Philippines - Country economic report : public sector resource mobilization and expenditure management

Ot-gO.% 'g -w;!,:,-` kN4 ON p. -n -,7M? .-FO CM 7B CD c
,, 51: at-,, tD ao
1,, PEe K
Pw -Q
VU to C+ CD C, oi rri td '' --ft t
-3 co Z O Pn to ED DO  CURRENCY EQUIVALENTS Average 1991 - US$1.0 = 26.7 Average 1990 - US$1.0 - 24.3 Average 1989 - US$1.0 = 21.7 Average 1988 - US$1.0 = 21.1 Average 1987 - US$1.0 = 20.6 Average 1986 - US$1.0 - 20.4 ACRONYMS AND ABBREVIATIONS APT = Asset Privatization Trust BIR Bureau of Internal Revenue BOC Bureau of Customs BOT - Build-Operate-Transfer DBM = Department of Budget and Management DBP Development Bank of the Philippines DOF - Department of Finance EO = Executive Order EPZA - Export Processing Zone Authority ETR 5 Effective Tax Rate GC = Government-Owned or Controlled Corporation GCMCC = Government Corporate Monitoring and Coordinating Committee GSIS Government Service Insurance System GRT = Gross Receipts Tax ICC = Investments Coordinating Committee LRTA Light Rail Transit Authority LWUA Local Water Utilities Administratior MTFP Medium Term Financial Plan MTPIP - Medium Term Public Investment Program MWSS Metropolitan Waterworks and Sewerage System NEDA = National Economic and Development Authority NEA - National Electrification Administration NFA = National Food Authority NHA National Housing Administration NIA = National Irrigation Administration NPC National Power Company NTRC x National Tax Research Center OIC = Omnibus Investment Code OPSP = Oil Price Stabilization Fund PCE = Personal Consumer Expenditure PNB - Philippine National Bank PNOC Philippine National Oil Company PNR = Philippine National Railways PPA - Philippine Ports Authority SPPBS = Synchronized Planning Programming and Budget System TIN - Tax Identification Number VAT = Value Added Tax  FOR OFFICIAL USE ONLY PHILIPPINES COUNTRY ECONOMIC REPORT PUBLIC SECTOR RESOURCE MOBILIZATION AND EXPENDITURE MANAGEMENT Table of Contents EXECUTIVESUMMRY-2* f . o . . . . . . . . . . . . . . . I. THE MACROECONOMIC CONTEXT . . . . . . . . . . . . . . . . . . . . . 1 A. Growth Experience since 1975 .... . . . . .. . ..... . 1 B. Savings and Investment Behavior . . . . . . . . . . . . . . . . 9 C. Implications for Development Strategy . . . . . . 22 II. PUBLIC RESOURCE MOBILIZATION . . . . . ............ . 24 A. Overview of Public Revenue System . . . . . . . . . . . . . . . 24 B. Tax Intake and Tax Potential . . . . . . . . . . . . . . . . . 33 C. Issues in Internal Revenue Collection . ..... . . . . . . . 38 D. Internal Revenue Administration . . . ............ . 44 E. Customs Administration . . . . . . . . . . . . . . . . . . . . 30 F. Non-Tax Revenue Issues . . . . . . . . . . . . . . . . . . . . 55 G. Enhancing Revenue Mobilization . . . . . . . . . . . . . . . . 56 7II. PUBLIC EXPENDITURE MANAGEMENT . . . . . . . . . . . . . . . . . . . 65 A. Budget Process Issu,s . . . . . . . . . . . . . . 65 B. Public Expenditure Patterns and Productivity . . . . . . . . . 69 C. Private Sector Role in Public Sector Projects . . . . . . . . . 79 D. Government Corporations: Financial and Management Issues . . . 83 SelectedBibliography .... . 99 StatisticalAppendix .... .. . . . . . . . . . ....... . . . . 101 This report was prepared by a team consisting of Farrukh Iqbal (Task Manager), Jeffrey Balkind, Daniela Gressani and Delfin Go. It is based in part on background reports prepared by A. Premchand, M. Woolley (IMF staff) and R. Manasan, M. Lamberte and E. de la Cruz (consultants). The help of the National Tax Research Center is gratefully acknowledged. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents rnay not otherwise be disclosed without World Bank authorization.  LIST OF TABLES, FIGURES & BOXES CHAPTER I TABLES Page 1.1 Key Macroeconomic Indicators . . . . . . . . . . . . . . . . . . . 2 1.2 External Finance Flows: 1970-89 . . . . . . . . . . . . . . . . . 3 1.3 External Environment: 1989-92 . . . . . . . . . . . . . . . . . . 5 1.4 Savings and Growth Rates . . . . . . . . . . . . . . . . . . . . 11 1.5 Comparative Savings and Growth Trends . . . . . . . . . . . . . . . 12 1.6 Selected Investment Data . . . . . . . . . . . . . . . . . . . . . 19 1.7 External Debt Outstanding, by Sector . . . . . . . . . . . 21 1.8 Debt and Taxes . . . . . . . . . . . . . . . . . . . . . . . . . . 22 FIGURES 1.1 Real GDP Growth Rate . 2 .. . . 2 1.2 Ratio of Public Sector Deficit to GNP . . . . . . 7 1.3 Ratio of Current Account Balance to GNP . . . . . . . . . . . . . . 7 1.4 National Savings & Investment Rates . . . . . . . . . . . . . . . . 11 1.5 Public Sector Savings . . . . . . . . . . . . . . . . . . . . . . . 15 1.6 Public sector Investments . . . . . . . . . . . . . . . . . . . . . 15 1.7 Personal Savings and Real GDP . . . . . . . . . . . . . . . . . . 16 1.8 Net Foreign Investment Trends . . . . . . . . . . . . . . . . . . . 20 CHAPTER II TABLES Page 2.1 Structure of National Government Revenues . . . . . . . . . . . . . 26 2.2 Revenue Importance of VAT . . . . . . . . . . . . . . . . . . . . . 29 2.3 Effective Tariff Rates: 1980-90 . . . ............. . 30 2.4 Buoyancy and Elasticity of Major Taxes . . . . . . . . . . . . . . 34 2.5 Tax Ratios in Comparative Perspective . . . . . . . . . . . . . . . 34 2.6 Potential Versus Actual Income Tax Collections . . . . . . . . . . 37 2.7 The Potential VAT Base in 1990 . . . . . . . . . . . . . . . . . . 37 2.8 Potential Versus Actual Number of Tax Filers . . . . . . . . . . . 39 2.9 Revenue Foregone from BOI Incentives . . . . . . . . . . . . . . . 42 2.10 Revenue Foregone from All Exemptions (1989) . . . . . . . . . . . . 43 2.11 BIR Accounts Receivable Collection Performance . . . . . . . . . . 46 2.12 Enforcement Through Warrant and Distraint Levies . . . . . . . . . 47 2.13 Workloads of Assessment Personnel . . . . . . . . . . . . . . . . . 49 2.14 Effectiveness of Tax Fraud Investigation . . . . . . . . . . . . . 50 2.15 Data on Seizure of Smuggled Goods . . . . . . . . . . . . . . . . . 51 2.16 Budget of the Customs Poline Administration . . . . . . . . . . . . 52 2.17 Import Inspection Performa.sce . . . . . . . . . . . . . . . . . . . 54 2.18 Data on Misdeclared Shipments . . . . . . . . . . . . . . . . . . . 54 2.19 Estimated Revenue Impact of Selected Measures . . . . . . . . . . . 63 2.20 Revenue Mobilization and other Objectives . . . . . . . . . . . . . 64  FIGURES 2.1 Structure of National Government Revenues . . . . . . . . . . . . . 27 2.2 Trends in Direct and Indirect Tax Shares . . . . . . .. . . . . . 27 2.3 Direct Taxes and Per Capita Income, 1989 . . . . . . . . . . . 33 2.4 Tax Ratios, 1976-1990 . . . . . . . . . . . . . . . . . . . . . . . 35 2.5 Gov't Revenue and Per Capita Income 1989 . . . . . . . . . . . . . 35 BOXES Page 2.1 Tax Reform Package of 1986 . . . . . . . . . . . . . . . . . . . . 25 2.2 Tax Measures Considered in 1991 . . . . . . . . . . . . . . . . . . 57 CHAPTER III TABLES Page 3.1 National Govt. Expenditures, by Economic Classification, obligation Basis, (Percent of GNP) . . . . . . . . . . . 70 3.2 National Govt. Expenditures, by Sectoral Classification, obligation Basis, (Percent of GNP) . . . . . . . . . . . . . . . 70 3.3 Status of Private Power Projects . . . . . . I . . . . . . . . . . 84 3.4 The Major Subsidy Recipients . . . . . . . . . . . . . . . . . . . 89 3.5 The Major Equity Recipients . . . . . . . . . . . . . . . . . . . . 90 3.6 Results of Perfurmance Evaluation . . . . . . . . . . . . . . . . . 93 3.7 Collection Efficiency . ... . . ........ .94 FIGURES 3.1 Govt. Expenditures, by Economic Classification . . . . . . . . . . 72 3.2 Govt. Expenditures, by Sectoral Classification . . . . . . . . . . 72 3.3 Public Investments by Source (%GNP) . . . . ... 75 3.4 Public Investments by Source (million pesos) ...... . 75 3.5 Public Investments by Sector (Share of Total) . . . . . . . . . . . 76 3.6 Public Investments (Percent of GNP) ... ..... 76 3.7 Flow of Funds to GCs . . . . . . . . . . . . . . . . . . . . . . . 87 3.8 Govt Transfers to GCs . . . . . . . . . . . . . . . . . . . . . . . 88 3.9 GC Transfers to Government . . . . . . . . . . . . . . . . . . . . 88 3.10 Financial Performance of GCs . . . . . 96 BOXES 3.1 List of Monitored Corporations and Respective Mandates . . . . . . 86 3.2 Financial Crisis and Reform of the National Power Corporation... . . . .... 98  EXECUTIVE SUMMARY 1. This report reviews recent macroeconomic trends and developments and suggests that medium-tern development strategy in the Philippines should focus on fiscal adjustment, emphasizing non-de t sources of public finance on the one hand and judicious management of public expenditures on the other. Developing a non-debt base of public finance will require greater policy attention to domestic resource mobilization. Improving the management of public expenditures will require better cost control and more realistic user charge policies for public services, as well as increasing the scope for private sector participation in public sector projects. 2. The report endorses the general direction of the government's current economic stabilization program which gives high priority to fiscal adjustment. As far as the medium-term adjustment program is concerned, it urges that strong consideration be given to the following three points. First, between revenue mobilization and expenditure reduction, Government should prefer the former since the latter is likely to be much more damaging to both short and long term growth. Second, Government should act on a broad front as far as revenue mobilization is concerned: at present, it does not have the luxury of choosing between expanding the base on which taxes are collected or improving the process of tax collection. Third, as far as public expenditure management is concerned, Government should pay special attention to improving the physical and financial performance of public enterprises. A. Background: The Adjustment Problem 3. The crux of the adjustment problem in the Philipines lies in certain structural weaknesses of the external sector on the one hand and the domestic resource mobilization system on the other. The external sector is characterized by high import dependence and a relatively slowly growing and narrowly based export structure. These aspects have condemned the Philippines to a history of recurrent balance of payments crises. Phases of economic recovery typically result in high import growth but relatively lower export growth, thus leading inexorably to payments problems. 4. It has proved difficult to resolve external payments crises by recourse to the domestic resource mobilizatioii system since this has structural rigidities of its own. This system is characterized by low rates of public and private savings. On the public side, the tax system is unable to generate the volume of resources needed to finance the desired level of expenditures partly because of deficiencies in design and structure and partly because of problems of tax administration and collection. On the private side, savings are low partly because of income and demographic developments, partly because of the impact of foreign savings and partly because of shortcomings in the domestic financial system. 5. The scope for managing balance of payments crises has been narrowed in recent years by certain developments pertaining to external and domestic  financing ability. On the external side, the availability of long term commercial credit has been significantly curtailed since the onset of the debt crisis in the early 1980s. Wnile official flows have increased somewhat, this has not been enough to compensate for the decline in commercial flows; the Philippines has experienced negative resource transfers in each of the last seven years. On the domestic side, Government has resorted increasingly to direct borrowing. Excessive reliance on domestic debL issuance, however, is not desirable because of the potential for driving interest rates up to damaging levels. The domestic capital market is saturated with Government paper and has recently had difficulty in smoothly absorbing large amounts of domestic public debt. 6. The above points can be illustrated by referring to macroeconomic developments during the 1980s. Several consecutive years of high growth financed by high external borrowing during 1975-82 led to a balance of payments problem in 1983-84. With the onset of a global debt crisis in 1982, further recourse to external borrowing was sharply curtailed. The subsequent adjustment process featured a deep recession over three years (1983-85) during which GNP fell at an average rate of 3.4%. Investment plunged from 272 of GNP in 1983 to around 14% in 1985. 7. The economy began to recover in 1986 and grew steadily over the next three years. Unfortunately, import growth was much stronger than export growth during this period (about 28% per annum compared to around 12% for exports) and a balance of payments crisis loomed again in 1990. Domestic policy slippage contributed to the gathering crisis: the real exchange rate was allowed to appreciate, monetary control was relaxed, and fiscal deficit targets were exceeded. With external financing still constrained, external debt servicing still relatively large and domestic revenue performance still inadequate (although considerably improved), Government resorted to domestic borrowing to finance growth. This, however, pushed up interest rates to very high levels, thus feeding the fiscal imbalance and propping up the exchange rate. A series of economic shocks during the year (drought, earthquake and Gulf conflict) made macroeconomic management even more difficult. So did the worsening external environment which saw the industrial countries slipping into slow growth and world trade volumes declining. 8. The Government took corrective policy actions and formulated a new stabilization program at the end of 1990. Progress has been made in 1991 towards achieving its goals. Fiscal imbalances have improved substantially and monetary growth has been gradually brought under control; as a result, inflation has decelerated to about 12 percent by end-1991, the current account deficit has been reduced, and international reserves of the Central Bank have recovered substantially. The cost of tight fiscal and monetary policy, however, has been the slowdown of growth, which is estimated below 1 percent in 1991. The Tasks Ahead 9. In view of the foregoing, the tasks for both short-term and medium-term macroeconomic management are clear. The short-term task is to continue with the stabilization program since the availability of development financing, from both external and domestic sources, depends on a stable macroeconomic environment.  - iii - The medium-term tesk is to develop further an adjustment program which will help shake loose the structural constraints that have hobbled macroeconomic management and hindered the achievement of sustained growth in the past. 10. Fiscal adjustment is central to both the stabilization and adjustment obiectives. A key target of the ongoing stabilization program is the reduction of the fiscal deficit to 2.72 in 1992. The man.aer in which this is achieved will be critical to structural adjustment. If it is achieved largely by cutbacks in capital outlays and without improvements in revenue mobilization, the country will achieve only temporary macroeconomic stability; political pressures and infrastructure needs will revive expenditure growth and lead inexorably to a renewed fiscal crisis. The key to successful and sustainable medium-term fiscal adjustment lies in (a) the enhancement of domestic revenues and (b) the careful management of public expenditures so as to maintain growth momentum while simultaneously respecting fiscal constraints. 11. An adu'itional important consideration is the degree to which the private sector can be harnessed in the pursuit of adjustment. This involves improving the environment for private savings and investment efforts. While this is not explicitly considered in this report, its importance should not be under- emphasized. Private savings and investment are large components of aggregate savings and investment and if they cannot be harnessed in support of the development effort, sustained high growth will remain out of reach. To accomplish this, it is important to maintain a credible macroeconomic policy regime, one that is consistent with stability objectives (especially fiscal adjustment) and one that is not subject to frequent reversal on political grounds. A credible macroeconomic policy regime can catalyze and sustain growth in private savings and investment. Private savings will rise to the extent that the stability of policy and outcomes preserves capital value and discourages capital flight. Private investment will rise as macro stability brings real interest rate decline in its wake and provides a less volatile planning environment to the investor. 12. The stress on fiscal adjustment in this report should not be taken to mean that measures to relax. the external constraint directly are not important. They are very important. Measures to promote market oriented exchange rates, low and uniform tariffs, minimal import restrictions, and institutional support for export-oriented activities are a sine qua non for achieving high and sustainable growth. A number of key measures in these areas have been taken in 1991-92, but need to be sustained to lay the foundation for long-term growth. Moreover. despite the pressures of an election year, the Executive and Legislature have cooperated in passing several important revenue measures, a promising test of policy resolve which needs to be maintained. B. Enhancing Domestic Revenues 13. The key question with respect to the public revenue system is whether or not it generates sufficient funds to finance the desired development effort, given certain constraints on government's ability to borrow. In the case of the Philippines, despite considerable improvement in recent years (the tax ratio rose  -v - from 10.3X during 1981-85 to 12.42 during 1986-1990), the system falls short in three respects. First, it generates less than what is required to fund development expenditures; this is shown by the fact that actual taxes collected were lower than the amounts targeted in each of the last five years. Second, Philippine tax effort lags that of its ASEAN neighbors and is less than what its relative development stage would lead one to expect. Third, the system generates less revenue than would be expected from considerations of national income and tax parameters; mn other words, actual tax collections are far less than potential tax intake. 14. Tax Collections versus Tax Potential. Calculations performed in this report indicate that low collection relative to potential is a feature of virtually every major tax. For example, it is calculated that in 1988 only 532 and 65Z of potential individual and corporate income taxes respectively were captured by the revenue system. VAT collections were short of potential by 50% in 1990. Significant tax evasion is reported for excise taxes also, although current estimates are not available. Finally, it is reported that failure to prevent smuggling robs the national treasury of at least around 20% of potential collections from import duties. These estimates of tax potential are necessarily imprecise but they provide a meaningful indication of the scale of the problem. Revenue Collection: Structural Issues 15. Base Erosion through Exemptions. A considerable amount of revenue is foregone through exemptions awarded to government and private corporations. While some of these exemptions are justifiable (e.g. tax and duty drawbacks as a tool to promote exports) others may impose a fiscal cost well in excess of actual benefits. For example, the evidence suggests that fiscal incentives awarded through the Board of Investments are not particularly effective in inducing "incremental" investments and succeed only in providing a windfall gain to those who would have invested anyway because of other business considerations; in the meanwhile, such incentives accounted for a potential revenue loss of around 2.82 in 1989. Given current fiscal needs it is recommended that the incentive system, especially that covering private investment, be redesigned to reduce the number of eligible sectors or industries. Government has recently reduced the number of industries listed in the Investment Priority Program; this is an encouraging initiative and should be intensified. 16. Base Erosion through Nonfiling of Tax Returns. Far fewer individuals and corporations file tax returns than one might expect from general considerations of income and employment growth and company formation in the country. For example, only about 28? of the potential number of individual income tax filers and 29Z of the potential number of corporate income tax filers actually filed returns in 1989. Even if these ratios are doubled, to account for possible liberalism in our calculations, the scope of the non-filing problem is still seen to be very large. It is admittedly very difficult to address the non-filing problem in an economy characterized by poor record-keeping and legal and fiscal impediments to discovery and prosecution. Nevertheless, a start can be made by computerizing the tax monitoring system to catch drop-outs with greater ease. 17. Base Erosion through VAT Exemptions. When the VAT system was introduced in 1988, several categories of goods and services were exempted and a single rate  of OZ was adopted; This too has contributed to reducing the size of the base on which taxes ~re raised in the Philippines. While a n4rrow base may have been sensible initially, in view of anticipated difficulties in introducing a new system, enough experience has been ined by now to justify the expansion of the VAT base. Among the categories of services that could be included in the VAT base are: hotels, caterers, common carriers, forest products, securities dealers, public utilities, communications, and entertainment. It should be noted that revenues would not necessarily be increased in all cases if the VAT rate of 10% is applieds for example, hotels currently charge a higher rate and there is no benefit of input crediting. Nevertheless, a widening of the base would be justified on rounds of tax neutralit' and easing of administration. 18. Underdeclaration of Incomes. Many of those who file do not report their taxable incomes accurately. This is particularly true of self.employed professionals and small businesses. The scope of the problem may be gauged from the results of a study ir which the presumptive income levels of doctors in Manila were calculated from a survey of case loads, consulting hours and fees: it was foand that thege were from four to fifteen times higher than the average income reported by doctors in the tax files of the Bureau of Internal Revenue (BIR). In part, such underdeclaration is possible because of the 'egal inability of the BIR to detect cases of evasion through the financial records of tax filers. The failure to catch such evasion, especially by the wealthy, involves a large direct revenue loss, of course, but also has a corrosive effect on public ethics with regard to tax obligations. 19. Given the potential revenue payoff of more accurate declarations, it is recommended that a presumptive income level system be introduced for certain categories of hard-to-tax groups. It is clear that several legal and administrative challenges will have to be surmounted before such a system can be widely implemented. Concerned groups will lobby strongly against such a system and initial assessments will be widely contested. On the administrative side, there is the difficulty of making sufficiently precise estimates and then updating them regularly. Neverth.eless, several countries (e.g. France, Turkey, Japan, Israel, Thailand) have adopte4 this system in one form or another; their experience suggests that the system is productive and that various anticipated difficulties can be dealt with satisfactorily. For example, to obtain the cooperation of target taxpayer groups, their professional associations could be involved as partners in the process of estimating presumptive income levels and setting procedures for revision and arbitration. 20. It is also recommended that the Bank Secrecy Law be amended to facilitate the accquisition of financial records by the BIR. The amendments would have to be made in such a way that a balance is struck between the authority's legitimate need to know and the individual depositor's need for protection from abuse. To develop such amendments, guidance could be sought from the tax and bank laws of other countries. Two considerations should be kept in mind. First, given present levels of evasion, depositors should not be protected to a higher degree than is normal for neighboring countries. Second, some concerns are likely to be exaggerated. For example, relaxation of bank secrecy laws is unlikely to create great financial disruption. There usually is no alternative for the "big" tax evader to using banks since informal institutions are unlikely to provide an adequate degree of security and convenience. Concerns about runs on banks can  'vi. - ( also be met through othtir means, such as by providing liquidity to affected banks tht"Ough Cenqral Bankc mechanisms.. 2iL. Overstatement of Deductions. There are several provisions in the tax code that make `it easy for taxpayers to overstate allowqable deductions. For example, sp(rises often file separate returns with each claiming full deductions for children, effectively doubling the value of the deduction. Deductions for certain business expenses, sv;ch as representation, promotion, and travel, are widely abused. The problem here lies not in the allowance of deductions per se but in the fact that they are not subject to reasonable ceilings. In this regard, recent legislation (viz. the simplified net income tax scheme) to impose iore rational ceilings on business deductions is to be welcomed. Revenue Collection: Administration Issues 22. Internal revenue administrati^n has improved in recent years as evidenced by improvements in tax effort as well as in the cost of collection which declined from 1.14 pesos for every 100 pesos collected in 1981-85 to an average of 0.83 pesos in 1986r9O. Nevertheless, there is considerable room for improvement. The key problem areas relate to collection, audit and tax fraud investigation. 23. Problems in Collection. Collection is characterized by a rather large overhang of accounts receivable. The number of such accounts amounted to almost 14% of the total number of income tax filers during 1986-90 while their value amounted to almost 19% of total BIR collections. The severity of this problem can be related to such causes as an inadequate taxpayer identification system, inadequate computerization of records, and weak enforcement follow-up. The first two causes make it difficult to generate an accurete master list of taxpayers, monitor this list for drop-outs and delinquents, and update the record of ,payments and adjustments. Manual tracking of accounts inevitably invites 'disruption and delay, whether accidental or by design. Weak enforcement follow- up encourages fraud since the threat of penalty is effectively reduced. Lack of personnel is typically cited as the chitf cause of weak enforcement. Often, however, the enforcement process cannot even be started because the defaulting taxpayer cannot be located, a problem related both to a poor monitoring system and lack of resources to investigate delinquencies. 24. The collection problem is being addressed by a change-over from the present system of multiple account numbers for different tax transactions to a system featuring a single tax identification number (TIN). This would help enhance data processing and reduce tax evasion. Data processing would be facilitated since fewer files would need to be opened, payments made could be easily recorded, and adjustments easily effected. Tax evasion would be discouraged to the extent that different transactions with BIR could be easily retrieved for cross-checking purposes. In addition, if the TIN were designed to identify spouses, this would help cut down on evasion through the practice of both spouses claiming full deductions for the same set of dependents. It should be noted that the full value of a TIN system can only be derived if tax accounts are properly computerized. In addition, computerization would enable improved monitoring of tax account status and improved interaction with other government agencies that have a revenue function.  -Vii - 25. Problems in Audit. The quality of assessment and audit suffers from the lack of appropriate case screening criteria. At present, no attempt is made to separate out cases which are likely to generate high post-audit adjustments from those whlich are unlikely to do so. Ae a result, workloads are heavy relative to available personnel and much of the assesement work consists of perfunctory desk review. Experience shows that certain types of returns rarely involve fraud and misdeclaration: for example, this is the case for returns filed by individuals who derive income purely from compensation and whose taxes are withheld at source. Legislation has recently been pcssed making it possible to consider such withheld taxes final in nature. As 8t th, they need no longer be part of the auditable pool, a development that it --pected to reduce BIR's workload and improve audit quality for the remaininL ,ases. It is recommended that case screening criteria continue to be developed with ar. eye to the economics of tax administration. 26. Problems in Tax rraud Investigatior. Tax fraud investigation in the Philippines is characterized by slowness of pace and ineffectiveness of results. These characteristics are related in turn to a number of personnel motivation problesa and to weak penalties In the law for tax fraud. Personnel motivation problems arise from low pay and heavy workloads. The BIR has found it difficult to hire competent tax law"yers at its present pay scales as most such lawyers have a more lucrative alternative in -rivate practice. The number of legal personnel in BIR has declined from 142 in 1983 to only 88 in 1990; as a consequence, there has been a sharp drop in the filing of civil and criminal tax cases. 27. Examiners are generally reluctant to get involved in tax fraud investigations because the process is a long drawn-out one with little prospect of a satisfactory outcome. Sometimes, such investigators are harassed and threatened with lawsuits; in such cases they are personally liable for expenses and penalties if they lose the case. Furthermore, courts do not give high priority to tax cases. And for cases that are heard and decided, the penalties typJ'ially awarded are considered unlikely to have a significant deterrent effect; civil penalties are light while criminal penalties are rarely imposed. In recent years, the number of cases filed has been less than 0.1Z of the total number of returns. 28. Recently, 20 courts in selected major urban areas were asked to handle tax cases on a priority basis; however, since these courts continued to handle non-tax cases as well, they did not significantly ease the backlog of tax cases in the jurisdictions concerned. Legislation has recentlv been passed to set up a system of dedicated, specialized tax courts which would only hear tax cases. This is a step in the right direction. As far as penalties are concerned, common sense suggests that evasion should decrease in direct proportion tc anticipated penalties. Raising fines and applying jail terms more frequently would have a deterrent effect on tax fraud. Legislation has also been passed recently to increase penalties for tax evasion; only experience will tell whether the enhanced penalties have a sufficient deterrence value, but this too is a step in the right direction. What remains to be done now is to protect tax personnel from potential harassment. Thus it is recommended that legal protection be given to tax personnel against lawsuits for actions taken in their official capiytj in such cases, these personnel should also have access to public legal and financial support.  viii - 29. Customs Administration and Smuggliag. Underperformance with respect to the collection of import duties is primarily related to inadequate surveillance and prevention of smuggling. Other problems, such as miodeclarLtion of dutiable value, are comparatively minor and are being adequately addressed by the pre- shipment, import inspection scheme; the incidence of misdeclaradion should be further reduced by the recent decision to globalize this scheme. 30, The value of smuggling into the Philippines is high and seems to be rising. Indicative estimates prepared by the Economic Investigation and Intelligence Bureau show that smuggling averaged around $754 million during 1979-85 or about 9? of average annual imports, and rose to $1.1 billion per year during 1986-90, or about 12Z of average annual imports. Apprehensions of smuggling attempts are few and far between. The value of apprehended cargo is typically a small fraction, around 4%, of the value of estimated smuggling. 31. Perhaps the most important cause of low apprehensions is the meagerness of the resources devoted to the prevention of smuggling. For example, customs water police have only one operational patrol boat. Even this is used not for patrol duties but as a means of transport for customs officials to board ships for cargo inspection. The bureau does not even have its own stripping machines for cargo inspection; instead, it relies on private firms for this function, an aspect that reduces the speed and flexibility with which the customs police can act. The budget of the customs police has decreased in nominal terms since 1986 and has declined as a proportion of the overall customs budget in every one of the last five years. 32. Given the seriousness of the smuggling problem, it is recommended that resources (both personnel and equipment) devoted to the anti-smuggling effort be considerabli increased; the recent award of a global import inspection contract to a private party should permit a reallocation of customs budget resources. Privatization of State-Owned Assets 33. Government has been implementing a rationalization program during the past five years among whose key features is the privatization of state-owned assets. While revenue generation is not the dominant rationale for this program, it is an important consideration. From the point of view of revenue mobilization, the program has been a partial success. During the last five years, a sum of around P44 billion has been recorded as gross proceeds from privatization. Of this sum, however, only P14.5 billion has beern remitted to the national treasury so far. The remainder is accounted for mainly by amounts held in escrow pending settlement of legal issues, sales transactions awaiting completion via debt- equity or debt-asset swap schemes, and amounts payable in installments. 34. For the period 1992 and beyond, privatization is projected to gross P42 billion of which Government's share is expected to be around P14 billion. The bulk of the sales, an amount of P30 billion, is expected to occur in 1992; of this, Government's share is expected to be about P9 billion. Privatization got off to a good start in 1992 with the successful auction of Philippine Air Lines, a "big-ticket" item which drew a higher-than-expected bid. Among other such items slated for 1992 is the sale of a further batch of shares in the Philippine National Bank.  - ix 35. The proceeds from privatization, both historical and projected, are small relative to total revenues. During 1987-90, .he cumulative remittances to the treasury amounted to just around 1.5% of total revenues. Except in 1992, future contributions will be even less significant, partly because the proceeds will taper off as the stock of privatizable assets shrinks and partly because total revenues (the denominator here) will increase through growth and inflation. Nevertheless, privatization should be encouraged for several reasons. First, it has intrinsic merit in that it leads to a more efficient deployment of overall resources. Second, in an environment of severe fiscal constraint, additional resources have a high marginal utility. Third, even if the direct revenue effect is limited, the indirect savings from disposing of non-performing assets and inefficiently managed public corporations can be quite substantial. Revenue Mobilization Priorities 36. The foregoing has indicated that additional revenues might be derived from improvements both in the design and administration of the tax system. Given current fiscal needs, the Philippines does not have the luxury of preferring one set of improvements to the other- the "optimal" revenue strategy would combine new taxesltax refinements with measures to improve tax collection. Action must be taken on all fronts. Recently, a number of important tax measures have been legislated. These include the imposition of more rational ceilings on business deductions, the simplification of the net income tax scheme, the restructuring of transfer taxes (on estates), the imposition of higher penalties for tax evasion and the setting up of specialized tax courts. In addition, a global import inspection scheme is due to be implemented soon. Among remaining measures, high priority should be given to the following on the grounds that they are likely to "produce" sooner: (a) reduction of VAT exemptions; (b) restructuring of vehicle registration taxes; (c) computerization of revenue administration; (d) introduction of a presumptive income level scheme; (e) enhancement of legal protection for tax personnel and (f) amendment of the Bank Secrecy Law. C. Improving Public Expenditure Management 37. While enhanced resource mobilization is a necessary condition for sustaining higher growth in the future, it is not a sufficient condition. It is important to go beyond raising resources to ensuring that they are used in an efficient manner and towards desirable ends. The experience of the Philippines during the 1970s provides a case in point. Public expenditures rose substantially during the 1970s, but they were not deployed in an efficient manner. Higher public expenditures were associated with substantial growth in the size of government, with an increase in subsidies for government services, and with a preference for large and expensive investments in sectors where comparative advantage was doubtful, to say the least. Consequently, these expenditures did not result in sustained growth and, to the extent that they were financed by external borrowing, did not generate the resources needed to meet debt-service obligations. This contributed to the fiscal and economic crisis of the mid-1980s, a crisis whose ramifications affect budgetary priorities and economic performance even to this day.  mx- Patterns and Product4vity of Public Expenditures 38. Public Sector Investment. The relative decline of public investment expenditure is a prominent feature of the pattern of public expenditures in recent years: total public sector investment, that is, including the investment of public enterprises and local government units, declined sharply from an average of 6.4Z during 1975-85 to only 3.92 during 1986-90. Both the national government and government corporations cut investment outlays as external flows declined (relative to the 1970s). There was, in addition, a dramatic shift in the sectoral composition of public investment: the share of the power sector dropped from 30Z of the total in 1975-85 to 19X in 1986-90 and that of infrastructure (roads, bridges, etc.) dropped from 66% to 53X. On the other hand, the share of education almost quiadrupled (from 2.22 to 8.4%) while that of health more than doubled (from 0.6Z to 1.8%). This shift was due to two factors: a reduction in the funds available to public enterprises (which are the major investors in power and water etc.) and a commitment to social infrastructure development made by the Aquino Government. 39. Effect on Overall Productivity. The above-described changes in the size and composition of the public investment programs could have significant implications for long-run growth in the Philippines. A proper assessment of these implications requires an understanding of the relationship between public expenditures and productivity growth. This relationship has long been the subject of debate. Proponents argue that such expenditures are critical to development in that they (a) make up for the lack of private capital and initiative to develop infrastructure (b) compensate for market failure in certain areas (e.g. public goods provision) (c) exploit large positive externalities in the social sectors and (d) stimulate private investment by raising aggregate demand as well as by creating supportive and enabling infrastructure. Opponents argue, however, that the evidence of market failure and lack of private initiative is weak and that public expenditures do not have a net stimulative effect for one or both of two reasons: they may depress overall productivity since they are typically associated with less-efficient and non-competitive (public) enterprises and they may crowd out private investments by increasing the cost and/or decreasing the availability of credit to the private sector. 40. Judgements with regard to the productivity of public investment in the Philippines must be based on cross-country experience since country-specific empirical analysis is hampered by lack of appropriate data and a commonly- accepted methodology. The tentative conclusions that might be drawn from cross- country evidence are (a) that pub,lic investment does increase productivity; (b) that operating and maintenance expenditures might be preferable to directly productive capital expenditures; a bias towards capital formation as opposed to capital maintenance could be detrimental to growth; and (c) that the positive growth effects of human capital formation (via social sector spending) might be greater than those of physical capital formation. In light of this, the reduction in the size of the public investment program witnessed in recent years should be a cause for concern; cuttin the public investment program and not spending enough on maintenance may only succeed in achieving short-run adjustment at the expense of long run growth. On the other hand, the shift in the composition of the public investment Drogzam in favor of human capital formation could, under appropriate macroeconomic management, improve the returns to the program.  - xi 41. Effect on Private Investment. Another channel by which public investment is thought to affect national productivity is by way of its effect on private investment. There are two considerations here, conveniently captured by the terms "crowding out" and "crowding in". Crowding out occurs when the effect of financing public expenditures through borrowing is to raise the cost and/or decrease the availability of credit to the private sector, thereby decreasing private investment. Crowding in can occur from the stimulative effect of public expenditures in one or both of two ways. There could be a demand-side effect as higher government expenditures raise aggregate demand, subsequently leading to an increase in private investment. There could also be a supply-side effect as public investments relax key infrastructure constraints thereby enabling higher levels of private investment. 42. Analysis for the Philippines suggests that public spending contributes to crowding out through financial markets. The higher the demand for public credit, the higher are interest rates and the lower is the level of private investment. In the 1980s, these links have been even stronger since constrained foreign savings meant that government had to satisfy more of its credit needs from the domestic market pushing up interest rates in the bargain. 43. Investigation also suggests that government investment outlays have a positive direct effect on private investment whereas spending through government corporations (GCs) has a negative effect. This cannot be neatly identified with Infrastructure and non-infrastructure investments since some GCs, such as the power and water utilities, provide much-needed infrastructure services. It does reflect the fact, however, that many GCs tend to compete with the private sector in various business areas and may therefore crowd it out. This is a plausible description of GC behavior in the late 1970s and early 1980s, when many of them were active in the manufacturing and mining sector and a few held monopoly positions in other sectors (e.g. coconut oil processing). 44. In any ctee, the mere provision of infrastructure may not be enough to generate and sustain private investment; there typically needs to be a supportive economic and political environment as well. This point can be illustrated from the experience of export processing zones in the Philippines. Starting in the early 1970s, several export processing zones were established; they were provided with basic infrastructure such as roads, water and power and benefitted as well from fiscal incentives and the availability of imported inputs at world prices. However, on the whole, their performance has been disappointing in terms of investment, employment and export growth, especially when compared to neighboring countries. This is due in part to a poor choice of location for several of the EPZs set up in the Philippines and in part to a policy environment that has generally supported import-substitution activities. A good example of the location problem is that of the Bataan EPZ, which is situated in a relatively remote area (about 200km from Manila) and has never performed to potential. Despite the availability of good basic infrastructure within the Bataan EPZ in the early years, it was unable to attract many investors. Over time, the lack of a growing revenue base led to cash shortages and inadequate expenditures on maintenance which led in turn to a deterioration of the basic infrastructure: all in all, it has turned out to be a poor public infrastructure investment. On the other hand, a similar investment made in the Mactan EPZ has turned out well primarily because this site is well located, being cheek-by-jowl with Cebu.  - xii - Private Sector Role in Public Investment 45. Three factors have led to increased discussion of a potential private sector role in the provision of public investment in the Philippines. The first is tbe continuing budget problem: there is not enough available to finance the desirable level of infrastructure investment. Second, the e is a mounting awareness that public investment quality has been poor and that one way to guard against this would be to transfer part of the risk, and part of the reward, of ensuring quality to the private sector. Third, much international experience has been gained with new mechanisms, such as the Build-Operate-Transfer (BOT) concept and its many variants, for involving the private sector in public investment activitiest such experienri has reduced the uncertainty for both public and private parties contemplating such ventures and thereby reduced their resistance to the idea. 46. The need for private participation is well illustrated by the example of the power sector in the Philippines. The National Power Corporation which Is the public entity entrusted with the task of generating power has been unable to generate the necessary funds to meet planned investment schedules; it faces charges of inefficiency, mismanagement and corruption as well. The power plants under its management have technical and non-technical losses that are above industry ste.ndards in Southeast Asia. Since 1989, there have been frequent power outages and it is estimated that output losses to industry on account of this amount to around $1.1 million per day. The investment requirements to close the demand-supply gap are overwhelming. During the period 1989-1999, it is estimated that 3,679 MWs of new capacity must be added; this would cost some $7.5 billion, of which $4.7 billion would be needed for new generating capacity alone. It is estimated that only 28% of this investment requirement can be sourced domestically and 722 must come from foreign sources, in the form of loans and/or equity. Since NPC is not in a position to service its loans without budgetary support and since the pressures of the fiscal deficit indicate that even the 28% projected local public funding may not be available, it is imperative that private financing be sought. 47. The Philippines has put in place a set of policies and guidelines to facilitate private sector participation in a broad range of public infrastructure investments. An Executive Order (E0215) was enacted in 1987 to permit such participation in power projects. One project has already been completed, one is under negotiation, and three more are under evaluation. A total capacity of 1660 MWs is to be developed under private contracts through 1995. In July 1990, an Act of Congress expanded the permitted scope of private participation to cover most infrastructure investments, including those in transport, communications, water supply, sewerage, education, health, and environmental management. Some improvements have been brought about as well in the set of laws governing foreign investment; this should make it easier for foreigners to participate in public investment projects. It is generally conceded that the political will, the executive enthusiasm and the appropriate legal framework are now in place. What is needed is experience with actual projects. With the benefit of such experience, guidelines and regulations can be fu-ther refined. 48. Some experience has already been gained. In 1989, NPC awarded a BOT contract to Hopewell Holdings Ltd. under which the latter committed to install  - xiii - a 210MW gas turbine, operate it for 12 years, and then transfer the ownership to NPC. The installation was completed rapidly and the project is running smoothly so far. The success of this project was due in part to the fact that Hopewell was an experienced regional project developer able to offer a very attractive package to NPC. For its part, Government provided an adequate guarantee for the purchase of power from the Hopewell project. Another proposed BOT project, this one for a co-generation (steam and electricity) plant to be set up by Cogentrix Inc., did not fare as well. It was abandoned after protracted negotiations chiefly because of a failure to agree on an appropriate power purchase plan. Other critical factors were the lack of a clear process regarding environmental and construction permits, lack of a strong local joint venture partner and failure to secure financing. The Cogentrix example suggests that there needs to be much learning-by-doing in this area, for both Government and the private sector. Given the complexity of BOT projects and the lack of experience with forward contracts in infrastructure services, rapid and positive resolution of deals should not be routinely expected and such ventures should not be viewed as a panacea for the public infrastructure problem. In particular, such initiatives should not distract attention from the need to improve the public sector's ability to provide the bulk of the infrastructure program. Improving Public Enterprise Performance 49. Public enterprises are relevant to fiscal adjustment to the degree that they depend on the national government budget for their financing needs. If this dependence can be reduced, the process of fiscal adjustment would be eased. There are two ways in which this is normally done. First, some government corporations (GCs) can be privatized: this generates additional government revenues and decreases future claims on the budget. The historical and potential yield of the ongoing privatization program has already been noted in a previous section. Second, for those enterprises that cannot be privatized, financial restructuring can be undertaken so as to develop alternative sources of funds and management improvements can be introduced so that costs are reduced, productivity is increased and lower claims are made on the budget as a consequence. The Philippines has attempted both financial and management restructuring for its GCs in recent years. 50. Financial Reform. Financial restructuring has taken the form of a sharp reduction in net budgetary transfers to GCs: net flows to GCs declined from a high of P33 billion in 1986 to P4 billion in 1990, or from 30% of expenditures to around 2Z. The decline was made up of a reduction in gross flows in the form of equity infusion and net lending as well as an increase in GC transfers to the national government in the form of dividends, interest payments and the proceeds of privatization. Restructuring has also involved a change in the nature of the financial link with GCs, a change which made the size and cost of transfers more transparent. Thus, subsidies have been preferred to equity or loans and, when made, loans have carried market interest rates. This approach has also been applied to tax exemptions. Most GCs (all but six) had their tax:exemptions revoked in 1986; since then, they have had to do without such privileges or apply to obtain relief from a Tax Expenditures Fund (TEF) set up in the budget. This effectively limits the size of tax exemptions to the appropriation cover for the TEF and makes them transparent in the sense that they are subject to periodic review by Congress.  - xiv - 51. Such restructuring has contributed to greater financial autonomy which is reflected in declining dependence on funding from the government budget: the ratio of government-sourced funds to total expenditures fell from 20Z to 72 over the period 1986-90. There was also a rise, on average, in the volume of cash generated internally as well as an improvement in operating ratios from 1986 to 1989 (but followed by a decline in 1990-91). It is generally believed that some GCs have managed themselves better once weaned from the public purse. 52. The above account of financial restructuring should be qualified by three considerations. First, there was little attempt at generating internal cash by a market-oriented adjustment of user charges; indeed, user charges declined in real terms over 1986-90. Second, there was a sharp cutback in the size of the overall capital expenditure program of GCs, a cutback that also permitted slower growth in related current expenditures. A far better outcome of reduced dependence on the budget would have been an aggressive attempt to control costs, raise user charges in line with inflation, and preserve the investment progrm. Third, the trend in the direction of greater financial autonomy was upset in 1991 on account of the deteriorating financial condition of the largest public corporation, the National Power Corporation. NPC was given financial support (loan plus equity) in the amount of almost 5 billion pesos, an amount equal to almost 602 of the original budget for transfers to all GCs. 53. Management Reform. Management reform in the GCs has been attempted through several means. First, the overall policy-making and monitoring body (the GCMCC) was reconstituted in 1987 with a clearer mandate. Second, technical assistance was made available to GCs to formulate corporate plans, undertake computerization of their accounting and management information systems, and carry out studies of their organizational needs and opportunities. Third, measures were adopted to increase operational autonomy of GCs. Fourth, a performance evaluation and incentive system (PEIS) was introduced, initially for five GCs (in 1989) and subsequently for another 23, with the goal being that of covering all eventually. 54. The results so far derived from the PEIS show that there remains much room for improvement of GC performance. Several GCs have consistently failed to achieve some targets. The general areas of concern and the main GCs affected are as follows (see inside cover for explanation of acronyms): -Collection of receivables: NEA, NIA, LWUA and MWSS. -Operating Ratio: NHA, NFA. -Systems Loss: MWSS, NEA, NPC. -Profitability: NEA, NIA, LRTA, PNR, NFA 55. The experience with profitability merits special consideration. The net income of GCs rose from P2.2 billion in 1985 to P6.3 billion in 1989, a handsome 30X annual growth in nominal terms or 212 in real terms. However, this is misleading to the extent that it includes government subsidies. Excluding the subsidies, the net income picture is quite different. It shows a decline in both nominal and real terms over the period 1985-90. Five of the 14 major GCs incurred consistent losses during this period: these are NIA, NEA, LRTA, PNR, and NFA. Since all but the NFA deal with the provision of basic services (irrigation, rural electricity, and transport), their poor financial performance is a cause for concern. These are not services whose costs are hard to measure  xV - or price or whose beneficiaries are hard to separate into subsidy-deserving and non-deserving groups. Poor financial Rerfomance for these GCs suggests that there has been inadequate reliance on cost-based user charges and insufficient control over cost and productivity. These deficiencies point to corresponding solutions, that is, better cost control and higher user charges. This conclusion is bolstered by the recent experience of the National Power Corporation, which suffered huge losses in 1991 due to failure to adjust user charges in a timely fashion; NPC management stands Accused as well of having exercised poor control over costs. 56. The question of control over costs and user charges is tied up with the degree of operational autonomy that individual GCs have. The degree of autonomy varies by category as well as by GC. There exist limits on the authority of GCs to make personnel, investment and financing decisions. For example, as far as the ability to hire and fire personnel is concerned, all GCs are bound by civil service rules and regulations; compensation levels are also governed by DBM guidelines. There is more flexibility on pricing decisions, although this varies by type of corporation. For example, most public utilities have to go through public hearings before price adjustments can be made; they are also subject to laws limiting rates of return on investment to 12%. The NPC and NFA, moreover, have to clear price adjustments with the Office of the President. NPC's financial crisis in 1991 was due largely to the failure of oversight authorities to allow user charge increases requested by NPC management. Among the major utilities, only the MWSS has full power to aujust rates without the approval of higher authorities. Since control over costs and user charges is shared between GCs and Government, responsibility must also be shared. Ultimately, Government must make the appropriate political decisions that will truly rationalize public sector operations and enable greater efficiency. Such decisions should not be confined to cost control and iser charges only: government should consider taking advantage of the fact that the services provided by some of the loss-making GCs are eminently privatizable. 57. In view of the foregoing, the following specific recommendations are made: (a) the subsidy program should continue to be re-examined with a view to determining desirable levels of subsidy for various activities and especially for food and irrigation, the two highest subsidy receiving activities at present; the review of the program should consider whether or not subsidies are in fact reaching deserving beneficiaries and whether or not subsidy-delivery costs could be reduced through improvements in management and adjustment of user charges; (b) the requirement for GCs to remit dividends to the NG should be implemented with as few exemptions as possible; the 10% mandatory dividend declaration rate (percentage to net income) should be increased to 15%, which is the average for firms listed on the stock exchange; (c) GCs should be given more operational flexibility particularly with regard to the adjustment of user charges; at the very least, they should be allowed to automatically raise user charges in line with inflation; in general, power to adjust user charges should be vested in the governing boards of GCs rather than in the Office of the President;  - xvi - (d) GCs should also be given more autonomy in matters relating to personnel hiring and firing decisions; present restrictions on such powers limit the ability of managers to improve manpower quality and reduce manpower costs and are inconsistent with the general objective of achieving a leaner but higher quality bureaucracy; and (e) systems to reward good performers and penalize poor performers among OCs should be pursued, following on Executive Order 486 of November 8, 1991, which provides for the payment of bonuses and incentives to staff of well- performing corporations. D. Combining Resource Mobilization with Efficiency 58. It has been argued in this report that a desirable development strategy would combine a shift towards non-debt sources of public finance and the use of efficiency criteria to guide investment. Reconfiguring the sources of public finance will require further improvements in revenue effort. Emphasizing investment efficiency will require further exposing the economy to the discipline of international competition through tariff reduction and a market determined exchange rate. There is no inconsistency in the long run between these objectives or the effects of the measures required to achieve them. In the short run, however, there may be tradeoffs. For example, import duties are an important source of revenues and tariff reductions will most likely decrease revenue flows to the treasury. Similarly, exchange rate depreciation which may be required to maintain international competitiveness will simultaneously increase the foreign debt-service burden of the public sector, thereby increasing expenditures. Given these possibilities, it is worthwhile exploring the quantitative dimensions of the tradeoffs between various measures that might be undertaken to promote domestic resource mobilization and economic efficiency. 59. Another, and recurrent, concern for the Philippines is the management of the current account deficit. Given a comparatively low level of investment efficiency and a tight external financing constraint, such deficits quickly threaten to produce balance of payments crises. Hence, it is important tce maintain a stable and low current account deficit; at present, in the context of the IMF-supported stabilization program, the Philippines is in the process of bringing down this deficit from the high and unsustainable level of 5.2Z of GNP reached in 1990. In achieving this objective, which translates into a reduction of foreign savings, domestic savings must be increased through fiscal adjustment. Policy formulation in this area requires knowledge of the quantitative links between various fiscal adjustment measures and the current account objective. 60. Such quantitative links were examined with the aid of a computable general equilibrium model of the Philippines. Four scenarios were simulated: (a) tariff reform as specified in Executive Order 470; (b) reduction of current account deficit by 1% of GDP; (c) an increase of 1Z in the investment rate and (d) a combination of selected efficiency increasing measures together with (a), (b) and (c). The efficiency measures comprise expansion of VAT coverage, elimination of all subsidies prov'ded by Government, abolition of the special 5% import levy, and liberalization of investment and trade (leading to an increase in import and  xvii - export elasticities). In each case, the model was used to calculate the required increase in domestic taxes or decrease in government consumption using 1990 as the reference year. 61. The results suggest that significant amounts of domestic revenues are needed in order to achieve in a sustainable fashion the medium-term objectives of greater efficiency, more growth, and a lower current account deficit, that is, the objectives implied by (a), (b) and (c) above. These objectives can be accomplished by raising domestic taxes by 38Z which implies an increase in the tax ratio of around 192. Given that the typical rate of increase in tax revenues in recent years has been around 22%, this suggests that special efforts are required to increase the elasticity and buoyancy of the tax system, that is, tax collections must be increased through a combination of reforms affecting the design and administration of the tax system. 62. of coura-, a reduction of government expenditures would ease the revenue enhancement required. The results of the simulation exercises suggest that such expenditure reductions will have to be of similar magnitude to the revenue enhancements, that is, around 38% for the medium-term objectives case. Efficiency or productivity enhancing measures would also help. A simulation combining selected expenditure reductions, revenue increases and productivity improvements (scenario (d) above) suggests that the required revenue mobilization can be cut to 192 by such measures. 63. The key conclusions are, therefore, the following: (i) a significant increase in revenues is required to meet the medium-term objectives of greater efficiency, higher growth and lower external deficit; this cannot be met by normal revenue increases associated with growth but must be obtained from improvements in tax collection; (ii) the required revenue effort can be moderated by improvements in productivity and reductions in public expenditures which are obtainable from such measures as higher user charges and better cost control for most public enterprises.  I. THE MACROECONOMIC CONTEXT 1.1 This chapter provides the historical macroeconomic context for relevant issues in domestic resource mobilization and management. It begins by reviewing, in Section A, the growth experience of the Philippine. in the last fifteen years and analyzing, in Section B, the role of two key variables, savings and investment, an. their public, private and foreign components. This provides the basis for outlining, in Section C, the thrust of an appropriate macroeconomic management strategy, a strategy that would be best placed to restore the Philippines to a sustainable growth path after the roller-coaster ride of the last fifteen years. Section C concludes with an overview of the organization of the report and the main issues dealt with in subsequent chapters. A. Growth Experience since 1975 1.2 The chronicle of Philippine growth since 1975 is essentially one of unsustainable expansion followed by inevitable contraction followed by an unsteady recovery (see Table 1.1 and Figure 1.1). The late 1970s were a boom period; real GNP growth averaged 6.1% during 1975-79. The first half of the 1980s was a bust period as the growth rate faltered to an average of 3.6? during 1980- 82 and collapsed to -4.8% during 1983-85. The deep recession of the mid-1980s effectively set per capita income back to the levels reached in the late 1970s. Since then, the economy has recovered but in a wobbly fashion; real GNP grew by 5.5Z during 1986-89 but then slid sharply during the next two years, to 4.2% in 1990 and virtually 0% in 1991. 1.3 The Late 1970s. High growth in the late 1970s was surprising coming as it did in the face of deteriorating terms of trade brought on by high oil prices and falling commodity prices. It was made possible by a strategy that emphasized finance rather than adjustment. Instead of adjusting to the oil price and terms of trade developments by restructuring its economic base, the Philippines ignored the new relative price signals and spent its way through a multi-year boom on the strength of foreign savings in the form of loans. There was a substantial inflow of such loans in this period and the external debt of the Philippines grew from $3.8 billion at the end of 1974 to $13.4 billion at the end of 1979. Much of the increment in external finance came from private commercial banks (see Table 1.2) which had become flush with 'petrodollar" deposits as a consequence of the sharp rise in the price of oil earlier in the decade. The increase in external finance supported an increase in aggregate investment, which grew at an average of 11% over this period as compared to 7% during 1960-75. Much of the increase in investment took the form of government investment, as government embarked on an ambitious program of public sector expansion: public sector investment jumped from around 5% of GNP in 1975 to over 11% in 1976 and averaged around 10% over the next four years.  -2- Table 1.1 Key Macroeconomic Indicators 1975-79 1980-82 1983-85 1986-89 1990 19911/ Real GNP Growth 6.1 3.6 -4.8 5.5 4.2 0.1 Current Balance/GNP -5.0 -6.8 -5.0 -0.6 -6.8 -4.5 Public Sector Balance/GNP/ -4.0 -5.3 -5.8 -3.7 -4.9 -2.5 Terms of Trade 141.9 112.0 105.6 120.9 109.2 n.a. Real Exchange Rate 101.8 90.0 101.8 92.8 86.8 98.8 Inflation 10.2 13.8 27.8 6.0 12.7 17.7 / Preliminary. -! Comprises deficit of national government and 14 non-financial public enterprises only; consolidated deficit was 7.8Z of GDP in 1990. Source: National Economic Development Authority and staff estimates. Figure 1.1 Real GDP Growth Rate 10--** - 4 p r 2 : e 0 L n -2- -4 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91  -3- Table 1.2 External Finance Flows: 1970-891/ (Millions) 1970 1975 1980 1985 1987 1989 Total Debt. Flows Net Flows 182 533 1619 1136 -161 1013 Interest Payment 44 118 1491 1756 1849 2183 Net Transfers 148 415 128 -620 -2011 -1169 Private Flows Net Flows 7 175 799 532 -770 -419 Interest Payment 15 33 231 473 747 1031 Net Transfers 7 142 568 59 -1516 -1450 Memo: Commercial Banks Net Flows 101 126 771 536 -631 -247 Interest Payments 29 30 372 518 765 1026 Net Transfers 71 96 400 18 -1397 -1272 I/ Data refer to all debt except for 1)70 and 1975 which exclude short-term debt. Source: World Debt Tables, 1990-91 1.4 This aggressive strategy undoubtedly produced high growth, but it also weakened the economy in several respects. First, it was based on borrowed money which had to be repaid; this reduced the room for maneuver available to economic managers who thereafter had to make policy decisions with debt-servicing obligations in mind. Second, much of the borrowed money was used to fund an expansion of the public sector; this did not help in generating foreign exchange revenues with which to meet debt-servicing obligations. Third, the easy availability of credit encouraged excessive leveraging not only in the public sector but also among private corporations. Fourth, this strategy was associated with an increase in investment in import-substitution activities supported by a highly protectionist industrial and trade policy regime. Indeed, it caused an increase in import-substitution activities to some extent by propping up the value of the peso. Fifth, it did not promote the country's competitiveness in that the foreign finance flowed largely to the non-tradables sector (see Table 1.8). Sixth, it initiated a weakening of the domestic savings effort and particularly the personal savings effort, which began a decline from a level of around 11% in 1976 to virtually zero by 1983. All these developments rendered the econom' more fragile, more unbalanced, and more vulnerable to shocks. 1.5 The Early 1980s. A series of shocks occurred during the early 1980s which exposed the weaknesses of the growth strategy of the late 1970s. The  -4- shocks included, among externally generated events, the second oil price shock, an increase in world interest rates, a recession in the industrialized countries, and a sharp decline in the availability of foreign loans. These shocks would have adversely affected growth performance in most countries but they had an exceptionally severe effect on the Philippines economy. Excessively leveraged firms, both public and private, began to fail as interest rates rose and loan supply tightened. Even export oriented firms were hard hit as the industrialized countries went into recession. A balance of payments crisis loomed as terms of trade deteriorated further and as external funding began to dry up. Two additional events, a financial scandal in 1981 and the assassination of Ninoy Aquino in 1983, contributed to the mounting loss of confidence in the economy. Capital flight ensued, businesses defaulted on their debts, and banks failed. The net result was a faltering of growth during 1980-83 and a recession during 1983-85. Ultimately, the economy adjusted to the various shocks in the only way it could, through a deep recession. 1.6 The Late 1980s. The economy began a recovery in 1986, and this gathered considerable steam during 1987-89 when real GNP grew at an average of 6z. Recovery was assisted by several developments. First, the change in government in 1986 restored private sector and foreign confidence in the economy. Second, larger flows of external finance resumed, although primarily from multilateral institutions. Third, the new government introduced significant reforms such as rationalization of the public sector, deregulation of key parts of the agricultural sector, and restructuring of taxes and tax administration. Fourth, macroeconomic management was sound, distinguished by a prudent monetary and fiscal stance and a responsible approach to external debt reduction. Fifth, the external environment improved in that oil prices and interest rates fell during this period while the industrialized countries, the destination of the bulk of the Philippines' exports, enjoyed relatively higher growth than in the early 1980s. 1.7 The recovery ran out of steam after 1989. During 1990, real GNP grew by around 4% while inflation increased to around 13%. During 1991, real GNP is expected to grow by a meager O.'Z; the first three quarters have seen negative growth. Investment collapsed in 1990, growing by only 5.6%; it subsequently declined by almost 17% in 1991. These developments were due in part to an extraordinary series of shocks that occurred over this period, including a coup attempt in December 1989, a prolonged drought in 1989-90, an earthquake in July 1990, the Gulf crisis in 1990-91, and a major volcanic disaster in June 1991. 1.8 The external environment turned unfavorable in 1990 and 1991 (see Table 1.3), thereby greasing the slide of the Philippine economy. Growth in the industrial countries decelerated from 3.3% in 1989 to 2.6% in 1990 and further to 1.32 in 1991. World trading volume declined correspondingly, from 6.9t in 1989, to 4.3% in 1990, and a meager 0.62 in 1991. Commodity price movements did not help either: oil prices jumped by 28.2Z in 1990 before falling back by 16.4% in 1991 whereas the prices of non-fuel commodities, including some of the Philippines' principal traditional exports, declined by 7.9% and 5.7X in 1990 and 1991 respectively. About the only favorable trend was that in world interest rates; six-month LIBOR softened from a level of 9.3 in 1989 to 8.4 in 1990 and further to 6.6 in 1991. The overall external environment was thus clearly unsupportive of high growth in the Philippines, and particularly so for export  -5- growth. Unfortunately, these external 4evelopments coincided with some domestic policy slippages. Table 1.3: External Environment: 1989-92 (Annual Percent Change, except as noted) 1989 1990 1991 1992 (Actual) (Projections) Output World 3.3 2.2 0.9 2.8 Industrial Countries 3.3 2.6 1.3 2.8 U.S.A. 2.5 1.0 -0.3 3.0 Developing Countries 3.2 1.0 -0.6 2.9 World Trade Volume 6.9 4.3 0.6 5.0 Commodity Prices Oil 21.5 28.2 -16.4 1.0 Non-Fuel -0.5 -7.9 -5.7 3.3 Interest Rates (level) 6 Month LIBOR 9.3 8.4 6.6 7.0 Source: World Economic Outlook, October 1991, IMF. 1.9 On the domestic policy side, the real exchange rate was allowed to appreciate (in 1989-90), thereby contributing to a widening current account deficit. Public sector expenditures rose sharply, as both current and capital outlays were increased beyond planned levels, thereby contributing to an enlarged fiscal deficit. Monetary policy was insufficiently tight, thereby allowing inflation to rise beyond target levels. In 1990, the fiscal deficit rose to 5.2Z of GNP, the current account deficit increased to 5.8Z of GNP and inflation rose to 14X. By the time control was re-established over key macro balances, a lot of damage had been done. A line of credit provided by the IMF (the EFF of March 1989) was canceled towards the end of 1990 as the program had gone significantly off-track, and negotiations for a new facility began. The Consultative Group meeting, scheduled for the summer of 1990 and necessary to build confidence through donor support, was postponed until a new agreement war reached with the IMF; the meeting ultimately took place in February 1991. The regular budget process was suspended while these negotiations proceeded: the 1991 budget, due to be passed in December 1990, was actually passed in June 1991. As a result, government expenditures were severely squeezed in the first two quarters and disbursements were governed by cash on hand rather than by the size of the appropriation cover.  -6- 1.10 In addition to exogenous shocks and policy slippages, macroeconomic performance during 1990-91 was affected by structural features of the Philippine economy. The two most important structural constraints that were operative over this period were the narrowness of the domestic capital market and the high import-dependence of the economy. The capital market failed to smoothly absorb the increase in domestic debt that was a key element of the recovery strategy. Domestic debt had to be increased for three reasons. First, because external finance, although flowing at higher levels than in the crisis years of 1983-85, was still limited: indeed resource transfers continued to be negative on a net basis. Second, because the public revenue system, while improving as a consequence of tax reforms introduced in 1986, was still yielding less than required to cover expenditures. And, finally, because expenditures could not be cut further from the already low levels reached in the crisis years; indeed, it made sense to restore them in order to "pump-prime" the economy. Unfortunately, the domestic capital market proved too thin to support this strategy: as domestic debt grew (from 192 of GDP in 1985 to around 22.4% in 1990) interest rates began to climb; the Treasury Bill rate averaged around 10% in real terms during most of 1990 and reached a high of 15Z in real terms (34Z in nominal terms) at one point. Such high rates were a signal that domestic borrowing needed to be curtailed. They complicated fiscal strategy since expenditures were rising largely to meet higher interest payments. They changed lending and investment priorities in that banks increasingly sh:fted to holding T-bills rather than making loans to business and businesses shifted to holding financial assets rather than making physical investments. 1.11 The import dependence of the economy also prevented the consolidation of the recovery along a sustained high growth path. As incomes and investment grew during 1986-88, imports increased sharply and at more than twice the pace of export growth. Th large differential in these two growth rates, and the absence of strong offs-tting trends in non-merchandise trade and transfers, led to an increase in the current account deficit after 1986 which, indeed, had been a surplus year. While the size of the deficits was quite manageable during 1986- 88, this was not the case thereafter and the balance of payments became a real constraint on policy and performance. This behavior of imports in particular and the current account balance in general has been true to type for the Philippines: high growth periods are associated with high rates of import growth which very quickly lead to balance of payments difficulties which, in turn, are defused by measures to curtail imports. Usually, such measures are deflationary; sometimes, they produce a recession. Exports play a secondary role; typically, they have been unable to prevent balance of payments crises or prevent the recessionary outcome of remedial measures. 1.12 The Stabilization Program in 1991. The growing instability of the economy prompted a fresh stabilization effort towards the end of 1990. While negotiations proceeded with the IMF, a special import levy of 92 was introduced as a temporary stabilization measure. Administrative measures were also taken to curtail government expenditures. A new program was agreed to with the IMF by February 1991 under which a fresh set of macroeconomic targets and policies  -7- Figure 1.2 Ratio of Public Sector Deficit to GNP 10 8 p e 8 r C 0 n 4 t 2 76 76 77 78 79 80 81 82 83 84 86 88 87 es 8g 90 Figure 1.3 Ratio of Current Account Balance to GNP 4 2- 0--- r e 4 -8 -10 7 7 - 80 - - - 76 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90  were adopted for an eighteen month period running to July 1992. The key targets for 1991 and 1992 were the following: the consolidated public sector deficit was to decline to 3.7% and 2.52 of GNP; the current account deficit was to be trimmed to 4.5% and 3.92 of GNP; base money was to grow at a maximum of 10.9% and 12Z while broad money was to grow at not more than 10X and 12% over the respective years. Given compliance with these targets, it was envisaged that inflation would be reduced to under 102 and growth of real GNP would recover to 4% in 1992. 1.13 The above targets were to be achieved through conservative fiscal and monetary policies. On the fiscal side, expenditures were to be cut selectively and subsidies arising from the operation of the Oil Price Stabilization Fund were to be eliminated; furthermore, the internal cash generation ability of government corporations, and in particular the National Power Corporation, was to be improved by appropriate increases in user charges. At the same time, revenues were to be raised through a combination of new taxes, refinements of existing taxes and measures to improve tax collection. On the monetary side, reserve requirements and other monetary measures were to be implemented in a manner consistent with the base and broad money targets. The current account objectives were to be achieved through the maintenance of a flexible exchange rate regime. Longer term external and efficiency objectives were to be met through structural reforms in trade and investment policies and through enhanced official and bilateral funding. 1.14 The stabilization program had a mixed experience in 1991. While certain key targets were met, others were not; furthermore, the record of administrative and legislative achievement during 1991 was not such as to assure the program's integrity in 1992 and beyond. Five areas of concern emerged. First, while the fiscal deficit declined, this occurred largely as a consequence of steep cuts in capital outlays; this does not bode well for medium-term growth given the poor state of infrastructure in the Philippines and the dire need for its rehabilitation and expansion. On a more positive note, the reduction of the fiscal deficit contributed to a significant decline in interest rates. If interest rates are maintained at low levels, this would provide an opportunity to increase capital outlays and thereby improve the quality of the fiscal adjustment effort. 1.15 Second, no significant revenue measures were legislated in 1991. Measures to introduce refinements in existing taxes and improve tax administration languished in the legislature while the idea cf introducing new taxes did not prosper at all. Meanwhile, the temporary import surcharge was cut from 9X to 52 leading to a further reduction in anticipated revenues. 1.16 Third, the overall fiscal program was compromised by the emergence of a financial crisis in the largest public enterprise, the National Power Corporation. This crisis diverted a significant amount of funds (about 5 billion pesos), originally set aside for other sectors and activities, to the NPC and its expenditure program. The most disturbing feature of this crisis was its origin in a reluctance to pass on the true costs of producing electricity to the consuming public through appropriate user charges. 1.17 Fourth, while the current account deficit target was met, this was due largely to a sharp drop in imports arising from the temporary import levy and  -9- from a sluggish economy; a pick-up in the economy in 1992, perhaps due to election related spending, may see a destabilizing rise in imports again. A bright spot was the modest growth of exports even in the face of a considerable slowdown in world trade. The decline in imports and a substantial increase in net foreign loan inflows (arising from previous reschedulings) led to a strong international reserves position and an appreciation of the peso. A persistence of this condition raises the prospects of a balance of payments problem emerging once again witn the next rebound of the economy. 1.18 Fifth, monetary targets were missed for most of 1991. Partly because of this, but also because of cost-push developments, the inflation rate did not decline significantly in the first three quarters during which it ran at almost twice the target level of 9Z. 1.19 Recent Developments. These concerns, and especially the lack of progress on the revenue front, prompted a re-examination of the stabilization program towards the end of the year. Some of the above-noted problems have now been resolved. Among important developments have been the following: some revenue measures have been passed, the financial position of NPC has been improved through the passage of a substantial rate hike, provision has been made in the fiscal program for inclusion of extraordinary expenditures related to the Mt. Pinatubo disaster, significant liberalization initiatives have been undertaken irn the area of trade, foreign investment and foreign exchange regulations, the inflation rate has come down to single digit levels, and tighter monetary management has been implemented. While some key revenue matters remain unresolved, the developments just noted provide a greater degree of assurance about the integrity of the stabilization program for 1992. B. Savings and Investment Behavior 1.20 A simple but useful way to characterize the growth process is to focus on the links between savings, investment and growth. Growth depends on investment which, in turn, depends on the supply of savings. Both savings and investment can arise from three sources: public, private and foreign. While this characterization appears simple, depending as it does on just a few variables, their inter-relationships can be complex. For example, domestic and foreign savings might substitute for each other. Or, public and private savings might be substitutes also; as more public savings are raised through the tax system, for instance, private savings might fall. Furthermore, public, private and foreign investments might also either substitute for each other or be complementary depending on the content and context of the investment. Finally, there is the issue of investment efficiency. Higher investment levels need not lead to more growth if the investment is characterized by inefficiency and waste. 1.21 The behavior of these few macroeconomic variables, and the causes and consequences of such behavior, go a long way towards explaining the pattern of growth observed in the Philippines since 1975. This is not to suggest that these are the only variables that matter, just that their interactions provide a simple, yet comprehensive, explanation for key aspects of the Philippine growth experience.  - 10 - 1.22 Savings-Investment Gap. An overview of the relevant data indicates that there was a large gap between national savings and aggregate investment in the late 1970s and early 1980s that was filled by the inflow of foreign savings (see Table 1.4). A sharp increase in the inflow of foreign savings into the Philippines can be detected in the mid-1970s when, in the space of just two years from 1973 to 1975, the ratio of such savings to GDP rose from -5.5% to 5.3% (see Figure 1.4). The ratio stayed at around 5% through the rest of the decade and rose to an average of 7.2% during 1982-83. Then it declined sharply, reaching - 5.3% in 1986, and averaging about 0.5% since then to 1989 before rising once again to 5.12 in 1990. As already shown in the preceding section, this pattern in the evolution of foreign. savings was a fundamental determinant of macroeconomic performance in the Philippines during the last fifteen years. As foreign savings rose in the late 1970s, investment rose and income grew at a high rate. As foreign savings declined in the early 1980s, investment fell and income growth decelerated into a full-blown recession. Since 1985, again with the help of foreign savings, investment and income growth have recovered, although in an uneven and unsteady fashion. The cumulative effect of the reliance on foreign savings has been an increase in the stock of external debt, from a level of $2.2 billion in 1970 to $17.4 billion in 1980 and further to around $30.5 billion in 1990. 1.23 The composition of foreign savings was different in different periods. Durinag the 1970s, a large proportion of the net flow originated in private loan sources, such as commercial banks; for example, during 1975-79, about half of net total flows were from private sources and a quarter from commercial banks. In the next phase, private flows began to decline in relative terms as a series of shocks battered the economy and confidence began to decline. Indeed, the only way to keep private flows up was to pay very high interest rates for very short- term money: the term structure of the Philippines external debt changed from one in which the snort-tenm debt averaged around 35% (in 1975-79) to one in which short maturities (mostly 90 day loans) accounted for 45% (in 1980-82). After 1982, the declining economy accelerated the exodus of private lenders and, in turn, this exodus turned the economic decline into a free-fall. 1.24 During the 1980s, despite the deterioration of the economy, official flows continued; indeed, they rose on average during the crisis years of 1983-85, compared to the pre-crisis years, and have risen steaeily since then. While the trend in official flows was opposite that in private flows, this was not enough to prevent a decline in net resource transfers once interest payments are factored in: such transfers have been negative in every year since 1980, reflecting both the large increase in the Philippines' debt stock and the higher average level of world interest rates in the 1980s.  - 11 - Table 1.4: Savings and Growth RatesA' Real GDP National Personal Foreign Year Growth Savings Rate Savings Rate Savings Rate 1970-74 5.1 23.2 8.3 -4.7 1975-79 6.5 25.2 8.9 4.6 1980-85 0.4 20.2 3.1 4.3 1986-90 4.2 16.5 3.4 0.4 Y Simple averages of ratios to GDP for all savings rates; note that data correspond to the 1972-based national accounts. Source: National Income Accounts Figure 1.4 National Savings & Investment Rates S0 25 p e 20, r 35____--0 - > - 0 n t 10 6- 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 858 87 88 89 90 -- Savings 1 Investment  - 12 1.25 The above summary of trends in foreign savings has several implications fcr the design of a medium-term development strategy for the Philippines. Future external financing is likely to consist primarily of official flows. Some voluntary private commercial flows may also arise in the context of debt-rescheduling and debt-restructuring exercises, but th-se are likely to be minor when viewed in the medium term perspective. As a result of this as well as a consequence of the debt-servicing profile on official loans accumulated over the last two decades, the level of future external financing, measured as the ratio of net flows to GNP, is likely to be sharply lower than was the case in the late 1970s. Such flows will not support an investment program of dimensions similar to those observed in the 19708. This suggests several options: the investment program could be kept small; more attention could be paid to investment efficiency so as to get a bigger bang from a smaller program; direct foreign investment could be encouraged; and, finally, alternative sources of financing, such as domestic borrowing or domestic revenue mobilization, could be harnessed. Savings Trends 1.26 National Sav'-ngs. The pattern that dominates the behavior of national savings over the period 1970-1990 is one of a high rate during the 1970s followed by a significantly lower rate during the 1980s. The national savings rate averaged around 25Z in the 1970s and only around 182 in the 1980s. The Philippines went from being a high savings economy to being a low savings economy, especially in relation to its ASEAN neighbors. The pattern of savings behavior is highly correlated with that of real economic growth, suggesting that it was the collapse of the growth rate and the decline in per capita income levels in the 1980s that led to the drop in the savings rate. For example, a major break in the savings rate occurred in 1981, heralding the trend observed through the rest of the decade; this break closely follows a major break in the growth rate in 1980. This relationship is corroborated by the pattern of income and savings behavior in the Philippines relative to that in other ASEAN countries: the Philippines experienced the biggest drop in real GNP growth rates as well as the biggest drop in national savings rates. Table 1.5: Comparative Savings and Growth Trends Saving Rates 1/ Growth Rates 2/ 1973-80 1980-87 1973-80 1980-87 Philippines 25.0 18.3 6.3 -0.6 Malaysia 28.8 26.9 7.5 4.4 Thailand 21.4 20.3 7.5 5.9 Indonesia 28.2 26.3 7.2 5.1 1/ Refer to current ratios of gross national savings to GDP. 2/ Refer to growth in real GDP. Source: SAVEM Tables, May 1991, World Bank  - 13 - 1.27 Econometric analysis confirms the importance of income growth to the generation of higher savings in the Philippines. It also reveals a negative relationship between national and foreign savings (Lamberte, et.al., 1991). This suggests that the easy availability of foreign savings in the late 1970s weakened the domestic savings effort and that the resumption of higher aid flows in the late 1980s may have a similar effect. To assess this result properly it is necessary to clarify two issues. First, what is cause and what is effect? Do foreign savings actually lower domestic savings or is the statistical correlation simply a reflection of the fact that foreign savings might come in when domestic savings are low. In the case of the Philippines, the relevant evidence suggests that domestic "need" in the sense of low per capita income or low savings was not the key determinant of foreign savings inflows. In the late 1970s, the period of sharp growth in foreign savings inflows, national savings rates were high and were supporting a reasonable level of economic growth. Furthermore, a large proportion of the relevant inflows originated in private sources and was not driven by the same motives as aid flows might. Indeed, these private flows may have been supply-driven in the sense that they represented an attempt by commercial banks to place rising "petrodollar" deposits. 1.28 A second point that needs to be clarified is the mechanism by which foreign finance might have depressed domestic savings. The most common intuitive explanation is that easy credit leads to a relaxation of the revenue-raising effort by the public sector while encouraging the undertaking of higher expenditures. This would depress public sector savings. The evidence for such a 'complacency effect" in the case of public savings is discussed below. 1.29 Public Savings. Public savings may be broken dorAn into the savings of the national government and that of public enterprises. Government savings rose through the 1970s and then began a gentle decline in the first half of the 1980s which turned into a collapse in the second half. The explanation for this pattern lies in several developments. The rise in the 1970s reflected a strong revenue effort brought about by changes in tax administration procedures: revenues fluctuated around 13.5% of GNP during 1975-80. The gentle decline in the first half of the 1980s reflected a deterioration in the revenue effort because of the deterioration of the economy; this decline would have been sharper had it not been for the fact that government expenditures were also cut significantly in this period. Finally, the collapse of government savings after 1985 reflects essentially the impact of higher interest payments on current expenditures; the sharp rise in current expenditures that resulted from the assumption of the foreign debt liabilities of various public and private entities offset even the effects of a much improved revenue effort over this period. 1.30 As far as national government savings are concerned, there does not seem to be a strong negative link between foreign savings and the national revenue effort. When foreign savings inflows were high in the late 1970s and again in the late 1980s, the revenue effort was high rather than low as might be expected from the complacency theory (see Figure 2.4). Rather, the operative link is between foreign savings and government expenditures: foreign savings were used to finance higher levels of current and capital outlays rather than to relax the revenue drive.  - 14 _ 1.31 Public Enterprise Savings. Non-financial public enterprises appear to have operated more in keeping with the "complacency" hypothesis: their savings, while generally low, fluctuated with the availability of foreign finance. When foreign finance was plentiful in the late 1970s, their savings hovered around 0%. As foreign finance became scarcer in the 1980s, savings performance improved; the savings rate rose from -0.8% in 1980 to 1.5% in 1988. Notwithstanding the above, the bulk of foreign finance was used to fund capital expenditures and not current expenditures. Financial public enterprises (i.e., government financial institutions) were also affected by trends in external finance, but in their case, there was a substantial drop in savings in the crisis years of 1984-87 as they were overwhelmed by a rising tide of loan defaults. Personal Savings: Trends and Determinants 1.32 Savings and Income Growth. Personal savings rose through most of the 1970s, from around 7% in 1970 to almost 11% in 1976; they then declined to 6.5X by 1981 and further collapsed to virtually 02 by 1987 before staging a modest and uneven recovery thereafter.1 Effectively, the Philippines went from being a relatively high-savings country in the 1970s to a relatively low-savings country in the 1980s, whether one looks at national savings or at personal savings. While slowly-evolving demographic factors, such as urbanization and changes in age structure, have played a role in shaping the long-run trend of the personal savings rate, macroeconomic factors have been the dominant determinants in the last 15 years. Among these factors was the collapse of income growth in the 19808; the effect of this on the personal savings rate may be confirmed both by visual inspection of trends (see Figure 1.7) and by rigorous statistical analysis. Clearly, then, the personal savings rate cannot be expected to recover to the average level experienced in the early 1970s without a substantial and sustained recovery of income growth. On the other hand, sustained income growth is doubtful without a higher level of savings to fund the required investment effort. There is, in effect, a vicious cycle contained in the endogenous relationship of income growth and savings. Breaking out of this cycle will require a consistent and credible macroeconomic policy regime sustained over a number of years that will raise incomes while reassuring private savers that their financial savings will not be eroded by inflation and currency depreciation. Without such a policy regime, income may not rise in a sustained fashion, and over the period that it does rise it may not generate a corresponding increase in productive savings as private savers, anticipating inflation and currency depreciation, will prefer to increase consumption and/or l/ Personal rather than private savings are analyzed here because the latter, as defined in published National Income Accounts, comprise the savings of three different types of savers, households and unincorporated businesses, public corporations and private corporations. Each of these groups has different behavioral propensities and is affected differently by government policy measures. Accordingly, it is difficult to interpret the behavior of the composite variable defined as private savings. Personal savings refers only to the savings of households and unincorporated businesses, a relatively uniform group as far as behavior is concerned.  - 15 Figure 1.6 Public Sector Savings 6 4 - _ p o 2- r C t 1 -21 76 76 77 78 79 80 81 82 83 84 85 86 87 88 89 - National Government Nonfinancial PEs Figure 1.6 Public Sector Investments 10 e 6 _ _ _ n 4 t 76 76 77 78 79 80 81 82 83 84 86 86 87 88 89 X National Government ^ Nonfinancial PEs  - 16 - put their money in inflation hedges (such as property) or foreign bank accounts (i.e. engage in capital flight). The latter activities, while forms of saving for the individual, are generally not considered to be supportive of domestic investment and growth; indeed capital flight is contractionary in its effects. 1.33 Effect of Foreign Savings. The income link, however, does not explain an important feature of the personal savings trend, this being the substantial decline during 1976-80 when income growth was still high, This is better explained by reference to the effect of foreign savings and changing rural-urban income distribution. Foreign savings may stimulate or depress personal savings depending on the quality of the investment program financed through them and the income expectations thereby generated.2 In the 1970s, foreign savings were used to finance an ambitious public investment program of relatively poor quality. However, private savers were by and large unaware of this fact; they saw income growth being sustained for several years and concluded that such growth was of a permanent nature. Accordingly, they decreased their savings, that is, they borrowed from expected higher future income to increase current consumption. This explains why the personal savings rate declined during 1976-81 even though income growth was high. As the economy deteriorated in the 1980s, and the adverse consequences of the 1970s external borrowing strategy became clearer Figure 1.7 Personal Savings and Real GDP Growth 1 5 'p n 0 _ X / t -10 I . I I . . . . . . . . 70 71 72 73 74 75 76 77 78 79 80 81 8283 84 85 86 87 88 89 90 Real GDP Growth -4- Personal Savings 2/ Another link between foreign and domestic savings operates through financial markets. Large inflows of foreign savings may depress domestic personal savings by depressing domestic interest rates. This link is unlikely to have been quantitatively significant in the late 1970s because deposit interest rates were influenced more by government decree than by market conditions in those years and the interest elasticity of savings is generally low.  - 17 - income expectations related to foreign savings changed but again to the detriment of personal savings; the decline of foreign savings in the 1980s was not met by an increase in personal savings in part because of the apprehension that fiscal constraints thereby generated would continue to depress income growth and destabilize the economy. In any case, the deceleration of income growth was the primary influence over the personal savings rate over this period. Investment Trends and Determinants 1.34 Aggregate Investment. Aggregate investment behaved similarly to national savings. There was a sharp increase during the 1970s (see Figure 1.4), when investment rates climbed from around 21% in the early part of the decade to around 30% in the latter part. This was followed by a sharp drop in the 1980s, to a low of 13% in 1986. The behavior of investment is related closely to the availability of funds: investment rates were high in the 1970s as both domestic savings and foreign savings (in the form of external finance) were high; investment rates collapsed during the 1980s as both external finance and national savings declined sharply. Indeed, during much of the latter half of the 1980s, aggregate investment has been constrained by the level of national savings as foreign savings have been low or negative. 1.35 Components of Investment. The share of the private sector in aggregate investment declined steadily through the 1970s (from 93% in 1970 to 74% in 1979), with an especially sharp fall in 1975-1'176 (from 82% to 752); it stayed more or less constant in the early 1980s (around 75Z) and has recently begun to rise (from 752 in 1986 to 81% in 1989). This pattern is consistent with the story told at various points above, that is that the role of the public sector expanded during the 1970s (fed by plentiful external financing) and declined during the 1980s (as external finance became scarcer); the more recent pattern is related also to the public sector rationalization program implemented by the Aquino Administration since 1986 which has sought to privatize several government corporations and to force many others to rely to a greater degree on internally generated funds to sustain their investment programs. 1.36 Private Investment. Private investment followed trends in overall investment with one difference; it began to decline earlier, in 1975, while overall investment was still high, thus mimicking the relationship between personal and national savings.3 Private investment declined in importance during the 1970s as its share to total investment declined while public investment grew rapidly both in level and share. This trend was reversed in the recovery following 1985. 31 The remarkable correspondence of the trends in private savings and investment (they both begin declining in 1975-76) suggests segmentation of the credit market, that is, that private investment could only be funded from private sources of credit. While this is a tempting hypothesis, it is not supported by evidence from credit distribution data. A more supportable hypothesis is one that links private investment to public investment through a crowding out process.  - 18 - 1.37 Trends in expected demand are usually the most important determinant of investment. If demand is expected to be strong and rising, investors are keener to put in the capacity to meet this demand. On the other hand, if demand is weak and falling, investment is typically put off. Empirical analysis confirms the importance of expected demand, as proxied by trends in GNP growth, to investment decisions in the Philippines. Thus, the sharp decline of the private investment rate in the first half of the 1980s and the recovery thereafter is consistent with the behavior of GNP over this period. However, this relationship, while important, does not yield a policy lever since income growth is an outcome not an instrument, and since investment and income are endogenously related. Clearer policy implications may be obtained from some other determinants of investment, such as the relative price of capital, the nature of industrial relations and the characteristics of public investment. 1.38 Relative Price of Capital. The relative price of capital depends on the price of investment goods relative to the price of output and the real interest rate. The decline of private investment since the mid-1970s does not appear to have been caused by an increase of the relative price of investment goods but rather by an increase in the cost of borrowing. While the domestic price of investment goods rose faster than the GDP deflator on average during the post-war period, it actually declined between 1975 and 1980. In particular, the effect of the large step devaluation of 1972 on the relative price of investment goods appears to have been exhausted by 1974. More recently, however, the large increase of the relative price of investment goods resulting from the exchange rate devaluations of 1983-84 reinforced the effect of the recession and contributed substantially to the drop of the private investment rate in 1984-86. 1.39 Data on lending rates suggest a substantial increase in real rates in the early 1980s and again in recent years; this has contributed to the rising relative price of capital since the mid-1970s and to the consequent decline of the private investment rate. Indeed, the increase of the relative price of capital in recent years has probably been the most important deterrent to the full recovery of private investment. This increase is related in turn to the high and rising public sector demand for credit in recent years. 1.40 Interest rate volatility has also increased in the 1980s in part due to the increase in uncertaintZ caused by the 'debt overhang"; this is reflected in volatility in inflation and exchange rates as well. Since investment is to a large extent irreversible, it is negative.y affected by volatility of key determinarsts such as interest rates. It is plausible to argue that the 'debt overhang' has kept private investment from growing at a faster rate. 1.41 Crowding Out Effect. There is a large body of evidence indicating that both the level and the composition of public investment affect private investment. There is evidence of this also in the Philippines. Although available data do not allow a clear distinction between infrastructure and non- infrastructure components of public investment until 1985, the assumption that most government investment has taken the form of infrastructure and that a substantial part of public corporations investment has taken the form of other productive activities is thought to be accurate, at least until 1988, when the  - 19 - government started to privatize and restructure the public corporate sector.4 Under this assumption, the government component of public investment -- i.e. infrastructure -- has provided a positive stimulus to private investment throughout the period, and in particular in 1975-1983, when it was substantially higher than in the rest of the post-war period. On the other hand, the public corporations component of public investment -- i.e. non-infrastructure -- has affected private investment negatively, in particular in 1976-81, when it averaged in excess of 5 per cent of GDP. Table 1.6: Selected Investment Data Real Gross Private Sector Investment Rate a/ Share bl ICOR cl (percent) 1970-74 22.2 88 4.2 1975-79 30.1 75 5.0 1980-85 24.7 75 8.4 1986-89 16.2 78 4.2 a/ Refers to simple averages of ratio of investment to GNP hb Refers to simple averages of ratio of private to total gross domestic fixed capital formation c/ Simple averages of incremental capital output ratios. 1.42 Trends in Foreign Investment. Unlike other ASEAN countries, foreign investment has not been a sizeable component of private investment in the Philippines. It did not exceed the level of 52 of private investment until the debt conversion program was initiated in 1987, and even then did not exceed 8 per cent. Meanwhile, over the last twenty years or so, foreign investment into the ASEAN region increased dramatically, first towards Singapore and Malaysia and more recently towards Thailand. Of the total inflows in the ASEAN region, the Philippines' share was the lowest averaging only 4.5 percent during the period under study. In contrast the averages for Indonesia, Malaysia, Singapore and Thailand were 15.4 percent, 30.3 percent, 40.0 percent and 9.8 percent, respectively. 4/ This should not be taken to imply that public corporations did not undertake infrastructure activities; among the largest public corporations, NPC and MWSS have carried out most of public infrastructure development in power generation and water and sanitation. Other corporations, as for instance PNOC and NSC, have however undertaken exclusively industrial activitiee.  - 20 - Figure 1.8 Net Foreign Investment Trends UWS Billion 12 10 - _ _ 8X 6 : 4 2 0 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 M Indonesia E Malaysla a Singapore e Thailand P Philippines 1.43 Foreign investment inflows in the Philippines appear to have responded positively to high income growth and competitiveness, as measured by the real exchange rate and the openness of trade policy. The important role played by the domestic growth rate is consistent with the prevalent domestic orientation of foreign firms according to BOI approval data. The positive response of foreign investment to the adjustment in exchange rates in 1970 and in 1983-84 -- in the latter case, anticipating the recovery of private investment -- and to import liberalization in 1981 was important as well to such flows. In addition, competitive real wages appear to have affected foreign investment. 1.44 The regulatory framework has also been of importance in shaping the direction and size of foreign investment flows into the Philippine economy. Entry to majority-owned foreign firms is prohibited in several areas.5 In addition, until the .enactment of the new Foreign Investment Act in April 1991, entry in the unrestricted areas was subject to approval by the Board of Investments. According to the pre-1991 regulations, entry to majority-owned foreign firms was to be allowed in export-oriented industries or in areas "not adequately exploited' by Filipino nationals, leaving substantial margins of discretion to BOI -- and uncertainty to foreign investors. In addition, foreign investors need to register with the Central Bank in order to be able to repatriate earnings or equity at a later date. It is expected that the new 5/ The Constitution limits to 40 per cent foreign ownership of land, and enterprises active in the exploitation of national resources and the operation of public utilities. The Nationalization Laws further restrict foreign ownership --. between 0 and 40 per cent -- in sub-sectors of the following industries: banking, government contracts, shipping, construction, and trade. Entry of foreign-owned firms is prohibited in fishing, rice and corn farming, retail trade, and mass media.  - 21 Foreign Investment Act will reduce the procedural difficulties faced by foreign investors and thereby encourage a larger flow over the 1990s. 1.45 Investment Efficiency. Perhaps the most critical aspect of investment in the Philippines is its efficiency. Developments in investment efficiency may be gauged, albeit in a rough fashion, from the behavior of incremental capital output ratios (ICORs). The data show that ICORs rose in the late 1970s, then increased dramatically through the mid-1980s, before declining sharply over the period 1988-90. The increase in the 1970s was a result of the public sector orientation of the investment push undertaken in the latter half of the decade; insulated from competitive pressures, such investment was not very efficient. The dramatic increase of the ICOR level in the 1980s was due to the collapse of real growth in the economy; the investment surge of the late 1970s failed to produce a basis for sustained growth and, when external finance was reduced in the mid-1980s, output growth collapsed. The decline in ICORs during 1988-90 reflects the effect of rising capacity utilization; output increased without large increases in investment. As of 1990, however, the slack in capacity utilization appears to have been exhausted; higher ICORs, reflecting a more typical level of investment efficiency for the Philippines, may be expected in the future, unless reforms are undertaken to ensure that incremental investment is of higher quality than before. 1.46 The sectoral allocation of external finance can also shed some light on the issue of efficiency. Table 1.7 shows that external finance tended to go into the non-traded goods sectors during the late 1970s but there was a shift away from these sectors after 1983: for example, from 1975-83, the share of non- traded goods sectors in the allocation of externally borrowed funds rose from 53Z to 812; thereafter, it declined to 432 in 1988. To the extent that the non- traded sectors are insulated from competition and display characteristics of relatively higher cost and lower efficiency as compared to the traded-goods sectors, one might conclude that investment efficiency was lower in the late 1970s than in the late 1980s. Table 1.7: External Debt Citstanding, by Sector (Percent) 1975 1977 1980 1983 1986 1988 Total, Economic Sectors 76.4 85.6 86.6 90.6 57.5 50.5 Traded Goods Sectors 23.1 17.1 9.9 10.0 8.4 7.7 Non-Traded Goods Sectors 53.3 68.5 76.6 80.6 49.1 42.8 Other Allocations 23.6 14.4 13.4 9.4 42.5 49.5 Total 100.0 100.00100.0 100.0 100.0 100.0 Source: Adapted from Lamberte et. al. (1991)  - 22 - 1.47 Investment and Tax Generation. The conventional justification for external borrowing is based on high returns to domestic investment. In the case of public borrowing, not only must the investments undertaken hasa a return higher than the cost of borrowing but they must be deployed in such a fashion that these returns accrue to the national treasury in the form of tax revenues. The available data show that the sectoral allocation of borrowed funds was not consistent with the objective of securing adequate flows of tax revenues: relatively undertaxed sectors obtained the bulk of the available external finance. For example, in 1979, over 60Z of outstanding external debt was attributable to the agriculture, power and services (transport, water, trade, defense and public administration) sectors; these sectors are all relatively "undertaxedt in the sense that their contribution to tax revenues is less than their share in GDP (see Table 1.8). On the other hand, manufacturing is an "overtaxed" sector but it accounted for only 6.6% of the outstanding debt. The allocation of borrowed funds changed over the 1980s in a manner more supportive of revenue generation: by 1987, only about 37% of the outstanding debt could be attributed to the three "undertaxed" sectors noted earlier. Table 1.8: Debt and Taxes Sector Taxation Share of Outstanding Index 1/ External Debt 1979 1987 1979 1987 Agriculture 0.3 0.4 2.9 3.2 Manufacturing 1.8 1.3 6.6 4.5 Power 2/ 0.5 0.5 28.4 14.2 Services 3/ 0.7 0.5 28.7 16.4 1/ The index refers to the ratio of the sector's share in total taxes generated to its share in GPP. 2/ Refers to electL:- .cy, gas and steam production. 3/ Refers to transport, communications, water, sanitation, trade, defense and public administration. Source: Adapted from Lamberte et al. (1991) C. Implications for Development Strategy 1.48 It is clear from the foregoing discussion that economic development strategy for the future should target both stabilization and adjustment. A stabilization program is again underway with the support of the IMF. Some problems in its implementation thrt have been identified here, such as the reconfiguration of electricity tariffs, have been resolved. Other problems, such  - 23 - as the legislation of permanent tax revenues to replace the special import levy, have not yet been resolved. The task for short-run management is clear: it is to continue with the existing stabilization program to meet the targets set for fiscal, monetar;' and external accounts. 1.49 The task for medium-term management is also clear: it is to pursue an adjustment program that will enable the economy to break away from the structural constraints that have repeatedly hobbled its performance. The discussion in preceding sections has provided a sense of what some of these constraints are and how they have operated. The lessons of the last fifteen years of macroeconomic experience may usefully be summarized as follows: (a) Growth financed through external debt is difficult to sustain. Servicing costs are exogenously determined and can rise rapidly while domestic policy flexibility is constrained by the need to finance the debt. Export orientation is a must in order to generate the foreign exchange to service the debt. The combination of high foreign borrowing and a protected, import-substituting economy can be disastrous. Moreover, foreign savings has tended to depress domestic savings, especially personal savings. (b) Growth financed through domestic debt is also difficult to sustain, primarily because the domestic capital market cannot smoothly absorb large amounts of debt in its present state. Interest rates rise fast to levels that exceed the returns to be reasonably expected from public investment. A wider and deeper capital market is needed before a domestic borrowing strategy can work. (c) Ensuring investment efficiency is critical to a growth strategy, especially if it is financed by debt. Inefficient investments were the nub of the economic crisis that hit the Philippines in the early 1980s. Expensive money was spent unwisely; the expenditures did not generate the resources with which to pay back the debt. 1.50 These lessons suggest that the way forward for the Philippines lies in a development strategy that emphasizes non-debt sources of financing and that seeks to improve the efficiency of investment, especially public investment. The elements of such a strategy are reviewed in the rest of this report. 1.51 Chapter 2 provides a discussion of public sector resource mobilization. The key elements of the tax system of the Philippines are highlighted and problems of structure and administration are identified, leading to recommendations for enhancing revenue collection. Chapter 3 provides a discussion of public resource management on the premise that sustainable growth requires not only a differently-configured public finance base but also a system that improves the returns from public expenditures. The elements of such a system are investigated through an examination of such issues as the productivity of public investment, the potential role of the private sector in implementing a public investment program, and the scope for greater efficiency in the financing and management of public enterprises.  - 24 - II. PUBLIC RESOURCE MOBILIZATION 2.1 From the point of view of resource mobilization the key question with respect to the public revenue system is whether or not it generates sufficient funds to finance the desired development effort, given certain constraints on Government's ability to borrow. In the case of the Philippines the public revenue system does not generate adequate funds. This chapter explains why. 2.2 The structure of this chapter is as follows. The first section provides, by way of introduction, an overview of the public revenue system, focussing on the principal taxes and their key characteristics. The second section provides evidence of the system's underperformance through a discussion of tax effort and of the gap between actual taxes collected and the tax potential in the Philippines. The third and fourth sections investigate the causes of low collection of internal revenues, highlighting deficiencies in both structure and tax administration. The fifth section discusses problems in customs administration. The sixth section diGeusses the revenue experience of privatization while the final section reviews proposals for enhancing revenue collection. A. Overview of Public Revenue Syste 2.3 An overview of the public revenue system might well start with a reference to the Tax Reform Package (TRP) of 1986. The TRP comprised a comprehensive effort to improve the efficiency and equity of the tax system without incurring losses in revenue. The TRP is widely regarded to have been successful especially in improving efficiency and revenue productivity. The accompanying box provides details of the program and its achievements. Structure of Taxes 2.4 Public revenues are raised from three main sources: internal revenues (55Z), trade taxes (25%) and non-tax revenues (20%). The chief sources of internal revenues are income, VAT and excise taxes. Import tariffs are the main trade tax, export taxes having been abolished several years ago. Among non-tax sources of revenue, the two most important are miscellaneous fees and charges and interest income from government deposits and advances. 2.5 In common with most developing countries, the Philippines raises the bulk of its public revenues from indirect taxes. In recent years (1986-90), indirect taxes have accounted for 682 of total tax revenues. This marks a shift in tax structure from earlier years (1976-85) when the share of indirect taxes was almost 5 percentage points higher (see Figures 2.1 and 2.2). This is largely due to improvements in the design and collection of direct taxes following from the tax reforms of 1986. In particular, it appears that the imposition of a final withholding tax on passive incomes from financial assets has improved collections. An additional factor behind the observed shift in tax structure could be the secular, though modest, decline in nominal tariffs during the 1980s.  - 25 _ BOX 2.1s TAX REFORM PACKAGE OF 1986 While tax reform has been a continuing activity In the Philippines, typically brought on by periodic flscal and economic crises, a major and notable attempt, generally referred to as the Tax Reform Package (TRP), was undertaken In July 1988 by the lncoming Aquino Administratlon. This reform attempt was comprehensive In scope and aim: It covered a wide range of revenue sources and tax administration measures and sought to achieve lmprovwments In the revenue productivity, efficiency and equity of the tax system. With respect to direct taxation, the TRP introduced the following major changoe: a. Rationalization of basis for Income taxation b Increase In final tax on passive incomes ce Elimination of double taxation of dividends d. Abolition of surtax on corporate incomes e Increase In personal exemptions f Setting of a flat 8a% corporste Income tax rate With respect to Indirect taxation, the TRP introduced the following major changes: a. Introduction of a consumption-type VAT of 10% b. Simplification of tax rates applying to public utilities c. Conversion of some excise taxes from specific to ad valorem basis d. Abolition of export taxes, except on logs e. Reduction of tariff rates f. Withdrawal of tax exemptions for most government and private entities With respect to tax administration, the TRP introduced the following major reforms: a. Croation of VAT administration units within SIR system b. Adoption of zonal valuation system for property tax matters c. Grant of tax amnesties covering unpaid Income and property taxes d. Administrative settlement of delinquent accounts *. Setting of clearer guidelines for selection of audit cases The TRP of 1988, and refinements thereafter, have improved the revenue productivity and efficiency of the tax system; this Is evident In the sharp improvement In tax revenue to GNP ratios in recent years (from 10.26% in 1986 to 13.7% In 1990) and In the declining cost of collection for BIR (from 1.1% in 1986 to 0.8% in 1989). VAT has rapidly become a significant source of revenues; moreover, it is yielding more than would have been expected from the taxes it replaced. While the grant of tax amnesties and the administrative settlement of delinquent accounts brought in less than expected, the amount collected (around 8 billion pesos during 1986-1990) was unprecedented; at the same time, these measures cleared a backlog of contested cases and released personnel for other functions. As far aS equity is concerned, however, the Impact of the TRP has not been significant: by most accounts, the pattern of tax lncidence, which Is generally regressive, remains unchanged from 1985.  - 26 - Table 2.1: Structure of National Government Revenues (#) Revenue Sourceo/Year 1976 1978 1980 1982 1984 1986 1988 1990 Direct Taxes 28.82 28.08 24.71 27.48 28.07 81.48 82.64 88.40 Income Taxes 24.89 24.44 28.26 24.39 26.69 28.94 80.28 81.85 Individual 10.22 18.11 10.99 11.11 8.88 8.98 8.77 10.41 Corporate 18.98 11.83 11.97 12.90 15.88 12.97 14.82 12.20 Financal 1/f 0.83 1.93 6.48 7.16 9.26 Miceallaneous V 3.18 2.70 2.29 6.27 4.26 4.79 4.42 2.98 Ind;rect Taxes 78.68 78.92 75.29 72.62 71.98 88.62 67.88 88.60 Excise Taxes / 17.16 18.82 16.98 17.10 18.69 24.96 22.14 18.72 Licence and Business Taxes 4/ 18.79 21.75 24.88 22.82 17.82 18.50 19.49 20.51 Imprt Duties S/ 27.90 26.64 24.65 24.77 27.01 19.71 19.42 21.:3 othre I/ 9.83 7.90 9.84 8.40 8.99 S.86 6.82 6.56f 1/ Includ, taxes on stock transactions, comnercial paper and bank deposits. 2/ Include transfer taxes, real property taxes, motor vehicle and trovel taxes. 8/ Include collections of the Bureau of Customs and tobacco Inspection fes. / Incl udo eo lectlons from Import trade, f I xed taxes, and selective percentage taxes (such as VAT). Include documentary stamps, sale of confiscatod ;oods and other income of the Bureau of Customs. / Include export and premium duties, documentary stamp taxes, franchise taxes, forest charges, and miscellansous Items. Source: National Tax Research Center The Principal Taxes 2.6 Individual Income Tax. While not the most productive single tax, this is an important tax: collections from this source amounted to 10.4% of total tax revenues in 1990. Among the notable characteristics of this tax are the following: (a) The top marginal rate is 35%; this is at the lower end of the range of rates observed in ASEAN countries. For example, Malaysila and Thailand have top rates of 40% and 55% respectively. Only Singapore, at 33%, has a lower rate. (b) Income derived from employment is treated the same, as far as tax rates and exemptions are concerned, as income derived from businesses and professions. However, the latter category enjoys the benefit of deductible business expenses, a benefit that is typically abusbd and that results not only in a lower tax take for the authorities but also in some inequity in the burden of taxes between those who are employed (for wages and salaries) and those who are self-employed. (c) Income derived from different sources is generally combined into one "global" figure for tax purposes, rather than being made subject to different "schedular" tax rates. One major exception to this rule is  - 27 - Figure 2.1 Structure of National Government Revenues Direct Taxes Direct Taxes Indirect Taxes Indirect 68% 1975-1985 1986-1990 Figure 2.2 Trends in Direct and Indirect Tax Shares 60 p e r e n t 20 -- . - - .- - - - - - - - - 7677 7879 80 81828384 8686 87 8889 90 nDirect Taxes * Indirect Taxes  - 28 - the assessment of a final withholding tax of 20Z on passive income, derived from such sources as interest earnings. 2.7 Perhaps the most notable aspect of this tax, however, is how little it generates relative to potential. For a variety of reasons, explored in more detail in the next section, this tax is an underperformer: it brings in far less than would be expected from the general parameters of income and employment growth in the Philippines. 2.8 Corporate Income Tax. This tax accounted for 12.22 of total tax collections in 1990, a share which has -luctuated very little in the last fifteen years. In 1986, the corporate tax was set at a uniform rate of 35Z for all taxable income: before that, two rates (25% and 35Z) had applied to different levels. In terms of tax burden, the present rate compares favorably with other ASEAN countries; in fact, it lies at the lower end of the range for these countries. Until 1986, corporations were also subject to a 25Z surtax on net income if they were found to be "improperly accumulating profits" by maintaining a low dividend payout policy. This surtax was abolished in the tax reform package of 1986; it had become redundant since the phasing out of the tax on dividends (done simultaneously via the TRP of 1986) removed a major impediment to the distribution of dividends. 2.9 As with the individual income tax, perhaps the most notable aspect of this tax is its underperformance relative to potential. For a variety of reasons, discussed in more detail in the n,jxt section, this tax also brings in far less than one might expect. 2.10 The Value Added Tax. The VAT was introduced in 1988 as a replacement for a number of complex, cascading sales taxes and is considered to be a worthwhile innovation in the Philippine tax system. According to Manasan (1990), it has improved both the efficiency and equity of the tax system by reducing differences in tax rates across commodity and input types and increasing, albeit modestly, the progressivity of sales taxes.1 2.11 From the standpoint of resource mobilization, the attractiveness of VAT depends on three considerations: its importance as a source of revenue, the extent to which it has generated more revenues than the taxes it replaced, and 1/ Manasan (1990) examined the efficiency of the VAT by calculating effective tax rates (ETRs) on the consumption of goods and services. Each ETR measured the direct and indirect taxes passed en to each unit of consumption of a commodity or service from successive layers of taxes on outputs and inputs in the production and sales process. She found that, after the introduction of VAT, (a) the average ETR was more than halved (from 14.4% to 6.52); (b) the dispersion of ETRs was substantially reduced (from a range of 1.3Z-34.6Z to 0.4Z-13.3Z); and (c) the difference between the nominal and effective rate - a strict measure of the extent to which inputs are taxes - was more than halved (from 7.2% to 3.3%). Zanasan also examined the equity consequences of the introduction of VAT by means of the Suits index, a summary measure of progressivity: this measure was found to decline marginally.  - 29 - the extent to which it is approximating its potential yield. The first two characteristics are discussed below while the third is considered later. 2.12 Table 2.2 shows the revenue of VAT relative to other taxes of the national government as a percent of GDP at market prices. In the first year of implementation, VAT revenue amounted to 1.6% of GDP. However, this represented collections over a 9 month period only; full year collections would have amounted to about 2.12%. During the next two full years of implementation, 1989 and 1990, VAT revenue amounted to 2.07% and 2.28% of GDP, respectively. These figures compare favorably with other sources of tax revenue reported in Table 2.2. For example, VAT collection was 57Z to 602 of the contribution of import duties (which amounted to 3.62 to 3.8% of GDP) and over 16% of the total tax effort (which stood at 12.4 to 14.12 of GDP). 2.13 The usefulness of VAT is partly a function of its revenue productivity relative to the taxes it replaced. The revenue from the indirect taxes replaced by VAT was estimated by NTRC (1990) at P 15.8 billion in 1987, the year before the reform. This comprised about 2.24% of the GDP of the same year. Using this criterion, VAT revenue in 1989 failed to replace the revenue of the other taxes by approximately 7.6%. In 1990, however, it exceeded the neutrality benchmark by 1.8%. Once administrative difficulties are resolved and the coverage of the VAT is increased, this tax will become one of the more stable and potent sources of revenue. As discussed in a later section, present collection is roughly half of the potential intake from VAT. Table 2.2: Revenue Importance of VAT (Percent of GDP) 1988 1989 1990 VAT 2.12 2.07 2.28 Import Duties 2.86 3.63 3.77 Other Indirect Taxes 3.46 3.01 3.61 Direct Income 3.35 3.67 4.44 Total Tax Revenue 11.26 12.38 14.10 Source: Bureau of Internal Revenue 2.14 Excise Taxes. Excise taxes are an important source of revenue; they have typically been the single most important category of revenue after income taxes. In 1990, excise taxes contributed around 19% of total tax revenues. While excises apply to numerous commodities, the bulk (around 90Z in 1990) of the tax take comes from only three product groups: alcohol, tobacco and petroleum. In various reform measures during the early 1980s, the form of most excise taxes was changed from "specific" or per physical unit to "ad valorem" or per value. This resulted in a steady increase in the share of excise to total revenues, as excise collections increased with economic growth and inflation. In 1990,  - 30 - however, the excise on petroleum products was changed back to a (lower) specific tax in a move designed to cushion the impact of rising oil prices; this resulted in a noticeable decline in the share of excise to total revenues. 2.15 Import Duties. In 1990, import duties contributed almost 22% of total tax revenues. While they remain a significant source of revenue, their relative importance has declined substantially: during 1976-85, import duties averaged around 262 of total revenues; this share fell to around 20X during 1986-90. This shift can be explained in part by the tariff reform program undertaken in the 1980s, which has reduced average tariff rates, as well as by the operation of the BOI incentives program which has increased the scope of duty exemptions. This shift is also reflected in the behavior of effective tariff rates (total duties collected/total import value) which have declined steadily during the 1980s (see Table 2.3). Table 2.3: Effective Tariff Rates: 1980-90 (Billion Pesos and percent) Particulars 1980 1982 1984 1986 1988 1989 1990 TOTAL IMPORTS 48.1 64.1 99.7 98.7 137.2 225.2 290.1 Dutiable Imports 37.9 54.3 61.8 70.7 103.1 148.0 187.2 Non-Dutiable Imports 10.2 9.8 38.0 28.0 34.1 77.2 102.9 Total Duties Collected 11.5 12.5 19.1 17.7 25.0 38.4 46.8 Effeutive Tariff Rates (Total Duties Collected/Total Dutiable Imports) 30.3 23 30.9 25.1 24.3 25.9 25 Effective Tariff Rates (Total Duties Collected/ Total Imports) 23.8 1.9.4 19.1 18 18.2 17.0 16.1 Source: Bureau of Customs Evaluation of Philippines Tax Structure 2.16 One way to evaluate the Philippine tax system is to compare its key features to benchmarks provided in a recent World Bank paper entitled: Lessons of Tax Reform. A comparison is provided below.  - 31 World Bank Reference Philippines System Taxes on Firms and Individuals a. Personal income taxes should be a. Maximum individual tax rate is 35%; subject to maximum marginal rate of tax base, however, is small. 50% and should be raised from wide bases b. Exemption levels should be high b. Exemption level of P12000 per (preferably twice per capita GDP) so married person (set in 1986) is low; as to exclude low income persons it is below per capita GDP as well as below annual income derived from daily minimum wage c. System should feature gradually c. Marginal tax rate schedule for rising marginal individual income tax individuals is progressive, but not rates steeply so d. Withholding taxes should be used d. Withholding taxes are applied to for wage and interest incomes, and wage and interest incomes; however, presumptive taxation should be applied self-employed escape tax net easily to hard-to-tax groups and presumptive taxation not used e. Company income taxes should have e. Corporate income tax has unitary single statutory rate comparable to rate of 35%, which is identical to ataximum personal income tax rate maximum personal income tax rate f. Tax preferences should be limited f. Tax exemptions are relatively broad in coverage and duration to cases in scope and long in duration; their where market imperfections cannot be justification in terms of market addressed more directly imperfections is weak g. Deductions and exemptions should be g. While statutory variation is small, equal across sectors and assets deductions and exemptions vary greatly across sectors in practice h. Indexing provisions should be h. Indexing only done for personal generally applicable exemption levels i. A 10-15% border withholding tax i. Profit remittances generally taxed should be applied to repatriation of at 15. profit, dividend, etc.  -32- World Bank Reference Philippines System Taxes on Goods and Services a. Pr.mary source of revenue should be a. VAT is major, though not primary, a single-rate VAT (between 10 and 20Z) source; VAT rate is flat lOX with with crediting provisions and zero crediting provisions and zero rating rating of exports and goods consumed for exports and food and agricultural largely by the poor products b. Excise taxes (with simple and non- b. Excise revenue largely raised from discriminatory rate structure) should alcohol, tobacco and petroleum; rate be applied to luxury goods structure discriminates against foreign producers c. Level and variation of nominal c.Level and variation of nominal tariffs should be low; exports should tariffs have declined in 1980s and are get inputs at world prices. now relatively low; exports are eligible for duty-drawbacks d. Export taxes should be eliminated d. Export taxes were abolished in 1986 2.17 The above comparison suggests that the basic structure of the Philippines tax system is sound; tax rates are not high and the number of bands is not exces3ive. Furhermore, reliance on indirect as opposed to direct taxes is not out of line with comparable countries; as Figure 2.3 shows, the Philippines is "normal" in terms of its relative reliance on direct taxes in the sense that it lies close to the regression line that best captures the relationship between the direct tax ratio and per capita GNP. 2.18 The we.aknesses of the Philippines tax system lie in the area of collection. Far less is collected than could be. This is so partly for reasons of system design which allow relatively high degrees of tax avoidance and evasion and partly for reasons of tax administration which result in less being collected than is legally due to the tax authorities. These issues are examined in detail in later sections. 2.19 Another shortcoming of the Philippines tax system is its regressiveness; richer people generally pay a smaller fraction of their incomes in taxes than poorer people. This characteristic is not explicitly discussed in this report. Existing studies indicate that while direct taxes are progressive, the combined burden of direct and indirect taxes remains regressive. They also indicate horizontal inequity within the income tax structure as certain types of income, e.g. compensation income, bear a heavier burden. It should be noted that that highly inequitable tax systems tend to fail at revenue mobilization also, typically because they engender a high degree of cynicism and resentment among taxpayers. Often the source of inequity is not the structure of statutory rates but the failure to make richer people pay a "just" share of national taxes through lapses in tax enforcement. This breeds a debilitating public ethic in which it is not considered immoral to cheat on one's taxes because of the system's perceived unfairness.  - 33 - Figure 2.3 Direct Taxes and Per Capita Income, 19a9 80 D e t R . a - ~ Phillpplnea t 20 0 0 a 10 15 20 25. GNP Per Capita (US$ 000) B. Tax Intake and Tax Potential Tax Ratios 2.20 Over the fifteen year period 1976-90, tax ratios in the Philippines (defined as the percentage ratio of tax revenues to GNP) have ranged from a low of 9.8? to a high of 13.7%. In recent years, there has been a marked upward trend (see Figure 2.3) as tax ratios have risen from 10.3Z in 1985 to 13.72 in 1990, the best ever achieved. The rising trend is also reflected in measures of tax elasticity and buoyancy: as Table 2.4 shows, these measures are much higher for 1986-90 than for the ten year period before 1986. The achievements of recent years stem from improvements in both the design and operation of the tax system. The Tax Reform Package of 1986 has clearly been successful in enabling the tax authorities to lay claim to a rising share of national income in the process of growth; this has been achieved, moreover, while keeping tax rates relatively low (e.g. 35Z top rate on income tax) and exemptions relatively generous.  - 34 - Table 2.4: Buoyancy and Elasticity of Major Taxes Type of Tax 1976-1985 1986-1990 Buoyancy Elasticity Buoyancy Elasticity All Taxes 0.89 0.53 1.33 1.07 Individual income tax 0.62 0.54 1.58 1.22 Corporate income tax 1.03 0.85 1.18 1.10 Excise taxes 0.96 0.32 0.78 0.77 License & business taxes 0.79 0.59 1.41 1.07 Import Duties 0.86 0.50 1.52 1.37 Source: National Tax Research Centre 2.21 Despite this positive trend in recent years, the tax effort of the Philippines falls short in two respects. It is lower than what is needed to fund development expenditures, given borrowing constraints. It is also lower than the effort demonstrated in neighboring countries. Figure 2.4 illustrates the first point; it shows that actual tax intakes have consistently been below planned or targeted levels in the last five years. Table 2.5 illustrates the second point; it shows that the Philippine has the smallest tax to GNP ratio (measured as an average over 1976-88) among ASEAN countries. Table 2.5: Tax Ratios in Comparative Perspective Country Average 1976-1988 Indonesia 19.06 Malaysia 22.13 Singapore 17.03 Thailand 13.73 Philippines 11.12 Source: Key Indicators of Developing Asian and Pacific Countries, ADB 2.22 Aniother way to demonstrate the second point, that the Philippines revenue intake is below average, is shown in Figure 2.5 which plots revenue to  - 35 - Figure 2.4 Tax Ratios, 1976-1990 14 -4 I'd p e~~~~~~~~For 2.6 4 - o I 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 Actual Tax Ratio -4 Target Ratio Figure 2.6 4 0 - . ... . . ... ... n 06101202 GNP Per Capita IUS$ 000)  - 36 - GNP ratios for thirty randomly chosen countries against their per capita GNPs for 1980 and 1989. The Philippines lies below the regression line which best captures the relationship between GNP and revenue intakes. Tax Potential 2.23 The tax effort of the Philippines also falls short in a third respect; it generates less revenue than would be expected from considerations of national income and tax parameters. This point can be demonstrated by making estimates of tax potential for different types of taxes. Such calculations are made and discussed below for the following taxes: individual and corporate income, VAT, and import duties. It is best to note at the outset that the calculation of tax potential is subject to much imprecision and methodological uncertainty. Different methods of calculating tax potential can result in widely-varying estimates. Nevertheless, one can still derive useful information from the range and trend of estimates. 2.24 Income Tax Potential. Table 2.6 presents estimates of tax potential for various years for individual and corporate incomes. For individual income taxes, the "gap" approach is used in which the estimate of tax potential is derived by applying statutory tax rates to an adjusted measure of personal income taken from the national income accounts. Adjustments are necessary to reconcile differences in the concept of income that is used in tax returns as opposed to that used in national accounts. For corporations, the estimate of tax potential is derived by applying the average amount of taxes paid per corporate tax filer to the total number of active registered stock corporations (adjusted by a factor of one-half to reflect the typical fraction of tax-exempt corporations). The results are notable. While there is variation from year to year, it appears that the revenue system generally captures about 50X of potential individual income taxes and about 60? of potential corporate taxes; the rest is lost, either to avoidance or evasion. It is quite evident that the potential tax yield is much higher than the actual tax intake.2 2.25 VAT Potential. An estimate of VAT potential is made by first calculating the potential VAT base as the sum of all taxable final demand expenditures. The most important element is the determination of the taxable personal consumer expenditure (PCE) by adjusting for unprocessed, agricultural, in-kind or home consumption, and other VAT-exempt consumption. Under the present guidelines for the Philippines, 52? of PCE is taxable. For the year 1990, this translated to a VAT base of P 522.5 billion Ssee Table 2.7). The corresponding potential revenue was P 52.2 billion or 4.62 of GDP at market prices. Relative to this percentage, VAT collection was short by 55% and 50% in 1989 and 1990, respectively. 21 It might be emphasized that the calculations pertain to the years before 1989. In recent years, tax collections have improved significantly; this may be due in part to improvements in tax administration and a reduction in tax evasion.  - 37 _ Table 2.6: Potential Versus Actual Income Tax Collections Potential Tax Actual Tax Actual/Potential (Million Pesos) (Million Pesos) (Percent) Individual Income Taxes 1984 10826 4476 41 1985 11015 5594 51 1987 7219 4681 65 1988 11805 6264 53 Corporate Income Taxes 1984 13311 8208 62 1985 11407 8441 74 1986 8673 4826 56 1987 13047 7518 58 1988 15083 9875 65 Sources: Manasan (1988) and Avila (1989). Table 2.7: The Potential VAT Base in 1990 1. PCE in 1990 P 843.2 billion less 2. 10Z for overestimation 84.3 equals 3. Adjusted PCE 758.9 4. Taxable PCE (= 522 of 13) 394.6 plus 5. Government consumption 104.5 equals 6. Total taxable consumption 499.1 plus 7. Residential construction 38.0 equals 8. Total domestic taxable expenditure 537.1 plus 9. Tourist expenditure 11.2 equals 10. Total taxable expenditure at market prices 548.3 less 11. VAT revenue in 1990 25.8 equals 12. Potential VAT Base 522.5 Source: For details of calculation see Version B in study done by IMF (Aguirre et al. 1987) in preparation for the VAT scheme.  - 38 - 2.26 It may be argued that the potential VAT revenue as estimated in Table 2.7 is on the high side. One problem is accounting for the exemption from VAT of small establishments with sales of P 200,000 or less. Most of these enterprises are small stores engaged in trading. If we assume that as much as 70? of the gross value added in trade must be deducted from the first estimate, the adjusted tax base is P 365.8 billion and the corresponding revenue is equivalent to 3.2%3 of GDP. Against this more conservative potential, VAT collection was still short by 35? and 29% in 1989 and 1990, respectively. Thus, even conservative assumptions generate a relatively large VAT potential as compared to present intake. 2.27 Moreover, the revenue potential of the VAT is affected by the level of the tax rate and the coverage of the tax base. There is no reason to believe that these aspects of the VAT could not be improved. An idea of the potential here may be gained from an international comparison of VAT experiences. The Philippines performance (revenue as percent of GDP) is still substantially below the numbers of several countries like Chile (8.1?), Brazil (6.5Z), Korea (4.0?), Turkey (3.1?), Mexico (3.2?), and others. VAT collections in these countries were attained by different design features. Some countries have multiple tiers in their VAT rate - for example, Chile (6 rates), Colombia (6), Uruguay (2), Mexico (4), and Argentina (3). Several countries have an upper rate greater than 10? - for example, Brazil (17), Uruguay (20), Mexico (20), Nicaragua (25), and Colombia (35). 2.28 Potential Collection from Imports. There are several pieces of evidence that suggest that the actual collection of import duties is below potential. Export-source data indicate that imports are understated in the Philippines; clearly, many imports enter but are not recorded. Indicative estimates prepared by the Economic Intelligence and Investigation Bureau (EIIB) show that smuggling averaged about 12? of the value of imports during 1986-89. This may be considered a conservative estimate. The prevalence of smuggling is also indicated by occasional physical apprehension of smugglers; the general view is that only a small fraction of the true smuggling is apprehended. The prices of many imported goods in the local market are reported to be lower than what they should have been if the right amount of duties had been paid. C. Issues in Internal Revenue Collection 2.29 Internal revenue collections are lower than their potential on account of deficiencies in both the structure of the tax system and in the tax 3/ This is about 13? higher than the NTRC (1990) estimate. The VAT base calculated in the NTRC study implied a potential revenue equal to 2.8? of GDP. This figure was arrived at from the production side by removing from total GDP the value added from VAT-exempt economic activities. However, half of the remaining sectors, including mining and manufacturing, were excluded to represent small establishments - a big subtraction. Also, no adjustments were made for the VAT-exempt value added of exports.  - 39 - administration system. This section provides a discussion of the major deficiencies. For clarity in exposition, the discussion of structural issues is separated into two parts, one dealing with the main sources of tax evasion and the other focussing on tax base erosion due to exemptions provided to industries and government corporations. Structural Issues 2.30 The tax base in the Philippines is much smaller than it could be on account of several factors including a high degree of non-filing of tax returns, considerable underdeclaration of income, exaggeration of allowable deductions, and relatively generous tax exemptions provided to industries and government corporations. 2.31 Non-Filing of Tax Returns. A major source of low tax collection is the non-filing of tax returns by individuals and corporations. This can be gauged by comparing the number of actual tax filers with the number of potential filers as done, for illustration, in Table 2.8 below. The potential number of individual tax filers is calculated as the product of the proportion of individuals with family income above P60,000 (obtained from Family Income and Expenditures Survey, 1988) and the number of people employed in the relevant year. The product of these two numbers gives us a conservative estimate of the number of people with taxable incomes since P60,000 demarcates the top two deciles of the income distribution. For corporations, the potential number of filers is assumed to be the number registered with the Securities and Exchange Commission. The data clearly show that far fewer individuals and corporations file tax returns than would be expected from general parameters of income and employment growth and corporate formation in the Philippines. Table 2.8: Potential Versus Actual Number of Tax Filers Year Potential Number Actual Number Ratio Individual Income Tax Filers (millions) 1988 3.9 1.2 31 1989 3.9 1.4 36 Corporate Income Tax Filers (thousands) 1988 115.1 46.9 41 1989 124.1 40.4 33 Source: Staff Calculations  - 40 _ 2.32 Underdeclaration of Incomes. This is another major source of low collections, It is particularly problematic in the case of the self-employed, such as professionals and businessmen, whose true incomes are very difficult to assess. The scope of the problem may be gauged from the results of an exploratory study reported by Yoingco (1991) in which presumptive income levels were calculated for doctors in Manila. The doctors were chosen from selected medical centers in Manila and a survey was made of their consultation hours, fees and patient traffic. The incomes thus calculated were between four to fifteen times higher (ranging from P416000 to P1500000) than the average income (P92000) reported by doctors in their tax returns. 2.33 One reason why many are able to get away with undeclaration of incomes is the inability of the BIR to verify actual incomes through other evidence. This inability is rooted in two main causes: lack of computer facilities and lack of access to the filer's banking transactions. The lack of appropriate computer facilities means that the BIR is unable to cross-check the records of other government agencies with which the filer may have conducted transactions. For example, the records of the Bureau of Lands may show that a filer has several properties of which the BIR is unaware; rarely is the BIR able to link the income and wealth records of people, even for those with obvious wealth. Another example would be the verification of business turnover and net income. Customs transactions might reveal that a business has much larger turnover than reported to the BIR, but the BIR is rarely able to cross-check with the BOC for lack of the appropriate information network. 2.34 The lack of access to banking records is even more detrimental to the BIR's prosecution ability. This is particularly so with respect to cases involving the relatively wealthy who, because of the scale of their business and wealth, are most likely to use the organized financial system for their dealings. Under Republic Act 1405 (passed in 1955), bank deposits are guaranteed secrecy; they cannot be examined by the BIR or any other agency unless a case has been filed in court. The problem, however, is that typically a case cannot be filed unless there is sufficient supporting evidence, evidence that typically cannot be obtained without access to bank records. The incentive to file cases against wealthy persons is limited also by tta fact that they can file suit for damages against BIR officers who then have to defend themselves at their own expense. The asymmetry of wealth, power and influence involved in such encounters leaves the tax officer at great disadvantage and serves as a major deterrent to zealous tax investigations. 2.35 Overstatement of Deductions. There are several provisions in the individual and corporate tax code that make it easy for the taxpayer to overstate allowable deductions and thus reduce taxable income. For example, a higher personal exemption allowances for "head of family" as opposed to "single" filer often tempts taxpayers to report a false status to claim the higher allowance. Another example is the allowance of deductions for children which leads to claims against fictitious offspring. These "pro-family" provisions have resulted in there being more families and children according to tax records than there are according to marriage and birth registrations. A third example is the facility of filing separate returns by married taxpayers. Typically, both spouses end up claiming full deductions for all children; NTRC estimates that this alone resulted in a revenue loss of 1.5 billion pesos in 1986-88.  - 41 - 2.36 Deductions for business expenses are also thought to be widely abused, especially among individual business tax filers. One study (Manasan, 1988) found that claimed deduction rates tended to increase with income, a finding that supports the hypothesis of abuse since the normal expectation would have been that of constancy in deduction rates across income levels. Furthermore, in the case of corporate income taxes, where the rate structure is less progressive and there is no great advantage to minimizing incremental taxable income, no such relationship was found. 2.37 The problem lies not in the allowance of deductions per se but in the fact that reasonable ceilings are not in place, and in the practical difficulty of verifying the accuracy of claimed deductions. In a country and society in which receipts are typically neither demanded nor offered, it is hard to verify deductions claimed for business representation, entertainment, travel and promotion. Similarly, it is hard to assess the validity of deductions claimed for expense-, related to depreciation, repair and maintenance of vehicles and equipment, which are used both for personal and business purposes. Consequently, it would be advisable to impose deduction ceilings in some cases and fixed ratios of deductions to income where ceilings may be difficult to justify. It might be noted that the Tax Reform Package of 1986 envisaged such ceilings but these were not enacted due to resistance from the business sector. Base Erosion through Exemptions Investment Incentives 2.38 Fiscal incentives for investment have been a significant feature of both the tax system and industrial policy in the Philippines for several decades. The current structure of incentives was put into place in 1987 under the legal cover of the Omnibus Investment Code and provides a range of exemptions to eligible investments, including income tax holidays, waiver of taxes and duties on imported capital equipment, tax credits on domestic capital equipment and tax credits in general. Initiatives have been launched in recent years to replace or supplement, as appropriate, some of these selective incentives with a system of universal incentives featuring accelerated depreciation and net operating loss carry over privileges. 2.39 The rationale for the system of incentives presently on offer is to compensate for market imperfections and to promote investment in general. In recent years. two criticisms have been levelled at this system. First, it is argued that fiscal incentives are not particularly effective in inducing investments. This argument is supported by evidence from other countries as well as the Philippines. In the case of the Philippines, it is argued that a host of other factors are more important in determining the level and composition of investment. Such factors include political stability, law and order conditions, government regulations, market potential, macroeconomic policy, and infrastructure quality. 2.40 Second, it is argued that the incentives impose a fiscal cost that may be well in excess of the benefits generated and that a poor economy can ill afford in any case. A. Yoingco, a former director of the NTRC, writes (see Yoingco, 1991) that incentives have "helped to decimate the productivity of the  - 42 - income tax system and have spawned avenues for tax evasion...we are giving away a lot in exchange for nothing." Table 2.9 provides some data on the revenue foregone from the availment of the main incentives implemented by the 80I in recent years. The numbers, which are conservatively estimated, range between four and six billion pesos and average around 3% of total revenues. The two most significant sources of foregone revenues are tax and duty exemptions on imported capital equipment and general tax credits. 2.41 Another source of revenue loss is the exemption of taxes and duties on imports brought into export processing zones. In 1989, this amounted to almost P2.76 billion pesos; added to the revenue foregone from incentives availed of by BOI-approved investments, this implies a combined loss equivalent to about 4.7Z of total revenues in 1989. Assuming that revenue losses from EPZA imports amount to about 22 of total revenues on averag? leads to an estimate of total revenue loss of around 5.5Z for a typical year. To be fair, however, it should be noted that these estimates of revenue loss, especially for export processing zones, are on the high side because, without the incentives, imports and business activity would be lower. Table 2.9: Revenue Foregone from BOI Incentives Type of Incentive Revenue Foregone (Million Pesos) 1988 1989 1990 1991 1/ Special Deductions 131 51 38 27 Income Tax Holiday - 33 72 64 Tax and Duty Exemptions on Imported Capital Equipment 1980 2640 3842 3167 Tax Credit on Domestic Capital Equipment 48 102 149 123 Tax Credits 1958 1261 1834 1513 Total Revenue Foregone 4117 4087 5935 4894 Total Revenue Collected (billion pesos) 112.9 153.5 179.5 230.8 Ratio of Revenue Foregone to Revenue Collected (percent) 3.7 2.7 3.3 2.1 1/ Projected. Source: Department of Finance  - 43 Exemptions for Government Corporations 2.42 The system of tax and duty exemptions applicable to government corporations has undergone several changes in the last decade or so. Upto mid- 1984, GOCCs enjoyed an array of fiscal incentives to help them achieve prescribed social and economic development objectives. These incentives varied from total exemption from the payment of all taxes, duties, fees, charges and assessments to exemptions from particular types of taxes such as real property tax. These various exemptions translated into an average revenue loss of around P6.4 billion or almost 15% of total revenue intake. Economic difficulties and the high fiscal cost of these exemptions prompted a re-examination of their rationale and the exemptions were withdrawn in mid-1984 in a rationalization move. 2.43 The rationalization attempt was half-hearted, however, and most GCs were able to recover their privileges either through discretionary Presidential action and through application to the Fiscal Incentives Review Board (FIRB). From mid-1984 to mid-1987, estimated revenues foregone on account of GC exemptions averaged P5.2 billion a year, or almost 8% of total revenues. 2.44 Another rationalization move was initiated in 1987. The fiscal privileges of all GCs were once again withdrawn and their restoration was made subject to clearer and more restrictive criteria. As a result, tax and duty exemption privileges were restored for only six GCs and the revenue foregone on this account now averages P2.76 billion per year or about 1.8X of total revenues. Table 2.10: Revenue Foregone From All Exemptions (1989) Source/Type Amount Ratio (Billion Pesos) (Z to GNP) BOI-Approved Investments 4.1 2.7 Export Processing Zones 2.8 1.8 Government Corporations 2.8 1.8 Others 13.0 8.5 Total Revenue Foregone 22.7 14.8 Source: Department of Finance 4I To provide an avenue for relief for GCs that were severely affeqted by the withdrawal of fiscal privileges, the government set up a Tax Expenditure Fund (TEF) to which GCs could apply for subsidies to cover eligible tax and duty expenditures. The TEF is funded by annual budgetary appropriations passed by Congress. The big advantage of shifting to this system of subsidies is that the true cost of fiscal privileges is made transparent and subject to legislative debate. In addition, an automatic ceiling is imposed on the amount of fiscal leakage by the size of the appropriations cover.  - 44 - D. Internal Revenue Administration 2.45 This section evaluates the adequacy of the formal provisions on tax enforcement and the effectiveness of the tax machinery in implementing them. It may be useful to note at the outset that the performance of the BIR has improved significantly in zecent years. This is evident not only in the rising trend of tax revenues to GDP noted in an earlier section, but also in data on the collection efficiency of the BIR. For example, while it cost an average of P1.14 pesos to collect P100 pesos during 1981-85, it only cost P0.83 pesos in 1986-90. Nevertheless, problems remain and improvements can be made. The principal issues of administration are discussed below under three heads: collection; assessment and audit; and tax fraud investigation. Problems in Collection 2.46 Among the more significant problems in collection of internal revenues are (a) deficiencies in payment and recording procedures and (b) existence of sizable accounts receivable. a. Payment and Recording Problem 2.47 The tax collection system of the Philippines features the use of authorized agent banks (AABs). Such banks are used as a measure of convenience to taxpayers since bank branch networks are much more extensive than the BIR office network and since the postal system is unreliable.6 Taxpayers file tax returns with BIR collection agents who then issue payment orders (POs) for the amount of taxes declared to be presented to an AAB. The MAB then issues corresponding confirmation receipts (CRs) for the actual amounts declared and paid via the POs. A reconciliation of POs and CRs is necessary to derive a current list of non-payment, deficient payment, or delinquent accounts. Such reconciliation has proved to be a source of confusion and delay: it is not unusual to find cases of fully-paid accounts being pursued by the BIR because they have not been purged from the delinquent account list. As part of its tax administration improvement program, the BIR plans to do away with the P0/CR system and to require instead that returns be filed and taxes paid directly to agent banks. Negotiations for loans and technical assistance to set up a fully electronic processing system are also being conducted (in part with the World Bank). 5/ Even with the present system of filing through banks there is a problem of "hijacking" of returns. Returns are lost after being filed but before the BIR establishes firm control over them; this typically occurs at the stage where returns are forwarded from regional offices to the district office for registration. This is thought to be an outcome of collusion between BIR personnel and taxpayers. Computerization would offer a solution in that the registration could be done at regional offices or wherever the first point of entry into the system is located.  -45 b. Problem of Accounts Receivable 2,48 Data on the number and value of accounts receivable (see Table 2.11) indicate that the problem is serious along both dimensions. The number of such accounts amounted to almost 14% of the total number of income and capital gains tax filers (average over 1986-90); this ratio would be even higher if exempt returns were removed from the denominator. The value of such accounts amounted to around 19% of BIR collections (on average over 1986-90). Furthermore, deficiencies in the monitoring system mean that it is difficult to tell which accounts have been delinquent for how long and what the status of enforcement action is at any given point in time. The data also suggest, however, that some improvement has occurred in the last five years. The ratio of such accounts to total tax returns filed declined from 15.92 in 1986 to 11.92 in 1990, while their value relative to total BIR collections declined from 24% to 10.2Z.6 2.49 The severity of the accounts receivable problem may be related to three main causes. These are excessive centralization, an inadequate taxpayer identification system, and weak enforcement follow-up on account of lack of personnel. 2.50 Excessive centralization. Responsibility and control over delinquent accounts greater than P20,000 in value rests with the central office of the BIR. In practice, this has meant that the central office handles the bulk of such accounts; in 1990, over 90% of all receivables were being handled by headquarters in Manila. It would appear that decentralization of control to regional and district offices might improve the monitoring of such accounts since such offices are closer to the problem. A very simple problem often dogs the prosecution of cases of delinquency: this is the lack of knowledge of the whereabouts of the delinquent taxpayer. This problem could be partly mitigated if monitoring and follow-up were made a district and regional office responsibility. 2.51 Inadequate taxpayer identification system. The Philippines lacks a taxpayer identification system that can generate an accurate master list of taxpayers. At present, different account numbers are used for different purposes; for example, corporations use different numbers for internal revenue and customs transactions and individuals use different numbers for income and real property tax transactions. It is also difficult to identify spouses and dependents from the existing account number system. The deficiencies of the present system can be judged from the following facts: (a) less than 202 of income and witthholding tax taxpayers currently use valid account numbers and (b) the master list has more than 11 million account numbers although only 2.5 million entities file returns. The issue of accurate taxpayer identification is 6/ Accounts receivable may be classified in three categories: overdue, notice and delinquent accounts. Overdue accounts are those for which no demand notice has been issued. Notice accounts are those for notices have been issued but payment not yet received. Delinquent accounts are those which remain unpaid after 30 days from the date of issuance of a demand notice. Strictly speaking, data on accounts receivable may exaggerate the true delinquency problem. In addition, because of an inadequate recording system, deletions from the list of accounts receivable may not be properly recorded.  - 46 - being addressed through the introduction of a single taxpayer identification number (TIN), similar to the social security number used in the United States system. The TIN system is being implemented on a pilot project basis in one revenue region in 1991 and will be extended to others soon. Table 2.11: BIR Accounts Receivable Collection Performance (percent) 1986 1987 1988 1989 1990 Number Accounts Collected to 8.3 16.2 8.8 13.0 6.8 Accounts Outstanding Accounts Collected to 15.9 15.8 12.9 12.9 11.9 Total Income Tax Filers Amount Amounts Collected to 1.2 1.7 2.1 6.2 4.2 Amounts Outstanding Amounts Outstanding to 24.1 22.6 22.7 14.5 10.2 Total BIR Collection Amounts Collected to 0.3 0.4 0.5 0.9 0.4 Tctal BIR Collection Source: BIR Annual Reports. Also see notes to Table 6 in Statistical Appendix. 2.52 Weak enforcement follow-up. The high degree of delinquency in the payment of taxes may also be partly attributed to the weak enforcement of penalties. The Philippines system requires the issuance of warrants of distraint and levy (WDL) to defaulting taxpayers following which property can be attached and recovery obtained. Data on the implementation of WDLs in recent years indicate a high degree of non-enforcement: for example, during 1986-90, only 53Z of WDLs issued were actually closed (see Table 2.14). Lack of revenue seizure agents is almost always cited as the principal impediment to more effective management of delinquent accounts. Furthermore, it takes time and resources to settle a case: a warrant has first to be served for a tax lien to be attached to property; notice of such lien has to be registered with the Office of the Register of Deeds where the property is situated; seizure of the property has to be effected; and sale thereof at public auction has to be made before the tax due is recovered. A basic problem faced is that the process cannot even be started in many cases because the defaulting taxpayer cannot be located.  - 47 - Table 2.12: Enforcement Through Warrant and Distraint Levies Total WIL Closed Number of _ CY WDL Issued Number Ratio to Total WDL Issued (Percent) 1986 17,464 12,746 73 1987 20,655 18,849 91 1988 19,929 6,696 34 1989 33,644 10,797 32 1990 16,827 8,535 51 108,513 57,623 53 Source: Bureau of Internal Revenue 2.53 A related issue pertains to the economics of prosecuting cases of default. Under present law, the BIR is required to follow up on virtually all cases of default, resulting in the absurd situation where cases involving as little as P100 (about the daily minimum wage) are also filed and followed. In 1990, as much as 93% of the number of accounts receivable involved amounts less than P20,000. Such enforcement is expensive. Moreover, the provision in the law that requires such prosecution was drafted over forty years ago and the trigger amount set then has by now become absurd because of inflation. While there is insuff'cient data to make a detailed analysis of the costs and benefits of prosecuting cases involving small amounts, it is generally agreed that the cost- effectiveness of BIR operations would be enhanced by setting the trigger default amount at a much higher level. Problems in Assessment and Audit 2.54 The main problems related to the assessment and audit functions of the BIR arise with respect to (a) the degree of centralization of authority and responsibility and (b) excessive workload. a. Centralization of Responsibilities 2.55 The headquarters office is deeply involved in the actual examination and Investigation of returns, especially for big taxpayers.7 This has advantages and disadvantages. Among the advantages is the fact that big cases typically involve complex business operations covering several regions. The 7/ BIR data show that in 1989 almost 82Z of the additional revenue from assessments came from cases handled by the national office while 18Z came from those handled by regional offices.  - 48 - audit of such cases often requires examiners with special technical expertise who are unlikely to be available in regional and district offices. Among the disadvantages is the fact that such responsibilities take up valuable staff time and other resources of the natinnal office, thereby affecting productivity in other mandated functions such as policy and review. 2.56 The centralization of the examination function is cited as motivation for under-assessment by regional offices so as to maintain jurisdiction over cases. In some cases, the transmittal of the investigation reports are deliberately delayed, leaving very little time for headquarters to conduct a thorough review of the case. This has often led national investigators to issue arbitrary assessments to beat deadline. b. Workload and Personnel Issues 2.57 The chief issue here can be seen as reflecting either an excessive workload for the existing staff or a serious personnel shortage for the existing work requirements. For example, data on the complement of assessment personnel and taxable returns filed in 1990 show that each assessment officer would have handled 765 cases if a 100% audit policy were followed; in some revenue regions, the potential workload is as high as 1114 per officer. Of course, 100% audits are not necessary and are not done in practice, but even a 10% audit policy, rather conservative for the Philippines, would require officers to handle 76.5 cases on average, which is on the high side. Under present guidelines for audit selection, exanmirners performing field audits are supposed to take on not more than 30 cases per year while examiners performing office audits are assigned a maximum of 90 cases. The general impression is that audit quality is poor. 2.58 One way to handle the caseload problem is to screen those cases that are unlikely to generate significantly higher assessments after audit. One category that would be a clear candidate is that of employed individuals who derive income purely from compensation and whose taxes are withheld at source. If this category is removed from the potentially auditable pool, a 10% audit rule would result in a workload of around 30 cases per officer (using 1989-90 data-- see table 2.13). Another way to improve audit quality is to set up a large taxpayers unit that would monitor such accounts intensively. Both of these measures would provide a better solution than adding more assessment officers and expanding t.de budget of the BIR.  - 49 - Table 2.13: Workloads of Assessment Personnel Year Number of Number of Cases per Cases Examiners Examiner ('000) 1986 1043 2641 39 1987 999 2564 39 1987 986 2711 36 1989 935 2838 33 1990 772 2933 26 Source: Bureau of Internal Revenue Problems in Tax Fraud Investigation 2.59 Tax fraud investigation in the Philippines is characterized by slowness of pace and ineffectiveness of results. These characteristics are related in turn to a number of personnel motivation problems and to weak penalties in law for tax fraud. Personnel motivation problems are reported to arise from low pay and excessive workloads. The BIR has found it difficult to hire competent tax lawyers at its present pay scales as most such lawyers have a much more lucrative alternative in private practice. The number of legal personnel in BIR has declined from 142 in 1983 to only 88 in 1990; there are some regional offices wlhich do not have any legal personnel. One consequence of this appears to be a sharp fall-off in the filing of civil and criminal cases by the BIR in recent years (see Table 2.14). 2.60 Examiners are generally reluctant to get involved in tax fraud investigations because the process is a long drawn out one with little prospect of a satisfactory outcome. Sometimes, such investigators are harassed and threatened. Often, they do not get the cooperation of other government agencies in *:sjn",lishing their case. Typically, the courts do not give high priority to tax cases. The civil and criminal penalties prescribed by law for tax fraud are considered too low to have a significant deterrent effect; civil penalties generally involve a 252 surcharge while crininal penalties limit imprisonment to a maximum of six years but typically to two years. Criminal penalties are rarely imposed and tax fraud cases rarely go to court. In a typical year, the value involved in tax fraud cases comes to only around 1.2? of total BIR collections whereas the number of cases filed is around 0.06% of the total number of tax filers.  50 - Tablo 2.14: Effectiveness of Ton Fraud Investigation Total Civil and Criminal Total (Tax Fraud) Cases Ratios fear -- DIR Textfilbrm Collection Number Amount Cases to Amount to (In Mill'ion) (Ml lon) Texfilore DIRCollectlon 1981. 2,701,482 18,827.20 974 271.80 0.04X 1.44X 1982. 2,680,880 20,087.00 81o 309.00 0.03% 1.64X 1998 2,792,742 21,462.70 1,078 84.00 0.04X 0.39% 1984 2,929,976 32,088.00 150 412.11 0.005% 1.28X 1986 2,832,748 41,858.20 1,887 379.40 0.08% 0.91% 188 2,462,614 48,986.90 2,068 714.30 0.08X 1.62X 1987 2,117,426 65,688.40 8,287 906.40 0.15X 1.66X 1088 2,475,979 83,415.60 4,540 1,297.60 0.16X 2.06X 1989 2,666,203 02,188.90 1,213 392.10 0.05% 0.48X 1990 2,719,404 104,108.70 910 968.60 0.03X 0.92% 28,076,392 489,381.80 10,696 V,724.21 0.08X 1.17# / Refer to income and capital goins filers except in 1981 and 1082 for which no data available on capital gains. Source: Bureau of Internal Revenue 2.61 The above-noted problems have led to two notable initiatives in recent years. First, the Supreme Court has designated special tex courts and prosecutors in some cities to speed up the handling of tax cases. Second, Congress is considering giving front line officials of the BIR higher salaries and incentives to enhance motivation and discourage corruption. E. Customs Administration 2.62 Underperformance with respect to the collection of import duties is related tp a number of issues in customs administration. These are discussed in this section under the following heads: surveillance and smuggling, classification and valuation, and examination and appraisal. Surveillance and Smuggling 2.63 The smuggling of dutiable or banned imported goods can take several forms including outright unrecorded entry, pilferage from customs custody and use of fake documents of duty payment. Smuggled goods are generally of the banned or regulated category, or those subject to highi tariff rates such as textiles, electronics, food stuffs, and motor vehicle parts. The value of such smuggling is thought to be high: indicative estimates prepared by the Economic Intelligence and Investigation Bureau (EIIB) showed that smuggling averaged $1.11 billion per year during 1986-89, or about 12% of average annual imports. The extent of smuggling may have increased in recent years: EIIB estimates suggest that the  - 51 - value of smuggling averaged $754 million during 1979-85, or about 9Z of average annual imports over the period. 2.64 Apprehensions of smuggling attempts are few and far between. The value of apprehended cargo is typically a small fraction of the value of total imports, typically less than 0.32, and about 4? or so of the estimated value of smuggling. While low, the value of apprehensions appears to be creeping ups it is reported that apprehensions rose from 2X of the total value of smuggled goods in 1987 to 4? in 1989. Table 2.15: Data on Seizure of Smuggled Goods Number of Estimated Total Ratio to Shipments Value of Imports Total Seized Seized Goods Imports (Million (Million Pesos) Pesos) 1987 104 129 98,697 0.13? 1988 95 220 124,108 0.18Z 1989 121 487 225,240 0.22Z 1990 101 410 290,082 0.14? Sources Intelligence Division, CIIS, Bureau of Customs. 2.65 Perhaps the roost important cause of low apprehensions is the meagerness of the resources devoted to the prevention and control of smuggling. For example, the customs police have only one operational patrol boat. Even this is used not for patrol duties but as a means of transport for customs officials to board ships for cargo inspection. The bureau does not even have its own stripping machines for cargo inspection; instead, it relies on private firms for this function, an aspect that reduces the speed and flexibility with which the customs police can act. The budget of the customs police has decreased In nominal terms since 1906, has declined as a proportion of the overall customo budget in every one of the last five years, and is now just over 112 of the customs budget (see Table 2.16). These resources are spent not on coasotal patrolling but on maintenance of security within customs premises.  - 52 - Table 2.16: Budget of Customs Police Administration (Million pesos) Year Customs CPA Ratio Budget Budget (2) 1986 21.8 5.9 27.3 1987 24.8 3.9 15.6 1988 29.0 4.1 14.3 1989 36.1 4.5 12.4 1990 38.0 4.4 11.5 Source: Bureau of Customs 2.66 Another factor sometimes cited as contributing to a weak anti- smiuggling effort is the legal/administrative provision which allows the Customs Commissioner to settle seizure cases directly without referring them to a court of law. The merit of this provision is that it accelerates the collection of revenue and decreases administrative costs by expediting cargo releases and by avoiding court proceedings. The demerit is that it can be abused, and if it is routinely abused it can weaken the enthusiasm of enforcement agents to pursue smuggling cases. 2.67 Surveillance is also made more difficult by the poor condition of port facilities. Dilapidated storage facilities contribute to the problem of cargo pilferage. The modernization of these facilities is, however, dependent upon another government agency, the Philippine Port Authority (PPA). Classification and Valuation 2.68 Classification and valuation are the bases for computing customs duties, taxes, fees, charges and sundry imposts on imported goods. As such, it is imperative that the classification and valuation system be simple and clear thereby making it less subject to discretion and more resistant to manipulation. One recent chanige (in 1988) has improved the system in the Philippines. This was tEi adoption of the Harmonized System of classification; this system is used by most of the country's trading partners and contains a clear-cut definition and description of goods, thereby lessening controversies in respect thereof. A major problem is the continued use of the hlome Consumption Value (tlCV) system of valuation rather than the Cost, Insurance and Freight (CIF) system. The CIP  - 53 - system is used by GATT members and is more transparent and less arbitrary than the HCV system.e 2.69 A further step taken to prevent misclassification and misdeclaration was the hiring of the Societe Generale du Surveillance (SGS) to make inspections of cargoes destined for the Philippines from nine ports in East Asia.9 An estimate of the extent of misdeclaration may be derived from the outcome of such inspections. In a period of almost four years, SGS conducted a total of 212,181 inspections; of these, 109,496 were dutiable shipments; of these, 69,002 were found to require uplifting in declared values. Thus about 632 of the shipments contained problems of classification and valuation (see Table 2.19). However, in monetary terms, the problem is less sevcre: upliftments of dutiable value amounted to only 1l.6Z of the total declared dutiable value while total additional duties payable to government amounted to about $125 million or less than 22 of total duties collected in this period. Part of the reason for this, of course, is the so-called 'honesty' effect: importers learn to provide accurate declarations once persuaded of the seriousness of the inspection process. 2.70 A similar impression, that the misdeclaration problem is minor in monetary terms, is given by data on shipments not inspected by SGS bu' apprehended by the Bureau of Customs on account of misdeclaration. These data show that, during 1988-90, the value of such shipments was typically less than 0.4% of the value of total dutiable imports (see Table 2.20). This reinforces the impression that the real problem is not misdeclaration but physical smuggling; consequently, customs resources should be reallocated to emphasize prevention of smuggling and away from classification and valuation functions. The globalization of the CISS should enable such a reorientation of customs priorities. 10 8/ An IMF report prepared in 1985 noted the following deficiencies in the HCV method used by the Philippines to assess dutiable values: it was arbitrary and in violation of GATT rules; it was administratively cumbersome; it discriminated among importers, penalizing those importing from neighboring countries; and it was illogical in that it made valuation independent of such features of normal business transactions as quantity discounts, discounts or premiums for payment arrangements, commissions, and market fluctuations. 9/ The inspection scheme, formally called the Comprehensive Import Surveillance Scheme (CISS), was started in 1987 with coverage restricted to three countries only, Japan, tiong Kong and Taiwan. In 1989, it was expanded to cover six other countries: South Korea, Singapore, Thailand, Malaysia, Indonesia, and Brunei. Macau was added to the list in 1990. In 1991, the CISS was made virtually global as SGS was awarded a contract covering import inspection from almost 100 ports of slhipment. 101 Of course, other possibiliti6s exists the value of misdeclaration may be low because of collusion between impo.ters and customs officials.  - 54 - Table 2.17: Import Inspection Performance (April 1987 - December 1990) Number of inspections conducted 212181 Number of inspections of dutiable goods 109496 Number of dutiable shipments on which increase in dutiable value obtained 69002 Total Increase in Dutiable Value 361.4 (million dollars) Total Increase in Duties and Taxes 125.1 (million dollars) Source: Department of Finance Table 2.18: Data on Misdeclared Shipments Year Total Apprehended Shipments Dutiablee Imports Value Ratio to Total (Million) Dutiable Imports 1988 103.114 118.514 0.18? 1989 148,013 582.922 0.39% 1990 187,202 269.150 0.142 Source: Bureau of Customs Examination and Appraisal 2.71 The examination and appraisal of imported goods may be conducted before or after each other depending on the filing of import entry and arrival of goods. At present, examination and appraisal are undertaken by two separate officials. The Examiner conducts physical examination of the imported goods while the Appraiser checks on the tariff classification and valuation based on the doctunents subm-tted. The Appraiser may, if he deems necessary, make an ocular re-inspection of the goods. 2.72 Given that it is difficult to make a proper assessment without actually seeing the cargo, it would appear that there is no good justification  - 55 - for splitting these two functions. It would be simpler and cheaper to combine these functions. One concern, however, is that this would vest a lot of power in a single officer and might make it easier for bribery to take place since only one person would need to be influenced. One way to reduce the bribery potential would be to rotate assessment officers at regular intervals; this would have the additional benefit of enhancing their skills and experience across a wider range of locations and goods. 2.73 Presently there is no independent check on duty assessments after payments have been made. The practice of audit review which is a feature of the collection of internal revenues might be adopted in the case of customs revenues also. Spot checks of documents and assessments after customs payments have been made could be undertaken as an additional deterrent to corruption. F. Non-Tax Revenue Issues 2.74 Non-tax revenues have typically ranged from lOZ-15Z of total revenues. The principal non-tax revenue sources are interest income from loans and advances, income from fees and charges for certain government services (e.g. forest charges), foreign grants, and proceeds from the sale of assets. In recent years, the last-mentioned source has been the most significant component of non- tax revenues. Accordingly, this section dwells on the revenue implications of the government's privatization program. Privatization of State-Owned Assets 2.75 Government has been implementing a rationalization program during the past five years among whose key features is the privatization of state-owned assets, Such assets fall into two categories: regular chartered public corporations and companies and properties (non-performing assets) acquired by government as a result of the restructuring of several government financial institutions in the mid-1980s. The disposition authority over the relevant public corporations rests with a number of other government corporations while that for the non-performing assets rests with the Asset Privatization Trust (APT). From the point of view of resource mobilization, it is the disposition of the latter group of assets that is more izportant since this results in direct and immediate flow of revenues (net sales proceeds) to the national treasury. The sale of government corporations does not typically produce revenue for the national goverrnent sinice the ownership rights rest with other government corporations. However, such privatization helps the national treasury indirectly since it typically improves the cash flow position of the disposing parent corporation and enables it to declare a higher dividend to the national government or seek reduced subsidies and other transfers from it. 2.76 Progress of ProRram. The rationalization program has been relatively successful in disposing of non-performing assets but not in privatizing public enterprises. By mid-1991, the APT had disposed of 230 out of the 399 accounts under its management for a sum of P25.2 billion pesos. Compared to this, by end- 1990, only 67 of the 123 government corporations approved for sale had been  - 56 - offered for sale and only 32 had actually been wholly or partially sold for a total of P8.6 billion. 2.77 From the point of view of generating cash to meet current needs, the program has been a partial[5;0r6'succ&bsits inception, it was expected that the program would yield about P27.4 billion pesos from the sale of NPAs by end-1991. The sales value so far is P25.2 billion pesos. However, only P14.6 billion pesos of this sum has been remitted to the national treasury so far. The remainder is accounted for mainly by amounts held in escrow pending settlement of legal issues, sales transactions awaiting completion via debt-equity or debt-asset swap schemes, and amounts payable in installments. 2.78 The program has been financially profitable for the national government: the cost of privatizing NPAs has been negligible compared to the proceeds thereof. The national government provided APT a subsidy of P84.3 million pesos during 1987-89, an amount equal to only 0.75% of the sales proceeds remitted to the Treasury in that period. Actual custodial expenses amounted to P295.5 million pesos during 1987-90, which was only 0.94% of the revenues generated from the privatized assets. 2.79 Potential Revenues from Further Privatization. For the period 1992 and beyond, privatization of NPAs is projected to gross P42 billion of which around P30 billion is expected in 1992 alone. Thus, the bulk of the privatization is projected to occur by the end of 1992. From the total, the national government is projected to receive proceeds of P14 billion, of which P9.3 billion 13 expected in 1992. 2.80 Assessment. The proceeds from privatization, both historical and projected, are small relative to total revenues. During 1987-90, the cumulative remittances to the treasury amounted to just around 1.5 of total revenues. Future contributions will be even less significant, partly because the proceeds will taper off as the stock of privatisable assets shrinks and partly because total revenues (the denominator here) will increase through growth and inflation. Nevertheless, privatization should be encouraged for several reasons. First, it has intrinsic merit in that it leads to a more efficient deployment of overall resources. Second, in an environment of severe fiscal constraint, additional resources have a hligh marginal utility. Third, even if the direct revenue effect is limited, the indirect savings from disposing of non-performing assets and inefficiently managed public corporations can be quite substantial. The game is worth the candle. G. Enhancing Revenue Mobilization 2.81 The foregoing sections have reviewed and analyzed the main causes of low tax collections in the Philippines. This section reviews some solutions that might be instrumental in Improving revenue mobili2ation. Some of these solutions formed part of the Government's proposed tax reform program and were considered in the legislature in 1991 (sea box 2.2). They are divided into two groups, those that would improve the design of the tax system and those that would Improve its administration.  - 57 - BOX 2.2: TAX REFORM PROPOSALS OF 1991 A sot of proposals was put forward by tho Administration for consideration by the Philippine Congress. The chief eloents of tho proposod tax reform packago were: (a) Simplifiod Not Income Tax System: this aims to limit deductions on business expenses and reduce applicable tax rates; this Is expected to improve horizontal equity, lower tax ratoe *nd generate a not rovonue guin. (b) Restructuring of Transfor Taxes: high rates on gifts and transfers (of estates and donations) are thought to have generated a high dogre of evasion and resultod in a very low tax take; this proposal sooks to reduce tho highest applicable rate on estate transfers from OOX to asx and tho number of brockots from sixteen to two. (c) VAT Rofinements: one proposed refinement is to expand the VAT base by bringing such proviously exompt sorvices as hotel, catering, and entertainment under VAT; another proposed refinoment Is to provide Incentives (in the form of a 4X shere of collections In excos of target collections) to local governments to asslot In Improving collections; a third refinoment Is to Issue advanco tox credits to zoro-rated porsons or activities so as to roduco the number of refund claims. (cd) Withholding Tax on Compensation Income: It io proposed that this tax be made a final one, to shift tho burden of adjustment of taxes duo end paid to employors and employoes, ond to absolve fixod-income senore of tho responsibility of filing tax returns; It is expected that this would ease tax compliance, docreaso BIRs costs of collection, ond permit e roallocation of BIR resources to the audit of more productivo casos. (o) Mosures to Improvo Tax Administration: the chlf measuroes bing proposed are tho creation of special tax courts dodicated to speedy disposition of tax caoss, tho incresso of penaltioe for tax fraud, and tho creation of confidontial positions at the BIR end BBC; thoso positions would carry higher salarios and bonusoe but would not enjoy tho protection of civil sorvice laws; It Is expected that the higher Income and lose socuro tenure would discourage frontline examinors and asoossors from collusion with tax evaders. In addition to the above, Govornment agroed with tho IMF on a fivo-point progrom to Improve tax administration. Tho program consists of meosures that can be lmploemnted rolativoly quickly and should doliver rosults soon enough to affect fiscal performance in 1992. Tho eloemnts of the program are: (a) Introduction of a singlo Tax Idontification Number (TIN) to roplaco the existing array of tax accounts. (b) Adoption of case screening criteria to distinguish between categorls likely to result in high post-audit adjustments and those unlikely to do so. (c) Collection of VAT on a monthly basis, Initially from large payers, to reduce collection lags. (d) Estabilshment of special unit to monitor filing and payment of taxes by large taxpayers. (e) Reform of tax collection system to move away from current procedures involving payment orders and confirmation receipts. Improving Tax System Design 2.82 Presumptive Income Levels. It has been shown that non-filing and underdeclaration result in significant loss of revenue and that it is difficult to verify the true incomes of self -employed professionals and businesses, the so- called Ohard-to-tax" groups. One way to attack underdeclaration is to estimate  - 58 - and apply presumptive income levels for different categories of professionals and businesses. Empirical evidence for the Philippines suggests that the degree of underdeclaration is high and there would be a large payoff to the adoption of a PIL system. 2.83 It is clear that several legal and administrative challenges will have to be surmounted before such a system can be widely implemented. Concerned groups will lobby strongly against such a system and initial assessments will be widely contested. On the administrative side, there is the difficulty of making sufficiently precise estimates and then updating them regularly, and further of managing contested cases. Nevertheless, several countries (e.g. France, Turkey, Japan, Israel, Thailand) have adopted this system in one form or another; their experience suggests that the system is productive and that various anticipated difficulties can be dealt with satisfactorily. For example, professional associations could be involved as partners in the process of estimating PILs and setting procedures for revision and arbitration; this should result in a system which has the sanction of the target taxpayer groups. 2.84 Business Deductions. It has been noted that the pattern of claimed business deductions exhibits characteristics which suggest that this facility in the tax law is being abused. This source of tax evasion could be controlled by setting ceilings on allowable business deductions, especially on those that are most open to abuse such as promotional, representational and travel expenses. Such ceilings were originally envisioned in the TRP of 1986 but were not introduced in the face of opposition by business groups. At the time such ceilings were proposed in 1986 it was estimated that the net gain in revenue would be around P200 million; on current levels of corporate income, the gain could be three times larger. 2.85 VAT Coverage. When the VAT system was introduced in 1988, several categories of goods and services were exempted. While that may have been a sensible decision at that time, in view of anticipated difficulties in introducing a new system, enough experience has been gained by now to justify the expansion of the VAT base. Among the categories of services that could be included in the VAT base are: hotels, caterers, common carriers, forest products, securities dealers, public utilities, communications, and entertainment. It should be noted that revenues would not necessarily be increased in all cases if Lhe VAT rate of 102 is applied: for example, hotels currently charge a higher rate and there is no benefit of input crediting. Nevertheless, a widening of the base would be justified on grounds of tax neutrality and easing of administration. 2.86 Fiscal Incentives. The fiscal incentive system, featuring a variety of tax waivers and credits to promote investment, is a notable source of revenue base erosion. Unfortunately, the fact that at present investntent is perhaps more sensitive to other considerations, such as political stability and the availability of infrastructure, renders the present system of incentives ineffective in terms of generating adequate levels and a desirable mix of investment. Consequently, there is a case for drastically re-designing the system. For fiscal considerations, it would be useful for the redesigned system to limit such incentives to a very small group of industries. Government has begun moving in this direction. The Investment Priorities Program list was  - 59 - reduced from 206 industries/sectors in 1990 to only 95 in 1991. If the process of pruning is continued in future years, this source of fiscal leakage will be reduced. Improving Tax System Administration 2.87 Bank Secrecy Law. It has been argued that large scale tax evasion is made possible mainly because of the legal inability of the BIR to detect cases of evasion through the financial records of tax filers. The failure to catch such evasion by "big fish" involves a large direct revenue loss, of course, but also has a corrosive effect on public ethics with regard to tax obligations. Given this, it is imperative that the Bank Secrecy Law be amended in such a way as to enable BIR to secure bank records without having to file a case in court. It should be made possible, for example, for the BIR to obtain bank records once it starts a tax fraud investigation. Concerns about potential harassment of taxpayers should be met in other ways, such as restricting access to only very senior BIR officers. Concerns about runs on banks should also be met through other means, such as providing liquidity to affected banks through Central Bank mechanisms. It should be noted, however, that concerns about financial disruption that might be caused by the lifting of bank secrecy may be exaggerated; there usually is no alternative for the "big" tax evader to using banks since informal institutions are unlikely to be able to provide an adequate degree of security and convenience. 2.88 Taxpayer Identification System. Data processing and taxpayer identification is hampered by the existence of multiple account numbers for different tax transactions with BIR. The BIR has already begun the process of replacing these multiple account numbers with a single tax identification number (TIN). This should help enhance data processing and reduce tax evasion. Data processing would be facilitated since fewer files would need to be opened, payments made could be easily recorded, and adjustments easily effected. Tax evasion would be discouraged to the extent that different transactions with BIR could be easily retrieved for cross-checking purposes. In addition, if the TIN were designed to identify spouses, this would help cut down on evasion through the practice of both spouses claiming full deductions for the same set of dependents. 2.89 Computerization. It should be noted that the full value of a TIN system can only be derived if tax accounts are properly computerized. In addition, computerization would enable improved monitoring of tax account status and improved interaction with other government agencies that have a revenue function. Government is currently receiving technical assistance from various international development agencies such as the World Bank, the International Monetary Fund and the Asian Development Bank, in preparation for a tax computerization effort to start next year. 2.90 Expansion of Withholding Scheme. Legislation has been passed recently to enable the BIR to treat withholding on compensation income as final in nature. Withholding at source has generally proved to be a productive and reliable source of revenue, whether it be final withholding on passive earnings or tentative withholding on compensation income. Experience has shown that the latter category of tax returns rarely involves fraud and seldom requires adjustments.  - 60 - Therefore, it is argued that the withholding from compensation income should be made final for at least those cases where this is the sole source of income. Employers would be required to withhold an amount equal to the tax liability; should differences arise, they would be settled between employers and employees directly through refunds or extra withholdings from the next year's salary. This would cut down on the number of returns to be processed, examined, and acted upon by the BIR thereby releasing personnel and resources for more active investigation of other returns. It is estimated that such a measure would reduce by 80X the number of returns to be filed and examined. 2.91 VAT Administration. Since VAT is a relatively new tax, there is room for improvement in its administration. Moreover, since VAT is a potent source of revenue, such improvements are likely to have a large payoff. The principal areas for improvement are: (a) Effort: VAT merits a larger administrative effort possibly through an upgrading of the VAT division in the National Office to a distinct Service and the enlargement of VAT units in regional offices through the hiring of additional personnel; (b) Decentralization: All VAT processing, audit, and investigation functions should be delegated to the regional and district offices, leaving headquarters only the mandated roles of review and evaluation, and policy and program development; (c) Education: A VAT information program should be continued so as to reach potential taxpayers who may not be filing out of sheer ignorance of the system and its requirements; (d; Monitoring: The list of VAT filers should be updated and verified periodically with a master list of business establishments in all revenue districts; (e) Cooperation: The cooperation of local officials should be enlisted towards VAT collection by allocating a certain portion of VAT revenues to local governments; and (f) Collection: At least for the larger taxpayers, VAT should be collected monthly rather than quarterly, as is presently the case; the amounts involved are large and the interest income foregone from staggered collections is sizable. Monthly collections are quite feasible from large taxpayers and entail modest administrative complications since large filers (i.e., those paying more than P100,000 per quarter) form only 3% of VAT filers but account for almost 90Z of VAT revenue. 2.92 Audit Case Selection. Another measure that has the potential for reducing workloads, increasing productivity and thereby reducing the cost of collection at BIR has to do with the selection of cases for audit and investigation. Under present law, the BIR is required to follow up on cases involving as little as P100 (about the daily minimum wage); in 1990, around 93Z of accounts receivable involved amounts less than P20,000 pesos. Following up on small cases is expensive. The criteria for selecting cases for audit and investigation should be adjusted so as to focus the resources of the BIR on relatively larger cases with prospects of higher recoveries. It is counter- productive, from a revenue point of view, to spend more resources on the collection process than might be reasonably expected as an outcome of the process. For small cases, only random audit and investigation should be conducted; this would keep collection expenditures low without removing the threat of audit and prosecution from small taxpayers. 2.93 Courts and Penalties. It has been noted that the relevant BIR personnel are often discouraged from filing civil and criminal charges in cases of evasion because the court process is long and tedious and very seldom results  - 61 - in suitable penalties being awarded to the tax evader. In this light, it would be useful to set up special tax courts and to increase the penalties applicable to tax evasion. Recently, 20 courts in selected major urban areas were tasked to handle tax cases on a priority basis; however, since these courts continue to handle non-tax cases as well, they have not significantly eased the backlog of tax cases in the jurisdictions concerned. What is needed is a system of dedicated, specialized tax courts which hear nothing but tax cases. As far as penalties are concerned, common sense suggests that evasion should decrease in direct proportion to auticipated penalties. The problem in the Philippines is that financial penalties are too low to be effective while the jail terms prescribed in the law are rarely applied. Raising fines and applying jail terms more frequently would have a deterrent effect on tax fraud; administrative and judicial discretion in this matter should be restricted so as to minimize the possibility of "negotiated" settlements. Legislation has been passed recently to raise penalties for tax evasion and provide special courts for tax cases. These are clearly steps in the right direction. 2.94 Frontline Cadre. Corruption is a major element of tax evasion. The roots of corruption may lie in all manner of socio-cultural considerations but they also lie in such matters as low pay for tax and customs officers. Low pay not only discourages competent people from seeking a career in the BIR or BOC but also encourages staff to increase their real incomes through collusion with tax evaders for a cut. One solution to this problem that is presently being considered is to create a special cadre of frontline officers (examiners and assessors) tn the BIR and BOC who would be entitled to higher pay and special bonuses but who would also be removed from the protection of civil service laws. It is thought that such a cadre would have iess incentive to be corrupt since their pay would be higher but also because, if cases of corruption occur, the officers concerned could be summnarily dismissed. In other words, the scheme contains both a carrot and a stick. Only experience will tell whether such a scheme will work and whether the carrot part or the stick part is the more effective element. Legislation was introduced in 1991 to implement such a scheme but it was not successful mainly because of concerns that it would give senior tax officers excessive discretion and set a bad precedent for the operation of civil service rules. 2.95 Control of Smuggling. It has been shown that smuggling is a far more serious problem in the Philippines than misdeclaration of import value. Conservative estimates place the value of smuggling at 12Z of total imports. Given the archipelagic nature of the country it is admittedly difficult to control smuggling; present efforts are, nevertheless, woefully inadequate. Resources (both personnel and equipment) devoted to the anti-smuggling effort need to be considerably increased; the globalization of import inspections should permit the necessary reallocation of customs budget resources. In addition, penalties for tax fraud via smuggling should be increased and court proceedings expedited. Quantitative Aspects of Revenue Mobilization 2.96 Several of the measures discussed in the preceding section formed part of a legislative tax reform package advocated by Government and considered by Congress in 1991. Among the ones that were approved are: the simplified net  - 62 - income tax scheme including a rationalization of deductible expenses for businesses, increased penalties for tax evasion, creation of special tax courts, restructuring of transfer taxes on estates and donations, creation of a unit to monitor taxes payable by large taxpayers, making the withholding tax oti income derived solely from compensation final in nature, and restructuring of forest charges. In addition, Philippine Air Lines was successfully auctioned, drawing a higher bid than anticipated. These developments helped to satisfy the IHF of the viability of the stabilization program for 1992. It is expected that the above-noted measures will contribute to the generation of an additional 20 billion pesos of revenue in 1992 and enable the government to maintain the public sector deficit at 2.7Z of GNP and public expenditures at the level of about 18Z of GNP. This level of expenditures is expected to be consistent with a growth target of 2.5Z for 1992. 2.97 The medium term fiscal picture, however, leaves much to be desired. Among important measures that were not approved in the last legislative session were: the extension of the VAT base, the rationalization of motor vehicle taxes and the approval of frontline "confidential" pGsitions at the BIR and BOC. The first two of these measures were of the "hard" variety, that is, measures whose revenue yield is relatively quick and predictable. The third measure is admittedly "soft" but is estimated by BIR to have a high yield. To improve the fiscal picture in the medium run, these and other measures suggested in this report should be put on the legislative table again. Some of these measures are needed to improve revenue flows but some have additional desirable features as well of promoting neutrality and equity as well. 2.98 Revenue mobilization must also be evaluated in relation to other development objectives, in particular, the objectives of achieving greater economic efficiency and of maintaining a sustainable current account deficit. In the long run, there is no inconsistency between these objectives and that of raising more revenues. In the short run, however, some tradeoffs might arise. For example, greater economic efficiency often requires the reduction of import duties but these are typically important sources of revenue to the treasury. Similarly, exchange rate depreciation might be required to maintain international competitiveness and a low current account deficit, but this tends to increase the external debt service burden of the public sector, thereby increasing revenue requirements. Policy formulation to achieve simulltaneously higher revenues and the other objectives noted above requires knowledge of the quantitative links between various revenue measures and other macro balances. 2.99 Such quantitative links were assessed for this report with the aid of a computable general equilibrium model of the Philippines Four scenarios were simulated: (a) tariff reform as specified in Executive Order 470; (b) redaction of current account deficit by 12 of GDP; (c) an increase of 1% in the investment rate and (d) a combination of selected efficiency increasing measures together with (a), (b) and (c). The efficiency measures comprise expansion of VAT coverage, elimination of all subsidies provided by Government, abolition of the special 5Z import levy, and liberalization of investment and trade (leading to an increase in import and export elasticities). In each case, the model was used to calculate the required increase in domestic taxes or decrease in government consumption using 1990 as the reference year. It -as assumed that prices and wages would be flexible enough to accommodate a real devaluation in each case;  - 63 - the maximum real devaluation implied under the first three scenarios was 3.62. It was also assumed that there would be no additional external financing. Table 2.19: Estimated Revenue Impact of Selected Measures (Billion Pesos) Annual Yield Expansion of VAT Base and other Refinements 8.1 Creation of Confidential Positions 7.5 Simplification of Net Income Tax 2.7 Restructuring of Motor Vehicle Tax 2.6 Increased Penalties for Tax Evasion 3.2 Restructuring of Forest Charges 0.5 Continuation of 5Z Import Levy 4.2 Sources Congressional Planning and Budget Office 2.100 Table 2.20 below provides estimates of the adjustment in domestic taxes and tax ratio required under the various scenarios noted above. The numbers suggest that significant amounts of domestic revenues are needed in order to implement the policies entailed by the objectives of greater efficiency, more growth, and a lower current account deficit, that is, a combination of (a),(b) and (c) above. These objectives can be accomplished by raising domestic taxes by 38Z which implies an increase in the tax ratio of around 19%. Given that the typical rate of increase in tax revenues in recent years has been around 22Z, this suggests that special efforts are required to increase the elasticity and buoyancy of the tax system, that is, tax collections must be increased through a combination of reforms affecting the design and administration of the tax system. 2.101 Of course, a reduction of government expenditures would ease the revenue enhan-ement required. So would measures that increase the productivity of the economy. This is done in the fourth simulation ncted above. Government expenditures are reduced to the extent that all subsidies offered by the national government are eliminated. Productivity is enhanced by liberalization of investment and trade. In addition, this simulation features an increase in the coverage of VAT to include most of the service sector (with an assumed collection rate of 602 of the actual collection rate in manufacturing). These measures, in  - 64 combination, reduzce the required domestic revenue effort by half; an increase of only 19Z is called for in this case. Table 2.20: Revenue Mobilization and Other Objectives Percent Change Domestic Taxes Tax/GDP 1. Reduction in Current 16.2 10.7 Account Deficit by 1Z 2. Tariff Ref3rm: E0470 11.4 0.6 3. Investment Growth of 11.8 8.2 Additional 5% 4. Total (1+2+3) 38.1 19.4 5. Total + Other Measures!/ 19.6 7.0 11 Other measures refer to: extending the -overage of VAT; eliminating subsidies, and liberalizing investment and trade. Source: Staff Calculations 2.102 The key conclusions are, therefore, the following: (i) a significant increase in revenues is required to meet the combined objectives of greater efficiezcy, higher growth and lower external deficit; this can not be met by normal revenue increases associated with growth but must be obtained from improvements in tax collection; (ii) the required revenue effort can be mV'erated by improvements in productivity and reductions in government expenditures; the next chapter considers how this might be done through reforms in the management of public expenditures.  - 65 - III. PUBLIC EXPENDITURE MANAGEMENT 3.1 While enhanced resource mobilization is a necessary condition for sustaining higher growth in the future, it is not a sufficient condition. It is important to go beyond raising resources to ensuring that they are used in an efficient manner and towards desirable ends. The recent economir history of the Philippines provides a case in point. Public expenditures rose substantially during the 1970s, but they were not deployed in an efficient manner. Higher public expenditures were associated with substantial growth in the size of government, with an increase in subsidies for government services, and with a preference for large and expensive investments in sectors where comparative advantage was doubtful, to say the least. Consequently, these expenditures did not result in sustained growth and, to the extent that they were financed by external borrowing, did not generate the resources needed to meet debt-service obligations. The result was a fiscal and economic crisis in the mid-1980s, whose ramifications affect budgetary priorities and economic performance even to this day. 3.2 If a similar outcome is to be avoided in the future, not only must the pattern of funding shift towards resources raised domestically but the size and composition of expenditures must be guided to a greater extent by the criterion of efficiency and cost-effectiveness. This necessitates consideration of issues related to the management of public expenditures which is accordingly the topic of this chapter. Since the central government and public enterprises account for the bulk of public expenditure, attention is focussed primarily on these and local government issues are ignored. 3.3 The chapter is organized as follows. The first section provides a brief discussion of budget process issues mainly to see if gross inefficiencies arise in the transformation of development plans to annual and medium term financial plans for public expenditures. The second section reviews recent trends in government expenditures and public investments and examines the factors that determine the productivity of such expenditures in the Philippines. Among these factors is a greater role for the private sector in delivering infrastructure services; this is the subject of the third section. The fourth section deals with public enterprises and considers issues related to their financial autonomy, operational efficiency and financial performance. A. Budget Process Issues 3.4 The budget process is important to the issue of public resource management in that it has the potential to change the content, increase the cost and delay the implementation of the public expenditure program. The extent to which this occurs in the Philippines is discussed below through a review of three stages of the budget process: planning, approval and implementation.  - 66 - 3.5 Budget Planning. On the whole, the budget planning process in the Philippines is sound. The process is a consultative one; the investment priorities of both central line ministries and local government units are sought and accommodated subject to fiscal constraints and national priorities expressed in the Medium Term Public Investment Program (MTPIP). One measure of the overall soundness of the system is that public investment outlays hbve tended to follow MTPIP sectcral allocation patterns even when fiscal constraints have prevented the attainment of matching levels of expenditures. Furthermore, recognizing that local government units are often incapable of drawing up plans and priorities, and also that some local investment programs may be inconsistent in a regional context, the system provides for relevant technical assistance as well as consistency checks through Regional Development Councils. This process has irproved local capabilities in many instances. thereby providing a base on which further decentralization can rest. 3.6 One recent innovation has reduced frictions in the planning process as well as the time needed to put together a final plan. For the 1990 budget, the Department of Budget and Management (DBH) prescribed ceilings to spending agencies and let them decide how to prune their submissions to fit under these; the earlier practice had been for DBM and the spending agency to prune submissions jointly, a practice that was both eacromonious and time- consuming. The ability to choose their own priorities (subject to broad national guidelines) has been welcomed by spending agencies and has speeded up the budget submission process. 3.7 Nevertheless, there are two areas of concern. One is that the present planning process does not yield a Medium Term Financial Plan (MTFP) of sufficient accuracy to reliably guide medium term budget decisions. The present MTFP is highly aggregative and tentative. Related to this concern is the fact that, whereas capital expenditures are carefully planned, selected and monitored, a similar degree of attention is not foeused on current expenditures which make up close to 80% of the budget. There tends to be little analysis of the individual components of agency current expenditure profiles as they relate to development objectives. In practice, this has meant a rather haphazard approach to budgeting for operations and maintenance expenditures. The MTFP could become a more useful document if it were less aggregative and more detailed in its coverage of expenditure categories and if it contained simulations of multiple scenarios with respect to the key determinants of expenditure trends. 3.8 The other concern is that the syste- may be growing too complex. Many agencies are involved and there is always the risk of coming up with inconsistent plans, priorities and schedules. Rather than sort out the inconsistencies at a later stage, it may be better to improve coordination earlier. Accordingly, Government has recently begun testing a coordination exercise called the Synchronized Planning, Programming, and Budget System (SPPBS). The SPPBS is intended to improve coordination among the budget, planning and revenue agencies in particular so as tc ensure consistency of budget plans with development goals and available finance. Implementing the SPPBS will be a difficult exercise, but one worth pursuing as long as  - 67 - attention is not diverted from the pressing short-term problems that confront policy makers. 3.9 Budget Approval. The budget approval process involves the review of the submitted budget by Congress. This process has revealed some tensions in recent years. There was a serious delay of about six months in the approval of the 1991 budget partly due to disagreement over expenditure priorities but largely over revenue mobilization strategy with Congress objecting to the tax increase measures proposed by the Government to meet the terms of a macroeconomic agreement worked out with the IMF. 3.10 One expenditure area over which differences have regularly arisen is external debt service. Congress has tried to set limits to this category in the budget whereas the Administration has argued that debt service falls under the cover of automatic appropriations and congressional authorization is not required. Despite the fact that the Government's position was affirmed by the Supreme Court two years ago, attempts to set debt service caps persist. While this may not be a problem with procedures per se but more a reflection of political sentiments on the part of some legislators, it has the potential to disrupt the approval process. 3.11 Budget Implementation. The procedures for commitment and release of funds affect the efficiency with which public resources are managed. If the procedures are cumbersome, program schedules can slip in critical ways and investment costs can rise. On the other hand, if procedures are too lax, funds could be spent for inappropriate purposes, again increasing overall program costs. The objectives of flexibility and control have to be balanced. Three innovations of recent years merit comment. First, the validity of authorizations was extended by a year, in effect giving agencies two years to make authorized expenditures. Second, the common fund approach was adopted whereby agencies were given the authority to shift funds from slow-moving projects to fast-moving projects. Third, fast release procedures were adopted whereby agencies were given cash to cover authorized expenditures even before Government had received disbursements from foreign creditors. These innovations have provided greater flexibility to agencies in managing projects. The last-mentioned measure has, however, created some problems for macro managers in that agencies have less incentive to submit documentation required by foreign creditors on a timely basis and the Treasury is left with an unpredictable lag between outflows and inflows. This has contributed to cash management problems. 3.12 In recent years, two aspects of cash management have drawn attention. One aspect is that of cash float. This became a prominent issue in 1990, at a time of great fiscal stress, when it was discovered that the cash float involved in the transmittal of funds from the DBM to spending agencies was averaging around 10 billion pesos. This was considered to be excessive in relation to the true cash needs of the spending agencies and was thought to be caused by high interest rates then carried by Government securities. These rates provided an incentive to depository banks to delay the transmission of funds to spending agencies in order to benefit from the spread between the interest rate they paid to agencies on such deposits and the return from investing in government securities. It is thought that some  - 68 - spending agencies also sought to improve their financial position by delaying their expenditure schedules and investing excess cash in T-bills. Once detected, the problem was speedily resolved. The number of depository banks was reduced from 7 (including conmercial banks) to 3 (including only government owned banks) and the average float was cut from 10 billion to under 4 billion pesos. 3.13 A second area of concern is that of the timeliness of cash releases to spending agencies. Many agencies complain that project schedules are disrupted and delayed by the failure of DBM to make cash releases on time despite the existence of appropriate authorization for the relevant expenditures. This problem has been especially acute in 1991 but appears to be related more to policy than to process. A widening fiscal deficit in 1990 led to a policy decision to limit expenditures to the availability of revenues on a monthly basis. Since anticipated revenue enhancement measures did not materialize, this led to a squeeze on cash releases to spending agencies; spending has been determined effectively by cash on hand rather than appropriation cover. Assuming that fiscal constraints necessitate continuation of this approach, it is useful to consider how its adverse effects might be controlled. 3.14 The main adverse effect of such a cash management approach is that it reduces the usefulness of work plans submitted by the agencies. If cash releases do not bear much resemblance to submitted plans over a couple of years, cynicism might creep into the process of formulating plans. Furthermore, contractors working on government projects would adjust their own cost and schedule submissions to take into account the greater uncertainty characterizing anticipated reimbursements. Such responses would tend to delay project start-ups, lengthen implementation intervals, and increase costs. Two steps might be taken to decrease the uncertainty involved and thereby mitigate adverse consequences. First, a core set of governmnt expenditures could be identified which will receive priority in financing. in other words, funds rationing would be selective across agencies. In this way, once it is known which projects fall into the core set, uncertainty with respect to their financial needs being fulfilled would be reduced. Second, revenue intake should be smoothened across months to meet anticipated expendicures through the use of borrowed funds. This can be done by allowing weekly Treasury auction sizes to vary in accordance with weekly or monthly revenue needs rather than be set at pre-determined levels, consistent with annual needs, as is the case at present. 1/ The intention here is not to engage in micro-management of each agency's spending but to have a pre-determined rule to allocate cuts across agencies; that is, not all agencies would receive the same across-the-board percentage cut. As far as intra-agency projects are concerned, it is best to leave "triagel decisions to the agency concerned since it usually has the best information on individual project status and importance.  - 69 - B. Public Expenditure Patterns and Productivity 3.15 There are some very clear trends and patterns in the recent evolution of public expenditures. These are discussed below in two sections, one dealing with the national government budget2 and the other dealing with public sector investments, including the capital expenditures of the national government as well as those of government owned and controlled corporations (GCs); the investment outlays of local government units are ignored as they are a small fraction of the total. For ease of exposition, the discussion is framed in terms of a comparison of recent pattetns for the period 1986-90 with those that prevailed over the period 1975-85. Government Expenditure Patterns 3.16 Economic Categories. Viewed in terms of economic or functional categories, the recent pattern of national government expenditures shows several clear differences with the earlier period: (a) Total government expenditures jumped as a proportion of GNP from 15.9Z to around 21Z. This was due entirely to (b) an increase in current operating expenditures from 10.9% to around 17Z; the share of capital expenditures declined from 52 to 3.8?. The rise in current expenditures was due primarily to (c) a sharp increase in interest payment obligations (from 1.22 to 6.1?) and an increase in personal services expenditures (from 3.82 to 5.3Z). The decline in the share of capital expenditures was accompanied by (d) a fluctuating trend in operating and maintenance expenditures which fell in the early part and rose in the latter part of the decade and by (e) a sharp fall in investment outlays from 2.12 to 0.8Z of GNP. 3.17 The sharp rise in interest payment obligations came about as a result of two factors. First, the national government assumed (in 1985) the external debt liabilities of several government and private entities for which it had issued guarantees and which were in the process of being restructured in the wake of the financial and economic collapse of the mid- 1980s. This assumption of liabilities transferred the debt service obligation to the national government, thereby resulting in a big increase in allocations for interest payments. Second, substantial depreciation of the peso relative to the earlier period increased the magnitude of the external 2/ The budget of the national government is organized in two parts, current and capital expenditures. There are no special accounts that are not included in these two categories. Operations and maintenance expenditures are included in current expenditures. The practice of specifying a development budget, listing all current and capital expenditures as have a developmental impact, is not followed in the Philippines.  - 70 - Table 3.1: NA71ONAL GOVERNMENT EXPENDITURES, BY ECONOMIC CLASSIFICATIONd OBLIGATION BASIS, (Percent of GNP) ........... ............................... ................. 1975- 1986- 1985 1990 1986 1987 1988 1989 1990 TOTAL 15.90 21.04 18.63 22.11 20.34 20.73 22.47 Current Operating Expenditures 10.88 17.21 12.59 18.14 17.53 17.72 18.49 Personal Services 3.83 5.34 4.64 4.48 5.30 5.41 6.23 Maintenance and other Operating Expenditures 7.05 11.87 7.95 13.66 12.23 12.31 12.26 Interest Payments 1.22 6.07 3.52 9.98 5.57 5.73 5.68 Transfers 1.13 1.12 1.00 0.84 1.02 1.14 1.41 Loan Repayment 1.91 2.16 1.05 0.00 3.09 2.89 2.81 Other MOE 2.78 2.52 2.38 2.83 2.55 2.55 2.36 Capital Outlay 5.02 3.83 6.04 3.97 2.81 3.01 3.99 Land, Land Improvements and Structure Outlays 1.51 1.08 0.49 1.07 0.91 1.11 1.51 Buildings & Structures 0.57 0.82 0.97 0.53 0.53 0.66 1.27 Equipment 0.26 0.28 0.10 0.17 0.34 0.36 0.34 Investment Outlay 2.13 0.81 2.03 0.67 0.29 0.58 0.81 Loans Outlay 0.54 0.84 2.46 1.52 0.74 0.30 0.07 Table 3.2 NATIONAL GOVERNMENT EXPENDITURES, BY SECTORAL CLASSIFICATION OBLIGATION BASIS, (Percent of GNP) 1975- 1986. 1'985 1990 1986 1987 1988 1989 1990 TOTAL 15.90 21.04 18.63 22.11 20.34 20.73 22.47 Total Economic Services 6.42 4.83 6.73 4.74 4.02 4.31 4.88 Total Social Services 3.00 4.03 4.10 3.42 3.64 4.08 4.61 Education 1.76 2.74 2.27 2.41 2.68 2.84 3.17 Health 0.54 0.65 0.53 0.57 0.66 0.65 0.75 Other Social Services 0.71 0.64 1.29 0.45 0.30 0.59 0.68 National Defense 1.67 1.32 1.25 1.24 1.50 1.34 1.26 Total Public Services 1.80 2.46 1.99 2.72 2.51 2.25 2.72 Debt Service 3.00 8.22 4.56 9.98 8.67 8.57 8.49 Unallocated 0.00 0.18 0.00 0.00 0.00 0.18 0.53 Net L.nding 0.49 0.68 2.45 1.09 0.60 0.17 0.04  - 71 - debt servicing obligation.: Recent years have seen positive developments, however, as declining international interest rates and successful debt rescheduling exercises have reduced interest payments from almost 10 of GNP in 1987 to around 5.7Z in 1990; in terms of shares of the annual budget, interest payments have declined from 45Z of the budget to around 25Z over this period. 3.18 Interest payments, however, were not the only factor behind the increase in current expenditures. Personal services expenditures also grew sharply as a result of numerous salary adjustments granted to government employees by the Aquino Administration and some increase in the number of employees. The salary adjustments include a 10? across the board increase in July 1986, a 5? increase for career executive positions and a 302 increase of rank and file employees in selected agencies in 1987, and increases implied by the Salary Standardization Act which came into effect in 1990. As a result, expenditures on this account have steadily risen from 4.6? in 1986 to 6.2? of GNP in 1990 or from 20? of the annual budget to over 27?. 3.19 The increase in personal service expenditures must be viewed in perspective, however. Even with substantial growth in recent years, such expenditures are not out of line in comparison with neighboring countries. The Philippines spends a lower proportion of its budget dn wages and salaries for government employees than Malaysia and Thailand. The bulk of recent increases was due to the Salary Standardization Act and represented a transitional phase; now that the Act is in place, personnel expenditures are expected to grow at a more normal rate. Furthermore, the public sector work force remains a small proportion of the total labor force. The number of public sector employees has only increased modestly, from 1.19 million in 1986 to 1.25 million in 1990; within this, the size of the national government employee roster increased from around 832,000 to around 845,000. Much of this increase occurred in field-level staff rather than central office staff, in part due to the "nationalization" of local level teachers. 3.20 The sharp decline in investment outlays and the lack of any increase in maintenance expenditures is worrying in view of the clear infrastructure and maintenance deficiencies in the country.4 These trends also suggest that the fiscal adjustment since the mid-1980s has been borne a/ The prominent role of the larger interest payment obligation now shouldered by the national government is reflected in two notable features of recent budgets: (a) government expenditures net of interest payments were virtually the same (around 14.52 of GNP) during 1986-90 as during 1975-85 and (b) primary surpluses were generated in each of the last five years as non- interest expenditures were systematically lower than revenues. 4/ Even the "other MOE" category in Table 3.1 may not be adequately disaggregated to show trends in 'true" maintenance expenditures. However, sufficiently disaggregated data collected from the key infrastructure agencies and corporations confirm the impression that maintenance expenditures remain inadequate. Such expenditures have averaged about 12? of capital outlays in recent years.  - 72 - Fig 3.1 National Government Expenditure by Economic Classification (percent of GNP) Peraonal Personal 6.3% 386% Interest Payments 1.2% Other Capital Othr-1990 Capitl 293.0% al~ neet Payments ivsmn F19 3.2 Natlonal Government Enventment Other Current investment 6.8% ~~~~~2.1% Other Current 1975-1986 Total * 15.9% 1986-1990 Total * 20.9% Fig 3.2 National Government Expenditure by Sectoral Classification (percent of G3NP) Economic 0oia 6.4% 4.0oEonomIc 48% social Publi 3.0% ~~~~~Other ..:...:..Oher 1.7% Pubi 1.8% ~Debt Service 3.0% Debt Service 1975-1986 8.2% Total a 15.9% 1988-1990 Total a 20.9%  _ 73 - primarily by these expenditure categories. Such a pattern of adjustment is not unusual and reflects the politics of budget formulation in resource- constrained countries. It is alarming, however, to the extent that it essentially buys short-term adjustment at the expense of long-term growth; in so doing, such a strategy may sow the seeds of its own failure in that foregone growth may constrain resource mobilization and thereby create additional pressure on the fiscal balance. 3.21 Much of the decline in capital investment outlays occurred during 1986-88 as the new government sought to bring the fiscal accounts under control; in these three years, investment outlays plunged from a high of 2Z of GNP (or 11Z of the budget) to 0.32 of GNP (or 1.4% of the budget). This dramatic decline was partly reversed during 1988-90 as investment outlays climbed to 0.8Z of GNP (or 3.6Z of the budget). 3.22 Sectoral Categories. Several other trends become evident when national government expenditures are viewed from a sectoral or functional perspective. From this perspective, (a) economic service allocations contracted from 6.4% during 1975-85 to 4.8% during 1986-90 while (b) social service allocations expanded from 3Z to 4% and public services rose from 1.8 to 2.5%. The most notable trend was (c) the sharp increase in debt service allocations which rose from 3% to 8.2Z. 3.23 The relative contraction of allocations to overall economic services was matched by declines in the shares of all but one sub-category namely, Agrarian Reform. The five-fold increase in the relative share of this sub-category from 0.08z of GNP to 0.42% reflected the Aquino Administration's priorities: land reform via the Comprehensive Agrarian Reform Program has been a major element in this administration's development strategy. 3.24 Another major element of recent development strategy has been an emphasis on eduration. This is reflected in the increased relative share of education expenditures from an average of 1.76% of GNP to 2.7Z. Indeed, education's share in the national budget increased to 14Z in 1990, making it the second most important budget category next to debt service. 3.25 The reasons for the growth in the importance of debt service as a budget category have already been mentioned. It is the largest claimant on the annual budget, having increased from around 19% during 1975-85 to 39Z during 1986-90. In recent years, however, its share of the budget has been declining, from a high of 45% in 1987 to a low of 38% in 1990. This declining trend reflects external debt reschedulings and reductions (through debt-equity swaps etc.) and a softening of world interest rates. Public Sector Investment 3.26 Analysis of overall public sector investment, that is, comprising the combined capital outlays of the national government, public enterprises, and local government units, reveals three notable trends. First, there was a sharp decline in public sector investment, from 6.4Z of GNP to around 3.9Z. Second, there was a shift in the relative importance of the national government and  - 74 - public enterprises to the overall program; the share of investment originating in the national government rose from around 39% to around 53? while that of public enterprises declined from around 59? to 43Z. Third, there was a change in the composition of public investment in that the share of social infrastructure rose relative to that of physical infrastructure. 3.27 The sharp decline in public investment took place in two stages: there was a plunge during 1983-85 from 7.3Z to 4.3? followed by a more gradual decline to 3.2% in 1988. The dramatic plunge during 1983-85 was a response to a sharp reduction in resources brought about by the combination of a deep recession and a decline in external finance. Growth in the public investment program continued to lag GNP growth in the first three years of the new government as the disruption of the bureaucracy through changes in personnel and procedures hampered project implementation. Since 1988, however, spublic investment has risen steadily, reaching a level of 4.6? of GNP in 1990. 3.28 The national government became a relatively more important source of investment spending than it used to be largely because, starting in 1986, it reduced the flow of equity and loans to government cJrporations thereby depriving them of their main source (other than foreign loanf,) of funds. Some government corporations have since diversified their funding sources (by raising user charges and cutting operational costs) and have begun rebuilding their investment programs. However, several including the largest government corporation and largest investor, the National Power Corporation, remain heavily dependent on national government and foreign funding for their investment programs: if such funding is not forthcoming, the investment program is cut. 3.29 The shift in sources of investment spending led to a shift in the sectoral composition of public investment. As one would expect, sectors such as power and energy did not grow as much as sectors such as roads and transportation since the former are dominated by government corporations and the latter by national government agencies. In addition, national government spending priorities changed towards greater emphasis on social infrastructure; so sectors such as education, health and public administration grew relative to economic sectors such as agriculture and ransportation. 3.30 Some of these sectoral shifts were dramatic. The share of investment in the power sector dropped from 30? of the total in 1975-85 to 19? in 1986-90 while that of infrastructure (such as roads and bridges) dropped from 66? to 53?. On the other hand, the share of education almost quadrupled (from 2.2Z to 8.4?) while that of health tripled (from 0.6? to 1.8?). 3.31 The above-described changes in the size and composition of the public expenditures and investment programs could have significant implications for long-run growth in the Philippines. A proper assessment of these implications requires an understanding of the relationship between economic growth and public expenditures. This is elaborated in the next section. 6/ By some measures the ratio of public investment to GNP stood at 5Z in 1990; the source of the discrepancy is not clear.  - 75 - Fig. 3.3: Public Investments, by Source (Percent of GNP) National Government 2.3% National Government 2.2% Local Gov'ts Local Gov'te 0.3% 0.2% Govt Corporations Govt Corporations 3*4% 1.8% 19 1975-1985 1988-99 Total - 6.0% Total - 4.0% Fig. 3.4: Public Investments, by Source (Million pesos) P million 60 80 . -- _ _ - - - - - . __.._...._ = 40 _ ---_ .d. -. s0 --- ;------- - ------------.- ------- 20 10 76 76 77 78 79 80 81 82 83 64 85 86 87 88 89 90 = National Government 1 GCs EDi Local Governments  - 76 - Flg. 3.5: Public Investments, by Seotor (Share of Total) Infrastructure 64%s Infrastructure / \llifl / | :I Others ii 111iliiii: 13% M :2::::: Other Other Social v~~~~~~Others OE 11% 12% Other Social Education '~cjuctio~~ Agriculture 8% Agriculture 2%18-19 17% 1986-1990 1975-1985 Fig. 3.6: Public Investments, by Sector (Percent of GNP) .~~~~~~~~~~ 0 7B 76 77 78 79 80 81 82 83 84 86 86 87 88 89 90 I Infrastructure E Agricuiture E3 Education E Other Soclal Q Others  - 77 - The Prodctivity of Public Expenditures 3.32 The role of public expenditures in economic development has long been the subject of debate. Proponents argue that such expenditures are critical to development in that they (a) make up for the lack of private capital and initiative to develop infrastructure (b) compensate for market failure in certain areas (e.g. public goods provision) (c) exploit large positive externalities in the social sectors and (d) stimulate private investment by raising aggregate demand as well as by creating supportive and enabling infrastructure. Opponents argue, however, that the evidence of market failure and lack of private initiative is weak and that public expenditures do not have a net stimulative effect for one or both of two reasons: they may depress overall productivity since they are typically associated with less-efficient and non-competitive enterprises and they may crowd out private investments by increasing the cost and/or decreasing the availabiitty of credit to the private sector. 3.33 Recent empirical work has not provided a definitive resolution of this debate: while some issues have been clarified, others remain murky. For example, some studies have found that national productivity is strongly affected by the rate of public investment (Aschauer, 1988,1989); indeed, public investment is found to have an even stronger impact than private investment. Others find a correlation between growth and public investment which is positive but not necessarily very strong (Barro, 1989). Still others find that the relationship differs by component of public expenditure. One recent study (Diamond, 1990) involving a cross-section of 38 developing countries found the following: (a) while aggregate public spending did not exert a significant influence on real growth, this was due to the offsetting effects of its components in that capital spending did raise real growth but current spending did not; (b) further disaggregation sv:ggested that capital expenditures on social services had a bigger impact on growth than capital expenditures on infrastructure and (c) directly productive capital expenditures had a negative effect on growth. 3.34 Some tentative conclusions might be drawn from the above. It is clear that the concepts of aggregate public spending and aggregate public investment spending are not particularly useful. Even if uniform and non-conflicting results had been obtained for such concepts, it would be difficult to use them as guides for policy. To do this, it is necessary to disaggregate and examine various components. In this regard, the Diamond results noted above imply that operating and maintenance expenditures might be preferable to directly productive capital expenditures (on infrastructure) as far as short-run growth is concerned; a bias towards capital formation rather than capital maintenance could be detrimental to growth. They also suggest that the positive growth effects of capital expenditures may come through human capital formation (via sopial sector spending) rather than through the traditional channel of physical infrastructure (roads, power, etc.). 3.35 Empirical results for the Philippines from an analysis conducted along the lines of Aschauer (1989) suggest that public investment does lead to an increase in productivity and that investment in infrastructure and economic sectors is more productive than investment in agriculture and social sectors. The effects are, however, quantitatively mtodest: a 12 increase in the ratio of public investment to GDP raises labor productivity by only 0.04% and the same  - 78 - holds, upon disaggregation, for the effect of infrastructure investment. Given that the approach used here captures only the short-run effect of public investments, the results generated should be viewed with caution. This is so especially for the results on the productivity of social sector investments: these typically bear fruit in the long run and have often been shown to have returns superior to those of many standard physical infrastructure investments. Effec. of Public Investment on Private Investment 3.36 Another channel by which public investment affects national productivity is by way of its effect on private investment. There are two considerations here, conveniently captured by the terms 'crowding out." and "crowding in". Crowding out occurs when the effect of financing public expenditures through borrowing is to raise the cost andlor decrease the availability of credit to the private sector, thereby decreasing private investment. Crowding in can occur from the stimulative effect of public expenditures in one or both of two ways. There could be a demand-side effect as higher government expenditures raise aggregate demand, subsequently leading to an increase in private investment. There could also be a supply-side effect as public investments relax key infrastructure constraints thereby enabling higher levels of private investment. 3.37 These potential effects of public investment have been examined in a number of empirical studies (see Blejer and Khan, 1984; Chibber and van Wijnbergen, 1988). The empirical consensus is that infrastructure investments crowd in private investment whereas non-infrastructure investments crowd it out. There is weak support also for the view that public spending crowds out private spending through its financial market effects. We report below the results of an exercise carried out to investigate such effects in the case of the Philippines. The exercise had two parts, one dealing with private investment as a whole and the other focussing on foreign investment. 3.38 Our results suggest that public spending could contribute to crowding out through financial markets. The higher the availability of credit to the private sector, the higher is private investment. From other sources it is also fairly well established that higher public expenditures are associated with a higher public demand for credit (given the relatively low yield public revenue system) which, by implication, results in there being less available for the private sector. 3.39 Our results also suggest that government investment outlays have a positive direct effect on private investment whereas spending through government corporations (GCs) has a negative effect. This cannot be neatly identified with infrastructure and non-infrastructure investments since some GCs, such as the power and water utilities, provide much-needed infrastructure services. It does reflect the fact, however, that many GCs tend to compete with the private sector in various business areas and may therefore crowd it out. This is a plausible description of GC behavior in the late 1970s and early 1980s, when many of them were active in the manufacturing and mining sector and a fe held monopoly positions in other sectors (e.g. coconut oil processing).  _ 79 - 3.40 In any case, the mere provision of infrastructure is seldom enough to genlerate and sustain private investment; there needs to be a supportive economic and political environment as well. This point can be illustrated from the experience of export processing zones in the Philippines. Starting in the early 1970s, several export processing zones were established; they were provided with basic infrastructure such as roads, water and power and benefitted as well from fiscal incentives and the availability of imported inputs at world prices. However, on the whole, their performance has been disappointing in terms of investment, employment and export growth, especially when compared to neighboring countries. This is due in part to a poor choice of location for several of the EPZs set up in the Philippines and in part to a policy environment that has generally supported import-substitution activities. A good example of the location problem is that of th, Bataan EPZ, which is situated in a relatively remote area (about 200km from Manila) and has never performed to potential. Despite the availability of good basic infrastructure within the Bataan EPZ in the early years, it was unable to attract many investors. Over time, the lack of a growing revenue base led to cash shortages and inadequate expenditures on maintenance which led in turn to a deterioration of the basic infrastructure: all in all, it has turned out to be a poor public infrastructure investment. On the other hand, a similar investment made in the Mactan EPZ has turned out well primarily because this site is well located, being cheek-by-jowl with Cebu. C. Private Sector Role in Public Sector Projectse The BOT Concept 3.41 In the past few years there has been much discussion of whether and how the role of the private sector can be enhanced in the delivery of basic infrastructure and services in developing countries, and the Philippines is a good case in point. As discussed above, the Philippine Government's budget has been hard-pressed to finance a minimum level of public investment expenditures. In order to provide for the infrastructure needs of economic growth without having always to rely on scarce public funds, governments have been turning to such arrangements as the Build-Operate and Transfer (BOT) Scheme. Under the BOT model, a private developer or contractor is awarded a concession to build, own and operate a given plant for a fixed period (say 15-20 years), after which time the firm transfers the project to the Government, which agrees to buy the plant at a pre-agreed price or formula (or it could even be free at that point in time). There are two variants of the BOT scheme, viz. the Build, Own, and Operate (BOO) arrangement, under which the foreign firm does not transfer the plant to the Government and instead continues to own and operate the plant in perpetuity; and the Build, Transfer and Operate (BTO) version, under which the foreign firm builds the plant, soon thereafte% transfers ownership to the Government, which in turn hands the plant back to the private firm for the latter to operate it on a concession basis for a given period. Although there are some e/ This section draws on a recent Bank study entitled: 'Private Investment in Power and Coal in Asia -- Selected Case Studies and Lessons," Report No. 9884-ASIA, November 1991.  - 80 - differences between BOTs, BOOs, and BTOs -- primarily the ownership question -- in this discussion the phrase BOT will be used generically to capture the other variants as well.7 3.42 There are three primary advantages of BOT projects each of which would seem relevant for the Philippines: (i) such projects can supplement infrastructure capacity without constituting a drain on scarce public investment funds. Thus, the BOT scheme can substitute for domestic resource mobilization for part of the public investment program; (ii) the bulk of the financing obligations, which consist of both equity and loan funds, shifts from the Government budget to the private sector. To the extent that the equity portion is substantial, the BOT arrangement is therefore also a means of increasing the flow of foreign investment funds; and (iii) where the project involves substantial operating eomplexities and/or risks, the BOT arrangement allows the Government to reduce its risks by passing these onto a qualified firm with international experience. Thus, given the size and complexity of most large- scale energy and transportation projects, if done on a BOT arrangement, the project can thereby secure access to up-to-date technical expertise from abroad on a continuing basis. However, technical innovation or large size are not necessary conditions of BOT projects. The Hopewell Navotas Gas Turbine project, is a good illustration of a successful, small BOT project. 3.43 There can also be disadvantages to BOT initiatives. They involve complex negotiations and axpenditure of time and resources by both parties. The few projects done so far in developing countries have not been "pure" BOT schemes; rather, the respective governments have been asked for substantial financial guarantees to cover the loan and foreign exchange risk for the private party. To the extent such guarantees are given, this dilutes a major advantage of the BOT concept for many developing country governments. Providing a loan guarantee increases external indebtedness and typically depletes the country's access to international credit, which is much the same as when a government borrows directly to fund its public investments. The merits of a BOT therefore rest firstly on the balance between the "costt of the guarantees provided by Government and the leveraging value (additional finance thereby secured) of the guarantees; and secondly on whether there are specific operational efficiencies to be obtained by having the project run by the private sector. The tradeoffs must be carefully weighed, including the possibility that protracted negotiations might not lead to a concluded deal. Where adequate financial risk transfer is resisted by the private party, a better arrangement might be to have the private company build the plant and then operate it on a concession basis. 7/ Although the BOT scheme has been extensively used in power projects in the United States, it is fairly new in the developing countries, having been applied recently in the Pakistan Hab River Power Project (project still to be implemented) and four other projects which have been implemented: the China Shajiao Power Project, the China An Tai Bao Coal Project, the Indonesia BHP-Utah Coal Project, and the Philippines Navotas Power Project. There have also been detailed discussions on proposed BOT projects in Turkey, several Latin American countries, and in Eastern Europe.  - 81 - The Potential for BOT Projects in the Power Sector 3.44 In the late-1980s it was apparent that the National Power Corporation (NPC) would be unable to meet electricity demand and by 1987 power shortages and brownouts became common. Clearly, NPC's investment program was insufficient to meet a situation of growing power shortages. In July 1987 Executive Order No. 215 was issued, which for the first time allowed private sector participation in the generation and transmission of power. Under this executive order, NPC's exclusive rights over power generation were withdrawn. In 1988 NPC issued notification of bid offers for private generation projects and it received bids to construct two power projects -- the Navotas Project and the Batangas Power Project. Annex A reviews the project experience, which is summarized below. 3.45 The Navotas gas turbine project, which was successfully completed in March 1991, involved the installation of a 210 MW gas turbine generation capacity, which was procured and installed on a BOT basis by Hopewell Holdings Ltd., based on a similar project that it had done in Shajiao, China. Under the project arrangements, Hopewell is to operate the project for 12 years, after which it will transfer the project to NPC. The Cogentrix Batangas project was planned to be about the same size -- 220 MW cogeneration capacity, consisting of four 55 MW. steam turbines, which were to obtain their steam from eight coal boilers. However, the project never went ahead. Hopewell's successful experience was in marked contrast to that of Cogentrix, and it is useful to examine why that was the case. 3.46 Critical factors in the success of Hopewell's Navotas project include its experience in implementing a similar project in China, an adequate financing package, and the fact that it was able to come to an agreement with NPC for purcb-?se of the power generated. Also, an innovation in the agreement provides that fuel input will be supplied by NPC under given efficiency parameters, which avoids the need to have fuel pricing indexation arrangements. 3.47 In contrast, the history of the Batangas Cogeneration Project presents some valuable insights into the potential pitfalls of BOT projects if their arrangements cannot be worked out satisfactorily. Discussions were drawn out and mixed signals and poor coordination at the government decision-making level played a key role in the sponsors' failure to proceed. There was no lack of laws and regulations governing private power generation. However, the inability of Cogentrix and NPC to come to an agreement on the purchase of power was the single most important reason why that initiative failed. There were also difficulties regarding the time it would take to obtain environmental and construction permits and Cogentrix lacked a strong local joint venture partner. Future Directions 3.48 Although groTith in demand for electricity fell from 9.4 percent in 1987 to under 5 percent in 1990, demand is still outpacing the ability of NPC to construct new power plants. Outages of thermal power plants is exacerbating the situation of severe shortages and blackouts, especially in Luzon and Mindanao. The output losses to industry on account of inadequate power are estimated to amount to about $1.1 million dollars per day. During the period 1990-1999, it is estimated that the Philippines will need to add about 3,700 Mws of new  - 82 - capacity costing some $7.5 billion ($4.7 billion of which is needed for new generating capacity). It is estimated that no more than a third of these investment requirements will be sourced domestically and 72 percent will have to come from foreign equity investment (such as BOT sponsors) and from foreign loans to the Government and NPC. Since NPC is already burdened with debt and has to generate from its own resource. about 30 percent of its capital expenditures, and since the pressures of the fiscal deficit indicate that much of this desired $7.5 billion of energy generation and transmission investment will have to come from foreign sources of finance, the decision to assign a larger role to the private sector in the generation of energy is a welcome and necessary step. 3.49 In August 1990 NPC awarded to Hopewell Holdings Limited (Hong Kong) the right to develop a 700 Mky coal-fired BOT power plant in San Juan Batangas on Luzon island. NPC had issued a request for proposals for the coal-fired project in November 1989, and prequalified fourteen firms for bidding. When ready (start-up schedul'd for January 1994), this project would be the second private power plant comp.'eted under E.O. 215. 3.50 Government also appro,ed in 1990 a proposal to build on a BOT basis a 440 MW geothermal project in Tongonan, Leyte. The plant, designated as a renewable resource production facility under E.O. 215, is expected to come on line in late 1997. It will serve the Metro Manila load center on Luzon, via an underwater transmission line. The project was originally included in the NPC expansion plan, and has been accorded "preferred pioneer status", entitling it to tax benefits and 100 percent foreign ownership. A series of mini-hydro plants with a total combined capacity of 11.5 MW have also been approved. These small projects will supply electric power to industrial users, cooperatives, and service utilities in the localities where they are to be built. 3.51 Table 3.3 summarizes the major BOT projects in the power sector that are under accreditation, including their stattus and planned completion dates. Even if only a half of the BOT projects are built, it would add a total of 830 VW of generation capacity in the country, which would amount to about 20 percent of the total estimated increase in generation capacity that is required over the period 1991-1999. Such an increase in generation would amount to about an increase of 30 percent over NPC's current generation capacity. This would represent a very sizeable expansion indeed, especially if it is to be provided entirely by the private sector using external sources of finance with no drain on the Government budget. Should the BOT scheme be successful in delivering the four projects listed in the table (all of which are larger than the Hopewell/Navotas project), it would represent a dramatic change in the arrangements regarding the actual financing of infrastructure programs in the country and it would represent a major achievement in the quest to attract foreign investment into the economy. 3.52 It is not intended that BOT be restricted to the power sector. In July 1990, an Act of Congress formally permitted BOT ventures in a broad range of public infrastructure investments, including those in transport, communications, water supply, sewerage, health, and environmental management. The fact that the first few BOT projects were in power is not an indication that there are no reasonable prospects in these other sectors. On the contrary, port operations, airport services, shipping, warehousing are all illustrations of  - 83 - where the private sector on a BOT basis (i.e. with 100 percent foreign ownership and 100 percent private risk) could offer a viable product or service, in an area where the public sector or the domestic private sector is not able to deliver a product/service of acceptable quality or standard. It may be that concession arrangements or service contracts will prove over time to be an easier and quicker method to pursue, especially in the case of where there are linear programming-type activities to manage, such as the mass movement of freight goods and passenger traffic. Foreign firms that are experienced in running ferry operations for inter-island shipping, or handling containerized packing for goods transferred at road-rail-port depots would be suitable candidates for soliciting BOT proposals. D. Government Corporations: Financial and Management Issues 3.53 To the extent that Government Corporations (GCs) continue to receive support from the Government in the form of subsidies, loans, grants or equity infusions, their economic and financial performance affects directly the management of public resources. Three issues are important to achieving greater efficiency in public enterprises: privatization, financial autonomy, and management improvements. The first issue relates largely to resource mobilization in that the sale of Government assets contributes to non-tax revenues; this issue was discussed in Chapter 2. In this section the public enterprise sector is analyzed, recognizing that the net flow of budgetary resources to public enterprises is in part determined by the productivity and efficiency of such enterprises. 3.54 Policy Framework. The ecunomic reform program of the Aquino Administration attached high priority to rationalization of the public sector. This became a matter of some urgency because the public enterprise sector had come to represent a large financial drain by the mid-1980s. For example, by 1984 budgetary contributions to GCs had increased so substantially that they accounted for 22.5 percent of the total budget, up from a share of less than 8 percent ten years earlier. Clearly, a situation of government corporations draining away nearly one-quarter of the budget was unsustainable. It was vital to wean public corporations away from fiscal dependence. To do this required making them financially and managerially independent and accountable for their own decisions.  Table 3.3: Status of Private Power Projects (as of September 1990) Project Sponsor Type Capacity Fuel Status Start-up (MW) Navotas Gas Turbine Hopewell Holdings Ltd. BOT 200 Diesel Completed 1115191 (Hongkong) San Juan, Batangas Hopewell Holdings Ltd. BOT 700 Coal Awarded July 1995 Coal-Fired Power (Hongkong) 1990, after Plant int'tnl bidding. co C o n t r a c t I ~~~~~~~~~~~~~negotiation ongoing Tongonan, Leyte a. Design power/Mitsui BOT 440 Steam Under Evaluation 1997 Geothermal Power (New Zealand/Japan) Plant b. ORMAT Inc. (USA) BOT Steam Under Evaluation 1997 Limay, Bataan a. International Power BOT 300 Bunker C Under Negotiation 1993 Combined Cycle Gas Company (USA) Turbine Power Plant b. Asea Brown Boveri BOT (Switzerland) c. General Electricl BOT Mitsui (USA/Japan) Batangas Coal-Fired IPCO (Singapore) BOT 220 Coal Under Evaluation 1994 Cogeneration Power Plant Total 1660  - 85 - 3.55 The public enterprise reform program was launched with the followirf' policy commitments: (i) there would be minimal government intervention in the runni:.g of the GCs; (ii) there would be a drastic reduction of support and special privileges for GCs; and (iMi) greater flexibility and autonomy for GCs would be encouraged as a prerequisite for accountability in performance. These commitments underpinned the following rolicy objectivess (i) there would be a larger role for the private sector in the economy, especially in manufacturing activities; (ii) the fiscal deficit would be reduced; and (iii) improved effectiveness of the GCs would be sought by insulating them from political interference. There was broad-based support for the GC reform program, both domestically and from the donor community. Financial Autonomy 3.56 The reforms affected th,e Government's financial links with GCs in a fundamental manner. The specific steps taken to encourage financial autonomy included: the removal of tax exemptions, imposition of restrictions on the issuance of Government guarantees for GC borrowings, imposition of interest on Treasury advances, curtailment of subsidy grants, and tighter enforcement of payment of dividends from those GCs making profits. As part of these measures. a Tax Expenditure Fund was set up in the NG budget to cover tax exemptions. 3.57 The above measures and the emphasis on productivity improvements at the entity level led to a reduction of GC dependence on the budget. The net flow of funds from the NG to the GCs, which had grown from 1.6 percent of GNP to 5.6 percent between the period 1976-86, declined to 1.5 percent in 1987 and dropped further to around 0.3 percent in 1988-89. This decline was made possible in large part by the assumption of external debt service liabilities of GCs by the national government; however, even after adjusting for assumed debt service, there is a decline in net flows to GCs suggesting that there has been a mearingful reduction in dependence on the budget (see Figure 3.7). This is reflected also in the trends of most individual transfer categories: equity infusions and net lending declined in nominal peso terms as well as relative to GNP (see Figure 3.8). 3.58 There was also a marked increase in transfers from GCs to the national government. The bulk of such reverse transfers was made up of interest payments, the government's share in the income of the Philippine Amusement and Gaming Corporation (PAGCOR) and dividend payments by some GCs (see Figure 3.9). The overall picture of improvement should be qualified by two considerations. First, subsidy flows did not decline, although they were directed at more deserving entities. Second, aggregate transfer trends are heavily influenced by transfers to the National Power Corporation, the largest public corporation, so much so that preliminary data suggest that, in 1991, transfers to it alone exceeded the reduction in transfers to all other GCs combined. The financial position of NPC deteriorated sharply in 1991 because of delays in adjusting user charges, which resulted in the NG having to provide around P5 billion of transfers (in loans and equity) to NPC. This represented about 50 percent of total budget support to 14 GCs. 3.59 Subsidies. Subsidy contributions to GCs are recorded as having risen in the period after 1986. This reflects a decision to make transfers more transparent. Transparency was accomplished in two ways, by including tax exemptions in the subsidy  - 86 - BOX 3.1: LIST OF 14 MONITORED GCS AND RESPECTIVE MANDATES GC MANDATE EPZA Export Processing Zone Authorit Operation and management of export processing zones LWUA Local Water Utilities Administration Water supply and waste water disposal outside Metro-Manila LRTA Light Rail Transit Authority Construction, operation and maintenance or lease of LRT systems MMTC Metro Manilh Transit Corporation Operation of bus transport services MWSS Metropolitan Waterworks and Sewerage Systems Water supply and waste disposal in metropolitan areas NDC National Development Corporation Holding company for developmental investments NEA National Electrification Administration Financial intermediary and procurement agent for electric cooperatives NFA National Food Authority Promotion of local grains industry NHA National Housing Administration Provision and maintenance of adequate housing NIA National Irrigation Administration Development and maintenance of irrigation systems NPC National Power Corporation Generation and transmission of bulk power PNOC Philippine National Oil Company Transporting, refining and marketing crude oil and petro products, and development of indigenous energy resources PNR Philippine National Railways Operation and maintenance of railways Source: Department of Finance  - 87 - Figure 3.7 Flow of Funds to Government Corporations 6 e r 0 4 e t 3 2- p I~~~~~~~~I I 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 - Net Flows 4 Less Debt Service figures and by preferring subsidies to other forms of transfers whenever justifiable, such as when Government impo7es on a corporation a pricing policy that results in financial losses. Between 1988 and 1989 there was a sharp jump in the level of subsidies, mainly to agencies carrying out some of the Government's socially-oriented programs, e.g. the National Food Authority (NFA), the National Irrigation Administration (NIA), the National Electrification Administration (NEA), the National Housing Administration (NHA) and the Land Bank. Table 3.4 below shows the major GC subsidy recip.-ents for the two periods 1980-85 and 1986-90. Of significance is that the subsidy to the Fertilizer and Pesticide Authority (FPA) was virtually eliminated, in keeping with policy reforms supported by a World Bank adjustment loan in 1984. Subsidies to the five other GCs mentioned above increased significantly and accounted for 58 percent of total subsidies to GCs in the 1986-90 period, compared to only 6.8 percent of total subsidies in the preceding period. The two largest recipients of subsidies were the National Food Authority and the National Irrigation Administration. 8/ Subsidy figures after 1986 include tax exemptions granted to various GCs. As such, they are not directly comparable to figures before 1986. Thus, it is difficult to say whether subsidies have actually risen over time.  - 88 - Figure 3.8 Gov't Transfers to Gov't Corporatlons p _ 0 r 76 77 - 8 ---------384868 8 8 9 0 n t~~~~~~~~~Egr 3. 0 N O _ I 0 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 Subsidy Equity ED Net Lending Figure 3.9 Gov't Corporatlon Transfers to Gov't 0.5 P 0.4 ------.-- e r c O 0.3 - ----- n t ID 0.2 -- - - .- . I G N 0.1 . P 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 MDividends M Share In PAGCOR ED Interest  - 89 - Table 3.4 The Major Subsidy Recipients (in million Pesos) 1980-85 Z of 1986-90 Z of Total Total Fertilizer and Pesticide Authority 1,504 27.0 15 0.1 National Kidney Foundation 353 6.3 0 0.0 Philippine Crop Insurance Corporati3n 198 3.6 144 1.3 Philippine Coconut Authority 171 3.1 250 2.2 National Food Authority 166 3.0 2,603 23.3 National Housing Authority 159 2.9 604 5.4 National Irrigation Admin'stration 49 0.9 2,227 19.9 Land Bank 0 0.0 570 5.1 National Electrification Administration 0 0.0 480 4.3 Others 2,968 53.4 3,888 39.2 TOTAL 5,568 100.0 11,161 100.0 Source: DOF-Bureau of the Treasury 3.60 Equity Contributions. The other major form of NG financial transfers to GCs is equity infusion. In the mid-1970s such equity contributions were around 1.5 percent of GNP, but by 1982 they had climbed to 3.1 percent. In 1985 they stood at 2.6 percent of GNP and then started declining and were only 0.2 percent in 1989. Three features of trends and patterns in equity transfers are worth mentioning. First, much of it went to government financial institutions. The Development Bank of the Philippines (DBP) was given a large share of total equity transfers through 1985, largely to stave off its bankruptcy from the accumulation of bad loans; in 1986-87, however, the inevitable could no longer be postponed and DBP was rehabilitated via the transfer of its external liabilities to the national government and a fresh infusion of equity. The Philippine National Bank alco experienced a similar bail out and equity transfer in 1986-87. Regarding NPC, in addition to the P555 million equity contribution mentioned in para. 3.57, the NG converted P1.96 billion of previous advances into additional equity of the corporation. 3.61 Second, equity transfers to the National Power Corporation have undergone large fluctuations. In the first half of the 1980s, NPC received a large chunk of total equity transfers, in part no doubt to support its massive investment in the Bataan Nuclear Power Plant. The Aquino Administration sharply reduced equity transfers to NPC during 1986-90, this time reflecting the overell cutback in infrastructure investment. In 1991, however, NPC was given an equit, infusion of P2.5 billion which was around 612 of the total equity transfer program for 1991 and considerably more than the amount provided in the previous five years. This demonstrates the extent of the deterioration that occurred in NPC in 1991. The need to prop up NPC's finances was a  - 90 - direct result of the failure to adjust its user charges. Since NPC is the largest public corporation, variations in its financial performance have substantial consequences for the public sector deficit. This points to the need to have timely adjustments of electricity tariffs in the future, both to keep the enterprise viable and to preserve overall fiscal health. Table 3.5: The Major Equity Recipients (in million Pesos) 1980-85 % of 1986-90 Z of Total Total Development Bank of the Philippinec 14,848 28.6 3,691 17.8 National Power Corporation 8,593 16.6 952 4.6 National Irrigation Admn 3,315 6.4 2,000 9.7 Philippine National Bank 3,050 5.9 3,750 18.1 National Food Authority 386 0.7 1,567 7.6 National Development Co. 4,265 8.2 216 1.0 Philguarantee 1,550 3.0 1,480 7.2 Others 15,901 30.6 7,023 34.0 Total 51,908 100.0 20,679 100.0 Source: DOF-Bureau of the Treasury 3.62 Third, aside from the above NPC infusion, the overall decline in equity contributions that occurred between 1980-85 and 1986-90 led to a slowdown of fixed investment by the GCs.9 Starved for capital, and operating under a policy framework that strongly discouraged investment on the part of those GCs that competed with the private sector (as in the case of PASAR and PHILPHOS) the restrictions on equity infusion put an effective brake on expansion of the GCs. To the extent that GCs embarked on investments in the 3'86-90 period, the entities were encouraged to fund their expansion/ modernization requirements from funds other than NG equity. Net lending to the GCs also declined mainly because the NG started to charge market interest rates on any advances extended to the GCs and also because in some cases the advances were converted into equity. 3.63 Tax Exemptlons. For a long time GCs had been given tax exemptions in lieu of equity contributions. In the late 1970s these tax exemptions grew to be substantial, amounting to as much as 4.3 percent of GNP in 1978. The exemptions were broad-based applying to import taxes, income taxes, withholding taxes on interest, real estate taxes, and sales taxes. In late 1984 the exemptions were cut back and amounted 9/ It might be noted that the investment program of GCs was also affected by a cash squeeze produced by declining access to external loans and increasing debt service obligations.  91 - to 1.8 percent of GNP and 1.0 percent in 1985. In 1986 the new government withdrew all exemptions, but the Fiscal Incentive Review Board (FIRB) -- whose function it was to determine the list cf GCs that could still be eligible to receive tax exemptions -- granted exemptions to six GCs -- the APT, PITC, the Duty Free Philippines, NPC, the Private Debt Restructuring and Payment Corporation and the FPA. The exemptions amounted to 0.4 percent of GNP in 1986. As for the rest of the GCs, they could claim relief through the Tax Expenditure Fund by means of an explicit and transparent budget appropriation. Also, GCs were charged fees for some services previously rendered free, such as the imposition in 1987 of fees for any guarantees provided by the Department of Finance on GC borrowings. 3.64 The issue of tax exemptions illustrates another problem with regard to the operation and viability of public corporations. If tax exemptions were not provided, user charges would have to be raised. This is particularly true for NPC which presently accounts for the bulk of tax exemptions among GCs. The fiscal drain of a GC must be seen not only in terms of transfers such as subsidies and equity but also in terms of revenue foregone by the Government through providing tax exemptions. Indeed, to the extent that such exemptions allow the charging of lower rates they contribute not only to fiscal stress at the macro level but also to distortions at the micro level, distortions which provide misleading signals to guide the investment program and the search for efficiency among GCs. 3.65 Dividend Payment Responsibilities. Before 1986, dividend payment1o responsibilities were loosely enforced. While enforcement efforts have improved since then, the results have not been exceptional. Of the 33 GCs from which dividends were expected, onl six paid any dividends in 1990. While non-payment was mostly due to the incurrinig of losses, other causes were also important. Among these was a special exemption given to those GCs (primarily the GFIs) who invested surplus cash in Treasury Bills. Other reasons for non-paymient of dividends involved delays and lags caused by auditing. The situation improved in 1991 as the number of remitters rose to 14. Management Issues 3.66 Public sector management notions have undergone a radical transformation in many developing countries in the past decade. From a situation in which productivity and cost considerations used to be largely ignored in the provision of many public services, the emphasis has shifted to privatizing public enterprises where feasible and striving for cost-efficient commercially-oriented operations where the service, for a variety of reasons, is likely to remain in public hands. Productivity measures and performance targets for public enterprises are being successfully implemented in several countries. The Pailippines began to experiment with a productivity-oriented approach several years ago under the auspices of a Bank adjustment loan (the Government Corporation Reform Loan). This approach involves the development of multi-year corporate plans and the setting of annual productivity targets. In the past three years the Performance Evaluation and Incentive System (PEIS) has been implemented in 5 large 10/ E0399 requires each profit-making GC to remit dividends equal to 102 of its net earnings. Prior to 1986, the requirement was a more modest 5%. E0399 does not apply to those GCs (such as NPC, NEA, MWSS) whose charter provides for other rules on the distribution of profits. It also does not apply to GSIS and SSS.  - 92 - public enterprises (NPC, MWSS, NIA, PPA, NHA) in 1989, twelve more in 1990 and another 17 more in 1991 for a total of 34 GCs. In addition, management reforms have been implemented in an effort to insulate GCs from political intervention, instill operating practices akin to those of the private sector, and improve corporate viability under a difficult macroeconomic situation. 3.67 A key institutional issue is whether the current policy framework is consistent with the goal of "corporatization" of the public enterprise sector -- defined as the ability of a public corporation to operate as a company, irrespective of whether or not it is to be privatized. Under the GC reform program, it was envisaged that some of the more efficient GCs needed to be able to grow and consolidate their operations before they could be sold as healthy, viable companies, and the ones that were not in such a condition would need to be revamped or upgraded in order to bring them to vendible status. The Government Corporate Monitoring and Coordination Committee (GCMCC) was reconstituted with a mandate to oversee this process and to implement the PEIS in the GCs designated to be retained in the public sector (responsibility for upgrading and monitoring of GCs designated to be privatized is left to the disposition entity). The GCMCC also has a key role in the institutional follow- up of the GC reform process and specifically in the review of financial autonomy, operational efficiency/productivity questions, and related financial and personnel issues. The GCMCC is also involved in *disseminating the experience under the PEIS program. 3.68 Experience with PEIS. The PEIS was formally begun in 1989. The GCMCC managed the evaluation prcgram, which was based on well-accepted systems of public enterprise monitoring that had been developed in the late 1970s and early 1980s in other countries. The PEIS involved adoption of common performance criteria (return on assets, operating ratios, debt service ratio, current ratio), which were used in the case of 15 GCs and specialized criteria (billing efficiency ratios NPC and MWSS, farebox ratios for PNR and LRTA), which were used in the case of 16 GCs. Refinements were made to the system, such as the introduction of weights in 1990 -- financial or profitability indicators were given a 40 percent weigbt. As stated above, the GCMCC started with modest coverage and was expanded in 1990-91. For 1992, ten more GCs will be included. Results are shown below for the five pilot GCs with whom the program was started; results for several others are available from Government. 3.69 Despite the modest targets, the GCs showed only a 43.2 percent success rate (number of performance criteria attained divided by the total number of criteria) in 1989 and 35 percent in 1990. The corporations with the highest attainment ratios were the Philippine Ports Authority (PPA) and Metropolitan Waterworks & Sewerage (MWSS), both of which attained nine out of the seventeen targets in 1989, whereas the National Irrigation Administration (NIA) attained only two out of 14 targets. The PEIS has several weak points: first, some indicators are more important than others for different corporations, e.g. collection indicators in the case of NHA and NIA, and billing efficiency in the case of NPC and MWSS; second, it is quite difficult to develop an adequate weighting system; and, third, performance may be affected by factors outside management control, such as natural disasters, or limitations in operational autonomy. It is useful, therefore, to supplement the impressions provided by the overall success rate measure with other, more specific, measures of operational efficiency. Two measures are discussed below, collection efficiency and financial performance.  - 93 - Table 3.6: Results of Performance Evaluation No. of Performance Not Success Government Corp. Year Criteria AtZained Attained Rate National Power Corp. 1989 10 4 6 40.0O 1990 10 5 5 50.OZ Metropolitan Wate- S&S 1989 9 6 3 66.7Z 1990 8 3 5 37.5Z National Irrig. Admin. 1939 8 0 8 0.0% 1990 6 2 4 33.3Z Philippine Ports Auth. 1989 9 6 3 66.72 1990 8 3 5 37.5Z National Housing Auth. 1989 8 3 5 37.5% 1990 8 1 7 12.5Z Total/Average 1989 44 19 25 43.22 1990 40 14 26 35.OZ Source: Department of Finance - GCMCC 3.70 Operational EfficiencY. A key aspect of the viability of any public corporation is its operational efficiency. While there are several different ways to measure this for different corporations, depending on the type of goods and services they provide, a useful and commonly-applicable measure is that of collection efficiency, which refers essentially to performance with respect to collection of accounts receivable. Table 3.7 provides some data on this for a number of GCs in the Philippines. 1wo aspects are worth noting. First, there is great variation in collection efficiency. Some GCs, such as NPC and MWSS, are excellent at collecting whereas others, such as NEA, NHMFC, and NIA are very poor at collecting; as a result, the laggard companies build up arrears which they are also very ineffective in clearing. Second, the collection performance of NPC has a great influence on the average; given its preponderant size among GCs, weighted averages of performance are typically dominated by the performance of NPC.  Table 3.7: Collection Efficiency 1985 1986 1987 1988 1989 1990 Average NPC 95.0X 81.0Z 89.7t 87.2? 81.22 84.0% 87.2Z LWUA - Current Accts 54.02 57.0% 58.02 39.1? 47.8? 43.22 49.8? Arrears 53.0? 47.0? 43.0? 2.9? 7.3Z 0.82 26.5Z NIA - Current Accts 39.72 40.3Z 41.5? 39.4? 41.7% 51.02 42.3? Arrears 5.4Z 4.4? 3.3? 2.5Z 2.8Z 50.0? 11.2Z MWSS 98.9% 93.7Z 75.5? 87.8? 83.32 71.02 84.7Z NEA - Current Accts 22.5% 20.3% 23.52 17.82 17.8Z 20.5Z 21.42 Arrears 15.2? 13.8? 20.9? 10.12 15.0? 16.2Z NHA 64.0? 15.OZ 47.0? 11.0? 48.02 56.0? 39.8Z HIGC - Current Aects n.a. n.a. n.a. 54.6? 72.5? 83.62 70.2% Arrears n.a. n.a. n.a. 38.02 41.02 73.4? 50.8Z NHMFC Current Accts n.a. n.a. 52.1% 38.8? 38.92 73.6?1/ 52.9? Arrears n.a. n.a. n.a. 19.2% 12.0Z 12.02 14.4Z EPZA 89.0X 82.OZ 84.0% 57.52 89.5% 77.8% 69.1? Non-Weighted Averagel/ 65.9Z 52.8Z 57.7? 47.0? 56.22 63.02 57.52 Weighted Average2/ 91.82 78.9? 85.0? 78.32 86.22 80.1Z 83.OZ 1/ Excluding arrears 2/ Weight based on 1987 receipts Source: Aquino, T. et al., Public Sector Resource Mobilization, and Department of Finance. 3.71 Financial Performance of GCs. The extent of improvement in the financial performance of GCs in recent years depends on how government subsidies are treated. Gross of subsidies, the operating ratio (receipts divided by expenditures) of the 14 major non-financial GCs shows clear improvement between 1985 and 1989; there was a decline in 1990 but this still left the ratio above its 1985 level. Net of subsidies, there is a similar picture of improvement luring 1985-89 but the decline in 1990 brought the operating ratio back to its 1985 level. The importance of subsidies is even greater when financial performance is viewed in terms of net income figures. In this case, financial performance clearly deteriorated as net income declined from 2.2 billion pesos in 1985 to 1.8 in 1989 (net of subsidies). The chief reason for the decline appears to be the failure to raise internal cash generation fast enough and  - 95 - substantially enough to cope with changing market conditions. User charges actually declined in real terms over this period. Note that several GCs benefited from the general decline in oil prices and interest rates over this period but were still unable to improve net income without a subsidy from Government. This suggests that much greater attention must be paid in the future to improvements in management of government corporations with a view to reducing costs and improving productivity. 3.72 Operational Autonomy. Measures to improve operational autonomy involved giving more flexibility on matters affecting personnel, procurement, investment decision controls, and negotiations with creditors. However, on the freedom to adjust user charges, some of the GCs are able to set their tariffs (e.g.MWSS), whereas others (NPC, NEA) do not have delegated powers. NPC and NEA have to obtain the approval of the Energy Regulatory Board for any rate adjustments, while other boards have the last say on user charges or rates on output/products in their sector. The transport rates of the Metro Manila Transit Corporation (MMTC) are subject to the same restrictions as the private sector transport companies. The NFA is another case of an entity having no power to adjust prices without approval of the National Government. 3.73 Other GCs, however, do have flexibility to adjust their rates. The financial institutions can adjust interest rates on lending and are not subject to any state controls. GCs, however, cannot seek external borrowings without government permission, because given the debt crisis that ensued in 1984, it is not possible to borrow in the absence of government guarantees. Permission to borrow from abroad has to be given by NEDA. NEDA's clearance involves assessment of the relative priorities and desirability of the project. Government policy is to try and refrain from providing guarantees and for private sector projects and to allow the multilateral institutions to secure sovereign guarantees for public corporate loans. On personnel matters the GCs are limited by the civil service coverage and by the regulations of the Department of Budget and Management (DBM). Given the power of the unions, the ability of managers in the Government corporations to 'hire and fire" does not truly exist. Aside from the shallow depth of the capital markets, this factor may be the dominant reason why the privatization of the big-ticket items has been slow to get off the ground, as the large corporations do not feel that they are in a position to implement wide-scale staff cutbacks. The ab!lity to send employees overseas for training is limited given the pressing financial position of many GCs. Where training grants are available, which are managed by NEDA, the possibilities are greater for matching training opportunities abroad with the needs to upgrade the skill mix of public sector employees. The foregoing points regarding operational flexibility and the previous discussion of financial performance suggests that the issue of operational autonomy, especially with regard to setting prices (user charges) and making personnel decisions (cost control), needs to be re-examined by policy-making authorities. 3.74 Some Conclusions. Three main conclusions may be drawn from the GC reform process so far. First, despite some improvements, government corporations remain significantly dependent on the national government; this is quite important to their profitability outcomes. Second, the financial and management performance of GCs has been weak in recent years, primarily in the areas of pricing (user charges) and cost control. It is important to stress the importance of cost control since consumers are more likely to accept higher user charges if they perceive clear improvements in operational efficiency. Third, the problems of critical entities, such as NPC, need to be distinguished from those of corporations in need of improvement but whose performance is considered not likely to jeopardize macroeconomic stability. This is  - 96 - Fig. 3.10: Financial Performance of %Cs a) Operating Ratio (%) 2 1.6 < 1.5 ----------- - - _ _ 1 1 1 ------------------------------- ~~~~~~~~~~_ __ .__ _ __ _. ___ 0.5 __ 1985 1988 1987 1988 1989 1990 Gross -4- Net of Subsidy 8b) Net Income (billion pesos) 6 4 24 - ------ --- - - --___ _ _ -21 1986 1986 1987 1988 1989 -* Gross 5 Net of Subsidy  - 97 - not the case with NPC, as the Government's cash infusion in May 1991 into that corporation amounted to nearly one-fifth of the projected national government deficit for 1991. With this in mind, a more detailed discussion of NPC is provided in Box 3.1. Selected Recommendations 3.75 Given the above analysis, the following recommendations are pertinent: (a) the NG should continue to review its resource allocation system to find out whether specific high-subsidy programs in food and irrigation should be maintained. The cost of the non-revenue earning social services provided by such GCs should be weighed against budget priorities. Subsidies should not be used to pay the costs arising from poor corporate performance and the lack of adjustment of user charges; recent initiatives to privatize the trading operations of the NFA are a step in the right direction; (b) the requirement for GCs to remit dividends to the NG should be strictly implemented. After many years of depending on NG funding, GCs should share their earnings with the shareholders. The 10% mandatory dividend declaration rate (percentage to net income) should be increased to 1SZ, which is the average for firms iisted on the stock exchange; (c) GCs should be given more operational flexibility particularly with regard to the adjustment of user charges. The decline in the real level of user charges should be reversed; (d) the PEIS should continue to be refined. Trends in key capacity indicators, such as the frequency of power brownouts in the case of NPC and the degree of congestion at fare broths during rush hours in the case of the LRTA should be monitored closely; (e) systems to reward good performers and penalize poor performers should be pursued in line with Executive Order 486 which provides for the payment of bonuses and incentives to GC staff in accordance with performance; (f) ways to allow GC managers more autonomy with respect to hiring, compensation, and financing decisions should be studied; the objective should be to improve the flexibility with which GC managers can respond to changing economic conditions; (g) the principle of public accountability should be infused in the corporate plan formulation and performance evaluation process. Thle GCs concerned should be encouraged to conduct public hearings to discuss corporate plans and present annual reports so that interested and affected sectors can participate in the discussions and contribute to recommendations based on the corporate plans.  - 98 - BOX 3.2: FINANCIAL CRISIS AND REFORM OF THE NATIONAL POWER CORPORATION NPC'e Financial Problems. As In all countries with significant Inflation, the rate of roturn for public ut17tltis in Philippines Is based on Its rovaluod fixod ausets. Partly because of differences of vliws on th, valuation pproach and the preliminary results, the assets wore not revalued during 1987-90. This mad it difficlt to obtain an accurate assessment of NPC's financial performanco until mid-1991, when an aeset appralsal value was approved by NPC'a Board. Based on this appralsal, NPC's rat of return has ben bolow the 8K covenanted under the World Bank's Energy Sector and Bacon Manito Projocto (Loans 8168-PH and 2969-PH), and was OX In 1988, 6X In 1989 and 2.7X In 1990. This is much lower than the 7.9X roported In the Audit Reports for 1989- 90 and it prompted the Bank to review the situation. NPC's debt service ratio has b"en about 0.9 and 1.0 (compared with a minimum 1.3 covenanted). The deterioration In financial performance in 1990 resulted In a current ratio of below 1.0, a ngative return on Its net worth (down from 5X In 1989), or Its total assts (down from 1K In 1989) ond not losses of P66 million. Moreover, although NPC's Audit Reports show a positive Internal cash genoration, this was due to debt rescheduling, and during the laost four years more than 20OX of the capital expenditures was financod by loans. This was clearly unsustainable. NPC's financial crisis resulted from: (a) Insufficient tariff adjustments In previous years, whose financial Impact wao not obvious due to the dolays In revalulng the fixed asosts and Improper accounting of the cash flow; (b) the accumulation of debt service aft r 1991 (due to prevlous debt reacheduling); (c) a substantial increas- in oil prices In 1990-91; (d) lower Increases In power sales (3-3.6X In 1990-91 compared with 5K in 1989); and (o) a 25 percent devaluation of the peso vie-a-vis the dollar In 1990. Tariff Probl-ms In 1991. A tariff Increase of PO.17/KWh was approved by NPC$s Board for Implementation from March 1991, but after 8 months of appeals to the Supreme Court (because of complaints that NPC's rate of return exceeded 10X) and lengthy public hearings by the Energy Regulatory Board, no decision has been taken to approve or deny the tariff Increase. Because of this delay, tho Increase In powor tariffs would have coincided with a 34X reduction In oil prices, that although already factored in the requested tariff was difficult to explain to the public at that time. On top of financial problom,, there are roported Irregularities and Inefficiencies within NPC that damages Its credibility and Its ability to bring about the tariff Increases. Under these conditions the tariff review became more political than technical. Slnce tariffs wore not Increased In 1991, all of NPC*s financlol indicators worsened, including its rate of return (1X), Its *slf-financing ratio (-34K), and Its debt service ratio (0.5). NPC would end the year with a negative not income of almost P4,000 million. To allow NPC to pay its operating costs and debt service, the Government had to provide P7 billion (about f270 million) of financing to NPC In May 199l. NPC s Reform Pro ram. A comprehensive Reform Program has now been agreed with the Government, which Incluaos (a) corporate improvements (an Immedlate Program and a Managoment Audit); (b) promotion of private sector participation In electricity generation and partial privatizatlon of NPC; and (c) several financial actions, which would be satisfactory. Particularly critical is the proposed tariff increase that needs to be Implemented In January 1992 (about PO.24/kWh), which *hould: (a) ensure at least a OX rate of return on revalued assets In 1992; and (b) provide a stisfactory cash position to NPC, without the need for further Governmont contributions.  - 99 - SELECTED BIBLIOGRAPHY Aguirre, C. et. al.: A VAT for the Philippines, International Monetary Fund, Fiscal Affairs Department, 1987. Aechauer, D.: Public Investment and Productivitv Growth in the Group of Seven, Economic Perspectives, Federal Reserve Bank of Chicago, 1989. Avila, A.: Measurement of Tax Evasion, National Tax Research Centre Staff Papers, 1989. Barro, R.: Economic Growth in a Cross-Section of Countries, Working Paper No. 3120, National Bureau of Economic Growth, Cambridge, Mass. 1989 Bureau of Internal Revenue: Annual Reports, various years. CowiConsult: Re-assement Study of EPZs and Industrial Estates, Report prepared for Dept. of Trade and Industry, Republic of the Philippines, 1991. Centre for Research and Communications Selected Papers on Tax Reform in the Philippines, 1990. De la Cruz, E. and G. Beltran: Issues in Government Corporation Performance, Mi:ieo, World Bank, 1991. Diamond J.: Government Expenditure and Growth, Finance and Development, Washington DC, December 1990. Lamberte, M.: Private Financial Savings and Direct Foreign Investment, Mimeo, World Bank, 1991. Lamberte, M., et.al.: External Financing, Structural Adjustment and Development, Mimeo, Philippine Institute for Development Studies, 1991. Manasan, R.: Tax Evasion in the Philippinea, Journal of Philippine Development, Volume XV, No. 2, 1988. Manasan, R.: An Assessment of Fiscal Policy in the Philippines, Working Paper Series No. 90-06, Philippine Institute for Development Studies, 1990. Muten, L. et.al.: The Tax and Customs System of the Philippines, International Monetary Fund, Fiscal Affairs Department, 1985. National Tax Research Centre: Annual Reporta, various years. National Tax Reeearch Centre: A Study on Tax Administration and Compliance in the Philippines, November 1986. National Tax Research Centre: Assessment of Tax Reforms, 1986-89, 1989.  - 100 - National Tax Research Centre: An Aeessment of the Value Added Tax SYstem of the Philippines, October 1990, Societe Generale de Survellancet Comprehensive Import SuPervision Service for the Philippines, Manila, 1990. World Bank: Leasons of Tax Reform, Board Paper, 1991. Yoingco, A.: Taxation in the PhilippineB: Foretellin2 of Things to Come# Asian-Pacifio Tax Investment Research Centre, Bulletin, February 1991.  - 101 - STATISTICAL APPENDIX List of Tables Page No. 1. Savings and Growth Rates . . . . . . . . . . . . . . . . . . 102 2. Public Sector Savings and Investments Rates . . . . . . . . 103 3. Selected Investment Data . . . . . . . . . . . . . . . . . 104 4. National Tax Effort: 1976-90 . . . . . .., . . . . . 105 5. Establishing PILs for Doctors . . . . . . . . . . . . . . . . 106 6. Internal Revenuet Accounts Receivable . . . . . . . . . . . 107 7. Workloade of Assessment and Collection Personnel . . . . . . 108 8. Budget of the Customs Police Administration . . . . . . . . 109 9. SGA: Value of Inspection Advice Notices . . . . . . . . . . . . 110 10. National Government Expenditures, by Economic Classitication . . 111 11. National Government Expenditures, by Sectoral Classification . . 112 12. Composition of Public Sector Investment (% Distribution) . . . . 113 13. Composition of Public Sector Investment (% to GNP) . . . . . . . 113 14. Public Sector Investment: (% Distribution) .. . . . . . 114 15. Public Sector Investment (% to GNP) . . . . . . . . . . 114 16. Net Flow of Funds between NG and GCs, 1976-90 . . . . . . . . . 115 17. NG flows to GCs by Categories, 1976-90 . . . . . . . . . . . . . 115 18. The Major Subsidy Recipie.ts .................. . 116 19. The Major Equity Recipients .... . .- ... . 117 20. Operating Ratios of Major Non-Financial GCs . . . . . . . . . . 118 21. Net Income Major Non-Financial GCs . . .. .... .... 118 22. Return on Assets Selected GCe s.. .119 23. National Government Revenues by Source . . . . . . . . . . . . . 120  - 102 - STATISTICAL APPENDIX Table 1s Savings and Growth Rates 1/ Real GDP National Personal Foreign Year Growth Savings Rate Savings Rate Savings Rate 1970 4.7 20.9 6.9 0.3 1971 3.6 20.8 7.2 0.1 1972 4.7 21.1 8.1 -0.5 1973 7.8 27.0 9.8 -5.5 1974 4.8 26.0 9.7 0.9 1975 7.3 24.2 8.4 5.3 1976 7.7 25.2 10.8 5.8 1977 6.0 25.5 10.5 3.3 1978 5.3 24.6 8.4 4.3 1979 6.1 26.6 6.5 4.5 1980 7.9 25.7 6.0 5.0 1981 1.0 25.5 6.5 5.1 1982 2.9 21.0 3.2 7.3 1983 0.9 19.7 0.9 7.0 1984 -6.1 14.3 0.9 2.7 1985 -4.4 15.0 1.1 -1.1 1986 1.4 18.2 2.4 -5.3 1987 4.5 16.0 0.1 0.1 1988 6.2 18.0 6.0 0.7 1989 5.4 17.5 6.3 1.1 1990 3.3 13.0 2.3 5.2 Source: National Income Accounts, 1972 base. 1/ Refers to ratios to GDP  - 103 - Table 2: Public Sector Savings and Investmonts Rates 1/ Savings Investmento NG PE NFPE 3/ NG 2/ PS NFPE 1975 1.9 0.9 0.2 3.1 4.3 2.0 1976 1.? 1.0 0.2 3.4 12.8 7.7 1977 1.5 1.1 0.2 3.3 7.2 5.0 1978 2.7 0.7 -0.1 4.0 5.3 4.2 1979 4.1 0.8 -0.1 4.2 7.3 4.6 1980 3.9 0.8 -0.8 5.1 4.7 4.6 1981 3.1 1.1 0.3 7.1 8.8 5.9 1982 2.2 1.2 0.5 6.4 5.9 4.3 1983 2.9 1.1 0.4 4.8 5.2 3.2 1984 2.6 0.3 0.7 4.4 5.9 3.5 1985 2.3 -3.3 0.7 4.1 2.7 i.1 1986 2.0 -3.6 0.1 7.1 1.5 1.0 1987 0.5 -3.1 0.9 3.4 -0.6 1.2 1988 -0.1 0.2 1.0 2.7 1.1 1.1 1989 1.0 -0.5 0.0 3.1 1.9 1.8 Source: Adapted from Lamberte, et al., (1991), Table 4.5.1. 1/ NG - National Government; PE - Public Enterpr'