfavorably against plantation work. A substantial proportion of the active population, especially the young who have no desire to work on small holdings, expressed willingness to be temporarily unemployed in their desire to migrate to towns in search of wage employment, steady income, better education and training, and job security. 3.52 The framing of the labor shortage issue in agriculture in relation to unemployment and the functioning of the labor market is thus misleading. The problem is clearly not due to some barriers preventing the unemployed from filling those jobs, but arises because some jobs are available that could be done profitably at low wages, but not at rates that will be sufficient to compensate local workers for the growing disutility of working in plantations and similar occupations. In short, as far as local workers are concernea, effective supply and demand is in balance, allowing for voluntary search/waiting unemployment. 3.53 The picture emerging from the aforementioned data is of a labor market that is highly competitive (albeit, with many imperfections). The data point to a high degree of economic integration, especially when compared with other developing countries. Workers are highly mobile, and relative wages do change depending on the supply and demand conditions in the various labor markets. E. Wage and Employment Adjustments 3.54 During the first half of the eighties, real wages increased at an average annual rate of 4.2%, with wages in manufacturing accelerating from an annual growth rate of 1.9% in 1970-80 to 10.8Z in 1980-85. In recent years, however, wages in the manufacturing sector appear to have moderated, with an apparent slight decline in 1987. Wage Policies in the Public Sector 3.55 Contrary to popular perception, the Government has effectively controlled its wage and employment growth. Government employment growth in 1987 was only 1% (para. 3.37), while the average government salary has come closer to that of other sectors. Table 3.15 reveals that in the second half of the seventies the average government salary rose very rapidly. As a - 153 - consequence, government compensation per employee relative to average remuneration per worker for the economy as a whole jumped from less than 1.15 prior to 1976 to 1.53 in 1977. With the relative wage at 1.45 at the close of the seventies, the wage difference remained substantial at 1.40 at the beginning of the eighties. Between 1980 and 1985, however, wages in the non- governmental sectors climbed rapidly, while the salaries of public employees stagnated. The effect was to bring the wage ratio back close to where it was in 1975. Table 3.15: COMPENSATION PER WORKER IN MANUFACTURING AND GOVERNMENT RELATIVE TO THE NATIONAL AVERAGE Annual 1980 M$ Ratio (M) (G) (N) Manufacturing Government National average M/N GIN 1967 3,713 3,260 3,205 1.16 1.02 1968 3,967 3,597 3,452 1.15 1.04 1969 3,874 3,648 3,537 1.10 1.03 1970 3,641 3,792 3,413 1.07 1.11 1971 3,628 4,222 3,579 1.01 1.18 1972 3,517 4,665 3,731 0.94 1.25 1973 3,012 4,144 3,471 0.87 1.19 1974 3,358 4,103 3,)eu U.94 i.i 1975 3,491 4,126 3,587 0.97 1.15 1976 3,741 6,186 4,541 0.82 1.36 1977 3,884 8,071 5,276 0.74 1.53 1978 3,957 7,947 5,276 0.75 1.50 1979 4,268 7,679 5,315 0.80 1.45 1980 4,414 7,760 5,554 0.80 1.40 1981 4,438 7,976 5,676 0.78 1.40 1982 4,975 7,433 5,831 0.85 1.28 1983 5,351 7,417 6,017 0.80 1.23 1984 6,548 7,322 6,267 1.05 1.17 1985 7,368 8,274 6,829 1.08 1.21 Source: Wong and Najimudin (n.d.). Productivity and Wage Adjustments 3.56 In general, the real wage in Malaysia has kept in step with labor productivity in the long run. Between 1950 and 1973, real per worker GDP grew annually at 2.3%. Remarkably, the MFLS data reveal that the time trend of monthly earnings of an average worker was quite close at 2.4% during the same period. Furthermore, estimates based on payroll data show that average yearly salary per worker was growing at 5% annually for 1970-80, roughly the same as the rate of increase of GDP per worker during those same years. As the growth - 154 - of output per worker decelerated in 1980-85, a roughly similar adjustment appeared to have taken place in the growth of compensation per worker. 3.57 Despite the coincidence between the growth rates of average real wage and labor productivity, the two do diverge in some years and on a sector- specific level. This is to be expected in a changing open economy that is vulnerable to shocks in the international economic environment. Lack of perfect foresight, high information cost, and possibly, restrictive institutional arrangements are underlying factors for the divergence. Consequently, the process of balancing wage and productivity is never smooth. One may lag behind at one point and then surge ahead of the other at another time in the process of catching up. 3.58 Analysis of the adjustment process is limited by the availability of reiiable information on wages. Hence, the analysis of wages is to a large extent judgmental. The data used here are drawn from a DOS study of workers' compensation which, by including retrenchment benefits, might reflect changes in occupational composition and number of days worked. Consequently, annual changes give false impressions about wage flexibility. This information is complemented by other data on average monthly earnings by occupation collected by the Ministry of Labor through its Occupational Wage Survey (OWS) of workers. While the OWS data are not subject to distortions due tc :etrench- ment payments and occupational changes, they nonetheless may be distorted by changes in the number of work days. However, the two data sources taken together should provide indications of broad movements (see statistical annex for tables). 3.59 The process of adjustment to maintain the productivity/wage balance is complex. In recent years, labor market correction has involved a combination of adjustment mechanisms including nominal wage adjustments and quantity adjustments designed to improve the productivity of the remaining employed labor vis-a-vis their wages. That there is some downward flexibility, at least for selected occupations, is illustrated by the OWS data which show that the monthly salary rate of workers in manufacturing for selected occupations rose substantially between 1980 to 1983, but declined in 1985 for a production supervisor and foreman and, probably, for typists and office clerks as well. It is interesting that while the rate remained the same for an industrial efficiency engineer, it rose for plant maintenance mechanics and unskilled labor. 3.60 Average monthly earnings of workers in occupations in those industries most affected by the recession have generally been reduced. According to the Ministry of Labor, between 1984 and 1986, the nominal monthly earnings of construction workers in 18 out of 23 occupational groups fell by substantial amounts. The prolonged recession in construction seems to have affected the wages of selected workers in related industries (e.g., the manufacture of concrete and cement) as early as 1985. Data for 1983 and 1985 indicate that downward adjustments in nominal wage rates may be part of the market adjustment. The OWS indicates that the monthly and daily wage rate of workers in selected occupations have dropped except for unskilled labor, which remained the same in 1985--the year when construction output fell but employment continued to rise. But as the construction industry suffered and - 155 - its employment fell, the wage rate of unskilled labor in the manufacture of concrete and cement may have also fallen. 3.61 In the oil palm industry, annual growth of nominal daily wage rates of factory workers in the estates were negative from 1982 through 1986, except during 1984 when the palm oil price increased by 50%. The nominal daily wage rate of workers fell drastically in 1985 and 1986, when the price of palm oil dropped by -26% and -48%, respectively. In the rubber estates, factory workers similarly exhibited strong (downward) flexibility of wages, and tappers took a cut of -15Z in 1984, although they experienced more years of wage gains than losses. 3.62 The flexibility of wages in the agricultural sector must be substantial since the basis of compensation in this sector is generally linked to the output and its price. There is also a widespread use of the traditional system of sharecropping and piece rate compensation method. Furthermore, where labor unions are active as in the plantations, the collective bargaining agreements on wages stipulate a system which links compensation with productivity. The system has three elements: (a) the basic wage; (b) an incentive pay that depends on the individual worker's output; and (c) a factor that is tied to the price of rubber or oil palm as the case may be, 3.63 Such flexibility is also present to some extent in other industries. Fringe benefits, bonuses, allowance and commissions constitute a significant proportion of total earnings. EPU data for the economy as a whole reveal that they constitute between 10% and 20X of total earnings, and 15% for male workers in 1984. This, however, is relatively low compared to Korea, where the share of bonuses and overtime payments in total compensation in 1981 was 30%, split evenly between the two. 3.64 Reduction in entry level salaries is a major element in the downward flexibility of wages. Data from two different sources are consistent with this hypothesis. One source is the Malaysia Employer's Federation (MEF) wage survey of "bargainable" non-executive positions, the other is an estimate of salaries of new members of the Employees Provident Fund (EPF). 3.65 The MEF data were not designed to establish wage trends but are suggestive nonetheless. Between 1984 and 1985, when the manufacturing sector suffered a decline of -3.8% in production and -1.9% in total employment, 14 of the 19 occupations surveyed show negative nominal wage change. Only three of them exhibited a positive change, while two remained the same. In 1986, when the economy began its partial recovery with output growing at 7.5% but employment declining by another -2.2%, the number of occupations showing wage declines decreased from 14 to 8 to become even with those exhibiting no change. Noticeably, the number with positive movement remained the same. But by 1987, when both output and employment were growing at 8% and 1.3%, respec- tively, 9 out of 18 occupations had wage increases and only 5 experienced a decrease. The other 4 showed no change. 3.66 While it is not possible to identify a trend in the overall average wage rate at entry level jobs from the MEF data, the EPF data reveal that the - 156 - average implicit salary levels of EPF members between 1986 and 1987 declined by 12.5%. This decline is apparent in Table 3.16, even after controlling for age. MEF data for the three largest groups of occupations (skilled, semi- skilled, and unskilled labor) also show the entry level (minimum) wage to be falling from 1985 to 1987 (Table 3.17). In addition, payroll data for manufacturing indicate a fall in nominal compensation per worker in 1987. 3.67 The conclusion then from the above discussion is that wages seem flexible downwards. This appears to be true particularly for entry level jobs. Table 3.16: DISTRIBUTION OF NEW EPF MEMBERS REGISTERED IN 1986 AND 1987 BY AGE AND SALARY (percent) All ages Less than 25 25-30 years Over 30 years Monthly salary 1986 1987 1986 1987 1986 1987 1986 1987 200 and below 41.8 55.0 43.8 57.7 25.3 34.4 40.3 47.8 201-300 27.0 22.4 28.2 22.9 19.5 18.4 22.4 20.9 301-400 15.0 10.7 15.4 10.5 14.0 13.3 11.6 9.2 401-500 4.4 3.2 4.0 2.7 7.1 6.4 5.7 5.7 over 500 11.8 8.8 8.6 6.2 34.0 27.4 19.9 16.5 Total 100 100 100 100 100 100 100 100 Average 256 224 Source: Employees Provident Fund (EPF). Table 3.17: MINIMUM WAGE OFFERS FOR WORKERS WITH LESS THAN THREE YEARS EXPERIENCE (1984 - 100) 1984 1985 1986 1987 Highly skilled 100 67.2 67.2 59.7 Semi-skilled 100 90.0 90.0 80.0 Unskilled 100 90.0 80.0 80.0 Source: Malaysian Employers Federation (MEF): (a) Survey on terms and conditions of employment of bargainable position, November 1984. (b) Wage/salary report, non-executive positions, October 1985 and August 1987. - 157 - Quantity Adiustments 3.68 The process of balancing productivity and wage in Malaysia involves quantitative adjustments. For various reasons outlined in Table 3.18, many firms may find it more profitable or cost-effective to cut back on hours of work or to retrench workers than to drastically cut wages in full proportion to a decline in demand. 3.69 Retrenchment (lay-offs) in manufacturing was substantial in 1985 and 1986. Employment growth fell by -1.9X in 1985 and by another -2.21 in 1986. Clearly, the adjustment in employment is a response to the persistence of falling demand. However, nominal and real wages in the manufacturing industry as measured by payroll data appear largely positive ezcept for 1987. The rate of increase from 1981 to 1986 is unlikely to be solely due to statistical distortions arising from the inclusion of retrenchment benefits. It most probably also reflects some wage stickiness. 3.70 The retrenchment of workers in 1985 and 1986 helped to improve the wage/productivity balance, though the per worker output is also the result of shifts towards more intensive production induced by high wage growth and government heavy industrialization policies. In 1980-85, average labor productivity grew at only 2.24S per year. This is apparently less than half of real wage growth (Table 3.19). In contrast, in 1985-87, per worker output grew at 9.871, or about five times the rate of growth of the real wage. Clearly, the increased productivity growth was due to the ability of the manufacturing industry to grow with less labor inputs or without a proportionate increase. In this regard, the value added elasticity of employment in the 1980s appears to have fallen to 0.40 from 0.71 in 1971-75 and 0.52 in 1975-80 (Table 3.20). That wage increases have pushed the economy towards less labor-intensive prcc^ases can be gleaned from Table 3.21, which shows constant wage elasticities of demand for labor that are much higher than the EPU estimates, which do not control for wage changes. The lower EPU estimates are to be expected, given the fast rate of wage increases and the relatively high wage elasticity of demand for labor indicated in the table. 3.71 In construction, employment adjustment is also evident. Demand has been declining since 1985 due to tighter fiscal constraints and excess building capacity. When demand fell by -8.41 in 1985, employment was rising at 2.11. But, when an additional reduction of -141 came in 1986, employment eventually dropped by -2.61, and with another output decline in 1987, the reduction in employment accelerated to -3.5Z. Meanwhile, productivity change decelerated from -11.41 to -3.0X. F. The Average Wage in a Two-Tier Labor Market 3.72 The preceding discussion strongly suggests the development of a two- tier labor market, which is a natural consequence of a decline in demand in an economy where the nominal wage is not fully flexible downwards. For various reasons outlined in Table 3.18, employers would rather lay off some workers to align productivity and wages rather than cut the salary of employees and avoid lay-offs. Thus, when demand falls, the labor supply essentially breaks into the incumbents who are retained by firms and the retrenched workers. The former are paid at the same wage rate as before, presumably plus whatever 1 wvy inmrr ^ao h v° Alrga Ay henn or.u4 iinnn. The lolef-off workers, on the Table 3.18: A SYNOPSIS OF ALTERNATIVE EFFICIENCY WAGE THEORIES Problems leading to Benefits to firm ary efficiency wage payments of high wages rking Imperfect observability Raise cost of job loss of worker effort level encouraging good per- and performance; formance; economize on monitoring is costly monitoring costs iover Firms must bea.r part of High wages reduce turn- turnover costs (hiring over costs if quit rate and training costs) is decreasing function of wages erse Imperfect observability Attract higher quality ection of worker quality and pool of applicants If performance more productive workers have better outside opportunities iological Morale and worker Improved work norms, feelings of loyalty to morale, feelings of firm depend on perceived loyalty to firms which fairness of wages raise productivity on threat Costs of replacing Maintain industrial existing workitorce gives peace or prevent union- employees barg;aining ization power rce: Katz (1986). - 159 - Table 3.19: PERCENTAGE GROWTH RATE OF WAGES, EMPLOYMENT AND LABOR PRODUCTIVITY IN MANUFACTURING Real wage Nominal wage Labor Productivity Employment Output Quarter Annual (quarter) (quarter) (annual) (annual) 1970-80 - 1.9 - 4.9 - - 1980-85 - 10.8 - - 4.1 - - 1981-84 6.6 10.6 12.7 4.8 6.3 - - 1985/86 5.9 - 6.4 5.1 3.9 - - 1985-87 3.2 - 3.9 4.4 4.8 - - 1981 9.4 - 19.1 .6 2.5 2.2 4.7 1982 5.0 - 11.2 - 8.2 -2.5 5.6 1983 5.5 - 9.5 5.4 6.0 1.9 7.9 1984 6.6 - 11.0 13.4 8.7 3.6 12.3 1985 5.3 - 5.7 -7.9 -1.9 -1.9 -3.8 1986 6.5 - 7.2 18.1 9.7 -2.2 7.5 1987 -2.4 - -1.1 3.2 6.7 1.3 8.0 Source: EPU, Department of Statistics, Monthly Survey of Manufacturing Industries. 3.73 The average wage (WA) observed in the market will be the weighted sum of the wages of the incumbents and the retrenched workers (WA and WR, respectively). That is, (1) WA = b*WI + (1-b)*WR where b is the ratio of the incumbents to total number of workers. The rate of change of the average wage (GWA) can be expressed as (2) GWA = GB*(WI-WR) + b*GWI + (1-b)*GWR where GB, GWI and CWR are the growth rates of b, WI and WR, respectively, and initially, Ca1. 3.74 Given the values of the parameters on the right hand side of equation 2, what happens to the average nominal wage? Assume that WI = 350 and WR = 225 in 1987 and that WI and WR rose by 1% over 1986. Given that the average duration of unemployment is roughly six months, suppose further that GB is 15%. Then, the rate of change of the average wage in 1987 would be about -5.01, while in 1986 it is zero. This exercise suggests that the average wage may have indeed started to decline and the reduction could be substantial. - 160 - Table 3.20: EMPLOYMENT VALUE ADDED ELASTICITY /a 1971-75 /b 1976-80 /b 1981-85 /c 1986-90 /c All sectors 0.65 0.43 0.45 0.46 Primary 0.54 0.30 0.05 0.11 Agriculture, forestry, livestock & fishing 0.48 0.39 0.12 0.19 Mining & quarrying -0.25 0.03 -0.92 -2.48 Secondary 0.73 0.52 0.60 0.52 Manufacturing 0.71 0.52 0.39 0.41 Construction 1.00 0.56 0.86 0.84 Tertiary 0.76 0.57 0.57 0.57 Electricity, gas & water 0.47 0.81 0.57 0.33 Transport, storage & communications 0.58 0.32 0.57 0.57 Wholesale & retail trade, hotels & restaurants 1.00 0.63 0.66 0.68 Finance, insurance, real estate & business services 0.74 0.64 0.74 0.52 Government services 0.69 0.56 0.46 0.53 Other services 0.71 0.64 0.71 0.70 /a Wage rate not held constant. /b Fourth Malaysia Plan, Tables 2-1 and 4-6. /c Fifth Malaysia Plan, Tables 2-10 and 4-6. Source: EPU. - 161 - Table 3.21: WAGE AND OUTPUT ELASTICITIES OF EMPLOYMENT USING NORMALIZED REGRESSORS, 1968-1985 C Wages Output R2 D-W Agriculture 3.09 -0.39 0.88 0.943 2.4 (8.2) -(3.2) (6.2) Mining -9.0 -2.10 3.77 0.748 2.0 -(3.3) -(5.5) (5.3) Manufacturing 0.03 -0.88 1.55 0.974 2.1 (0.1) (-4.4) (8.9) Construction 1.08 -la31 2.04 0.987 2.0 (0.6) -(1.8) (3.5) Note: Elasticities are computed using a double log function of the form in Eti I Ci ln Otl where Wt1 and 0 ° are the log of wages and output respectively. Figures in parenttheses denote t-values. Source: Wong and Najimudin (n.d.). G. Industrial Relations 3.75 Industrial relations in Malaysia are generally peaceful and favorable. There are no heavy pressures for major wage increases from organized labor, and strikes have been few in recent years. Employers are given protection to exercise their common law rights in running their business, and labor unions are not very aggressive, though they provide valuable services to their 0.6 million members. 3.76 The process of collective bargaining is working well. The success rate of collective bargaining is roughly 50% at the stage of direct bilateral negotiations, 25% at the stage of conciliation (assisted negotiations) and only the remaining 25% need to be referred to arbitration for settlement. Of the deadlocked negotiations, 502 are referred to arbitration voluntarily at the joint request of both labor and management. 3.77 A salient feature governing collective bargaining at present is that fundamental management functions are prescribed by law as management prerogatives which have been classified as non-negotiable matters. These include recruitment of workers, termination, promotion, transfer and allocation of duties. The Industrial Relations Act (1967) also prohibits collective agreements in pioneer enterprises, during their first five years of - 162- operation, from providing better conditions of service or employment than those legislated under the Employment Act. It has been pointed out that because of the broad definition of pioneer status under the Investment Incentive Act (1968), the coverage of this regulation is wide. 3.78 There is no legally built-in mechanism for a wage-price spiral in Malaysia. Capacity to pay is accepted as the overriding criterion in nego- tiation for wage increases. A material rise in the cost of living or the existence of higher wages in comparable undertakings would merely establish a prima facie case for a wage increase. But the actual proposal for a wage increase would have to be determined by the financial condition of the particular employer concerned. In this regard, the Industrial Court has ruled that the rate of wage increase for cost of living adjustment cannot exceed two-thirds of the growth rate of the consumer price index. 3.79 Trade unions have shown restraint and willingness to respond to labor market conditions and the firms' profitability. Informed sources report that some are even prepared for a wage freeze. Tnis is illustrated by the percentage increase granted by the Industrial Court on collective agreements in recent years. In 1986, the average rate of increase of the basic salary schedule was 4Z for three years (not annually), which is the standard length of contract in collective bargaining agreements. For manufacturing, the rate was slightly higher at 4.25%. By 1987, the average rate fell to 3.2Z, with the manufacturing industry registering an even lower rate of 2.5%. 3.80 Nevertheless, several current practices may not be conducive to wage flexibility in the short to medium run. First is the incorporation of predetermined automatic increments into the salary scale of workers, based on seniority or length of service with no explicit linkage to performance. Second, once granted, increases of all forms are built into the basic wages with no provision for coping with a downturn in business activity. The Employee Provident Fund and Other Labor Compensation Laws 3.81 Malaysia's minimum wage legislation (Wage Councils Act 1947) covers a very small segment of the labor market and has no significant impact on the economy. There are, however, other laws regarding labor benefits that might have a bearing on employment, the most important being the Employees Provident Fund (EPF) Act and the Employment Act of 1955 which stipulates rules on overtime and termination (severance) benefits. 3.82 The Employees Provident Fund. The EPF Act seeks to provide a provident fund that an employee can have at his disposal at the time of his retirement or other circumstances as provided for in the Act. It covers all persons employed under a contract of service regardless of occupation and salary. (Self-employed persons are allowed on a voluntary basis to become members.) Under the present provisions of the EPF Act, once an employee has commenced work for the same employer without a break in employment for a period of more than one month, both the employee and employer are required to make monthly contributions towards the EPF according to prescribed rates. Currently, the employee contribution is 9% of the wage, while his employer's is 11%. The contributions go the employee's own account, which the EPF Board - 163 - annually credits with interest payments as may be determined by the Board. In 1983-87, the interest rate paid was 8.5%. 3.83 A member's EPF accruals are paid to him under the following conditions: (a) attainment of age 55; or (b) incapacitation; or (c) permanent emigration; or (d) death. A member can also withdraw a portion of his EPF credit upon reaching age 50, or for a housing loan. After contributing for at least five years, a member is permitted to withdraw 45% of his EPF credit or an amount not exceeding 10% of the value of the house he wishes to purchase, whichever is the lower up to a maximum of M$20,000. 3.84 The employment affect of the EPF contribution is difficult to assess quantitatively. Assuming that a ringgit at a given discount rate contributed to EPF has the same value as the ringgit given by the employer as part of the workers's wage, then the EPF levy should not affect employment, although the composition of the workers's total remuneration would change, with regular take home pay decreasing. For older persons, such an assumption is probably valid. For young job entrants, however, there is probably a stronger preference for cash in hand vis-a-vis the promise of money in the distant future. Hence, there is no equivalence between the cash wage and the EPF package, and the EPF levy most likely has a negative employment impact. It reduces demand for young workers, but the shift, if any, in the standard wage- labor supply schedule in response to the EPF benefits might not be sufficient to keep the same employment level. 3.85 While demand elasticity estimates (see Table 3.21) point to values in the neighborhood of -0.5 or larger (in absolute term), the long-run own- wage elasticity of labor supply, on the whole, has been reported to be close to zero. In the short run, however, during a recessionary period labor supply may be very elastic. These numbers suggest that the short-run increase in the demand for labor in response to an 11% increase in the wages paid by employers can be as large as 6%. 3.86 That the wage (take home pay) remuneration may not be equivalent to the future sum to be received from EPF is indicated by the following market phenomena. First, employers are strongly opposed to any increases in EPF contributions. If the EPF levy has no effect other than the composition of the compensation package, employers should be neutral about it. Second, there have been reported abuses of members withdrawing EPF money under the pretext of a housing loan. 3.87 Overtime and Termination Benefits. Under the Employment Act of 1955, for any overtime work carried out in excess of the normal hours of work, an employee must be paid at a rate which is not less than one and a half times his hourly rate. Barring certain exceptions, an employee's normal hours of work cannot exceed more than eight hours in one day spread over a continuous period of ten hours inclusive of any period or periods of leisure, rest or break. 3.88 Regarding retrenchments, an employee continuously employed for a period of not less than 12 months is entitled to receive termination or lay- off benefits. The minimum amount of termination benefits for various years of - 164 - continuous service with a company is equivalent to about 4-8% of salary. As an illustration of the order of magnitude involved, two electronics and metal companies which recently retrenched 119 workers paid about M$3,500 each. 3.89 Quantitative assessment of the effect of overtime and severance pay on employment is not available. Qualitatively, it can be argued that severance pay probably reduces the firm's incentive to retrench, but also encourages the firm to lengthen search before hiring or to terminate new entrants before becoming eligible for termination benefit. Similar potential effects may be generated by a wage premium on overtime work. Apart from the substitution of capital for labor, high overtime pay provides an incentive for a firm to use more laborers working at normal hours and less of them doing overtime work. In this regard, as noted in the previous discussion of the distribution of employed persons by hours of work, the indication is that overtime work appears to have declined, most likely as a consequence of the recession. This is a trend that is likely to be generated in a regime of a high overtime wage premium. 3.90 To summarize, while industrial relations are working well, a few concerns need to be addressed. First, under collective wage agreements, wage increases are generated by a simple increase in length of service, without explicit regard to productivity and performance. Second, except for the rubber and oil estates, collective agreements on labor compensation do not sufficiently use incentive payments or a performance (profit) based bonus system. Third, EPF contributions affect the employment of the young. And, fourth, the overtime premium may prevent some firms from attaining a socially efficient combination of productive inputs, with firms rationally using more capital-;itensive processes than they otherwise would without the overtim pay rule. H. Manufacturing Labor Costs and International Competitiveness 3.91 Underlying the issue of wage flexibility is the concern that Malaysia may be losing its competitive edge in the international market. The Ministry of Labor and other government agencies are aware of the issue and are now reviewing the laws on labor compensation. An earlier World Bank report also warned of the acceleration of wages in Malaysia and its effect on labor cost in relation to trends in other exporting countries. 3.92 Table 3.22 presents an updated comparative index on manufacturing unit labor costs for Korea, Taiwan, Singapore, Hong Kong and Malaysia. Between 1980 and 1986 Malaysia's labor competitiveness was eroded vis-a-vis that in Korea, Taiwan and Hong Kong. During the period, Korea had a roughly constant labor cost, while Hong Kong's tended to decline. - 165 - Table 3.22: MANUFACTURING UNIT LABOR COST /a (in US$, 1980=100.0) Korea Taiwan Singapore /b Hong Kong /c Malaysia /d 1975 46.7 54.2 77.1 - 58.8 1976 61.2 59.6 69.3 - - 1977 74.4 68.5 73.5 - - 1978 88.6 70.5 81.5 - - 1979 107.0 84.3 88.3 - - 1980 100.0 100.0 100.0 100.0 100.0 1981 96.4 111.3 113.4 96.8 - 1982 104.8 112.2 132.0 107.1 - 1983 106.4 107.7 143.9 81.3 124.0 1984 98.9 120.7 140.6 73.8 145.1 1985 101.5 116.2 146.9 84.3 - 1986 101.9 120.4 131.9 81.3 148.1 1987 /e 120.4 141.2 - - 140.6 1988 7? 147.0 169.3 -- /a Unit labor cost in US$ = wages (labor productivity x exchange rate) where labor productivity is measured as real value added per average employment and the exchange rate is expressed per US$. /b Singapore wages are August weekly earnings for all manufacturing workers. For 1985-86, they are assumed to have increased at the same rate as wages for production, transport and other manual workers. Average employment is estimated as the average of the two year end figures. /c Wages and employment for Hong Kong are the average of March and September figures. Real manufacturing value added is obtained by deflating nominal value added by the GDP deflator. 1986 real value added is estimated from the industrial production index using the historical relationship during 1981-85. /d Wage estimate is based on per worker compensation from the survey of manufacturing. Value added per worker is derived from the Economic Report 1987/88. /e Estimates based on incomplete data. 7? Forecasts based on the assumptions that, in 1988, the average exchange rates against the US dollar will be appreciated by 5.0% from the end of 1987, and that wage and labor productivity increases will be roughly the same as the annual averages during the 1980s. Korea's nominal wage increase for 1988 is assumed to be 12% from the level settled through the labor strifes in the previous autumn. - 166 - 3.93 In 1987, Korea experienced a very rapid increase in labor cost, while Malaysia's unit labor cost index fell. Another huge increase is forecasted for Korea in 1988. If the trend towards wage moderation continues to prevail as expected, the foreign exchange rate does not appreciate substantially, and labor productivity growth of manufactur:ng remains positive, Malaysia should be able to recover lost ground to Korea in 1988. With respect to Taiwan, which experienced a dramAtic rise in labor cost in 1987 and will probably see another hefty increase in 1988, Malaysia should also be in a good position to gain in competitiveness. Singapore, however, was able to improve its competitiveness vis-a-vis Malaysia in 1986 through reduction in the payroll tax rates which caused the unit labor cost index to fall substantially. 3.94 Malaysia has thus recovered lost ground in its international competitiveness except in relation to Hong Kong, and prospects are good that its labor force will become even more competitive. However, its competitive- ness remains precarious, and will be maintained only if the ongoing labor market adjustment is allowed to run its course, productivity is improved, and the ringgit does not appreciate substantially. I. Government Measures Against Unemployment 3.95 The Government has attempted to reach three target groups through its various initiatives to address unemployment: (a) retrenched workers, con- sidered a high priority because of their family responsibilities; (b) secondary school leavers, since prolonged unemployment at the start of their working life could adversely damage their long-term work prospects; and (c) graduates, due to the potential social and political consequences, in addition to the perceived wastage of human capital. 3.96 In its Economic Report 1987/88, the Government lists the following measures which it has taken to deal with the unemployment situation: (a) implementing the special low-cost Housing Scheme which should generate more jobs; (b) deregulating the hawker industry by liberalizing licensing proced- ures and requirements in the retail and petty trade sector; (c) revitalizing small-scale and rural cottage industries which are labor-intensive; (d) encouraging the private sector not to retrench workers and to implement the concept of job-sharing; (e) implementing small rural projects including repair and maintenance jobs; (f) instituting a system of guest workers to meet labor shortages in specific sectors and to resolve the problem of illegal immigrants; - 167 - (g) embarking on crash programs to train and retrain the unemployed in marketable trades both as wage earners and self-employed; (h) revieving the labor laws to allow labor costs to be more competitive so as to encourage more labor-intensive industries; (i) alienation of land to the private sector including individuals and corporations for cultivation of viable commercial crops which will also create more jobs; (j) encouraging private investment to expand, thereby generating more opportunities for workers; (k) improving information on the labor market and job placements to those unemployed; and (1) revitalizing vocational training to equip workers with greater skills. 3.97 Government also began to implement a program for graduate employment in late 1987 which includes programs for special training, specific profess- ional skills, private sector attachment, education enhancement, farming, public sector employment and central supporting services. 3.98 It is difficult to evaluate the impact of these measures since many of them have only recently been implemented and data are not available to assess their effectiveness. Under the Special Housing Scheme, 80,000 low-cost housing units will be built over a three-year period. It is estimated that about 60,000 jobs will be generated per year. However, the employment impact of the housing scheme has been lower than expected due to the employment of foreign workers, the majority of whom are unskilled or semi-skilled. Despite this proviso, the government program is headed in the right direction. Government's promotion of small socially profitable rural projects involving repair and maintenance jobs would seem appropriate in view of the retrenchment of an estimated 60,000 Malaysian construction workers between 1984 and 1986. Deregulating the hawker industry is also well advised by supporting recent trends towards self-employment. On the other hand, government encouragement of the private sector not to retrench is essentially based on moral suasion and is unlikely to be effective. Encouragement of private investment and the revitalization of labor-intensive small-scale industries are also appropriate policies in principle, although certain problems will probably need to be resolved. 3.99 The most potentially important step by Government is the review of labor laws. The fruitfulness of the review depends, of course, on the nature of the revisions to be recommended. As discussed in paras. 3.88-3.90, there is considerable scope for revising labor-related laws in a way that would promote greater employment and productivity. The Government also emphasizes training and education in dealing with the problem of employment. However, as will be discussed in the following sections, there is a need to improve the finance and delivery system for human resource development. - 168 - J. Education and Training 3.100 Employment and wage growth issues cannot be dealt with properly without some attention to human capital investment and their linkages to the labor market. Consequently, even though a full discussion of the relevant issues is not possible here, this section will, nevertheless, coment on selected aspects of education and training, with a focus on the issue of the social productivity of higher education and public sector training programs. The Social Productivity of Higher Education 3.101 From an economic perspective, the seriousness of the graduate unemployment problem lies not in the graduates' being unemployed but in the productivity of their work. First, they can always find some kind of job, at least through self-employment, if they wanted to. Second, they are neither poor nor disadvantaged; in fact, they are well off enough to be able to finance a more extended voluntary job search. Furthermore, the number of graduates without jobs is relatively small compared to other groups, though the former are much more vocal and influential. 3.102 The tertiary level unemployment rate has clearly stabilized, as indicated in earlier discussions. Two factors have contributed towards the deceleration of graduate unemployment. First, based on anecdotal evidence, the graduates seem to have scaled down their job expectations (or reservation wage). It is interesting to note that private firms are being encouraged by government officials to hire unemployed graduates at supposedly a "bargain wage" of M$500. Second, those under the bonded scholarship program were released faster by Government from their obligation to serve in the public sector. 3.103 Beneath the unemployment problem, however, is the larger issue of the social productivity of investments in higher education embodied in both the unemployed and employed graduates. This is of greater significance because of the magnitude of the human capital involved. Graduates may be employed but the productivity of their jobs, as reflected by their wages, may not be commensurate with the enormous amount of resources spent on their education. Evidence for this is the low social rate of return to tertiary education. Table 3.23 shows a social internal rate of return (IRR) circa 1983 of 4% for bonded government scholars and 7.6% for others. These rates of return to higher education may be lower now due to the current looseness of the labor market. This conjecture is consistent with the fact that in academic year 1987/88 the number of applicants for first degree courses fell from 30,900 during the previous year to 27,600. - 169 - Table 3.23: PRIVATE AND SOCIAL INTERNAL RATES OF RETURN ON HUMAN CAPITAL FORMATION IN MALAYSIAN UNIVERSITY SYSTEM BY RACE AND TYPE OF GRADUATES Non scholarship Scholarship holders Race Private IRR Social IRR Private IRn Social IRR Malay 0.122 0.076 0.173 0.040 Chinese 0.116 0.071 0.189 0.043 Indian 0.108 0.065 0.189 0.048 E. Malaysian 0.091 0.050 0.200 0.051 Others 0.096 0.056 0.189 0.048 All 0.122 0.076 0.173 0.040 Source: IPT/UM Graduates' Survey, 1983. Mehmet (1986). 3.104 The Covernment heavily subsidizes tertiary education, as indicated by a comparison of the private and social rates of return. The private IRn is very high for bonded government scholars at 17%, which is four times higher than the social IRR of 4.0%. Nonrecipients of government scholarships got a 12% private IRn, a rate that is 1.6 times higher than the 7.6% social IRR. The high private rates of return coupled with limited local supply of higher education and to financial and other constraints have led to excess demand. In academic year 1986/87, for example, the intake was only 8,595 out of the 30,900 applicants for a first degree course. 3.105 The social rate of return for government scholars is low partly due to the high cost of studying abroad. At one time in 1980, the number of Malaysian students overseas was estimated to be 39,908, compared to the 20,764 students enrolled in degree-level courses in local universities. It is esti- mated that one scholarship abroad is equivalent to seven to ten local scholar- ships, depending on the field of study. Average scholarship grants for study in the US or the UK are about M$30,000 and M$35,000 per year, respectively. 3.106 The total amount of resources devoted by Government for tertiary education is substantial. In 1987, recurrent expenditures of government institutions of higher education totalled M$653.2 million compared to the M$4.04 billion budget of the entire Ministry of Education (Table 3.24). Expenditure for education abroad was M$1.177 billion in 1985 up from M$1.1 billion in 1985. Much of this was due to the bonded scholarship program of the federal government and other statutory bodies, which support both local and foreign studies. Some costs are due to persons studying abroad on their own because they could not enter local universities. In 1987, the Public - 170 - Service Department (PSD) alone budgeted M$460 million to support 30,000 students. MARA for the same year allocated M$174.6, down from M$192.6 million during the previous year. Table 3.24: BUDGET FOR RECURRENT EXPENDITURES: HIGHER EDUCATION VOCATIONAL/TECHNICAL, AND MOE TOTAL (in M$) 1985 1986 1987 1988 Ministry of Education (MOE) 3,768 4,046 4,044 4,037 (including higher education) Higher education 570.2 638.6 653.2 665.5 Vocational/technical 72.6 80.0 84.1 85.1 Source: EPU. Ministry of Education. 3.107 The cost recovery rate in local public institutions of higher education is quite low. While some fees are charged, they are relatively small. Fees are also not generally differentiated to reflect the large variations in unit cost across fields of studies and excess demand conditions. Thus, while unit recurrent costs by courses range from M$5,000 to M$30,000, fees are M$600 for all courses except medical and some other applied sciences, for which the charge is M$700. 3.108 Scope for improving the rate of return to higher education in Malaysia is indicated by a comparison with the Philippine experience. Estimates of the social rate of return for tertiary education in the Philippines show that in 1985, the internal rate of return was 11.3% (social) and 11.7% (private). Malaysia would be expected to have a higher social IRR because it has a much more developed and dynamic economy, and has relatively fewer graduates from local colleges and universities. Furthermore, while estimates for the Philippines were obtained when its economy was depressed, the Malaysian data reflect a tight labor market with rapidly rising wages. 3.109 The different IRRs can be partly explained by the fact that the Philippines depends on the private sector more heavily in providing tertiary education, and its cost is largely borne by the student and his family. The Malaysian situation is exactly the opposite. The Government's direct provision of local higher education is predominant, with very little partici- pation by the private sector. Malaysia also invested heavily in foreign edu- cation while the Philippines did not. - 171 - 3.110 Recently, due to the budgetary crunch, the Malaysian Government shifted its financial aid program from scholarships to student loans, with 80Z of financial aid to be in the form of loans and 20Z as scholarship grants. The student loans, which are still highly subsidized, appears attractive. There is no bonding requirement and loans are interest-free. Depending on the field of study and, therefore, on the amount of the loan, the payback period is anywhere from five to fifteen years. The borrower pays the total amount borrowed, except if he obtains a cumulative grade point average of 3 for professional and 3.5 for nonprofessional courses or better. In that case, he pays back only 25Z of the loan. The attractiveness of the loans is indicated by the experience of PSD in 1987, when it received 15,000 applications for the 5,000 student loans it made available. 3.111 While the policy shift is a move in the right direction, it must be complemented by other institutional reforms. First, the administration and planning of the financial assistance program for higher education remains very fragmented. Second, inasmuch as the clientele of the financial assistance program will largely work for the private sector (unlike in the past when they were intended for the public sector), the role of PSD has become obsolete and inappropriate. Third, as time passes, the problem of collection of loan repayments will become more prominent. Enforcement of loan repayments will be important. On one hand, given the generous conditions of the loans, repayment is necessary for the maintenance of the revolving fund. On the other hand, if the chances of having to pay for the loan are very small, then the incentive to use it efficiently declines. Hence, an institution is needed with a capacity to handle this activity. Training and Public Sector Programs 3.112 Government allocates substantial amounts for industrial and vocational training programs. During 1981-85, about M$680 million (141 of the total allocation for education and training) was spent on these programs. In 1986- 90, the expected allocation is about M$1,196 million, about 211 of the total allocations for education and training. Public training institutions are expected to produce 144,000 trained persons. 3.113 The efficiency of local training institutions has been seriously questioned with skill mismatches seen as a major problem. For example, in Table 3.25 which presents the distribution of persons trained by public industrial training institutions by status of employment, 781 of trainees who graduated in 1984 were either unemployed or employed in a job that is unrelated to their training (up from 501 in 1981). At the same time, the MIDA Industrial Trends Survey for the same period indicates that over 60% of companies perceived the lack of skilled labor as a factor limiting their output. The 1982 Tracer Study undertaken by the MO1 found that in 1982, a year of relatively good employment prospects, about 66 percent of the respondents who graduated from public sector training institutes were either unemployed or engaged in jobs not related to their training. Data from an in- house survey by the MOE Vocational/Technical Division also indicate that of the 1981-83 secondary vocational graduates interviewed, 33% of those with wage jobs are not employed in areas for which they have been trained, 36% are unemployed, and the rest are roughly evenly split between self-employment and - 172 - further study. These data may not be ty pical of the secondary vocational education system in as much as they include only 8 out of 25 schools. 3.114 Training at public training institutions is generally provided free of charge. In fact, substantial allowances are given to trainees during their training period in some programs. For example, M$1,530 is provided as pocket money for a year in one of the Ministry of Labor industrial training programs. The cost of training per student per year is therefore substantial, with the major regular programs of MOL industrial training institutes costing as much as M$4,000 - 5,000, including both capital and recurrent expendi- tures. Informed sources feel that because training is provided free of charge, excess demand and wastage are high. In one program, because of the provision of pocket allowance, 15,000 applied for the few hundred positions available. At the same time, lack of motivation to learn is recognized as a problem. The current failure rate remains high at 30 percent, although it has declined from 60 percent in 1986 as a result of an increase in entry qualifications in some courses. Table 3.25: PERCENTAGE DISTRIBUTION OF RESPONDENTS (INDUSTRIAL TRAINING INSTITUTES) BY STATUS OF EMPLOYMENT, 1981-84 Graduates in year 1981 1982 1983 1984 Employed (related to field training) 49 35 30 22 Employed (unrelated to field training) 19 15 16 15 Unemployed 31 50 54 63 Total respondents 100 100 100 100 (180) (338) (292) (648) Sources: (1) Kajian Mengesan Lepasan Institut Latihan Perindustrian, 1981-83 (October 1986). (2) Kajian Mengesan Lepasan Institut Latihan Perindustrian, 1984 (September, 1987) (3) EPU. - 173 - K. Policy Recommendations Policy Framework 3.115 Macroeconomic reflation in the form of an expansion of government expenditure and activity is unlikely to be an appropriate policy for addressing the employment issue. With the labor market well on its way to making the necessary adjustments to balance wages and productivity, conti- nuation of the policy to consolidate the gains from economic recovery is a reasonable policy to adopt. Reflation of the economy through some kind of macroeconomic fiscal policy to create jobs may be counterproducive by sending the wrong signals on the need to continue the process of adjustments in wage determination and job choice. It will also artificially tighten the labor market at the expense of the manufacturing and other private productive sectors. 3.116 Malaysia is not a labor surplus economy with a near zero opportunity cost of labor. It is, therefore, important that employment policy initiatives should be looked at in the context of the economy as a whole. At the very least, government measures to alleviate unemployment must be linked to productivity. Generation of low productivity jobs will not be sustainable, and will reduce overall productivity and hurt prospects of more rapid economic growth in the long term. Hence, public programs to create employment should be carefully assessed on the merit of their overall social profitability and not merely on their employment impact. In this regard, care should be exercised by the Government in implementing its Special Housing Program to stimulate employment and the job subsidy program for unemployed graduates. 3.117 While indications of wage reduction are welcomed as signs of market flexibility and movement towards stronger international competitiveness, lack of a reasonable growth of wages in the future would conflict with the social desire to raise labor earnings, particularly among poor, low-skill workers. Adjustments to lower wages, though necessary, is a bitter pill. There is thus a need for a return of the economy to a reasonable wage growth regime in the medium term. To be sure, increases in wages have to be in step with produc- tivity growth since rising wages without corresponding productivity growth could easily dissipate Malaysia's chance to be in a much stronger competitive position vis-a-vis other East Asian NICs. It can also lead to higher unemployment. 3.118 The problem is, however, just this: labor productivity is falling. Unless this is turned around, the real wage will also decline in the long run. The emerging policy issue from this broader perspective, then, is how to keep unemployment low and maintain international competitiveness, even as the economy must in the future be returned to a regime of rising labor earnings. The resolution of this issue will clearly require a combination of policies to improve allocative efficiency and total factor productivity growth for the whole economy and the manufacturing sector in particular. In essence, however, it means a strategy of efficient employment growth and development of human capital. The key word, it should be stressed, is efficiency in the employment and development of human resources. - 174 - 3.119 Specifically, the strategy for efficient growth of employment, productivity, and labor earnings should in principle involve three important elements: (a) a movement towards a neutral factor pricing policy; (b) improve- ment of the efficiency of human capital investments; and (c) enhancement of labor market flexibility along with improvement in risk-sharing between workers and firms. 3.120 Various laws currently tend to subsidize return to capital and make labor more expensive in Malaysia. Hence, movement towards factor pricing neutrality is important to improve both employment and efficiency. The EPF levy and provisions in the Employment Act regarding overtime pay are examples of such laws as are the investment incentives laws. These laws are now being reviewed by a cabinet Committee under the Deputy Prime Minister as part of Government's response to the employment problem. In addition, the Government is rightly re-examining the national masterplans for industry and agriculture which, to increase output per worker, call for a strategy of reduced labor use and greater dependence on physical and hu,a- capital-intensive technology. Regardless of the suitability of the plans in the past, the underlying strategy is now questionable given current labor market conditions. 3.121 The second element of the proposed strategy calls for reforms in the delivery and orientation of public training programs and government financial assistance for manpower development. If training is to play a significant role in stimulating productivity, it must be relevant to and focused on the firm since national productivity and competitiveness are dependent upon the decisions of tens of thousands of individual firms. However, the traditional public sector training system in Malaysia (as in many countries) is poorly suited to this task. The majority of public training institutions have no formal linkages with the private sector or industry, and there is very little consultation taking place with industry when specific training programs are formulated. 3.122 Malaysian economic planners now widely accept that reform of government training is needed before it can play a significant economic role in improving productivity and competitiveness. In this regard, senior government officials have articulated the desirability of greater private sector anticipation, accountability, and sensitivity to labor market conditions. Recently, institutional arrangements for the government training system have increased private sector involvement in the planning and implementation of training programs and policies. The government has approved the formation of the Natior.al Vocational Training Council (NVTC) through he reorganization of NITTCB. All public sector training institutions will come under the purview of NVTC. With representation in the council, in the Trade Advisory Committees and in the Standards and Certification Committee, the private sector together with government will be involved in the planning and development of training activities, including the determination of training needs, trade standards, and the structure of the curriculum and syllabus. 3.123 While important, mere representation of the private sector in these bodies is not sufficient to improve the linkage between firms and publicly- supported training. The views of the private sector representatives selected may not represent the variety of views in the industry. The representatives - 175 - may also not be conversant with the training issues pertaining to the needs and conditions of thousands of firms and workers. Hence, while there is appreciation in Government for greater market orientation and private sector participation, the idea of improving the external and internal efficiency of publicly-supported training programs needs a more effective operational strategy* 3.124 With regard to higher education, there is need to strengthen the link between its beneficiaries and the financial burden, without undermining government objectives for economically disadvantaged, yet deserving, students. This is important not just for mobilizing resources but, equally important, for reasons of improving efficiency and effectiveness. While the shift towards student loans is well advised, other measures need to be under- taken to improve the effectiveness and efficiency of the financial assistance program since the scholarship and loan programs continue to be fragmented and issues remain regarding the appropriateness of existing institutional arrangements that were created when the financial assistance program was geared towards the development of high-level manpower for employment in the public sector and when collection of loan repayments was not an issue. But with the scaling down of the government sector and the shift towards student loans and a consequent need for efficient collection of repayments, the fragmentation and appropriateness of current institutional arrangements for manpower development must be seriously reconsidered. 3.125 The third element of the proposed strategy looks to the future and further elaborates on the desirability of a manpower development policy that is focused on the firm and its workers. Even though retrenchment of workers as bottomed out and nominal wages have declined to some extent, it is nevertheless desirable to develop greater flexibility in the system. Malaysia is a dynamic, open economy that is bound to be vulnerable in the future to demand shocks and technological changes. Many jobs will need to be reconfigured, if not eliminated. In this regard, job security and flexibility could well be served simultaneously, if firms and workers can be encouraged to accept labor policies that emphasize continuous on-the-job training and profit-sharing. 3.126 The rationale for this approach depends on two concepts associated with the development process in Japan. The first concept is the importance of linkages among the turnover rate, continuous on-the-job training, adaptability to technological and demand changes, and productivity growth. Data reveal that workers from firms with continuous on-the-job training have much lower turnover rates and a steeper growth of earnings with respect to education, experience and tenure. Such firms are also associated with more rapid tech- nological change and total factor productivity growth. The relationships among these variables are simultaneous in nature. For example, while a low turnover rate makes it more profitable for firms to finance on-the-job train- ing, greater firm-specific training also reduces the incentive for companies to lay off workers whose training they have paid for. It is not clear what factors lead some firms (and not others) to prefer labor policies that emphasize continuous training and more permanent employment. Nevertheless, from the policy standpoint, the operational issue is simply to create efficient incentive structures that would induce management and workers to prefer such labor policies. - 176 - 3.127 The Japanese experience also indicates the inability of the traditional tools of macroeconomic policy to address employment issues. Keeping the unemployment rate low through manipulation of fiscal and monetary aggregates is achievable only at the cost of rising inflation. It has been cogently argued that the macroeconomic policy dilemma ultimately goes back to the wage system of paying labor. Hence, it may be better addressed by focusing on the microeconomics of employment, in particular on the wider use of profit-sharing or a similar approach to labor compensation. Recommendations 3.128 In light of the foregoing, the following policy measures are suggested: (a) Exemption from EPF contributions for workers less than 25 years old. This measure, which is a move towards factor price neutrality, is expected to increase demand for young workers who suffer most from the negative effect of the EPF levy. The proposed amendment should not be viewed as a countercyclical macroeconomic policy measure but as a means of reducing the two-tier labor market. With the basic wage not expected to increase in response to the EPF amendment (an unlikely event considering the looseness of the labor market at present), the proposed measure could raise demand for young workers by 6-11%, assuming a wage elasticity of demand in the neighborhood of minus one. A concern about this policy suggestion is the substitution effect between young and older workers. Information on this issue is not available at present. Nonrtheless, total demand for workers is likely to increase due to the availability of lower cost labor. A study of the issue is desirable and given the function of labor market and the incipient economic recovery, the impact of the policy is not to raise the unemployment of the older workers; it is rather expected to narrow the wage rate betwen the young and old. Hence, the policy appropriate in that it helps correct the distortion revealed by the development of a two-tiered labor market. (b) A mandate for EPF to use its funds and institutional capability to promote workers' training and education. This approach would allow EPF members to borrow for skills development, up to a certain fraction of their individual accumulated EPF credits. A system similar to the one proposed are already in use and the Philippines. Due to its higher level of development and different cultural milieu, however, Malaysia needs to carefully develop its own implementation strategy. (c) An EPF tax deduction to firms for 50% of the firm's cost of training a worker. In implementing this proposal, three provisions are recommended: (i) the total amount of deductions should not exceed 1% of the worker's basic wage; (ii) if the worker is laid off, EPF should be reimbursed by the firm for an amount to be credited to the worker's EPF account. (The amount should be proportional to the tax deduction availed by the firm and the time lapsed between the date of training and date of retrenchment); and (iii) firms avail- ing of these deductions should have a labor-management committee to oversee the quality of training provided to the workers. The committee would he the conduit of workers' suggestions and complaints about training matters; consequently, it should be given powers commensurate with its responsibility - 177 - to protect workers' interests. The type of costs to be deducted from the EPF tax should include the opportunity cost of attending training sessions. This should make the scheme more effective. It has been noted that time release allowing workers to attend training is common among enterprises. In general, the detailed specification of what costs can be deducted can begin with those listed in the Guidelines for Double Deduction Incentive for Approved Train- ing. The objective of the proposed measure is to promote continuous on- the-job training, and to the extent that it reduces unit labor costs, it should also help to correct whatever factor price distortion may have been created by the EPF payroll tax on firms. This scheme should improve the firms' ability to adjust to changes in demand and technology. Eventually, it could contribute significantly towards faster factor productivity growth and a lower long-run unemployment rate. An alternative approach to encourage enterprise-based training is the levy/grant system, which is common in Latin America. In this system, a 1% to 2Z payroll tax is levied by firms with employment above a certain size. These taxes are used by a national training authority to fund training programs. The authority is often governed by a tripartite board, consisting of representatives from labor unions, government and management. Firms which want to train their workers must submit their proposals to a highly central- ized bureaucracy under the national training authority. The experience with this model and similar highly bureaucratized systems is in general not very satisfactory. First, especially for small enterprises, the cost of applying for a training grant could be prohibitive. Second, levy/grant schemes according to Grabe (1981) tend to institutionalize training and counteract efforts to encourage training on the job. In this regard, Crabe points to the tendency of the national training bodies established in conjunction with the levy system to become so bureaucratized that the number of administrators sometimes exceeds the number of teachers and instructors. This is not to suggest that the levy/grant system should be completely dismissed. In fact, it may be useful to examine the experience of Singapore's Skill Development Fund Board and determine its suitability for Malaysia. The Malaysian Government has recently approved a double deduction tax incentive to encourage private sector commitment and active involvement in manpower development for the manufacturing sector. In this scheme the Ministry of Labor must approve the training program to be eligible for double deduction. Furthermore, the deductions are to be made from the firm's income tax. The effectiveness of the scheme is highly doubtful, since there are so many tax incentives currently available (many of them in favor of capital) that for many firms double deduction of training expenses from the firms' income tax is redundant. The experience in the past regarding the labor util- ization credit as an investment incentive favoring employment should be indicative of the limitation of the double deduction incentive scheme. There were very few takers. Consequently, the labor utilization credit was abol- ished. - 178 - (d) Liberalization of the legal provision for overtime pay. Amendment of the Employment Act is proposed to allow firms to offer a system of profit- based overtime premium pay in lieu of the legally imposed rate, provided that, unless such an agreement is reached by the worker and management, the present provision of the law applies. The proposal adds flexibility to labor market contracts and permits greater use of a productivity-inducing overtime premium that may be beneficial to both labor and management. (e) Development of a performance-based contractual system of implement- ing government training objectives. This proposal would let public training institutions compete for clients among themselves and private training enterprises. The increased demand for training expected from EPF involvement in training finance, if approved, should make implementation timely. Operationally, this approach would deliver publicly funded training through a system of performance-bared contracts, whenever feasible. In this regard, it is important that fundi.g decisions and the selection of those to be trained is separated from the government training delivery system itself. An example of such an approach is the system followed by the California Employment Training Panel (CETP). Started about five years ago, CETP has an appointed council (Panel) to administer funds transferred from the state unemployment insurance system for job training in California. The Panel was conceived as an experiment to determine whether public training funds could be used effectively to improve economic efficiency and ease the dislocations brought about by accelerating economic change. The Panel, which does not train workers itself, took the role of an agency that contracts with a variety of delivery systems. It uses traditional government and business/labor training programs, and contracts with public and private schools. Direct contracts with employers were also negotiated, with the selection of the delivery system left to the choice of the firm that would be employing the trainees. The Panel pays a training contractor on the basis of the number of people trained, hired and retained (for at least 90 days after the end of training) on the job for which they were trained. A fixed fee is negotiated for every trainee, and payment is made only when all three criteria are satisfied. The effectiveness of the system was found to be high. The unemployed experienced a substantial increase in earnings one year after undergoing a training program funded by the Panel. Significant wage increases were also observed for trainees who were in danger of losing their jobs. Finally, the average incidence of unemployment fell dramatically for those without employment at the start of training and for those with jobs but were in danger of being laid off. It is not suggested here thqt Malaysia should copy the CETP. Due to institutional differences, it must develop and test its own performance-based contractual system for implementing government training objectives. Consequently, resarch and pilot programs on this issue should be given high priority. (f) Reorgnization of the Government's financial assistance program for studies in higher education. There is a need to consolidate the implemenation - 179 - of the fragmented student assistance programs in an institution that has a comparative advantage in running it and whose function revolves around the workers' earnings and their future incomes. In this regard, the EPF is the most natural institution on which to base the Government's financial assistance program. EPF's functions revolve around the present and future incomes of workers, including those who would have borrowed for their education and training. With an expanded mandate to invest in the education and training of workers (including young members of their family), a more central role for EPF would become even more compelling. First, it would have a better administrative capacity to enforce and collect loan repayments, given that it is now hooked up to a payroll deduction system that encompasses the formal labor market. Second, it has a data base which it can develop to provide regular information about wage trends and differentials by occupation and education, which are important for market-oriented manpower planning. Finally, with EPF making its own funds available for the training of its members and the young members of their families (assuming the previous proposals are approved), better coordination would be necessary to avoid excessive debt burden and default. This organizational reform can be more easily achieved through EPF. Under this institutional arrangement, MARA's objectives of providing financial assistance to the Bumiputeras can also be served and even strengthened. If so desired, there is nothing to prevent Government from providing an allocation rule based on F. Under this institutional arrangement, MARA's objectives of providing financial assistance to the Bumiputeras can also be served and even strengthened. If so desired, there is nothing to prevent Government from providing an allocation rule based on ethnicity. I