Report No. 7208-MA Malaysia: Matching Risks and Rewards in a Mixed Economy (In Three Volumes) Volume 1: Main Report October7,1988 Country Operations Division Country Department 11 Asia Reion FOR OFFICIAL USE ONLY Document of the World Bank This report has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY 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 = 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 cDP = Gross Domestic Product GNP = Gross National Product HICOM = 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 = Malaysian Industrial Policy Studies MOF Ministry of Finance MPE = Ministry of Public NEP New Economic Policies NFPE 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 = Spacial Low Cost Housing Program VAT = Value Added Tax FOR OMCIL USE ONLY MALAYSIA Volume I: Main Report Table of Contents Page No. EXECUTIVE SUMR ....... .....................i Io INTRODUCTION*ooo.ooo*eoo*oo* *.*. *.*a* 1 II. RECENT ECONOMIC DEVELOPMENTS, 1981-87 ........................ 3 A. The Expansion of 1981-843.............. ................. .. 3 Fiscal Expansion ........ 6 The Construction Boom........................ .......... 9 Loss of ............................ 10 B. The 1985-86 Adjustment Period............................ 15 External Shocks ......... 16 Internal Shocks ............... *9e 19 Fiscal Turnaround ......................... 21 C. The Beginning of Recovery--1987..........................o 23 Factor Market Adjustmentsj u s t m e n tso.................... 26 D. The Emergence of Structural Weaknessesk.................. 26 Long-run Constraints on Public Deficits. cits.....**. 27 Rising Wages and Falling Profitso...... 000000004-00*0 32 The Mismatch between Risks and Rewardswards......**. 35 Struccural Unemployment.9..... .. . ... ............. 42 E. Overall Assessment... 46 III. MEDIUM-TERM GROWTH PROSPECTSO............. 50 A. Growth Potential--The Supply Side..... 51 Efficiency and Productivity Growth. o w th............. 52 Policies to Improve Productivitytuctivi...ty....... ... 60 Public Enterprises.t.e.., rs....s... 60 Tariff Reformf.o.*** .................. 66 Industrial Licensing and Restructuring.............. 67 Corporate Income Tax Reoo r m 71 Investment Tee n ds 74 Oil vs. Non-oil Private Investment.................. 76 Investment by Ind u s tr.y...o ... ...... ........ 76 Policies to Encourage Private Investment.. * * 80 Structural Polces.. i c ie...s.....o 81 Foreign Direct Investment...vse......... t...... ..... 89 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contants may not otherwise be disclosed without World Bank authorization. -ii - Page No. B. Growth Outlook--the Demand Site.......................... 91 Private Consumption.................................... 92 Government Spending .............................* 94 Off-budget Government Programs...... .................. 101 Overall Impact............................. 102 IV. ECON)MIC OUTLOOK AND INTEGRATED POLICY RECOMHENDATIONS ....... 103 A. Economic Outlook... **** ** **........ ......**.. 103 The External Environment...... ......... ............... 103 Prospects for Exports and the Balance of Payments ...... Ill The Domestic Environment........e...... ............ 113 Major Uncertainties. ................... ................. 116 B. Integrated Policy Recommendations ....................... 116 Consistency in Policy Programs. ........................ 118 Business Confidence.*.*.. ... .......... . 119 Amendments to the EPF.................................. 119 Public Financesu.................................... 121 Conclusion ........ > ~~~~~122 Appendix 1 - Assessing the Viability of Non-oil Primary Commodity Exports ...... 124 Appendix 2 - Construction of the Flow of Funds Table................ 126 Appendix 3 - EPF: Life-Time Accumulatio- by the Individual ........ 129 The Base Case........................................ 129 Appendix 4 - Methodology for Calculating Total Factor Productivity Growth ................................ 136 Notes on Data Used ................................... 137 Reslt ............................................... 137 Appendix 5 - Methodology Underlying the Calculation of the Marginal Effective Tax Rate (METR) ................... 140 Note on Assumed Project ............................. . 142 Appendix 6 - The Determinants of Private Fixed Capital Formation.... 143 Construction of the Data ............................. 143 Appendix 7 - Private Consumption in Malaysia***...................... 145 TABLES 2.1 Key Macroeconomic Developments, '981-87 2.2 Composition of Change in Demand 2.3 Consolidated Government Accounts 2.4 The Composition of Public Investment, 1978-87 2.5 Selected Statistics in Property Markets, 1981-87 2.6 Manufacturing Unit Labor Costs in East Asia 2.7 Major Commodity Export Prices, 1981-87 2.8 The Structure of Exports, 1981-87 2.9 Adjustment to Shocks, 1984-87 2.10 Sustainable Long-Run Public Sector Deficits 2.11 Flow of Funds: Overview, 1980 - iii - 2.12 Flow of Funds: Overview, 1985 2.13 Flow of Funds: Overview, 1986 2.14 Nominal Rate of Return by Institution, 1981-87 3.1 Labor Productivity Growth 3.2 Size Structure of Manufacturing Firms 3.3 Intercountry Comparison on the Role of SHEs 3.4 Comparative Profitability Ratios in Public and Private Firms 3.5 Average Tariff Levels 3.6 Marginal Effective Tax Rates in Selected East Asian Countries 3.7 Local and International Prices of Steel Bars 3.8 Price Trends of Industrial Properties and Land in Selangor State 3.9 Difficulties in Borrowing from Banks 3.10 The CIC-Guided P/E Ratio 3.11 Capital Gains for New Issues, 1981 3.12 Estimation of Actual Equity Generated in Industrial Activities 3.13 Inflow of Foreign Direct Investment 3.14 Japan's Direct Investment to Asian Countries 4.1 External balances of East Asian Countries 4.2 Non-Tariff Barriers Against Imports from Selected Asean Countries 4.3 Non-Tariff Barriers Protectionism in Selected Commodities 4.4 Malaysia - Balance of Payments 4.5 National Accounts in Constant 1978 Prices 4.6 National Accounts Growth Rates at Constant Prices FIGURES 2.1 Real Wage and Labor Productivity 2.2a Korea: International Competitiveness 2.2b Malaysia: Real Exchange Rate Movements 2.3 Net Fiscal Impulse, 1972-87 2.4 Holdings of Federal Government Debt 1987 2.5 International Comparison of Public Debt 2.6 Decomposition of GNP by Income Sources, 1972-87 3.1 Efficiency in Resource Use in Malaysian Manufacturing Industries 3.2 Composition of Capital Formation 3.3 Personal Income and Consumption 3.4 Total Expenditure, Subsidies and Transfers HAP IBRD Map No. i- , R OOREWORD The preparation of this report benefitted greatly from the assist- ance and cooperation provided by the Government of Malaysia. In particular, the author would like to thank officials in EPU, the Ministry of Finance, the Ministry of Trade and Industry, MIDA, EPF, CICU, Ministry of Public Enter- prises, Bank Negara, the Ministry of Labor, and the Department of Statistics for their generous help in providing data and explanations to the mission. The report was prepared by a World Bank mission which visited Malaysia from November 23-December 14, 1987. The mission members were: Homi Kharas (team leader); Ken Ohashi (finance); Vic Paqueo (employment); Gerardo Sicat (public finance); Yoon Je Cho (private investment), all from the World Bank, and Hafiz Shaikh (consultant, nonfinancial public enterprises). The report was discussed with Malaysian government officials in August 1988, and subsequently updated. The views expressed, however, are those of the authors and are not representative of the positions of the Malaysian government. EXECUTIVE SUMMARY A recovery in 1987 has led Malaysia out of its worst growth period since independence. Along with growth, a record surplus in the current account of the balance of payments and negligible inflation are concrete demonstrations of the successful macroeconomic adjustment undertaken since 1982. At that time, the country was beset with serious macroeconomic imbalan- ces, notably a public sector deficit of 191 of GNP and a deficit on the cur- rent account of the balance of payments of 141. External borrowing and debt were accelerating, and the global environment of high interest rates, depressed commodity prices and slow trade growth was unpromising. Government moved forcefully to address its macroeconomic problems and, although the over- all public sector deficit and total public debt remain high today, the degree of adjustment achieved in the face of unfavorable external conditions is a striking tribute to pragmatic policy making, and has ensured that Malaysia remains in the ranks of the most creditworthy developing countries. Strong growth in 1988, estimated at about 6-7Z, serves to reinforce the impression of successful macroeconomic performance and policy-making. Although encouraging, recent performance should not be interpreted as grounds for complacency nor as a signal that the need for adjustment is over. A clear distinction must be drawn between cyclical and structural issues; much of the recent gains can be attributed to cyclical, external developments--most importantly, improved demand and prices for commodities and a boom in the global electronics industry. Yet the external environment can easily sour, and the experience of the setbacks to the New Economic Policy (NEP) targets of poverty eradication and restructuring during the 1985-86 recession shows that reliance on aggregate demand management alone cannot ensure the achievement of a high, sustainable growth rate. Structural adjustment must therefore be continued. The challenge for policy makers is to manage interventions in the economy to achieve restructuring objectives in the context of a turbulent world environment, while taking advantage of the efficiency gains to be obtained from liberalization of markets. In addition to the challenge of meeting NEP targets, there are two economic challenges that merit urgent attention. First, despite the signifi- cant progress made in reducing the public sector deficit since 1983, the level of the deficit as a percent of GNP remains cause for concern. The recent increase in the deficit can be traced to lost revenues following the recession and the general decline in tax buoyancy from falling commodity prices and profits. The collapse in oil prices has been especially significant in this regard. The policy challenge in this area is to reestablish a growth- orientation to fiscal policy without increasing the size of the overall deficit: increasing revenues in a non-distortionary manner, improvement of the efficiency of nonfinancial public enterprises (NFPEs) and the gradual elimination of other implicit and explicit subsidies and transfers that add little to growth, are examples of how this can be achieved. Second, factor market distortions and other administrative and institutional constraints have led to a slump in private investment growth and the emergence of sizeable unemployment. During 1988, both these indicators - ii - have improved, but the levels of investment and unemployment are still far from those prevailing in the healthy growth period of the 1970s. The recovery has also underscored the lessons of the recession--that growth depends impor- tantly on maintaining the competitiveness of manufactured exports and on support of business confidence through relaxation of industrial licensing practices and other rules of compliance. The liberalization measures and new incentives introduced since 1986 have met with considerable success in attracting foreign investment and boosting export-oriented business. However, more can be done to broaden the base of investment funds, encourage new entrepreneurs and spur private sector initiatives to boost aggregate domestic demand further over the medium term. The policy challenge in this area is to fully unlock the gains of past reform measures through a further rationaliza- tion of measures that remove constraints and enhance competitiveness and business confidence. Recent Macroeconomic Developments Since 1980, Malaysia's macroeconomic performance has been irregular, with growth, the current account, investment, unemployment and the public sec- tor deficit all exhibiting sharp fluctuations. Following the oil price rise in 1979/80, government embarked on an ambitious expansion of development proj- ects, driving total eapital formation to a peak of 38Z of GNP. Although this protected the economy against the global recession in 1981-82, it led to unbalanced growth and the emergence of high fiscal and current account defi- cits. Since then, the Government's policy dilemma has been how to cure exter- nal and internal financial imbalances without derailing growth. Expansionary fiscal policy, coupled with a property boom, sustained growth through 1984. Rising private savings, in part due to the increased coverage of the Employee Provident Fund and in part to high real interest rates, helped bring the current account deficit down to more manageable pro- portions. But other pressures were accumulating: real wages rose faster than productivity as a result of the tightness in the labor market generated by fiscal expansion and the construction boom; the real exchange rate appreciated in line with the US dollar, eroding the competitiveness of manufacturing relative to other East Asian exporters by as much as 50Z; property prices had soared out of line with household incomes; and total public debt was approach- ing the same level as GNP. The recession in 1985 and 1986 resulted from a combination of external and internal factors. Terms of trade losses threatened both the cur- rent account and public finances. When government responded by first increas- ing public savings and then cutting investment, it withdrew the demand stimu- lus that had buoyed the economy in the early 1980s. At the same time, the property boom collapsed and exports slowed. In response, monetary policy was eased and the ringgit depreciated. Nevertheless, private investment declined as a result of worsening profitability, sustained momentum in wage growth, high levels of government investment in directly competitive areas, and a drop in foreign investment; and unemployment started to accelerate as labor absorp- tion in the public sector and in construction slowed. The result was a highly successful adjustment in public finances and in the balance of payments, but at the inevitable cost of growth. - iii - The recovery in 1987 has reversed the deterioration in economic performance. Growth was led by manufacturing (13%) and agriculture, forestry and fisheries (7.5X); within manufacturing, the largest contribution to growth was in the export-oriented sector of electronics, textiles and apparel, and wood-based products, and within the primary sector, sawn timber and log production and cocoa were largely responsible for the good growth perfor- mance. Amongst other key commodities, rubber output was up, but palm oil production declined. The service sector also grew, with trade, transport and storage responding favorably to commodity price improvements. Only construc- tion continued to slide, by 6%. Aggregate domestic demand, particularly private sector demand, has been sluggish and remains, in real terms, below the level prevailing at the start of the Fourth Malaysia Plan (1981-85). Low demand, rather than robust growth, is also primarily responsible for the current account surplus. Private non-oil and gas investme.nt, for example, has fallen below 10 of CNP. Initial indications are that growth in 1988 will accelerate to 6-7%. The leading sector is manufacturing which continues to enjoy strong export demand growth. But the base of manufacturing expansion is expected to broaden into resource-based industries, such as food, oils, beverages and tobacco, domestic consumer-oriented products, notably cars, and construction materials, such as cement. On the demand side, public and private investment are expected to lead the recovery and, although exports will grow rapdily, imports will grow even more quickly, negating any net demand stimulus from international trade. Overall, the large trade surplus achieved in 1987 should be maintained, permitting a continuation of the policy of prepaying external debt. The main cause for concern in the short-run, then, is the state of public finances, which will once again suffer from falling oil prices--if budgeted investments are actually made, the deficit could again exceed 8% of CNP. To meet the challenge of converting recent short-term growth sutcess into sustainable growth, policymakers must address further structural impedi- ments to growth which have been deferred, and on occasion accentuated, during the macroeconomic adjustment program of the last few years. For example, the capital-intensity of manufacturing has grown and the demand for labor has shrunk since 1980, partly as a result of programs to reduce the burden of corporate income taxes, designed to encourage investment, and to increase Employee Provident Fund contributions, designed to boost savings. For example, there has been a bias in factor pricing, resulting from capital- cheapening corporate income tax provisions and high payroll taxes on labor designed to boost private investment and savings. As a result, employment growth in the 1980s in manufacturing has only been 1.3% per year while value added has grown at 6%. Similarly, problems have emerged in human and physical capital accumulation due to rigidities and rationing in education and training systems and the expansion of non-financial public enterprises (NFPEs). Last, legitimate efforts to protect the economy against external shocks have contri- buted to an inequitable distribution of the burden: formal sector wage earners have been protected while corporations, new entrants and informal workers and the unemployed have been hurt; and savers in low risk schemes, including ASN and EPF, have enjoyed protected returns, while investors have suffered. In addition, use of public credit, employment generation schemes, electricity - iv - pricing discounts, tax incentives and other implicit subsidies have been used to protect selected groups but at the cost of larger fiscal deficits, which will impose a burden on the whole economy in the future. To rectify these problems and allow government to build on its successful adjustment program, reform in four major areas is proposed: (a) support for private investment; (b) reform of NFPEs; (c) stabilization of public deficits, debt and spending levels; and (d) employment and human resource development. For the most part, the recommendations made below with respect to these areas are justified on microeconomic efficiency grounds. Taken together, however, they would also help resolve important macroeconomic dislocations expressed in the falling share of profits in national income, in the sluggishness of private demand and in the rise in unemployment. As such, the recommendations should be viewed as a comprehensive package, consistent with desired macroeconomic trends and designed to remove impediments to sustained growth. Such an approach would complement the policy reforms which Government has recently initiated but which have yet to show significant results in broad-based growth, investment and employment generation because of unintended spill-over effects into other areas. Structural Issues and Prospects for Growth The best hope for medium-term growth lies in the manufacturing sector. A symbolic milestone in Malaysia's transition towards an industrial- izing nation w4s passed in 1987, when this sector overtook agriculture as the largest in the economy. The structure of the industrial base is potentially vulnerable however. The most rapidly growing segment, oriented towards exports, is narrowly concentrated in electronics and garments. With limited linkages to the rest of the economy, this sector is heavily dependent on competitive wages. In US dollar terms, however, wages have been volatile as a result of domestic labor market adjustments and exchange rate fluctuations. Furthermore, technological changes and competition from other developing country exporters of manufactures pose an increasing threat to Malaysia in these products. Another segment of manufacturing, oriented inwardly, has grown within a policy environment that features variable effective rates of protec- tion, easy access to inexpensive credit, handsome fiscal incentives and restrictions on new entrants caused by industrial licensing. In addition, Government has itself become a large direct investor in major industrial proj- ects. Although domestic industrial growth was also rapid in the early 1980s it was inefficient and costly. Capital and intermediate good inputs grew faster than output, and total factor productivity growth turned negative in the 1980s. Employment growth was also disproportionately low compared to output growth as the capital intensity of production increased, and unemployment has become a major social concern. Problems in the manufacturing sector were initially disguised by high levels of demand, temporarily created by a government spending and property boom in the early 1980s, but were exposed during the ensuing reces- sion. A sustained decline since 1979 in profitability caused by wage growth in excess of productivity and by high real interest rates contributed to a v steady drop in private non-oil investment. Government cutbacks in infrastruc- ture investment and expansion of directly productive investment have further deterred private investment. Finally, sharp increases in the costs of capital goods relative to other countries were caused by the property boom and by high regulated prices for steel and cement. The drop in private investment, which bottomed out in 1987, has occurred despite a steady rise in private savings and a resultant increase in liquidity and low interest rates in the banking sector. Thus, although private savings reached 30X of GNP in 1987, these were not transformed into productive, private investment largely because of a structural shift in the composition of savings. Since 1980, private savings have been increasingly generated by the household sector rather than by corporations, and are channelled into high return/low risk assets which finance government deficits. The redistribution of savings from corporations to households has also constrained private investment by introducing a mismatch between the risk profiles of savers and investors. This has occurred because a sizeable portion of household savings is intermediated by nonbank institutions, such as provident and mutual funds, insurance companies and cooperatives. These intermediaries are unwilling to bear the risk associated with lending for productive activities. They there- fore purchase government securities and other high quality assets and reduce the availability of equity and other risk-finance to the corporate sector. The development of a broader range of financial intermediaries and instru- ments, including well-capitalized venture capital funds, is required to channel private savings back towards domestic investors. More recently, the conditions for investment have improved markedly, as wages have fallen, profits grown and the real effective exchange rate depreciated. Government has also liberalized industrial licensing and regula- tions governing foreign direct investment, and reduced administrative delays in the issuance of licenses. The early estimates for 1988 show that these have reversed the downward trend in private investment. This momentum could be sustained over the medium tenm, if remaining constraints are removed and confidence in the economy strengthens. Such constraints include heavy corpo- rate contributions to the Employee Provident Fund (EPF), crowding out by public enterprise investment, the risk mismatch in the flow-of-funds, and the restrictive legal basis on which corporations must base overtime, bonus and other benefit payments to labor. A dominant feature of industrial growth over the last decade has been the emergence of the NFPE sector to center stage. In contrast to early growth in the sector based on acquisition of existing assets, recent growth has been driven by investment in new projects at a rate that has given Malaysia one of the largest NFPE sectors amongst non-socialist economies. The sector is dominated by a few large firms, of which some are extremely profita- ble (PETRONAS) and others big money losers (HICOM). Although these dominate the financial aggregates, beyond them is a complex of over 700 smaller compa- nies, who to all practical intents and purposes operate outside the purview of the Federal Government. These smaller companies are particularly inefficient and together represent a sizeable drain on overall public sector finances, -vi - amounting to over M$800 million per year out of a total of M$2 billion nega- tive net profits of all loss-making NFPEs in 1986. Although Government has acted swiftly to reduce the deficit brought about by the collapse in petroleum revenues and, earlier, by the sharp expan- sion in public investment, it has done so largely through cutbacks in real expenditure. Although this was a desirable course to take due to the short- term inelasticity of revenues, an overreliance on cuts in capital projects rather than in subsidies and transfers has proved problematic. In fact, development spending has been a blunt instrument for achieving short term fiscal policy goals, because of the lags involved in planning and implementing investment projects, and the lorng gestation of projects once initiated. During the 1980s, the effect of public investment policies was to lock in large public deficits, as spending (particularly by NFPEs) could not be halted until 1985 without leaving projects unfinished with wasted earlier expendi- ture, and to cause an overheating of the labor market in conjunction with an in-phase construction cycle. Similarly, although government initiated a Special Low Cost Housing Program in 1986 to combat the recession, the full effects of this are likely to be felt only now in 1988, when the recovery is already firmly in hand. Further initiatives by government to accelerate the recovery through additional spending should be avoided as these tend to amplify the natural business cycle on both the up-swing and the down-swing, to crowd-out private investment and to use capital resources sub-optimally. One consequence of high public investment and large overall deficits has been an exponential rise in total public internal and external debt. While the government has successfully managed external debt by limiting new borrowings, refinancing old, high interest rate debt, and prepaying other loans, thereby maintaining the debt service ratio at very comfortable levels, it has steadily built up internal debt to finance expenditures. When debt is expressed as a ratio to GNP, Malaysia has one of the highest levels of indebtedness of any country, developed or developing. The burden of servicing this debt is substantial (interest alone will likely amount to 10% of GNP by next year) and subject to sharp fluctuations in response to global changes in interest and exchange rates. Although this is not as yet a problem due to accass to new domestic resources from the EPF, over time a growing interest burden will steadily limit policymakers' room to maneuver. Moreover, as a result of these contractual obligations, Government is increasingly using revenues derived from petroleum to finance current expenditures rather than replacement assets. In addition to debt service, Government is faced with rising financial claims to support statutory bodies, other government agencies, explicit and implicit subsidies, pensions and the like. Overall, over 40 cents out of each dollar of public expenditure go towards financial transactions (including a build-up of assets in some statutory agencies) rather than towards current or development expenditures on real goods and services. In fact, the burden of Adjustment in public finances has been disproportionately great on real as opposed to financial transactions. The slowdown in the rate of expansion of the public sector has forced new entrants into the labor force, particularly college graduates, to look elsewhere for jobs. Other sectors, however, have not been creating jobs and, since 1984, wages for new entrants have fallen by perhaps one-half. - vii - Because wages for existing employees have been sticky downwards in nominal terms, and inflation has been minimal, a sizeable wedge has emerged between the wage rates for workers with similar experience and education. This wedge has persisted over time because of the high costs of firing and because of the unwillingness of some firms to move to a ts.-tier labor market with its atten- dant problems of low morale and productivity. Given the dispersion in wage rates, there are significant benefits to searching for a high paying job, especially for the young and relatively well-educated who do not need to work immediately. The unwillingness of this group to work for low wages and their preference to look for appropriate employment has contributed to a steady rise in unemployment, even in years of rapid output growth. Falling wages for new entrants have not been immediately reflected in declining average wages. Because of the nature of negotiated wage agreements and the slow pace at which low-paid workers have entered the labor force, the average wage has continued to rise in conjunction with unemployment. More recently, the dynamics of wage adjustment have shifted and average wages have started to decline. Along with expected output growth, this should lead to a fall in the unemployment rate. The lengthy process of labor market adjustment reflects a signifi- cant degree of short-term rigidity in existing labor contracts and the absence of direct linkage to productivity, price or profitability developments, except in the case of plantation workers. It also reflects rigidities in the ability of higher education, vocational and on-the-job training systems to respond to changes in the structure of demand for labor, due in part to heavy subsidies in these systems and the consequent wedge between social and private rates of return. Given this, Government efforts to reduce unemployment through Keynesian-type public spending programs are unlikely to succeed in creating an efficient deployment of human resources. These factors suggest that the challenges facing Government in terms of generating growth and employment over the medium-term are considerable. There are, however, two important advantages favoring the country. First, as both product and factor markets seem to operate quite well in Malaysia, albeit slowly, price adjustments may provide an effective mechanism for equilibration if left to operate in a stable macroeconomic environment. Already, wage adjustments and interest and exchange rate developments are helping restore profitability and slow the increase in unemployment. Second, Government has recently undertaken significant reforms in the system of industrial licensing designed to reduce the cost to firms of doing business. With these in place, the rewards to overcoming the remaining structural weaknesses should be considerable. Policy Objectives and Recommendations A coordinated strategy will be required to sustain growth over the medium-term. Government's broad objectives should be to increase efficiency in production, support private investment, remove the bias against labor and improve human resource development. In so doing, it must reduce the size of the government deficit to a sustainable level, and avoid the large fluctua- tions in spending that have characterized the recent past. Achievement of - viii - these objectives, combined with a continuation of prudent policies to avoid large macroeconomic imbalances, would create an environment in which a dynamic private sector can be expected to thrive. Policies to Improve Efficiency. Efficiency improvements can be expected from policy reforms which increase the competitiveness of the indus- trial sector and promote greater neutrality and transparency in the incentive system. Government is already moving towards these ends, with positive results reported by monitoring agencies. To strengthen this process, this report recommends consideration be given to the following measures: (a) early attention to privatization and/or cl ure of small and loss- making public enterprises; (b) a performance evaluation system to restructure the incentives facing NFPEs. Such a system would require enhanced monitoring and data collection, and should be implemented in a hierarchical fashion, with the federal Government monitoring major holding companies which would in turn monitor their subsidiaries; (c) an intensified public/private dialogue un industrial restructuring to focus on mechanisms for sharing risk, including promotion of long-term credit, venture capital funds and linkages with foreign investors; (d) avoidance of new tariff protection, except on a transitional basis, and, in the medium term, a reduction in the 210 protective tariffs with rates greater than 30Z to reduce the wide dispersion in protec- tion across sectors; (e) introduction of an exemption from the development tax for companies complying with NEP; while such a deduction already exists, its impact is muted by the existence of a broad array of other tax credits and deductions. These should be accredited only against income tax payments and not against the development tax. (f) reductions in corporate contributions to the EPF to minimize the bias towards capital intensity in production; and (g) the reform of the corporate income tax to reduce capital-cheapening tax holdings, depreciation and investment allowances but with compensation in the form of lower nominal statutory tax rates. Policies to Support Private Investment. Several of the measures to improve efficiency will also boost private investment by improving the confi- dence of the business community in government's commitment to private sector- led development. For example, lower corporate contributions to EPF will provide an important stimulus by increasing the share of profits in national income and restoring international competitiveness. In addition, Government should consider: - ix - (a) reorienting public expenditure towards maintenance and necessary infrastructure projects, while restraining directly productive investments; (b) reducing regulated prices of cement and steel closer to border prices, and liberalizing other administrative procedures that lead to inflated investment costs; (c) supporting the generation of equity finance by reform of the Kuala Lumpur Stock Exchange with a focus on regulations governing the determination of stock prices for new issues, take-overs and mergers and, in the medium-term, developing venture capital funds that could, in part, tap the long-run capital available in provident funds and insurance companies; and (d) encouraging the development of a more flexible wage compensation package by linking overtime and bonus payments to productivity or profitability. Policies to Support Employment and Human Resource Development. Reforms in this area should focus on removal of the bias present in the cur- rent system against labor intensive production, against the young, and against efficient human capital formation. Individual enterprises should be encouraged to provide greater on-the-job training, so that changes in the structure of output are accommodated by retraining of workers not retrench- ment. A closer partnership between the public and private sectors in shaping training courses should also be supported through a sharper delineation of their roles, with the private sector taking the lead in the delivery of training and the public sector taking the lead in establishing direct and indirect access to finance. In addition to reducing the corporate contribu- tion to the EPF and in legislating more flexible wage contracting, Government should consider: (a) a voluntary exemption from the EPF for these under 25 years of age; (b) improvements in the delivery of vocational and on-the-job training programs through establishment of appropriate incentives and compe- tition; (c) deductions from corporate EPF contributions for 50% of firms' costs of approved training; and (d) reorganization of the Government's financial assistance program for higher education through implementation of higher user-charges coupled with an enhanced student loan program. As the primary institution responsible for financial capital accumulation, the EPF could also become a central element in improving accumulation of human capital. It should mandate withdrawals and make loans for approved training programs with tangible connections to future jobs, while leaving the delivery of such programs to the private sector and corporations. It should also - x - implement a student loan program to cover higher educational fees. By bring- ing to bear detailed knowledge as to the market rates of return to education, available from its data base, EPF can assist in the process of making curric- ula and course development more in tune with market signals. Government should also develop a performance-based contractual system to implement its training objectives, whereby Ministry of Labor and EPF would become providers of training finance, while public and private courses would compete for students and charge user-fees to cover costs. The institu- tional responsibilities of these agencies would be to ensure that funding would only be released once a cycle of training, hiring and retention on the job was completed. Such programs have worked successfully elsewhere to allow firms and workers to adjust to demand and technological changes without uanec- essary retrenchment and to raise wages and employment prospects for the unemployed and those in danger of losing their jobs. Policies to Stabilize Public Debt, Deficits and Spending. As gross debt levels have reached high proportions by international standards, Malaysia should act to stabilize the debt by reducing overall public sector deficits. It should also avoid as far as possible sharp cycles in development spending occasioned by external shocks. Reductions in the deficit should be aimed at decreasing overall spending by reducing subsidies and transfers and by stricter control over development expenditures to weed out unproductive projects. This may involve a change in the composition of spending as, in some instances, reducing distortions may even involve higher expenditures. This is likely to be true in the deregulation of cement and steel prices and in the removal of overly restrictive budgetary constraints on maintenance. It would be useful, therefore, to undertake these reforms in tandem with a pro- gram of raising revenue, through broadening the sales tax base or introduction of a VAT, and cutting expenditures in other areas. In addition, where possi- ble Government should move off-budget the operations of departments and statu- tory bodies directly providing public services, by increasing user-charges and initiating greater self-reliance in funding. One example noted above is in selected educational programs. To achieve these objectives, public finance must be placed on a more solid and less volatile footing by: (a) cutting explicit and implicit subsidies to statutory bodies, govern- ment agencies and others, such as those in transport, housing, rice production, distribution and milling, and education; (b) restricting new lending, lending guarantees, equity contributions, tax exemptions and other forms of financial support for loss-making public enterprises by full or partial divestiture, liquidation and NFPE incentive reform; (c) reducing debt service costs and exposure through more aggressive internal and external liability management, including use of swap operations and, potentially, commodity bonds; (d) unifying the sales, excise and service taxes to reduce administra- tive costs, perhaps in the context of a VAT implemented at the retail stage; and - xi - (e) shifting the basis for grants to statutory authorities from a pro- gram to a consolidated budget basis to avoid the current situation where recipients of Federal grants for specific programs also build- up financial assets based on surpluses in other activities. Conclusion Implementation of these reforms would give the economy a solid foun- dation for long-term economic growth. From a macroeconomic perspective they would put more purchasing power in the hands of the private consumer and more profits in corporate treasuries, expanding private domestic demand which is the sector that must lead growth over the medium-term. The recommended reforms would also induce greater employment, particularly amongst the young and more highly educated, and more labor-intensive growth. The measures proposed to reduce the public deficit would also help sustain private demand and would reduce pressures on domestic capital markets to mobilize funds to purchase government debt, releasing funds for private investment. From a microeconomic perspective, the reforms would introduce greater neutrality into factor and output prices, more efficient risk-sharing in the economy, more effective physical and human capital formation and a better matching between individuals' expected earnings and expenditures in retirement. Consistency between desired macroeconomic and microeconomic outcomes would be achieved if the reforms were implemented as a package rather than as individual items. I. INTRODJCTION 1.1 From an economic perspective, the Malaysian economy, as it stands today, is a seeming paradox. There is excess supply in all the major factors of production: unemployment has risen to worrying levels in rural and urban areas and in all skill categories; liquidity in the banking system is ample and short-term interest rates on Treasury bills and interbank loans have declined to very low levels; finally, an increased supply of land is being idled. In some countries, these trends could be ascribed to major distortions in the economy or deficient aggregate demand. This is not the case in Malaysia, however. The trade regime is one of the most open of all developing countries, international capital flows in and out freely, the exchange rate floats according to supply and demand, inflation is low, domestic capital markets have been liberalized and labor unions are moderate; furthermore, public expenditure and deficits are high. Yet, seemingly embarrassed by the richness of potential, real GDP and consumption per capita have stagnated for the past three years, and are only showing signs of a vigorous expansion in 1988. 1.2 Any judgment of economic performance on the basis of short-term growth rates is unquestionably narrow. Malaysia has made great strides in reducing poverty, in restructuring society and in adjusting to some of the most severe external shocks to afflict any developing country. Between 1980 and 1982 for example, the terms of trade deteriorated by 17%, wiping out about 5% of national income per year and leading to sizeable deficits in the current account of the balance of payments. The adjustment from a deficit of 14% of jIGP in 1982 to a surplus of about 8% of CGP in 1987 is a remarkable feat. But the successes in adjustment will quickly pall if growth does not resume and with it the Government's ability to achieve employment and other objectives of the New Economic Policies (NEP). 1.3 A short-term hiatus in growth need not be of great concern, espe- cially when viewed against the background of Malaysia's strong past growth. In present circumstances, however, a more detailed look at the causes of stagnation is warranted because of its coincidence with the Government's shift in growth strategy towards a concentration on industrialization. As part of its policy program, Government has expanded public investment and employment, mobilized new domestic savings, rapidly built up public indebtedness and supported manufacturing growth through directed credit, tax holidays and exemptions, and labor training programs. Despite these measures, there are indicators of worrisome trends that go beyond the aggregate growth statis- tics. Private investment has declined and with it efficiency growth and technological upgrading. Public enterprises have outgrown their institutional control structures and are an increasing financial burden on the Treasury. Public expenditures have had to be sharply curtailed to contain rising public debt. Finally, unemployment has steadily risen and the most dynamic economic sector, manufacturing, has generated far fewer new jobs in the 1980s than its output growth rate would warrant. A fundamental issue, therefore, is whether these trends imply that the institutional and policy framework needs adjustment or whether they are transitory phenomena occasioned by political uncertaincies and unfavorable international economic developments. - 2 - 1.4 The first stage in implementing Government's growth and industrial- ization strategy was to secure the requisite resources for investment. The fundamental theme of resource mobilization has been addressed in two previous World Bank reports, "Development Strategies and their Financing" (1985) and "Industrializing a Primary Producer" (1986). Malaysia's high savings rates and secure external creditworthiness are testaments to the Government's successes in managing this stage. As the immediacy of resource mobilization issues has receded (with the important exception of managing continued high public deficits), attention in this report is turned to the issues associated with the allocation of resources and the transformation of accumulated financial capital into an efficient and dynamic industrial base. To a signi- ficant degree, these issues are concerned with mechanisms for balancing rewards to investors and savers commensurately with the degree of risk under- taken. The report, therefore, analyzes investment, financial intermediation, and wage-setting processes which all involve risk-sharing at the firm and sectoral levels. It also addresses the macroeconomic response to external shocks, stemming largely from terms of trade changes and involving fiscal, monetary and exchange rate policies that apportion the burden of adjustment among the public and private sectors, wages and profits, and households and corporations. 1.5 As recent events have diverged significantly from what was anticipated in the Fifth Malaysia Plan (5MP) for 1986-90, major Government initiatives to address the problems of low private investment and high unemployment are taking place without benefit of an overall macroeconomic framework. Accordingly, an update of macroeconomic performance and medium- term prospects is provided in Volume I of this report. The analysis first retraces developments in the Fourth Malaysia Plan (1981-85) as this heralded the start of the industrial growth strategy, and ther. identifies several significant structural weaknesses, rooted in the period of the FMP, which were brought into full view during the recession of 1985-86. Next, the prospects for medium-term growth are discussed, from both supply and demand perspectives. The last chapter contains a set of integrated policy recom- mendations to sustain growth. Volume II of the report contains chapters on public enterprises, public finance and employment. These chapters contain more detailed analysis on which the policy recommendations and medium-term scenario in Volume I is based. -3- II. RECENT ECONOMIC DEVELOPMENTS, 1981-87 2.1 The period 1981-87 covers the Fourth Malaysia Plan (1981-85) and the start of the Fifth Plan (1986-90). During this period, the economy underwent sharp reversals in growth performance and macroeconomic policies, brought on by external shocks and by the country's accelerating structural transformation from primary conodity production to industrialisation. The vicissitudes of the period are illustrated by a series of landmark events: the most severe recession since the country's independence arose in 1985-86, during which real per capita national income fell by 18X; the value of manufactured exports surpassed the major primary commodity exports for the first time in 1987; federal government savings turned negative for the first time in 1986; the merchandise trade account recorded the first negative balances for decades in 1981 and 1982 but achieved the largest surplus on record in 1987; and the country started to reduce its external debt through a program of prepayments starting in 1987. 2.2 A review of the turbulent events of the 19809 provides valuable lessons for assessing Malaysia's current growth prospects, new strategic directions and the efficacy of government policy instruments. These lessons should prove helpful to the Government in its handling of economic planning and management, particularly now that events have overtaken the assumptions and projections of the Fifth Plan. Under these circumstances, the policy agenda outlining how best to achieve the objectives of job creation, struc- tural transformation, poverty eradication and restructuring of society must be rethought. It is hoped that this review will assist in this process. This chapter, therefore, outlines the major economic trends of this period--the "forced" expansion of 1981-84, the major adjustment and subsequent recession of 1985-86, and the recovery of 1987--and concludes by identifying current macroeconomic problems rooted in this earlier period and the structural features underlying these problems. A. The Expansion of 1981-84 2.3 At the turn of the decade, the Malaysian economy entered a new phase of growth marked by a sharp deterioration in the terms of trade and the emergence of sizeable deficits in the current account. The average impact of external price developments during 1981-83 was to reduce national income by 4.52. Moreover, the slowdown in international economic activity dampened demand for exports. On the supply side, real growth was negatively affected by the plateau reached in important agricultural commodities, especially rubber and timber, but this was offset by a boom in the exploitation of petroleum and natural gas. 2.4 Given these concerns, the Fourth Plan targets were correctly set below the actual performance of the 1970s. Real GDP growth of 7.6% was projected, largely based on manufacturing expansion (11), but also including sizeable contributions from construction (9%) and government (9%). Two key components of this forecast were: (i) continued rapid expansion of manufac- tured exports, especially textiles and electronics and (ii) moderately strong growth in domestic demand (5.9%) led by non-oil private investment (11%). - 4 - 2.5 The Plan envisaged an almost total reliance on domestic resources to finance growth. The foreign financing component was to average less than 0.2Z of total investment. Fiscal and monetary policy would be oriented towards resource mobilization and would channel the funds towards private investors. The private sector would spearhead growth in productive capacity while public sector investment would be geared to poverty alleviation, restructuring of society and improving the quality of life. 2.6 As indicated in Table 2.1, actual events quickly diverged from the Plan forecasts. The declining terms of trade experienced in 1981-82 limited private sector investment. Government, however, compensated for the short- fall by launching an accelerated public investment and expenditure program to sustain real growth at an artificially high level. Although this protected the economy against the global recession of 1981-82, the cost of this policy was a soaring current account deficit. The sectoral balance envisioned in the Plan could not be maintained; government services and construction expanded rapidly, while the rest of the economy became weaker. Excess supply emerged in both sectors, reflected in ballooning public deficits and an overhang of unsold properties. Table 2.1: KEY MACROECONOMIC DEVELOPMENTS, 1981-87 FMP target Actual 1981-85 1981-85 1981-84 1985 1986 1987 Terms of trade index 1980=100 /a 99.5 80.2 80.4 79.2 76.0 79.3 Current account surplus (Z GNPT -0.1 -8.5 -10.2 -2.4 -0.3 8.3 Overall public sector deficit (X GNP) -6.7 -15.4 -17.6 -5.7 -10.6 -8.4 Real growth: GDP 7.6 4.7 6.7 -1.0 1.2 5.2 Agriculture 3.0 2.8 2.9 2.5 4.0 7.5 Manufacturing 11.0 5.3 8.6 -3.8 7.5 12.7 Construction 9.0 3.7 8.1 -8.4 -14.0 -6.0 Government services 9.0 5.3 6.5 2.1 4.3 4.0 /a 1970 weights. Source: Fourth Malaysia Plan, 1981-85; Economic Report of the Treasury, Ministry of Finance; Bank Negara Malaysia, Annual Report, 1987. 2.7 The large current account deficits in 1982 and 1983 were primarily a reflection of high gross fixed capital formation by the public sector. National savings remained high, at over 25Z of GNP. Investment has since declined, starting with reductions in Federal government development expendi- ture, then extending to the NFPEs as projects initiated in the early 1980s were completed, and culminating with the collapse in private investment in 1985 and 1986 as recession took hold, competitiveness deterioriated and commodity prices plunged. Between 1983 and 1987, gross fixed capital forma- tion has fallen from 38Z to 24Z of GNP. Although public sector savings have also declined, private savings growth has more than offset this, reaching a rate of 291 of CNP in 1987. Thus, of the twenty percentage point turnaround in the current account between a deficit of 121 in 1983 to a surplus of 81 in 1987, fourteen percentage points are accounted for by lower investment and the remainder by higher private savings. 2.8 The Plan had allowed for a rapid growth in both exports and imports, but overall the net external sector was expected to contribute little to total demand growth. The engine of growth was to have been domestic private demand, particularly private investment. While domestic demand did indeed lead growth through 1984 (Table 2.2), the composition was not as expected. The role of public sector consumption and investment was proportionately almost double that originally envisaged, while private demand grew more slowly. When the recession year of 1985 is included, the fall-off in private demand shows up even more starkly; private demand growth over 1981-85 was only one quarter of total demand growth, rather than the three quarters of growth anticipated. The public sector and the net trade account provided the impetus for growth. More recently, domestic demand has actually fallen, and growth has only been realized because of export performance and a reduction in imports. Table 2.2: COMPOSITION OF CHANGE IN DEMAND (X of change in CDP) FMP target Actual 1981-85 1981-85 1981-84 1984-87 Domestic 97.8 75.1 102.4 -276.3 Private 74.1 28.1 57.7 -161.9 Public 23.7 47.0 44.7 -114.4 External 2.2 24.9 -2.4 376.3 Exports 57.1 73.2 68.4 305.3 Imports /a -54.9 -48.4 -71.3 71.0 /a Negative sign indicates growth in imports. Source: Fourth Malaysia Plan, 1981-85; Economic Report of the Treasury, Ministry of Finance. 2.9 While the shrinking private sector role in economic growth was initiated by the terms of trade decline in 1981 and 1982, thereafter several interlocking factors combined to generate a situation where incremental private activity was jeopardized by excess capacity, falling profits, rising -6- costs of new expansion, public encroachment into private spheres, and loss of competitiveness in trade caused by rising unit labor costs in manufacturing. To some extent, these problem. were accencuated by the reflationary expendi- ture policy of the general Government and public enterprises, coupled with accomodating monetary and exchange rate policies. Fiscal Exp!nsion 2.10 The scale of fiscal expansion in 1981 and 1982 was unprecedented. Based on actual and expected future increases in revenues from petroleum, government launched an ambitious series of development projects. Public development expenditure almost doubled as a share in GNP, from 14.3Z in 1976- 80, to over 271 in both 1981 and 1982 (Table 2.3). Although expenditures were partially financed by a surplus in NFPS operations, attributable to profits generated by PETRONAS and its subsidiaries which reached 5.7Z of CNP in 1982, the consolidated public sector deficit swelled from 81 of CNP in 1979 to 221 and 19S in 1981 and 1982 respectively. Table 2.3: CONSOLIDATED COVERNMENT ACCOUNTS (1 of GCP) 1976-80 1981 1982 1983 1984 1985 1986 1987 (average) Federal and State Governments Revenue 29.2 32.8 33.1 33.2 32.5 36.3 37.0 31.0 Current expenditure 26.1 30.9 30.5 30.6 29.5 30.9 35.6 32.4 Operating Surplus 3.1 1.8 2.6 2.6 2.9 5.4 1.4 -1.4 Operating surplus of nonfinancial public enterprises (NFPEs) 1.0 4.0 5.7 6.6 6.8 7.8 4.5 4.5 Consolidated public operating surplus 4.1 5.8 8.3 9.2 9.7 13.4 5.8 3.1 Development expenditure 14.3 27.6 27.2 26.2 22.9 19.1 16.4 11.5 Federal Government TEX DU 230.5 16.8 12.4 110.5 TO. 7.4 NIPES 2.3 4.2 6.7 9.4 10.5 8.6 5.8 4.1 Overall deficit -10.2 -21.7 -18.9 -17.0 -13.2 -5.7 -10.6 -8.4 Source: Ministry of Finance, Economic Report of the Treasury; Sank Negara, Annual Report, 1987. 2.11 Public investment was mainly oriented towards infrastructure (Table 2.4) with transport, education and irrigation expenditures doubling between 1980 and 1982. In many instances, these projects were justified as much by their contribution to domestic demand as by their future expected returns. Irrigation for padi production for example, has very low or negative returns when evaluated at border prices, but has strong linkages to rural activity in the construction phase. Similarly, many transport projects, including the Sandakan airport, were overdesigned, lowering their rate of return but increasing their short-term contribution to aggregate demand. In addition, expenditures on defense were raised, from M$713 million in 1979 to M$2.1 billion in 1082. 2.12 The current expenditures of the Federal Government also rose rapidly in the early 1980s, accounting for 26X of GNP in 1981 compared to 23X in 1979. The most rapidly growing items were for operating expenditures in the provision of economic services in agriculture and rural development, commerce and industry (see statistical Appendix, Table 3.2). General administration expenditures also soared, from M$752 million in 1979 to M$2.1 billion in 1982, as a result of higher wages and an expansion in the number of civil servants. Health and education programs were also expanded. Last, debt servicing, largely comprised of interest payments, accelerated as total public debt and interest rates rose; at M$2.7 billion in 1982, debt servicing charges represented 16.3Z of federal operating expenditure, compared to M$1.3 billion and 12.72 in 1979. 2.13 In 1982, a process of structural adjustment was initiated to redress the twin deficits in the public sector and the external accounts. As shown in Table 2.3, the fiscal adjustment was primarily based on two factors. First, the operating surplus of the NFPEs rose with the coming on stream of liquified natural gas exports and the expansion in petroleum production and exports (Table 2.7 below). The NFPE surplus rose steadily in nominal terms and as a percent of GNP until the collapse of petroleum prices in 1986. The second element of fiscal adjustment was in development expenditures. The share of GCP accounted for by Federal Government expenditure in 1984 and 1985 had returned to the level of the late 1970s, around 12Z of GNP. This was accom- plished by a one-third cut in nominal development expenditures, concentrated in defense and commerce and industry, and, in 1987, in infrastructure (roads and irrigation). Social services, including education, health and housing were largely spared. 2.14 Although Federal government development expenditure was reduced after 1982, overall public sector investment remained high through 1984 because NFPE investment continued to grow until then. The expansion of the NFPEs was partly a function of the fact that these expenditures were not directly controlled by the Treasury, and partly because of the long gestation of the projects, which implied that, once initiated, investments could not be easily halted without considerable wastage. During 1982-84, investment by NFPEs grew by 392 annually, and by 1984 the NFPEs accounted for half of all public investment--much higher than their 102 share in 1979. Table 2.4: THE COMPOSITION OF PUBLIC INVESTMENT, 1978-87 1978-80 1981 1982 1983 _ 1984 1985 1986 1987 M$ bln 1 H$ bln 2 MS bln M$ bln 2 MS bln X MS bln 2 MS bln Z MS bln Z Infrastructure 3.8 83.1 7.5 81.1 9.6 84.7 10.6 84.4 10.0 83.1 10.1 82.8 10.3 91.6 6.1 70.8 Land development /a 0.5 11.2 0.7 7.2 0.9 7.7 0.6 5.2 0.6 5.3 1.2 9.8 0.4 3.5 0.2 2.5 Social sectors 0.4 7.8 0.8 9.1 1.1 10.1 1.1 8.7 1.1 8.9 1.0 7.8 1.1 10.2 1.1 13.4 Transport and utilities 2.1 46.4 3.9 41.4 5.7 50.4 6.4 51.0 7.3 60.5 6.7 55.0 5.8 51.9 3.8 43.6 1 Others 0.8 17.7 2.2 23.4 1.9 16.5 2.4 19.5 1.0 8.4 1.2 10.2 2.9 26.0 1.0 11.3 0 Directly Productive 0.8 16.9 1.8 18.9 1.7 15.3 2.0 15.6 2.0 16.9 2.1 17.2 0.9 8.4 2.5 29.2 Commerce and industry 0.3 6.3 0.7 7.1 0.9 7.7 1.1 9.0 1.5 12.3 1.4 11.7 0.2 1.5 1.5 18.0 Housing 0.2 4.2 0.6 6.4 0.4 3.3 0.5 4.2 0.3 2.1 0.3 2.4 0.3 2.4 0.3 3.7 Agriculture 0.3 6.4 0.5 5.4 0.5 4.3 0.3 2.4 0.3 2.5 0.4 3.1 0.5 4.5 0.7 7.5 Total 4.6 100.0 9.3 100.0 11.4 100.0 12.5 100.0 12.0 100.0 12.2 100.0 11.2 100.0 8.6 100.0 /a Including drainage and irrigation. Source: II__ -9- The Construction Boom 2.15 The counterpart to heavy government investment was a continuation of the boom in construction, which had started in 1977. Driven by civil works projects financed by windfall receipts from oil, the construction sector expanded by 9.71 during 1981-84. The sector was, however, unable to expand as fast as demand, so prices rose. By 1982, price increases had added 35X to construction costs relative to 1978. By comparison, the producer price index for local production had risen by only 21.6X. 2.16 Adding to the cost of investment in the early 1980s, and contribut- ing to the boom in construction, was the steady rise in the price of land and property, especially in urban areas. Office rental rates in Kuala Lumpur, for example, as reported by the Valuation Department of the Ministry of Finance, more than doubled between 1980 and 1982 (Table 2.5). This indicates the considerable tightness in the property market that triggered the boom. On the supply side, obtaining land alienation licenses and constructing office or residential property became an easy route to instant wealth. Investors realized both value added from construction and capital gains on the land and property, wAich together yielded handsome returns. With tangible collateral, commercial bank lending for property also soared, facilitating both the expansion of const,u.tion and accommodating price increases. On the demand side, however, growth was more modest as real income gains and private con- sumption spending slowed. The ratio of housing prices to household income, an inverse measure of the affordability of housing and hence of likely demand, grew from 3 to 5 between 1979 and 1982. 2.17 The result of these forces was a speculative cycle of considerable amplitude, both for prices and for construction output. Price changes have not served to equilibrate the property market. At the end of 1984 there was a uizeable stock of unsold residential and commercial properties which, despite low prices, has yet to be worked off. The current vacancy rate for office space in Kuala Lumpur is an estimated 201. - 10 - Table 2.5: SELECTED STATISTICS IN PROPERTY MARKETS, 1981-87 1981 1982 1983 1984 1985 1986 19871e Office rentals in Kuala Lumpur /a 30 46 38 28 23 17 13 (Z change) 36.4 53.3 -17.4 -26.3 -17.9 -26.1 -23.5 Housing values /b ('000 ringgit) 73.6 80.3 80.8 83.8 79.7 70.1 n.a. (X change) 23.3 9.1 0.6 3.7 -4.9 -12.0 Household income /c ('000 ringgit) 16.2 16.0 16.6 18.9 16.1 15.2 n.a. Housing credits /d (bn. ringgit) 7.0 9.4 11.3 13.7 16.5 n.a. n.a. Housing v^lue/household income 4.54 5.02 4.87 4.43 4.95 4.61 Memo: CPi (O change) 9.7 5.8 3.7 3.9 0.3 0.7 1.5 /a Ring5it per month per square meter. 7r Average value of single-story terrace houses in a sample of cities. 71 Private disposable income per household. 7T Total loans for housing including credit institutions, commercial banks, finance companies, government housing loans, cooperatives and others. /e Preliminary. Source: Property Market Report, various issues, Ministry of Finance; Bank Negara Malaysia, Annual Reports; World Bank, "Malaysia: Housing Sector Study," 1988. Loss of Competitiveness 2.18 The simultaneous expansion in construction and in government services brought about a tightening of the urban labor market. At the same time, the productivity growth in agriculture associated with a switch from rubber to oil palm began to slow, and the downward trend in agricultural employment which had been observed during the Third Malaysia Plan (1976-80) reversed itself. Overall unemployment rates fell from 5.7Z in 1980 to 5.11 in 1982. 2.19 These changes in the labor market brought about a rise in real wages relative to productivity. The traditional measure of the real wage, the nomi- nal wage deflated by the rise in the consumer price index, did not grow faster than in the 1970s. Thus, it was not the case that workers were enjoying acce- lerating increases in their living standards. What did happen, however, is - 11 - that the real wage grew far faster than productivity (Figure 2.1).]/ From the perspective of a c3mpany, it is the relationship between the real wage and labor productivity that is one important factor in determining competitive- ness, not the real wage per se. The reason for the divergence between the two indices can be traced to labor market developments and the negative trend in productivity during this period. 2.20 Other countries in East Asia also suffered from the phenomenon of domestic wages rising faster than productivity (Table 2.68. But whereas they were able to hold down unit labor costs in US dollar terms by aggressive de- valuations, the rise in unit labor costs in Malaysia was aggravated by the Government's exchange rate policy, resulting in a loss in international compe- titiveness. In Korea, Thailand and Taiwan, unit labor costs in manufacturing denominated in US dollars were roughly level with their 1980 values in 1984 and 1985. In Hong Kong, they were over 202 lower. Only in Singapore and Malaysia is the trend sharply upwards, with a rise of 40-50Z in just four years. 1/ The index of labor productivity used in Figure 1 is an index of the ratio of value added to employment in the whole economy and in the manufactur- ing sector in the two panels, respectively. When the ratio between value added and output is constant, such an index also represents output per worker. The comparison between productivity and wage rates is a useful indicator of the wage cost involved in production. Real wage grwoth at the same rate as productivity growth implies a constant proportion of total wages to output in constant ULC. When wage growth exceeds produc- tivity growth, the wage bill per unit of output is increased and, ceteris paribus, there is upward pressureon costs and a decline in competitive- ness. - 12 - Figure 2.1 Real Wage and Labor ProducUvity 130 -aIYalF t990-100 120 - fa Labor ProductM * Rea Wage 110 - 100 90 -8 80 70 60- 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 Year Real Wage and Labor Productvity Mwmniurtuling hIdusry, 1980=100 170 - 004% B Labor ProductMly 160 - * Rea Wage / 150- / 140 - 130 / 120 - 110 100 90 80' _ 70 - \u , so 71 72 73 74 ;S 76 77 78 79 80 81 82 83 84 85 86 87 Year - 13 - Table 2.6: MANUFACTURING UNIT LABOR COSTS IN EAST ASIA (1980 = 100) 1981 1982 1984 1985 1986 1987 A. Unit Labor Cost in Domestic Currency Korea 108.1 126.2 131.2 145.4 147.8 163.0 Taiwan, China 113.9 121.9 132.7 128.5 126.4 124.8 Singapore 111.9 131.9 140.0 151.0 134.1 nea. Hong Kong 108.8 130.7 116.0 132.0 127.4 n.a. Malaysia 109.7 125.3 154.8 178.3 169.9 153.9 Thailand 110.8 109.9 114.9 121.4 n.a. n.a. B. Exchange Rate (LC/US$) Korea 112.1 120.4 132.7 143.2 145.1 135.4 Taiwan, China 102.3 108.6 109.9 110.6 105.0 88.4 Singapore 98.7 99.9 99.6 102.8 101.7 n.a. Hong Kong 112.5 122.0 157.1 156.6 156.8 n.a. Malaysia 105.8 107.3 107.7 114.1 118.6 115.8 Thailand 106.6 112.3 115.4 132.6 n.a. n.a. C. Unit Labor Cost in US$ (A/B) Korea 96.4 104.8 98.9 101.5 101.9 120.4 Taiwan, China 111.3 112.2 120.7 116.2 120.4 141.2 Singapore 113.4 132.0 140.6 146.9 131.9 n.a. Hong Kong 96.8 107.1 73.8 84.3 81.3 n.a. Malaysia 103.6 116.8 143.8 156.3 143.3 132.9 Thailand 104.0 97.9 99.6 91.6 n.a. n.a. 2.21 The contrast in exchange rate developments between Korea and Malaysia is indicative of totally different approaches to exchange rate management. In Korea, particularly since 1980, the nominal exchange rate has been adjusted so as to offset any increases in unit labor costs made in domes- tic currency terms (Figure 2.2a). Thus, purely domestic factors influenced the exchange rate. By contrast, Malaysia pursued a more accommodating ex- change rate policy which closely reflected international events and which displayed the typical characteristics of an oil exporter. As the dollar soared in 1980-84, the real exchange rate in Malaysia appreciated (Figure 2.2b). More recently, it has depreciated in line with the US currency. The real appreciation in the value of the ringgit, however, combined with the rise in real wages, was a blow to Malaysian competitiveness, particularly in manufacturing, from which it is only slowly recovering. 2.22 The different exchange rate experiences between Korea and Malaysia reflect inherently different systems for dealing with international capital flows. Korea has a closed capital account and effective central bank control over the nominal parity of the won. Malaysia has an open capital account and - 14 - Figure 2.2a KOREA: INTERNATIONAL COMPETITIVENESS 17- _(1008100) o ULC In Korean Wan 160- + WanAU.S.$ o o ULCInU.S6 140 140 130 120 110 1 " " / '''''''' P * a..*,,, ;"" 100 a.4.00 90 o0- f 1979 1980 1911 1982 1983 1994 1S85 1996 1997 Yeaw Figure 2.2b MALAYSIA: REAL EXCHANGE RATE MOVEMENTS (1998=100) lo~~~~~~to .50-. 140 - REER 130 got - t20- 70 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1901 1982 1993 1984 1995 196 1997 YEAR a- - 15 - a floating exchange rate regime. The competitive problems in Malaysia arose not because of exchange rate mismanagement pE se--exchange rates were left to float during this period-but because, given exchange rate movements, the nominal wage rose too rapidly. The Taiwan example illustrates how exchange rate movements similar to those experienced in Malaysia need not have resulted in a loss of competitiveness had wage increases been kept down. 2.23 One policy that sustained the ringgit in the early 1980s was the Government's decision to utilize foreign borrowing to finance its investment program. In 1981 and 1982 net public foreign borrowing amounted to 7.5Z and 9.61 of GNP, respectively. Over the period of forced expansion, 1981-84, foreign savings on average financed 28Z of total investment compared to the plan forecast of 0.22. A substantial proportion of the foreign borrowing (571 in 1983-84) was undertaken by public enterprises. This recourse to external funds helped these agencies escape the surveillance and discipline that could have been imposed by the Federal Government had there been a greater reliance on the Treasury as a source of funds. Overall, net foreign inflows more than compensated for the current account deficits being registered. Central Bank foreign exchange reserves built up steadily, and pressure to depreciate the exchange rate was temporarily diverted. 2.24 Rising real wages and an appreciating currency imposed a profit squeeze on manufacturing which was reinforced by a large shortfall of private consumption and investment below expected levels: the former as a result of real income losses due to the prolonged terms of trade deterioration, and the latter because of declining economywide profits and rising global interest rates. Furthermore, one source of rapidly growing domestic demand, the investment binge of the NFPEs, failed to benefit local industrialists since much of the public enterprise investment was highly import-intensive, with foreign machinery and equipment purchases closely linked to financial and technical assistance being sought from abroad. Against this background, it is not surprising that manufacturing growth during 1981-84, at 7.6Z, fell far short of its target of 11Z. B. The 1985-86 Adjustment Period 2.25 In 1985, government efforts to adjust the structure of public finances started to bear fruit. The Federal Government deficit narrowed and, for the first time in the 19809, the overall expenditure of the NFPEs was brought below their revenue. Government efforts were oriented towards improv- ing efficiency and productivity in government services. At the same time, monetary policy was loosened to facilitate growth and provide liquidity for private investment. Interest rates were brought down and a New Investment Fund of M$l billion was set up to channel funds into the directly productive sectors. The policy mix therefore shifted towards a tight fiscal/loose monetary stance in sharp constrast to the strategy in previous years which had targeted loose fiscal policy at growth maximization and tight monetary policy at managing the balance of payments and inflation. 2.26 The actual performance of the economy in 1985-86 was mixed. On the external side, the terms of trade declined sharply with the result that real - 16 - national income per capita fell by 6.5% and 11.2Z successively.11 Export receipts also fell with weakness in commodities coinciding with a downturn in manufactured exports. Despite this fall, however, imports were reduced by even more (-101 in 1985, concentrated in investment and intermediate goods (Statistical Appendix Table 4.7)) and considerable progress was achieved in reducing the current account deficit. This improvement, coupled with easing international liquidity, assured access for Malaysia in foreign capital markets. A program of external debt refinancing was instituted and Malaysia became the first developing country to obtain a syndicated credit at a rate below LIBOR, and one of the first developing countries to obtain a credit rating from the major US agencies. 2.27 Internal developments were considerably less favorable. Real GDP stagnated for two years, the worst performance since independence. The conso- lidated public sector deficit rose again in 1986, reaching 10.6Z of GNP, as revenues from all sources declined: income taxes fell in response to the lower income in 1985; indirect tax receipts fell as imports and sales declined; and petroleum-based taxes and royalties collapsed with the fall in oil prices. Unemployment also started to rise significantly, as labor- intensive sectors in commerce, government and construction were badly hit, and reached 8.5Z by 1986. External Shocks 2.28 The immediate cause of Malaysia's recession was a collapse in the prices for its major exports (Table 2.7). While the diverse mix of commodi- ties exported had in the past provided a solid hedge against price fluctua- tions, the problems of 1985 and 1986 were caused by a simultaneous reduction in prices for petroleum, palm oil, rubber, sawlogs, tin and cocoa. Based on an analysis of historical post-war commodity prices, however, it appears that the probability of a shock of such magnitude or larger occurring is quite high--one year out of every three is liable to see a deviation from trend of M$2 billion in primary (non-oil) commodity exports due to price fluctuations alone (see Appendix 1). This analysis suggests that Malaysia will in all probability be faced with similar situations in future. In light of the economy's vulnerability to external events, a smoother pattern of real govern- ment spending is desirable in order to minimize sharp oscillations in the economy. 2/ The official terms of trade index uses 1970 weights and consequently accords petroleum a very low weight. The income figures above are calculated using data on current and constant 1978 price exports and imports. - 17 - Table 2.7: MAJOR COMMODITY EXPORT PRICES, 1981-87 1981 1982 1983 1984 1985 1986 1987/a ------(M$/ton) - … A. Prices Petroleum (US$/barrel) 39.0 36.3 30.7 29.3 27.6 14.8 18.2 Palm Oil 1,153 829 984 1,583 1,046 579 720 Rubber 2,503 1,927 2,344 2,308 1,919 2,100 2,420 Tin 32,150 30,535 30,088 29,343 28,711 16,089 16,915 Sawlog. ($/cu meter) 156 175 149 166 141 151 186 Sawn timber ($/cu meter) 360 352 371 368 409 435 451 Cocoa n.a. 3,440 4,000 5,110 5,000 4,700 4,357 S. Export volume growth rates -------- (percent) - …-- … Petroleum -9.9 18.0 18.8 16.0 1.2 12.5 -4.0 Palm oil 9.9 14.9 7.9 1.6 8.6 33.9 -6.1 Rubber -2.7 -7.2 13.4 1.8 -5.9 1.3 6.9 Tin -14.7 -26.9 17.5 -30.7 44.9 -29.7 22.9 Sawlogs 4.4 21.1 -3.0 -9.5 15.8 -2.9 20.7 Sawn timber -10.1 9.1 11.8 -17.9 -3.2 7.9 30.8 Cocoa n.a. 36.5 -0.7 15.6 23.2 30.1 48.5 la Estimates. Sources Economic Report of the Treasury, various issues, Ministry of Finance. 2.29 To some extent, the impact of the terms of trade decline was offset by rising commodity export volumes, particularly in petroleum, palm oil and cocoa. Other commodities had a more mixed performance, with rubber and timber exports declining in 1985 and then rebounding in 1986, while tin and sawlogs posted strong real gains in 1985 before volumes retreated back in 1986. Nevertheless, the total value of major commodity exports fell by almost one- quarter in nominal terms by 1986 relative to 1984. 2.30 One result of weak commodity exports was to increase the importance of manufacturing exports. These had doubled between 1981 and 1984 (Table 2.8), largely on the basis of rapid growth in electronics exports, particu- larly semiconductors, and have roughly reached parity with commodity exports. In fact, Malaysia's manufactured export performance had been rated among the best of all developing countries in the 1970s. The World Bank's World Development Report 1987 shows Malaysia to be the sixth best developing country exporter of manufactured goods in the 1970s after better known leaders - 18 - Table 2.8: THE STRUCTURE OF EXPORTS, 1981-87 1981 1984 1986 1987 M$ mln % MS mln % M$ mln % M$ mln Z Commodities 20,807 76.8 26,471 68.5 20,460 57.3 24,914 55.1 Petroleum 6,918 25.5 8,733 22.6 5,400 15.1 6,290 13.9 Palm oil 2,710 10.0 5,194 13.5 3,346 9.4 3,250 7.2 Rubber 3,712 13.7 3,495 9.1 3,182 8.9 3,915 8.7 Tin 2,138 7.9 1,163 3.0 651 1.8 839 1.9 Saw logs 2,473 9.1 2,774 7.2 2,872 8.0 4,286 9.5 Sawn timber 971 3.6 1,169 3.0 1,305 3.7 1,770 3.9 Other 1,885 7.0 3,943 10.2 3,704 10.4 4,563 10.2 Manufactures 6,302 23.2 12,156 31.5 15,264 42.7 20,267 44.9 Electrical machinery, appliances and parts 3,017 11.1 6,749 17.5 8,491 23.8 10,976 24.3 Transport 308 1.1 569 1.5 518 1.5 722 1.6 Food and beverages 600 2.2 777 2.0 960 2.7 1,223 2.7 Textiles, etc. 785 2.9 1,141 3.0 1,643 4.6 2,286 5.1 Wood and wood products 473 1.7 416 1.1 537 1.5 851 1.9 Rubber products 83 0.3 107 0.3 155 0.4 243 0.5 Chemicals 417 1.5 452 1.2 760 2.1 917 2.0 Minerals and metals 221 0.8 433 1.1 714 2.0 1,074 2.4 Other 398 1.5 1,512 3.9 1,486 4.2 1,975 4.4 Total 27,109 100.0 38,627 100.0 35,716 100.0 l5,ot 100.0 Source: Bank Negara Malaysia, Annual Report; Economic Report of the Treasury; EPU. - 19 - such as Korea, Hong Kong and Taiwan (China). But whereas other countries, particularly the newly industrializing countries (NICs), diversified their manufactured export base while registering rapid growth, Malaysia's manufac- tured export structure has become highly concentrated, with the volatile electronics and electrical machinery subsector accounting for over half of all manufactured exports. 2.31 The rate of growth of manufactured exports slowed in 1985 and 1986 as a result of two factors. First, there was a dramatic shake-out in the global semiconductor industry which had been a consistently strong growth performer until then. Second, compounding this effect, was the loss of competitiveness in manufactured exports that further slowed growth across the spectrum of exports, particularly in transport, minerals and metal., wood and wood products, food and beverages and other products. 2.32 The decline in commodity export prices in 1985 and 1986 affected real GDP growth in two ways. First, output responded directly to price changes. This effect, however, does not seem to have been very important as the volume of commodity production and exports remained strong in 1985-86, except for tin, logs and rubber. Second, decreased prices indirectly affected output by reducing purchasing power and hence domestic demand. In Malaysia, however, the impact of terms of trade changes on consumption also seems to have been limited. In 1985-86, the decline in private consumption associated with price shocks was about 6X, or about one fifth of the fall in domestic demand. Consumption did not respond immediately to the negative terms of trade because private income was protected through growth in the public deficit and because private coumption in general only reacts partially to single year changes in income.- Overall, the Malaysian trade sector was a major source of strength during 1985-86, contributing 17.2 percentage points to rell GDP growth, shared evenly between export expansion and import compres- sion., Part of the source of the recession must, therefore, be sought for in developments other than external shocks. Internal Shocks 2.33 Economic performance in any year is a function of a complex mix of cyclical and structural trends, set in motion by past developments, by behavioral adjustments reflecting fresh expectations about the future, and by policy changes mirroring the current situation. Each of these elements imparted a negative impulse to the economy in 1985-86 which cumulatively resulted in the recession. 3/ See Appendix 6 for a discussion of the consumption function. 4/ The import compression, largely of intermediate and capital goods, served to protect domestic output from the full force of declining consumption and investment; part of the decline in demand fell onto a decline in demand for foreign goods. There was, however, little visible import substitution during this period. - 20 - 2.34 Two intartwining ba*es- the kursting of the property market bubble and thie dowatura- in the overall bu':z -.,_cls-coincided to turn Malaysia's adjustment program iuto a fult-fi'cli.A and ultimately very painful reces- sion. Signs of overheating in tht p:operty market had clearly emerged by 1984; property prices had risen a record rates, over twice as fast as the consumer price index over the preceding decade. Housing investment which had hoverel,around 5Z of GNP in the mid-1970s, rose to about 9% in the early 1980s.-' The price and output developments pushed the supply of housing well beyond underlying demand. New construction was facilitated by easy credit for property development based on its proven track record of good collateral and capital gains, which were optimistically extrapolated into the future. Thus, in common with other speculative bubbles, the property boom thrived on opti- mistic expectations for the future. These expectations were dashed when global commodity prices fell, New housing construction w:as promptly halted. Coupled with the decline in government civil works projects and the collapse in demand for commercial property as corporate profits were squeezed, the construction sector went into a tailspin which has still to bottom out, losing one fifth of its real value added by 1986. 2.35 At the same time as the property bubble burst, the business cycle turned down. Along with other countries, Malaysia's economy has fluctuated around a consistent upward trend. One characteristic of this cycle is that the upswing is associated with wage rates rising faster than productivity. As this occurs, corporate profits are hurt, investment declines, and growth suffers. An adjustment peried of slower growth, during which wage gains are minimized and productivity improved, is required to restore balance to the economy. As illustrated in Fig. 2.1 above, the wage-productivity relationship had shifted out of balance by 1984. While a *number of factors, including the shift towards capital-intensive output, inefficient investments and falling capacity utilization, were responsible for the fall in productivity growth, the result was a growing share of wages in total value added and a decline in profits. Caught in a cash-flow squeeze, corporations trimmed their invento- ries--stock adjustments in 1985 alone reduced real output growth by almost three percentage points. 2.36 Structural factors were also at play in reducing growth in 1985-86. The tremendous growth impetus that had been provided by petroleum and gas exports flattened out. Tin production was also severely disrupted with the suspension of the Kuala Lumpur Tin Market. Weaknesses also emerged in the financial services sector, necessitating Central Bank intervention through direct control in some instances and tighter regulation, monitoring and supervision in other cases. 5/ These data on housing investment are drawn from "Malaysia: The Housing Sector" World Bank Report No. 7292-MA. They are based on survey data of the stock of housing and are substantially above official figures based on the flow of new residential construction. The latter, however, also show the same pattern of a two-thirds rise in the ratio of housing investment/GNP between the late 1970s and the peak in 1982. - 21 - 2.37 Growth performance also reflected changes in domestic policy. A strategy of facilitating a real effective exchange rate depreciation through changes in nominal parities was adopted in 1986. The Malaysian ringgit fell by 72 against the US dollar and, as the dollar continued its decline in relation to other major currencies, the real effective exchange rate depre- ciated rapidly. The depreciation was an appropriate response both to the terms of trade decline and to the overvalued level to which the currency had risen, along with the US dollar, up to 1984. The legacy of the 1980-84 currency shifts had been large external deficits, which threatened to widen again in an environment of low commodity prices and a weakening of most manufacturing subsectors. Adjustment was required for long-run growth but it involved short run costs. In the short run, the exchange rate depreciation depressed consumption by an estimated 8Z. It is likely that this effect oucurred through a shift in the composition of income from the high-consuming, self-employed group involved in nontraded sectors such as trade and commerce, towards a higher-saving formal sector group engaged in manufacturing and agriculture. One indicator supporting this hypothesis is the increased contribution per member to the Employee Provident Fund (+5%) in a year which saw average incomes fall by 6Z. Fiscal Turnaround 2.38 Perhaps the most important policy shift, however, was in the fiscal accounts. The overall consolidated public sector deficit fell to 7.8% of GNP in 1985 despite the cyclical factors that would typically have led to a widening of the deficit during a year of depressed growth. The NFPEs as a group accounted for about 4.5 percentage points of the reduction in the overall public deficit/GNP ratio. Their revenues exceeded their total expen- ditures for the first time since 1981, largely as a result of a cutback in capital expenditure, and increased profits from faster oil extracti and the coming on stream of liquified natural gas. The net fiscal impulse,- the portion of the deficit attributable to policy changes rather than cyclical factors, was highly negative in 1985 (Figure 2.3). In fact, expenditure and tax changes together imparted an 11% reduction in domestic demand. The magnitude of this change, particularly in comparison to earlier years, reflects the Government's success in bringing the expenditures of NFPEs under control. As discussed above (Table 2.4), the cutbacks in investment were largely in defense and transport. By 1986, investments in irrigation, commerce and industry were also pared back. 2.39 The negative fiscal policy impulse did not last long. The fall in tax revenues from petroleum and other sources, coupled with a levelling off of government expenditures, led to a resurgence of the deficit. At the same time, the NFPE accounts also returned into the red as PETRONAS' surplus declined. Exacerbating the NFPE difficulties was the rise in the value of the yen and the consequent surge in debt service obligations. In addition to these structural elements, which account almost entirely for the increase in 6/ See Volume II, Chapter 2, for a detailed discussion of the methodology involved in the computation of the net fiscal impulse. Figure 2.3 - 22 - Net FiscaL Impulse, 1972-87 Net Fiscal Impulse Se-mtN'tiity Anq"si 0.1 - 0.08 C Net Fiscal Impulse 0.06 - 0.04 0.02 0 0 ' -0.02 -.0.04- . . -0.06 / 0 W-0.08 -0. 1 -0.12 -0.14 --0.16 C onsolidated -0.18 ~~~~~~~~~~~deficit/GNP -0.218 -0.22 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1966 1987 a Al + A2 0 A3 V CD/GNP 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 10 potential GNP growth assumption; and A3 assumes 81 growth but is otherwise based on the same assumptions as Al. - 23 - the deficit, the Government undertook several discrete policy changes. An Anti-Recession Committee was formed to create employment and growth and improve the government accounts. As a result of the recommendations of this Committee, the Special Low Cost Houaing Program was launched, to build 240,000 units over three years. In addition, a total of M$2 billion was restored to the initial cutbacks in the Fifth Plan development budget. Overall, fiscal policy became neutral in 1986. 2.40 Monetary policy was accommodative throughout this period, and nominal interest rates began to fall in line with global developments (see Statistical Appendix Tables 8 and 5.8). Real interest rates, however, rose as inflation was squeezed out of the system by the recession. Both the deposit rate and the prime lending rate increased in real terms, aggravating the drop in domestic demand. To assist in recovery, the authorities implemented new credit programs including the New Investment Fund (NIF) and a small industries development fund. They also reduced administrative requirements in the operation of the Export Credit Refinancing scheme and increased the scope of its operations by moving from a positive to a negative list approach in determining eligibility. These measures helped boost credit supply, which rose at an annual rate of 10Z, but were not effective in restoring momentum to the all-important private investment sector. In fact, econometric analysis indicates that credit is not a significant determinant of private investment in Malaysia. The loosening of credit policy, theref 9e, did not prove to be an effective instrument for stimulating the economy., C. The Beginning of Recovery--1987 and 1988 2.41 The economy bottomed out in the second half of 1986 and strengthened throughout 1987. As in the downturn, the upswing coincided with a change in the terms of trade. The provisional estimate for real GDP growth in 1987 is 5.4%, based on a strong performance in the manufacturing and agricultural sectors, particularly in the fourth quarter of the year. The balance of payments also improved, with the current account registering a surplus of 8.1S of GNP, based on substantial gains in export revenues. This permitted the authorities to prepay some M$2 billion in external debt and to build up external reserves by almost M$3 billion. The only negative items in the economic performance were the budget deficit and unemployment growth numbers. The overall consolidated public sector deficit improved to 8.4% of CNP but only because of a temporary shortfall in spending on development projects. Unemployment rose to 9.1%. In addition, evidence that the share of workers earning less than M$200/month--the minimum compiled figure--has surged from 42% to 55% suggests that the number of underemployed or part-time workers 7/ This conclusion is based on regressions reported in Appendix 6. Other analysts have found the interest rate to be a significant determinant of investments but the effect is typically weak and sensitive to model specifications. Although funds such as the New Investment Fund have been in high demand, this can be attributed to the below-market interest rate charged. It is likely that these funds substitute for other commercial bank credits. - 24 - has risen sharply, indicating an even worse aggregate employment situation than shown by the unemployment rate. 2.42 Recovery in output growth and the current account surplus indicate that the process of macroeconomic adjustment was largely completed in 1987. This adjustment can be viewed from both an external and internal perspec- tive. As shown in Table 2.9, the current account surplus reflects a rapid real growth in exports of 11.3%, coupled with price improvements that together boosted the share of exports in GNP to a record 70X. At the same time, it also reflects private sector savings far in excess of private investment. 2.43 On the trade side, both commodity and manufactured exports rose sharply; the value of total merchandise trade increased by 26.5% over 1986. Commodity export growth was led by double-digit volume growth in tin, logs, timber and cocoa, and by strong price increases in petroleum, palm oil, rubber, tin, logs and timber (Table 2.7). The higher palm oil price followed a sharp increase in demand with the entry of India into the market as a result of drought damage to that country's oil seed production. Log and timber exports, particularly from Sarawak, were mostly destined for Japan, stimulated by the construction and housiug boom that has followed the appreciation of the Japanese yen. Manufactured export growth was also strong, partly as a result of improving competitiveness following the ringgit's depreciation. In ringgit terms, the biggest increases in manufactured exports were posted in the electronics, electrical machinery, apparel and textile sectors, but the gains were broad-based (Table 2.8). Smaller industries, such as rubber and wood- based products, also showed strong export gains, and some domestically- oriented sectors, such as steel and food and beverages, were also able to export in limited quantities. 2.44 On the internal side, adjustment was evenly divided between the public and private sectors. The overall public sector savings/investment gap, which had rebounded to 11.3Z of GNP in 1986 as a result of the falling revenues from petroleum, was reduced by about 3 percentage points of CUP by a reduction in public fixed investment. This redug ion was concentrated in energy, transport and miscellaneous investments_ Meanwhile, the private sector saved an estimated 16.6% of CNP more than it spent, helping finance the public deficit and the current account surplus. The private surplus rose as a result of sharply higher savings. These stemmed from the fact that income gains from higher commodity prices were saved, and because the fall in wages and increase in unemployment helped Aepress private consumption. As a result, 8/ The current revenues and expenditures of the general government both declined as a ptlportion of CNP in 1987. Income tax revenue was nega- tively affected by the low profits registered in 1986, particularly by oil companies. Falling oil prices also account for lower export duties on petroleum and for the reduction :n the operating surplus of nonfinan- cial public enterprises. Despite this, public savings did not decline as much as might have been expected because current expenditures were held constant in nominal terms and also declined in relation to CNP as a result of growth in the economy. - 25 - aggregate domestic demand continued to fall by 1.11 in 1987. Even this modest decline, however, represents an improvement over the 10.8Z decline in domestic demand registered in 1986. Table 2.9: ADJUSTMENT TO SHOCKS, 1984-87 (X of CNP) 1984 1985 1986 1987/a Change in terms of trade (Z) 6.9 -4.5 -15.6 7.4 Exports (goods and nonfactor services) 58.2 59.2 61.8 69.5 Growth rate (Z) 13.8 0.4 17.6 11.3 Imports (Goods and nonfactor services) 56.2 51.9 54.9 54.7 Growth rate (Z) 6.5 -10.0 -2.7 6.2 Current account -5.3 -2.4 -0.2 8.1 Net external financing 5.3 2.4 0.2 -8.1 Public sector -6.5 -6.0 -11.3 -8.5 Savings /b 9.7 11.1 5.8 3.0 Investment 16.2 17.1 17.1 12.5 Private sector 1.2 3.6 1;.5 16.6 Savings 21.0 16.2 21.5 28.1 Investment /c 19.8 12.6 10.1 11.5 /a Estimates. 7T Total public sector current surplus. 7T Includes change in stocks. Source: Economic Report of the Treasury; 1987 Staff Estimates. 2.45 Domestic-oriented manufacturing, while still sluggish at 5.3Z growth compared to 19% for the export-oriented sector, was the swing performer in the recovery. That is, it was changes in the domestic-oriented manufacturing growth, relative to its 1986 performance, which allowed overall manufactured growth to improve in 1987. Much of this improvement, however, may be cyclical in nature as domestic demand in 1987 was sluggish, and declined somewhat. Thus, production gains may have gone towards replenishing inventories which had been depleted during 1985 and 1986. This does not, however, imply that growth is likely to slow in 1988. On the contrary, preliminary indications suggest an acceleration of manufacturing growth and a more broad-based parti- cipation by subsectors such as transport, cement, oils and tobacco products which had missed the 1987 recovery. In the longer-term, improvement in domestic demand, particularly in private consumption and investment, is necessary to sustain a diversified, rapidly expanding manufacturing sector. - 26 - Factor Market Adjustments 2.46 The provisional estimate of macroeconomic performance in 1987 suggests a strong momentum to growth that should carry over to 1988. At a microeconomic level as well, the signs are encouraging. The high degree of flexibility exhibited in Malaysia's credit and labor markets in 1987 is espe- cially noteworthy. As a recession is typically a period of adjustment for firms, and of a shift of resources from less efficient to more efficient subsectors, its depth and length will be partly affected by the speed with which factors of production are reallocated. Evidence of well-functioning factor markets, therefore, augurs well for a solid-based recovery. 2.47 Following a reduction in statutory reserves and liquidity require- ments, the availability of domestic credit eased substantially in 1987. The continued drop in investment led to a slight fall in commercial bank loans outstanding through the middle of the year. Coupled with a growth in deposits associated with higher real disposable income, domestic interest rates fell sharply in real and nominal terms. The three-month fixed deposit rate fell to its lowest level in over a decade (2.5Z). This responsiveness of interest rates to liquidity seems to have assisted in boosting both consumption and foreign investment inflows, alleviating the economic downturn. 2.48 Labor markets too have demonstrated a degree of flexibility. High unemployment has contributed to a levelling-off of wage demands; in new negotiated contracts covering a three-year period, only nominal increases in wages have been recorded. The wage for new entrants has been falling since 1985. However, average payments to all employees continued to rise through 1986 (by 7.2% in that year), reflecting built-in escalators in old contracts and payments made for retrenchment. The average wage finally stabilized or declined in 1987, and should now begin to fall rapidly, restoring profita- bility and encouraging output increases in all sectors. D. The Emergence of Structural Weaknesses 2.49 Although the short-term prospects for growth of the Malaysian economy are excellent, they are heavily conditioned on the current buoyant external trade and commodity price outlook. To translate this into a longer- term sustained growth, there must be a recovery of private investment and a decline in unemployment. These have shown encouraging signs of improvement in 1988, but, viewed in a historical context, the levels of private investment and the rate of unemployment still reflect inherent structural weaknesses in the recovery. The most important of these are: (a) an inability to alter the structure of public finances to achieve a sustainable deficit in the face of large external shocks; (b) growing subsidies to alleviate private income shortfalls and, hence, a fall in overall profitability and in private invest- ment; (c) a mismatch between a pool of savings seeking low-risk returns and the economy's need for risk-taking entrepreneurial investments in new produc- tive activities; and (d) rigidities in the labor market that have created a two-tier wage system and sizeable structural unemployment. - 27 - Long-run Constraints on Public Deficits 2.50 High public sector deficits must be financed by new borrowing from domestic or external sources, money creation, or a run-down of previously existing assets. Analysis of the prospects for these financing sources is a useful method for evaluating the size of the public sector deficit that can be sustained over time. In Malaysia, recourse to money creation (or borrowing from the Central Bank) has been minimal in the past few years; government policy has expressly been to avoid such financing as it can lead to inflation- ary pressures in the economy. Furthermore, after an initial surge in 1981 and 1982, borrowing from international capital markets has also been reduced. In fact, with the cost of external money above that of domestic funds (particu- larly on debts incurred in the pre-1982 high interest environment), a program of prepaying external debt has been initiated. Asset draw-downs have only recently become important, and are clearly not a long-run source of funding. Thus, the bulk of deficit financing has taken the form of domestic borrowing, largely from non-bank institutions such as the Employee Provident Fund, other pension funds, and Petronas (Fig. 2.4). 2.51 To date, debt management in Malaysia has focused on external liabi- lities. Through a program of refinancings and the initiation of prepayments, total public external debt service has been kept at a manageable and commen- dably low level: after rising to 17.61 in 1986 as a result of falling commo- dity exports, the debt service ratio has been brought down again and should dip below 15% in 1988. In comparison with other developing countries, Malaysia has remained amongst the most prudent and creditworthy of foreign borrowers. 2.52 Through its debt management policies, Malaysia has been able to avoid the short-term debt crises afflicting other developing countries. It retains superior access to foreign credit markets and, because of the govern- ment's minimal resource to the banking system and the effective mobilization of incremental private savings in non-bank financial entities, public borrow- ing has not resulted in high interest rates, as observed in other developing countries like Turkey and Brazil. Over the long-run, however, the accumula- tion of high public deficits into high overall debt will pose increasingly difficult debt mangement problems for policy makers. Indeed, Malaysia today has one of the highest public debt/GNP ratios in the world (Fig. 2.5), when both domestic and foreign indebtedness are combined. As a result, the interest on public debt now absorbs a huge 10 of GNP per year. 2.53 The build-up of public debt would be less important if the economy were growing at a rate faster than the interest rate. In that case, the growth in GNP is typically a lower bound for the growth in financial assets of an average middle-income developing country. Thus, the amount of new borrow- ing that the public sector could do by just maintaining its share of total financing would more than offset the amount of interest that would have to be paid. In fact, the noninterest public sector accounts (i.e. the primary balance, discussed in Volume II, Chapter 2) could show a sizeable deficit without any tendency for debt/GNP to rise. When the growth of GNP slows below the interest rate, however, interest payments cannot be covered by new borrow- ing without a dangerous climb in the debt/GNP ratio. Instead, the primary HOLDINGS OF FEDERAL GOVERNMENT DEBT 1987 Bitewal (5.7%) Multilater (3.4%) Social Security EXTERNAL DEBT (38.5%) Foreign Market INTERNAL DEBT s~~~~~~~~~~~~~~~~~~ni Other Domesfic lnstitons (2.5%) Corporations (5.7%) Ofer Financial Insfftiuions Central Bank (2.0%) (19.0%) ca*.41974a 11% INTERNATIONAL COMPARISON OF PUBLIC DEBT 150 - 140 - 137.1 130 - ;3 120 - DomesUc Debt 110 106 105.6 ZI Foreign Debt 100- 90 87 CL80 70 ~~~~~~67.1 ' 60 9. 58 55.8 55.4 55.3 52 50 447 44.4 4 40 - 00 30 6- 25 20 - 104359 0 r~~~~~~~~~ . 'T 00~~~~~~~~~~4C4 - 30 - balance must be brought into surplus to generate the resources to pay interest. Malaysia is in this latter situation. 2.54 Given the sizeable spread between the interest rate and GNP growth, the debt/CNP ratio in Milaysia will continue to rise unless the primary deficit is brought into surplus. The urgency for undertaking adjustment quickly rather than permitting further rises in the debt ratio stems from two factors. First, the already high public debt/GNP ratio suggests that the room for further increases is limited without eventually being faced with the unpleasant choices now confronting many highly-indebted developing countries and developed countries--namely, the need to reduce deficits in an environment where entitlement programs and contractual spending obligations are so great that major cuts in discretionary programs must be made to have any impact on the budget. Second, the fact is that as the debt ratio is allowed to rise, the primary surplus required to eventually stabilize further growth becomes larger. That is, the cost of postponing adjustment is to mandate more contractionary fiscal policies in the future, prejudicing long-term growth. 2.55 A quantitative evaluation of the trade-offs involved is presented in Table 2.10 which shows the results of some simulation exercises in which sustainable deficits, measured inclusive and exclusive of interest payments, are calculated. The results depend on alternative long-run debt/GNP ratios (shown in each row), and on t47 differential between interest rates and growth rates (shown in each column).- For example, if the initial debt/GNP ratio is 140X, and the interest rate exceeds the growth rate by 2 percentage points, a primary surplus of 2.91 of GNP would be required to maintain the debt ratio unchanged. Lower surpluses could cause an increase in the debt ratio, while higher surpluses would lead to a decrease. These results make allowance for the fact that not all deficit financing leads to debt creation. A certain portion may be monetized in a growing economy with low but positive infla- tion. Another adjustment is made for valuation changes in external debt which can affect the debt/GNP ratio even in the absence of new flows of external borrowing. 2.56 The figures in the table illustrate the trade-off between short-run expansion and long-run contraction in fiscal policy. For any positive interest/growth differential, the higher the debt ratio is allowed to grow through expansionary short-run fiscal policy, the more contractionary the long-run primary surplus must be to eventually stabilize the debt ratio. The figures also demonstrate that when growth slows and interest rates rise, a major contraction in fiscal policy is required to prevent a surge in the debt/GNP ratio. This is what has happened, internationally as well as domes- tically, in the 1980s. There has been a shift from an environment where growth exceeded the interest rate by about two percentage points in the late 9/ These exercises are, of course, only indicative as they treat important variables, such as the interest rate and the growth rate, as given, whereas in reality both will interact with the level of the public deficit. See Vol. II, Chapter II for additional details on the methodology. - 31 - 1970s to one where the interest rate now exceeds growth by about the same margin. For debt/CNP ratios in the range 130-140X, the impact of this change in the economic environment is to mandate a six percentage point reduction in the public deficit/CNP ratio in order to avoid further debt build-ups. Table 2.10: SUSTAINABLE LONC-RUN PUBLIC SECTOR DEFICITS (percent of CNP) Interest-growth differential -2 -1 0 1 2 Debt/CNP ratio (2) (percentage points) 120 Primary deficit 2.9 1.5 0.2 -1.2 -2.5 Overall deficit 12.5 11.1 9.8 8.4 7.1 130 Primary deficit 3.1 1.6 0.2 -1.3 -2.7 Overall deficit 13.5 12.0 10.6 9.1 7.7 140 Primary deficit 3.3 1.7 0.2 -1.4 -2.9 Overall deficit 14.5 12.9 11.4 9.9 8.3 145 Primary deficit 3.4 1.8 0.2 -1.4 -3.0 Overall deficit 15.0 13.4 11.8 10.2 8.6 150 Primary deficit 3.5 1.8 0.2 -1.5 -3.1 Overall deficit 15.5 13.8 12.2 10.5 8.9 160 Primary deficit 3.7 1.9 0.2 -1.6 -3.3 Overall deficit 16.5 14.7 13.0 11.2 9.5 Note: Interest rate assumed fixed at 8%; inflation at 21; US dollar depreciation at 3Z. Allowance is made for some non-debt monetary financing consistent with inflation (seignorage) and for the valua- tion adjustments associated with currency realignments on external debt. Negative sign indicates a surplus. Source: Bank staff calculations. 2.57 The interest rate in Malaysia is now expected to exceed the growth rate by about 2 percentage points. To stabilize the debt ratio at current levels of 137Z of GNP, the 1988 deficit would have to be reduced to 8%. This seems to be an unrealistic target. A more reasonable phased deficit reduction should aim to stabilize the debt ratio by 1991 at a level of 145% of GNP. This would imply that the primary deficit should shift from its current level of -2.7X to a surplus of 3.0% by 1991--a substantial, but not unrealistic fiscal adjustment. 2.58 The simulation exercise demonstrates that the problems faced by the Malaysian Government in achieving fiscal adjustment, i.e., in reducing the deficit to a level that is sustainable without an exponentially increasing - 32 - debt ratio, have become more acute because of the slowdown in growth. At the same time, the slowdown in growth is related to the way in which the adjust- ment has been made. Adjustment can be achieved through tax increases, reduc- tions in subsidies or expenditure cuts in different areas, but each approach has a different impact on growth. In Malaysia, the bulk of adjustment was carried out by cutting expenditure, particularly development expenditure, with a disproportionate reduction in infrastructure spending. This was, however, the most costly form of adjustment in terms of foregone growth since infra- structure investment is an important determinant of private investment (see para. 3.87 below). As a result, the aggregate reduction in domestic demand was a multiple of the reduction in public investment. 2.59 From a growth perspective, the least costly way of reducing a deficit is by increasing government revenues and reducing subsidies, as long as these can be done in a nondistortionary fashion. Measures such as broaden- ing the tax base, eliminating exemptions, and increasing lump-sum or land taxes in line with property revaluation, are examples of such nondistortionary taxes. Unfortunately, the tax effort in Malaysia over the period 1982-87 has not been successful in these areas. For example, it is estimated that 75Z of domestic manufacturing production and imports are exempt from the sales tax, ane that the average effective rate is less than 2Z of the value of sales. More and more distortions have arisen through capital-cheapening investment allowances, discretionary granting of exemptions, and a narrowing of the sales tax base. These coexist with high statutory corporate income tax rates and development taxes. Some progress, however, has been made in the 1988 budget towards improving the contribution of the tax effort towards fiscal adjust- ment. 2.60 Maintaining expenditures and using higher revenues to cut the deficit helps growth because the private sei8yr is thought to have a lower propensity to spend than the public sector. - The same arguments suggest that cutting income transfers and subsidies to the private sector would be as effective as raising taxes, and potentially even better because such measures will often be associated with the removal of distortions that have been artificially created by public programs. As in the case of taxes, however, there has been little success in this regard until recently. For most of the 1980s, government outlays for items not directly related to real consumption or investment rose sharply despite declining interest rates. By 1987, only 60% of all government financial expenditure added directly to domestic demand for real goods and services. The remainder was dissip4ted in transfers, subsidies, loans, debt service and other financial obl-gations. Cutting back on these areas would be less damaging for growth then cutting direct expendi- ture. Rising Wages and Falling Profits 2.61 In the 1970s, Malaysia was most vulnerable to external price developments in its major agricultural primary commodity exports. However, 10/ The classic "balanced-budget multiplier" in Keynesian economic analysis. - 33 - mechanisms have evolved in the modern plantation sector to distribute the gains and losses from commodity price swings in an equitable fashion between profits and wages through explicitly incorporating productivity clauses and price bonuses in wage contracts. This has serv d to protect profits during a downswing and to reward workers during the upswing of commodity price cycles. 2.62 More recently, the price of petroleum has become a major component of the overall terms of trade. However, the built-in stabilizers present in agricultural commodity contracts are not prrsent in petroleum or gas produc- tion. The recent collapse in energy prices led to a sharp reduction in national income, concentrated in a decline in profits and indirect taxes. In effect, the Government absorbed the terms of trade losses, and has not yet distributed the burden over the rest of the economy. 2.63 While profits have been declining, wages have been increasing. The changing distribution of national income in favor of wages is shown in Figure 2.6. It is apparent that the wage share was roughly stable in the 1970s, at about 50Z of national income, but has since risen sharply, espe- cially in the last three years of declining or stagnant income. The data in Figure 2.6 is obtained as follows. Average compensation to wcrkers is obtained from the Ministry of Labor. This is multiplied by total employment taken from the Treasury's Annual Sconomic Report to give the wage bill. Indirect taxes are broken out from Federal Government revenue. Profits are derived as a residual, using the national accounts identity between GNP and the sum of wages, profits and indirect taxes. Such a measure imputes an average wage to all workers, including the self-employed. In this sense, the measures are not comparable to national accounts where no imputation of returns to labor for the self-employed is made. It is interesting to note, however, that national accounts figures also show a rise in the share of wages in value added from 38% in 1978 to 401 in 1983. Furthermore, profitability data from the Financial Survey of Companies reinforce the finding of a collapse in profitability after 1984. 2.64 In 1985 and 1986, both wage rates and the total wage bill continued to rise, despite growing imbalances in the labor market reflected in rising unemployment. Several factors influenced this rise. First, labor contracts, which are typically negotiated over a three-year cycle, still reflected the generous awards that were typical of the tight labor markets of the 1982-84 boom. Thus, although the marginal wage for new entrants turned down in 1985 compared to 1984, average real wages in all sectors (with the possible excep- tion of private services) continued to increase. Second, the high cost of retrenchment built into labor contracts implied a temporary surge in wage payments during the recession years when lay-offs were taking place. Third, other elements of inflexibility, such as fixed overtime and bonus allowances, EPF contributions and seniority provisions, also imparted an upward bias to wages. Finally, although new nominal wage increase have been held to minimal levels, the absence of significant inflatiorc has meant that workers have not suffered substantial eros;on in their real wages. The downward stickiness of nominal wages for existin, workers has precluded a rapid real wage adjustment except for new entrants. DECOMPOSMON OF GNP BY INCOME SOURCES Malaysi, 1972-87 80 - 74.2 Profit 71.8 70 - 69.8 Es Indirect Tax 65.2 65.9 wjage 59.7 60- 55.6 51.4 50- 40- C ~~~~~~~~~~~~~~36.2 31.1 30 - ~~~~~~~27.0 21.9 21.6 ooo, 20 -18.1 13.8 1 0 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 X rasc os~~~~1: - 35 - 2.65 The cyclical factors associated with the terms of trade decline in 1985-86 accentuated a structural trend of a rising wage share that had its roots in the early 19809. During this time, government jobs accounted for 20X of all new jobs created in the economy. As the average government real wage was consistently at least 25% above the economywide average, there was a shift towards a growing fraction of high-paid labor as the government sector expanded. Furthermore, the Government's countercyclical expenditure policy during 1980-84 reinforced the absorption of workers in labor-intensive sectors such as construction. As a consequence, the overall elasticity of employment with respect to value added in this period was greater than one. Wage gains also outstripped productivity gains. This combination of a greater growth in employment than growth in production and higher real wages has generated the rise in the wage share. 2.66 The wage share has also been affected by changes in government policy responses to external shocks. External shocks were formerly absorbed through flexible agricultural contracts, but with the underlying structural movement in the economy towards the volatile oil and gas sector and the greater emphasis on countercyclical spending, external price shocks have recently been absorbed by the Govcrnment. At an aggregate level, this is observed from the correlation between the wage share and the terms of trade index. For the period 1974-80, this correlation coefficient is 0.06-that is, the terms of trade had almost no effect on income distribution in the economy, a sound testament to the effectiveness of contracting techniques in the plantation sector. For the period 1980-87, however, the correlation between the terms of trade and the wage share is -0.61--that is, a fall in the terms of trade was strongly associated with a rise in the wage share, an indication of the protection afforded wage earners against external shocks. Such protec- tion, however, can only be temporary. As Government debt and debt service build up, the fiscal stance must contract and the terms of trade losses be passed through into the economy. 2.67 The corollary of a rising wage share is a falling share of profits. Around 1980, profits represented about one third of national pro- duction; by 1985-87 they had fallen to one fifth. The fall in profits is the major explanatory factor behind the recent fall in investment; changes in profits explain about 80X of changes in private investment (see paras. 3.83 ff. below). With the recovery of profitability in the second half of 1987 and 1988, following the terms of trade improvements, wage declines and the ringgit depreciation, there is good reason to expect a recovery of investment. To sustain this in the medium-term, however, will require more flexibility in wages if and when the external environment deteriorates again. The Mismatch between Risks and Rewards 2.68 For the economy as a whole, total investment must equal domestic savings plus foreign savings. The close relationship observed between profits and investment, however. does not imply an equally close relationship between domestic savings and investment. One major weakness apparent in the Malaysian economy is the seeming inability to transform high levels of private domestic savings into equally high levels of private domestic investment. Instead, private savings in the 1980s have been used to fund substantial public sector - 36 - deficits and, more recently, to permit a build up in external reserves aylto finance privatge sector capital outflows and foreign asset accumulation.- 2.69 A useful tool to analyze the final disposition of private savings is a flow-of-funds account. Although this is unavailable in sufficiently dis- aggregated form in official statistics, a flow-of-funds matrix can be compiled from data in a variety of sources. The disadvantage, of course, is that the data may not be mutually consistent and, for some items, may not exist from primary sources. In the latter situation, estimates can sometimes be derived by the residual method, utilizing the property of the matrix that total sources of funds must equal total uses. This, of course, is not a fully satisfactory solution, as it tends to accumulate all data errors and inconsis- tencies into a single entry. The numbers shown in the flow-of-funds tables included here should therefore be treated as illustrative of general trends rather than as definitive in their own right. Detailed data sources and a description of the methodology used in compiling the tables are presented in Appendix 2. 2.70 The overall flows-of funds for 1980, 1985 and 1986, respectively are presented in Tables 2.11-2.13. The year 1980 was taken as the most recent period before serious internal and external imbalances developed and serves as a proxy for a "normal" year. More recent years, 1985 and 1986, represent a period of distress, especially for the private corporate sector. The differ- ence between the flows in the early and recent periods shows the impact of adjustment policies and external shocks oa domestic financial intermediation. 2.71 In the flow-of-funds tables, the economy is divided into three sectors: public, private non-financial, and financial intermediaries. Where possible, these are further subdivided: general government and public enter- prises; corporations and households; the banking system, provident funds and other nonbank financial intermediaries. This last category includes insurance companies, farmers' organizations, the Permodalan Nasional Berhad (PNB), Tabun Haji (Pilgrim's Fund), building societies and cooperative, development and savings banks. An external sector is also added to complete the circulation links. 2.72 For each sector of the economy, estimates are made for savings and fixed capital formation. The difference bettween these items represents a net accumulation of financial assets which can be further broken down into changes in loans or deposits with different financial intermediaries, and holdings of cash, securities, equity and other financial instruments. 2.73 The major trends between 1980 and 1985-86 are striking. The first row in the tables shows that gross national savings inc-eased perceptibly over 11/ The foreign asset accumulation by the Malaysian private sector is a normal response in an economy with open capital markets, to the desire for diversification. The ratio between domestic and foreign assets will change according to interest rate and exchange rate expectations, leading to capital outflows or inflows according to underlying conditions. Table 2.11: FLOW OF FUNDS4 OVERVIEW, l980 (M$ btllion) Private Public sector nonfinancial sector FNnancial Intermedlaries Cor- Banking Provident External Gross Government NFPEs /a porations Households system fund Other sector Change national Use Source Use Source Use Source Use Source Use Source Use Source Use Source Use Source in stock Total savings Nonfinancial Savings - -2.6 - - - 10.5 - 7.5 - 0.0 - 0.0 - 0.0 - 0.8 - 16.2 15.4 Capital outlays 6.2 - - - 8.9 - 1.5 - 0.0 - 0.0 - 0.0 - - - -0.4 16.2 - Financial Savings-capital inv. -8.8 - - - 1.6 - 6.0 - 0.0 - 0.0 - 0.0 - 0.8 - - -0.6 Net financial Inv. -8.8 - - - 1.6 - 6.0 - 0.0 - 0.0 - 0.0 - 0.8 - - -0.6 Total Financial Uses and Sources 1.2 10.0 - - 9.2 7.6 8.3 2.3 10.0 10.0 2.0 2.0 2.2 2.2 3.1 2.3 - - Depgults 8anking system 1.0 - - - 2.1 - 2.2 - - 6.0 0.3 - 0.4 - - - - Provident funds - - - - - - 2.0 - - - - 2.0 - - - - - Other - - - - - - 1.3 - - - - - - 0.8 - 0.5 - Currency - - - - - - 0.6 - - 0.6 - - - - - - - Loans Banking system - 0.1 - - - 5.3 - 1.4 7.0 - - - - 0.2 - - Provident fund - - - - - 0.1 - - - - 0.4 - - 0.3 - - Other financial - - - - - - - 0.2 - - - - 0.2 - - - Direct - 1.2 - - 1.3 - - - - - - - - 0.1 - - Government - - - - - - - - - - - - - - - - Foreign - 1.6 - - - - - - - 1.0 - - - - 2.6 - Securities Banking system - 1.0 - - - - - - 1.0 - - - - - - - Provident fund - 1.1 - - - 0.1 - - - - 1.2 - - - - - Other financial - 0.2 - - - 1.3 - - - - - - 1.5 - - - Direct - - - - 0.5 - - - - 0.5 - - - - - Equi ty Banking system - - - - - - - - - - - - - - - _ Provident fund - - - - - 0.1 - - - - 0.1 - - - - - Other fLnancial - - - - - - - - - - - - - - - Government 0.2 - - - - - - - - 0.2 - - - - - - Direct - - - - - 0.3 0.7 - - 0.5 - - 0.l - - _ Other Domestic - 4.8 - - 4.8 - 1.5 0.7 0.7 0.7 - - - 0.8 - - Foreign - - - - 0.5 - - - 1.3 0.5 - - - - 0.5 1.8 Stock change - - - - - 0.4 - - - - - - - - - _ /a A breakdown by NFPEs is not available for 1980. These are Included in the corporate sector. Table 2.12: FLOW OF FUNUS: OVERVIEW, 1985 (M$ billion) Private Public sector nonflnancial sector Firancial intermediaries Cor- Banking Provident External Gross Government NFPEs porations Households system fund Other sector Change national Use SourcUs ese SourcSource ource Use Source UTs- Source Use Source Usou rcurce Use Source in stock Total savings Nonfinancial Savings - 1.4 - 5.5 - -3.0 - 15.7 - - - - - - - 1.7 - 21.3 19.6 Capital outlays 7.2 - 5.0 - 7.4 - 2.4 - 0.4 - 0.2 - 0.5 - - - -1.8 21.3 - Financial Savings-fixed inv. -5.8 - 0.5 - -10.4 - 13.3 - -0.4 - -0.2 - -0.5 - 1.7 - - -1.8 Net financial inv. -5.8 - 0.5 - -10.4 - 13.3 - -0.4 - -0.2 - -0.5 - 1.7 - - -1.8 Total Financial Uses and Sources 1.3 7.1 4.0 3.5 - 10.4 15.7 2.4 10.2 10.6 3.9 4.1 2.9 3.4 7.1 5.4 - - Deposits Banking system 0.5 - 1.2 - - - 3.7 - - 5.8 -0.2 - 0.6 - - - - Provident funds - - - - - - 4.1 - - - - 4.1 - - - - - Other - -- - 3.9 - 0.1 - - - - 1.0 3.0 - - Currency - - - - - - 0.2 - - 0.2 - - - - - - - Loans Banking system - 0.1 - -0.7 - 5.7 - 2.4 7.5 - - - - - - - Provident fund - - - 0.1 - - - - - - 0.6 - - 0.5 - - Other financial - - - - - - - - - 0.2 - - 0.2 - - - Direct - - - - - - 0.4 - - - - - - 0.4 - - Government - - - 0.9 - - - - 0.9 - - - - - - - Foreign - 1.0 - 1.0 - 0.2 - - - - - - - - 2.2 - Securities Banking system - -1.9 - 0.1 - - - - -1.8 - - - - - - - Provident fund - 3.4 - 0.1 - - - - - - 3.5 - - - - - Other financial - 0.4 - 0.5 - 1.2 - - - - - - 2.1 - - - Direct - 1.8 1.8 - - - - - - - - - - - - - Equity Banking system - - - - - - - - - - - - - - - - Provident fund - - - - - - - - - - - - - _ _ _ Other financial - - - - - - - - - - - - - - - - Government 0.8 - - 0.6 - - - - - 0.2 - - - - - - Direct - - - 0.3 - 0.3 1.9 - - 0.8 - - - 0.5 - - Other Domestic - 2.1 2.1 - - 0.0 1.5 - - 0.5 - - - 1.0 - - Foreign - 0.2 -1.1 - - 1.8 - - 3.5 2.9 - - - - 4.9 2.4 Stock change - - - 0.6 - 1.2 - - - - - - - - - - Table 2.13: FLOW OF IFUNDS: OVERVIEW, 1986 (M$ billion) Private Public sector nonfinancial sector Financial intermediaries Cor- Banking Provident External Gross Government NFPEs porations Rouseholds system fund Other sector Change national Use Source Use Source Use source Use Source Use Source Use Source Use Source Use Source in stock Total savings Nonfinancial Savings - -0.2 - 2.2 - 0.1 - 14.9 - - - - - - - 0.8 - 17.8 17.0 Capital outlays 7.4 - 3.8 - 5.6 - 1.6 - 0.2 - 0.2 - - - - - -1.0 17.8 - Financial Savings-capital inv. -7.6 - -1.6 - -5.5 - 13.3 - -0.2 - -0.2 - - - 0.8 - - -1.0 - Net financial inv. -7.6 - -1.6 - -5.5 - 13.3 - -0.2 - -0.2 - - - 0.8 - - -1.0 - Total Financial Uses and Sources 0.4 8.0 1.0 2.6 -1.0 4.5 16.1 2.8 9.0 9.2 4.0 4.2 3.5 3.5 6.5 5.7 - - - Deposits Banking system -0.1 - 0.8 - -1.0 - 3.3 - - 3.3 0.1 - 0.2 - - - - - - Provident funds - - - - - - 3.9 - - - - 3.9 - - _ _ - _ _ Other - - - - - - 3.3 - - - - - - 1.2 - 2.1 - - - Currency - - - - - - 0.4 - - 0.4 - - - - - - - - - Loans Banking system - 1.3 - - - 0.3 - 1.9 3.3 - - - - -0.2 - - - - - Provident fund - - - - - - - - - - 0.2 - - 0.2 - - - - - 0 Other financial - - - 0.4 - 0.6 - 0.9 - -0.1 - - 1.8 - - - - - - Direct - - - - - - 0.4 - - - - - - 0.4 - - - - - Government - - - 0.8 - - - - 0.8 - - - - - - - - - - Foreign - 1.2 - 0.2 - 0.2 - - - - - - - - 1.6 - - - - Securities Banking system - -0.1 - - - 0.1 - - - - - - - - - _ _ _ _ Provident fund - 3.4 - 0.1 - 0.1 - - - - 3.6 - - - - - - - - Other financial - 0.4 - 0.3 - 0.5 - - - - - - 1.2 - - - - - - Direct - 1.3 1.3 - - - - - - - - - - - - - - - - Equity Banking system - - - - - - - - - - - - - Provident fund - - - - - 0.1 - - - - 0.1 - - _ _ _ _ _ _ Other financial - - - - - - - - - - - - - - - - - - - Government 0.5 - - 0.3 - - - - - 0.2 - - - - - - - - - Direct - - - - - 0.2 -0.6 - - 0.5 - 0.3 - -1.6 - - - - - Other Domestic - - - 0.2 - 0.3 5.4 - 0.2 1.9 - - 0.3 3.5 - - - - - Foreign - 0.5 -1.1 - - 1.4 - - 4.7 3.0 - - - - 4.9 3.6 - - - Stock change - - - 0.3 - 0.7 - - - - - - - - - - - - - - 40 - the period, as expected in a time of rising national income. The general government marginally increased its "savings", from minus M$2.6 billion to plus M$1.4 billion in 1985 and minus M$0.2 billion in 1986. These figures include adjustments made for the portion of development expenditure that does not represent capital formation, and for other balancing items such as profits of Bank Negara. They therefore differ from the standard figures for public savings that reproduce the overall public sector's operating surplus or deficit. The corporate sector, however, showed a sharp decline in savings. In 1980 the NFPE surplus was only about M$0.3 billion, but the private corpo- rate sector had savings of about M$8.7 billion. In 1985 and 1986, however, private firms appear to have been dissaving--that is, running down financial assets by more than they increased capital formation. In contrast, households on average doubled their savings in 1985/86 relative to 1980. 2.74 The substitution of household savings for corporate savings is a distinctive feature of the 1980s. It is consistent with the trend noted above towards a rise in the share of wages in national income. It is also sympto- matic of a period when the returns generated by financial intermediaries, such as EPF, PNB and LUTH, have been consistently high and positive in real terms. These returns accrue to the households owning unit trust shares; their increase over the period has added substantially to household income and saving. 2.75 Associated with the fall in corporate savings is a fall in corporate investment. In 1980, the private corporate sector was able to fully fund its own investments. While making significant use of the banking system, the corporate sector deposited and withdrew roughly balanced amounts. By the mid- 1980s, they were heavy net recipients of funds. Furthermore, throughout the 1980s, formal equity markets played a very small role in raising capital. The implication is that corporate balance sheets must have deteriorated signifi- cantly over this period, making it increasingly difficult for lenders to find commercially viable companies without a heavy overhang of existing debt. 2.76 As the source of savings shifted, the structure oc financial inter- mediation has also adjusted. Despite a 40% rise in nominal incomes, the banking system intermediated the same nominal amount of new flows in 1985/86 as in 1980--around M$10 billion. In 1980, this represented 70% of total new financial intermediation; by 1986 it had fallen to 55%. The big gainers were the EPF, which doubled the size of its annual flows in the five-year interval, and other intermediaries, which grew by 50%. The latter trend illustrates the growing importance of deposit-taking cooperatives. Recent scandals and financial failures among these institutions attest to the danger of rapid realignment in financial structures prior to the establishment of proven systems for monitoring and regulation by the Central Bank. 2.77 The rationale behind the growth of EPF and other nonbank financial companies and the stagnation of the banking system is explained by comparing the returns to individual investors in these entities. Table 2.14 compares the annual dividend for PNB/ASN and EPF with the savings deposit rate of commercial banks, the total return on the Kuala Lumpur Stock Exchange (KLSE) and GNP growth, which serves as a proxy for some other diversified investment portfolio. In terms of the size of the return and the stability of the yield, - 41 - the trust funds clearly offer preferred investment instruments. Investors, particularly those in PNB/ASN, have been well protected against the downturn in the economy, but at the same time benefited handsomely in the boom years of 1981-84. In addition, the trust funds have strong backing from the Federal Government, further reducing their risk, and legislated privileges which permit them access to profitable transactions at minimal risk (for example, PNB's take-over prerogatives). Table 2.14: NOMINAL RATE OF RETURN BY INSTITUTION, 1981-87 (X) 1981 1982 1983 1984 1985 1986 1987 PNB/ASN 20.0 18.0 18.0 17.2 17.2 10.0 10.0 EPF 8.0 8.0 8.5 8.5 8.5 8.5 8.5 Commercial bank Savings deposits 7.0 6.5 6.0 7.5 6.0 6.0 4.0 20 + year government securities - - - 8.6 8.6 8.6 7.6 KLSE 13.9 -18.9 40.8 -19.3 -17.6 12.8 0.5 (ONw dividend) (4.6) (4.6) (3.0) (5.1) (5.5) (4.7) (4.1) Memo: GNP growth 8.2 7.4 9.2 13.9 -3.2 -8.3 5.9 Source: Economic Report of the Treasury; PNB and EPF Annual Reports; Kuala Lumpur Stock Exchange. 2.78 Government's involvement in the trust funds, particularly PNB, is closely tied to the broader objectives of the NEP, under which funds are mobilized from the Bumiputera community to purchase assets in trust for the community. However, the policy of implicitly guaranteeing high returns at low risk, which was thought necessary for such mobilization, has built up the trust funds at the cost of introducing distortions in the structure of finan- cial intermediation. It also seems likely judging from the flow-of-funds tables, that this policy has diverted funds that would otherwise have gone into the banking system rather than creating a pool of new savings. The common practice of banks' extending loans to individuals for the purpose of purchasing ASN units illustrates the same tendency. 2.79 The major implication of the shifting structure of funds is that the supply of risk-capital to corporations is being squeezed. Nonbank financial institutions such as pension funds and insurance companies have a strong - 42 - proclivity to invest in low-risk assets such as government bonds. The EPF has also established a small housing finance subsidiary (MBSB) to enter the lucrative mortgage market. The commercial banking system has also strongly increased its mortgage activity, reflected in a rise in lending to households in the flow-of-funds tables. In 1986, however, the banks expanded their loans to the corporate sector by only M$300 million. Liquidity is available--the difficulty is in finding creditworthy borrowers prepared to pay a sufficient premium over bench-mark long-term Treasury bonds to cover the additional risk. 2.80 Interest rates in competitive capital markets, such as those in Malaysia, are primarily determined by two factors: the risk-free interest rate and a premium that must be added to compensate for the riskiness of any particular loan. In Malaysia, the risk-free interest rates on medium- and long-term bonds have been maintained at attractive positive levels which, despite the surge in funds seeking low-risk investments, have not come down because of the administrative process for determining rates in the absence of a well-functioning secondary market in government securities. This in turn raises the cost of funds for all other medium-term loans. Meanwhile, the shift of funds to institutions less willing to take risk has increased the premium charged on risky assets. As a result, funds are available for low- risk activities--witness the dramatic fall in short-term lending rates and interbank loans in 1987--but not for higher risk activities such as capital formation in manufacturing. 2.81 In the future, as the EPF continues to expand, the mismatch in the flow of funds will likely worsen unless corrective measures are instituted. As the Government brings its deficit under control and starts to issue even fewer securities and at lower rates, the rising cash surpluses of the nonbank sector will be hard pressed to place their funds in prudentially sound areas. Yet institutions such as the Provident Fund cannot be expected to become venture capitalists; this is too speculative and, as occurred in the Philippines, can compromise the financial integrity of the institution and of individuals' retirement benefits. Instead, sound institutions such as the development banks should be encouraged to tap these funds through their own bond issues and to be more aggressive in taking on commercial risk for an appropriate premium. Structural Unemployment 2.82 The stubborn increase in the unemployment rate since 1982 suggests that market mechanisms on their own may be insufficient to restore full employment. Furthermore, the fact that unemployment has continued to rise in years of rapid growth (1983, 1984 and 1987) as well as years of slower growth, indicates that the problems are of a structural nature combined with cyclical factors associated with the recession. In such an environment, classical Keynesian-type measures to reduce unemployment by stimulating demand are 12/ This is appropriate, given the prudential obligations of these entities. In the case of EPF, governing regulations require at least 70% of total investments be held in government securities. - 43 - unlikely to be successful and may even disrupt and slow-down equilibrating market forces. It is unlikely that higher demand will make an immediate significant dent in unemployment because (i) high costs of retrenchment have created a situation of labor hoarding in the work-place, so increased output would be met through greater productivity not increased employment and (ii) search activities would remain high. 2.83 Care must be taken in interpreting unemployment numbers in Malaysia, especially when making international comparisons, because of differences in coverage across countries. In Malaysia, the unemployed include those not actively looking for a job, but who would be prepared to work if the right opportunity (or wage) should arise. These account for over one-half of current unemployment. They are mostly recent school-leavers; the incidence of unemployment is highest for those completing secondary education, and lower for those with no formal education and primary school leavers. This latter group are largely from poor households and cannot afford to be unemployed. Despite the considerable attention focussed on them, unemployment rates are also low for college graduates. Overall, the unemployed tend to be young (about 75X aged less than 25), reasonably skilled, and prepared to wait for appropriate work. The incidence of unemployment is almost the same in rural and urban areas. 2.84 Some indication of the nature of unemployment can be gleaned from the Labor Force Surveys. About 80Z of the middle and upper secondary school leavers were "waiting for answers" to previous job enquiries, rather than giving as reasons for their unemployment "unqualified" or "no suitable job available." Given that tne duration oi unemploymenL is lusi than si wnths in two-thirds of the cases, and that a very small proportion of unemployment (perhaps 1X) is due to retrenchment of existing workers, the data strongly suggests that the problem is not one of an aggregate shortage of jobs, but rather of extended "search" as adjustment in wage expectations and specific job opportunities occurs. 2.85 There are three reasons why unemployment has risen in the past five years: (i) a fall in demand for labor, a consequent development of a two-tier labor market characterized by a large disparity in wages for similar occupa- tions and, hence, a need to explore alternative job opportunities; (ii) a demographic bulge in the increase of young entering the labor force amid bottlenecks in the higher education system preventing the absorption of 18 to 19 year-olds into universities; and (iii) a reluctance by the young to accept work at wages half of those received by a five-year older cohort. 2.86 A disparity amongst wages for similar jobs primarily reflects the timing of entry into the labor market. Given the very low inflation rates that have prevailed recently, even small or zero increases in the nominal wage have been insufficient to bring down the average real wage of workers hired before 1984. These workers were able to obtain relatively high wages because of a temporary tightness in the labor market following the construction boom and rapid expansion in government. With the onset of a recession and the decline in the terms of trade in 1985-86, it was necessary for real wages to fall to share the burden of income losses. The burden of adjustment, however, fell almost entirely on new entrants to the labor force as incomes of those in - 44 - existing jobs were protected through contracts. The upshot has been the emergence of a two-tier labor market, wherein workers employed on the basis of salary scales prevailing before 1984 have substantially higher wages (about double, abstracting from the three years experience differential) than current new entrants for equivalent work. 2.87 Not all firms are prepared to implement two-tier compensation systems. Problems in terms of worker morale and low productivity can arise. In Government, abandoning the principle of equal-pay-for equal-work runs counter to a central philosophy of promoting equity. Thus, a differential in wages across jobs has arisen, and the young spend time in searching for the best paid job. Given the existing rates of retirement, illness/death, and other factors affecting the turnover of "old" workers, the wedge between wages is likely to persist for some time. The wedge is also sustained by the high contractual costs of firing (on average six months pay), that prevents firms from substituting low wage new workers for more expensive old workers. 2.88 In normal circumstances, the education system helps keep down the need to search for appropriate work by matching the skills required by indus- try with those supplied to the student body. The education system, however, has been unable to keep pace with the rapid change in job markets. One recent feature, for example, has been a sharp compression of wages across skill levels, eroding the profitability of secondary education and vocational training, but these programs have been slow to adapt curricula. The current educational structure is geared to a regime of relative wages that has been superseded by recent economic developments. Furthermore, the slowdown in the rat- z. go-wth of uniL.ersity space bas exacerbated a bulge in the rate of entry to the labor force of higher secondary school leavers and a correspond- ing unemployment incidence for this group almost four times higher than that for new college graduates. 2.89 Given the definition of unemployment in Malaysia, even those not actively looking for a job are counted as unemployed. In many instances, low wages are likely to be a natural deterrent (the "discouraged worker" effect). Hence, it is not surprising that unemployment should rise in a period of declining wages, particularly when the difference between initial wage expectations and the realities of the job market is as great as it is today. As wage expectations gradually adjust, however, unemployment numbers will also show a decline. 2.90 Although unemployment is expected to decline gradually as strong growth and wage adjustments act together to boost labor demand, government can assist in speeding up adjustment and in creating more resilient and flexible labor markets in several ways. The most immediate measure would be to reduce the wedge between old and new workers. One attractive way of achieving this is to eliminate EPF contributions (on the part of both employees and employers) for those under twenty-five. The impact of this is likely to be to increase firm demand for labor, to increase the wage received in hand by the employee, raising his willingness to work, and to speed up the process of assimilating young workers into the labor force. The disadvantage, from the point of view of public finances, would be a reduced net investable surplus in the EPF. This would, however, be small--only about 10.9% or M$500 million. - 45 - 2.91 This measure should not be viewed simply as a countercyclical device, although its adoption at present will have a timely effect in stimu- lating demand for young workers. It has additional long-term merit. One consequence of the growth/interest rate developments that have taken place is that the profile of the expected earnings stream of a worker has been shifted; a greater portion of life-time income is being made available after retire- ment, while a correspondingly smaller proportion is available at younger ages. In fact, under current circumstances, a worker who enters the labor force at age sixteen, and receives normal incremental wage increases as his experience grows, is likely to hf37 as much, or more, income in retirement as at any time in his working life.-, As expenditures are typically much lower during retirement than in middle-age, particularly as health care is free in Malaysia, the current contribution rates of the Provident Fund are excess- ive. A policy which exempted young workers up to age 25 would create an earnings time-profile more in line with desired life-cycle expenditures. 2.92 A second measure to reduce structural unemployment would be to give to EPF a mandate to use its funding and institutional capability to promote workers' training and education. One approach is to simply allow members to borrow, up to a certain fraction, against their accumulated contributions to finance training in selected courses, both formal and on-the-job. An alterna- tive is to make funds available to employers with the provisions that: (a) the worker agrees with the training program and (b) if the worker is laid off following training, he should be reimbursed an amount proportional to the funding provided and the time elapsed between training completion and retrenchment. A specific proposal in this regard is to allow enterprises to deducL 5Z ox' the cost of tcaining from the annual contributions to individual retirement accounts which must be made under present law. As the worker receives the benefits of training in the form of higher wages, he stands to gain considerably from participation in firm-sponsored training programs. Simulations indicate that this would not be expected to alter the EPF surplus to any significant extent. A 25 year-old worker who withdrew one year's total EPF contribution (i.e., 20% of earnings) to invest in training with a rate of return of 15%, would have a retirement income less than 1% lower than without undertaking the initial EPF withdrawal. 2.93 The use of workers' provident, pension and social security funds to finance training and education is not new. Singapore, the Philippines and, in a somewhat different institutional context, Mexico all have such provisions. The idea is to create incentive structures that would make it in the employers' and workers' self interest to rely less on retrenchment to adjust to changes in demand and technology and to depend more on continuous training and automatic adjustment of labor compensation with productivity change. In essence, the EPF scheme would help reduce the risks associated with on-the-job training faced by both workers and their employers. 13/ Retirement income is estimated to be 20-year annuity value of accumulated life-time EPF contributions and interest earned thereon, based on retire- ment at age 60 (see Appendix 3). - 46 - 2.94 Over the medium term, unemployment will only be held down by a wage compensation system that is more flexible than the current one. One way of introducing flexibility is to amend the Employment Act to allow firms to offer profit-based or productivity-based overtime and bonus pay in lieu of the legally imposed fixed rate. In this way, the total compensation package can be adjusted more quickly as macroeconomic conditions warrant (see Volume II, Chapter 3). E. Overall Assessment 2.95 In the 1980s, Malaysia was faced with unprecedented commodity price movements, a general slowdown in international trade and growth and high real interest rates. At the time, the Government's strategy was based on an aggressive expansion of public investment, both in infrastructure and in directly productive activities. This quickly became financially untenable as large deficits emerged in the external current account and in fiscal accounts. High inflation imported from the rest of the world into Malaysia's open economy generated further financial instability. Since 1982, the Govern- ment's policy dilemma has been how to cure the external and internal financial imbalances without derailing economic growth. The various combinations of fiscal and monetary policy tried have been highly successful in bringing down inflation rates and restoring the current account to surplus. The cost, however, has been a low and uneven growth pattern in the last few years, ballooning public debt, and a worrying consistent decline in real private investment. 2.96 By the standards of other developing countries, Malaysia has weathered the shocks of the 1980s remarkably well. As a major primary commo- dity exporter, the magnitude of the external shocks with which it has had to deal have been disproportionately large relative to national income. Yet by its own standards set in the 1970s, current economic performance has been weak. Then, too, the country was hit with substantial external shocks, but these barely interrupted an impressively stable grcowth path. Why the difference? 2.97 Our analysis indicates that changes in the source of external shocks, in government policy responses, and in the external and internal environment all played a part. In the 1980s, the Government may have over- reacted to external shocks, underestimated response lags in the economy and miscalculated the size of short-term/medium-term trade-offs, thereby accen- tuating rather than dampening the cycles. As both product and factor markets seem to operate well in Malaysia, albeit slowly, price adjustments may have provided a more effective mechanism for equilibration if left to operate in a stable macroeconomic environment. 2.98 One clear lesson from the 1980s is that countercyclical public expenditure cannot be used to raise growth rates permanently, and it can be used to increase current growth only at the expense of future growth. Higher expenditure in one year must be offset by reduced expenditure (or higher taxes) in future years if public finances are to remain manageable. Given this, large changes in public expenditure are undesirable, with efficiency losses likely both during upswings (when poor project selection, overdesign - 47 - and inadequate supervision cause waste and cost overruns) and downswings (when important expenditures, such as for maintenance are cut back). Malaysia is now paying for the higher growth it enjoyed in 1981-84 by having to reduce public expenditure. In the future, the cost will become even more apparent as the deficit is finally brought under control. 2.99 The high cost of past countercyclical policies is due, in part, to recent changes in the international environment. The cost of such policies is minimized when growth rates are high and interest rates are low. Then, the future reduction in government spending is not so burdensome, relative to total demand in the economy, and does not have to be so large to accommodate debt service needs. Currently, both these variables-growth and interest rates-are at disadvantageous levels, and therefore a countercyclical policy is less attractive and should be less used. 2.100 Another factor reducing the effectiveness of countercyclical policy is the growing degree of unpredictability in commodity prices. Knowledge about whether a price ci.snge is permanent or temporary and about how long a cycle is likely to last is a key ingredient in implementing an effective countercyclical program. Yet this information is increasingly more difficult to assess. Commodity prices, which in the past have been closely correlated with Winess cycles in the OECD, have gradually broken free of this relation- ship._ Petroleum prices have also grown more volatile. Taking a conserva- tive posture with respect to comodity prices would seem to be a more prudent policy than fine-tuning a response. In addition, because terms of trade cycles are so unpredictable and have such important financial consequences, the use of the development budget as the major instrument for macroeconomic adjustment to external price shocks is inappropriate. Other policy tools, including new instruments for debt management (see below), could be useful. 2.101 Efforts by the Government to respond to external shocks through changes in real expenditures have had undesirable side effects by creating inflexibilities in the macroeconomy and postponing adjustment. One example of this inflexibility is the evolution of wages. A combination of factors, including expansionary public expenditure, the Government's role as a wage leader, downward stickiness in nominal wages and low inflation rates, have led to a surge in wages in the key manufacturing sector which is far beyond the increase in productivity levels. This has also occurred in other countries, like Korea and Singapore, but to a lesser degree. In Korea, however, rising domestic unit labor cost" have been effectively offset through exchange rate policy, preventing any erosion in international competitiveness. In Malaysia, by contrast, the real exchange rate has moved along with the US dollar/ Japanese yen exchange rate, rather than in response to terms of trade, income growth or international competitiveness. The recession in Malaysia was compounded by the fact that relative to Korea, Thailand and Hong Kong, Malaysia's competitiveness had eroded by 50% in just four years. 14/ P. Drucker, "The Future of the World Economy," Foreign Affairs, 1986. - 48 - 2.102 A corollary of the spurt in wages was a rise in the share of wage income in total national income and a significant decline in profitability. This was supported by the Government's countercyclical policy and reinforced by structural policies designed to further NEP objectives. Through expansion in nonfinancial public enterprises, promotion of financial institutions such as PNB and EPF to mobilize funds from small savers, and other mechanisms, the Government has favored wage-earners and savers and penalized corporations and entrepreneurs. Other structural policies such as the increase in EPF contri- bution rates also contributed to the rise in the wage share. 2.103 While temporarily protecting wage incomes in the shortrun in the face of declining national izicome, government policies have led to a fall in invest- ment, an inefficient pattern of resource allocation and a reduction in indivi- dual welfare. 2.104 Private investment has collapsed since 1984 despite a reduction in real interest rates, ample credit in the economy and a liberalization of government licensing requirements and bureaucratic procedures. The most important deter- minants of private investment are profits and public sector infrastructure, both of which have declined dramatictlay since 1984. Profits have been squeezed by falling national income - and a rising wage share, while public investment in infrastructure has borne the brunt of efforts to reduce the consolidated public sector deficit. 2.105 At the same time, scarce investment resources are also being less efficiently allocated. Private savings have shifted from the retained earn- ings of the corporate sector to savings by individual households. Additional intermediation is required to transform these savings into productive invest- ment. In practice, and to a very significant extent in 1987, domestic private savings have not been channelled into domestic investment but into capital flight and the accumulation of private foreign assets. This has resulted in part because financial intermediation is increasingly taking place through nonbank sources, such as provident and mutual funds, insurance companies and cooperatives, and less through the banking system. 2.106 Some degree of inefficiency is inevitable whenever attempts to restruc- ture society and redistribute income are made. If the redistribution is successful, however, the benefits of stability may well outweigh the costs involved. Nevertheless, it seems unlikely that the Government's policy of protecting wage earners' incomes from external shocks in the 1980s has increased substantially the general welfare. A strong indication of this is that the Malaysian economy has not witnessed a major boom in real private consumption, despite the growth in wages. Instead, real per capita consump- tion steadily dropped through the 1980s, and has been just as volatile as GNP. 15/ The fall in national income may be attributed to a decline in external commodity prices and to a weakness in domestic demand following reductions in commodity prices and government expenditure. The latter will appear as falling sales and slower orders for firms. - 49 - 2.107 Overall, the challenge of converting financial successes in mobilizing domestic resources and restoring health to the balance of payments into a basis for stable medium-term growth and employment generation still remains. Government is, however, to be commended for its macroeconomic and micro- economic adjustment efforts. The latter include liberalization of industrial licensing, easing of administrative procedures, strengthening of bank regula- tion, privatization and tax changes, all designed to reduce the cost to firms of doing business. Detailed sectoral and subsectoral plans have also been formulated to coordinate public and private actions. With these measures already in place, the rewards to overcoming the remaining structural weak- nesses highlighted above could be considerable. - 50 - III. MEDIUM-TERM GROWTH PROSPECTS 3.1 Although it is now virtually certain that Malaysia will be unable to achieve the targets for income and employment envisaged in the Fifth Malaysia Plan, there is considerable scope for regaining satisfactory income growth extending well into the 19909. The principal source of growth must be the industrial sector led by manufacturing. Rapid growth in this sector, however, is conditional on three develypments. First, the trend in total factor pro- ductivity (TFP) growth rates,t which were negative in the early 1980s, must be reversed and the rates restored to their positive levels of the 1970s. Second, private investment in manufacturing, both from domestic and external sources, must be resurrected. Third, demand for a broad range of manufactured products must emerge. This implies broadening the export base beyond elec- tronics and textiles, and revitalizing domestic demand. 3.2 New initiatives should be taken in each of these areas. A restora- tion of efficiency requires further policy initiatives in privatization, in removal of distortions in effective protection rates, in industrial restruc- turing at the subsector level and in firm-specific training and skills upgrad- ing. To energize private investment, profitability must be restored and the exchange rate/wage/productivity relationship carefully monitored to maintain international competitiveness. Lastly, if consumer demand is to play a lead- ing role, a greater share of income needs to be retained as discretionary cash-in-hand rather than as forced savings. 3.3 With suitable policy adjustments, the government is in a good posi- tion to reduce the uncertainties surrounding future growth. The advantageous short-term prospects for Malaysia, based on rising commodity prices and residual strength in global economic growth, should provide a valuable breathing space which would allow for systematic and gradual policy innova- tions. The greatest dangers to fulfilling the country's growth potential lie in impatience and complacency. Impatience could steer the authorities into an ill-advised expansion in fiscal policy in an effort to reduce unemployment or spur private investment, if these do not improve quickly in spite of moderate growth; while complacency in the face of an accelerated short-term recovery could postpone adjustment and leave the economy vulnerable to a future commo- dity price downturn or global recession. 3.4 The discussion of medium-term growth prospects that follows looks first at the potential for aggregate output from a sectoral perspective and then at an aggregate level, from an analysis of productivity growth and new capital formation. Policies to strengthen performance in these areas are identified. Next, the demand side is considered, particularly the prospects for private consumption and government spending. Finally, supply and demand 1/ The total factor productivity growth rate is a measure of efficiency gains in production. It is measured as the residual between the actual growth in output and the contribution to output growth of growth in capital, labor and intermediate inputs. - 51 - are linked together, and placed in the context of international economic developments, in order to arrive at a consistent set of macroeconomic projec- tions. A. Growth Potential -- The Supply Side 3.5 In the short-run, growth prospects are moderate to good. A conser- vative estimate for 1988 would be in the range 6-7X. Agriculture is unlikely to repeat its strong performance in 1987, but should grow by 2-3X, largely on the basis of greater rubber, palm oil and cocoa output that should follow the recovery in commodity prices and run-down in buffer stocks. Padi production, however, is likely to remain depressed. Mining output should also increase sharply as a result of the decision to raise petroleum extraction from 494.2 to 540.0 thousand barrels per day. This, in and of itself, will add one per- centage point to the 1988 growth rate. It does not, however, contribute to higher sustainable growth as petroleum production will plateau out at the new level. 3.6 Manufacturing growth holds the key to future performance. It is reasonable to expect potential growth of at least 10% for this sector over the medium-term. During the 1987 recovery from recession, manufacturing growth was brisk, at about 12Z. Much of this, however, is attributable to the very low levels to which output had sunk during 1986 and consequently cannot be sustained. Sectors such as wood and iron and steel benefited substantially from this rebound phenomenon and in future would grow much more slowly. Some export-oriented sectors, including semiconductors and rubber products, also enjoyed a sudden suree in demand and output of over 30X that may not be replicable at such rates over the medium-term. 3.7 On the other hand, domestically-oriented manufacturing has lagged the general recovery, growing at only 5.31 in 1987, and can be expected to pick up steam in 1988 and beyond only if consumer demand recovers. Big-ticket industries, including beverages, tobacco, cement, petroleum products, cars and trucks were sluggish in 1987, but show signs of improving strongly in 1988. A continuation of this trend is necessary for the overall health of the manufac- turing sector. 3.8 Other sectors in the economy tend to follow overall GDP, rather than lead the recovery. Transport, trade, utilities, finance and other services are closely linked with the main productive sectors. They have the potential to grow if material product grows. 3.9 The main drag on potential growth in the medium-term will continue to be construction and government services. Construction output has already fallen by a quarter from its 1984 peak, but is unlikely to recover strongly while government development expenditure and NFPE investment continue to fall. The brightest light for the short-term prospects of the sector is in housing, particularly at the low end of the market. The accelerating imple- mentation of the SLCHP should give the construction sector a boost in 1988 and reverse three years of declining output. Cross-country evidence indicates that housing investment typically rises as incomes grow for countries at Malaysia's development stage. Once the current overhang of more expensive - 52 - houses is worked off, therefore, growth in housing activity could again be a leading sector in the economy. Government services, however, are unlikely to rebound as fast. Pressure on public finances will be apparent over the medium-term and, with interest and maintenance expenditures rising faster than GNP, the public sector wage bill will be further squeezed. Efficiency and Productivity Growth 3.10 A sectoral analysis of potential growth runs the risk that the implicit assumptions about use of resources by each sector are inconsistent with the economy-wide supply of resources. An alternative methodology for assessing growth potential is based on the growth of inputs into production and the efficiency with which those inputs are used. Traditionally, econo- mists decompose value added growth into capital and labor growth, with the residual representing efficiency changes. Few countries can grow rapidly on the basis of new factor inputs alone. A simple numerical example best illus- trates this point: assume the ratio of the capital stock to GDP is four; assume further an economic depreciation rate of 2Z per year and a labor force growth of 3Z. Then a total investment/GDP ratio of 20X would be required just to maintain the averagg capital per worker constant and, hence, to generate growth of 3Z per year._/ If investment were to be raised by an additional 10 percentage points, to 301 of GDP, growth would still be only 5X (assuming a substantial return on capital of 201). 3.11 This arithmetic clearly highlights the importance of efficiency improvements in generating high levels of growth. In developed countries, the broadest measure of total factor productivity (TFP) growth has accounted for between 40-65Z of GDP growth. In industrializing developing countries, the range has been significantly smaller, at around 10-20Z for successful manufac- turers. These differences are partly a statistical phenomenon, reflecting the difficulties inherent in measuring labor and capital. An aging, more experienced and better educated work-force, characteristic of developed countries, actually produces more labor than the growth in the number of employed would suggest. Once adjustments of this nature are made, the resi- dual TFP growth in developed countries declines. 3.12 Cross-country comparison of TFP growth rates are fraught with problems caused by differences in methodology, data, sectoral coverage and time periods. For developing countries, estimates for manufacturing TFP growth are 2.31 in Hong Kong (1960-70), 3.81 in Singapore (1957-70), 3.51 in Korea (1960-70), 3.61 in Taiwan (1960-70), 0.6 to -1.2% in the Philippines (1956-70 and 1971-80), 1.7% i Indonesia (1975-82), 0.7% in Thailand (1963-76) and 2.11 in Turkey (1963-76)._ 2/ Given the assumptions above, gross investment would need to be 5Z of the total capital stock or 201 of GDP. 3/ See A. Krueger and B. Tuncer, "Estimating Total Factor Productivity Growth in Developing Country," World Bank Staff Working Paper No. 422; and "Philippines: Issues and Policies in the Industrial Sector," World Bank, 1987. - 53 - 3.13 Malaysia does not collect the data required to do a detailed calcu- lation of the sources of economy-wide growth. One particularly important missing variable is the capital stock in each sector or in the economy as a whole. There does exist information on employment in selected sectors, however, and this permits calculation of labor productivity. Although such estimates are useful, they fail to distinguish increases in output per worker that arise from greater inputs of other factors from increas.* in output stemming from efficiency improvements. They must therefore be interpreted with caution and considered in the light of other information. Estimates of labor productivity growth are shown in Table 3.1. There is a trend towards a gradual decline in labor productivity growth since 1975-80, accentuated during the recession, but continuing into 1987. Total labor productivity growth in the 1980s has been less than one-half of its average growth in the 1970s. Considering that investment rates have been higher in the 1980s, this suggests that total factor productivity may also have declined. Table 3.1: LABOR PRODUCTIVITY GROWTH (z) 1971-75 1975-80 1980-84 1984-86 1986-87 1971-80 1980-87 CDP 3.6 4.7 3.7 -1.5 -0.1 4.2 1.7 Agriculture 4.1 5.5 3.1 2.7 -1.8 4.8 2.2 Mining -1.3 9.9 13.1 16.1 -2.3 4.7 11.6 Manufacturing 2.8 0.4 4.5 3.8 6.6 1.5 4.6 Construction -6.3 2.0 1.1 -11.0 -3.1 -1.8 -3.1 Government services 5.0 1.7 6.5 1.1 -0.5 3.2 3.9 Source: Staff calculations based on data in Economic Report, 1987, Ministry of Finance. 3.14 A substantial part of the decline in labor productivity growth is due to agricultural performance. During the 1970s, the rapid shift from rubber to palm oil helped boost agricultural labor productivity, both by rais- ing output and by reducing labor input requirements. With this transition now largely completed, agricultural productivity growth has slowed considerably. 3.15 The other important sector to show a decline in productivity is construction. Particularly during the recession years of 1985-86, output in construction has fallen far faster than employment. One possible explanation may be that a significant adjustment has taken place in the number of hours worked per person. Some evidence suggests that there has been more work- sharing to avoid layoffs; in this case, productivity numbers based on hours of work rather than numbers of employees would show a marked improvement over the figures in Table 3.1. A second possible explanation is that the statistics on construction output are severely underestimated by their failure to capture small firm activities adequately. Inconsistencies between construction output - 54 - data and housing completion reports point to a potentiai)y severe bias in officially published data on the construction industry.- 3.16 Two important sectors show a considerable increase in labor produc- tivity in the 1980s--mining and manufacturing. Mining improvements reflect the coming on stream of petroleum and gas production. The manufacturing story is less transparent. What is clear is that the considerable growth in manu- facturing value added in the 1980s (6X annually) has been accompanied by a growth in employment of only 1.31 per year. By contrast, growth in employment in manufacturing in the 1970s was about 10% per year, roughly comparable to an 11X growth in output in this period. 3.17 The growth in labor productivity in manufacturing may be explained by two phenomena: a growth in capital assets per worker or an increase in efficiency. Separating between these two is critical in evaluating manufac- turing performance as output growth based on large investments may not turn out to be efficient. Indeed, a preliminary analysis based on survey data collected by the Department of Statistics suggests that the contribution of capital to output growLh in the early 1980s was heavy and that overall efficiency declined (see Appendix 3). Although these results should be treated with caution because of the residual-based methodology and inherent data problems, they do suggest that the strategy of promoting heavy industry and of channeling substantial investment resources into industrial chemicals, refineries, iron and steel and cement has contributed less to output and employment growth than if those same resources had been reallocated to more dynamic sectors. 3.18 Another indicator of the efficiency of manufacturing can be derived by comparing the capital and labor inputs used by different subsectors to produce one unit of value added. The results, averaged for the years 1982-84, are shown in Figure 3.1. The data is based on the Department of Statistics' Survey of Manufacturing Companies. In general, those subsectors that use less of both capital and labor to produce a unit of value added are c gsidered more efficient, in technological terms, than those using more inputs." The curved line in Figure 1 roughly separates efficient sectors (those below and to the left of the line) from inefficient subsectors (those above and to the right of the line). In economic terms, however, two additional criteria are required to determine efficiency. First, depending on the structure of wages and 4/ See the World Bank's "Malaysia: Housing Sector Study" (1988) which concludes that the flow data on new residential construction and the stock data on completed housing units cannot be reconciled. One or other of the :es is likely to be systematically biased. 5/ In order to be valid, productivity data should only be considered in years where there is not excessive underutilization of capacity. Thus, adding more recent years (1985-87) would not be appropriate. It must be borne in mind, however, that several of the capital-intensive subsectors had not reached full maturity by 1982-84. Measuring their efficiency at initial stages of production could lead to some bias. Figure 3.1 : USE Capital per UnRt IN MALAYSI STRIES of ValueAdded IE ev 9 1 2- i 3.2 - O IRON& STEEL 3- 2.8 - OGLASS 2.6 -INDUSTRIAL CHEMICALS 2.4- 2.2 - \ POTTERY 2- \ NON-FERROUS METALS PETROLEUM 1.8 - ° REFINERIES NON-METALIC o MINERALS 0 PAPER 1.6 - 0 FOOD 0 TEXTILES \ O~~~~~~~~~ PLASTIC PRODUCTS OF 1.4 - &PETROLEUM a COAL 1.4 0 WOOD PRODUCTS 0 TRANSPORTATiO 0 FABRICATED 1.2 0 RUBRMETALS BEVERAGE LEATHER I OTHER MACHINERY MANUFACTURING O FURNITURE OTHER CHEMIC FOOTWEAR 0.8 PRODUCTS 0 o o ELECTRICAL SCIENTIFI & TOBACCO T° PRINTING MACHINERY MEASURINGEQUIPMENT WEARINGAPPAREL 0.6 - I I I I I I I - I I 0 20 40 60 80 100 120 140 Labor per Unit of Value Added a.t441g74b - 56 - interest rates, some combinations of capital and labor inputs are preferable to others. In Malaysia, it seems plausible that efficient subsectors are those that are labor-intensive and those that only require small increases in capital as labor is reduced. Subsectors such as rubber, electronics, printing, machinery, scientific and measuring equipment, furniture, footwear and wearing apparel fall within this category. Subsectors such as industrial chemicals, minerals and metals, and refineries are likely to be less efficient economically, even though they economize on the use of labor, because of their heavy use of capital investments. 3.19 The second qualification to be made is that the market prices on the basis of which value added is calculated for the purposes of the figure may not correspond to appropriate social prices. Tobacco is one example of a subsector which appears to be very efficient in private terms, but this is because of extremely high levels of effective protection. If efficiency were to be evaluated at border prices, tobacco, automobiles and cement, amongst other sectors, would appear to be considerably less efficient. 3.20 One implication of the substantial apparent differentials in the efficiency of resource use in Malaysia is that considerable gains may be expected from reallocating resources away from inefficient sectors such as pottery, glass and iron and steel, towards more efficient sectors. In order to accomplish this, it must be recognized that all subsectors should not be protected at the same time. Rather, the objective should be to engineer a set of incentives that induce resources to flow into the most efficient sectors. 3.21 The analysis of productivity reveals some disturbing features of Malaysia's manufacturing system. First, accelerated growth in fixed assets in the sector has been associated with a decline in the growth of gross output. This is related to the development of highly capital-intensive industries in the 1980s. Second, these capital-intensive industries have performed poorly and are characterized by excess capacity, declining productivity, and low rates of labor absorption (although to some extent the data may also reflect the long gestation periods inierent in capital-intensive processes). Third, while capital is flowing freely to some subsectors, others, including tex- tiles, footwear and scientific instruments, have failed to even replace depreciated eqgi,pment. This bodes ill for future productivity gains in manufacturing.- Fourth, productivity problems are widespread throughout manufacturing. They are worse in the capital-intensive sectors, which also typically have a high share of publicly controlled firms, but extend to other sectors including the fast-growing exporters of electronics and garments. 3.22 Small-Scale Firms. One consequence of the capital intensive manu- facturing growth strategy adopted has been a decline in the size of the small- scale sector in Malaysia and a rise amongst medium-scale firms. The latter 6/ Note, however, that this is not always an indicator of inefficiency. In some textile plants, for example, old equipment has been well-maintained and is the basis of cost competitiveness. Nevertheless, the fact remains that productivity is not growing in these subsectors. - 57 - are more capital intensive than either small or large firms, and it is felt, are generally less efficient, capturing neither the flexibility of smallness nor economies of scale from large operations. The decline amongst firms with less than 100 workers is also thought to be both cause and symptom of an industrial structure that features limited backward integration from the large exporting and final good producer to smaller producers of intermediates. 3.23 Table 3.2 shows the size structure of manufacturing firms for 1978 and 1985. Firms with less than 100 employees represent about 80% of estab- lishments, but only 30% of employment, 20% of value added and 19X of fixed assets. In all categories, these firms' importance in manufacturing had declined since 1978. Growth has instead been more rapid in medium-sized firms, employing 100-200 workers, although large firms have also increased their share in value added and fixed assets. Table 3.2: SIZE STRUCTURE OF MANUFACTURING FIRMS (%) Number Total of estab- Value Value Fixed Total employ- employment lishments of output added assets ment size group 1978 1984 1978 1985 1978 1985 1978 1985 1978 1985 Below 50 - 16.14 66.3 58.4 12.6 8.8 11.7 9.3 10.3 8.5 50 - 99 - 14.56 16.3 19.3 13.6 12.4 14.2 11.1 12.5 10.6 Below 100 - 30.70 82.6 77.7 26.2 21.2 25.9 20.4 22.8 19.1 100 - 199 - 17.16 8.7 12.6 18.5 25.9 17.8 18.6 16.3 18.4 200 and above - 52.14 8.7 9.7 55.3 52.9 56.2 61.0 60.9 62.5 Source: Survey of Manufacturing Industries, DOS, various issues. 3.24 The decline in small firms has brought Malaysia closer to the industrial structures of Korea and the Philippines, two countries with acknow- ledged high business concentration, than to that of Japan (Table 3.3). Although the figures in the table are not directly comparable because different countries compile statistics on size distribution of firms according to different criteria, the table does indicate that Japan has been able to generate much higher value added and employment from its small industries than Malaysia. Development of small-scale enterprises therefore may offer one important avenue for improving industrial efficiency. 3.25 In every country, financing, marketing and technology upgrading are amongst the major problems faced by small-scale firms. In Malaysia, financing Table 3.3: INTERCOUNTRY COMPARISON ON THE ROLE OF SMEs Number of establishments Employment Value added Total employ- Malay- Phil- Malay- Phil- Malay- Phil- ment group size sia /a Korea Japan ippines sia Korea Japan ippines sia Korea Japan ippines 4 - 99 81.7/b 90.3 92.7 97.8 30.7 32.6 55.5 35.3 20.1 19.6 38.6 12.8 100 - 199 10.4 (5.0) - 0.9 17.2 (12.2) - 8.0 16.3 (9.9) - 8.3 100 - 299 /c - 7.1 2.5 - - 21.1 16.4 - - 18.1 17.3 - 200 + 7.9 (4.7) - 1.3 52.1 (55.1) - 56.7 63.6 (72.1) - 78.9 300 + /c - 2.6 4.8 - - 44.0 28.1 - - 62.3 44.1 - /a Figures for Malaysia are for 1984. Figures for other countries are for 1982. /b Includes less than four employees. /c Malaysia and the Philippines do not collect data in the 100-299 range. - 59 - seems to be the most acute.71 Furthermore, in contrast to Korea, Japan and Taiwan where large downstream firms provide financing, quality control and technical assistance to upstream small and medium firms, these links are largely absent in Malaysia. In fact, the industrial structure in Malaysia is not conducive to backward integration for two reasons. First, the preponder- ance of multinationals and foreign companies in Free Trade Zones implies that a major segment of manufacturing probably already has a well-developed upstream linkage with subsidiaries or affiliates abrcad, and there is little incentive to develop domestically-located upstream firms. Second, the compa- rative advantage of Malaysia still lies in assembly activities, especially in the electronics and apparel subsectors, where the scope for upstream linkage is small because of limited comparative advantage in intermediate produc- tion. As the manufacturing structure evolves towards other final goods, however, including wood, rubber and other resource based products, and auto- mobiles for example, the scope for integration should increase. 3.26 In parallel to promoting backward linkages, small-scale firm deve- lopment could be encouraged by an orientation towards exports. Although the proportion of firms surveyed by Fong, who exported any portion of their output, increased from 16X in 1980 to 29% in 1985, this is still a small proportion relative to Taiwan, China or the Philippines. Lack of: information on foreign markets is a distinct handicap to small-scale firms and the lack of a marketing infrastructure is keenly felt. Japan and Korea both have public bodies as well as large private marketing companies that provide necessary information on foreign markets snd technologies. Efforts to develop such bodies in Malaysia have been tried in the past without great success, but in part this is seen as due to institutional reasons concerning the composition of the companies rather than as symptomatic of an inappropriate approach. 3.27 Currently, Government offers a variety of incentives to encourage the development of small-scale enterprises. These include a sales tax exemp- tion for sales of M$100,000 or less per year; stamp duty exemption for lease agreements with rents below M$2,400 per month; a 5% abatement of adjusted income for corporate income tax purposes; credit guarantees and a Special Loan Scheme for small business; and direct access to credit through the Bank Pembangunan Malaysia Bhd. and the Malaysian Industrial Development Finance Corporation at an interest rate of 7.75%. 3.28 One problem with these schemes, however, is that they are geared towards very small enterprises and serve different purposes. For example, tax exemptions also act as a means of reducing administrative costs of collection from small businesses. There is, therefore, a diffused focus to Government efforts. There are currently ten ministries and nineteen government agencies 7/ According to a survey of 167 firms by Fong Chan Onn (1987), 50% of firms reported financial shortages as their most serious difficulty and 27.5% as the second most serious (eF4-er marketing). See "Changes in the Industrial Structure and the koLe of Small and 1' -m Industries in Asian Economies: the Case of Malaysia," Institute of eloping Economies, Tokyo, February 1987. - 60 - involved in the promotion of small-scale firms. Each uses differeng,defini- tions of smallness. For example, the Credit Guarantee Corporation -_ regards firms with paid-up capital and reserves of less than M$200,000 as small (M$100,000 for non-Bumiputera firms); the corporate income tax abatement is for companies with shareholders' funds not more than M$200,000 and fixed assets of not more than M$1 million. Exemptions from the licensing require- ment of the Industrial Coordination Act are granted to firms with share- holders' funds equal to or less than M$2.5 million and 75 employees. To streamline activities and incentives, a new Coordinating Committee for the development of small-scale firms has been established within the Ministry of Trade and Industry. Their definition of a small-scale industry is one with fixed assets below M$250,000. This may be perhaps overly restrictive, as it would tend to cover only firms employing less than about 10 workers, and leaves an important segment of manufacturing, firms employing 10-200 workers, without any government support. Policies to Improve Productivity 3.29 Studies of total factor productivity growth in other countries indicate that the most rapid efficiency gains occur when barriers to competi- tion, such as tariffs, quotas and other protective devices, are absent; when there is flexibility in the allocation of factors and neutrality in the wage/ interest rate ratio; and when readily available funding and an appropriate risk-reward structure induces investments in new machinery and equipment. In Malaysia, these criteria are affected by: (i) the role played by NFPEs; (ii) the tariff regime; (iii) the systems of industrial licensing and restruc- turing assistance; and (iv) the corporate tax structure. Policy reform in these areas can help generate productivity improvements. (i) Public Enterprises 3.30 The public enterprise sector has grown to perhaps one-quarter of the Malaysian economy, one of the highest ratios in the nonsocialist world. Compared to the private sector, it has lower profitability (Table 3.4), but these measures, being financial, do not necessarily indicate comparative effi- ciency. Differences between public and private firms in debt/equity ratios, capital-intensity and the pursuit of noncommercial objectives will account for some of the observed profitability. Nevertheless, it is likely that low efficiency is one explanatory variable of the public sector's disappointing financial performance. 8/ The CGC provides guarantee cover for up to 60% of commercial bank loans to small-scale firms. Loans up to M$50,000 may be made without collateral, and up to M$200,000 can be provided to a Bumiputera company (M$100,000 for a non-Bumiputera Company). - 61 - Table 3.4: COMPARATIVE PROFITABILITY RATIOS IN PUBLIC AND PRIVATE FIRMS /a (unweighted averages) 1981 1983 1985 Manufacturing Private 0.086 0.065 0.049 Public 0.011 0.022 0.024 Construction Private 0.016 0.057 0.008 Public 0.047 0.039 0.022 Mining /b Private 0.29 0.28 0.23 Public 0.19 0.17 0.17 /a Profitability is measured as net after-tax income as a ratio of total assets. "Private" firms are based on DOS surveys of all companies. They therefore are an average of public and private firms. /b Private mining profitability is taken as a weighted average of tin mining (0.25) and other mining (0.75) to be comparable to the public sector estimates. Source: DOS Financial Survey of Companies; CICU data on NFPEs. 3.31 Efficiency differentials between public and private firms are present in at least three broadly-defined activities: project evaluation and choice; implementation and operation; and the decision to close down ineffi- cient firms. There are several instances where, ex post, the public sector performed poorly in project evaluation and choice. The selection of large- scale, loss-making enterprises such as Sabah Gas, Sabah Forests, Perwaja steel, Proton Saga and Kedah cement is evidence of optimistic forecasts of prices and demand and of a failure of the apparatus for investment decision- making. However, with the decision to prohibit large incremental NFPE projects, additional policy reform of evaluation and choice criteria has become less urgent. 3.32 Data on operating efficiency for public enterprises are not immediately available. One issue is to differentiate between efficiency and profitability; the data on net profitability indicate that only about 55% of NFPE's have had positive profits in each of the years 1982-86. This aggregate, however, disguises important sectoral variations. In general, public enterprises in the transport and extractive sectors performed better - 62 - than those in agriculture, manufacturing or services. Public construction firms were mostly profitable in the early 1980s but, with the slump in construction activity, are now amongst the worst performers. These figures are unweighted--that is, they are based purely on the number of firms reporting profits or losses, without regard to whether they are large or small. There is, however, a distinct difference in performance according to size. Small NFPEs--those with less than M$5 million in paid-up capital-- account for about two-thirds the total number of NFPEs, and seem to perform considerably less well than larger NFPEs. 3.33 Negative profits may arise from operational inefficiency or from large payments of taxes or interest which reflect firms' capital structure, depreciation and other accounting factors. For the NFPEs, however, 87% of firms with negative net profits also had negative operating profits. In aggregate, the debt/assets ratio is low--one-third--and interest payments are not an undue burden. In selected instances, however, particularly for manu- facturing companies established in the early 1980s, the financial burden is heavy. These companies were set up with debt/equity in the proportion 70:30, with much of the debt obtained from abroad. The appreciation of non-dollar foreign currencies and the dilution of equity as a result of negative operat- ing profits have combined to generate increasingly leveraged financial capital structures and heavy debt service burdens on these companies. 3.34 Several factors appear to have contributed to the poor performance of the NFPEs. First, there has been inadequate evaluation of management stemming from a diffuse control structure, particularly for smaller NFPEs. This has occurred because of a lack of clarity and priority in objectives, which have ranged beyond profitability to income distribution and restructur- ing of society, pursuit of socioeconomic and employment objectives, develop- ment of backward regions and industrial diversification. While each objective may be appropriate for a public enterprise, the inability to quantify and prioritize objectives has made the task of evaluation more complex. In many instances, such as in the Ministry of Public Enterprises, the evaluation function has been dissolved and left entirely in the hands of the Board of Directors. 3.35 Second, poor performance may be ascribed to the incentives faced by NFPE managers. Not only are there few rewards for good performance, but there is also an incentive to utilize all available funds within the company rather than remitting the resources back to the Treasury in the form of taxes or dividends. In fact, besides PETRONAS, very few NFPEs or statutory authorities have ever paid a dividend on government equity. An additional disincentive to profit-making and retention in NFPEs is the threat of takeovers. As part of the program to redistribute assets to Bumiputeras, the PNB has acquired several companies at prices that typically reflect the cost of the assets rather than the market value of the company. Quite appropriately, the PNB searches out the most profitable companies for takeover, which are precisely those most coveted by the parent organization. In such an environment, some cross-subsidization, or provision of uneconomic services, then serves to minimize the whreat of transfers. - 63 - 3.36 The third source of inefficiency in the NFPE sector stems from their ability to return to Government for additional loans, equity infusions or protection against closure, mostly to ensure that the social and employment- related functions continue to be undertaken. As long as the economy was performing well, resources were provided ungrudgingly. In the present resource-constrained environment, however, there is more pressure to quantify and evaluate the cost effectiveness of provision of these noncommercial services. 3.37 NFPE reforms. This analysis of performance indicates that reform in the incentive system--broadly defined to include the setting of objectives, financial accountability and the institutional structure of control--and in the financial structure of corporation balance sheets should be considered. Two mechanisms to achieve such reform can operate in tandem: a performance evaluation system for NFPEs, and a strengthening of the program of privatiza- tion. 3.38 Privatization is currently conceived of as a broad umbrella covering many forms of institutional change in the organization of NFPEs. In general, however, the overriding principles are an orientation towards a corporate structure and culture and an emphasis on profitability. While this is suit- able for many of the public enterprises, particularly those under Federal ownership, it is less amenable to handling the requirements of social objec- tives that lie behind the creation of many NFPEs, notably those under State Economic Development Corporations and State Agricultural Development Corporations. Because of the difficulties attached to privatization of these types of enterprises, high priority should be accorded the development of a performance evaluation system to provide correct signals to NFPE managers, set objectives, determine targets and review performance. The current data collection system introduced by CICU, however, is still young and evolving and needs additional refinement to become an adequate base on which to develop an evaluation system. One prime requirement is the ability to differentiate between financial profitability, which may be determined by many factors outside the control of an enterprise manager, and efficiency for which he should be made responsible. As a first step to evaluate efficiency, a data breakdown between prices and quantities is required. This would include variables additional to those readily accessible by CICU. For example, volume of sales, volume and value of material inputs, employment and total wage bill figures should be collected and analyzed, at a minimum. Details on transfer payments such as sales taxes paid on inputs or outputs as well as other taxes and exemptions would also be desirable. Full design of the data requirements, however, should be closely coordinated with and tailored to the development of the performance evaliation system. 3.39 The success of a performance evaluation system depends on an ability to difterentiate clearly between commercial and noncommercial objectives. The tendency to burden NFPE with noncommercial objectives must be resisted or, at least, the costs involved in their pursuit must be calculated and due compen- sation made to the company. This principle has already been attempted in the case of the railways, for uneconomic passenger service, and for some services performed by NFPEs under the Ministry of Public Enterprises (MPE). Such a system should be administered by newly created evaluation units within ICU and - 64 - the MPE. Significant staff reallocation and training would be required, including perhaps the transfer of some managers of NFPEs to the evaluation units, and the functional responsibilities of these agencies should shift towards evaluation and away from information collection. This latter largely duplicates work that is done by CICU in an efficient and centralized fashion. 3.40 To be effective, the performance evaluation system should be combined with an appropriate incentive system, with cash and prestige bonuses to reward good performance and the threat of dismissal to penalize poor per- formance. Other countries, notably Pakistan and Korea, have implemented such systems with success. Prestige awards, such as Manager of the Year, are use- ful instruments to promote the philosophy of efficiency at relatively lc*,# cost, if they are implemented at the highest level of government and accom- panied by generous publicity. The other side of the coin is the penalty that should be imposed for poor performance; the Board of Directors should be en- couraged to use the threat of replacing management more aggressively in the face of clear underperformance of agreed-upon targets, corrected for external factors by the evaluation units. 3.41 The remaining priority area for government reform is to act more quickly to close inefficient enterprises. As a result of access to Treasury funds in emergencies, justified in part on the basis of socio-economic objec- tives, there are few mechanisms to enforce the closure of inefficient public enterprises. During the period 1981-86 only 32 out of a total 867 NFPEs were liquidated; others were in various stages of closure, suspension, dormancy, and receivership. By contrast, the private sector is probably much more effi- cient and speedy about liquidation, with creditors unwilling to inject new funds into operations of doubtful profitability. While some good companies are unquestionably forced into bankruptcy, overall the private sector system avoids the error of compounding initial resource allocation mistakes much better than the public sector. 3.42 The performance evaluation system should be viewed as the logical extension of current government efforts to restructure NFPEs. Government has set up a special unit in the Treasury which has, as a first step, investigated the options for a few large inefficient firms. As a result, restructuring packages for Perwaja steel and Pernas, for example, have already been imple- mented. The second phase is to expand the unit's functions to cover all other Federal NFPEs. As these have almost no noncommercial objectives, they are most easily privatized. The third phase of the policy is to gradually reduce reliance on Federal budgetary support for State controlled NFPEs. 3.43 In the medium-term, this strategy will have to be complemented by an institutional mechanism for monitoring and evaluating performance. For example, the restructuring of the large NFPEs has been carried out as a result of perceived losses and mismanagement, but monitoring of the new structures has not yet been institutionally introduced. Furthermore, while it is rela- tively straightforward to identify and focus on the biggest NFPEs, the larger problem is how to tackle the smaller, inefficient NFPEs. Reliance on privati- zation will impose a considerable burden on the government to evaluate proposals from private individuals. In particular, as divestiture will often involve government protection, guarantees and debt write-downs, it may not - 65 - always be the best solution. Liquidation may be preferred from an economic point of view in cases where the costs of continued operation outweigh the benefits. An evaluation system would be a useful mechanism for deciding on the merits of divestiture versus liquidation and would further introduce transparency and objectiveness into the size of budgetary support offered to non-privatized Federal and State NFPEs. 3.44 A cornerstone of the Government program to improve efficiency of NFPEs is through privatization. The Privatization Master Plan will address the precise modalities of this operation. To date, however, despite the priority given to privatization, only modest progress has been made; a total of 71 companies have been privatized, and 45 more are under consideration, but only 5 have been of significant size, accounting for over 90X of the priva- tized assets. Moreover, none of these have been truly exposed to private market forces. Even where private sector appointed management is in place, there are monopoly privileges, government contracts and guarantees in place that reduce financial pressures to increase efficiency. In fact, the most successful privatizations--those of HAS and MISC--have served primarily to recapitalize company balance sheets. The improvements in profitability that have occurred in these companies after flotation can probably not be attri- buted to the privatization but to other market forces. Management personnel have remained the same. Over the medium-term, however, the clear commercial objective for MAS and MISC should help retain high efficiency in operation. 3.45 There are several hurdles to overcome in order to accelerate the pace of privatization and to ercourage complementary new private investment. First, many of the companies have large losses and are therefore unattractive to potential buyers. Second, in many cases, the agency involved is unwilling to provide to potential investors the information necessary to formulate a bid. Third, government has legitimate fears of retrenchment and of a cutback in politically or socially desirable services (e.g. railways) and hence is unwilling to relinquish control. Fourth, there is as yet no adequate regula- tory framework to govern potential monopoly situations. Fifth, there may be different objectives amongst Federal, State and Local governments all of whom may be involved in particuiar projects. Sixth, because of the thinness of capital markets and the guidelines imposed on new equity issues through the Capital Issues Committee, the scope for rapid privatization at fair market price may be limited. One example of the limitations on institutional inves- tors (pension funds, insurance companies, etc.), who are perhaps the largest source of domestic private capital, is the requirement that they only hold Trustee stocks, i.e. those which have paid dividends f r at least five years. Few NFPEs have ever paid Treasury a dividend;9 hence, none of the other NFPEs would be eligible as a Trustee stock. These obstacles must be addressed before privatization can be expected to contribute to improved effi- ciency on a major scale. 9/ Dividends were paid to the Federal Government by the following: PETERONAS, MAS, Tabung Haji, LLN, Pernas International Hotels, FELDA, FELDA Oil Products and MISC. MAS and MISC have already acquired the status of blue-chip or tgrestee stocks on the KLSE. - 66 - (ii) Tariff Reform 3.46 The average level of tariff protection in Malaysia is relatively low compared to other developing countries. Average nominal tariff rates in 1985 were about 13.6Z while the same figure was 23.0Z for Indonesia, 21.9Z for Korea, 28.5Z for Philippines and 34.02 for Thailand (Table 3.5). The weighted average level of effective protection in manufacturing was calculated to be about 23Z in 1982 by the Malaysian Industrial Policy Study of 1985 (MIPS). Relative to other developing countries, this is still not high, although there is a wide dispersion in effective protection, between as well as within indus- tries, that is a source of concern. There are also signs that the dispersion may be growing; the reason is that the involvement of public firms in sectors such as steel, cement and automobiles has led to an increase in nominal and effective protection rates in these sectors in the early 1980s. In some cases, such as cement, tariffs and quotas may be viewed as an anti-dumping device, but anti-dumping should be considered a short-term measure and should not be invoked over the medium-term as it has been. Table 3.5: AVERAGE TARIFF LEVELS (unweighted averages; percent) Proportion of import items Year Average nominal tariffs subject to import restrictions Indonesia 1980 28.08 1985 23.0 25.1 1987 23.0 15.6 Korea 1980 25.0 31.4 1985 21.9 12.3 1988 18.0 4.6 Malaysia 1980 11.6 1 less than 5Z 1985 13.6 Philippines 1980 43.1 37.0 1984 28.1 36.1 1986 28.5 16.7 1988 -- 10.0 /a Thailand 1981 31.0 } less than 52 1985 34.01 /a Of which 4.7% are due to health, safety and security reasons. 3.47 The precise impact of effective protection on efficiency is hard to measure quantitatively. Because EPR calculations use statutory tariff rates and aggregate input-output data, there is likely to be a substantial variation - 67 - in the level of protection across firms within industries and over time. The existence of other significant elements of protection, through government procurement practices, licensing, exemptions from indirect taxes and other tax advantages, further clouds any attempt to measure the true level of protection in an industry and to relate it to performance. Nevertheless, a comparison of EPRs at the two-digit level is revealing. When firms are ranked according to the ratio of gross operating margins to fixed assets, the mean effective protection rates of the m8y profitable firms is 20Z lower than the EPRs of the less profitable ones., Profitable firms, in addition, employed half as much capital per worker as less profitable firms. Thus, rather than generating surplus profits to induce investors into new sectors, EPRs appear to be compensating for unprofitable firms' inability to compete, and to be subsidizing capital. It is also noteworthy that EPRs are not selectively b;psed toward capital-intensive sectors or labor-intensive sectors, which s I be expected if their purpose was to correct for some factor market C. rtion. In fact, they appear to be haphazard and arbitrary. .48 Reform of the tariff structure is under consideration by the special Action Comittee on Incentives. In undertaking reforms, the Committee should be guided by the principle that tariffs should be viewed primarily as a device for protection. Other objectives, including revenue collection, should be accorded second priority as they can be pursued more effectively with other fiscal policies. Specifically, the burden of revenue-raising should be shifted to the overall indirect tax system which would levy taxes equally on imported and domestically manufactured goods. Once the indirect tax system is implemented at the retail level, rather than at the manufacturing level, administration of indirect taxes on imports could be further separated from tariff collections. 3.49 A focus on the protective nature of tariffs necessarily implies that reforms should be directed at rationalizing the EPR structure rather than the nominal tariff structure. There may not, however, be major differences in practice, if the objective is limited to reduction in the highest EPRs, con- centrated in a small number of industries. These are almost always associated with a combination of high nominal tariffs on output and low or zero tariffs on inputs. A reduction in high nominal tariffs is the best way to reduce high EPRs, as tariffs on inputs and capital equipment create a cascading sequence of economic costs borne both by downstream producers and by consumers of the final product. Attention should therefore be focused on the 210 BTN codes (out of a total of 6,457) governing i-ports with statutory nominal tariff rates greater than 30X. (iii) Industrial Licensing and Restructuring 3.50 Reform of the tariff structure would only lead to greater efficiency if complemented by greater competition within industries. Such competition is 10/ On notable exception is tobacco, where both EPRs and profits are high. Excluding the tobacco sector, the "profitable" sectors' EPR's would be 36X lower than those in other sectors. - 68 - affected by the system of industrial licensing and other elements of indus- trial policy, including restructuring efforts associated with implementation of the Industrial Master Plan (IMP). Licensing requirements have recently been substantially liberalized, and bureaucratic delays reduced. Neverthe- less, further improvements could still be made. 3.51 An industrial license is required for the addition of most new prod- ucts or processes, and for the import of machinery which adds to capacity or reduces manpower. Licenses only apply to those companies with M$2.5 million or above in equity or more than 75 full-time employees. They are typically approved as long as the equity and employment criteria of the NEP are met. There are also a number of automatic exemptions that have recently been introduced. For example, an exemption is given to foreign companies exporting more than 80% of their output. An exemption from the equity guidelines is also given to foreigners exporting at least 50% of output, or thcse with 350 full-time Malaysian employees, provided their domestic sales do not compete with existing products. In addition, no licences are required for expansion or diversification activities for export-oriented companies or for those fulfilling the 30X Bumiputera equity requirements. In 1987, about 30% of all domestic expansion and diversification programs were automatically approved. 3.52 In theory, the issue of licenses may be based on a variety of factors, including supply conditions in an industry. If implemented strictly, these could have serious consequences in limiting competition. In practice, however, licenses are granted liberally as long as NEP criteria are met. The intent of the licensing system has become, effectively, to monitor industrial development rather than regulate it. Nevertheless, in practice, the licensing system may have had side-effects which could have reduced industrial efficiency in at least three ways. 3.53 First, the need to apply for licenses for new machinery can slow the rate of technological upgrading. This partly reflects the additional costs and uncertainties in terms of time of processing applications through cumber- some bureaucracies. In many instances, minor improvements and upgradings may not be deemed worthwhile. In addition, because licensing involves different agencies with different approval schedules, it can unwittingly create protec- tion for certain activities at the expense of others. For example, an imported consumer electronics product may only require a single license, whereas a domestic manafacturer who needs to import several components may require several licenses. Thus, even if import-substitution production were efficient, it may be unprofitable relative to imports of the final good because of the additional licensing requirements. 3.54 Second, because the licensing and incentive system can be used to favor regional and parochial interests, it has served to prevent economies of scale in certain industries. The palm oil processing, cement and metal fabrication industries stand out as examples where lack of economies of scale have led to inefficiencies. Biases against large, efficient projects arise because: (a) locational preferences favor several small, dispersed projects, rather than a single large one; (b) expansion investments are less favorably received rhan new projects in other areas; and (c) it is difficult to find Bumiputera investors necessary to meet the NEP criteria for large projects. - 69 - 3.55 A third source of inefficiency stemming from the licensing system is the constraints imposed on more efficient firms' attempts to force out less efficient firms. Particularly in industries with sizable excess capacity, new entrants may be discriminated against and old, inefficient plants continue to operate. In some cases, licenses can act as significant barriers to entry and allow firms a quasi-monopoly position. This also depresses incentives to innovate and upgrade machinery. 3.56 Recent efforts at liberalization of industrial licensing and at reduction of administrative delays and bureaucratic duplication are welcome measures. To foster greater efficiency gains, the priority areas for reform should be: (a) eliminating the protective elements of licensing by streamlin- ing procedures and focusing on safety, health and environmental concerns rather than economic concerns; (b) abandoning de jure as well as de facto concerns over industrial capacity and location; c movement over the medium- term towards price-based incentives to comply with NEP objectives rather than licensing approvals. A system of offering exemptions against the development tax for companies that comply with NEP regulations would be the preferred solution. In this case, other tax credits and deductions should only be allowed against income taxes. 3.57 Government actions to liberalize licensing requirements for small- scale and foreign export-oriented companies are important steps in the right direction. Government has raised the threshold equity level below which licensing exemptions are granted to M$2.5 million. The full benefits of this, however, will not be felt because of the associated criterion of less than 75 full-time employees which must also be met for firms to be exempted from licenses. The relative magnitudes of these ceilings do not correspond with average conditions in Malaysian firms. On average, a firm with eqyjyy of M$2.5 million would be likely to employ 150-200 full-time workers. The employment ceiling should, therefore, be raised to this range to permit a larger number of small-scale firms to be exempt from licenses. This would also bring the Malaysian categorization of small and medium-scale firms closer to other East Asian countries' definitions, such as Japan, Korea and Taiwan, China. 3.58 Licensing is an integral part of overall industrial policy which is increasingly focused on industrial restructuring. Action programs for indus- trial restructuring on a subsector basis are being prepared under the auspices of the Industrial Master Plan. Assistance for restructuring is also to be made available from a new M$500 million Industrial Adjustment Fund. In both the provision of finance and of fiscal incentives, care should be taken not to 11/ MIDA data indicate that a firm with equity of M$2.5 million would have, on average, fixed assets of M$5 million. Based on the 1985 manufacturing survey of the DOS, firms with fixed assets between 1-5 million ringgit had an average asset per worker ratio of 20,918. Those in the 5-10 mil- lion ringgit range had an asset per worker ratio of 33,547. The range 150-200 workers would be appropriate for a firm with 5 million ringgit assets and an asset per worker ratio between 25,000-33,000. - 70 - jeopardize the goals of long-run efficiency growth. The sectoral task force reports indicate that protection is being sought for the textile sector, foun- dries, shipyards, upstream automobile parts, upstream electronics products, ceramics, roofing tiles, clay pipes, furniture and fish canning. Meanwhile, export-oriented sectors such as palm oil and rubber processing are pushing for reduced tariffs on intermediate inputs and export duties on raw-material exports. 3.59 Assistance to ailing companies is often desirable because the social costs of bankruptcy can be high. Bankruptcy, however, is also the major threat that induces companies to strive for maximum efficiency improvements. In undertaking industrial restructuring, it is important to bear in mind that all industries may not necessarily be competitive. The objective should be to assist those industries that can become competitive at world prices in a reasonable span of time while allowing others to fail. Mechanisms, such as tariffs and other protective devices, which ensure the financial security of companies but which do not differentiate between efficient and inefficient firms, should be avoided in favor of more automatic systems which only provide transitory assistance. 3.60 There are several alternative models in the world by which efficient companies are steered through difficult times. In the United States, for example, manufacturing companies have relatively high equity participation. In essence, temporary reductions in company profitability are borne by share- holders; the company does not feel a squeeze in liquidity because the fixed costs of servicing debt obligations are reduced. On occasion, extra liquidity is available by cutting dividend payouts. This system is diametrically opposed to that in Korea. There, banks, under guidance from government, have long-term credit relationships with companies and maintain credit lines open to cover transitory losses. In other countries, risky ventures are undertaken directly by government which has the ability to provide additional finance in bad times, or in joint ventures with multinational foreign companies. This was the model followed until recently by Malaysia. Dissatisfaction with the progress towards an efficient and dynamic industrial structure, however, has led to a reconsideration of government's role and a search for an alternative mechanism for sharing risk with private entrepreneurs. 3.61 The current program of industrial restructuring should be oriented towards development of a sustainable long-term institutional structure which would allow companies leeway to steer through one or two difficult years, like 1985 and 1986. At present, the industrial restructuring initiatives contain many useful sector-specific interventions to enhance efficiency. They also include, however, generalized measures to protect industries, through tariff protection, low-cost loans and other fiscal incentives. These can offer short-term survival for industries, but are not supportive of longer-term efficiency objectives. Given Malaysia's open capital markets and liberalized financial system, reliance on commercial banks for sharing industrial risk, as in Korea, seems unrealistic. That system is closely tied to tight government controls on domestic financial markets. Instead, it seems preferable to uti- lize Malaysia's developed financial system and to underpin industrial restruc- turing with a stronger equity base. This would involve removal of impediments in the stock market (see paras. 3.107-3.109), and greater initiatives in - 71 - encouraging long-term credit and corporate bond markets, venture capital funds, joint ventures, leasing arrangements and other corporate financial structures that are more flexible than existing arrangements. (iv) Corporate Income Tax Reform 3.62 The statutory corporate income tax rate in Malaysia, at 40Z, is amongst the highest in East Asia. There is, in addition a 5% development tax. These, however, give a misleading picture of the true effective tax rate be- cause of the significant deductions allowed for depreciation, investment allowances, and income abatement. A five-year tax holiday, applicable to com- panies with pioneer status, further reduces the tax rate. In addition, favor- able treatment of capital gains and an offset against personal income taxes of corporate taxes paid on dividend receipts act to lower the tax burden. 3.63 The complexity of evaluating the overall corporate income tax struc- ture stems from the diversity of incentives and the differences amongst firms in their availment of incentives. Nor can the cost of the various allowances be quantified because only sketchy data exist. Survey data do indicate, how- ever, that pioneer status is the dominant element, accounting for three-quart- ers of the tax breaks issued between 1958-82. There is, furthermore, a gene- ral impression that the corporate income tax is no longer a buoyant revenue raising instrument and indeed significantly distorts the incentives to invest. 3.64 One useful way of assessing the impact of corporate taxes is through the use of a simulation model. It should be emphasized, however, that this involves use of data which are purely hypothetical, albeit based on the actual parameters of the tax system and on the average capital /omposit ion of invest- ment projects. One model developed in the World Bank 2 computes a cash-flow stream before taxes and after taxes, making allowances for depreciation, inte- rest deductions, loss offsets, capital gains, personal income tax credits against dividend payouts and other features. The percentage difference in the rate of return generated by the before-tax and after-tax streams is designated the marginal effective tax rate (METR). 3.65 Simulations show that Malaysia's actual METR is significantly lower than its statutory rate, and about the same as other Asian countries (Table 3.6). The actual rate depends on the percent of investment financed through debt. These calculations, however, only take into account the minimum allow- ances accorded all companies. Once additional possibilities for deductions and exemptions are included, the rate can drop dramatically. For example, a pioneer company with a 5-year exemption, would face a much lower METR, of only 15Z, calculated for a typical manufacturing project, financed 50% by debt. If losses from a project can be used to offset income earned elsewhere by a corporation, then the METR falls to just 10%. Although Malaysia limits the possibilities for offsetting losses in one company against profits in a subsidiary or parent by prohibiting group relief taxation, offsets are 12/ See A. Pellechio, G. Sicat and D. Dunn, "Taxation of Investment in East Asian Countries," DRD Discussion Paper No. 261, World Bank, March 1987. - 72 - feasible within a company, for an expansion or diversification project, for example. If, in addition, pioneer status is extended from 5 to 10 years as often occurs, the METS falls to a negligible level of only 4%. These simula- tions indicate that for significant numbers of new companies, especially in broadly defined sectors such as manufacturing for which pioneer status is readily obtainable, the effective tax rate on domestic companies is suffi- cientfi lov7 that it most likely does not serve as a deterrent to new invest- ment., {A fact, given the reduced tax burden after the major incentives are taken, the value of minor credits and deductions, such as the old labor utilization relief credit and the double deduction for training, is so small for many companies that the impact of such measures is likely to be small. That is, the tax system has become very unresponsive to fine-tuning. Table 3.6: MARGINAL EFFECTIVE TAX RATES IN SELECTED EAST ASIAN COUNTRIES (Z) Marginal effective tax rate Country Statutory rate All equity 50X debt financing Malaysia 40.0 32.0 20.5 Singapore 40.0 28.4 15.2 Philippines 35.0 40.4 31.9 Indonesia 35.0 41.6 34.1 Thailand 35.0 24.9 18.6 Japan 33.3 39.2 29.4 Korea 30.0 33.1 24.6 Taiwan 25.0 31.9 28.2 Hong Kong 18.5 17.3 9.6 Source: Pellechio, Sicat and Dunn, op. cit. 3.66 Pioneer status is only one of the mechanisms by which companies can lower their effective taxes, although it appears to be the most popular. Under the Promotion of Investment Act, 1986, companies could opt for an Investment Tax Allowance, which permits deduction from the adjusted corporate income tax base of a certain percent of investment expenditures made in that year. The percent, which can be up to 100%, is determined by the Ministry of Trade and Industry. The ITA tends to be favored by large, capital-intensive 13/ This corresponds to findings by D. Usher, "The Economics of Tax Incentives to Encourage Investment in Less Developed Countries", Journal of Development Economies, June 1977; and Dr. Abdullah Tahir, "Tax and Investment Incentives in the Less Developed Countries: A Case Study of the Manufacturing Sector of Malaysia", Ph.D. dissertation, American University, Washington, D.C., 1984. - 73 - projects. Alternatively, companies can apply for Reinvestment Allowances and accelerated depreciation allowances under the provisions of the Income Tax Act. These options ensure that a broad range of companie: do indeed enjoy substantial relief from Malaysia's high nominal statutory corporate income taxes. 3.67 The low effective tax rate on capital may have contributed to an erosion of the tax base. It may also have had an unintended consequence in biasing projects towards more capital-intensive techniques. Although the tax system has been moved away from an explicit pro-capital bias, through removal of the links between the length of a tax holiday and/or the size of investment tax allowances with the magnitude of investment, an implicit bias still remains. The simulation model shows that if corporate contributions to the EPF are viewed as an implicit tax on labor, then there is a sizeable increase in the effective tax rate associated with greater use of labor. In fact, imputing the full EPF corporate contribution as a tax on the corporation leads to a rise in the METR from 15S to 30Z for a project where wages are 34Z of gross value added (the manufacturing sector average). A project with a 50% labor share, but the same before-tax rate of return, would be subject to a METR of 44%. By lowering the tax on capital through tax holidays and the dividend offset, while taxing labor through the EPF, the simulations indicate a strong bias towards increasing the capital-intensity of production in Malaysia, as observed in the 1980s. 3.68 The bias in the tax system towards capital-intenisity is most likely unintended, and is the primary source of inefficiency introduced via the corporate tax system. The problem is not that tax rates are too high or that they dissuade investment-in fact, a crude measure of the average tax rate, derived by taking the ratio of non-oil corporate income tax to the previous year's gross profits, hovers around 12% in the 1980s, with no signs of decline. The issue, rather, is one of efficiency in production. The growing capital intensity of manufacturing has sharply curtailed employment growth in this sector. Furthermore, some types of investment in vehicles and in enterprises where capital gains rather than current income is the motive, are favored by the tax codes, while other investments, including machinery and equipment are in effect taxed more highly. Similarly, because of the dividend offset, domestic companies are favored over foreign companies, whose citizens may not benefit from such treatment. 3.69 Simulations are based on purely hypothetical exercises. A financial analysis carried out by the International Finance Corporation for a potential aquaculture project in Sabah, however, shows how lax the corporate tax system is in practice. Their analysis indicated that the expected economic rate of return would be 18%, while the financial rate of return would be expected to be 17X. The difference, of one percentage point, is the net impact of taxes on profitability. Clearly this is insignificant in any corporate decision- making. 3.70 As the corporate income tax is playing neither an important revenue- raising role, nor enhancing efficiency, an overhaul is desirable. Given that other countries in the region, notably Singapore, are moving towards corporate tax structures with lower nominal rates, Malaysia would be well-advised to - 74 - follow suit. It must take care, however, not to exacerbate the bias towards capital-intensity in so doing. For example, simulations show that the METR would be reduced from a base rate of 6% to -6X if the statutory tax rate were lowered from 451 to 30% with a full offset provision. Thus, any reduction in nominal corporate tax rates must be countered by removal of exemptions and investment allowances. 3.71 One final point is worth noting about the current corporate tax system. Because of the low effective rate associated with pioneer status, there is little gain to companies from additional tax incentives. Thus, several tax devices designed to further specific government objectives, such as the labor utilization relief, locational abatement of income, abatement of income for compliance with NEP and double deductions for labor training, have proven to be highly ineffective. A tax structure with fewer exemptions would permit more effective, selective tax incentives than the present system. Investment Trends 3.72 When new capital formation is efficiently allocated, it can comple- ment factor productivity growth rather than substitute for it. In Malaysia, however, following an investment boom in the early 1980s, there has been a sustained decline in new capital formation, particularly in the private sec- tor. Private sector gross fixed capital formation in 1987 amounted to just 13.51 of GDP; excluding oil and gas investments, the figure falls to 11Z. This is still well below the average rate during the 1970s which remained relatively constant until 1981 (Figure 3.2). 3.73 Current levels of private investment are barely sufficient to cover annual wear-and-tear and obsolescence, let alone increases in the capital stock per employed worker. Considering that about one-fifth of private non- oil and gas investment goes into residential housing, only some M$8.6 billion or 10.91 of GDP actually went towards new plant and equipment in 1987; about one-quarter of this came from foreigners. This decline in capital formation would be of less concern if it reflected a fall in construction activities, following the overbuilding of office space in 1980-84. The data, however, indicate that the trend is exactly the reverse. Spending on machinery and equipment has fallen even faster than total capital formation. Machinery accounted for 50% of investment at its peak in 1981 and has since seen its share decline to 45% in 1986. While this is due in part to a slow-down of investment in the capital-intensive NFPEs, it underscores the decline of private investment in productive assets. 3.74 At an aggregate level, insufficient resources are now being devoted to capital formation to permit a resumption of rapid growth in the long-term. At its peak in 1981-83, Malaysia was devoting over 371 of GNP to new capital. In part, this is an abnormal figure reflecting two cyclical phenomena: first, large-scale investments in the infant oil and gas sector, which accounted for over 5% of GNP; and second, an increase in public infrastructural spending which grew to over 13% of GNP from a historical level of about 81. With investment in these areas returning to more normal long-run sustainable levels, the investment/GNP ratio required for 6Z GDP growth would be around 28-30%. This figure is significantly higher than the investment ratio during COMPOSION OF CAPITAL FORMATION (% of GNP) 0.4 0.35 / ~~~~Total gross ie 0.3 /capi formaton 0.25 0.2 Private GFCF U 0.15 k P. -nn 0.1 i ~~~~~~~~~~~~~~~~~~~Foreign direct 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 0.05 k y | !~~~~~~~~~m s - 76 - the booming 1970s, but is nevertheless appropriate given the transition towards manufacturing as the leading grcwth sector. In general, manufacturing has a higher capital/output ratio than agriculture (excluding land), so the shift in growth from agriculture to manufacturing will require greater invest- ments. If the private non-oil sector in Malaysia were to regain the invest- ment spirits of the 1970s and return to a level of capital formation of 15% of GNP, both the composition and overall magnitude of investment would be brought close to a long-run equilibrium. 3.75 Data showing the sectoral decomposition of investment, especially of private investment, is incomplete and has been gathered from several different sources, including EPU, Treasury, DOS and MIDA. For the most part, surveys are only available upto 1985, precluding analysis of the most recent period of the collapse in investment. Despite the data discrepancies and incomplete- ness, some broad trends emerge: the boom in the early 1980s favored nontraded goods sectors such as construction, trade and financial services. Growth in fixed assets in manufacturing, while steady, has been less spectacular. Investment in the important commodity sector, too, was flat. (a) Oil vs Non-oil Private Investment 3.76 The breakdown of private investment between oil and gas and other sectors is shown in Statistical Appendix Table 7.9. Oil and gas investments include both upstream exploration, development and production activities and downstream processing. During 1985 and 1986, the investment in oil and gas, particularly downstream activities, has been substantially reduced, falling from M$3,190 million in 1984 to M$1,219 in 1986. Non-oil and gas investment also fell but not as fast as oil investment. It fell from M$10,155 million to $6,421 million. As a result, the share of oil and gas investment in total private investment fell from 23.9% in 1984 to 16.0% in 1986. Oil and gas investment is expected to recover somewhat in 1988, however, owing to plans for big gas projects, but then would likely stagnate or decline in the medium- term. (b) Investment by Industry 3.77 Decomposition of investment data by industry can only be done indirectly. There are several sources which provide indicators of investment trends in each industry. One is the Department of Statistics which conducts annual surveys of Manufacturing and Construction, and a census of Stone Quarrying and of Rubber Estates. However, these censuses do not cover all industries and are only available at an aggregate level for the years 1978, 1979, 1983 and 1985 (Statistical Appendix, Table 7.8). Another source is the Financial Survey of limited companies conducted by DOS, which covers companies whose annual revenues are $M5 million and above. This reports the investment in fixed assets by each industry, and covers more than 4,500 firms in Malaysia. The data do not correspond to investment in the national income accounts sense since they include the purchase of used equipment and land. The data also do not include investment in certain other sectors such as housing and real estate. Nevertheless, they should at least reflect the trends of investment in selected sectors. - 77 - 3.78 The breakdown of investment by industry is shown in the Statistical Appendix Table 7.7. The financial survey covers about 701 of private invest- ment. Although the yearly data fluctuate as the number of respondents varies, the broad trends are similar in the two data bases. The data show that the largest share of investment (about one-third) goes into manufacturing. Mining has, on average, accounted for another quarter, with petroleum and gas repre- senting 951 of the sector. Agriculture and trade each gave a further 101. 3.79 Investment in manufacturing had grown steadily in the early 1980s, although it fell slightly in 1985 and, based on MIDA approvals, may have i fallen by another 10% in 1986. Year-to-year figures, however, must be inter- preted with caution as they can be strikingly affected by a few large ; projects. According to approvals data, during the period of 1981-86, indus- trial chemical and chemical products got the highest share (14.4%) of invest- ment approval followed by nonmetallic products (13.5X), food manufacturing (9.9Z) and petroleum and coal industry (9.3%). Electrical and electronics and transport equipment also had a steady and substantial amount of projects * approved. During 1985-86, food manufacturing, and petroleum industries had big increases in project approvals, while wood and wood product and basic metal industries experienced a decline. 3.80 The investment data signify that major structural shifts in the Malaysian economy are occurring. Three notable features stand out as contri- buting to the observed decline in private investment. First, there is a plateau or decline in investment in traditional commodities, including rubber, palm oil, petroleum and tin. Partly this can be attributed to the decline in commodity prices; but partly it signifies the completion of major phases in development--the decline in importeice of tin mining and the completion of the rapid expansion phase of palm oil and petroleum. In agriculture, only cocoa investments are growing rapidly and these aie still small in size. 3.81 The second feature is the interruption of the growth in manufactur- ing investment after 1980. This coincides with the period of rapid apprecia- tion of the real effective exchange rate and the growing entry of public enterprises into manufacturing. In terms of subsector approvals, large public enterprises in fertilizer, cement and automobiles account for the dominance of these sectors in project approvals. Sectors where the private sector is more active were less buoyant, except food manufacturing. This sector, however, enjoys very high protection because of the agricultural diversification program and is consequently highly distorted. There are a substantial number of bankruptcies in the food sector that have resulted from the particular structure of incentives prevailing on both the manufacturing and food produc- tion sides. 3.82 The third feature concerns the rapid development of the nontraded sector during the early 1980s. Construction, wholesale and retail trade and banks all expanded fixed capital at a rapid rate. The extension of new branches by the banking system was initiated by substantial liberalization of licenses by the Central Bank. Total commercial bank branches grew from 509 in 1978 to 837 by 1986. These sectors are now all less likely to grow as rapidly as before. - 78 - 3.83 Determinants of Investment. What are the determinants of private investment and what policy options are open to government to support it? To address this question, it is useful first to take a macroeconomic perspective on investment. This gives an interpretation of how iT u stment moves in response to changes in other macroeconomic variables.- A more detailed sectoral and firm-level perspective can then be used to evaluate how policy reform may influence the major variables of significance for investment. 3.84 In the analysis below, private domestic investment is isolated by subtracting public investment and foreign direct investment from total invest- ment recorded in the national income accounts. Public investment is derived from budget figures. It includes the investment of federal and state govern- ments, statutory bodies and NFPEs. Prior to 1980, only 10 major NFPEs are included. After 1980, coverage is extended to 40 NFPEs. This does not, however, create a major distortion in the data because prior to 1980 there was relatively little capital formation by public enterprises. 3.85 Data for foreign direct investment is taken from the balance of payments. The series includes net new flows of investment plus reinvested earnings of foreign corporations in Malaysia, plus changes in other foreign liabilities of foreign companies. For the most part, foreign investment has gone into separate export processing zones and is subjected to a very differ- ent set of incentives, tax regulations and capital market restrictions. For these reasons, it is treated separately from domestic investment. 3.86 Regression analysis was used to explain the determinants of private investment. Several factors were found to have a significant impact. The most important of these is the level of profits in the economy. The higher the level of profits, the higher the level of investment. The regression results suggest that about one-third of all profits are reinvested. The cred- ibility of this result is strengthened by the fact that this is very close to the proportion of shareholders' funds in total liabilities computed in the financial survey of the Department of Statistics. To a sizable degree, the recent collapse in private domesLic investment can be explained by the fall in profits in the economy as national income has shrunk and the wage share risen. 3.87 Profits are important for investment decisions for three reasons. First, they indicate the returns that may be expected from new investment, assuming the current environment remains constant. Second, they generally follow the domestic business cycle and are a proxy for capacity utilization rates. During a boom when capacity is fully utilized, profits rise. Thus, through signaling tight capacity, high profits indicate the need for capital expansion. Third, profits provide the revenue with which to undertake capital expansion and the basis on which leverage through borrowed funds may be obtained. Confirmation of the importance of this is given by data in the Business Expectations survey conducted by the DOS which indicates that 80-90Z of capital investment by firms is financed through their own funds. 14/ The detailed methodology is presented in Appendix 5 to this Volume. - 79 - 3.88 The second most important variable in explaini ^ private investment is government investment. Based on details of the sectoral composition of public development expenditure, government investment has been decomposed into two elements. Production investments are those which directly contribute to value-added in the future through their own activities: agriculture, housing, commerce and industry are examples. Infrastructure investments are those where returns are only realized through complementary activities: land devel- opment, drainage, irrigation, health, education, water supply, transport and energy. The analysis demonstrates that private investment responds strongly and positively to infrastructure investment. As public investment in infra- structure has been sharply curtailed in recent years, private investment has also declined. 3.89 Production investments have no such complementary effect with private investment. In fact, the analysis indicates that s.me substitution occurs--public investment in areas like housing development, steel and cement crowd out private investments in these fields. Such substitution takes place in several ways; it may be in the same industry or in industries with upstream or downstream linkages. For example, public investment in cement and steel production has created excess capacity and effectively barred potential private investment in these sectors. To avoid losses throughout the industry, prices have been controlled and set to cover average domestic costs. During the 1980s, these have been on average 20-50% higher than the potential c.i.f. import price. Thus, all downstream industries, especially construction, have been negatively affected. In some instances, such as automobiles and housing, public enterprises have competed directly with their private counter- parts for a share of the domestic market. In still other instances, govern- ment procurement practices have favored other public firms, squeezing out upstream private firms. 3.90 At the macroeconomic level, the substitutability betweea public and private firms, while evident, does not appear strongly in the data. Partly, this may be because public production investment is still small compared to overall private investment. In some large sectors such as agriculture, com- merce and trade, the public sector has only a token presence. It is in manu- facturing that the greatest competition exists. In fact, the size of public corporations in the manufacturing sector is about equal to the size of private domestic corporations. 3.91 Exogenous factors also enter into investment decisions. The pull of domestic demand is captured through an accelerator term that measures the change in GNP. This term has imparted a consistent downward impulse to investment in recent years which should now be reversed as the recovery takes hold and strengthens on the back of more favorable commodity prices. Competi- tiveness, proxied by the real effective exchange rate, is also important. The recent real effective depreciation of the currency, following the US dollar's fall, has helped prevent an even larger drop in private investment in 1986 and 1987. - 80 - Policies to Encourage Private Investment 3.92 Private investment is partly subject to autonomous forces, over which the government has little control. The short-term prognosis for these is positive. Favorable terms of trade, rising incomes, falling wages and interest rates and a depreciating currency will serve to boost private invest- ment without any intervention by gover-nment. Recently, Government has under- taken several new initiatives to encourage investment. Liberalization of industrial licensing, reduction of administrative delays, and new and expanded credit schemes are examples. The full impact of these measures has yet to be felt, but initial indications are that more needs to be done to support private investment over the medium-term. A concerted effort by government to coordinate macroeconomic, structural and firm-level policies is required. 3.93 Macroeconomic Policy. At a macroeconomic level, government policy should be aimed at restoring international competitiveness, which would likely involve raising the share of profits in national income to historical levels and reorienting government expenditure to support private investment. Government influences the distribution of income between profits and wages in two ways. First, as it is a large, perhaps dominant, force in formal labor markets, public pay and employment policy sets the tone for other wage agree- ments. The labor market is presently adjusting to the excess wage awards of the early 1980s, and it is important that government assist this adjustment by continuing to ref 1e,t market signals in the wage increases announced for public employees.- 3.94 One important determinant of wages should be labor productivity in manufactur.ng, in order to ensure that international competitiveness is main- tained. Because of Malaysia's policy of floating the exchange rate and its open capital markets, the real effective exchange rate is likely to continue to be volatile in the future. It cannot be used as a direct instrument to counteract any incipient wage pressures on competitiveness, except to a limi- ted degree via mana&ement of foreign exchange reserves. In such an environ- ment, wage negotiations take on more importance and a greater degree of flexi- bility in wages is required as a buffer against international shocks and exchange rate developments. Government should encourage the greater use of variable compensation packages in the private sector, which have a significant portion of annual payments dependent on productivity and profitability mea- sures. 3.95 The second important action government c&r take is to reduce EPF contribution rates. The empirical evidence from Malaysia and other countries on the incidence of EPF contributions between profits and wages does not demonstrate in convincing fashion that lower EPF contributions would necessar- ily result in higher profits. Nevertheless, there is a clear coincidence between the time of increased EPF contribution rates and the period of declin- ing profit shares that suggests a causality. Civen current conditions of 15/ More detailed recommendations on government interventions in the labor market are discussed in Volume II, Chapter 3. - 81 - excess labor supply, reflected in the unemployment numbers, it is likely that a cut in EPF contributions would increase demand for labor, reduce the average compensation package paid by corporations, and generate increased output and profits. 3.96 Another macroeconomic adjustment that needs to be made is in the composition of government expenditure. In recent years, the bulk of adjust- ment to the financial problems facing the public sector has come from a reduc- tion in infrastructure Investment. This was appropriate to some extent, as infrastructure investment had risen to unreasonably high levels during thg early 1980s; for example, the recent World Bank transport sector report - concludes that the Malaysian transport system, except for some additional road requirements, is adequate to meet expected needs through 1995 without major new investments. Other big-ticket infrastructure spending, such as on irriga- tion for padi, has also been appropriately scaled back. But in some cases, sueh as communications, flood control and drainage, infrastructure cuts and postponement have probably contributed towards the worsening climate for private investment. Furthermore, there are indications that reduced mainte- nance expenditures have already begun to take their toll in the form of deter- iorating infrastructure. Urgent actions are required, particularly in road maintenance, to salvage the benefits from past infrastructure investments. Reorienting government expenditure towards maintaining and improving infra- structure would boost private investment. Structural Policies 3.97 Many of the structural policies to improve efficiency, discussed above (paras. 3.29-3.61), will also have a favorable impact on private invest- ment. In addition, government policies could help reduce further the costs of investment and the risks associated with investment by reform of pricing policy and regulation of construction and construction materials and institu- tional corporate funding structures. 3.98 Construction Costs. Since 1979, the prices of steel bars and cement have been under control. These are the two most important materials for construction, accounting for 25% to 75% of the material used. Construction, in turn, is about one-half of fixed capital formation. Table 3.7 compares the local controlled price of steel bars with an estimate of the cost of imported steel bars, derived by taking the Japanese f.o.b. export price and adding 20% for freight, insurance and handling. The imported price is expressed as a range from -10% to +10% of this estimaL;. t:o compensate for different products. Since 1982, the estimated domestic price has been about 50% greater than the import equivalent price. Since local prices 3re ex-railyard, while the import price contains a generous provision for delivery on-site, the true difference to an end-user may be even greater. 16/ "Malaysia: National Transport Policy Review," The World Bank, 1988. - 82 - Table 3.7: LOCAL AND INTERNATIONAL PRICES OF STEEL BARS Year Range of local prices Range of International Prices Difference (M$/ton) /a (M$/ton) /b (X) /c 1980 730 - 818 772 - 944 +6% - +151 1981 830 - 917 787 - 962 -4Z - +51 1982 917 - 1,024 614 - 751 -331 - -25% 1983 891 - 995 559 - 683 -37X - -31Z 1984 891 - 995 589 - 720 -34% - -28% 1985 891 - 995 603 - 737 -32% - -261 1986 891 - 995 613 - 749 -31% - -25X 1987 891 - 995 552 - 675 -38X - -32% /a The prices of steel bars with diameters of 3/8", 1/2", and 7/8", ex rail yard in Peninsular Malaysia. /b Staff estimates. Based on export prices of reinforced bars from Japan plus 201 for freight, insurance, and handling charges. The range repre- sents -10% to +101 of the average price. /c Percentage difference of the international prices over the local prices. Source: Staff estimates; Master Builders Association, Kuala Lumpur. 3.99 Undertaking a similar price comparison for cement is less reliable because of the greater relative importance of transport costs on which little information is available. Nevertheless, available evidence suggests that the local price range of M$160-180 per ton may be 20Z to 30% higher than the potential import cost. While import prices may reflect dumping due to excess capacity in East Asia, the medium-term nature of the price differentials suggests that tariff and quota protection have been costly, and that a more efficient solution would involve deregulated prices and other forms of indus- trv prCtection. Efficient import substitution would be reflected in a narrowing over time of the differential between domestic and international prices. This has clearly not happened. The issue of how long to protect a fledgling domestic industry against import competition, whether "dumped" or not, must be addressed. 3.100 The cost of regulated steel and cement prices has been twofold: first, there is a substantial cost associated with the inappropriate levels of added capacity and with the failure to allow a shake out in yM,ch the limited demand would be met by the most efficient domestic produces; - second, there is the downstream cost of higher investment prices that is borne by the whole economy. A deregulation of steel and cement prices should, therefore, be 17/ See Vol. II, Chapter 1 for a discussion of the cost differentials between private and public producers of steel and cement. - 83 - considered. The impact would be to lower the cost of investment by 4-52. That is, the same dollar amount of investment would purchase a 4-5X increase in real capital formation, twice the gain actually recorded in 1987. On the other hand, prices are regulated in order to cover the operating costs of public enterprises producing at low capacity. Deregulation would require further government subsidies to the industry and cause a higher deficit unless a program of consolidation and restructuring to improve efficiency 8as to be implemented at the same time. 3.101 Other government policies also affect the cost of investment. The most important of these is land-use regulations. The complexity of prccedures to alienate land for industrial use lengthens the time period for investments to mature and slows down the response time of new investment to market signals. As a result, chauges in the investment climate are felt in sharp speculative price movements for industrial land and factory buildings. Table 3.8 shows a range of price movements for industrial land and property in Selangor State. Because these prices vary greatly according to location, size and condition-- variables which may not be represented equally in the samples on which the annual valuation reports are based--they should be treated with caution. The trend, however, is clear: the same factory building which cost M$100 in 1981 might have cost over N$280 by 1985 and be back to its 1981 price at present. These swings imply that often the financial success of an investment has had more to do with the timing of the venture than with the efficiency of the manufacturing process itself. Some States have already liberalized land-use regulations, and have established Committees or Task Forces on Industrialization to expedite approval of applications for land and factory buildings. Within some industrial estates, approval is now very rapid. In other areas, however, more concerted efforts are required to coordinate land- use licensing, industrial site planning, infrastructure, and other code and licanse requirements. Speeding up the elapsed time from project initiation to production would help stimulate investment by reducing uncertuinty and by generating a more competitive investment process which would iron out price cycles and force producers to compete on the basis of the efficiency of pro- duction rather than on the ad hoc timing of investment decisions and license procurement. - 84 - Table 3.8: PRICE TRENDS OF INDUSTRIAL PROPERTIES /a AND LAND /b IN SELANCOR STATE (Z change over previous year) 1982 1983 1984 1985 1986 1987/c Property n.a. +20 +5 to +10 0 -24 to -30 -30 to -40 Land +5 to +15 +20 to +40 +20 to +30 0 -15 to -20 -30 to -35 /a Factory buildings without equipment installed. rb Vacant land developed for industrial use. T7 Estimate. Source: Property Market Report, various issues, Ministry of Finance; preli- minary 1987 estimates from Selangor branch, Evaluation Department, MOF. 3.102 Funding Structures. Perhaps because of the dominance of government and foreign corporations, the private domestic manufacturing sector in Malaysia has not relied on credit financing mechanisms for expansion. Most fixed investment is financed through own funds, that is retained earnings or new equity--only about 20Z is borrowed. This has resulted in a very conservative debt structure; overall, the manufacturing sector had a debt/equity ratio of 1.3 in 1985, compared with 3.6 in Korea, 3.2 in Japan, 2.4 in the Philippines and 2.1 in Turkey.- This is perhaps the main reason why monetary condi- tions, such as availability of domestic credit and the average real lending rate, do not seem to have had a significant impact on aggregate private domestic investment. 3.103 Government attempts to stimulate private investment through the financial system have focused on credit programs and directed credit guidelines for commercial banks. Special funds, such as the New Investment Fund, the proposed Industrial Adjustment Fund, the Export Credit Refinancing Scheme, and schemes to foster small-scale enterprises and the Bumiputera community have had, as one objective, the infusion of credit into manufacturing. Yet despite the success of these schemes, in terms of the amount disbursed, the share of total credit allocation by the banking system to the manufacturing sector has slipped from 20Z in 1980 to 15Z in 1986. This suggests that while government- induced lending schemes have been attractive, perhaps because of the favorable interest rates charged, they have likely substituted for other banking system credits. 18/ Comparator country data refers to 1983. - 85 - 3.104 The problem does not appear to lie with the banks per se. The survey of manufacturing companies by Bank Negara suggests that overall tightness in credit has eased subs-.v-ially since 1981, and that banks' lending attitudes have become .ncreasingly tolerant and less tight (Table 3.9). During 1987, there is additional evidence of ample liquidity in the banking system and falling domestic lending rates which now average about 8% in real terms, but little demand for loans. Table 3.9: DIFFICULTIES IN BORROWING FROM BANKS (X of total firms surveyed) Credit situation 1981 1982 1983 1984 1985 1986 Overall Funds Position Easy 11 13 19 20 14 22 Not so tight 35 55 51 46 46 44 Tight 54 32 30 34 40 34 Banks' Lending Attitude Easy 7 16 24 23 22 31 Not so tight 42 58 52 43 45 41 Tight 51 26 24 34 33 28 Source: Manufacturing Survey, Bank Negara Malaysia. 3.105 The financial problems ;acing firms, therefore, seem not to be in tradi- tional credit markets, but in the market for risk capital and other long-term financing. Given the volatility of manufacturing profitability in an economy where the real exchange rate and domestic wages swing with terms of trade changes and other international developments, there is a need for a conserva- tive debt/equity posture. Well-managed companies are unwilling to take on more debt without additional equity or retained earnings. 3.106 Structural problems in obtaining long-term and risk capital are at the heart of the financial constraints on manufacturing growth. Unless these are addressed, there is little prospect that the scaling-back of public investment will be replaced by a sustained surge in private investment. The demand for risk capital c-n be indirectly gauged by the phenomenal growth in the cor- porate securities market until 1984. In that year, M$2.4 billion, or 10% of total investment, was raised in corporate securities compared to M$0.2 billion (1.3X of investment) in 1980. But fraud and scandals at home and in Singapore have completely undermined this market. Adding to the problem is the lack of access to long-term credit; development banks are small, comprising only 2-3% of financial assets, and there are no strong relations between banks and industry as in countries such as Japan and Germany. - 86 - 3.107 Three factors have created a squeeze on risk capital that has damaged manufacturing investment in the 1980s. The first, a decline in corporate profits, has already been touched upon above; the second is the distortions generated by the regulations governing new public offerings; and the third is related to the equity requirements of the NEP. 3.108 Although the market capitalization of the KLSE is one of the highest in Asia, the annual t.iading value is small in comparison. This indicates that the secondary market for shares is quite small, and that large blocks of shares are closely held. It also contributes to the volatility of stock prices, with the result that the market is viewed by investors primarily in speculative terms and not in terms of providing long-term capital for invest- ment. Thus, over the last decade, the number of companies listed on the KLSE has only risen by 21. 3.109 A major source of thinness in the market is the unwillingness of many successful companies to go public. A Capital Issues Committee (CIC), chaired by the Governor of the Central Bank, in practice controls the new issue price of corporate stock in a range that varies by sector (Table 3.10). For manu- facturing, the price of a new issue can range between 4.0 and 5.5 times projected maintainable pretax earnings. Given the fact that the average market P/E ratio was 32 in December 1987 (after the global crash), the subsidy element in these guidelines is evidently substantial. As a result, few new issues are brought to market. Those that are, are heavily oversubscribed and usually see a meteoric rise in value. Table 3.11 shows the initial price and the price after five business days of the five new issues during 1987; the capital gains ranged from 108Z to 398x. Table 3.10: THE CIC-GUIDED P/E RATIO Price/Earnings Sector Ratio Trading/services 3.5 - 4.0 Property 3.5 - 5.0 Manufacturing 4.0 - 5.5 Contracting and construction 4.0 - 5.5 Transportation 4.0 - 6.0 Tourism 4.5 - 6.0 Plantations 5.0 - 6.0 Insurance 5.0 - 7.0 Finance companies 6.0 - 8.0 Commercial banks/merct, -t banks 8.0 - 12.0 Source: Ministry of Finance. - 87 - Table 3.11: CAPITAL GAINS FOR NEW ISSUES, 1987 Karket price at end of fifth day X Offer price following listing gain (M$) (NO) MISC 2.40 5.00 108 Dreamland Holdings 1.00 3.10 210 Sports Toto 2.00 9.95 398 Southern Bank Bhd. 2.20 5.05 130 Georgetown Holdings 1.00 2.86 186 Source: "Malaysian Securities Market Departments," M. Merican, processed paper presented at a conference on Recent Securities Market Developments, Kuala Lumpur, October 19-20, 1987. 3.110 The CIC also fixes prices of shares that may be swapped in a merger agreement (ior both companies, whether currently listed or not), and prevente companies from acquiring temporary loss-making or nonincome-generating assets, since guidelines specify that there should be no earnings dilution. It also governs Rights issues. CIC policies, taken in combination, serve to reduce the supply of listings on the exchange, thereby perhaps boosting the prices of listed companies and generating windfalls for selected investors in new issues, but penalizing new entrants to the market and investors unlucky enough not to win the lottery that rations applications for new shares. 3.111 The ytderpricing of new shares is not a phenomenon exclusively observed in Malaysia,- -nor is it always a negative symptom. Regulators have a responsibility for protecting investors and fledgling underwriters and ensuring a smooth market. When the underpricing becomes excessive, however, important public policy and social welfare consequences arise. First, there is a burden on firms' financial structure as they receive less from a new issue than otherwise. Second, where public corporations are involved, the system is open to charges of selling public goods at artificially low prices. Underpricing is partly a result of administrative control, but also reflects the institutional structure under which shares are allocated. Given underpricing, shares become oversubscribed; but this raises investor's costs as they must finance large credit lines with only a low expectation of being allocated new shares. In order to enjoy a fair return, then, the degree of underpricing is accentuated. Several schemes could be considered by the 19/ See "The Underpricing of initial public offerings in Singapore: Public policy issues and possible solutions". J. Lim and A. Saunders, New York University Working Paper No. 487, August 1988. - 88 - authorities to reduce initial underpricing and the ensuing volatility of share prices (both up and down). A competitive auction could be used, as in France. The disadvantage of this is that the underwriting risk is borne by the issuing firm. Alternatively, an over-allotment option could be given to firms, as in the USA, where the underwriter would have the option to expand the quantity of offered shares in the event of oversubscription. 3.112 The third constraint in company generation of risk capital stems from the implementation of the NEP. In part, this is linked to the guidelines on public listing of new securities. Any firm which goes public must meet NEP criteria of at least 30Z ownership by Bumiputeras, in addition to the minimum 10-15% that must be publicly offered. Thus, a non-Bumiputera investor can retain only just over one-half of incremental shares from a new listing. This exacerbates the subsidy element implicit in the CIC guidelines. A more serious problem, however, emerges for private companies because of problems in matching the incentives of Bumiputera and non-Bumiputera investors with the NEP equity guidelines. With the rapid development of an entrepreneurial Bumiputera class, there is a growing incentive towards 100% or majority- controlled Bumiputera firms. These firms benefit from various government loan, tax and preferential procurement privileges and have the managerial sophistication and business experience to thrive. On the other hand, the small Bumiputera saver, who does not want to participate actively in a busi- ness venture, has access to a series of mutual funds (such as ASN and LUTH), which offer returns commensurate with equity investments combined with minimal risk. There is, therefore, a shrinking pool of Bumiputera investors prepared to participate as minority joint-venture partners, and a gradual shrinkage of new equity mobilized from non-Bumiputera Malaysians. According to MIDA survey data, the actual new equity generated from non-Bumiputera sources ha fallen precipitously since 1984 and turned negative in 1986 (Table 3.12)± ±I As non- Bumiputera Malaysian residents still own about ;0% of the manufacturing share capital of limited companies, their participation in future growth is essen- tial. 20/ Note however that other sources of data, such as proposed called-up capital, do not show the same decline. Data weaknesses, therefore, do not permit firm conclusions to be drawn. - 89 - Table 3.12: ESTIMATION OF ACTUAL EQUITY GENERATED IN INDUSTRIAL ACTIVITIES (million ringgit) 1980 1981 1982 1983 1984 1985 1986 Bumiputera 394.9 428.9 683.8 633.6 783.1 1,327.4 904.8 (Z) 28.60 35.58 37.41 43.42 37.80 73.35 73.91 Non-Bumiputera 566.8 358.9 576.3 642.5 855.8 25.4 -85.3 (X) 41.05 29.77 31.53 44.03 41.31 1.40 -6.97 Foreign 419.1 417.7 567.8 183 432.8 456.9 404.7 (Z) 30.35 34.65 31.06 12.54 20.89 25.25 33.06 Total 1,380.8 1,205.6 1,827.8 1,459.2 2,071.7 1,809.8 1,224.2 Note: Data are only for projects approved by MTI and are laot representative of the entire industrial sector (small projects are excluded). Sample may also suffer from non-response bias. Some reclassification occurred in 1986 as a result of improved survey coverage and detail. Source: MIDA. 3.113 Government policies in supporting investment financing should shift from their current emphasis on directed credit towards support for mobilizing risk capital and long-term credit. Reform of stock market procedures is a high priority, particularly for large firms. It takes on added importance in the light of the privatization program. It is unlikely, however, that medium and small companies will benefit in the medium-term form such reforms. These companies would be best assisted through the development of venture capital companies which, through equity participation, could help achieve NEP objec- tives and help ease the problems faced by non-Bumiputera investors in comply- ing with the NEP at reasonable cost without giving up operational control. Foreign Direct Investment 3.114 Foreign direct investment remains crucial in stimulating industrial growth. The foreign-owned corporate industrial structure is about one-third of the total industrial base. At around M$2 billion annually, foreigners contribute about as much to total industrial capital formation as do private domestic investors. Because of its prevalence in FTZs and its export- orientation, foreign investment reacts to a different set of incentives than private domestic investment. External factors, such as international competi- tiveness and past levels of foreign investment are important, whereas internal factors like domestic demand and public investment are less relevant. As expected for exporting companies, the level of the Malaysian real effective exchange rate is also highly significant. - 90 - 3.115 Regulations governing foreign direct investment have recently been eased. Export-oriented foreign companies have been exempted from the NEP, to all intents and purposes. The impact of this liberalization, however, is still uncertain. While there has been a modest increase in foreign direct investment in 1987, much of this can be attributed to an impulse stemming from currency movements and from the fall in interest rates. Overall, foreign investment into Malaysia remains depressed. In fact, after a boom in 1981-83, foreign direct investment in Malaysia has collapsed to half this level (Table 3.13). This pattern is generally shared by Singapore and the Philippines, and seems to be attributable to the decline in major commodity prices, discouraging resource-based investments especially in oil and gas, and to the worsening competitiveness of these countries vis-a-vis other East Asian economies. - Table 3.13: INFLOW OF FOREIGN DIRECT INVESTMENT (Million SDRs) Phil- Singa- Malaysia Thailand ippines pore Indonesia Korea Taiwan 1970-74 (ave.) 184 73 2 241 77 69 52 1975-79 (ave.) 362 52 92 444 258 59 61 1980 718 146 -82 860 138 6 128 1981 1,073 249 146 1,195 113 86 128 1982 1,266 175 14 1,260 205 62 94 1983 1,179 327 98 931 270 65 139 1984 778 394 9 862 221 109 196 1985 684 159 -11 959 304 227 335 1986 452 219 86 574 221 365 278 Source: IMF, Balance of Payments Yearbook, and the Central Bank of China. 3.116 DeveJoped countries seem to find Taiwan and Korea relatively more attractive as they adjust their global production and marketing strategies. These economies have fairly strong industrial skills and technological sophis- tication which permit them to absorb new techniques easily. Coupled with excellent work ethics, a well-developed infrastructure and a sizable domestic market, there are clear advantages to investment there. Japan, for example, is expected to rapidly deepen intra-industry specialization with the Asian NICs in sectors such as electronics and machinery. It may also move the pro- duction of many finished products, for Japanese and export markets, out of Japan. 3.117 The shift towards the NICs is particularly apparent in the project commitments of Japanese direct foreign investment (Table 3.14). Malaysia has clearly lost out in the battle to attract new Japanese investors despite recent moves to liberalize the economy. Hong Kong has been the greatest beneficiary, but Indonesia, Singapore, and even Thailand have attracted - 91 - increasing numbers of Japanese foreign investors in recent years. Malaysia seems to be more successful in encouraging second-tier foreign investment, from the foreign-ezchange rich "four tigers." Much of this is likely to be geared to reducing trade frictions on the part of these countries, particu- larly vis-a-vis the United States. With its open economy, passage of the Copyright Act and small bilateral surplus with the United States, Malaysia is in a strong position to challenge successfully future attempts at protection- ist legislation aimed at products from its shores. The strategy to attract foreign investment should build on this national advantage. This implies that great care must be taken to avoid "shell" companies that simply reroute pro- duction, sometimes without the goods even being transshipped through Malaysia. It also implies that Malaysia's success in trade diplomacy, in both multilat- eral and bilateral form, will become a more critical component of its indus- trial strategy. Table 3.14: JAPAN'S DIRECT INVESTMENT TO ASIAN COUNTRIES (US$ million) 1984 1985 1986 1987 (1-6) Number Amount Number Amount Number Amount Number Amount Hong Kong 119 412 105 131 163 502 131 644 Korea, Rep. of 57 107 75 134 111 436 99 295 China (Taiwan) 68 65 68 114 178 291 138 136 Singapore 108 225 110 339 85 302 83 259 Indonesia 82 374 62 408 46 250 33 397 Malaysia 63 142 60 79 70 158 36 37 Thailand 76 119 51 48 58 124 79 91 Source: Ministry of Finance, Japan, "Statistics of Applications for Overseas Direct Investment." B. Growth Outlook--the Demand Side 3.118 The potential output growth reflects Malaysia's capacity to produce. As the past few years have shown, however, actual output may fall well below potential output. During the 1980s, growth was achieved almost entirely on the basis of external demand. Only 29Z of the incremental production between 1980-87 was absorbed domestically; of this, 23 percentage points were absorbed by the public sector. The miniscule level of private demand, which grew at an annual average rate of only 0.4% in 1980-87, must be raised for effective long-term growth. The central problem for demand management is, therefore, clear: with sluggish private investment, cutbacks in government spending to reduce the deficit and modest growth from the external sector, how will growth materialize? - 92 - Private Consumption 3.119 In four of the five years since 1983, private consumption growth has lagged behind national income growth. This has been a period of generally tightening fiscal policy, with the major exception of 1987, which has resulted in a decline in the share of private disposable income in GNP. Furthermore, it has been a period of expanding coverage of the EPF. Until 1980, only about 3% of private disposable income was unavailable for consumption, locked in EPF deposit accounts. By 1987, this had risen to 8a. 3.120 To some extent, the failure of private consumption to track CNP growth more closely is related to government's and individuals' attempts to protect against external shocks. As discussed above, government has attempted to smooth private income through countercyclical wage and fiscal policies. The analysis of private behavior, however, suggests that this has been unnec- essary. Even with fluctuations in their income, the private sector tends to smooth out its consumption expenditures; both past and present income are found to be important determinants of 2i7asumption. This accords with the economic concept of permanent income.- Furthermore, transitory income caused by fluctuations in current income from its trend level does not affect consumption. Such transitory income is typically associated with terms of trade changes. Thus, when GNP growth is linked to temporary movements in the terms of trade, consumption does not necessarily follow suit (Figure 3.3). 3.121 Despite the fact that contributions to the EPF and income earned on accumulated EPF assets are not available for consumption directly, the data indicates that some substitution between funds occurs. That is, the buildup of financial wealth in the EPF encourages individuals to reduce their savings out of discretionary income. In this way, total private consumption reacts to the wealth created in the Provident Fund. Another interpretation is that the breakdown of total private income between discretionary and nondiscretionary components reflects a shifting distribution of income. As a higher fraction of income goes to wages, EPF contributions and surplus increase. If the pro- pensity to consume out of wages is higher than out of profits, then there should be a close correlation between the EPF surplus and consumption, as found empirically. 3.122 It is difficult to identify quantitatively the precise impact on consump- tion of income saved in EPF accounts because of the variety of channels through which the influence is transmitted. It is reasonable to suppose, how- ever, that there is some impact on consumption associated with a substitution of income between cash-in-hand and earnings in EPF accounts. A policy that permitted individuals greater control over their own resources would give an important stimulus to consumption. 21/ See S. Bhalla, "The Measurement of Permanent Income and its Application to Savings Behavior," Journal of Political Economy, Vol. 88, August 1980, for a discussion of the relationship between permanent income and observed annual incomes. PERSONAL INCOME AND CONSUMPTION Mnlaysla, 1973-87 80- 3 GNP 70 - O Private Income + Discretionary Income A Permanent Income 60 X Private Consumption 650 ~40- 30- 20 10 7 7 73 74 75 76 77 78 79 80 81 82 83 84 as 86 -X - 94 - 3.123 Two other important determinants of consumption are the real effective exchange rate and the real deposit interest rate. As is typically the case, a real depreciation of the currency negatively affects consumption; the declining value of domestic assets depresses spending. There also appears to be a response to the real deposit interest rate, but the effect is quite small quantitatively. A one percentage point reduction in deposit rates boosts consumption by only 0.41. 3.124 With significant changes in interest and exchange rates having already occurred in 1987, more stability in these variables is expected for 1988. Thus, most of the growth in private consumption will come from real income growth. Given the recent pattern of income growth, a further growth of, say, 61 in 1988 would be expected to support a 6.5X growth in consumption. The long-term relationship between consumption and income, however, suggests that it is unrealistic to expect consumption growth to lead income growth in Malaysia, without measures being taken to reduce the significant forced savings now in place. Government Spending 3.125 Real government spending in 1988 for operating purposes should stabilize close to 1987 levels. This is an inevitable consequence of the need to restore stability in public finances and the need for control over NFPE expen- ditures. This would still not be the end of adjustment. If public investment expands as a result of "catch-up" to the shortfall in spending in 1987, however, the deficit would grow again. A steady reduction in the deficit until 1991 is likely to be necessary, reaching a taraget level of about 81 in 1991. 3.126 The impact of government deficit reduction on growth will iepend on the mechanics of adjustment. Regardless of how the deficit is brought down, however, one implication is that private disposable income will fall as a share of GNP. Because much of the anticipated growth in 1988 will be absorbed by government in the form of higher revenues for PETRONAS and higher taxes, even a 61 increase in nominal GNP will only generate a 2.3% increase in discretionary private income. 3.127 Government's continued reliance on investment cutbacks in response to its financial difficulties is particularly harmful to growth. During 1987 there was substantial underspending of the development budget. Some increase is called for, along with a strategy of holding the line on other expenditures, while waiting for growth to restore balance in public finances. 3.128 Within the permissible deficit, there will have to be important changes in the composition of spending. Because of a likely increase in both total debt and interest rate levels, interest payments on total public debt will rise faster than growth; they could be M$2 billion more in 1991 than in 1987. Maintenance expenditures, particularly on roads, will also have to rise by about M$500 million per year. There may be some loss in revenue collection if oil prices remain at current weak levels. Any gains from a broadening of the sales tax base (or the introduction of a VAT system) could be offset in the short-run by small transitional losses that would occur in any reform of - 95 - the corporate tax structure. In fact, because of these offsetting revenue effects, it would be advisable to link the introduction of a VAT with corpo- rate income tax reform. Overall, however, the potential for major nondistor- tionary revenue increases is small and so fiscal stability depends on Government's ability to hold total expenditure, ot;her than maintenance and interest, below the growth of output. 3.129 Three areas should receive priority in the search for lower ' vernment expenditures: (i) subsidies and transfers; (ii) debt management; and kii) reform of loss-making public enterprises. A detailed budgetary analysis of these is beyond the scope of this report, but order of magnitude calculations suggest that progress in these areas may be sufficient to meet fiscal objec- tives without requiring further cuts in real government spending beyond 1988. 3.130 Subsidies and Transfers. If all subsidies in the budget were direct and transparent, such as those listed under current expenditures in the budget, it would be straightforward to tabulate and track down government subsidies. In the case of Malaysia, however, there are sizable implicit subsidies. These operate, for example, through budgetary transfers to agencies for purposes of equity capitalization or loans which never get repaid and through funding operational expenditures for activities which do not recover costs through user charges. A direct estimation of these indirect subsidies is almost impossible; it would involve estimating the difference between market- determined costs of government-provided services and user prices. Especially in the context of the NEP, government programs in agriculture, housing, trans- portation, credit, health care, education and industry provide some element of subsidy. 3.131 The major explicit government subsidy program up to 1983 was in the area of petroleum products (see Volume II, Chapter 2). With the reform of this program, explicit subsidies have fallen sharply to total just M$500 million in 1987. This decline, however, may mask an increase in implicit subsidies. 3.132 One way of calculating implicit subsidies and transfers is to subtract total government expenditure, recorded in the budget, from government demand for real goods and services as recorded in the national income accounts. While there are some clear inconsistencies in moving between accounts in this fashion, due to differences in coverage and in concept between the cash-based budget figures and the accrual-based national income data, the trend is indicative of what is happening to subsidies. 3.133 Figure 3.4 shows the pattern of total government expenditure and the subsidy and transfer component defined as above. Since 1979, the portion of all government spending--Federal, State, local and enterprises--that has gone towards expenditure on real goods and services has been only about 60%. The other 40Z has been absorbed in implicit subsidies and transfers. Of this, less than one-half is accounted for by net interest payments, pensions and gratuities; the remainder should reflect implicit and explicit subsidies. Controlling these expenditures is the most effective way of achieving sustain- able public finances. In general, the subsidy programs induce distortions in the economy that restrict growth and efficiency. - 96 - Figure 3.4: Total Expenditure, Subsidies and Transfers 40~ 30- 25 I 20- to 19701971 197219731974197S19076177197019,791g01s1 1 gt 1 937OtI F O EUP .S$__ - 97 - 3.134 Four government programs stand out as the largest and least cost- effective recipients of public subsidies: the support of padi farmers; subsidized housing credits; subsidized transport; and subsidized education. 3.135 Total support to padi farmers currently runs around M$200 million per year, approximately equal to half the total income of padi farmers in 1987. In addition, there are implicit subsidies through the absence of full recovery of water charges, fertilizer distribution and loans to cover the operating losses of LPN. These would be even higher except for the cross-subsidization of milling and padi price support programs by taxing imported rice. The overall savings from a rationalization of the rice production, milling,2 distribution and import system could be in the order of M$500 million.-_ 3.136 The Treasury has four types of lending programs for civil service housing. These bear an interest rate of between 4-6% which, when compared with commercial mortgages, contains a subsidy of about 51% on average. If the current annual expansion of M$l billion per year in outstanding loans were to continue, the subsidy equivalent would be M$510 million ver year. While this reflects the expected savings to the Treasury from moving to market-based lending rates, it does not reflect the savings from reducing the size of the loan program. The latter is given by the differential between the rates at which the Treasury borrows and on-lending rates. Under current market condi- tions, this is a much smaller, but still significant, figure of M$100-200 million. 3.137 Reform of the Housing Loan Program is appropriate because with the development of a secondary mortgage marke- and increasing liquidity in the banking system, the rationale for the program seems less evident now than at its inception. Raising interest rates on housing loans has proven to be difficult, however, as civil servants perceive such loans as part of their overall compensation package. The preferred option would be to sell the existing portfolio of loans to a housing finance company such aa MBSB, and to use the cash proceeds, which would probably only cover some 50% of the face value of the loans, to retire existing debt and to compensate civil servants who have not yet ta2r out loans. Following this, the program could be abolished entirely.- 3.138 The transport sector is another major recipient of government funds. Malayan Railways (KTM) has consistently run on an operating loss and its overall net income loss, after accounting for interest expenses, reached M$77 million in 1986. Although the government has not directly subsidize KTM's operations, the railway is totally dependent on Government for its survival. Government has provided loans totalling M$514 million between 1981-85, has reimbursed KTM for 50% of its losses on uneconomic passenger services, has exempted it from all sales taxes and duties since 1982, and has released it 22/ The World Bank and EPU are currently studying this issue; a report is expected shcctly. 23/ See "Malaysia: Housing Sector Report," World Bank 1988. - 98 - from its pension liabilities. One striking feature of KTh's performance is that despite its sustained losses, it has increased its fixed assets in service by 502 since 1981. Government should act quickly to reduce KTM's annual cash drain by (i) reform and decentralization of procedures to set tariffs; (ii) reform of employment policies and absorption of railway workers into other government departments; (iii) reappraisal of uneconomic passenger service, particularly where alternative modes of transportation exist; and (iv) modification of the balance sheet, includi' write-down of debt, as a precursor to corporatization and privatization.- 3.139 By far the largest government subsidy program is in the education sector. Total expenditures on education amount to over $M4 billion. These include teacher salaries, support to universities, scholarships for overseas students and public training institutions, mostly of a technical nature. The riajor potential areas for savings are in the latter three areas. 3.140 University education in Malaysia is heavily subsidized. Cost recovery through user-charges, which has recently been introduced, only suffices for 2.5-10Z of the marginal operational costs. Furthermore, there is a wide disparity between the cost of alternative programs which is not fully reflected in differentials in user charges. Although a student loan program is now in operation, its collection potential has yet to be tested and ample opportunity exists for conversion to a grant. 3.141 The financing structure in university education has had several adverse consequences. Lack of funds has constrained potential expansion. Some students go abroad on scholarship or private funds, but many are unable to participate in further education and enter the job market. This has created a bulge in the supply of 15-19 year olds in the labor force and, consequently, the highest incidence of unemployment is in this group. The excess demand for university education can be measured by the degree of rationing of entry applications--only one in three applicants are successful nation-wide. 3.142 Increased use of user charges in universities, coupled with an enhanced student loan program to defray all cash outlays, would achieve several desir- able objectives, without jeopardizing government support for the principle of affordable and universally-accessible education. Particularly if loans were moved off-budget to a separate agency (see Volume II, Chapter 3), there would be a positive cash-flow benefit to the Treasury. 3.143 Similar problems exist in government training programs, especially vocational training. There, only one out of ten applicants is enrolled. Despite this, there is a high drop-out rate, with some 60X of the student intake failing to graduate. Even for those that do complete the courses, there is little difference in their subsequent ability to obtain jobs relative to those not trained. Only 5% of trainees are thought to obtain jobs using skills developed in the courses. To assist in student placement, the Ministry of Labor now offers firms an additional inducement--payment of wages for one 24/ See "Malaysia: National Transport Policy Review," World Bank 1988. - 99 - year for on-the-job training in firms. There is a substantial element of wastage in these programs of both budgetary and manpower resources. Closer linkages with the private sector and industry are desirable, going well beyond the current advisory and consultative role played by the private sector in the Manpower Development Board and the National Industrial Training and Trade Certification Board. Greater industry participation, sensitivity to labor market conditions and accountability of course coordinators must be developed. 3.144 Debt Management. Interest payments on public debt are likely to total 11.41 of GNP by 1991, or about one-third of public expenditure. Sizeable amounts of principal on domestic currency debt are also starting to fall due and will grow exponentially in the medium-term. The public sector can hope to generate considerable savings on its debt service obligations through more active debt management. 3.145 Active debt management is not a new concept for Malaysia. The country aggressively and successfully refinaned much of its high interest external debt in 1983/84, reducing interest and smoothing principal repayments. It has recently embarked on a program to prepay M$5 billion of foreign loans in 1987 and 1988. Such a program is attractive because of the relatively low levels of domestic interest rates, particularly for short-term bills, and the favor- able foreign exchange reserves position. The Government has announced expected savings from this program to total M$528 million, which implies an average interest rate reduction of 2.5Z. This is unlikely to represent true savings, however, because if the ringgit continues to appreciate against the US dollar, as it has of late, then currency movements would anyway lead to reduced debt servicing in ringgit terms. Given the size of the debt, existing programs should be thought of as only a first step in a mor! 5omprehensive liability management system that Government should develop.- 3.146 The volatility in world interest and exchange rates is directly reflected in volatile debt service payments and obligations. Recently, for example, despite a prepayment of US$1 billion, the external debt of Malaysia continued to rise marginally because of the appreciating value of yen-denominated debt. Such problems beset many countries and multinational corporations. As a result, there has been an explosive development in financial mechanisms to alleviate risk-eurodollar futures, options on eurodollar futurs, forward rate agreements, and interest rata caps are amongst the new instruments used for short-term hedging, while futures and options on futures for Treasury notes and bonds are used for medium- to long-term hedges. These are useful to lock in a given interest rate, or to reduce the uncertainty associated with the interest rate to be paid on future, prospective borrowings. 3.147 Beyond these well-developed instruments, there are additional, more fledgling markets, such as those in commodity bonds, that could prove useful for Malaysia, particularly given the reliance of government revenue on petro- 25/ The World Bank is prepared to assist Malaysia and other developing countries in developing the necessary expertise for active liability management. - 100 - leum prices. Commodity bonds link either or both interest and principal payments to a physical quantity of the commodity. For example, if the govern- ment had issued petroleum-based commodity bonds, its debt service obligations would have declined along with the price of oil, relieving the pressure on the budget to some extent. Commodity bonds have been used by the French and Mexican governments, antd by a variety of corporations. In the 12 months ending October 1987, there were some 45 commodity bond issues, two of which involved public entities. Most were liLk7d to gold or silver, but others included links to oil, copper and zinc.-2 3.148 The scope for commodity bonds in Malaysia is particularly great because of the country's dependence on primary commodities. Partly, such bonds can be used internationally, to hedge against commodity price movements. At present, the scope here is limited because of the thinness of the international market and the lack of liquidity in commodity bonds. Of greater potential is the development of a domestic commodity bond market. Long-term institutional investors, such as EPF, may be willing to hold such bonds because of the potential for greater returns. In this way, the fortunes of commodity prices would be directly and broadly transmitted to individuals throughout the country in an efficient manner. By contrast, individuals are now affected by commodity prices, for example for oil, through indirect channels such as cut- backs in government expenditure and declining employment and job prospects. This "real" adjustment is much less efficient than the proposed financial adjustment. 3.149 Reform of Loss-making Public Enterprises. In 1986, over half of all public corporations on which data exist posted losses in their net profits account. There were 311 such firms operating in every segment of the economy. In addition, a half-dozen companies, including the railways (KTM), the Urban Development Authority, and Sabah electricity, that are not corpora- tized are known to have made losses. The cumulative total of these losses equalled about M$2 billion. This figure is larger than the sum of all the profit-making public enterprises excluding Petronas and its two largest subsidiaries. 3.150 The net loss figures quoted above are an underestimate of the true cost of operations of these companies. Government has sunk some M$33.9 billion in total assets into the loss-making companies, most of which must realistically be considered as a write-off. The loss-making firms carry only a slight interest burden, (M$600 million--they have total outstanding loans of only M$9.8 billion, low in comparison to the size of assets), and pay almost no taxes (M$60 million) or dividends. In several instances, including the important examples of steel and cement, financial losses have been minimized by setting regulated prices above world prices for these goods. In other cases, such as railways, important rental income is earned from non-operating profits. In all but thirty-nine cases, companies with negative net profits also had negative operating profits. In addition to ftunding these losses, 26/ See "Commodity Bonds: A Risk Management Instrument for Developing Countries," T. Privolos, World Bank (processed), November 1987. - 101 - Government or parent NFPms must also provide financing for additional capital expenditure. 3.151 The implication of these statistics is striking. First, remedies that rely solely upon financial retructuring and tax breaks are unlikely to provide a cure. The greater problem lies in efficiency of operation as a result of some combination of poor management, excessive employment, poor technology or inappropriate location. Privatization would only be suitable if losses stem from the first two causes. Otherwise, liquidation must be considered. 3.152 Second, despite the recent Government initiative to set up a unit in the Treasury to study alternatives for dealing with the ten largest loss-makers, there remains a sizeable probi v that is essentially excluded from public sector scrutiny and controls._, These latter cost the Treasury some M$800 million in 1986. Thus, while Government efforts are correctly focused on the largest firms, it cannot afford to set aside the problems of smaller firms. Introducing incentives for privatization and liquidation as part of a perfor- mance evaluation system should be considered as a mechanism for dealing with small loss-making firms. 3.153 To a significant degree, the Treasury must continue to bear losses as a result of interest payments, labor pensions and other commitments, even if a firm is liquidated. A lower bound on the savings to the exchequer is the total operating loss of companies that are candidates for closure. This totalled M$420 million in 1986. To the extent that (a) some revenues may be realized from the sale of assets; (b) some non-operating expenditures would also be reduced on closure; and (c) some of the interest burden would be shared by other equity holders, the savings could substantially exceed this figure. Off-budget Government Programs 3.154 In addition to its budgetary operations, government influences the economy through a series of off-budget programs. The most important of these are credit programs and the Special Low Cost Housing Program. But a variety of other activities, such as setting of smallholder agricultural prices, are also important. The tendency has been to use these programs in a counter- cyclical fashion. There is some danger, however, that implementation of these programs generates distortions and amplifies business cycles in an unwarranted fashion. 3.155 The Special Low Cost Housing Program (SLCHP). The SLCHP proposed that 240,000 low cost housing units be constructed by private developers in three years, starting in mid-1986. There are no explicit budgetary provisions or formal mechanisms for achieving the target, but considerable incentives are 27/ Amongst the largest loss-makers are Perwaja Steel, Perusahaan Otomobil Nasional, Kedah Cement, Sabah Gas, Sabah Forests, Sabah Shipyard, Malaysia Mining, Malaysia Shipyard and Engineering, UMW Corporation, Malayawata Steel and Mamut Copper Mine. - 102 - given in the form of revised building and planning standards, provision of low cost land by State governments and liberal rezoning permissions. Because of implementation lags and the sequencing of land acquisition, land development and construction, the number of completed units and investment in the first year of the program has been quite small--32,000 units complete1 by mid-1987 for an estimated investment of under M$600 million. Many units, however, in the intermediate stages of completion will be finished during the current year, so the pace of implementation is expected to pick up substantially. An additional M$1 bilion over the first year investment level should be forth- coming, giving a sizeable boost to the economy in 1988. 3.156 Despite its merits in terms of the social benefits from low-cost housing and the creation of a housing supply system that is more responsive to demand, the SLCHP has run afoul of a typical macroeconomic problem besetting public countercyclical programs; the need for a macroeconomic stimulus may have long passed by the time the program comes fully on stream. The effects of a major stimulus in 1988 are likely to be mixed as recovery is already well in hand, primarily because of commodity price increases and government's fiscal expan- sion in 1987. If investment is already recovering, then instead of adding incremental investment, the SLCHP may substitute for investment in other areas and may postpone adjustment in the labor market. It would be advisable to extend the life of the program so as to phase investments towards the next few years when external stimuli may be absent or negative. Overall Impact 3.157 The overall impact of the measures decribed above would be to place public finances on a more solid and sustainable footing. The savings identi- fied could amount to M$1-2 billion, with almost no reduction in real govern- ment consumption or investment. More importantly, however, the changes would signify a completion of government's fiscal adjustment, which has started with the phasing down of spending on goods and services and should now continue into the more sensitive area of grants and subuidies. The proposed measures would instill a greater degree of financial accountability at a decentralized, program level, and make for a more effective delivery of government services. - 103 - IV. ECONOMIC OUTLOOK AND INTEGRATED POLICY RECOMMENDATIONS A. Economic Outlook 4.1 A growth scenario for the future and the corresponding investment, savings and balance of payments developments is projected below on the basis of supportive policy reform, as outlined above. In the short-run, the most critical variable to sustain recovery is private investment. In the medium- term, private consumption will play a more significant role. The emphasis on the private sector as the engine of demand growth arises because of the limitations on government imposed by current financial constraints and because net ezports should not be expected to improve beyond the record levels posted in 1987. The External Environment 4.2 Having accounted for 72X of total demand growth in the 1980s, the evolution of the external environment is obviously critical for Malaysia. The major structural adjustments underway in the developed world and in parts of the devcloping world have consequences for trade, commodity price, exchange rate and capital flow behavior that will seriously affect Malaysian prospects for industrial growth. 4.3 The Global Setting. In its most recent World Development Report, 1988, the World Bank forecasts growth in the developed countries of 2.31 over the next eight years, inflation of 41 and trade growth of 42. The composition of growth, however, is expected to adjust-the recent historical pattern of export-led growth in Europe and Japan coupled with consumption-led growth in the USA will likely reverse itself. Japan, and Western Europe to a lesser extent, will become more internally oriented while the USA will become more outwardly-oriented. Given the fact that the US market is the most open in the developed world, the expected growth of manufactured exports from developing countries is likely to slow down to 5.81. 4.4 Sluggish growth in industrialized countries also implies that commodity prices will remain weak over the short-term, failing to keep pace with general inflation. Added to short-term pressures on petroleum prices, this suggests that Malaysia's overall terms of trade may deteriorate slightly in 1988 before steadily improving on the strength of palm oil, tin, rubber and petroleum prices. In volume terms, petroleum exports will grow substantially in 1988, but thereafter palm oil, cocoa and LNG are the only commodities with moderate growth prospects in the 3-4Z range. Environmental pressures and forest depletion will limit timber exports, particularly of logs; and slow demand for rubber, tin and petroleum is expected to limit future growth of these commodities. 4.5 Because Malaysia's manufactured exports are so concentrated in electronics and textiles, their prospects may well differ from the overall trend in developing country manufactured exports. The short-term prospects in both these industries are excellent, but medium-term trends portend rapid changes in the structure of trade and nature of comparative advantage of LDC exporters like Malaysia. - 104 - 4.6 Electronics. Malaysia's electronics industry developed very rapidly from virtually nothing in the early 1970s to the point where it accounts for over half of manufactured exports. The output structure is dominated by components (80-85Z of output) and within that subsector, semi-conductors have predominated accounting for 80-90% of component output. Malaysia has not developed a sizeable electronic equipment industry, either for consumer or industrial electronics products. The industry's growth has been inextricably linked to investments by multinational semiconductor firms which cam to Malaysia in search of cheap labor for the assembly of semiconductors. 4.7 Trends in the industry worldwide since the early 1980s raise doubts whether this growth pattern can continue for the medium term. Of critical importance is the continuing automation in semiconductor assembly. There is evidence that an increasing share of semiconductor assembly is shifting back towards the developed countries. The trend is most striking in Dynamic Random Access Memory chips (DRAMs), the largest single segment of the semiconductor market. Also acting against developing country procedures is the continuing shift toward more complex, application specific integrated circuits within the industry's prodtuct mix. These lower volume, high-cost products depend on close coordination between consumers and producers. Finally, there is a trend toward use of surface mount components, which involve high density packaging and assembly technology, in a wide variety of electronic products. necessi- tating expensive automated equipment. These worldwide trends suggest that Malaysia's competitive advantage based on low labor costs will be increasingly less important. As the experience of the Philippines shows,- loss of competitiveness can quickly hurt the industry. 4.8 In the short-run, however, growth in traditional semiconductors should continue to be rapid. These components have benafitted from the ever- increasing scope of applications, especially into consumer durables. Malaysia should continue to benefit from this given two advantages relative to competi- tors such as the Philippines: political stability and the existence of semiconductor testing facilities that ensure high quality. Nevertheless, there will be stiff competition from other countries as currently, there is sizeable global overcapacity in production of 64 K and 256 K semiconductors. 4.9 Global trends do not spell the end for the Malaysian electronics industry but they do suggest that a new strategy is required in order for the phenomenal growth in exports of semi-conductors to continue over the medium- term. The first element of such a strategy must be to induce foreign companies to make the required investments to produce new products at the necessary qualitv levels. The second element should be to exploit the advantages in quality that Malaysia possesses over lower-cost producers such as the Philippines, due to its large semi-conductor testing facilities. The third step would be to take greater advantage of proximity to Singapore to develop niche products that are closely tailored to the requirement of the I/ The Philippines experienced rapid growth in electronics exports in the early 1980s, which came to an abrupt halt when political instability and an appreciating real exchange rate caused a pull-out by multinationals. - 105 - Singapore market. Finally, the marketing strategy should foster closer integration with producers of engineering goods, automobiles and consumer durables at home and abroad. 4.10 The advantage of developing a "niche" strategy is that it can be implemented relatively quickly, at low cost, and with full focus of government and industry resources to ensure success. In the rapidly changing global technological and marketing environment, this has the added attraction of not requiring specific long-range planning and long gestation investments which would be difficult to modify even if future conditions warrant. Overall, it appears that the factors underlying the successful semiconductor export drive from Malaysia over the past decade are rapidly disappearing. In their place are new opportunities for growth, but these require building on strategic advantages, inducing a new round of major investments in plant and equipment, and exploring direct marketing involvements with new consumerp. Success in this new environment is uncertain; what is clear is that simple provision of an undistorted location in Free Trade Zones and an abundant supply of relatively skilled but cheap labor will no longer be a sufficient guarantor of exporting success. 4.11 Textiles. The textile industry is also subject to major changes in the structure of world trade. The industry currently employs about 142 of the manufacturing workforce and accounts for 11% of manufactured exports. Malaysia is a relatively small textile producer by world standards, but has been highly successful in the garment sector, essentially commissioning work for foreign clients on the strength of its well qualified and low cost labor force. The country has a competitive supply of polyester fibre, but has to import virtually all its cotton and has a surprisingly small spinning capa- city, which shows considerable age (three-quarters of the machines are more than 10 years old). Weaving capacity is also relatively small and almost 90Z of it is older than 10 years. 4.12 Whereas the textile industry has catered to the domestic market under relatively high effective protection, the garment industry has developed predominantly as an enclave industry for exports, where local labor is added to imported inputs. The garments sector has continuously modernized and expanded its share in the lucrative markets of the USA and the EEC. Now, however, the industry is increasingly subject to pressures from several directions. Overall world consumption of fibres is projected to grow considerably slower in the future (1.62) than it has in the past (4.62 p.a. since 1950). Furthermore, quotas under the Multifibre Agreements (MFA) have been steadily tightened. For the products included under the MFA, Malaysia falls within the category of countries for which growth is limited to 62. Because of quota restrictions, there has been growing surplus capacity in a number of producing countries, which is resulting in increasing competition for the more limited non-quota market. In addition, recent and ongoing investments into highly efficient and largely automated equipment, particularly in developed countries and NICs, are likely to cause an even stronger shift of supply of high quality, high fashion textiles and garments to those countries. Textile labor costs have also risen quickly in Malaysia and are now well above those in some of its principal competitors--China, Indonesia, Thailand and Pakistan. - 106 - 4.13 In this increasingly competitive market, the ability to survive and expand seems closely related to the type of equipment in use and the efficient management of a highly automated process. Such an operation requires tech- nical and managerial skills of a high order, a dependable labor force and a reputation for quality. Garment production in the upper and middle price brackets demanao excellent design, the ability to produce in small lots and a well developed marketing system built around respected brand names and with developed channels for dealing with numerous retailers in importing countries. Increasingly, fashion models can be designed with the help of personal computers; cutting, which already is semi-automated on a commercial scale, can now be fully automated and computer-guided. Robots for sewing and making-up are still in an infant stage, but they are also expected to become commercially competitive during the 1990O. The ease and speed with which pattern changes can be introduced under this technology is a critical competi- tive advar.tage that may cause a gradual return of garment manufacturing into industrialized countries, particularly for mass-produced high fashion items with short life-cycles. 4.14 One additional modernization trend is in the marketing of textiles and garments. Electronic media and telecommunication systems will permit a much closer link between manufacturers of textiles and apparel and clients in wholesaling and retailing, with greater attention paid to high inventory turn- over and lower carrying costs. This will further detract from the advantage of cheap lyor that is still the basis for the garment industry in developing countries._ 4.15 While the long-term prospects for textiles and garment exports from developing countries are moderate at best, Malaysia has some important medium- term advantages. Foremost is the shift in quotas in the US from an item-by- item approach to a comprehensive agreement that permits a 6Z annual volume growth in exports. This permits greater flexibility in filling quotas and encourages a shift to higher value-added products. In addition, stricter quota restrictions on Malaysia's major competition, including Japan, Korea and Hong Kong, limiting their growth to just 12 per year, should help divert investment from these countries to Malaysia. The new quota system has also assisted the US industry to reach full capacity and this, combined with the restrictions on imports, is liable to generate upward pressure on prices that would benefit small exporters. 4.16 Protectionism. Protectionist legislation provides the most severe threat to Malaysia's manufactured export potential. Legislation pending in the US Congress would seek to cut textile and garment quotas to 1% growth per year across-the-board. Recent atLempts to relabel palm oil are also a reflection of this trend. Despite these potential threats to market access for Malaysian products, however, the country is in excellent shape to turn protectionism into a benefit rather than a cost. 2/ See Kurt Hoffman, "Clothing, Chips and Competitive Advantage: The Impact of Microelectronics on Trade and Producticn in the Government Industry," World Development, March 1985. - 107 - 4.17 This somewhat paradoxical conclusion emerges clearly if protectionism is viewed from a political economy perspective. There is a growing tendency to use trade policy for bilateral ends, demonstrated by the favorable treatment afforded Caribbean and African countries, and the LLDCs. The other side of the coin is the trend towards retaliation against selected countries on a bilateral basis. The four Asian NICs are all being pressured in this regard. 4.18 Garment and automobile exports are examples of two sectors where Malaysia has benefited from global protectionism. In the textile and garments sector, it is probable that without a quota system, or if MFA quotas are "globalized" in the next round of trade negotiations, the low-cost producers in China and India would sweep away the rest of the developing country competition in the future, leaving behind only small high-fashion niches for others to exploit. Similarly, the presence of voluntary export restraints on Japanese automobile exports to the United States, apd the danger posed to Korea of a too successful market penetration, makes it more probable that the Proton Saga could hope to gain significant sales. 4.19 Malaysia is in an excellent position to ward off protectionist threats against its products, even as other countries become more exposed. First, by virtue of its trade structure, Malaysia has a comparatively small bilateral surplus with the UP\. By contrast, other East Asian economies have seen their surpluses grow to unprecedented heights (Table 4.1). This trend is continuing despite US efforts at adjustment. While the overall trade deficit in America is declining in real terms, these gains have been made mostly at the expense of Western Europe rather than the NICs. 4.20 A second favorable factor in Malaysia is the openness of domestic markets. The high share of imports in GDP and low tariff barriers on imports gives the country an edge in seeking favorable treatment for new quotas in the future. A third factor is the importance of semiconductor exports which are largely free from trade restrictions. Finally, passage of the Copyright Act in 1987 and openness in the financial, insurance and other service sectors in Malaysia should also keep the country in good standing in its trade relations with the USA. - 108 - Table 4. 1: EXTERNAL BALANCES OF EAST ASIAN COUNTRIES (US$ billion) 1983 1984 1985 1986 1987 Malayfia Current account balance -3.5 -1.7 -0.7 -0.3 2.3 Trade balance wit' US -0.3 -0.1 0.1 0.3 n.a. Korea Current account btiance -1.6 -1.4 -0.9 4.6 9.8 Trade balance with US 2.0 3.6 4.3 7.3 9.0 Taiwan, China Current account balance 4.4 7.0 9.2 16.2 20.0 Trade balance with US 6.7 9.8 10.0 13.6 16.0 Hong Kong Current account balance -0.6 1.4 1.9 1.5 n.a. Trade balance with US 4.4 6.3 6.5 8.1 n.a. Singapore Current account balance -1.0 -0.7 -0.0 0.5 0.2 Trade balance with US -0.3 0.6 0.8 1.4 n.a. Thailand Current account balance -2.9 -2.1 -1.5 0.1 -0.2 Trade balance with US -0.3 -0.1 0.4 0.3 n.a. Source: Directory of Trade Statistics, IMF, various issues. 4.21 Malaysia will not be able to escape unscathed from protectionism. The shifting structure of trade will bring it face-to-face with markets in the EEC and Japan that rely much more heavily than the United States on non-tariff barriers (NTBs) (Table 4.2). Still, in comparison with other East Asian countries, and LDCs in general, Malaysia faces less protectionism in these markets. To see the sectors where a shift in markets is likely to be diffi- cult, Table 4.3 presents the percent of imports of selected major commodities that is subject to NTBs in the major importing countries. The table shows that traditional exports such as garments and textiles are heavily regulated in all major importing regions. Other products, however, such as shellfish, prepared fish and rubber, which are free from NTBs in the US, are faced with significant barriers in other markets. Imports of products such as wood, lumber and cork in Japan and footwear in the EEC and also largely governed by NTBs. Thus, diversification of exports in these areas will not be easy. Table 4.2: NONTARIFF BARRIERS AGAINST IMPORTS FROM SELECTED ASEAN COUNTRIES, 1984 (US$ million) Imports subject to nontariff barriers Share of Fiscal Volume con- Import Control of Technical total measures trol measures authorization price level barriers All Total imports Amount % Amount Z Amount 2 Amount % Amount % NTBs Imports (2) lmports of .1apan from: Malaysia 23.1 1.6 43.1 3.0 16.3 1.2 0.0 0.0 331.8 23.5 371.1 1,414.0 26.2 Phtlippines 261.7 18.7 69.6 5.0 72.5 5.2 0.0 0.0 686.7 49.1 725.9 1,397.7 51.9 Thailand 74.7 7.3 123.5 12.0 195.7 19.0 0.0 0.0 804.2 78.2 833.0 1,028.7 81.0 Indonesia 655.97 5.9 47.0 0.4 245.7 2.2 0.0 0.0 4,329.9 38.9 5,042.9 11,123.7 45.3 All LDCs 1,707.9 2.2 2,095.0 2.7 3,727.4 4.9 0.0 0.0 17,941.7 25.5 20,534.5 76,205.1 26.9 Imports of _ EEC from: Malaysia 445.6 20.8 63.4 3.0 138.9 6.5 8.' 0.4 0.02 0.0 827.6 2,139.8 38.7 Philippines 278.9 28.2 49.7 5.0 93.8 9.5 22.4 2.3 0.0 0.0 610.2 990.0 61.6 Thailand 723.0 49.2 608.2 41.3 805.1 54.7 65.2 4.4 0.26 0.0 1,112.7 1,470.9 75.6 Indonesia 299.5 24.9 68.26 5.7 134.8 11.2 4.8 0.4 0.20 0.0 584.4 1,203.7 48.5 All LDCs 5,640.6 5.0 16,398.2 14.5 11,317.2 10.0 3,209.6 2.8 249.6 0.2 39,510.8 113,214.8 34.9 Imports of USA from: Malaysia 1.90 0.1 0.02 0.0 0.06 0.0 153.9 5.7 0.0 0.0 156.2 2,682.1 5.8 Philippines 169.4 7.0 130.6 5.4 0.02 0.0 27.2 1.1 0.0 0.0 255.0 2,417.9 10.5 Thalland 114.4 8.7 20.9 1.6 0.01 0.0 250.0 19.0 0.0 0.0 370.6 1,316.6 28.1 Tndonesia 7.0 0.1 0.0 0.0 0.0 0.0 200.4 3.7 0.0 0.0 219.5 5,381.2 4.1 All LT)Cs 5,594.9 46.0 3,658.2 3.0 251.7 0.2 4,840.1 4.0 319.8 0.3 10,037.3 122,356.8 8.2 Souirce: IJNCTAD, Data Base on Trade Measures. - 110 - Table 4.3: NON-TARIFF BARRIER PROTECTIONISM IN SELECTED COMMODITIES /a Wood, Prepared lumber, Foot- Garments Textiles Shellfish fish Rubber cork wear Importer Japan 82.4 71.0 100.0 100.0 100.0 31.2 0.0 EEC 62.8 100.0 100.0 99.8 n.a. 8.7 23.7 USA 90.3 99.7 0.0 19.1 0.0 0.0 0.0 /a Percent of exports from Malaysia to major markets covered by non-tariff barriers in 1984. Source: UNCTAD data base on trade measures. 4.22 Competitor Behavior. Malaysia's manufactured export performance depends heavily on its competitiveness relative to neighboring East Asian countries. As noted above, the competitive position deteriorated substan- tially in the mid-1980s but has been improving since 1985. A major impetus will come from additional pressures on the unit labor costs of the successful NICe. Large current account surpluses in those countries are being trans- formed into foreign exchange reserve accumulation and pressures to appreciate the currency. This has already occurred to some extent in 1987 in Taiwan and Koreat and is likely to accelerate in 1988. In addition, wage increases in local currency terms in those countries are starting to outstrip productivity gains. Dollar-denominated unit labor costs in Korea and Taiwan could rise by 20-30Z in 1988. Similar pressures are being felt in Hong Kong and Singapore, although these have, to date, been resisted by the authorities. 4.23 The NICs will also come under increasing pressure to adjust because of their large trade surpluses with the United States. In addition to exchange rate appreciation, these countries face increasing threats of non- tariff barriers including voluntary export restraints, anti-dumping charges, restrictive quotas and abolition of the generalized system of preferences (CSP), not to mention even more discriminatory trade bills pending in the US Congress. 4.24 Malaysia stands to gain from these external developments both in terms of its ability to compete directly with the Asian NICs and through the foreign investment inflows it could capture in the future. The NICs themselves have become major exporters of capital anu will look increasingly favorably on a location providing relative safety against protectionism, low- cost labor and few distortions affecting export-oriented industries. If Malaysia continues to qualify in this regard, the recent growth in foreign direct investment could accelerate. - lII - Prospects for Exports and the Balance of Payments 4.25 Taking into account the international setting, aggregate exports in 1988 are forecast to grow by 9.11 in real terms and by 15.4% in nominal terms (Table 4.4). Export revenue growth would be led by petroleum (6.6%), manufactures (28.6%), and palm oil (22.91). In the latter case, however, most of the growth is anticipated to come from price increases. After 1988, therefore, prospects for export growth are less favorable. Petroleum exports are likely to plateau at the higher 1988 level; rubber, tin and sawlog volume growth is also expected to be flat. The sole remaining areas of dynamism will be in palm oil, manufactures and other, minor exports. Overall, 1988-91, nominal export growth would decline to 8.3%, and maintain this level approximately thereafter. 4.26 The 1987 trade surplus of US$5.8 billion is projected to slowly decline in the medium term. Imports would grow faster than exports as intermediate inputs for manufacturing pick up and as the structure of domestic demand shifts towards private consumption and investment (see paras. 4.30-4.40 below). Nevertheless, because exports start from a 1987 base that is 50Z greater than the import level, the merchandise goods balance is projected to deteriorate systematically very slowly over the next decade. 4.27 The services accounts, on the other hand, are projected to continue to deteriorate. Non-factor service payments, principally related to insurance and freight of Malaysia's expanding trade, are projected to grow in real terms faster than overall GDP. reflecting the growing overall openness of the economy. Meanwhile, non-factor service receipts are projected to expand at a modest 4% per year. Should Malaysia initiate a drive to develop tourism or to increase the capacity of its shipping fleet, better progress on non-factor service receipts could be expected. In addition, net factor income payments are aLso anticipated to rise as growing profits in the foreign corporate sector are repatriated. Interest payments, however, are likely to decline in the medium-term, as a result of Government's prepayment program. They would only rise thereafter if a conservative foreign exchange reserve management policy was followed. That is, if reserves are maintained at over six months of merchandise imports, additional foreign borrowing would become necessary to achieve balance in external payments after 1990. This is the scenario indicated in Table 4.4. On a net basis, however, interest payments would not increase as earnings on reserves would also rise. 4.28 The projected scenario shows a considerable long-term improvement in Malaysia's external debt position. Total medium- and long-term debt would fall from 74% of GDP in 1987 to 72% by 1997. The overall debt service ratio, temporarily raised by the prepayment program, would fall from 18.5% to 9%, of which only 3.2% represents interest outflows. These indicators all point to the continued strong creditworthiness position held by Malaysia, which should enable the country to obtain the finest of terms on bank credits and to move into new, more suitable and diversified, financial markets for its development requirements. - 112 - Table 4.4: MALAYSIA - BALANCE OF PAYMENTS (US$ millions at current prices) Revised Projections 1987 1988 -1989 1990 1991 1992 Exports of goods & NFS 20,154 23,262 25,456 27,011 29,551 32,376 Merchandise (FOB) 17,711 20,603 22,600 24,106 26,505 29,183 Nonfactor services 2,443 2,659 2,856 2,905 3,046 3,193 Imports of goods & NFS 15,830 19,166 22,310 24,335 27,017 29,967 Merchandise (FOB) 11,823 14,664 17,283 19,096 21,387 23,909 Nonfactor services 4,007 4,502 5,027 5,238 5,630 6,058 Resource balance 4,324 4,096 3,146 2,677 2,534 2,409 Net factor income -2,043 -2,140 -2,060 -2,049 -2,077 -2,054 Factor receipts 715 803 812 808 880 976 Factor payments 2,758 2,943 2,872 2,856 2,957 3,030 (interest payments) 1,428 1,466 1,391 1,329 1,336 1,307 Net current transfers 143 139 135 137 136 136 Current account balance 2,424 2,094 1,221 765 593 491 Long-term capital inflow -420 -1,636 -735 733 1,234 1,470 Direct investment 576 618 647 632 638 643 Net LT loans -996 -2,253 -1,382 101 596 827 Total other items (net) -854 -459 -486 -486 -486 -487 Net short term capital -955 -56 -83 -83 -83 -83 Capital flows N.E.I. -10 -403 -403 -403 -403 -403 Errors and omissions 101 0 0 0 0 0 Changes in net reserves -1,149 0 0 -1,012 -1,341 -1,475 Source: Staff estimates. - 113 - The Domestic Environment 4.29 Despite the rapid growth in exports that is forecast, especially of manufactured goods, it is domestic demand that will have to absorb all of Malaysia's incremental output over the medium-term (Table 4.5). Net of imports, the trade sector will surely impart a negative component to total demand growth. This follows from the fact that Malaysia cannot expect to maintain or increase its trade surplus beyond the record levels posted in 1987. If it were to try, it would face the same protectionist pressures that currently beset other East Asian manufactured exporters, and the same pressures towards an appreciation of the ringgit. Nor could this be absorbed through a build-up of foreign exchange revenues as these already stand at substantial levels. 4.30 This logic indicates that Malaysia must itself make a significant adjustment as a result of the change in its external accounts. Furthermore, the troublesome overall fiscal deficit, combined with high public debt levels, imply that the public sector will be ill-advised to take the lead role in driving domestic demand growth upwards. Instead, it is the domestic private sector that must sustain the recovery. The base case scenario presented below envisages 6% growth for 1988 continuing at this level over the medium-term. 4.31 Private Investment. The economic fundamentals driving private investment are extremely favorable. During 1988, private investment should benefit from a cyclical upswing as well as from positive medium-term develop- ments. These should combine to generate a growth that could reach 15-20% over 1987. 4.32 A strong cyclical effect should be felt in 1988. This stems from the lag between the decision by firms to invest and the actual implementation of a project. A foretaste of this effect is given by the Business Expecta- tions Survey that indicetes a major surge in corporate managers' plans to undertake new capital expenditures. The strength of the 1987 recovery, juxta- posed against the weakness of growth in 1986, suggests that the impetus to investment imparted by the business cycle may be substantial--in the order of M$l billion. An equivalent amount of stimulus will also be derived from the Special Low Cost Housing Program. 4.33 The medium-term fundamentals also bode well for investment. Both real interest rates and the real exchange rate are at favorable levels. After several years of slackness, oil and gas investment is likely to rise by perhaps M$500 million. Foreign direct investment should also contribute an additional M$100 million. The real key, however, lies in the boom in profits that appears imminent given the trend towards moderate growth in national income and the increasing share of profits in this income. 4.34 Based on the recent dynamics of the labor market, average real wages in 1988 are likely to decline, perhaps by as much as 3%. This shift reflects the interaction of two trends. On the one hand, wage growth for existing workers and new entrants will continue to be negligible, or only slightly positive, due to the pressure of unemployment. On the other hand, an increas- ing fraction of the labor force will be comprised of workers whose salary and - 114 - Table 4.5: NATIONAL ACCOUNTS IN CONSTANT 1978 PRICES (M$ million) Prelim. Projections 1987 1988 1989 1990 1991 1992 Gross domestic product m.p. 61,008 64,975 68,907 73,003 77,406 82,143 Import duties less imported bank service charges -525 -462 -490 -519 -550 -584 Agriculture 13,323 13,629 14,038 14,459 14,893 15,340 Industry 23,420 25,976 28,365 30,720 33,303 36,139 (of which Manufacturing) 13,323 15,567 17,435 19,353 21,481 23,844 Services 24,790 25,831 26,994 28,343 29,760 31,248 Resource balance -10,630 -10,344 -9,156 -7,904 -7,042 -6,211 Exports of GNFS 41,706 45,487 48,269 51,199 54,403 57,957 Imports of GNFS 31,076 35,142 39,113 43,296 47,361 51,746 Total Expenditures (A-B) 50,378 54,630 59,750 65,100 70,364 75,933 Total Consumption 36,392 38,961 42,330 45,969 49,585 53,443 Private consumption 26,666 29,131 32,302 35,641 38,843 42,164 General government 9,676 9,831 10,027 10,328 10,741 11,278 Gross domestic investment 13,986 15,669 17,420 19,131 20,780 22,490 Fixed investment 13,627 15,606 17,224 18,926 20,559 22,253 Changes in stocks -359 63 197 205 220 237 Memo items Terms of trade adjustment -2,141 -2,834 -3,642 -3,141 -2,601 -2,050 Gross national income 54,878 58,155 61,601 66,136 71,059 76,417 Gross national product 57,019 60,989 65,242 69,272 73,660 78,467 - 115 - compensation base was determined after the 1985 collapse of new-entrant wages. It is this rising proportion of low-wage workers that will bring average wages down in 1988, and will initiate a fall in the share of total wages in national income and a rise in the share of profits. 4.35 In the short-run, employment will respond positively to the fall in wages. It will also increase as a direct result of output growth. In the most probable scenario, total employment would rise by 4.5% in 1988, faster than the rate of growth of GDP. Nevertheless, the total wage bill would only rise by 1.51 over 1987. This implies that total profits will surge by almost 1OX, providing an additional boost to investment. 4.36 The five major positive forces behind investment noted above--the cyclical demand increase, the SLCHP, renewed oil and gas interest, higher foreign direct investment, and the profits bonanza-are not all independent phenomena, but to some extent represent different ways of accounting for the same expenditures. Overall, private investment should grow by around M$2 billion. With public investment remaining approximately constant, this would generate a 141 growth in total domestic investment. 4.37 Over the medium-term, investment growth should continue to outstrip GDP growth, although by smaller amounts. Fixed investment growth should average about 9Z over the next decade. This would be made possible only as a result of private consumption growing faster than GDP, providing the demand stimulus to support investment. Additionally, public investment should start to pick up once the public finance situation stabilizes in 1991. Foreign direct investment should also continue to be a prime mover in investment growth over the medium-term. 4.38 Private Consumption. As is the case for investment, private consumption should also benefit from a strongly positive cyclical effect which would outweigh the negative structural forces affecting private disposable income. Private consumption behavior in Malaysia appears to respond to the trend in long-run sustainable income which is currently beset by two counter- vailing forces. On the one hand, private disposable income is being squeezed by the trend towards reducing the fiscal deficit. The expected improvements in federal tax collections as a result of elimination of exemptions and loopholes and the measures to improve public enterprise operating surpluses have, as counterpart, a decline in the share of national income remaining in private sector hands. A 6% growth in GDP in 1988 would only generate a 2.3% growth in private disposable income, after allowance is made for the addi- tional expected public sector take. On the other hand, as consumers' memories of the 1985 recession recede, they are likely to revise up their estimate of long-run sustainable income, closer to the levels represented by 1987 and 1988. Overall, long-run income growth will be 5-6%. On this basis, private consumption is forecast to grow by 7-9Z. As general government consumption will be restrained, with forecast growth of 1.6% , total consumption growth would be only 7.11 in 1988 (Table 4.6). 4.39 Over the medium-term, consumption can grow faster than GDP, without jeopardizing the nation's finances. Malaysia posted a savings rate of almost 33% of national income in 1987, a performance largely responsible for the - 116 - substantial surplus in the current account of the balance of payments. If this were to be slowly brought down to 28%, it would provide the leeway for the necessary consumption growth over the medium-term. Within this level, however, public savings should recover its previous positive contribution. Thus, private savings could fall even more steeply and private consumption rise even faster at the rates indicated in Table 4.6. Major Uncertainties 4.40 The macroeconomic scenario presented abov~:e is just one of many possible outcomes. It represents a balance between the need for restraint in public expenditures and the desire to assist more rapid growth. By stressing the role to be played by the private sector, however, the scenario is inevitably speculative in nature. Two additional sources of uncertainty should be stressed--public sector policy and the external environment. 4.41 In order to generate sustained growth, the base case scenario imposes short-term constraints on public expenditure that are then relaxed in the medium-term. This is desirable to bring public finances under control. In the short-term, it would be feasible for the public sector to be more expansionary, but the resulting build-up in debt would make future growth less likely. Investment response to public demand, furthermore, would involve a sectoral allocation of resources that would differ from that induced by private demand growth. Hence, inefficiencies would begin to emerge, slowing growth in the long run. 4.42 The other major source of uncertainty lies in the international environment. Stock market volatility in the OECD points to considerable variability in expectations for future growth. As indicated above, commodity prices are forecast to improve after 1990, providing some leeway for real consumption growth to expand rapidly without jeopardizing the balance of payments. In the absence of such improvement, the trade surplus could dis- appear much more quickly than envisaged. To some extent, this could be offset by stimulating even more rapid export growth in manufactures. It would also, however, imply that growth and consumption should be slowed to bring down imports. B. Integrated Policy Recommendations 4.43 Although Government has undertaken substantial reforms and initiated new adjustment programs aimed at reducing market distortions in several areas, the results to date have been muted. For example, government has liberalized industrial licensing, moved on privatization, reduced bureaucratic delays in licensing, engineered a fall in interest rates and so forth, but until 1988, when the external environment improved, there was little response from private investment. This should not, however, be taken as a sign that these distortions were unimportant or that a more active interventionist stance is required from government. Rather, the preceding analysis suggests that the lack of response is a function of overlapping constraints. In many instances, Government has moved in an ad hoc fashion without sufficient consideration of how policy reform in one area can offset programs in another area. In order to achieve the growth scenario outlined above, the benefits of past reform - 117 - Table 4.6: NATIONAL ACCOUNTS GROWTH RATES (X) AT CONSTANT PRICES 1988 1989 1990 1987-92 Gross domestic product m.p. 6.5 6.1 5.9 6.1 Agriculture 2.3 3.0 3.0 2.9 Industry 10.9 9.2 8.3 8.9 (of which Manufacturing) 14.0 12.0 11.0 11.7 Services 4.2 4.5 5.0 4.8 Exports of GNFS 9.1 6.1 6.1 6.6 Imports of GNFS 13.1 11.3 10.7 10.7 Total Expenditures 8.4 9.4 9.0 8.6 Total Consumption 7.1 8.6 8.6 8.1 Private consumption 9.2 10.9 10.3 9.7 General government 1.6 2.0 3.0 3.1 Gross domestic investment 12.0 11.2 9.8 9.9 Fixed investment 14.5 10.4 9.9 10.1 Change in stocks -82.5 212.0 4.2 5.0 Memo items Gross national income 6.0 5.9 7.4 6.8 Gross national product 7.0 7.0 6.2 6.6 I -8 - should be unlocked by rationalizing er*' .orv`ving policies and how further reforms structured in a comritementary f6jsbioit tu contribute to objectives of efficiency, macroeconomic stability and growtk3 Consistency in Policy Programs 4.44 A few examples of areas where go-ernment programs conflict with each other, reducing the effectiveness of each, illustrate the need to approach policy reform from a well-developed economy-wide perspective. One of the most important concerns the interaction between policies for mobilizing savings and programs to promote investment. 4.45 There are two significant ways that Government has intervened to promote savings. One is through the "forced saving" contributions to the Employee Provident Fund, the Social Security Organization, the Armed Forces and the Teachers' Provident Funds. Between them, the surpluses genorated by these funds is about 7Z of GNP. In addition, there are government-sponsored savings schemes, including ASN, which pay high returns, made possible through implicit Government subsidies, yet involve very low risk. These funds primarily channel savings towards the public sector. 4.46 In an environment of heavy direct government intervention in the economy, it was appropriate to institutionalize mechanisms to generate govern- ment financing. These captive funds have been primarily responsible for the success achieved by Government in running large overall deficits to finance investment without recourse to excessive monetary financing, pressure on the banking system or an overload of external debt. Presently, however, the funds are continuing to increase in size while Government financing requirements are falling as a result of the decision to reduce the direct role of the public sector in the economy. Given the absence of effective mech-anisms for channelling these funds to the private sector, an imbalance in the flow-of- funds has arisen, and is expected to persist into the future. The magnitude of the imbalance is difficult to predict. Although the largest fund, the EPF, makes some projections of its future surpluses, these are inadequate for use in macroeconomic planning as they implicitly assume no population growth after 1990 and no growth in average income over time. Nevertheless, it is apparent that the mechanisms to generate savings act against the ability to transform savings into investment and against government programs to encourage private investment. 4.47 A second example of policies that counteract each other is to be found in industrial and labor legislation. On the one hand, Government policies to promote industry have favored capital-intensive investments. This is true of the design of the corporate tax structure, wherein the most important incentives--tax holidays, investment and depreciation allowances- interact to subsidize capital. Equally, it is true of the programs for industrial diversification and the push towards heavy industry development by NFPEs. These, by nature, have been highly capital-intensive. On the other hand, Government has tried to promote labor utilization (unsuccessfully) through tax incentives and (more successfully) through a reorientation of government expenditure towards labor-intensive projects. - 119 - 4.48 Similarly, Government attempts to promote private investment through liberalization of licensing have been stymied by the offsetting impact on private investment of public enterprise operations and of various wage promo- tion and labor protection programs of government. Inflexibility in wages, public countercyclical policies and taxation of corporate employment via EPF contributions have served to reduce profitability and deter private invest- ment. 4.49 A final example concerns privatization of NFPEs and controls on the domestic capital market, especially regulation of the Stock Exchange. Amongst other objectives, privatization is designed to bolster Government finances and to restructure the balance sheets of public corporations where operating losses, heavy investments and strong currency external debt revaluations have increased debt/equity ratios. Yet both these objectives are difficult or costly to meet in an environment of strict controls on the price at which new shares can be issued, and take-overs or mergers performed. 4.50 Malaysia's growth potential can best be unlocked through packaging policy reforms in a cotuplementary fashion, avoiding the conflicts mentioned above. Emphasis should be on microeconomic reforms that enhance efficiency which, taken together, simulate a macroeconomic adjustment program. An example of this would be removal of mandatory EPF contributions for the young. This would foster a more efficient labor market and improve the matching between the earnings-expenditure distributions for individuals while also stimulating private demand. On the other hand, short-term macroeconomic remedies that do not foster efficiency should be avoided. These include reflationary public expenditure, subsidized directed credit programs and generalized tariff reform to protect ailing industries. Business Confidence 4.51 With the diminished role of the public sector and the difficulty in expanding net exports beyond their record levels in 1987, private sector demand will play a leading role in growth. For this to occur, business confidence must be strengthened. It is only in a dynamic private environment that government programs such as privatization can be expected to yield the benefits envisaged. Government can send a clear signal to the private sector by reforms which clearly indicate official intentions to withdraw from the provision of gocds and services in areas where private activities could be encouraged. Privatization, reduced government spending, lower subsidies, liberalized industrial licensing, corporate income tax reform and tariff reform would all demonstrate the consistency of Covernment's commitment to growth and efficiency that is necessary to foster confidence in the economy. They would also indicate the strategic directions envisaged for economic policy after 1990. Amendments to the EPF 4.52 Because of its size and importance in the labor market, several recommended reforms involve the EPF. Clearly, reduction in EPF surpluses must be made contingent on an equal reduction in public deficits in order not to burden the banking and monetary systems. As discussed below, these public - 120 - sector reforms should also be tailored towards efficiency and growth-enhancing measures. 4.53 Reforms affecting the EPF in four areas should be considered: a lowering of average contribution rates by the corporate sector; a voluntary exemption for those under 25 years in age; a deduction from corporate EPF contributions for 50% of firms' costs of approved training; and a mandate to withdraw funds for approved education and training. Hypothetical simulations indicate that the current corporate contribution to EPF, at 11%, imparts a significant bias against labor that may be in part responsible for the low output elasticity of employment displayed by the manufacturing sector in the 19809. The coincidence between the period of high manufacturing emplcyment growth and low EPF contributions in the 1970s, on the one hand, and low employment growth and high EPF contribution rates in the 1980s, on the other hand, also suggests a casual relationship. Reduction in corporate EPF contributions would improve microeconomic efficiency at a firm level and help bring Malaysia's wage costs more in line with other East Asian competitors. 4.54 Exemption for the Young. The voluntary exemption from participation in EPF by the young is recommended as an instrument to improve the employment prospects of this group, to reduce the wedge between the take-home pay of new entrants and those hired prior to 1984, and to induce the young to gain valua- ble work experience, even at lower overall wages. In theory, the exemption could reduce the effective wage costs to the firm of hiring a new entrant by as much as 20%. In the current slack employment market, the actual effective wage reduction may approach this. As the labor market tightens, however, as is expected in 1988 and 1989, che exemption should result in higher take-home wages for the young. This a'-oup probably discounts retirement savings at a very high rate, so the loss of accruals in the EPF would not be of great concern. Thus, the total perceived compensation package would rise, inducing a greater willingness by the young to work. In so doing, valuable work experience is gained that should reflect on later career earnings. This measure is expected to benefit more highly educated, secondary school leavers and new college graduates in particular. 4.55 Financing of Training. One of the most severe drains on the Government budget comes from its support of higher education and vocational training programs. A mandate to permit use of EPF funds for such purposes, as well as approved on-the-job training programs, would help take these programs off-budget completely, either through privatization of programs or through greater use of user-charges. The EPF should have institutional interest in this: the support of human capital formation is a natural adjunct to finan- cial capital accumulation and physical capital formation that are already supported by the institution (the latter through the Housing withdrawal scheme). Further, it has a natural comparative advantage in administering a student loan program. It could automatically deduct repayments from individuals' future contributions, avoiding the problems of collection that bedevil other loan schemes. Also, through its detailed information base on wages, the institution could provide valuable inputs into an understanding of the rates of return to alternative education and training schemes, career development paths and such-like. It would thereby bring market signalling closer to the discussions on curricula development and course preparation than the present system allows. - 121 - 4.56 As far as on-the-job training is concerned, EPF participation could emerge in two forms. First, it could participate in an Employment Training Panel, that would approve courses for which payment by an individual from his personal accumulated contributions to a private training contractor could be made. Approval would only be granted when an individual has been trained, placed and retained in a job (say for 90 days) for which training was undertaken. Second, similar approval could be granted for within-firm training. Experience suggests such a system can significantly help the unemployed and sharply reduce retrenchment rates for those employed but in ianger of being laid off. Such a system adds competitiveness to the delivery of training systems that is absent under current vocational training schemes. It also energizes firms to provide on-the-job training for employees and induces them to meet the dislocations brought about by accelerating economic and technological change via retraining rather than retrenchment. 4.57 Implementation of these measures would have four desirable macro- economic consequences: (i) it would boost employment overall, and selectively for the young; (ii) it would reduce the EPF surplus, add discretionary income into workers' hands and stimulate private consumption; (iii) it would reduce the effective tax on corporations, raise profits and spur private investment in socially desirable labor-intensive sectors; and (iv) it would reduce the public sector deficit without reducing access to education services. Public Finances 4.58 Reduction in EPF surpluses would not be appropriate without measures to reduce the government deficit. One problem that is faced, however, is that the deficit seems intractable without growth, as real expenditures have already been slashed to the bone, while growth may remain elusive if the deficit is cut by traditional fiscal means. From an efficiency stand-point, government may actually need to increase expenditures, especially for maintenance of the sizeable infrastructure network that was created in the early 1980s. On the tax side too, reform of corporate income taxes and introduction of the value-added tax in place of the sales tax are both warranted from an efficiency point of view but may both involve revenue losses during a transitional period. 4.59 Rather than introduce new distortions or inefficiencies through use of the existing tax structure, imposition of new tariffs or cut-backs in real expenditure, this report suggests attention be focused on reducing explicit and implicit subsidies to public authorities, statutory bodies, NFPEs, and the like. One such measure has already been discussed, namely the raising of user charges in education and vocational training coupled with a broader student loan program. Other measures include targeted cuts in the Treasury housing loan program, in rice growing, milling and distribution subsidies, and in the railways, to mention a few. Of these, only padi coupons and the fertilizer subsidy appear explicitly as budget subsidies. Others are listed only under individual public programs. Identification of subsidies is, therefore, complicated. 4.60 One continued drain on the budget, and a reason for sluggish overall economic performance, is the non-financial public enterprise sector. This - 122 - sector contains several hundred loss-making units whose operations largely escape institutional oversight. The financial losses may even disguise their true costs of inefficiency as additional relief in the form of tax exemptions, price fixing, public procurement and tariff protection is often first pro- vided. Government has the capability to study the top ten or so largest money losers. The larger problem, however, is how to tackle the myriad of smaller companies and how to ensure that even those that are financially solid are operating efficiently rather than benefiting from specific privileges. The report recommends (i) privatization and (ii) establishment of a performance evaluatior system for NFPEs. The latter takes on additional importance if the pace of privatization proceeds slowly, or where privatization is unlikely to result in efficiency improvements because of specific monopoly, guarantee, or economics of scale issues that preclude competition even for a privatized company. 4.61 In spite of recent improvements in the public deficit, total public debt in Malaysia remains higher than in most other countries. To some extent this is offset by considerable public assets. Nevertheless, interest expenses are still expected to amount to over 1OZ of GNP over the medium-term. This is not only a significant cost on the budget, but it also exposes the public sector to considerable volatility in its finances. In addition, the revenue side of fiscal accounts is also volatile, depending to a large extent on petroleum and other commodity prices. Until now, government has reacted to its financial situation by adjusting real expenditures, particularly develop- ment spending. This is, however, a highly inefficient way of responding to external shocks. Government should use new financial instruments to reduce its average borrowing costs and to reduce the volatility that it faces. Such financial hedging can be pursued both with respect to external debt and internal debt. 4.62 Stability in public expenditures and reduced fiscal deficits will allow for more efficient public investment planning. They will also provide stability in the macroeconomic environment that is necessary to stimulate private investment. Unlike in other countries, Malaysia has not witnessed much crowding-out of the private sector as a result of public sector borrowing from the banking system and a resultant rationing of credit or high interest rates. Yet the public sector has displaced p- -late investment in important ways; first, by direct competition in productive activities; and second, by bidding up investment costs--of labor, materials and land-through policies of protection, licensing and unwartanted expansion. A better delineated role for the public sector, perhaps outl;ned in the Privatization Master Plan, would address these issues. Conclusion 4.63 The recommendations outlined in this report would provide the basis for respectable, sustainable medium-term growth in Malaysia. As a result of unfavorable prospects for real growth in rubber, tin, petroleum and forestry, the projections suggest a slow-down in overall growth relative to the 1970s, even with a rapid growth in manufacturing. As manufacturing expands, however, and if efficiency and productivity improve, growth could well exceed the conservative scenario depicted. In fact, with an improved private sector - 123 - orientation, Malaysia could well position itself to enter the ranks of the New Industrializing Countries by the end of the sixth Malaysia Plan. It is from this perspective that the recommendations focus on efficiency. They should not, therefore, be thought of as dependent on growth, nor should a desire for more rapid growth lead to substantial deviation from the policy agenda outlined here. Given its strong external position, Government has the ability to force growth to a higher level for a few years but, as the experience of the 1980s shows, such performance cannot be sustained indefinitely and serves to undermine the diversity and flexibility that private sector development brings, qualities that are crucial to long-run growth in a world with rapidly adjusting trade and industrial structures. Appendiz 1 -124 - Page 1 of 2 MULAYSIA ASSESSINC THE VARIABILITY OF NONOIL PRIMARY COMNODITY EXPORTS 1. Malaysia has a diversified portfolio of primary commodity exports, the most important of which are rubber, palm oil, sawlogs and timber, tin and cocoa. Diversification helps to limit the volatility in export earnings when the prices of individual commodities are uncorrelated (or negatively correlated) with each other. The ability to diversify, then, depends on whether the same underlying forces affect all commodity prices (such as a global recession, for example), or whether unrelated coamodity specific events (such as a drought in India driving up palm oil prices, or a frost in Brazil boosting cocoa prices) are more important determinants of prices. 2. The analysis below is based on monthly commodity prices from June 1973 to December 1987. A longer historical series is available, but the resulting variance in the rate of change of prices proved to be unstable over time. Consequently, the shorter time period is used as this more accurately depicts the randomness in commodity prices today. All commodity prices were deflated by a manufacturing unit value index, to remove the global inflationary component. We have assumed that the resulting monthly percentage change in real prices is driven by a Wiener process, that is: dp= -ldt + o dz p where i is the expected trend increase over time in the price and a is the standard deviation. 3. The means, variances and correlations of the percentage change in real prices was calculated and found to be as below. In general, the correlations are small, indicating that commodity specific events mostly determine price fluctuations. The most volatile conmodity price has been palm oil, but rubber, logs and cocoa also exhibit sizeable swings. On average, commodity prices have been trending downwards with the exception of logs, for which real prices have risen by an average 6X per year. - 125 - Appendix I Page 2 of 2 Table A.1: MEANS, STANDARD DEVIATIONS AND CORRELATIONS OF COMMODITY PRICE GROWTH (Z) Mean S.D. Correlation Matrix Rubber -2.775 23.48 Palm oil -0.149 32.37 .0104 Logs/timber 6.005 22.88 .0566 .1609 Tin -2.386 17.47 .0886 .1038 .0904 Cocoa -1.065 23.51 .0515 .0705 -.0291 .1587 /a Figures are annualized and refer to the growth of prices. 4. An example illustrates the table's contents. Rubber prices on average have fallen by 2.8Z per year, but two-thirds of the time (one standard deviation), the annual change in rubber prices has been between-26Z to +21X. Through diversification, however, the volatility exhibited by individual commodity export earnings has been reduced. Based on the value shares in exports in 1984, the ringgit value of the sum of these commodity exports (totalling around M$13 billion) is less volatile; the standard deviation of the rate of change in export earnings due to price changes is calculated to be 15.76Z, and the mean increase is 0.22X per year. This implies that one-third of the time, export earnings are likely to differ from their expected value by more than M$2 billion (one standard deviation), and one-sixth of the time, there will be a shortfall of this magnitude. Data 5. The commodity prices are defined as below: (a) Rubber: RSS No. 1, US, in bales, New York, spot; (b) Palm oil: Malaysian, 52 bulk, cif NW Europe, nearest forward shipment; (c) Logs: 1973-76, Philippines, Lauan, 3.6 m or more x 60 cm, Tokyo delivered, importers sale price to wholesalers. Beginning 1977, Malaysian, Meranti, Sabah SQ Best Quality, sale price changed by importers, Japan; (d) Tin: 1973-September 1984, official settlement price, Penang, ex-smelters for Straits, Min. 99.85Z purity. Beginning October 1984, Kuala Lumpur Tin Market settlement price for Straits; (e) Cocoa: ICCO Daily Price, average New York and London, nearest 3 future trading months. Appendis 2 - 126 - Page 1 of 3 MALAYSIA Construction of the flow of Funds Table 1. Malaysia does not produce flow of funds tables for the whole economy. We have constructed approximate tables for 1980, 1985 and 1986. The year 1980 was taken as the year before serious internal and external imbalances developed. Although 1979 may have been a better year for it, data limitations precluded that choice. The years 1985 and 1986 represent the period of distress, especially in the private corporate sector. 2. Overall Structure. The economy is divided into four real sectors: Government (Federal, State, and local governments, plus statutory authorities), NFPEs, private corporations, and households. The first two form the public sector and the last two the private sector. For 1980, however, it is not possible to separate NFPEs from private corporations. The financial intermediaries are divided into the banking sector (commercial banks, finance companies, merchant banks, discount houses, and Bank Negara), provident funds, and other financial intermediaries. Included in "other financial intermediaries" are insurance companies, National Savings Bank, cooperative banks, Bank Rakyat, development banks, Building Society, farmers' organizations, PNB, and Tabun Haji (Pilgrims Fund). The flow of funds table was constructed as follows. 3. Overall Savings and Fixed Investment. The overall savings figures, gross national savings, are available from the national accounts. The consolidated public sector account gives current surpluses of the Government and NFPE sectore (but Government only for 1980). Making allowances for the difference between the development budget (on which the consolidated public sector account surpluses are based) and the fixed capital formation figures in the national accounts, true public sector savings are derived. The balance of gross national savings is generated by the corporate and household sectors. 4. Total private sector savings and fixed capital formation are distri- buted to the household and corporate sector as follows. The corporate sector figures are derived as residuals. (a) Using financial statistics from various subsectors of the financial system, total financial investment and financial liabilities of the household sector are identified and the net financial investment estimated. Household savings in the EPF include both the excess of contributions over withdrawals and interest earned on investments. It is derived from changes in the end period amounts outstanding to credit of contributors. (b) Fixed capital formation by the household sector is estimated on a simple assumption that new residential housing investments are equivalent to total fixed capital formation by this sector. The remainder of private fixed capital formation is assigned to the corporate sector. Appendix 2 - 127 - Page 2 of 3 (e) Adding the net financial investment and fixed capital formation, the total savings of the household sector are estimated. (d) Subtracting household savings from total private savings, corporate sector savings are estimated. (e) Specifics. For the Government sector, a broad breakdown of financing is known from the consolidated accounts. Specifics, for instance about which sector holds government bonds, are availabLe from Bank Negara Annual Reports and Quarterly Bulletins. Some transactions that have no obvious counterparts in other sectors are assigned to the NFPE sector (see below). 5. For the financial sector, uses and sources of funds are available from Bank Mega-a publications by the type of assets (not by economic sectors). The available aggregate statistics, e.g., for the banking sector, are incorrect in that sources and uses of funds within the sector are no netted out. After correcting for this, the sources and uses according to economic sectors were identified, with an element of guesswork. In this process, the retained earnings of Bank Negara was added to Government savings and private sector savings were reduced accordingly. The counterpart to the Government savings is the increase in equity investment by Government. 6. For the UPPE sector, both sources and uses include items that are more or less balancing items. For instance, unidentified domestic loans to Government were all assigned to the NFPE sector, for PETRONAS is the most likely source of such financing. Likewise, liquidation of assets by the Government that does not show up in the financial sector statistics is assigned to NFPEs. Finally, the changes in deposits belonging to NFPEs is the balancing factor against the changes in corporate deposits. (The total deposits of business enterprises are known. First, the private corporate sector portion is estimated by using this as one of the fine-tuning entries to ensure adequate funding for the corporate sector. Then the NFPE portion is obtained as the residual.) 7. Transactions with the foreign sector that are not already captured are added to the picture. Incidentally, the errors and omission line in the balance of payments is believed to consist of two elements: Bank Negara's capital gains from foreign exchange holdings and unrecorded private capital movements. The latter is treated as an increase or decrease in the household sector's deposits held abroad. 8. Changes in inventories (negative in all three years) are recorded as a source of financing for the NFPE and corporate sector. 9. Some Cautionary and Technical Notes. The table fails to capture the direct equity investment from the household to private corporate sectors. However, since the two sectors are linked directly through the estimate process for the corporate savings, any increase in such direct investment will increase household savings and reduce corporate savings by the equal amount. Therefore the balance between sources and uses of funds will remain intact. - 128 - Appendix 2 Page 3 of 3 10. In the flow of funds tables, sector breaidowns are shown for five types of financial transactions2 deposits, loans, securities, equity, and other. For deposits, the sectors refer to the direction of use, and for the others, the sectors refer to the direction of source. 11. For 1986, "other financial" registered a large decline in capital and reserves and a large increase in "other" source of funding. They were both assigned to households. Appendix 3 - 129 - Page 1 of 7 EPF: Life-Time Accumulation by the Individual 1. The accumulation of assets by individuals in tIeir Employee Provident Fund accounts is the single most important source of savings by individual members. In the aggregate, the EPF is also a major source of national savings. The magnitude of EPF surpluses and the size of individual retirement accounts depend importantly on macroeconomic variables, which determine wages, the growth of wages over time and interest rates, as well as on microeconomic variables such as the contribution rate, the age of entry and of retirement, training and education. This Appendix explores how individuals and the EPF might be affected under alternative scenarios. The Base Case 2. Estimatins Life-Time Income Profiles. The EPF has readily available data on average contributions per member for each age. These are listed in Column 1 of Table 1 below. The average income for this group may be derived by multiplying this figure by 5 (as the contribution rate is 201). Any individual, however, may have a different educational level than the average EPF contributor. To derive an individual-specific profile, therefore, it is necessary to correct for this education effect. Data on average income for a High School Certificate graduate, taken from the Department of Statistics' Income and Labor force survey, 1984, was compared with that for the average EPF contributor and found to be 1.458 times higher (for an MCE graduate, the adjustment factor was 1.076). Thus, average income for a HSC graduate, shown in column 3, is derived by multiplying the average EPF member's income by 1.458. An HSC graduate is assumed to start work only at age 18. 3. The cross-section data, showing wages by age, do not exactly replicate an individual's expected earnings, because they do not factor in potential productivity growth. Over the long-run, wages in Malaysia have moved closely in line with productivity growth. In the base case scenario, we assume productivity growth of 1 percent per year. The overall expected income stream, incorporating productivity growth, is given in column 4 of Table 1. Note that the impact of inflation on higher nominal wages is not included--all data are implicitly expressed in 1986 ringgit. 4. Accumulation Process. Given the expected life-time income profile, an individual's annual contribution can be calculated (20Z). This is credited to an account (column 7) at the end of the working year. Intetest is earned on the outstanding balance at the beginning of the year. Interest is expressed in real terms rather than nominal terms. In the base case, a real interest rate of 3.0 percent is used. This is the average real interest rate in Malaysia between 1970-86 (3.02Z), excluding the two high inflation, outlier years of 1973 and 1974. Interest is also credited to the individual's account. 5. Benefit Level. The individual is entitled to withdraw his accumulated contribution at age 55. As this is expressed as a stock amount, however, it is difficult to compare it with the income cash-flow earned prior to retirement. We assume that the individual converts his contribution into a Appendix 3 Page 2 of 7 - 130 - Table 1: INDIVIDUAL EPF CONTRIBUTION PROFILE Overall income growth - IZ p.a. Real interest rate - 3X Education level - HSC Average EPF Income Expected Expected Accumula- contribution adjustment Average incom contri- Interest ted con- Age in 1986 factor income stream bution earned tribution (1) (2) (3) (4) (5) (6) (7) 16 185 0 0 0 0 0 0 17 237 0 0 0 0 0 0 18 323 1.458 2,355 2,355 471 0 471 19 387 1.458 2,821 2,849 570 14 1,055 20 485 1.458 3,536 3,607 721 32 1,808 21 584 1.458 4,257 4,386 877 54 2,739 22 703 1.458 5,125 5,333 1,067 82 3,888 23 835 1.458 6,087 6,398 1,280 117 5,284 24 991 1.458 7,224 7,669 1,534 159 6,977 25 1,155 1.458 8,420 9,027 1,805 209 8,991 26 1,345 1.458 9,805 10,617 2,123 270 11,385 27 1,517 1.458 11,059 12,095 2,419 342 14,145 28 1,672 1.458 12,189 13,464 2,693 424 17,262 29 1,805 1.458 13,158 14,680 2,936 518 20,716 30 1,897 1.458 13,829 15,583 3,117 621 24,454 31 1,958 1.458 14,274 16,245 3,249 734 28,437 32 2,011 1.458 14,660 16,852 3,370 853 32,661 33 2,056 1.458 14,988 17,401 3,480 980 37,121 34 2,094 1.458 15,265 17,900 3,580 1,114 41,814 35 2,125 1.458 15,491 18,346 3,669 1,254 46,738 36 2,151 1.458 15,681 18,757 3,751 1,402 51,891 37 2,171 1.458 15,827 19,120 3,824 1,557 57,272 38 2,187 1.589 15,943 19,454 3,891 1,718 62,881 39 2,200 1.458 16,038 19,765 3,953 1,886 68,720 40 2,211 1.458 16,118 20,063 4,013 2,062 74,794 41 2,221 1.458 16,191 20,355 4,071 2,244 81,109 42 2,231 1.458 16,264 20,651 4,130 2,433 87,673 43 2,231 1.458 16,264 20,857 4,171 2,630 94,474 44 2,231 1.458 16,264 21,066 4,213 2,834 101,522 45 2,231 1.458 16,264 21,277 4,255 3,046 108,823 46 2,231 1.458 16,264 21,489 4,298 3,265 116,385 47 2,231 1.458 16,264 21,704 4,341 3,492 124,218 48 2,231 1.458 16,264 21,921 4,384 3,727 132,329 49 2,231 1.458 16,264 22,141 4,428 3,970 140,727 50 2,231 1.458 16,264 22,362 4,472 4,222 i 9,421 51 2,231 1.458 16,264 22,586 4,517 4,483 158,421 52 2,231 1.458 16,264 22,811 4,562 4,753 167,736 53 2,231 1.458 16,264 23,040 4,608 5,032 177,376 54 2,231 1.458 16,264 23,270 4,654 5,321 187,351 55 2,231 1.458 16,264 23,503 4,701 5,621 197,672 Notes: (1) Source: EPF. (2) Actual average income for HSC level workers divided by average of (1) for ages 18-55. (3) - (1) x (2) x 5. (4) - (3) multiplied by (1.01) raised to the power of (age - 18). (5) - (4)/5. (6) - previous (7) x real interest rate. (7) - previous (7) + (5) + (6). Appendix 3 - 131 - Page 3 of7 20-year annuity-that is, we calculate the constant amount an individual can withdraw from his account every year, such that the total fund is exhausted after 20 years. Naturally, in the years prior to exhaustit'i, interest is still earned. The annuity figure gives an estimated retirement income which can be compared to income during an individual's working years. 6. Sensitivity to Interest and Crowth Rates. A comparison between retirement income and peak working year disposable income is shown in Table 2 below for various interest and growth rates. In the base case (row 2), the annuity value of the accumulated contribution of M$197,672 is M$13,287. This is 62% of the highest disposable income during working years (income shown in Col. 4 of Table 1 less the 9Z individual contribution to the EPF). With a lower real interest rate of 2.5% real per year, the annuity falls by 12%. At a higher interest of 3.25%, however, close to the average between 1977-86, the annuity rises. At 1987 interest rates of 6%, the annuity would exceed the highest working income (not shown). 7. Two other sensitivities are shown in Table 2. If productivity growth averages 2X per year rather than 1Z, the annuity income rises sharply, by 22Z. However, the higher wages earned imply that the highest working income grows by even more, by 44Z to a level of M$30,795. Thus, the ratio of retirement income to working income falls as growth rises. This is true at all interest rates and at different education levels. Finally, Table 2 shows the impact of making the more realistic assumption that individuals work beyond age 55 upto age 60 and, even if they no longer participate in the EPF, behave as if they do. The additional 5 years of work and contributions raises the annuity income by 29 percent to M$17,130 in the base case. The ratio to highest disposable work income also rises to 76 percent. S. The sensitivity analysis demonstrates that retirement incomes are likely to have risen relative to highest disposable working income in the 1980s as a result of two main macroecinomic changes: (i) the rise in real interest rates and (ii) the slowdown in wage and productivity growth. At present contribution levels, particularly if the realistic assumption is made that individuals work until they are 60, retirement incomes from accumulated EPF contributions are likely to exceed the highest disposable income attained by an individual during peak working years. This is considered to be an anomaly of the system. 9. Two schemes are recommended for reducing individual retirement incomes and total EPF surpluses: allowing withdrawals for training and exemptions for those under 25. First, assume an individual is allowed to withdraw one year's contribution (i.e. 20X of annual income) to finance training, which yields a rate of return of 15 percent. The change in accumulated contribution will arise from the deduction for training, and from the higher contributions in future years resulting from higher wages after training. The net effect depends on the age at which training is undertaken. Table 3 shows the net effect is to reduce the accumulated contribution by a relatively small amount. The amount approaches 1 percent for training at age 25. It is about the same regardless of the educational level of the trainee. Appendix 3 Page 4 of/ - 132 - Table 2: RETIREMENT INCOME AND WORKING INCOME Annuity Highest Annuity Real Accumulated payments disposable as Z of interest rate contribution (20 years) income (HDI) HDI M%- (M$) Base Case 2.5 182,102 11,681 21,387 54.6 3.0 197,672 13,287 21,387 62.1 3.25 206,088 14,174 21,387 66.3 2% Annual Wage Growth 2.5 223,569 14,341 30,795 46.6 3.0 241,609 16,240 30,795 52.7 3.25 251,337 17,287 30,795 56.1 Age 60 Retire- ment 2.5 231,478 14,849 22,478 66.1 3.0 254,855 17,130 22,478 76.2 3.25 267,652 18,409 22,478 81.9 Appendix 3 - 133 - Page 5 of 7 Table 3: IMPACT OF EPF WITHDRAWAL FOR TRAINING Real interest Accumulated contribution rate (Z) No training Training M$ $ -- (in 1986 M$) - Training Age - 20 2.50 135,547 135,161 -386 -0.28 3.00 147,296 146,743 -553 -0.38 3.25 153,645 153,006 -639 -0.42 HSC 2.50 182,102 181,579 -523 -0.29 3.00 197,672 196,951 -721 -0.36 3.25 206,088 205,252 -836 -0.41 Training Age = 22 MCE 2.50 135,564 134,958 -606 -0.45 3.00 147,296 146,489 -807 -0.55 3.25 153,645 152,722 -923 -0.60 HSC 2.50 182,102 181,304 -798 -0.44 3.00 197,672 196,605 -1,067 -0.54 3.25 206,088 204,867 -1,220 -0.59 Training Age - 25 MCE 2.50 135,564 134,507 -1,057 -0.78 3.00 147,296 145,943 1,353 -0.92 3.25 153,645 152,125 -1,520 -0.99 HSC 2.50 182,102 180,693 -1,409 -0.77 3.00 197,672 195,866 -1,805 -0.91 3.25 206,088 204,058 -2,029 -0.98 Appendix 3 - 134 - Page 6 of 7 10. Impact of exemption of those under 25. Another option is to exempt those under 25 from the EPF. Alternative scenarios with exemptions for just employers and both employers and individuals are shown (Table 4). With a total exemption, an individual's retirement income would only be brought down by 9.71, from M$13,287 in the base case to M$12,114 with the exemption. The exemption from employers only would lead to a reduction of just 5.1Z. The results for alternative interest rates are also shown. As the interest rate rises (falls), the reduction in the retirement annuity rises (falls). 11. Civen that an exemption would only marginally affect the retirement income of the young, an additional issue is the macroeconomic effect on total EPF contributions. The data on contributions by age was not made available to the mission directly. Nevertheless, from data on the age distribution of the population, the EPF participation rate by age, and the number of contributing members, the aggregate contribution of those under 25 (i.e. age 24 and below) can be estimated. This amount is M$448 million in 1986 or about 14.21 of total contributions in that year. If only employer contributions are exempted, the total reduction would be M$247 million. If the exemption age were lowered to under 21, the loss to the EPF would be M$127 million. Appendix 3 Page 7 of 7 - 135 - Table 4: UNDER-25 EXEMPTIONS FROM EPF Annuity Highest Annuity Real Accumulated payments disposable as Z of interest rate contribution (20 years) income (HDI) HDI (X) (M$) Employer Exemption 2.5 173,946 11,158 21,378 52.2 3.0 188,079 12,642 21,387 59.1 3.25 195,686 13,459 21,387 62.9 Full Exemption 2.5 167,272 10,730 21,387 50.2 3.0 180,230 12,114 21,387 56.6 3.25 187,176 12,874 21,387 60.2 Appendiz 4 - 136 - Page 1 of 4 Methodology for Calculating Total Factor Productivity Growth Following the Gollop and Jorgenson (1986)11 methodology, a production function may be written: Q = F(K,L,X,t) Where Q is gross output, K is capital input, L is labor input, X is intermediate good input and t is time. With profit maximization, value share of each input is equal to the elasticity of output with respect to each output: = L _aln Q L qQ a ln L ..PkK a ln Q K qQ a ln K S= - = a In Q x qQ 1 n X Where q is price of output and PL, PK and P. are the respective prices of the inputs. Differentiate Q with respect to t, d ln Q = 3 ln Q d ln K + a In Q d ln L + a ln Q d ln X + 1 a ln Q d ln t a ln K dt a In K dt a ln X dt t a ln t S d ln K + S d ln L + S d In X + S k dt L dt X dt t where St is the residual and rate of total factor productivity change. For changes between discrete points in time, this can be approximated by: la Q -ln Q = [IK [n Kt- ln Kt-l i [In Lt - ln + S [ln Xt ln Xt-4] + St I/ Gollop and Jorgenson (1986), "Productivity and Growth in Sectoral Output in the Us", Harvard Institute of Economic Research, Discussion Paper No. 1217. Appendix 4 - 137 - Page 2 of 4 where SK [S S I and so on Notes on Data Used All the data are from surveys or censuses of manufacture published in the Yearbook of Statistics. The capital input data reflects the survey's assessment of fixed assets. While this is not satisfactory, alternative methods for estimating the capital stock by summing gross investment for example cannot be used with the existing data. The profit share is derived from the diff rence between gross value of output less cost of inputs less wages and salaries. The labor input data reflects wages and salaries paid which should take into account changes in age, sex, hours worked and educational composition. Results Survey data collected by the Department of Statistics tabulate output, input, fized assets and wage statistics for 28 manufacturing subsec- tors and for the sector as a whole. By comparing the growth in output between two dates with the weighted growth in each input, a residual measure, defined as the total factor productivity change, is obtained. This procedure was followed for a series of subperiods. The results are suumarized in Table A3.1. Although these should be treated with caution because of the residual- based methodology And inherent data oroblems. two important trends emerge. The first is that the growth of labor inputs, defined as the salary and wage bill, has been extremely low in both periods, and the contribution of labor to output growth has been correspondingly negligible. The second trend is the reversal in annual total factor productivity growth from 3.8Z per year in 1975-79 to -1.91 annually in 1981-84. These figures imply that while productivity change contributed over 121 to output growth in the late 1970s, comparable to the long-run performance in other newly industrializing counties, its contribution became negative in the 1980s. - 138 - Appendix 4 Page 3 of 4 Table A3. 1: GROWTH AND CONTRIBUTIONS TO GROWTH IN MANUFACTURING (X) 1975-79 1981-84 X contri- X contri- Growth /a but-on /b Growth /a bution /b Total output 22.9 100.0 9.2 100.0 Capital input 3.4 11.2 5.0 52.2 Labor input 0.8 2.6 0.0 0.3 Intermediate inputs 18.2 73.9 6.3 66.0 Total factor produc- Livity change /c 3.8 12.4 1.9 -18.5 /a Annual average compound growth rate. 7i Contribution of input to increment in output. 7T Residual. The statistical reason for the negative productivity growth is clear: fixed assets in manufacturing in the 1981-84 period accelerated to an annual growth of 51 per year. Based on the average profitability of capital, this should have produced over 50X uf the actual output iacrease. In addi- tion, under normal circumstances, an increase in fixed capital should be asso- ciated with an improvement in total factor productivity as technology is improved and techniques of production are adapted to local conditions. Instead, the growing capital intensity in manufacturing seems to have substi- tuted for technological advance. For the manufacturing sector as a whole, efficiency improvements can be explained partly as a result of developments in each subsector of manufac- turing, and partly as a result of a change in the composition of resource allocation. When resources are shifted from a less efficient to a more effi- cient sector, the overall efficiency is improved even if each sector performs the same as previously. To distinguish between these two effects, TFP growth rates were calculated separately for 27 subsectors identified in the DOS Survey of Industrial Companies.u' During 1975-79, 23 out of the 27 subsectors had positive total factor productivity growth. The most dynamic sectors were textiles, apparel, wood, nonindustrial chemicals, petroleum and coal, plas- tics, glass, cement and scientific instruments. By 1981-84, however, only ten subsectors had positive TFP growth. Of these, only leather and print had a faster TFP growth rate than in 1975-79. Thus, the decline in productivity growth was almost universal across subsectors. 2/ One subsector, nonferrous metal basic industries, was excluded for lack of comparable date over time. - 139 - Appendix 4 Page 4 of 4 It is possible to separate systematically the effects of specific subsector performance from the effects of shifting resources towards less efficient sectors (Table A3.2). The overall decline in manufacturing TFP growth is entirely attributable to a deteriorating performance within indus- tries. After posting sizable gains in the late 1970s, the sector-specific TFP growth trend was reversed in the 1980s. Over the decade 1975-84, each indus- try saw efficiency decline on average by 2.52 per year. Although the decline was widespread, in quantitative terms the problems in manufacturing are con- centrated in a few specific subsectors. In terms of gross output, three of the largest subsectors are industrial chemicals, petroleum refineries and iron and steel. These are also among the most capital-intensive sectors in manu- facturing. Each of these three had very substantial declines in productivity in the early 1980s, associated in each case with a rapid expansion in capacity that far outstripped actual production gains. If these subsectors are excluded, the remainder of the manufacturing sector would have only averaged a within-sector 21 loss in productivity over 1981-84. Table A3.2: A DECOMPOSITION OF MANUFACTURING TFP GROWTH (1) 1975-79 1981-84 1975-84 TFP growth TFP growth TFP growth All manufacturing 3.8 -1.9 -0.2 Within industries 6.7 -7.2 -2.5 Between industries -2.9 5.3 2.3 Major subsectors Industrial chemical 2.8 n.a. n.a. Petroleum refineries 4.2 -11.4 -2.6 Iron and steel 3.8 -21.0 n.a. n.a. TFP change in greater than 100l per year. Source: Bank staff estimates. - 140 - Appendiz 5 Page 1 of 3 Methodology underlying the calculation of the marginal effective tax rate The calculations presented in this Appendiz use tax, financing, and economic factors to compute the cash flows that can be ezpected from an investment project. The cash flows combine the individual streams for each asset (credits, depreciation, replacement investment) with the income generated by the project and other project-wide flows (e.g., debt service payments) to yield a single before-tax cash flow and single after-tax cash flow for the entire project. These two cash flows are used to compute NETR for the project as a whole. The before-tax (BTCF) and after-tax (ATCF) cash flows can be suims rized as follows: (1) BTCF.i - - Ei dki + Ri- Inti - Prini + NetSalesi (2) ATCF. = BTCF. - t(l+s) [Ri - Depi - InvDed. - IntDed. - Carryover i CapGain + BalAdjij + cDiv where Ei a amount of equity used to finance the investment in period i d economic depreciation rate for capital stock K = capital stock in period i Ri = investment income, net of operating costs, in period i Inti - interest payment in period i Prini = principal payment in period i NetSalesi = sales proceeds in period i t = statutory tax rate s = surtax rated Depi = depreciation allowances in period i InvDedi = investment deductions in period i IntDedi = interest deductions in period i Carryoveri = carryover losses in period i CapGaini = capital gains in period i - 141 - Appendix 5 -141- ~~~~~Page 2 of 3 BalAdji = balancing adjustment in period i c - dividend credit rate Divi = dividends in period i The equity used to finance the investment is a positive amount in the period before the investment generates income, i.e., Eo > 0, and is zero afterwards. The rate of return underlying the METR calculation is a return to equity. Replacement investment is undertaken which reduces the before-taz cash flow. Repia cement investment equals the amount of economic depreciation for each asset.i' The investment project starts generating income in period 1. Investment income is the,marginal increase in net revenue that results from the investment project., Investment income keeps pace with inflation because replacement investment preserves the productive capacity of the project. Interest and principal payments on debt used to finance the invest- ment are subtracted from the before-tax cash flow. When the investment is bOLuP Labe prtrcede &K are ucd Lu ; vesvte-; CSMh cas 1 f'v". N.Łts-A oe equal the sale price minus the payoff of the balance of a loan, if applicable. The after-tax cash flow equals the before-tax cash flow minus taxes paid and plus credits. The statutory tax rate, t, plus any surtax rate, s, are multiplied by taxable income to yield the regular tax liability. Taxable income is given by the term in brackets in equation (2). Taxable income, in its most basic form, equals investment income minus depreciation allowances, investment deductions, and interest deductions. Investment deductions are given in addition to depreciation and are intended to serve as an incentive Zor investment. They may be given for the project as a whole or on an asset- specific basis. A positive taxable income may be reduced by losses being carried forward. If taxable income is negative and full loss offset is not assumed then the loss is carried forward. When the asset is sold, any taxable capital gains or balancing adjustment are added to taxable income. 1/ As the investment project includes three asset3--buildings, machinery and equipment and vehicles--then dK represents the sum of the depreciation rates for each asset times their respective capital stocks. 2/ "Net" is taken to mean net of wages and purchases of intermediate goods and services. - 142 - Appendix 5 Page 3 of 3 The last term in equation (2) refers to any special tax treatment for dividends. In parcicular, a credit may be given on dividends at the personal level to offset the taxation of dividends at the corporate level. This credit reduces effective corporate taxes and, therefore, is added to the after-cash flow in equation (2). If b denotes the real before-tax rate of return and a, the real after-tax of return, then the margin effective tax rate is given as follows: (3) METR - (b - a)/b This formula has been 3jsed extenuively in the literature on the taxation of capital investment.3 Note on Assumed Project The calculations are based on a real interest rate of 101, and a project which also yields 10. Inflation is assumed to be 3Z. The initial investment is comprised of 13.62 land, 32.651 buildings, 48.91 machinery and equipment and 4.852 vehicles. These parameters are taken from the share of each in total fixed-assets in manufacturing reported in the Survey of Manufactures. 3/ A. Auerbach, "Taxation, Corporate Financial Policy and the Cost of Capital," Journal of Economic Literature (September 1983), for a review. - 143 - Appendiz 6 Page 1 of 2 The Determinants of Private Fixed Capital Formation Regression analysis has been used to analyze the determinants of private domestic fixed e;apital formation in Malaysia over the period 1972- 87. The results of the regression, which account for 99% of the variance in the level of investment, are presented in Table A.1 below. Private domestic investment is found to depend strongly and positively on profits, government infrastructure investment, the real effective exchange, and an accelerator term measured as the lagged change in GNP. On the other hand, foreign direct investment and government productive investment affects private domestic investment negatively. Table A.l: REGRESSION COEFFICIMNTS FOR PRIVATE INVESTMENT (Dependent variable is real private domestic investment) Variable Coefficient T-statistic Profits 0.344 14.5 R2 0.986 Government infrastructure investment 0.253 6.3 D.W. = 2.101 Real effective exchange rate 18.363 2.6 Sample - 1972-87 Accelerator 0.087 3.5 Foreign Direct Investment -0.351 -2.5 Government Productive investment -0.009 -0.0 All the variables were first calculated in nominal terms, and then deflated by the consumer price index. The regression was also run in first- difference and percentage-change terms with very similar results to those reported above. This robustness to specification changes gives added confi- dence in the results. Other explanatory variables were also introduced but found to be insignificant. The most important of these, from a theoretical standpoint, were the real average lending rate of commercial banks and the real supply of domestic credit. Both of these have been found to be signi- ficant determinants of investment in other countries. Construction of the data Host of the data were constructed from officially published data sources. Private domestic investment is the residual of total capital forma- tion less public fixed investment less foreign direct investment. Public in- vestment is calculated on the basis of development expenditure. The series is published in the Economic Report of the Treasury, but contains some inaccura- cies. First, it is derived from budgetary cash accounts not from accrual accounts. Second, it includes expenditures on land and used equipment pur- chases which are not elements of gross fixed capital formation in the national income accounts. Third, the coverage of public enterprises is incomplete and - 144 - Appendix 6 Page 2 of 2 changes over time. One major caveat is that oil and gas investments are not included in public investment. Public investment is further broken down into two components, one, labelled infrastructure investment, includes investment in sectors such as transport, utilities, land development, education and health. The other component, labelled productive investment, reflects those investment which directly output destined for final consumption such as com- merce and industry, agriculture and housing. Foreign direct investment is subtracted from private investment be- cause it mostly enters free trade zones and is oriented towards export mar- kets. The series is obtained from the balance of payments and includes the retained earnings of foreign corporations. Profits are constructed as the residual of GNP less the wage bill less net indirect taxes. The wage bill is calculated by multiplying sectoral wage data, available from the Ministry of Labor, by total employment in each sector. Gross import taxes are taken as a proxy for net indirect taxes; a direct estimate is not available because of lack of data on implicit subsidies. The accelerator term is equal to (CNP_. - CNP_2), where the subscripts refer to time periods. The real effective exchange rate is a trade-weighted average of Malaysia's bilateral real exchange rates with six major trading partners, where the real exchange rate is the nominal exchange rate deflaLud by relacive price movements in each country's wholesale price index. - 145 - Appendix 7 Page 1 of 2 Private Consumption in Malaysia Private consumption growth in Malaysia has not followed national income growth closely throughout the 1980s. Yet this does not imply that private consumption has been haphazard. It suggests rather that factors other than income have been important. The analysis below indicates that five major factors account for the divergence between consumption and national income. First, private disposable income because of the varied Government tax bite, the fluctuating performance of public enterprises and the programs instituted by government to shield the private sector from terms of trade losses. Second, consumers base their spending decisions on some notion of long-term sustainable income, and not current income alone. Third, a growing fraction of private sector income has been absorbed into forced savings accounts in Provident Funds and Social Security organization and is not available for consunption. Fourth, private consumption has been sensitive to the real effective exchange rate, which meaunres Malaysians' wealth holdings in terms of an international standard. Fifth, private consumption responds to the real interest rate. Estimation of a consumption function in Malaysia is complicated by the Łact that there does not exist any primary data on private disposable income. Some analysts have used GNP as a proxy. An alternative is to construct a series on private income by making use of the national income accounts identities. First, note that private income must equal private consumption (PC) plus private savings. Private savings equal national savings less public savings. National savings equal total investment less net foreign savings (the current account deficit (G.AD)). Public savings equal the consolidated public sector deficit (G) plus public investment. Using these definitions, private disposable income (PY) is equated to: PY = PC + I + AS - CAD + G p where I is private fixed capital formation and AS is the change in stocks; and botR G and CAD are positive when there is a deficit. Private income is then further broken down into a discretionary por- tion (PDY) and a non-discretionary portion by subtracting the investable sur- plus of the Employee Provident Fund each year (DEPF). Discretionary income is converted into a measure of permanent income (PPY) by using a weighted average of PDY over three years, such that PPYt = 0.37 * PDYt + 0.33 PDYt-l + 0.3 PDYt_2. The weights are taken from Bhalla's derivation of weighting schemes corre- sponding to a 102 discount rate. The real effective exchange rate (REER) is a trade-weighted average of Malaysia's bilateral real exchange rates with six major trading partners, where the real exchange rate is the nominal exchange rate deflated by relative price movements in each country's wholesale price index. Finally, a real interest rate variable is included (RRD), defined as the three-month commercial bank fixed deposit rate less the percentage increase in the CPI. Regression results are shown in Table A3.1 below. All nominal variables are deflated by the CPI, the sample period covers 1973-87. Esti- - 146 - Appendix 7 Page 2 of 2 mation is done using two-stage least squares because the method for computing private income implies a correlation between income and the error term in consumption. This proves to be an important qualification because some variables which were significant using OLS estimation techniques, such as the terms of trade index, close their significance once the TSLs procedure is used. Instruments used for income are the logs of real exports, real invest- ment and real money supply (M2). Other variables were also initially in- cluded, but found to be insignificant. These were money supply (a proxy for wealth) and transitory discretionary income, defined as current income less permanent income. Table A3.1: CONSUMPTION FUNCTION REGRESSION RESULTS Sample 1973 - 1987; dependent variable is log of real private consumption. Variable Coefficient T-Stat Constant 2.7503193 4.6627620 Private permanent income 0.7716098 9.8732375 Change in EPF assets 0.0940418 2.5533511 Real effective exchange rate -0.2917560 -4.3690617 Real deposit rate -0.0043107 -2.4060107 R-squared 0.995367 Mean of dependent var 10.07587 Adjusted R-squared 0.993513 . S.D. of dependent var 0.261491 S.E. of regression 0.021060 Sum of squred resid 0.004435 Durbin-Watson stat 1.708910 F-statistic 537.0771 Log likelihood 39.66240 IBRD 1937L MALAYSIA PINES < H ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~PHILIPPINES THAI LAND Rood, / Stu ,Sea Railways Airports Rivets Mountainous Areas '. -State Boundaries H3 Internalitionl Boundaries ~~* 0 ~~00 200 300 400 - s I N D O N E S I A <& @t + a . v t A z / M A ~~ t A ~ x, / P _0' \ l 1011 - 112- 116- KItOMETERS P A KLA .nqm BRUNEI -ABAH 4fl Sop10- INDONESSA ~~~~~~V ~~~~~( A RA A K too nT. WoPTa.*T M d.-P*.U* ~~~-' ~~~~~ - -/--- 's~~~~~~~~~~~.to,dIP r. dTO. Ut INDONESIA - I N DO0N ESI A AUGUST I