Report No. 14680-PH Philippines Public Expenditure Management for Sustained and Equitable Growth (In Two Volumes) Volumne I September 5,1995 Country Operations Division Country Department I East Asia & Pacific Region O-4JWari. - 4 Z 4? %. T'ss I~ - z~~~ This report has been prepared by Hisan Shishido (task manager). Major contributors were:, Douglas Hartle, Jeffrey Hammer, David Steedman, Stephen Howes, John Arnold, Antoine Schwartz, Syed Husain, Rosario Manasan, Pierre Markowski and Domenico Fanizza. Anjum Altaf, Shaikh Hossain, Renee Santiago, Rencie Padernal, Benjamin Diokno, and Soniya Carvalho also contributed. Nam Pham prepared statistical appen- dix. Peer reviewers were Homi Kharas and Malcolm Holmes. Deep appreciation goes to Bob Killoran who has patiently put this report together. Charu Vasil and Harry Travis provided editorial assistance. . PHILIPPINES Public Expenditures Management for Sustained and Equitable Growth Table of Contents Page No. Executive Summary ..........................................................i Introduction ......................................................... .I 1. A REVIEW OF FISCAI AGGREGATES ........................................................3 National Goverment .........................................................3 Consolidation of the Public Sector .........................................................9 Government Owned and Controlled Corporations ......................................................... 11 2. THE GOVERNMENT'S ROLE AND RESOURCE ALLOCATION ISSUES ........................................... 16 Introduction ..................................................... 16 Determinants and Effects of Aggregate Expenditure Pattems ............................... ...................... 16 Efficiency and the Role of the Public Sector in the Economy .................................. ................... 18 Equity in Public Expenditure ..................................................... 21 Omissions ..................................................... 25 Policy Recommendations on Sector Expenditure ..................................................... 26 TRANSPORT SECTOR ................................................. 26 POWER SECTOR ................................................. 28 WATER SUPPLY ................................................. 31 AGRICULTURE AND NATURAL RESOURCES ................ ................................. 32 EDUCATION ................................................. 35 HEALTH AND NUTRllION ................................................. 38 3. DEVOLUTION AND LOCAL GOVERNMENTS ................................................... 41 Background and Rationale ................................................... 41 Issues - Unintended and Intended Consequences of Devolution ................................................... 43 Principles for Improvements ................................................... 51 Policy Direction and Recommendations ................................................... 52 4. CIVIL SERVICE REFORM ................................................... 55 Introduction .................................................... 55 Background Analysis ................................................... 55 (a) Size, Salaries and Cost .................................................... 55 (b) Management Issues ................................................... 57 (c) Reform Initiatives .................................................... 59 Rationale and Objectives ................................................ 60 (a) Rationale ................................................ 60 (b) Objectives ................................................ 63 Strategic Issues ................................................ 63 (a) Political Commitment and Institutional Arrangements ................................................ 64 (b) Streamlining .................................................... 64 (c) Salary Strategies ................................................... 65 (d) Ways of Changing the Civil Service Size and Skills Mix ................................................... 66 (e) Personnel Management .................................................... 68 S. BUDGET PLANNING AND MANAGEMENT ................................................... 69 The Formal Budgetary Cycle .................................................... 69 Recent Budgetary Initiatives: Palliative or Cure? ................................................... 73 Issues .................................................... 74 Policy Recommendations ................................................... 75 ANNEX TO CHAPTER 5 RECOGNITION AND MANAGEMENT OF CONTINGENT AND OTHER LIABILITIES ........................ 82 Introduction ................... ................................................ 82 Definitions ............................................................... 83 Management and Control of Contingent Liabilities .............................................................. 83 (a) Objectives ......................................................... 83 (b) Budget and Expenditure Management Process . . ..................................................... 84 (c) Valuation ......................................................... 85 (d) Tuning of Expenditure Recognition .............................................................. 86 (e) Minimizing Losses .............................................................. . 86 Management and Control of Actuarially Determined Liabilities . ............................................................. 87 (a) Objectives ............................................................... 87 (b) Budget and Expenditure Management Process .............................................................. 87 Accountability ............................................................... 88 Conclusion ............................................................... 88 ATTACHMENT (Definitions) ............................................................... 90 LIST OF TABLES. CHARTS AND BOXES CHAPTER l TABLES 1-I Fiscal Aggregates 1980-94 (Percent of GNP) ........................ .........................................4 1-2 Trends in Public Expenditure in Asian Countries (As a Percentage of GDP) ............... .....................4 1-3 National Govermment O&M Spending by Selected Sector (% of GNP) .................... ........................7 1-4 Budgetary Inflexibilities -Mandated Expenditures (Percent of GNP) ...................... .......................8 1-5 Programmed and Actual Revenues and NG Investment (Billion) ...................................... ................9 1-6 Public Investment By Sector (% of GNP) ................................................................. 10 1-7 Consolidated Public Sector Deficit (billion peso, and % of GNP) ............................11......................... 1 1-8 Total Internal Cash Generation of Major GOCCs (In Million) ............................................. .............. 12 1-9 Net Income of Major GOCCs (Million) ..................................................................... 12 1-10 GOCC - National Government Financial Flows (Cash Basis) ......................................................... 14 1-11 NPC: Net Financial Flow with the National Government (In Million) . . 16 1-12 National Government Accounts - Medium Term Projections (Percent of GNP) (Including CB-BOL) ......................................................................... 18 CHARTS 1-1 Public Expenditures 1976-1994 ..........................................................................5 1-2 Mandated and Productive Expenditures ......................... ................................................7 CHAPTER 2 I&DLFs 2-1 Determiinants of Total Expenditure by Functional Category 1975-1993 ............................................ 17 CHARTS 2-1 Distribution of Taxes and Expenditure by Regional Income .............................................................. 22 2-2a Potential Incidence of Public Housing Program ......................................................................... 24 2-2b Substandard Housing by Regional Income ........................................................................ 24 CHAPTER3 TAfLE 3-1 Net Transfers to LGUs (Nominal million pesos) ........................................................................ 45 3-2 Consolidated Expenditures of National and Local Governments: Selected Items (% of GNP) ......... 45 3-3 Sectoral Allocation of National and Local Government Expenditures (% of GNP) .......................... 46 3-4 Expenditures by Sector, All Local Government Units: 1991, 1993 (% of GNP) ............................... 48 3-5 Collection Rates of Real Prperty Tax (percent) ........................................................................ 49 3-6 Total, Social and Health Expenditures Per Capita by LGUs 1991 and 1993, in nominal pesos ......... 50 BOXES 1 Civil Service Reform in Individual Agencies: the Case of the Central Bank ..................... ................... 59 2 An Audit of DECS ......................................................... 61 3 Competent Bureaucracies and Rapid Economic Growth ............................... .......................... 62 CURRZENCY EQ-UlIVALENr (as of June 1995) Currency Unit = Peso () US$1 = 25.58 1 = US$0.039 FTISCAI, YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS AA - Advice of Allotment DA - Department of Agriculture ABM - Agency Budget Matrix DAR - Department of Agrarian Reform ADB - Asian Development Bank DBM - Department of Budget and Management ARB - Agrarian Reform Beneficiary DECS - Department of Education, Culture and Sports ARC - Agrarian Reform Communities DILG - Department of Interior and Local ARF - Agrarian Reform Fund Government ARO - Allotment Release Order DOE - Department of Energy ATO - Air Transport Office DOF - Department of Finance DOH - Department of Health BESF - Budget of Expenditures and Sources of DOST-PCIERD - Department of Science and Financing Technology BLT - Build, Lease and Transfer DOTC - Department of Transport and BOI - Bureau of Investments Communication BOT - Build-Operate-Transfer DPWH - Department of Public Works and Highways BSP - Bangko Sentral ng Pilipinas (Central Bank) DSWD - Department of Social Welfare and BWSA - Barangay Water Supply Association Development CAG - Corporate Affairs Group (in Department of EMK - Effective Maintenance Kilometer Finance) EO - Executive Order CALABARZON - Cavite, Laguna, Batangas, Rizal, EPZ - Export Processing Zones Quezon ERB - Energy Regulatory Board CAR - Cordillera Autonomous Region ESC - Education Service Contract CARL - Comprehensive Agrarian Reform Law CARP - Comprehensive Agrarian Reform Program FAP - Foreign Assistecl Projects CB-BOL - Central Bank - Bureau of Liquidation FAPE - Fund For Assistance to Private Education CCPAP - Coordinating Council of the Philippine Assistance Program GAA - General Appropriations Act COA - Commission On Audit GARO - General Allotment Release Order CS - Civil Service GASTPE - Government Assistance to Students and CSC - Civil Service Commission Teachers in Private Education CSPD - Consolidated Public Sector Deficit GDP - Gross Domestic Product GFI - Government Financial Institutions GNP - Gross National Product NEDA - National Economic and Development GOCC - Government Owned and Controlled Authority Corporations NFA - National Food Authority GOP - Government of the Philippines NG - National Government GPEP - Grains Productivity Enhancement Program NHA - National Housing Authority GSIS - Government Service Insurance System NIA - National Irrigation Administration NIS - National Irrigation System HYV - High Yielding Variety NMYC - National Manpower and Youth Council NPC - National Power Corporation IA - Irrigators Association NWRB - National Water Resource Board IADP - Integrated Area Development Project ICG - Internal Cash Generation O&M - Operation and Maintenance IPP - Independent Power Producers ODA - Official Development Assistance IRA - Internal Revenue Allotment OECF - Overseas Economic Cooperation Fund ISF - Irrigation Service Fee OPSF - Oil Price Stabilization Fund KPA - Key Production Areas PAL - Philippine Air Lines PCGG - Presidential Commission For Good LGU - Local Government Units Government LPG - Liquefied Petroleum Gas PD - Presidential Decree LRTA - Light Rail Transit Authority PDO - Port District Offices LWUA - Local Water Utilities Administration PMO - Port Management Offices PNOC - Philippine National Oil Company MIAA - Manila International Airport Authority PNR - Philippine National Railways MICT - Manila International Container Terminal PPA - Philippine Port Authority MMTC - Metro Manila Transit Corporation MOOE - Maintenance and Other Operating RA - Republic Act Expenditures MTADP - Medium Term Agricultural Development SARO - Special Allotment Release Order Plan SEF - Special Education Fund MTPIP - Medium Term Public Investment Program SSL2 - Salary Standardization Law II MW - Megawatts SSS - Social Security Systems MWSS - Metropolitan Waterworks and Sewerage SUC - State Universities and Colleges System SWIP - Small Water Impounding Project NAIA - Ninoy Aquino International Airport TFS - Tuition Fee Supplements NALGU - National Assistance to Local Government NCA - Notice of Cash Allocation VAT - Value Added Tax NCEE - National College Entrance Examination NCR - National Capital Region WD - Water District NEA - National Electrification Administration WFP - Work and Financial Plan NEAT - National Educational Achievement Test E X E C U T I V E S U M M A R Y PHILIPPINES Public Expenditure Management for Sustained and Equitable Growth This report discusses how Fiscal management remains The findings of this report public expenditure management potentially precarious. Raising outline a framework for managing can be further improved in the additional tax revenues has public expenditures in the Philip- Philippines, focusing on four become increasingly difficult pines to achieve sustained and areas: whereas past under-spending in equitable growth. While recom- (i) the role of the public sector in infrastructure and social sectors mendations are discussed in more the environment of the has enormously increased the detail below, this framework can expanding private sector; expenditure needs in these sectors. be briefly summarized: (ii) fiscal devolution and roles of Fiscal balance was restored in 1994 * Rationalize expenditures that the national and local govern- mostly because of higher have not sufficiently achieved ments; privatization proceeds, lower objectives. Examples include domestic interest NFA consumer subsidies and a (iii) building a rates and lower civil service Thiere are thus urgenit needs to public investment good portion of spending on ffiat can than anticipated. ~~~~~~state universities and colleges. that can einhaon)ce efficien)cy of fiscol than anticipated. Social safety nets need better maginage nioi(igeftieiit by pTioeitiznng pigh targeting. Tax expenditures, changing maaeetblpirtzn ar, the high which amount to some 30 Govern- expenditures anid inmproving privatization billion, need to be reviewed ment role; budgetory maniugermientt proceeds, are based on clear priorities. and Udifficult to sustain. and) better There are thus # Improve budget procedures for (iv) better There are thus better fiscal forecasting and budgetary management that urgent needs to enhance efficiency control. It is vital that a identifies and ensures funding of fiscal management by prioritiz- medium-term fiscal framework of priority activities, and ing expenditures and improving with improved projections of incorporates contingent and budgetary management. The very fiscal aggregates be used as a actuarial liabilities within the nature of national government basis for budget planning, and fiscal framework. expenditure patterns is changing, that agreement with Congress TIhe Government of the Philip- however, due to the evolving role be reached on this framework pines has been successfully of the public sector as the private so that agreed priorities can be implementing liberalization sector role expands; and devolu- implemented. It is also critical measures and increasing the role tion which is changing the relative that the cost of contingent of the private sector in the share of responsibilities between liabilities, particularly for economy. These efforts have, by the national and local govern- government guarantees, be 1994, contributed to regaining ments. Improving public expendi- incorporated. macroeconomic balance and ture efficiency also requires restoring income growth based on examining who manages the * Strengthen fiscal devolution private investments and exports. public sector (the quality of the t rough increasing autonomy Sustaining robust economic civil service) and how public gnac ntability foreoa growth rates and addressing the resources are managed (budgetary governments. Local govern- country's development needs, processes) in addition to where ments need full autonomy In however, will require significant resources go. need to be made accountable public as well as private invest- and face hard budget con- ment. Pop I E X E C U T I V E S U M M A R Y straints. Local government private sector competitively. This them. This distinction can be transfers can be improved by has been made possible through subtle in the social sectors. Public taking into account equity technical development and provision of curative health considerations and providing decomposing what used to be (which can be provided by the incentives to implement considered a natural monopoly private sector) increases the projects that have high exter- into a number of contestable overall services in poorer regions nalities (particularly in the activities. The most notable but not in the richer regions. In social or environmental areas). example is that the power sector the latter, public provision merely Much needs to be done to has been unbundled into genera- crowds out private services. A strengthen local government tion, transmission, and distribu- careful review is needed not only capacity to carry out these tion. The Build-Operate-Transfer what services should be provided roles. (BOT) arrangements with private by the Government but also where * Focus on improving civil power generators were instrumen- these services are provided for service quality and efficiency. tal in overcoming the recent equity reasons. Similarly, the As the role of the government power crisis. Other examples Government efforts to achieve changes, needs for the civil include toll roads, telecommunica- universal secondary education by service change: from direct tion and water plants. Further- subsidizing those with no public provider of services to facilita- more, part of government services school access to go to private tor for efficient private activi- such as road maintenance and schools has reduced the supply of ties. It is important that refuse collection can be contracted true private schools, and created a government be staffed by an out to the private sector. class of hybrid schools which call able and professional cadre The Government, however, is themselves private but are depen- that can carry out this role. aware that the private participa- dent on the government subsidies. Civil service reform that tion is not a panacea for the These subsidies have also diverted improves inter alia incentives government's fiscal problem in all resources from important primary for higher level and profes- sectors. The private sector schools. sional staff - improving overall "cherry picks" - investing prima- RElArlIVE ROLES OF efficiency - is an important rily where returns are high and NATIONAL AND l.OCAl plank of overall fiscal reform. the pay-back period is short. They (G\VERNMEN rIS maximize the financial return THE ROLE OF TH E rather than the return to the In 1992, the National Govern- GOVERNMENT society in general. This may ment started the process of The public sector still has a increase the cost of services devolving major fiscal responsi- significant role to play despite the significantly. The Government's bilities for health, agricultural expanding role of the private regulatory oversight is thus extension, forest and watershed sector into areas previously important especially where management, rural infrastructure considered the monopoly of the private participants face limited and some welfare programs to public sector. The simple ground (or no) competition; regulation of local government units (LGUs), rule is that the Government needs prices or service levels is required while increasing resource trans- to intervene either as a service It is also the responsibility of the fers. The objectives include: to provider or regulator when an government that the privately improve local service delivery, and efficient market solution is un- provided portion of the service to reduce perception of political likely to exist if left to the private network (transport, power trans- alienation outside the Metro sector or when the equity impact mission, water supply) should be Manila through increased local is significant. effectively integrated in the autonomy. Some services that used to be overall network. It is too early to judge the true provided only by the public sector Finally, public service provision consequences of devolution as can now be provided by the needs to complement private data are limited, but a number of activities and not to substitute for Pup2 E X E C U T I V E S U M M A R Y observations can still be made. * make LGUs accountable and areas, by allowing cost First the allocation of transfers to harden the budget con- sharing with the national individual LGUs is not based on straints within which they government. the LGUs' fiscal needs, fiscal must operate by removing CIVIll SERVICE , REFORM effort, fiscal capacity or the uncertainties (both amount amount of devolved responsibili- and timing) on transfers to The ongoing evolution of the ties. Thus there are winners and them, stopping NG funding of public sector due to losers in terms of financial re- devolved activities, monitor- privatization and devolution sources, without significant ing LGU borrowing, and provides a unique opportunity correlation to economic or social resisting pressures for national for carefully redefining its roles attributes of individual LGUs. government bail-out of LGUs and priorities, and accordingly Second because there is no in debt service difficulties. identifying ways to improve incentive for LGUs to raise more * strengthen LGU capacity for public sector management. revenues or implement activities fiscal planning and imple- The Government has seized importarnt for the nation as a mentation especially to ensure this opportunity and started a whole, local revenue efforts are that LGU budget systems are c ivil service reform process weakening and problems are coordinated with national called streamlining the bureau- occurring in implementing systems; there is a clear role g The goal of this exercise projects (both foreign assisted and for the national overnment in to improve efficiency of the locally funded) whose benefits forovidingtechnical assistance central bureaucracy which, by spill over to more than one here f Goversrnent's own admission, jurisdiction. Third LGUs are suffers from "structural and spending larger amounts in the . develop a good statistical systemic dysfunction". social sectors, but spending on base for deriving the amount Objectives. dhere are five key economic services (irrigation, of transfers to each LGU, objectives for civil service roads, etc.) appears to be suffer- especially in the area of reobjciem frcvl-evc ing. The National Government's population. willingness to continue to fund While the current system of a) redefine the role of the public these devolved functions seems to income transfers is working - and sector and the civil service be a crucial reason for the LGUs' there would be some political followed by the restructuring reluctance to allocate their own difficulties in adjusting it - the and reorganization of depart- resources. Fourth. because Government may at some time ments and agencies; devolution makes fiscal monitor- wish to modify it. Two potential b) identify and reconfigure as ing of the consolidated public improvements are highlighted required the civil service skills sector more difficult, the LGUs here: mix; need to confront strict budget * incorporate concerns about c) improve incentives to attract constraints so that the overall equity and local resource and maintain desired skills; budget discipline be kept. Finally. mobilization in the allocation d) establish controls over the size an early indication is that the mechanism of resource and cost of the public service; devolution may be increasing transfer by considering an and inequity; this needs to be moni- equalization grant based on tored and steps taken to deal with the tax base and revenue e) develop an integrated system the problems. efforts. for buanagng personnel budgets, pOsitions and Recomumendations. Several . provide incentives for fund- individuals. steps could be taken to improve ing critical activities for a the functioning of the existing small number of selected In order to implement effective system: civil service reform, the Govern- ment needs to design an inte- grated reform strategy that incorporates all five objectives Pop9 3 E X E C U T I V E S U M M A R Y outlined above. The present costing of the various options, ing recruitment and wage reform effort falls short in estab- would help achieve and structures. lishing the desired skill profile, maintain the desired skills BUDGEIARY PROCESSES improving incentives and install- mix, which will probably ing a system of personnel manage- include a higher proportion of The budget processes are at least ment. These aspects are crucial, professional staff. Such a partly responsible for the inad- because the civil service is increas- reform would allow higher- equate coordination between ingly expected to be a catalyst for level civil servants to be development planning and project an economic transformation, and remunerated at rates more and program implementation; for the composition and quality of the compatible with their private the selection of some low priority future civil service should be sector counterparts. At lower projects over higher priority suited to this new role. levels, salaries are relatively projects; for inordinate delays in Recommendations To achieve competitive with the private implementing high priority the objectives discussed above, the sector. A move away from the projects; and for a bias against reform strategy must integrate the existing policy of concentrat- adequate maintenance of existing needs for: clear definition of ing wage increases at lower facilities with a consequent public sector roles; identification levels would be highly decline in public services. of the skills mix and organiza- desirable. Issues There needs to be an tional structure required for c) A combination of volunta overall agreement on the fiscal carrying out the defined roles; retirement and targeted framework and national priorities transparent and fair implementa- retrenchment of surplus staff among members of the Congress, tion of restructuring and reorga- would complement the reform central departments of the Execu- nizing; appropriate incentives to of the incentive system in tive, and line departments be- achieve the desired skills mix; and achieving the desired skills cause, especially in a democracy, a reform in personnel manage- mix. This approach would be the budget process entails the ment with a view to reinforcing far superior to reliance only reconciliation and compromises of and sustaining these improve- on natural attrition. a multitude of conflicting inter- ments. While the Government is d) Personnel managemen ests. The central departments making efforts to form this reform, based on a review of (DBM, NEDA, DOF) usually take strategy, the following actions the current allocation of a lead in preparing the budget would contribute to the success of management tasks among proposal. The Congress then its implementation: mentral agencies, would imposes different priorities by a) Strong political commitment significantly contribute to the adding activities of its own choice from the highest level of improved functioning of the (often in the areas of local roads leadership is required, and the civil service. Management and irrigation, and state colleges Office of the President needs would also benefit from an and universities). The Congress to take an increased lead in improved common human has to delete some other activities this process. The existing resources data base and a because by constitution it cannot Presidential Committee system of objective perfor- increase the peso amount of the responsible for the exercise mance evaluation. budget. The congressional and its Technical Secretariat e) In the restructuring the public deletion often includes foreign may be reinforced by the sector organization, it is assisted projects. addition of representatives important that the Govem- The Executive (DBM) often from CSC and advised by ment provide GOCCs and counters Congressional amend- outside experts as necessary. certain regulatory agencies ments by rationing cash in such a b) A reform of the incentive with more autonomy regard- way as to favor the activities to system. based on a detailed which it accords highest priority. Pop 4 E X E C U T I V E S U M M A R Y Revenue shortfalls relative to the are under no incentive to critically economic work should be estimates used in the budget are a assess their own proposals as carried out as objectively and good reason to limit cash releases, NEDA will eventually assess reliably as the state of our but the threat of shortfalls (and them, but NEDA does not have understanding permits. not the actual shortfalls) is enough enough human resources for (b) The cost of contingent for the Executive to introduce and assessing all of a variety of liabilities such as those maintain cash rationing during a projects in depth, in a short time. associated with the BOT-type fiscal year. Furthermore, con- Implementation delays frequently agreements needs to be taken flicts can often exist between the occur because the decision to into account as a real cost to priorities assigned to projects by proceed is often little more than a the budget when these the central agencies and the decision to begin the planning that liabilities are created (Annex priorities favored by the line should have taken place earlier. to Chapter 5). departments. Because cash Recommendations A funda- (c) Based on the Fiscal Frame- releases are in essence fungible mental measure to ameliorate work, responsible central between projects within a same these issues is to shift the conflict departments carry out devel- department/agency, the line resolution on the national priority opment programming. This agencies could also alter the from the "behind closed doors" activity will determine the spending priority thus subverting forum where cash rationing takes allocation of funds among the prionties established by the place, to Congressional debates major functional categories of President about the appropriate size of the spending assisted by propos- Cash rationing also has the public sector and the appropriate als from line departments and effect of delaying implementation allocation of spending among the in accordance with the agreed of high priority projects because broad functional areas of national priority. line departments face great interest. This would shift the uncertainty concerning the flow of emphasis away from the limited (d) The Fiscal Framework and funds that will be forthcoming. flow of funds to government in Development Progranmming Given the uncertainty, some the short-term towards the should be the subject of a loint agencies try to safeguard their medium term economic ouflook Resolution of the Congress at month by month cash allocations and the policy options available to the beginnng of tde term of for the payment of wages and government. In so far as the every Congress and voted salaries first before they start to "procedures" are concerned, prior to each tabling of the spend on more discretionary however, the medium-term fiscal President's Expenditure budget items. framework recommended below Budget. Congress would be Finally, there are technical may appear similar to the existing the two documents, after problems in project preparation- five-year development planning docume after project appraisals are often framework. The key difference is appropriate modification. jeopardized due to the insufficient how such procedures are imple- made by Congress must be trained staff in line departments mented. financed by reductions in whose time is also demanded for (a) the National Government appropriations within the keeping track of appropriations adopts a three-year Fiscal same functional category. It and allotments. They need to Framework, which incorpo- should be understood that prepare work plans in a hurry for rates annual (quarterly for the congressional and other projects added by the Congress. first year) projections of: fiscal changes that are inconsistent Allocation of responsibility for revenues based on existing with the approved Framework project planning within the taxes, expenditures (current and the Program will be Executive also poses problems. and capital), deficit, net loans vetoed. Departmental subver- The line departrnents, who and advances and net cash sion of the approved frame- originate these project proposals, requirements. Underlying Pp gs E X 1 C U T I V E S U M M A R Y work should also be disci- binding on development efficiently need to be identified, plined. programming. then privatized. Light rail transit, (e) Once the Fiscal Framework . toll roads, some functions at ports, and Development Program- (g) There is a need for a comput- and airport management are ming are in place and ap- erized accounting system among them. The National proved, a department in defined in a manual, and 'Railway can also be considered for charge of Budget Operations operating departments must privatization after its network is Management (DBM), prepares be staffed with qualified shrunk and financial losses budget and implement it accountants capable of reduced. efficiently. Continued use of The same system should be Depending on the private sector the baseline budgeting is pre sied for lola res for transport service provision recommended. It should be pesrmi on lol dauton'of needs to be accompanied by the noted that with the appropri- to introduction of more efficient data-expenditures should at ate Fiscal Framework in place, least be classified as proposed, pncing systems-involving the there should be far less need appropriated authorized, reduction of cross-subsidies- to racon cash than earliert and committedad ual.The especially in the port and railway thus complicated methods of committed and actual. The sub-sectors, and increased regula- allotment and cash releases computer system must be tion of safety and environmental can be largely abolished, robust with much backup and protechon. The Government except in large capital projects accessible to poorly trained needs to develop a regulatory where cash releases should be personnel. agency that is not involved in commensurate with the SECTOR RESOURCE provision of services. Finally, progress of the project prepa- ALLOCATION ISSUES appropriate staff down-sizing at ration/implementation. In In the transport sector, the the center needs to be made as addition, the department National Government is consider- responsibilities are devolved to should: (i) establish rules for ing reducing its network to an local governments and the private procurement and personnel essential core that can be main- sector. management (with the Civil tained to a high standard. Eirst, In the power sector, reforms are Service Commission); and (ii) small ports, airports and local already going on to "unbundle" conduct project monitoring to roads, need to be devolved to local the sector into generation, trans- evaluate the operational governments who would also mnission and distribution. perform-tance of other depart- review what is sustainable. Most Privatization of thermnal genera- ments. Management audits small roads are already under tion has started recently with the should also be conducted by local jurisdiction, but the National power crisis, and is expected to an appropriate agency. Government is appropriately continue albeit at a slower rate. (f) Organizationally, those who limiting further the national NPC is being spun off to subsid- establish the Fiscal Frame- highway that it would maintain to iaries based on regions or func- work and Development about 60 percent of the existing tions with a view to be sold to Programming should work network. Road maintenance private interests. Distribution has very closely together-they should be carried out by private been largely private, but except for could even be located in one contractors rather than through MERALCO, the sector comprise agency But should these two the force accounts. The National more than 130 small and finan- functions be divided into two Railway, which suffers from large cially weak distributors. The agencies, it needs to be made financial losses, also needs to latter need to be consolidated into clear that the financial con- reduce its size to a commuter rail about 15 large units that can share straints, which are dominated system around Manila. Second market risks with the generators. by macroeconomic consider- transport responsibilities that the For the time being, the high ations, were absolutely private sector could provide more voltage transmission and hydro- Pop i E X E C LJ T l V [ S U M M A R Y electric generation are recom- mance-dependent competitive of the poor to quality education. mended to remain largely in the allocation of resources. Cost-effectiveness also needs to be public sector. The public sector Agriculture has suffered from improved. By doing so, the pubiic also needs to focus its attention on slow growth recently, with serious sector spending would better improving the regulatory frame- implications for the already high, complement the private contribu- work, implementing fair and rural poverty incidence. A tion. transparent load dispatch system, coordinated approach to address It is recommended that, first, and reforming pricing so that it this issue has been difficult measures to improve resource incorporates time-of-day demand because major sector agencies availability in the sector be changes and appropriate stand-by have employed different develop- introduced, and, second, the and wheeling charges. ment strategies. Poor quality rural resources use need to be more cost In the water sector, the objective infrastructure and feeble support effective and targeted. The should be to rely increasingly on services have been serious impedi- measures to improve resource cost reduction, demand manage- ments to rural growth; it is critical availability include: school ment and increased efficiency in to boost rural productivity by construction and maintenance can production and distribution. A improving rural transport and be devolved to local governments; strategy that can introduce irrigation facilities, and agricul- State Universities and Colleges incentives consistent with such an tural research and extension. should depend on full cost objective is the commercialization Since many of these functions recovery and decentralized and private sector participation have been devolved to local financial decisions; and budget (PSP) initiative. Towards this end, governments, strengthening their management at the DECS be the transitional steps are (i) rapid implementation capacity is of high rationalized so that resources are introduction of PSP alternatives in priority. Improving infrastructure, ensured to reach schools rather the large metropolitan centers together with removing uncertain- than lost at administration (as a (Metro Manila, Metro Cebu, ties associated with the implemen- large segment of non-personnel Davao - the government is already tation of the Agrarian Reform, recurrent resources are). To acting on this initiative); (ii) would promote private rural enhance cost effectiveness and regrouping of smaller municipali- investments. Finally, protection of improve quality and equity of the ties into regional service areas natural resources, especially sector services: the Government large enough to yield necessary improved watershed manage- needs to move away from propor- economies of scale in manage- ment, is a priority for sustainable tionality (equal per student ment, lower transaction costs and growth. expenditures in every province) to attract private sector participation; Regarding the social sectors, the targeting resources towards the (iii) setting up an independent public sector spends 2.7 percent of most disadvantaged areas to regulatory body to undertake GNP on education, a sum that address the priority issues of financial and other contract could be used more cost-effec- equity. The resources spent on related regulation of private tively. The country as a whole achieving universal secondary service operators; and (iv) in the spends a decent 6-7 percent of education by giving grants to near future, making efforts to GNP, indicating large private students to go to private schools is reduce excessively high non- presence in the sector as well as not reaching the poor; the rich use revenue water in Metro Manila willingness to contribute to the resources to go to better through increased maintenance education at the household level private schools for which they and possibly a major network (in part due to the inadequate would have paid anyway and the rehabilitation. For rural areas, a funding by the public sector). system provides perverse incen- program of sequenced, demand- The effectiveness of the public tives for creating low-quality driven infrastructure investments sector contrbution should be private schools dependent largelv are needed coordinated by the improved by focusing on quality on public subsidies. Rather than public sector and based on perfor- of basic education and the access attempting to make this grant available to the general public, - - - -. . ........... . - -------- -------- - ----- E X E C U T I V E S U M M A R Y again, it should be targeted to the poor and the per-student spend- ing be increased. The health sector has been devolved and it is too early to make assessment on how spend- ing has been changed-the sector thus is not reviewed here. It can be noted, however, that the health sector expenditures before the devolution was particularly progressive favoring poorer regions. With the devolution the allocation is estimated to have become more regressive. Three other issues can be noted here. First, the National Government still has a role to play in more preventive medicine nation-wide, given the extemalities involved. Second, public provision of curative health care, which is more a private good than preven- tive care, is still important in poorer regions while in richer regions such services merely crowd out the private clinics. Third, the nutrition subsidy through the general consumption subsidy of the National Food Authority is not targeted to the poor and is a very expensive nutrition intervention. It is recommended that the program be phased out and more targeted and cost effective nutrition intervention programs put in place. Pp I, PHILIPPINES: PUBLIC EXPENDITURE MANAGEMENT FOR SUSTAINED AND EQUITABLE GROWTH Introduction The Philippines has at last enjoyed accelerating economic growth and macroeconomic stability in the last two years. GNP grew by 5 percent and inflation was contained at 7 percent in 1994. Foreign capital inflows have increased severalfold reflecting strong investor confidence in the country's policy stability. The events in Mexico early this year did not significantly affect this confidence. The Ramos Administration's growth-oriented program focusing on macroeconomic stabilization and further structural liberalization has finally, it is hoped, placed the economy onto a sustainable growth path. The Administration's success in effectively solving the severe energy crisis of 1992 and 1993 has also boosted the image of the country. Macroeconomic gaps have also narrowed. The national government (NG) deficit was reduced from the average of 3.6 percent of GNP during 1986-88 to 1.4 percent in 1993 and then to a mere 0. 1 percent in 1994. Consolidated public sector deficit was the lowest ever recorded at 0.6 percent of GNP in 1994. In the same year, overall balance of payments recorded a surplus due to stronig exports and foreign inflows. To sustain this recent momentum, the Government of the Philippines (GOP) must still maintain investor confidence by tackling three issues: improve infrastructure services; address the high incidence of poverty; and still sustain sound fiscal management. This is a serious challenge. Significant investments are needed in physical infrastructure and human capital to off-set the past under-spending. Furthermore, high poverty incidence, which slows human capital development and often threatens political stability, needs to be addressed with increased vigor. Resource requirements in these areas are enormous. But, the Government faces significant opposition in raising additional tax revenues, and budget does not allow much flexibility as interest payments and wages are each running around 5-6 percent of GNP. Furthermore, recent fiscal improvements were partly aided by higher privatization proceeds and lower domestic interest rates than expected. These are events which are unlikely to recur. Tight budgets, difficulties in increasing revenues, and significant resource requirements for infrastructure strengthening, and poverty alleviation with human capital development all point to the urgent needs of the GOP to improve efficiency in allocating and managing its own expeniditures. This calls for prioritizing public expenditures so that only activities with high eco(nomic returns are funded (and implemented) and wasteful spending removed. In prioritizing public expenditures, a number of issues should be examined. First, it is crucial that the evolving role of the public sector vis-a-vis the private sector be carefully taken into account. As a result of ongoing liberalization, the private sector is expanding its activities into areas that used to be considered a public sector domain such as infrastructure and social services. With careful assessments of public sector roles, there will be opportunities to free up resources for more essential and so far under-funded services. Also, proper regulation will lower the cost to the economy of private provision of services in markets where competition is limited. Second, the -2 - GOP's etiorts to devolve major responsibilities to local government units (LGUs) need to be taken into account in determining the roles of the NG and LGUs. Third, the civil service needs to be strengthenied, with appropriate financial and other incentives, so that the Government with the emerging new roles and the priorities can be managed efficiently. Finally, the budget management process can be improved to ensure timely implementation of priority activities. Given this background, this report reviews the patterns and management of public expenditures, and makes recommendations for further improvemenus. The report is divided into two volumes. Volume I will focus on the expenditure management issues, specifically: * review the overall fiscal patterns in the recent past and provide a medium-term fiscal framework (Chapter 1); * discuss the role of the public sector in view of the expanding private activities and the need to correct market failures for efficiency enhancement and equity improvement (Chapter 11); * examine how devolution is changing the spending pattern of the National Government and local governmnent units, and discuss how management of inter-government relationship can be improved (Chapter 111); - discuss how the public sector management can be improved in terms of both human resources and financial management. Given the evolving role of the Government, we will first discuss the desirable reforms of the civil service so that the civil servants in the future will be well adopted to managing the new public sector (Chapter IV); and * discuss desirable changes in the way budgets are prepared and managed by the National Government (Chapter V). Chapter V also contains an annex that discusses how to manage contingent liabilities of the Government and how they should be incorporated in the overall fiscal framework. Priority recommendations for improving management of major sector expenditures (transport. water, power, agriculture and natural resources, education and health) appear in Chapter 11 in a summary form. Detailed discussions of sector issues appear in Volume 11. 1. A REVIEW OF FISCAL AGGREGATES 1. 1 This Chapter reviews aggregate fiscal figures for the National Government (NG) and the public enterprises called Government Owned and Controlled Corporation (GOCCs), to provide a framework for discussions for the chapters that follow. 1.2 The analysis highlights four issues. First, since mid-1980s, the budget has become increasingly inflexible because non-discretionary spending suppressed productive spending on investment and maintenance. Second, recent improvements in both NG and consolidated fiscal balances are due primarily to high privatization proceeds, which may not recur in the future. This was aided in 1994 by the low interest rates on treasury bills. Third, the aggregate financial performance of the GOCC sector has not improved significantly since the start of the 1988 reform. But, while improvements in individual GOCC's financial performance remain important, the direction of sector restructuring, for example, the role of the private sector, has come to dominate how each GOCC should be reformed. Finally, future resource requirements are high because of the past under-investment in infrastructure and social sectors. The NG transfer to local government units (LGUs) are also increasing substantially. Given these expenditure needs, fiscal management will continue to be tight in the medium-term, and the Government needs to be selective and efficient in what it chooses to do. National Government 1.3 The Government of the Philippines (GOP) has been an active participant in the country's development, as reflected by the growing share of public expenditures in GNP (Table 1- 1, Chart 1-1). From 1976 to 1984, overall government spending was relatively stable, around 14 to 15 percent of GNP. But starting with the trough in 1983-84, real spending has grown to an annual average of 19 percent of GNP during 1990-94. This represents an annual average nominal growth of 17 percent over the period of 1976-94. The average annual spending levels in the Philippines are comparable to those of other Asian countries: public expenditure in neighboring Asian countries averaged about 22 percent of GDP during the 1980-90 period (Table 1-2). In general, public expenditures are positively associated with income per capita. Although there is no right level of spending for a country given its level of income, the GOP spending is below the amount predicted for its income level in 1992. In aggregate terms, therefore, it cannot be said that the GOP spending level obviously exceeds the appropriate amount. The key issues are how much GOP spends relative to its revenues, and whether spending is efficient or not. - 4 - TABLE 1-1: FISCAL AGGREGATES 1980-94 (PERCENT OF GNP) 1980-82 1983-85 1986-88 1989 1990 1991 1992 1993 1994 Total Revenues 13.1 12.0 14.3 16.7 16.7 17.4 17.5 17.1 19.9 otw Tax revenues 11.5 10.6 11.7 13.4 14.0 14.4 15.1 15.1 15.3 Curretit 9.9 9.3 13.1 15.6 16.4 15.5 15.5 14.7 16.6 expenditures Capital 6.7 4.8 4.7 3.2 3.7 4.0 3.9 3.8 3.4 Expenditures Infrastiuctuie etc. 2.9 1.4 1.2 1.4 1.2 1.4 2.2 1.7 2.4 NG Deficit -3.5 -2.0 -3.6 -2.1 -3.4 -2.1 -1.2 -1.4 -0.2 Source: DBM TABLE 1-2: TRENDS IN PUBLIC EXPENDITURE IN ASIAN COUNTRIES (AS A PERCENTAGE OF GDP) Country 1980 1985 1990 Couintry average (1[980s) Bangladesih 11.7 11.8 13.2 12.4 India 18.7 20.7 17.4 19.8 Indonesia 23.7 19.6 19.0 20.7 Korea 21.5 18.5 16.8 18.4 Malaysia 32.3 32.3 29.3 30.2 Pakistan 22.7 23.5 23.2 23.4 Philippines 15.5 14.0 20.4 16.7 Sri Lanka 31.0 33.3 28.2 30.9 Thailand 19.8 19.6 15.5 17.9 Viet Nam NA NA 24.1 21.3 Asian average 20.8 21.5 22.4 22.6 Source. World Bank data base. Country Economic Reports. 1.4 Structure and Composition of Expenditures Two salient features underscore the recent trends in public expenditure pattern: (a) the share of current expenditure in total expenditure - 5 - has increased since mid-1980s primarily driven by rising interest payments; and (b) capital expenditures and maintenance spending have been suppressed since mid-1980s. The structure of expenditures has evolved from the high public investment-cum-high deficit of the early 1980s to the low investment-cum-low deficit today. The turning point was the financial crisis of 1983-85. The stabilization efforts pushed down capital and maintenance spending while interest payments started to rise due to the increased relative reliance on domestic financing of budget deficits. Wage payments also increased gradually since 1983-85. GOP revenues increased significantly after the tax reform introduced in 1986-88, from 13-14 percent of GNP to today's 18-19 percent . But recurrent expenditures rose faster due primarily to the continued rise in interest payments, shrinking public savings from 3 percent of GNP in early 1980s to 0.3 percent by 1990. Investments and maintenance outlays never recovered to the pre-crisis levels lest the fiscal gaps threaten another financial crisis. Chart 1-1 Public Expenditures 1976-1994 I .20- 15 I h er it apital 10 15 0 ..... . . . . -: ntena n 0 0 Source: DBM 1 .5 Current expenditures have increased and become more inflexible because a greater portion of the outlays is now mandated or non-discretionary. In the early 1980s, current expenditures were about 10 percent of GNP, and were equally divided between wages and operations and maintenance. From the mid 1980s, they started to increase and are now about 15 percent of GNP. This rise in current expenditure was driven most dramatically by an increase in interest payments which increased from 1-2 percent of GNP to 6 percent of GNP by 1989-90 as the Government borrowed to finance a significant portion of budget deficit. The wage bill, which had been compressed severely in the wake of the financial crisis, rebounded and increased to a level higher than the pre-crisis level by 1988 with the implementation of the first round of the Salary See World Bank (1992, 1993) for description of the tax reform. Standardization Act. Concurrently, during the 1980s, the number of personnel employed by the national government also increased almost by 50 percent cumulatively (although the GOP managed to stabilize the size of the civil service during the 1990s). As a result, personnel expenditure rose from 3.8 percent of GNP in 1980 to near 6 percent in the early 1990s. Furthermore, the Local Governmenit Code of 1991 has mandated the NG to transfer an amount exceeding 10 percent of the total expenditures (2-3 percent of GNP) to the local government units (LGUs) from 1994, adding to the noni-discretionary portion of the NG spending. 1.6 These non-discretionary recurrent expenditures tend to crowd out more discretionary expenditures, namely investment and maintenance outlays for existing capital stocks that play a catalytic role for private investments. Table 1-4 indicates the amounts of these intlexible or mandated as compared with the total revenues each year, and Chart 1-2 shows how mandated and non-discretionary expenditures (wages, interests and LGU transfers) have suppressed more productive expenditures (investments and maintenance) from the mid-I 980s on. 1.7 Both public investments and maintenance outlays suffered unsustainable cuts. Capital expenditures by the NG have slipped from above 5 percent of GNP in the early 1980s to less thlan 3 percent in the late 1980s. Much of the decline was due to strict limits on the overall public sector deficit imposed by stabilization programs. The improved budgetary situation enabled some restoration of capital expenditure, to a close to 4 percent of GNP in 1993 and 94, and prompted a rise in investments in infrastructure and in capital transfers to public bodies. These transfers largely comprised outlays to GOCCs, local governments and contributions to the Comiprehenisive Agriculture Reform Program (CARP). NG's own capital formation (primarily roads, irrigation, agriculture, education, and health) has been only sliglhtly over I percent of GNP mlost of the years since the 1983/85 crisis. Furthermore, consolidated public investments (NG & GOCC) plummeted from 7-8 percent of GNP to about 3 percent between 1984 and 1989 (Table 1- 6). As discussed later, these cuts in the productive expenditures have deteriorated infrastructure stocks and, as seen in the power crisis, led to an eventual breakdown of services, costing the econlomily Imluch more than the nominal savings when cuts were made. 1.8 O&M spending has also been reduced. It is included as budget item called MOOE --maintenance and other operating expenses. This item used to be as high as 5-6 percent of GNP. It has now fallen to less than 2 percent of GNP, below what is considered by many as the minimum necessary . To make things worse, what is classified as MOOE in the government accounts in various sectors includes many other expenditure items whichl often consist of adminlistrative expenses unrelated to maintenance (education) or mostly wages for casual workers (roads), and not the cost of proper physical maintenance. 1.9 Table 1-3 is an attempt to derive "pure" O&M spending by subtracting from the total MOOE debt service components, labor cost, and MIOOE spending in defense on an obligation This is derived based on simple assumptions. If capital - output ratio is within the range of 4-5, and public sector capital stock is 20 percent of the total , public sector capital stock can be estimated as about equal to the national income. According to Heller (1979) the appropriate operation and maintenance varies between 30-50 percent of ihc stock (health), to 8-10 percent ( roads) to I percent (buildings). 2 percent should be considered the bare iininiuiii if these figures are remotely robust. - 7 - basis. From 1989-90, the O&M spending has declined by 0.4 percentage points of GNP, and considering that MOOE of the Agrarian Reform increased by 0.3 percent of GNP in 1994 alone, the outlays in remaining economic and social services appear to have declined more than half a percent of GNP since 1990. Particularly worrisome are the declines in O&M in education, which has not been devolved, and in transport for which the GOP expressed correctly that operation and maintenance has the highest priority. Chart 1-2 Mandated and Productive Expenditures 14 12 - .: 10~ Z 8 6 4 2 0 (0 CD 0 (N -; o C - CD -O co OD r Mndated Productive TABLE 1-3: NATIONAL GOVERNMENT O&M SPENDING BY SELECTED SECTOR (% OF GNP) ______ 1980 1983 1986 1989 1990 1991 1992 1993 1994 Pure O&M Spendn 2.251 .1 1.95 2.20~ 2.23 2.3 19 18 .2 Social Services 0.65 0.81 0.76 0.74 0 72 0 79 0.6-9 0. 59 0.5-3 40-2. 0.27 -0.27 -O.- -- -2-7 w Education 00.37 0.2 0.42 34 0.32 Health 0.27 0.33 0.26 0.30 0 29 0.32 O031 0.23 0.20 Econiomic Services 1 07 0.81 0.71, 0.89 0.95 0.701 0.66 0.59 0____3 o/wAgrarian Reform 0.02 0.02 0.021 0.22 0.30 0.0 0.05 0.05 0.35 Agriculture ____ 0.09& 0.10 0.091 0.15 0 19 __ 0211 0.18 01 01 ~~~~~~~0r_ _ _ _- 01 012 Natural Resoe 0.1 1 0.1 0.08 0. 08 7 009 0.07 0.06 0.08 Transp and Comm.lt711 0.50 0.48 0.34 032 0.29 0.291 0.301 0.20 Transp and Comm 0.71 | 0.50 048 0.341 03 .9 029 0 02 Source: DBM I.10 Tax expenditures The tax expenditures, foregone revenues from various tax exemptionis and credits, are recently estimated to be about P 30 billion a year. BOI incentives (EO 220 anid EO 226) amount to nearly a half of this, of which 50-60 percent (P 6-10 billion) are exemptionis of tariff and duty on imported capital equipments. In addition, special laws and other EOs exemiipt many private and public bodies, including NPC, from various tax and tariff payments. - 8 - The direct cost of these expenditures is the foregone revenues, but more indirect cost, i.e., creation of distortions, should not be ignored either. The GOP's Task Force on Tax and Tariff Reform has looked into different costs of various exemptions. A continuous examination of costs and expected gains of tax expenditures is needed so that the incentive structure will be rationalized and streamlined. Furthermore, management of remaining incentives needs to be more transparent. 1.11 The budgetary release mechanism has been cited among factors that tended to slow the implementation of public investments because these discretionary expenditures are subject to curtailment even after the Congress has approved the budget. This happened especially when revenue shortfalls were foreseen (Table 1-5). (Improvements are being introduced in 1995, but it is too early to tell if there has been marked improvement--see Chapter 5.) The key issue is that the Department of Budget and Management (DBM) has released resources to line agencies only as 3 revenues were collected to adhere to the overall deficit targets established under stabilization programs. This practice has increased uncertainty faced by the line agencies regarding the total annual amount and the timing of resource availability. Such uncertainty has discouraged the agencies from spending resources on discretionary items for the first several months of the fiscal year (fiscal year is January I-December 31) until agency managers are sure non-discretionary items (e.g., wages) can be fully paid. Given only the first three months of a fiscal year are the dry season most suitable for construction work (i.e., the resources released during the rest of the fiscal year could not be invested as productively ), this practice has made it very difficult to effectively implement public investment or mnajor rehabilitation activities even when resources were available, ex post. TABLE 1-4: BUDGETARY INFLEXIBILITIES--MANDATED EXPENDITURES (PERCENT OF GNP) 1988 1989 1990 1991 1992 1993 1994 Wages 5.2 5.6 5.8 5.7 5.4 5.3 5.6 Interests 5.8 6.0 6.6 5.9 5.8 5.1 5.3 LGU Transfers 0.4 0.4 0.4 0.5 1.2 1.8 2.2 Total Mandated 11.4 12.0 12.8 12.1 12.4 12.2 13.1 Revenues 14.3 16.7 16.7 17.4 17.5 17.1 18.9 Diff. bet. Total Mandated and Revenues 2.9 4.7 4.1 5.3 5.1 4.9 5.8 Source: DBM 3 This has also had the side effect of making the DBM highly powerful; their fund release patterns, rather than the intenlion of Congress or other executive departments, have determined the real budgets (see Chapter 5). - 9 - TABLE 1-5: PROGRAMMED AND ACTUAL REVENUES AND NG INVESTMENT (P BILLION) 1988 1989 1990 1991 1992 [ 1993 1994 Programmed Revenues 122.4 145.0 179.5 206.4 278.9 284.2 319.2 Actual Revenues 112.9 152.4 180.9 220.8 253.1 260.3 325.6 Programmed NG 16.1 24.7 35.3 47.9 53.1 57.3 45.7 Infrastructure and other Investment Actual NG Infrastructure 15.2 20.9 29.1 37.6 46.1 40.2 42.4 aild other Investment Consolidation of the Public Sector 1 12 Consolidated Public Investment The behavior of consolidated public investments (Table 1-6) has mirrored that of the NG's investment, but its composition shifted reflecting the priorities of each administration. It declined rapidly starting with the 1983/84 financial crisis from 7.4 percenlt of GNP (1980-83 average) to 3.4 percent (1984-87 average), and started to increase to above 4 percent of GNP only in the 1990s. In 1984, imnmediately after the start of the crisis, the Marcos administration reacted by reducing social investments from 0.53 percent to 0.22 percent of GNP. The education sector took a particularly hard hit. Investments in the transport sector also were decimated, but the administration struggled to keep energy and power, and water sector investment at as high a level as possible. The Aquino administration took a different strategy; it increased social sector spending and road investments--but sharply curtailed investments in other infrastructure such as power, energy and water precipitating the start of the later infrastructure crisis. The Ramos administration, which inherited the legacy of the Aquino administration-- weak infrastructure and high poverty incidence together with a more liberalized policy environment--first invested lieavily in physical infrastructure (power and transport), and is nlow turning to address the social issues. - 10 - TABLE 1-6: PUBLIC INVESTMENT BY SECTOR (% OF GNP) Year 1980-83 1984-85 1986-88 1989-91 1992-93 1994 Administration and (Marcos) (Financial (Aquino, (Aquino, second (Ramos with (Most Events Crisis) first half) half) power crisis) recent) Public Adniinistiation 0.18 0.17 0.25 0.32 0.27 0.49 Social Investmenit 0.59 0.30 0.73 0.65 0.19 0.64 Education 0.16 0.12 0.25 0.31 0.12 0.33 Health 0.04 0.03 0.05 0.06 0.02 0.08 Othe rs 0.39 0.16 0.42 0.28 0.05 0.23 Economic Investment 6.60 3.45 1.93 3.76 4.31 4.86 Power/Energy 2.24 1.20 0.51 0.94 1.65 1.89 Water 0.34 0.26 0.11 0.18 0.13 0.08 Trspt/Co:nm 2.35 0.31 0.85 1.69 1.94 2.14 Others 1.67 1.68 0.46 0.95 0.59 0.75 Total 7.37 3.92 2.91 4.73 4.75 5.99 1.13 Consolidated Public Sector Deficit The GOP has also taken measures to maintain 4 the consolidated public sector deficit (CPSD) at a sustainable level (Table 1-7). On average it has beeni 1-2 percentage points higher than the NG deficits (an average of 2.2 percent of GNP over 1989-93, the 1994 figure is preliminary). There have been three institutions, beside the NG, that suffered significant deficits in the past: the Central Bank, the OPSF, and the National Power Corporation (NPC) a public sector supplier of electric power. Among these, the GOP has recently dealt with two major sources of deficit: the Central Bank; and the OPSF. The Central Bank losses were dealt with in 1993 by transferring poor quality liabilities into another entity CB-Bureau of Liquidation (CB-BOL), and by making sure that the new Central Bank (Bangko Sentral ng Pilipinas-BSP) would have positive income and resources for independent monetary management. While this central bank restructuring did not substantially change the total consolidated deficit, it has made possible more transparent deficit accounting and more efficient monetary management by the new central bank, BSP. The OPSF deficit has been narrowed recently by timely adjustments of oil product prices, and there is a plan to make the oil price adjustments fully automatic CPSD includes financial accounts of the NG, the Central Bank, the Oil Price Stabilization Fund (OPSF), the Government owned and controlled corporations (GOCCs), Government Financial Institutions (GFIs), and social security institutions. - 11 - eliminating the need for the OPSF altogether. The remaining deficit center is the NPC. Given the serious needs for reliable power supply, the NPC should continue to invest, especially in transmlission activities, and the NPC may have to suffer deficits for the next five years. 1.14 The improved CPSD in 1994 is a combination of factors including increased tax efforts by the national government. But, again, what really turned the tide was the privatization proceeds, which amounted to P 62 billion, that improved both NG and GOCC accounts. A decline in the operation and maintenance spending during the 1990s has also contributed to the smaller CPSD. TABLE 1-7: CONSOLIDATED PUBLIC SECTOR DEFICIT (BILLION PESO, AND % OF GNP) 1987 1988 1989 1990 1991 1992 1993 1994 Total Public Sector -12.1 -25.9 -35.1 -54.3 -26.5 -27.0 -41.1 -10.4 Balance (percent of GNP) -1.8 -3.3 -3.8 -5.0 -2.1 -2.0 -2.7 -0.6 National Gov't -16.7 -23.2 -19.6 -37.1 -26.3 -15.9 -21.9 -1.8 GOCCs 0.2 2.9 -3.0 -16.4 -7.4 -12.3 -25.6 -13.; OPSF -8.5 -1.5 10.1 5.4 -7.9 2.6 BSP/CB/CB-BO1. -10.7 -15.9 -20.9 -21.9 -21.2 -21.8 -16.0 0.3 GFIs 1.0 1.8 2.7 3.1 2.4 3.8 6.1 2.7 SSS/GSIS 5.4 4.4 8.3 10.0 8.2 8.6 11.6 0.4 Others/Adjustment 10.8 4.1 5.9 9.5 7.7 5.2 12.6 -1.6 Government Owned and Controlled Corporations. 1.15 During the 1970s, the GOP established a large number of GOCCs to, inter alia. improve provision of infrastructure services. The GOCC sector borrowed one fifth of domestic loanable funds and absorbed almost half of external loans to the country during 1975-84. Yet, the sector suffered from a number of inefficiency and its value added contributed a mere 1.8 percent of GNP. By 1984, the GOCCs numbered more than 300, and required budgetary support that neared a quarter of national government spending. Recognizing how serious the problem had become, the GOP undertook a far-reaching sector reform during the 1980s through privatizing/liquidating more than 200 GOCCs, and strengthening performance of the remaining GOCCs. An oversight mechanism of GOCCs was also established within DOF. All performance is now evaluated based on "contracts" between the GOCC and the NG. This section briefly reviews performance of the GOCCs from 1989, a year after the latest reform efforts started, to 1993, and examinle if the benefits of the reform has been sustained. In addition, Volume 11 has detailed - 12 - discussion on GOCCs in transport (PNR, LRTA and PPA); power (NPC): water sector (MWSS and LWUA); and agriculture (NIA and NFA). TABLE 1-8: TOTAL INTERNAL CASH GENERATION OF MAJOR GOCCs (IN P MILLION) 1989 1990 1991 1992 1993 All 14 GOCCs 7426 5493 9780 13102 11426 (Percent of GNP) (0.8) (0.5) (0.8) (0.9) (0.8) Withlout NPC -18 2170 6550 2673 4321 (Percent of GNP) (0) (0.2) (0.5) (0.1 (0.3) Source: Department of Finance (Corporate Affairs Group - CAG) TABLE 1-9: NET INCOME OF MAJOR GOCCS (P MILLION) 1989 1990 1991 1992 1993 Export Processing Zone (EPZ) 3 3 65 46 63 Authority Local Water Utilities (LWUA) 24 41 30 157 117 Administration L-ight Rail Transit (LRTA) -350 -445 -585 -472 -370 Authority Metro Manila Transit (MMTC) 81 -275 -116 -47 -344 Corp. Metropolitan Waterworks (MWSS) 494 792 1069 1194 1072 and Sewerage System National Development (NDC) 722 -344 206 139 750 Company National Electrification (NEA) -630 -468 -335 -459 -332 Administration Nationial Food Authority (NFA) -1785 -1026 -2761 -3090 -1373 National Housing (NHA) 20 115 61 26 7 Authority National Irrigation (NIA) -39 -67 -31 -156 -37 Administration - 13 - Nationial Power (NPC) 1661 -65 -2930 4118 1565 Co rpo ration Philippine National Oil (PNOC) 1277 1714 2034 2500 3476 Co. Phlilippine National (PNR) -170 -284 -296 -285 -336 Railway Philippine Ports Authority (PPA) 381 358 194 396 418 Memoit): Nominal GNP 914,126 1,082.947 1,266,070 1.385,562 1,519,814 Source: Department of Finance (Corporate Affairs Group - CAG) 1.16 Financial performance of the 14 major GOCCs, listed in Tables 1-8 and 1-9, shows limited improvements during the early 1990s. Their internal cash generation (ICG) more than doubled from 1990 to 1993 although a large portion of the ICG was generated by NPC, which enjoyed numerous implicit subsidies and tax exemptions; and to PNOC and NDC, which benefited from proceeds of privatization of their subsidiaries. Positive profits continue to be elusive for many GOCCs with no trend of improvement. Even those with net positive profits, the 6 return on assets remained below 2 percent in all GOCCs but PNOC and MWSS . Their performanice was limited despite the fact that the market demand for services provided by GOCCs has been on the rise in recent years--demand for passenger and freight traffic, and for electricity and water connections, for example, have consistently shown an upward trend. There is thus an ample evidence that the limited financial performance of GOCCs has been due to supply side constrainits. 1.17 The principal constraint faced by GOCCs is the inadequate financing for rehabilitationi and expanding capital stocks due to a number of reasons. First, to a large extent, they suffer from inability to adjust the low user charges for improved cost recovery; and inefficienit collection of existing user charges. For example, the Philippine National Railway (PNR), and Light Rail Transit Authority (LRTA) have not adjusted their fares since 1989 and 1991. respectively, despite their persistent net loss position. The Philippine Ports Authority's tarif'fs were frozen from 1985 till 1994, and National Irrigation Administration (NIA)'s irrigation fee lhas been constant since 1975 (note that part of the fee is in kind). NPC and Metropolitan Waterworks and Sewerage System (MWSS) have recently adjusted tariff rates, but they still have large annlual financing requirements. Second, collection rates are low: for Local Water Utilities Adminiistration (LWUA), it is 64 percent, and for NIA, 45 percent. Third, the problem is compounded by large system losses due to poor network maintenance and pilferage: 58 percent of water MWSS produces no revenue, 43 percent for NIA and 20 percent for National Electrification 5 Thcsc GOCCs receive most of the NG transfers, and are monitored by the Ministry of Finance. MWSS appears to be under-spending on maintenance and their physical assets under-estimated. Thus, its actual late of return may be lower. - 14 - Administration (NEA). Finally, poor levels of service also discourage users from paying user fees creating a vicious circle of poor service and poor funding. 1.18 Because of these problems, many GOCCs still have to depend on the NG resource transfers to them. The data on net flow of funds from NG to the major GOCCs appears to suggest that it hias been reduced significantly during the last two years on accounts of increased tax payments (Table 1-10 first row). But, it is misleading because improvemenits were due to the increase in tax payments by NPC in 1992 and 1993. NPC's tax exemption status was confirmed by the supreme court in August 1994, and NPC has requested refund of tax payments made before that date amounting to P 15 billion. NPC has already received P 8 billion in 1994. If this transaction is taken into account (see below on NPC), the recent reduction in the net NG payments to GOCCs was small (Table 1-10, second row). Excluding NPC, the GOCC sector received from the NG, net, P 6.4 billion in 1989 and P 4.3 billion in 1993. This is a slow improvement. TABLE 1-10: GOCC-NATIONAL GOVERNMENT FINANCIAL FLOWS (CASH BASIS) 1989 1990 1991 1992 1993 Net Flow (with NPC) -5341 -5273 -7711 457 -2058 Percent ot'GNP (0.6) (0.5) (0.6) (.03) (0.1) Net Flow (without NPC) -6395 -4236 -4686 -2109 -4281 Percent of GNP (0.7) (0.4) (0.4) (0.2) (0.3) From GOCC to NG (without NPC) Tax payments 391 765 1338 980 1919 Interests 2918 1004 72 903 82 Dividends 80 197 130 219 629 Total (to NG) 3389 1906 1540 2102 2630 From NG to GOCCs (without NPC) Equity 2078 2677 1544 200 1290 Net Lending 3461 1418 1450 1754 1681 Subsidies 4245 2047 3194 2257 3939 Total (from NG) 9784 6142 6188 4211 6911 - 15 - 1. 19 To eliminate the vicious circle (poor services leading to poor user charge collection, and thus poor maintenance) and to establish a robust base for economic growth with improved provision of infrastructure services, many GOCCs have embarked upon ambitious investment programs. The 14 GOCCs trebled capital expenditures from P 12.7 billion in 1989 to P 40.5 billion in 1993. NPC alone increased its investment from iP 6 billion to P 27 billion during the same period to address the critical power shortages. If NPC is excluded, ICG in the aggregate financed 63 percent of the GOCCs' investmenit between 1989-1993. Domlestic borrowing finaliced 33 percent and external borrowing the remaining 4 percent. 1.20 Many of these corporations will want to pursue ambitious investment programs to alleviate the infrastructure bottleneck as part of the modified Medium Term Development Plan. Projected capital expenditures are to the tune of P 20.3 billion for 1995 and e 32.5 billion for 1996 excluding NPC. But, as discussed in Chapter 2 and in Volume 11 there is a need to examine the role of the public sector and appropriate investment programs for each sector. For exam1ple, LWUA's role should depend on the desirable size of water supply schemes--whether they should be the size of existing water districts in general or they should be regionalized to take advantage of scale economies. If the GOP is to privatize MWSS, its investment programii needs appropriate adjustmients. The National Food Authority (NFA)'s generalized price subsidy has been ineffective in alleviating poverty while being a heavy financial drain on the public sector. The GOP needs to contilnue to phase out NFA's subsidized activities and divest the associated assets. Another examlile of a GOCC needing a fundamental structural reform is the PNR which has suffered from contilnuous and significant losses despite repeated efforts to revitalize with donor assistance. It should probably slhrink its network to commuter rail around Manila and theni be privatized at an appropriate time. Finally, the National Housing Authority (NHA ) should einphasize housing for the poor. and accordingly adjust its investment program. 1.21 NPC-NG Relationship NPC alone constitutes a major public finance issue; its operatinig deficit in 1994 was 1.2 percent of GDP despite direct and indirect support from the NG (Table 1-11, See Background Paper Section VI for details). NPC's finalicial situation will likely remain precarious until 1998 when the deficit is projected to narrow to 0. I percent of GNP (due to the decreased need for investment) even when rather optimistic revenue assumptions are employed. Three maiii reasons are at the roots of the NPC financial difficulties. First, the need to overcome ihe recenit power crisis caused a significant increase in investments and operating costs. Capital outlays reached 2 percent of GDP in 1994, more than twice as much as in 1989. Operating cost also increased because purchases of power from independent producers incr-eased substantially lrom 1992. Second, despite the adoption of an automatic formula, prices have not beenl promliptly adjusted to reflect changes in fuel costs and exchange rate movements resultilIg in increased NPC deficits. Third, a number of investment projects which would yield returnis over a long temporal span have been financed with loans with a much shorter term increasling slhort term financial difficulties. 2. THE GOVERNMENT'S ROLE AND RESOURCE ALLOCATION ISSUES Introduction 2. 1 Fiscal policy and public expenditure decisions are critical tools the Government has in achieving the goal of higher income growth while maintaining social justice. This chapter reviews the GOP's role in resource allocation. In the first section of the Chapter, three general points are made: (i) the expenditure level and pattern should be determined to achieve long term goals of development and not be used as a short term counter-cyclical maneuver; (ii) the role of the Government, in the presence of the active private sector, should be carefully reviewed with the relative cost of private provision of public services in mind; and (iii) the Government could productively provide even private goods when the equity impact is significant. In the second half of the chapter, issues specific to major sectors are reviewed and recommendations for reforms and/or changes in expenditure patterns are summarized. (The details of sector discussions appear in Volume 11.) 2.2 The key sector messages are as follows: (i) in the transport sub-sectors, the GOP needs to continue to devolve parts of responsibilities to local governments and the private sector while maintaining strong network planning and a regulatory framework; (ii) for utilities, the GOP's focus on privatizing "production"* of power and water is in the right direction, and the GOP can increase efficiency further by taking advantage of scale economies through consolidating distribution systems; (iii) the GOP can focus on productivity increase in rural areas by investing in rural infrastructure and research. Also, the uncertainty regarding how and when the Agrarian Reform will be completed has allegedly reduced private investment because of the ban on changing land use in the areas designated by the reform program. Reduction of this uncertainty will thus likely contribute to increased private rural investment; and (iv) in social sectors, it is important that the poor get access to quality education and health services, and that the service provision be done in a more cost effective manner. In addition, strengthening technical and administrative capacity of LGUs is of a highest priority, because many of these responsibilities have been devolved to them. Determinants and Effects of Aggregate Expenditure Patterns 2.3 It is difficult to try to impute motives to the setting of priorities by decision makers; expenditure decisions are intensely political and the result of substantial compromise. Attempts, however, can be made to find some of the factors which go into the choices. Table 2-1 shows regression analyses of sectoral shares of budget on the rate of growth of tile previous year (tc see how much expenditure changes as a result of good times versus retrenchment) and the debt service tas a fraction of GDP) requirements of that year. These factors together are important determinants of the overall budget envelope. 2.4 At this relatively aggregate level, spending on economic services (infrastructure, irrigation, etc.) represents a distinctly counter-cyclical maneuver as reflected in the negative coefficient for growth. These expenditures tend to rise in bad times--the GOP intentionally "leans - 17- against the wind" by supporting the economic services as a way to speed up the resumption of growth. 2.5 The effect of debt repayment is far and away most pronouniced in the share of economic services, indicating a strong substitution between debt obligations and physical infrastructure. This corresponds to the observation above that there has been a long term decline in infrastructure investment as a result of the debt crisis and high debt service obligation that ensued. Combining these results indicates that there is indeed substantial rebuildinig and new public investments to be done to redress the legacy of the debt-ridden years. However, this should be done according to a well planned strategy of providing public infrastructure and not be varied as a countercyclical tool. 2.6 In contrast, other expenditures (education, health, defense and public administration) tend to respond to transitory incomes, and increase in periods of faster growth. They can thus be considered more along the lines of consumption goods. Although the relationship is weak for any one of the categories taken individually, together the relationship is quite strong. Interestingly, the share of expenditure going to social services did not fall with higher debt repaymiienit. This is likely due to the change in regime which had a greater coinmitment to social services and the simultaneous increase in debt service payments. It is to the credit of the governmnent that such social sector spending were relatively protected from the need to repay loans. TABLE 2-1: DETERMINANTS OF TOTAL EXPENDITURE BY FUNCTIONAL CATEGORY 1975-1993 Iiidepeindent variable Economic Services Social Services Derense Public Service Growth of the Previous -.099* .032 .017 .031 Year (.049) (.028) (.022) (.019) Debt payment as % of GDP -.242* .074* -..020 .079* (.074) (.034) (.042) (.025) { R2 .62 .16 .56 .48 Autocorrelation .463 .045 .803 .229 (standard errors in parentheses, * = significant at 5 percent) 2.7 Using spending on economic services to stimulate short-termi growth in bad times, however, has not been productive. A statistical analysis indicates no correlation between shares of expenditures on economic and social services on the one hand and next year's economic growth on the other. In fact, economic services had the weakest correlation with future growth. Thus, to the extent that one category of expenditure was favored relative to other parts of the budget as a means to stimulate growth, not much came of it. While not a definitive analysis, there is no evidence to support a position other than using the government investment (and its rates of returns) and expenditure programs to achieve their long-term goals and not to vary these targets for the sake of short term stabilization. - 18 - 2.8 In addition, to the extent that the social sectors are used both as a redistributive device and as a means of correcting market failures, there is no reason to compromise that mission for macroeconomic goals. Indeed, regional public expenditures in education and healthi appear to significantly contribute to improved human capital and medium term growth of the region contrary to the limited significance of the aggregate figures. Both recurrent and capital expenditure on education increases significantly the share of students completing primary education, and health expenditures also contribute to reduction in infant mortality rates -- but this effect is limited to poorer regions (see Background Paper Section VIII) suggesting that in richier regions public health expenditures are substitutes for private health care. Efficiency and the Role of the Public Sector in the Economy 2.9 Tight budgets, expensive revenue collection, and a multiplicity of claims on the public purse, all make it essentia; that government activity demonstrate its ability to add tangibly to the economy. Government furnishes goods and services ranging from pure public to pure private. In deciding where to cut and where to expand expenditures, it is crucial to determine which kinds of activities will substitute for private provision and which are truly additive. Within each sector it is possible to decompose budgets into three sorts of categories: * those which are for pure "public" goods, that is, goods which cannot be provided by the private sector. The principal reason .vould be that it would not be possible to refuse services to those who do not pay fees (within practical limits of the cost of exclusion). Tolls on local surface roads, for example, are impractical. In the most extreme example, it is impossible to exclude anyone from the benefits of national defense against foreign invasion. Here, the justification of government expenditure is completely clear and particular investmenlts onlly need to satisfy the social profitability rules in project evaluation. * those in which the private sector can participate but not at desirable levels. They would need to be subsidized (or taxed) or provided with a regulatory framework. The cost of the regulatory needs of the sector is but a small fraction of the provision itself. Electric power generation, for example, needs the regulatory framework for selling into the national power grid (which can be public or, with more careful regulation still, privately owned and operated). However, the administrative costs of monitoring and enforcing the contracts governinlg those sales are trivially small compared to the welfare cost of raising the funds needed. The current trend toward Build-Operate-Own (or Build-Operate-Transfer) schemes is a miajor recognition of this oni the part of the government. Control of communicable disease is another such category - there will be private care for patients but it may come too slowly and with too little thoughit to prevention. * those which are essentially private goods that are disproportionately conisumed by the poor. Providing them is a means of transferring resources to the poor in lieu of direct monetary redistribution. A similar analysis was carried out for infrastructure at a national level. The result is miixed probably reflecting long geslation and lumpiness of investment. - 19 - 2.10 The GOP has done well in allowing private investors to take part in the infrastructure and social areas. But, there is room for improvements. Many goods which are essentially private in nature are publicly provided. These goods are ones for which charging appropriate user fees is possible and practicable and those who do not pay fees can be excluded from consuming them. Electric power generation is the best example of formerly public services that can now be privately provided, and progress has been made in extending the scope of the private sector in other infrastructure areas. Other examples would include housing (which features promiiiently in the government's agenda for social reforms) and higher education. Further, there is substantial variation in the extent to which these goods and services favors the poor. This opens the possibility to specify more clearly the purpose of these budget items and to focus resources better on those expenditure categories which are better suited to the redistributive function. 2.11 In deciding on where to cut and where to expand public expenditures, it is crucial to determine which kinds of activities will replace private provision and which truly add to the overall service level. Each sector is characterized by activities with varying characteristics on this dimension. Three clear examples may serve to illustrate. 2.12 Health. Expenditures by the Department of Health (and its local equivalents) run the range from purely public goods to purely private. On the public end of the spectrum are such activities as pest control and certain forms of sanitation disposal. The government has no choice but to provide them. If the government doesn't, no one will. 2.13 In a middle ground are services with high external effects. Infectious disease control is critical since people's own demand for services will not necessarily take into account the possibility of infecting others. Services which combat consumer misinformation are also critical in health. The value of preventive care is frequently underestimated by the public and no private provider has much of an incentive to correct this misperception. 2.14 Finally there are services which are entirely private in nature (such as the treatment of noni-comiimunicable disease). The public sector tends to provide this activity due mostly to market failures in insurance. However, much of the insurance problem could be handled through better regulation. 2.15 It is essential that those activities which are public in nature are provided sufficient funding and are not diluted by competing claims on resources which have less rationale for governmnent intervention. Results in the health section below (paras 2.85-91) indicate that different kinds of public expenditure can have very different effects on the health status of the population with highl externality activities as well as subsidized treatmentfor the poor being the most effective. On the latter point there was significant progress during the 1980's. On the former, perhaps too little had been spent on preventive services. With devolution, the NG's role has been circumscribed substantially; interregional transfers in health per se are no longer feasible. As the general role of the NG in health care delivery has diminished, specific efforts should be bolstered in remaining areas which are of NG concern. Activities which cross jurisdictional boundaries such as infectious disease control should have the highest priority. 2. 16 Education. In 1989, the GOP decided to expand access of the public to secondary education by subsidizing students with no access to public facilities to go to private secondary - 20 - schools. This is certainly a laudable goal but it may have been an expensive policy in relation to the aggregate increases in educational attainment. The comparison should be with net increases in enrollment as a result of the policies implemented, rather than increases due to secular trends in income and the simple passage of time. Since secondary education is more of a private good (students can be excluded if they don't pay fees) and there had been a substantial private sector in secondary education before the reforms, there is the distinct risk that added public expenditure simply shifted the burden of payment to the government from individuals. 2.17 Indeed, it appears that this measure increased (i) subsidization of those who would have opted to go to private schools anyway, and at the same time, (ii) created a third type of secondary schools that are called private schools but dependent on this government subsidies and provided low quality education services. It is difficult to see if this measure has contributed to the objective of universal secondary school enrollment because of the difficulties in disentangling the effects of price changes accompanying the reform from general secular trends. That the extension of secondary education came at a higher cost to the public is a serious problem. This is not to say that secondary education is a bad thing, only that it may come at the cost of reduced resources for primary education. It is at lower levels of education that more genuine external effects are generated for the society. Further, primary education is one of the most consistent (across numerous international studies) means of transferring income to the poor. There is still ways to go to achieve universal primary education. The added cost to DECS of having to handle secondary education, by substantially replac;ng a functioning private sector, compromises the more critical goal. 2.18 In higher education, this argument is stronger still. The returns to a higher education are almost exclusively private and are reflected in higher wages for the graduate. The vibranicy of the private university sector indicates substantial substitution possibilities for publicly provided subsidies via the state universities and colleges. In relation to other countries in Asia, the costs of higher education are much lower, precisely because of the competition in the private sector (Mingat and Tan). 2.19 The costs of higher education and the difficulty of borrowing against future earnings may justify government interventions. But this should be in the form of loans (with full expectation of repayment) and not subsidies. The benefits, if any, of public subsidy would have to be on the grounds of income distribution. The state university program has been one of the more extreme examples where political influence, Congressional insertions, and supercession of DECS decisions have outweighed technical considerations. 2.20 Power and Infrastructure. Again, investments and expenditures covered by DPWH and NPC span the range from almost pure public goods (surface roads where tolls are infeasible) to services with large network externalities (power grid operation) to purely private goods such as power generation and electricity use. After very rapid progress on privatization in the power sector in response to shortages, subsequent actions have been much slower. The subsidies to NPC both explicit and implicit still distort private decisions. As pointed out in the recent report on private opportunities in power (World Bank, 1994), while there is an important role for the coordination of the electric power grid, generation and sale of electricity are essentially private activities. The costs borne (other than those caused by pollutioni) and the benefits received are fully excludable and rival. While some regulatory effort is necessary, the legitimate role of the - 21 - governm1ienit in raising investment funds and in supplying services at less than long run marginal cost is nil. 2.2 l The decision between building infrastructure and buying services should be kept actively in mind whenever infrastructure services are needed. Buying services can help obtain the desired products, since investment plans will not be interrupted and civil servant wage obligations will not overwhelm line agencies at the expense of needed operational expenditures. Similarly, contracting out services and making good on the contracts can assure that actual outputs are achieved as in road maintenance by protecting non-wage operational inputs form being squeezed by adjustments coming late in the year. The recent decision by Congress to reduce road maintenance through private contracts thus needs a review. Once the public's role In improving allocations is clear, there should be clearer mandates for following through on the obligation. Privatization, corporatization and purchases of goods/services rather than supplying through the civil service can all help. 2.22 Cost of private provision The private sector, however, does not provide a panacea for the GOP. The private sector will make investments in physical infrastructure when and only when it expects a high probability of earning an acceptable rate of return. In addition, most private investors require a pay-back period of three to five years. These limit the types of investments which can be made by the private sector. They may also create distortions because investments will be made to maximize financial return to the investor rather than to optimize the benefits to the country. It is particularly true when positive externalities are associated with the investment --for examiiple, the private sector will find it hard to invest in a mass transit system that reduces conigestioni and air pollution. 2.23 Even when the private sector does decide to invest in infrastructure, the cost of private capital should be evaluated carefully; it has higher cost of capital than when the government raises resources (because of additional risk incurred, high returns to competing investments available to them, and shorter maturity periods). Exceptions occur when the capital available to the governmetnt is limited by both insufficient fiscal revenues and limited access of the government to capital markets. In such a case, the private sector access to funds may become truly additive to governmi1enit investment. However, when governments do access capital markets through treasury bills and bond issues, it is essentially competing for the same sources of funds at the margin with the private sector. Equity in Public Expenditure. 2.24 It is difficult to determine with great accuracy who gets what from the public purse. Only relatively indirect methods are available to see to what extent the poor in the Philippines benefit more from public expendituie than do others. There are few surveys of national coverage which both identify the degree of use of public services and houselhold per capita income - the two ingredients necessary to pin down the distribution story. Here some inferences are made on the basis of regional data - per capita subsidy costs and average regional income. While this informationi cannot determine whether poor people in poor regions receive their share of expenditure or if the benefits accrue to wealthier people within poor areas, some inferences can be miade. - 22 - 2.25 Ensuring that more of a particular spending reaches the poor relative to higher income groups, however, may be too stringent a test for judging the redistributive capacity of expendliture. The real comparison is with the incidence of the marginal tax revenue needed to pay for the increased service. If the tax system is very progressive, meaninig that a large fraction of expenditure is funded by the relatively wealthy, then the system can support a number of activities which, while not received more by poor than non-poor, are still net transfers due to their funding meclhaniisml . Chart 2-1 Distribution of Taxes and Expenditure by Regional Income 3000 2500 -4--Tax j -iSn : ; ; 7 | : t-- Roads q 2000 -i--X Health t Basic Ed Sanitation 1000 Water LU 500 0 -a 1 3 5 7 9 11 13 15 17 19 21 23 Per Capita Regional Income Lower Part of Chart Enlarged 450 4 300 t) 300 -F 0 : i ealth ot 250 6 _ ~ - = 5 : XBasic Ed 2! 200 _ Sanitation , 150 - - . . a :xS O -e--Water 100 1 3 5 7 9 11 13 15 17 19 21 23 Per Capita Regional Income 2.26 The Philippines tax system is roughly proportional (i.e., everyone pays something quite close to 10 percent of their income in taxes) due to the reliance on itidirect taxes (which are - 23 - slightly regressive and contribute 70 percent of the total collections) modified by personal and business taxes (which, while a smaller fraction of collections, are considerably more progressive, see World Bank 1993). Chart 2-1 compares the per capita tax payments with receipts of various 2 expenditures per capita based on data for the 1980s and up to 1992 . In many of the expenditure categories, absolute progressivity does not hold throughout the income distribution. But, lower income regions do receive more from the spending programs than they on average pay taxes to support these programs. 3 2.27 The clearest pro-poor programs are health and primary education (prior to devolution). The area which is least progressive (at least cross regionally) is public works (mostly roads). Since much of infrastructure has characteristics of public goods. this may not be alarming. Its pattern is skewed largely by the concentration of DPWII activities in the National Capital Region. Public infrastructure with supports economic activity will naturally be more productive in the more industrial parts of the country. While there is always a need to monitor the specific services within the infrastructure aggregate to make sure opportunities for unbundling privatizable services are taken, this becomes even more important when distributional issues are considered. The reduction of subsidy to privatizable infrastructure in urban areas will reduce the regressive tendency of expenditure evident in the data. 2.28 But, after the devolution started, both education and health expenditures allocation 4 (GAA) lost their pro-poor patterns--richer regions received relatively more allocation (Background Paper Section VII and VIII). For health, positive correlation between the poverty and per capita sector expenditure by NG ends as early as 1992. The rate of decline in health resource allocation is higher in poorer regions, while the NG expenditures have beeni protected from the effect of devolution in richer regions. Regarding basic education, the regressive pattern emerges from 1994, and worsens in 1995. When 1994 data was examined by the level of education, all levels of education have negative correlation with poverty incidence and that the primary education spenidinig was unfortunately most regressive. As discussed below, these regressive NG spending patterns might very well be aggravated by devolution, because the IRA transfers to local governmenits were not implemented based on the income redistribution needs in mind (Chapter 3). 2.29 A major focus of the government's social priorities has been in the housing sector. Looking more closely at that sector reveals both opportunities and risks in addressing the needs of poor people through public subsidies. Charts 2-2 a and b shows two items taken from descriptions of the National Housing Program. The first is the number of housing units "needed" by region. Data is presented in regression form in order to facilitate comparison - that is, the curves represent estimated expenditures per capita based on the region's mean income. Education results need to be interpreted with care. Spending per student is rather constant across regions especially at the primary level (see Chapter 10). The "per capita" progressivity may merely mean that there are relatively more school age children in poorer regions. Also, it is found that the poor have limited access to quality education (Chapter 10). That the NG spends more on poorer regions may not ensure that poor families benefit from this spending pattern. GAA is used partly because of difficulties to obtain detailed actual figures for recent years, and partly because GAA expresses "intention" of the government towards income distribution. - 24 - This is an estimate based on the number of substandard units and projected growth of housing demand (from natural increase and from in-migration). The difference between the overall demand for housinlg and that component of it which is due to actually substandard housing is quite striking in terms of the income of the regions generating that demand. If housing policy could be directed specifically to substandard units (without, of course, producing an incentive to keep housing unlits substandard!) rather than spread uniformly over the projected increase in housing units, the degree to which the poor could benefit would rise substantially. In fact, the way in which the majority of housing subsidies is being administered - through the granting of mortgage subsidies - it is likely that the amount of money flowing to the public through this channel is going primarily to the non- poor, and probably to higher income groups. No data cross-checked with household incomlle is available to confirm this, however. Chart 2-2 a Potential Incidence of Public Housing Program 20 X, 5 ... . 0~ 10 0 5000 10000 15000 20000 25000 30000 Income per Capita of region Chart 2-2 b Substandard Housing by Regional Income I 7 6 5 :~4 cn3 z~2 I0 8 8.5 9 95 10 10.5 Reglonal Income per capita (in logs) - 25 - 2.30 Policy implications of distributional analysis. Government intervenes in the economy for reasons unrelated to helping the poor. However, expenditures which have little to do with correcting market failures should have a beneficial distributional profile in order to justify public subsidies. There are two ways to affect the amount and the distributional characteristics of the subsidy: by manipulating the fee charged for services (such as water, electricity, university education or health care) or by changing service characteristics and thereby alter the pattern of use of the service (e.g. location of facilities, comprehensiveness of network hook-ups). The distributional profile helps to determine the most effective way of improving delivery of services to the poor. If the profile is not progressive: either poorer people get less than richer or, worse, poorer people get less than their contribution to revenues, and fees should be raised. The poor are not hurt disproportionately and the subsidy is reduced. Extension of services to more of the population will reach the poor and change the distributional profile over time. Having access to safe water at a cost is better than having no access at all. Omissions 2.31 Inappropriate public expenditures are easier to identify and fault than failures to spend money where it would do more good. Put another way, public expenditures designed to remedy private sector market imperfections can be evaluated, while complete market failures, as in the case of public goods, may be more important but likely to be overlooked. Environmental protection and pollution abatement are two areas that warrant more attention. 2.32 Environmental protection will not be done adequately without public involvement. The main environmental issues have been reviewed in Bank and other documents. The governmenit has been increasing the allocation to the Department of Environmiiient and Natural Resources at a faster rate of growth than the overall budget (39 percent versus 4 percent from 1992 to 1994) and it is difficult to set specific numerical goals for this increase. However, the steadily increasing congestion and pollution of Metro Manila indicates substantial scope for stronger efforts. It is not necessary that such effort be more expensive. Much pollution abatement can be accomplished with taxes, either on industrial effluents or on the use of automobiles in center city. Reliance on tax instruments are usually more efficient than other methods of pollution control such as enforced standards (World Bank Discussion Paper #287) for two reasons. First, they allow individual producers to find the best way of reducing pollution or otherwise paying for the costs they impose. Second, by raising revenue instead of spending it, there will be substitution of an efficient tax for the distortionary taxes currently in place. The recent report, "Philippines Pollution,,,,," recommends such a tax based solution along with a description of how best to implement it. 2.33 Beyond the tax solution, alleviating congestion is likely to require real investments. Evaluating transit systems and infrastructure investments which change urbanl form and traffic flow should be a high priority. - 26 - Policy Recommendations on Sector Expenditure 2.34 Key recommendations for each of the major sector are summilarized below. More 5 detailed examinations of sector expenditures appears in Volume II of this report. Crucial questions that need to be asked in deriving the recommendations are as following: * Is the mix of sector spending appropriate for the economy'? Does it complement, supplement, or comipete with the current or possible role of the private sector? Is this NG spending consistent with the ongoing devolution process? * Where there is redistributional intent, is the sector spending benefiting the poor more than upper income groups? * Are the expenditures achieving the outcome which were intended, or are they inefficient? TRANSPORT SECTOR 2.35 Introduction Although the transport sector includes a diverse set of modes and networks, a common set of policies improve the efficiency of government expenditure in the sector. A key factor is to reduce the size of networks that the NG should build and maintain so that the quality of services can be as high as possible. Specifically, the GOP needs to consider: (i) agreeinig on appropriate division of responsibilities between the NG and LGUs; (ii) enlarginig the scope for transferring transport services to the private sector; (iii) imlproving network planning that takes into account LGU and private sector participation; (iii) addressing the issue of pricing and regulation; and (iv) down-sizing staff according to the reduced role of the government. 2.36 Devolution to Local Governments The first two concern improvements to the transport network. In the case of the roads, railroads, ports and airport. this means a reduction in the size of the network to be maintained by the central government to an essential core which the governmiiienlt can afford to maintain to an adequately high standard. First, decisions need to be made as to which part of the network would be transferred to LGUs. Then LGUs would have to make the same decision as to wlhat is sustainable. However, this transfer should also include transfer of the source of revenues needed to miaintain the local network. This implies an explicit transfer of taxing and/or pricing authority. The local governments would then be faced with balanicing the level of tax or price which the populace would accept with the size of the local network which could be maintainied to an acceptable standard. In the case of roads, most of the network imiaintenance work for non-national roads has already been devolved. Furthermore, the Power sector and health sector are not discussed in Volume II because strategies ot these sectors have been discussed in recent World Bank reports (World Bank 1994a, World Bank 1994b). - 27 - governmenit is considering to focus the maintenance work on the core artery network of about 14,517 km out of the total national roads of 26,554 km. The devolution of smaller ports to local governments should reduce PPA's responsibilities while allowing local governments to make the changes necessary to improve their performance. 2.37 Private Supply The second issue is to transfer selected responsibilities for providing transport services to the private sector to take advantage of its demonstrated ability to construct and maintain infrastructure and to operate and maintain transport services. The Government has already started doing this successfully in power. The government must first examinie its role as a provider of public infrastructure and essential services and then identify those functions whiclh the private sector could perform more effectively. However, the GOP would continue to insure access to the infrastructure and optimal pricing of the services so as to maximize the economic benefits derived from the network. The GOP should also insure that the individual services and infrastructure provided by the private sector are effectively integrated with the rest of the network so as to maximize the efficiency of the network. For example, toll roads should be well integrated as part of the highway network. Much more road maintenance needs to be carried out by private contractors rather than through force accounts--a trend whiclh recently was reversed by Congress. In addition, the Philippine National Railway (PNR), which has been too small to take advantage of scale economies and has been making persistent losses, needs to be reformed. One alternative is to convert it to a short-distance commuter rail linked witlh the rest of Manila's urban transport and, possibly be privatized. Finally, the Government should continue to act as a provider of low-cost funds to finance the development of transport infrastructure. To do this, it should diversify thie sources of funds, to include not only tax revenues, ODA and government loans but also bonds which are either issued or guaranteed by the central government and public offerings of corporatized transport functions. 2.38 Network Planning A related issue is the improvement in network planning. In the case the urban transit and roads, this means integrating the individual rail transit systems (Light Rail Transit Systems and privatized and commuterized PNR), interfacing them with the existing transit and public road transport systems and developing the urban roads as components of a network rather than individual links. In particular, the returns to the third LRT system would be enhanced significantly if it were to be effectively linked to first and second systems of LRT. In the case of the ports, this means merging the ports located within a metropolitan area and including the private ports in the planning for additional capacity requirements. In the case of the air transport, this would mean establishing a system of municipal and private airports which would serve the .iaeler markets and private aviation. 2.39 Pricing and Regulation The fourth issue is to introduce more efficient pricing of transport services. The rationalization of pricing is especially important for the ports, railroad and airports because the present system has created an elaborate system of cross-subsidies favoring specific markets by setting prices far below the users' willingness-to-pay. This system prevents the government from applying basic financial management criteria such as equating costs with prices or maintaininig a minimum level of surplus. The most effective way to control prices and their related costs would be through greater reliance on market forces. The benefits of deregulation and increased competition have been felt throughout the transport sector. The last vestiges of price regulation (and regulation of the frequency of service) are being eliminated from the domestic - 28 - shipping and have been partially removed from the air transport sector. The railroad and the port sector have not benefited from this autonomy with the result that they continue to provide a lower quality at prices which do not cover the costs. 2.40 The phasing out of the government's role in economic regulation of the transport sector should be complemented by a strengthening of its regulation of safety and environmental protection. As the private sector involvement increases and the forces of competition encourage a reduction in operating costs, it will become more important for the governmenit to insure that a imiiimluI standard of sal'ety is mzaintained and that the reduction in financial costs does not create a greater cost in terms of environimlental degradation. In order to avoid a conflict of interest, it will be necessary to develop regulatory agencies which are not involved in providing transport services. This imiplies removing these responsibilities from the PPA, ATO, PNR. and perhaps, from the DOTC. 2.41 Staffing The final issue to be tackled is the downsizing of the government agencies involved in transport. There are situations where there is obvious over-staffing not only through the in(discrimninate use of casual labor but also because of the failure to reduce the staff as their responsibilities have decreased. The amount of staff required will continue to decrease as their responsibilities are devolved to the local governments or transferred to the private sector. For examiple, DPWH's labor force should be smaller with the increased contract maintenance and sh.ould he miore professional so that better network planning can be made. Similarly, PPA is over- stafl'ed with administrative personnel. There is a need to reduce this layer of staffing, and conltract/casual workers as responsibilities are devolved to LGUs and the private sector. For countries such as Malaysia , Singapore and Thailand, this downsizing has been relatively easy because tlheir rapidly expanding economies absorbed the excess labor. While the Philippines has not yet attained this level of growth, there is some scope for voluntary separations due to the combination of low morale and expanding employment in the private sector. There is also some scope for the transfer of skilled personnel in situations where the private sector assumes responsibility for the activities which they were involved in. There is the iminediate option for eliminating casual labor. Hlowever, a small percentage should be maintained since some of the casual labor fill positions whicil are critical. For the remaining staff, the government must rely on attrition and early retirements but without stopping to recruit younger, more enthusiastic and better educated personnel to take on the changing responsibilities of the governillelit agencies. POWER SECTOR 2.42 Introduction The Philippines has experienced a major crisis of electricity supply in 1992-93, with blackouts averaginig seven hours per day being common in maniy regions across the country. Facing serious economilc losses, the GOP sought to supplemenit the capacity of the National Power Corporation (NPC), the Philippines' monopoly supplier of electricity, with private power developimient. As a result, the crisis ended by early 1994, and it becamie clear that the growing involvemilenit of private power producers will result in a marked transformation of the power genieration subsector fromnl nearly 100 percent public in 1991, to nearly 80 percent private by 1998. Since the 100 percent of distribution utilities are already private, the private interest will have comie to dominiiate the power sector. - 29 - 2.43 The Government's major power sector goal continues to be to meet all future capacity requirements in collaboration with private developers and private capital. Including long- deferred replacement of aging plants, capital requirements for generation during 1994-2000 are about US$ 10 billion or more; for the same period, investments in transmission should exceed US$3 billion. Neither the GOP nor the private sector, acting alone, can mobilize such amounts. Therefore, the GOP and the private sector must continue collaborating constructively for the foreseeable future. 2.44 Issues and Constraints To realize this collaboration, there is a need to bring in effective competition and appropriate risk sharing between suppliers and distributors. In addition, there is a need to formulate appropriate roles of NPC, GOP agencies, and the private sector. Competitioin and risk sharing have been inhibited by captive relationships, franchise restrictions, and mloniopolies. NPC is suffering from over-worked equipment and lost public credibility on account of the recent power crisis. Yet, NPC is still the dominant supplier in the sector, purchasing for resale more than 80 percent of planned private sector-produced electricity and assuming most of the inarket risk as well. The Government would prefer that NPC pass the rnarket risk to the distributors (who are all private). but most are so weak and fragmented that few private power genierators are yet inclined to deal with them directly. 2.45 Another constraint on NPC is that it should not serve as sponsor of the private power producers, developer of its own generation capacity, wholesaler of electricity, and also simultaneously act as (i) an impartial planner and operator of the transmission system and (ii) formulator of rules of dispatch. In the future, when supply is ample, multiple suppliers will need unliase(d access to the transmission system. Therefore, a separate transmission company, indepenident of any supplier, which a dispatch entity that is competent to establish fair and transparenit rules, needs to be developed. 2.46 Proposed Sector Structure The GOP realizes that to achieve this goal, NPC needs to be restructured and Government agencies strengthened to capture the benefits expected to ensue from privatization. The distributors apart from MERALCO need to pursue market strength througLl consolidation. The desirable sector structure would include: largely private generation with restructured and part privatized NPC; consolidated distribution utilities; and a backbone transmission company which can be publicly owned at lirst. 2.47 NPC's generation facilities should be spun-off, as proposed by GOP, to wholly- owned subsidiary regional companies functioning as holding companies for existing plants and developers of new state-sponsored thermal generation facilities. Alternatively, NPC could create even mlore subsidiaries, based on functional or geographic considerations, to manage its generation facilities. These subsidiaries' should enable them to follow commercial operating practices, and once soundly operated, sell their shares to private interests. NPC will continue to have a vital headquarters function in the future. In the near term, it should establish a sound basis for spinning off the transmission company and regional subsidiaries. Once the subsidiaries can obtain financing on their own merits, the parent will act more like a holding company; it will (i) assist the subsidiaries with cash management, (ii) collect interest and dividends from the subsidiaries; and (iii) meet its remaining liabilities. - 30 - 2.48 Consolidation of 130 small distribution utilities into 15 or fewer continuous units is recommenieiided because real competition among suppliers depends on the development of more commiiercially-viable buyers. Such a consolidation can lead to needed efficiency reforms, and larger distributioni utilities can then take appropriate market risks. The proposed consolidation, however, poses imaniy political problems; important entrenched interests are driving hard to maintain the status quo. Therefore, GOP needs to focus on an interim arrangement, whichi (i) progressively reduces the extent of market risk assumed by the Government; and (ii) provides appropriate incenltives for consolidation (including subtransmission networks of 69kV). 2.49 For its part, the public sector must focus on maintaining an enabling environment, within which the private sector can compete without undue constraints. The GOP's capacity to create and maintain a favorable climate for private sector-led growth also needs strengthening througlh: (i) establishing clear policies and rules for independent power generation, which in turn requires enhanced GOP (DOE) capacity for policy formulation and rule-making; and (ii) adjusting the legal and regulatory framework. 2.50 The following three functions should remain at the national level: Power Systemn Planning (this function should be shifted from NPC to Department of Energy (DOE) once the latter has developed the necessary capacity); Hydroelectric Development; an(d Backbone Transmnission Svstemns and Dispatclh. The last, the backbone transmission system could still be regionlalized. However, responsibility for the system should be retained at the national level, so that the priority for interconnection will not be subordinated to regional concerns. The need to ensure supplier-s unlrestrained access to the transmission system provides a compellinig rationale for creatilng a transmission company that is distinctly separate from any supplier, including state enter- prises inivolved in power generation. For the same reason, this proposed new company should also owII anid operate all facilities for dispatch. The dispatch function must be executed fairly and efficienitly. The participation of the Government, NPC, the private power producers and distributors in the fornulation of a dispatch entity and dispatch policy can be an outstanding opportunity to reachi long-termi consensus on those principles and enable a competitive environment for private sector participation. 2.5 1 Reg,ulatory Framework . A major feature of the proposed sector structure is increased relianice on effective competition as a regulator, although DOE, ERB and NEA will contilule to be involved in regulation and monitoring. Because the recommlienided structure anticipates and supports direct contact between suppliers and distributors, it also has the important advantage of relying less on Government regulation than several alternative frameworks that were considered. ERB would continiue regulating the distributors' tariffs since there would effectively continue to be protected monopolies. To strengthen their capacity, ERB, DOE, and NEA would still iieed to undergo a substantial institution building effort and staff strengtlhening due to the increased size and technical complexity of the work load. 2.52 Implementilng a fair and transparent load dispatch systeml is the most important residual ut'ture roles in the power sector for the Government. Load dispatch should be linked to opel atioin of the transmission system; that linkage appears to have advantages in the Philippines, w here n(o precedent exists for collaborative pooling among competing suppliers. As substantial additionis to capacity come on line, the dispatcher will determine the order with which plants are - 31 - brouglht on line so as to minimize consumers' costs. However, provisions will be needed to honor NPC-executed PPAs that guarantee a high off-take for some relatively high cost electricity. 2.53 Pricing Reform. Higher voltage power tariffs should by structured to reflect costs by: (i) unbundling fixed and variable costs within tariffs, and (ii) introducillg time-of-day differentials, at least for high and medium voltage consumers. Tariffs also need to explicitly include costs of stand-by capacity. Moreover, at the wholesale level, power pricing in different re- gions should reflect the true costs of generation and transmission. Current ihter-regional cross- subsidies have served as a substantial disincentive to private power producers interested in locating plants in the Visayas or Mindanao. Finally, separation of transmission costs and development of wheeling charges as part of the tariff for energy can be developed and later applied to the bulk transfer of electricity between buyers and sellers. In order to encourage private power producers to site their plants at favorable locations without losing the flexibility to identify the best possible selling arrangements, an acceptable framework for wheeling charges over the transmission and subtransmission systems is needed. 2.54 Taxation Taxation of fuels for power generations needs to be rationalized. To enable all suppliers of electricity to pay the same amount for fuels while shouldering the fuel supply risks for themselves, fuel tax exemptions should either be extended to the private power producers or be eliminated altogether, which would require 10-15 percent increase in retail electricity prices. Given this tariff increase comes without concomitant improvements in service, and that there would not be loss of existing tax revenues, the exemptions for now should be expanided to the IPPs. This policy, however, needs to be reviewed as the sector reforms are implemented. WATER SUPPLY 2.55 Introduction. Inadequate maintenance of water supply infrastructure and sustained under-investment in sewerage and sanitation have contributed to water shortages and environmental degradation. As public health concerns become prominent, fundamental reform of the sector and a new strategy for achieving it become more pressing. Decentralization, private sector participation, and integrated water resource management--including integration of water supply and sanitation-- would constitute the defining elements in shaping the new strategy. 2.56 Maior Issues Water supply is comprised of approximately 500 Water Districts (WDs). 1000 municipal supply schemes and 4,000 Barangay pipes water supply schemes (serving peri-urban areas), all independently managed and operated. These are supported and funded througlh different public sector institutions--MWSS for the Manila Metropolitan Area; LWUA in other urbani areas; and DPWH for rural areas. Most of these schemes are small, financially unviable, and without economic incentives or the force of effective regulation to control leakage and costs, and maintain quality service. Many supply schemes and their customers are caught in a viscous circle of attempted recovery of excessive costs through higher tariffs following and precedinig yet further private tapping of alternative (ground water) sources. 2.57 There are two GOCCs in charge of water supply, LWUA and MWSS. MWSS has been unable to reduce non-revenue water to acceptable levels over many years while LWUA has had problems recovering loans from WDs and has limited capacity to support all the WDs under its jurisdiction. The current pattern of incremental expansion of coverage based on internal resources - 32 - of Water Districts or small loans from LWUA often leads to sub-optimally engineered systems based on small diameter pipes extending out in a "spider" formation. To avoid high sunk costs, sector improvement plans should be based on network investments compatible with appropriate master plans. Furthermore, while incremental investments is supply have made it possible for the quantity of water supplied to increase, minimal investment in infrastructure for waste water disposal leaves only 10 percent of the population in Metro Manila and 2 percent outside connected to sewer service. and the price is beginning to be paid in increasing pollution of water resources, recourse to more distant and costly sources of raw water and higher morbidity rates due to waterborne diseases. 2.58 Recommendations for Future Policy Changes Improving sector performance will depend on cost reduction, demand management and increased efficiency in production and distributioll, all elements of commercialization and the private sector participation (PSP) initiative. The PSP strategy includes: (i) rapid introduction of GOP-supported PSP alternatives in the large mletropolitan centers (privatization of MWSS, and systems in Metro Cebu and Davao) with contracts tailored to their specific circumstances; (ii) consolidation of administration of smaller schemes into regional service areas, with scale economies in management and the potential to attract private sector participation in due course; (iii)credible regulation of private service operotors, and (iv) improved maintenance or major network rehabilitation in Metro Manila in order to sharply reduce distribution losses, prior to or as part of privatization of MWSS. For rural areas, a programii of sequenced, demand-driven infrastructure investments is needed. This should be coordiniated by the public sector, with resources allocated competitively on the basis of pertormanice. AGRICULTURE AND NATURAL RESOURCES 2.59 Background Agriculture accounts for almost one-fourtlh (22 percent) of the GDP and almiiost half (45 percent) of the labor force. But, the health of the sector has been uneven. After strong growth of over 5 percent p.a. during the 1970s, growth was negative during the economic crisis of 1982-85. During the second half of the 1980s, the sector recovered and grew by 3 percent a year. only. Recent growth declined again to 0.9 percent. Several factors account for the slowdowni in sector growth after the 1970s. These include: an overall downward trend in internlationial commliodity prices affecting the Philippines' traditional export crops, continuous deterioration of intersectoral terms of trade, macro-econom-nic constraints, and a series of natural calamities. Uneven growth and the uncertainty associated the Government's principle rural poverty initiative, the Comprehensive Agrarian Reform Program (CARP), has led to under-investment in the sector, which, in turn, resulted in weak infrastructure, and feeble support services . With other sectors unable to absorb a growing labor force, poverty is now concentrated in rural areas, where it is estimated at over 50 percent. Rural poverty, in turn, has led to many political and social problems, and had an adverse impact on natural resource management and the environment. 6 The public sector spending in agriculture and natural resources have been about 5-9 percent of the total government spending , i.e., the share of resources the Philippine agricultural sector receives from the NG is mIucI less than the GNP share the sector produces. The ratio of these shares is 0.27 for the Philippines as opposed to 0.6 for Thailand. and 0.4 for Indonesia. - 33 - 2.60 The Government is challenged to address simultaneously rural poverty, the continuiug depletion and degradation of natural resources and the environment, and rural infrastructure and support services. While increased spending during the 1980s on natural resource and environimental management was appropriate and effective, other responses to rural problems have not been so. Most CARP expenditures have been for landowners' compensation; price supports were not targeted on the poor; and irrigation, rural infrastructure, and support services have been left wanting. 2.61 The sector's problems were compounded by devolution to LGUs of agricultural extenision, rural infrastructure, (including communal irrigation and tertiary roads), and some environimental protection and natural resource management. Either because of the implementation capacity constraint of the LGUs or because of the LGU's' own preference, some of these activities, suclh as communal irrigation and integrated social forestry, have been almost at a standstill since devolution in 1992. Existing Investment Plans 2.62 Level of Public Investment Recent average annual public expenditures on agriculture, including agrarian reform of about P 20 billion are insufficient to maintain growth, reduce poverty and manage natural resources and the rural environment. Public underfunding follows a decline of 58 percent in lending for agriculture by the banking system over 1981-92. To reverse this trend, the Medium-Term Public Investment Plan (MTPIP) for 1993-98 projects a level of P 67 billion (excluding the Agrarian Reform Fund (ARF), which is outside the regular budget), rouLglly doubling of the sector's share in total public investment from about 5.6 percent in 1993 to 10.1 percent in 1998. 2.63 Composition of Investment Annual budgets of some P 4 billion each for the Department of Environment and Natural Resources (DENR) and the National Irrigation Adminlistration (NIA) for the remaining years of the MTPIP (1995-98) represent an appropriate doubling of 1994 allocations. Furthermore, the proposed phasing out of expenditures by the National Food Authority (NFA) on interventions in the grains market (price support) is a welcome change in Government policy. On the other hand, the ARP, which is outside the MTPIP, will continue to command a disproportionate share of total proposed rural expenditures. A huge amount (63 billion) is projected for landowners' compensation by 1998, without assurance of reinvestment by the landowners in those rural areas. Similar imbalances are projected elsewhere. For example, the Department of Agrarian Reform (DAR) projects P 20 billion for support services and infrastructure for one million beneficiary families over 1994-98. Yet, during the same period, the DENR must make do with only P 15.6 million for capital outlays under the MTPIP to benefit four millioni families living in the uplands, and also sustain responsibilities for bio-diversity conservation, forest protection and management, reforestation, watershed protection, and pollution control. - 34 - Recommendations 2.64 For the rural sector to contribute more to the economic growth and to alleviate its poverty, the Government must work specifically on: (i) enhancing rural productivity through investilng in rural infrastructure and research; (ii) increasing private investment in rural areas by, amonig otlhers, reducing uncertainties associated with the Agrarian Reform; and (iii) protecting environmiiienit and natural resources. Also in pursuing these objectives, it needs to (iv) increase coordilnation of development strategies among sector agencies, and strengtheni institutional capacity of implementing agencies. 2.65 Enhancini? Rural Productivity The public sector needs to invest more on rural infrastructure and in research and extension. These have high externalities and could be powerful in raising rural productivity--be it of agriculture or non-agriculture--infrastructure by lowering transport cost of inputs and outputs or increasing availability of irrigated water, and research and extensioni hy improving yields or introducing new crops that have high market potentials and value added. Given that many of these have been devolved to LGUs, the NG needs to consider providing selected LGUs with technical assistance in preparing and implementing infrastructure programs and extensionl services. In addition, it is urgently needed to develop a framework that allows cost sharing between NG and LGUs of devolved functions with high externalities (see Chapter 4) and that incorporates LGUs into foreign assisted projects. The NG also needs to lead in demand - driveni agricul,ural research and dissemination of research results in collaboration with the private sector. TIhe GOP need to make sure that the extension services carried out by LGUs are linked with the research activities. 2.66 Increasing Private Rural Investments Private investment increases with higher productivity and higher policy stability. It is thus important to implement measures discussed above to raise rural productivity, in particular, improving rural infrastructure. In addition, private investors require stable policy environment in which entrepreneurs can respond flexibly to market forces. In this regard, the vivacity of private investors has been dampened by the uncertainties associated with the future of the Comprehensive Agrarian Reform Program (CARP) especially as, unce land is designated by CARP, its conversion of use is difficult until land redistribution is compilete. To make things worse for investors, the CARP has been moving slowly and increasing ftuLre uLicertainties due mostly to: its high cost; limited impact on the poor without adequate rural infrastructure; and political opposition from existing land owners. Only about a quarter of the targeted 10.3 million ha over 1987-98 has been distributed, and the deficit for the next four years is projected to by over e 70 billion--resources that can also be well spent on rural infrastructure. 2.67 It is thus critical that the Government reduce these uncertainties. The DAR is alrea(dy considering, innovative policy changes that would be required in continuing the reform eftorts with reduced uncertainties. These changes include: "recalibration of targets"; imposition of a progressive agricultural land tax; allowing either direct selling or volunitary land transfer between landowilers and farmers without the intervention of the Government; and transforming the ARF into a revolving fund. 2.68 Environm1ental and Natural Resource Protection In the environment and natural resource imaniagenient, watershed management deserves more emphasis. Both institutional changes - 35 - and increased investments are necessary in this area. Given the extent of watershed degradation in the country, it is recommended that a long-term, programmatic approach be developed, with the contribution of external financiers, to watershed rehabilitation and protection clearly planned and monitored. The GOP also needs to address the absence of a national-level body to plan, coordinate and monitor a comprehensive and integrated approach to waterslhed managemelnt 2.69 Coordination of Development Strategy. In order to achieve greater efficiency in resource allocation to agriculture, there is a need to integrate three distinct development approaches currently being followed: the DAs growth strategy based on comparative advantage being under its Medium-Term Agricultural Development Plan (MTADP) --- essentially a top-down approach which picks "winners" for rice, corn, commercial crops, livestock, and fisheries, in "key production areas (KPAs)"; the DAR's bottom-up approach, involving participation by the local communities and NGOs, primarily focused on poverty alleviation, represented by the Agrarian Reform Community (ARC) concept; and another bottom-up, participatory approach of the DENR, primarily focused on sustainable natural resource management. In mid-1994, the DA and DAR reached an agreement that when the KPAs and ARCs existed in the same area, the ARCs will serve as the main conduit for tlle DA's program. A similar agreement is needed among DA, DAR, DENR and local governinenits regarding their respective jurisdiction. Further, the MTADP itself needs significant adjustments since some of its physical and financial targets are unrealistic. IN addition, the weaknesses in national water resources planning and allocation needs to be addressed through strengtlheniing the National Water Resources Board. EDUCATION 2.70 Constitutionally, education commands the highest priority in the allocation of Central Government resources. Indeed, net of debt service and IRA, education absorbs close to 30 percent of available GOP resources, a very high proportion by international standards, but less than 3 percent of GNP. So, despite the effort of the GOP, public education in the Philippines is under- financed because total budget resources are limited, the burden of national debt is large; and the central government has devolved responsibilities and resources to LGUs. 2.71 Due to increasing demand and unmet financing requirements, the public education system is responding in two ways which are incompatible with quality education. First, student/teacher ratios are increasing, and essential non-salary operating expenses are being deferred. Under-financing has particularly affected public secondary education. Second, inadequate NG resources has stimulated rapidly increasing education finaiicing from LGUs and private sources. The share of public education financed from local sources has risen rapidly from 2 percent in 1990 to 8 percent of the total. Nevertheless, substantial variation in local resources among regions and population groups portents growing inequity in public education among those same regions and population groups. 2.72 The bulk of LGU financing goes for MOOE (43 percent), but also includes personniel expenditure (25 percent). Private education financing takes two forms. Private schools, accounting for substantial share of total post-elementary student enrollment (33 and 79 percent at secondary and tertiary level respectively), are the more visible form. The other is the parental contribution, in the form of user fees (miscellaneous school fees, contributions to the PTA, the cost of bvoks and other school materials, and transportation). All in all, total commitment of national - 36 - resources for education is estimated to be close to 7 percent of GNP. Despite this non trivial national investment, considerable progress still needs to be made in improving both the quality of the education system and its relevance for national development. Equity 2.73 There are large disparities between and within regions in elementary sch0ool enrollment and retention, and these differences are associated with poverty and the quality of schooling. Generally, the poor liave more limited access to quality education and are less likely to renmain enrolled to completion. Overall, two of three children entering first grade complete the elementary cycle. But there is near universal completion in the affluent areas, and over 70 percent noni-comiipletion in the poorest provinces in Mindanao and the Visayas. Surveys among out-of- school children confirm the strong incidence of economic factors on school participation, particularly among post-elementary age groups. Also, even within the public schools the quality of public education is comparatively lower in the poorest areas, where salaries are lower, facilities poorer, and school management weaker. These differentials are exacerbated with fiscal decentralization and devolution, as LGUs serving higher income families can more readily provide extra support to public schools. As a result, there is a danger that the education quality gap between rich and poor areas will increase, further widening existing regional disparities in school comipletioni rates and learning outcomes. 2.74 Present criteria and procedures underlying DECS' regional budget appropriations are based on a nominally equal amount per student and do not aim at counterbalancing the regressive impact of fiscal decentralizationi on local public education finalicinig, not to speak of redressinig disparities in public elenientary education of longer standing. In fact, there is no DECS strategy for actively targeting resources towards the most pressing needs, the most disadvantaged areas or population groups. The issue is essentially political. There appears to be great reluctance on the part of the national and local legislature to face the implications of the country's considerable geographical, cultural and social diversity, and accept discriminatory use of national resour-ces tor the sake of addressing priority regional needs. As a result, DECS is largely deprived ol the authority and control required for effective social targeting. 2.75 Attempts to social targeting has instead promoted the Government to subsidize lower incomile youth to attend private secondary-schools and colleges throughi the DECS' GASTPE program (Government Assistance to Students and Teachers in Private Education). The program chianiniels tuition fee supplemenits (TFS) for students, and funds educational service contracting (ESC3 with private schools. Althougih innovative in design, and with close to fifty percent participationi of private secondary students, the program has a major flaw. For a given budget allocation. the GASTPE programii maxinizes the number of beneficiaries by setting the benefits per beneficiary at a level which is substantially below the actual cost of private schools. As a result: (i) TFS studenits are required to pay the difference between subsidy and actual cost, which eliminates a large segimient of the target population; and (ii) thie better private schools, unable to collect their full operatiilg cost from ESC benetficiaries, are withdrawing from the schemie. The ESC scheme, in its presenit forin, thus fosters the emiiergence of a third tier secondary schoolinig system: private, but low cost/(quality for the subsidized students from lower income groups. - 37 - Cost-effectiveness 2.76 Opportunities for cost savings in the operation of the public elementary education systemn are not apparent. Effectiveness can be improved by reducing grade repeaters and drop-outs. However, this will require substantial initial investment in quality improvements, which might be funded by reducing the cost to the national budget of the rather inefficient State Universities and Colleges and improving the effectiveness of budgetary processes in DECS. the expenditure monitoring system. In improving cost effectiveness, it is important to protect the MOOE, non- salary recurrenit expenditures which cover educational materials and textbooks. Governmiienit strategy 2.77 The Government intends to address quality and equity of basic education in the imnmlediate future by (i) increasing the length of the elementary cycle from five to six years; and (ii) focusing on the 19 Priority Provinces, with an integrated basic education component as recoimmilended for the Presidential Task Force for the Fight against Poverty. It is to be hoped, however, that this strategy will not lead to neglect of poverty-related education priorities in non- selected provinces, or foster uniform approaclhes to widely different regional needs and 7 circuLmstances. / Main Recommendations 2.78 The following recommendations reflect two basic objectives. First, within the pnrevailing inter-sectoral allocation of the national budget, more resources ought to be made available to DECS, so as to enable that department to fulfill its increasingly demanding mandate. Secondly, given national priorities with respect to education, DECS should seek more effective ways of spending its resources. Increasing available resources 2.79 Increasing resource availability for DECS can be achieved in two ways; i.e., by: (i) increasing DECS' share in the total education sector budget; and (ii) rationalizing expenditure within DECS' budget. 2.80 Some of these measures would include: (a) devolution of some of its present responsibilities (e.g. school construction and maintenance) to LGUs. DECS would provide selective compensatory financing to disadvantaged LGUs. This support would have to be granted in ways that do not discourage local taxation efforts, nor the extent of local resource commitment to education in this sense some equalization fund, providing matching grants, might be considered (see Chapter 4); (b) phasing out of the burden of State Universities and Colleges on the national budget by promoting fuller cost recovery, decentralized financing, and greater financial autonomy for tllese institutions; (c) rationalization and consolidation of public sector teacher education into These issues are being addressed in the proposed Third Elementary Education Project, identified jointly by DECS and the World Bank, which would promote bottom-up participatory planning processes geared to meet specific needs in targeted areas. The latter would not be limited to the 19 priority provinces. - 38 - fewer, more cost-effective programs; (d) extension of career opportunities within the teaching profession so that quality teachers stay in teaching rather than moving on to higher paying administrative positions; (e) DECS' elementary and secondary school construction program, to the extent that it is not devolved to LGUs, needs to be reassessed to emphasize proper maintenance of the existing stock; and (f) better coordination between budget preparation and execution, more transparency in budget coimmitment, and disbursement and monitoring procedures for achieving greater cost-effectiveness in the various programs and operations Effective use of resources 2.81 With more resources at its disposal, DECS should promote universal achievement of quality for basic education, particularly at the elementary level while targeting a higher proportion of expenditure towards the most disadvantaged areas and population groups. 2.82 In particular: (a) DECS needs to improve the extent and efficiency of expenditure targetling, away from the prevailing principle of strict proportionality -- per student, per teacher, per school, or other mechaniistic norms uniformly applied throughout the country -- towards addressing priority needs identified at the basis of the education system. This implies decentralization of sector management; (b) Overall budget appropriations for MOOE should increase at elementary and secondary levels, but regional allocations should again be based on actual needs; (c) Improving the quality and retention of public elementary sclhools will require lenygtlheninig the elementary cycle from five to six years, hence a considerable additional commiiiiitmiienit of DECS resources; finally (d) A case can be made for subsidizing elementary school participationi for children from the poorest households. Also, the GATSPE benefit should be targeted to students from lower income households. HEALTH AND NUTRITION 2.83 This section briefly reviews issues on the healthl and nutrition. The brevity is due, for the health sector, to the fact that it may not be the best tilme to discuss the public finance in detail given that the massive chaniges in the sector caused by devolution have yet to work themiiselves out. Furthermore, the World Bank (Philippines, 1993) has already discussed the role ot the cenitral government in the health sector. On the nutrition, there is also an ongoing effort of the Governmllent and the World Bank to assess poverty. Therefore, this report only touches on a few issues of interest which reach beyond the sector. Health 2.84 Following observations can be made on health: (i) in so far as regional allocation of public sector resources is concernied, the expenditure pattern lhad shifted from regressive to progressive in around 1986; (ii) the devolution has likely increased inequity (see chapter III below) altlhouhll it is too early to tell accurately what has happened; (iii) public expenditures in improving generail sanitary conditions improves the population's health status measured by infant mortality rates. anid importantly (iv) public expenditure in health improves the healtlh status of the poor, but not the riclh. - 39 - 2.85 Equity in Public Health Expenditure The changes in health expenditures toward higher progressivity from the period 1987 to 1991 stand in contrast with the allocation patterns in the other sectors. The others tend to exhibit a rough degree of proportionality between beneficiaries (per student in school, per capita, etc.) while the health budget indicated a much stronger pro-poor bent. It should be noted that the change in 1987 toward more equitable care was associated with a shift from preventive to curative. Devolution had reversed this and made the DOH focus more on the preventive care since curative care takes a disproportionate share of devolved activities. 2.86 Within regions, the pattern of use of public facilities across income classes (as opposed to residents of regions with different average incomes) is not as clear. The most recent inforilationi on the subject comes from the 1992 National Health Survey. Use of public facilities (for curative and patient - initiated preventive care) appears to rise with per capita household income up to levels of about 3000 pesos per month and fall thereafter. Indeed, the rate of increase in use seems to be faster than the rate of increase in tax collection up to the middle ranges of the incomze distribution. This indicates that the health care system on the curative side, at least, is progressive in the sense of transferring money from the relatively rich to the lower and middle classes but does not reach the poorest groups any more than it does the middle classes. Combining the regional and per household information, the largest benefits to recipients of health services (relative to payments made in support of these services) was in the poorer areas and within these areas, the lower to middle income residents. 2.87 Effectiveness of Public Expenditures in Health The main reason why government healtlh services are important is that there are numerous ways in which the private markets related to healtlh (quite robust in the Philippines) tend to fail to provide efficient levels of services. While the overall profile of health problems of the country has been changing (away from infectious diseases towards the chronic, degenerative diseases of richer countries), there is still a large, unifiiislhed, agenda for addressing the infectious diseases and other burdens on poor people for which public action is essential. 2.88 An numerical analysis of the effectiveness of public expenditures in the health sector for 1984 to 1991 shows that health status (here infant mortality is used as an example, Background Paper Section VIII) is affected by three categories of factors: personal income; public intervent?'on that improves general sanitation condition (safe water sources, sanitation services and immunllization rates) with their obvious externalities; and the provision of cliical care which makes up the bulk of the Department of Health's budget. 2.89 The results show that (i) income is obviously a key factor in improving health status: (ii) in general the first category of intervention (safe water, better sanitation, and more immuinization of children) is quite important nation-wide. Those services which have few acceptable private alternatives are the ones with the highest public payoff: and (iii) when the analysis is made more region specific, health care expenditures are quite potent in their impact on infant mortality but only when spent in poorer regions . At higher levels of income, the effect of public expenditure on health disappears entirely. This last result is important as it indicates that the public provision of medical care (which can be provided by the private sector) could well substitute for private provision of the same good. In the poorer areas where private provision is less reliable (and expensive enough to be out of reach of the poor for anythiig but minor services), - 40 - the public provision actually adds to the poor's budget and demand for services in ways which can translate into needed and effective medical care. 2.90 Devolution and Health Policy With newly devolved services, it is important to pin down the cause of the previous success in pro-poor provision of the health service. If the real benefits were due to a public health system which was well attuned to its clients, there may be little cause for alarm. The formerly NG workers who have been assigned to local health units would still, presumably, be able to identify local needs. The risk would be that health services as a whole might unduly lose out in conflicts with other local claims on the budget. On the other hand, if the underlying cause of these results is the level of overall health service budgets, devolution could be quite harmful by limiting the amount of services available in poor areas relative to the very progressive allocation which had been policy in the pre-devolution period. 2.91 Further issues which devolution brings to the fore are also related to the appropriate role of government; this time, the appropriate role of national governmenit. Once again, the primary question is: what can (national) government do which cannot be expected to be done by the people themselves or by the local governments. The main candidates are intectious disease control and health education campaigns of various types (smoking, AIDS and nutrition primarily) where many of the benefits of the activity will accrue to people outside of the jurisdiction in which the healthl expenditures are borne. Either because effects directly transcend political boundaries (as in pest control operations), or migration spreads the effects of interventions across jurisdictions, these policies may be underprovided by units with local responsibilities. NUTRITION 2.92 The overall malnutrition incidence in the Philippines continues to be high relative to its income level. This is due to (i) the limited resource allocation for nutritional interventions overall; and (ii) spending of modest amount without targeting. 2.93 The Philippines, with one third of children undernourished, spends 0.03 percent of GNP on nutrition intervention, an extremely low amount. Indonesia, Malaysia and Thailand spends more relative to GNP although they have less incidence of malnutrition. In India where two out of three children are malnourished, 0. 18 percent of GDP is spend on nutrition. Per child expenditures were four times more in India than in the Philippines. 2.94 One key expenditure for nutrition intervention is the NFA's grain consumption subsidy. NFA purchases 2-10 percent of the domestic production of rice and distribute at a subsidized price. The program is highly inefficient because it costs P 2-3 to subsidize P I to consumers, and, in addition, it is a general untargeted subsidy, the Metro Manila region which has 0.3 percent of the poor receive more than 20 percent of subsidized rice. It is estimated at most 50 percent of the truly poor have had some access to the subsidized rice. The cost of transferring P I to the poor then would be P 5-6. 2.95 Recognizing this and other efficiency problem, the GOP has started a program to streamline the NFA and phase out consumption subsidy. Instead, more targeted nutrition intervention programs are being considered. It is a highly commendable direction that the GOP is taking. The benefit to the poor would increase the earlier the GOP can take this action. 3. DEVOLUTION AND LOCAL GOVERNMENTS 3.1 Following the enactment of the Local Government Code (LGC) in 1991, a number of fiscal responsibilities have been devolved to local governments and the national government transfers to them were increased. This chapter reviews the consequences of devolution for fiscal management, spending patterns and equity . Detailed discussions of the background and the implementation of devolution appear in Background Paper Section 1. The GOP should be commended for the smooth implementation of this highly complex process in a very short period of time. 3.2 This chapter makes recommendation that, if adopted, would further increase the advantages of decentralization of decision making achieved as a result of devolution while better serving certain national goals. First, to make the LGUs accountable and face hard budget constraints, the NG needs to: remove uncertainty on the amount and timing of the LGU transfers; stop funding devolved activities; and monitor LGU borrowing strictly, without giving LGUs an expectation that the NG would bail them out of any possible debt service difficulties. Second, the NG may consider a reform in the allocation mechanism of resource transfer among the LGUs to incorporate concerns about equity and local resource mobilization by considering an equalization grant based on the tax base and revenue efforts. Third, another part of the resource transfer should be based on the cost share principle on a selected small number of activities that the National Government consider critical. Finally, many LGUs have insufficient capacity in fiscal plamring and implementation especially in view of the need to coordinate with the National Government. The National Government may consider giving technical assistance to LGUs in these areas. It should be kept in mind, however, that these recommendations are based on limited expenditure data by the LGUs; the whole devolution is only a few years old and 1994 was the first year of full implementation. Background and Rationale 3.3 The intent of devolution was to improve the well being of the people by empowering local voters to change the kind, quantities and qualities of the public services they receive from their local authorities. More specifically, there were three rationales: First, The Philippines has substantial spatial variations in physical conditions, economic circumstances and social attributes. It was accepted in the LGC that some types of services could be better delivered by local authorities who could take into account these differences and provide services that suit local needs and preferences better than the central government that is predisposed to provide a uniform bundle of services throughout the country. Second, less national government intervention would make it possible for local residences to hold locally elected officials accountable for their actions. Third, a higher degree of local autonomy, and hence local participation in collective decision making, was believed to reduce feelings of political alienation among residents and policy nakers outside the Metro Manila Region. 3.4 Devolved Functions As of January 1994, the following functions have been devolved from central departments to LGUs: - 42 - Department of Agriculture (DA): agricultural and fishery extension services, regulation and research; procurement and disbursement of seeds; purchase, expansion and conservation of breeding stocks; and construction, repair and rehabilitation of water impounding systems. Department of Environment and Natural Resources (DENR): forest management services; mine and geo-science services; environmental management services; and reforestation, integrated social forestry, and watershed rehabilitation projects. Department of Health (DOH): provincial health offices and district, municipal and medicare community hospitals; purchase of drugs and medicines; implementation of primary health care program; field health services; aid to puericulture; and operation of 5-bed lhealth infirmaries. Department of Public Works and Highways (DPWH): repair and maintenance of infrastructure; and barangay road, water supply and communal irrigation projects. Department of Social Welfare Development (DSWD): program for rebel returnees; barangay day care centers; and poverty alleviation programs in low-income municipalities barangays and depressed urban barangays. 3.5 Basic education is also considered as a candidate for devolution. Especially, school building construction is specifically mentioned in the LGC as an activity to be devolved. But, so far no education activity has been officially devolved by Congress. Given the diversity of cultural and other backgrounds of various areas in the Philippines, the GOP may also consider devolving the basic education so that localized decision making will be possible. 3.6 National Government Transfers to LGUs Under the LGC of 1991, the total amount of the unconditional block grants to the LGUs, called Internal Revenue Allotments (IRAs), were significantly increased. All smaller NG transfers to LGUs were absorbed into the new IRA and the IRA itself was increased from a maximum of 20 percent to a fixed 40 percent of internal revenue collections for the third preceding year. This transfer now amounts to 2-3 percent of GNP per year. Also, natural resource related taxes, that previously were shared by the LGUs and the NG, are now devolved to the LGUs where they are physically located. These latter devolved revenues amount to about we I billion in 1992 (0.1 percent of GNP) prices. Of the total IRA, 23 percent is shared by 77 provinces, 23 percent by 62 cities, 34 percent by 1,542 municipalities, and 20 percent by 41,891 barangays. Allocation among individual LGUs within the same category of LGUs (provinces, cities, and municipalities) is determined as follows: 50 percent is allocated based on population; 25 percent is allocated by the land area; and the remaining 25 percent is divided equally by all LGUs of the same category (called an equal share) 3.7 This formula gives greater weights to land area and equal share and less to population than in the past. This is because the GOP wishes to transfer more resources to the sparsely populated (and presumably poorer) areas and less to more densely populated (and more After initial adjustments, the allocation among barangays will be: population, 60 pcrccni; and cqual sharc, 40 perccnt from 1995. However, each barangay with at least 100 inhabitants will be provided a minimum allocalion of P80,000 per annum. - 43 - developecl) areas. The GOP also hopes that such an allocation rule will, at the margin, slow the rural-urban migration, by improving the quality of life in less populated rural areas. As set forth below, however, the available evidence strongly suggests that, because the services devolved to the cities were considerably less costly than the additional IRA funds transferred to them, the cities, on balance, gained more from devolution than the provinces and the municipalities. The attempt to improve rural relative to urban public services to discourage migration from the former to the latter was tlhus negated, in whole or in part. 3.8 The formula for allocating the IRA among levels of local governments and among individual LGUs was based on no other economic grounds. Therefore, the amounts transferred bear no necessary relationship to the actual cost implications of devolved functions. Nor do they take into account the capacity of local governments to raise their own resources or to carry out devolved functions. 3.9 Also notably absent from the devolution policy was the imposition by the National Governmlenit of any conditions upon local governments (such as provision of essential services or mobilization of local revenues) that must be met if they are to receive their formula-determined shares of national government revenues. 3. 10 One of the objectives of the devolution was precisely to make the intergovernmental transter system both transparent and predictable. Yet, LGUs still do not have certainty as to when and h1ow much transfers they will receive because the budget proposals and some directives (e.g., Local Budget Memorandum No. 22 dated June 15, 1994) contain instructions that diverge from the LGC. This has resulted in lower fund flows to local govermuents than the Local Government Code of 1991 stipulates. 1993 releases to LGUs were 82 percent of the appropriated amount. In addition, the budget is usually passed in December (or even January of the ensuing fiscal year), while the LGUs need to prepare their fiscal plans by September, without the knowledge of the exact amount of their IRA allocation for the next year. Issues--Unintended and Intended Consequences of Devolution 3.1 1 As a result, there are at least six consequences that should concern the GOP: * Total transfers to the LGUs greatly exceed the cost of devolved functions, thus providing a large iiumber of LGUs with windfall gains of additional discretionary income and jeopardizing the NG budget; * The IRA allocations to individual LGUs are not based on their need or capacity, and created winnler-s and losers among LGUs; * Regarding sector spending, LGUs are allocating more resources on social services at the expense of economic services, maybe because the NG has not established a clear role for the l GUs and continues to fund some devolved economic services; * Devolution has made it difficult for the NG to provide LGUs with incentive to implement development projects that have high benefit spill-over effects or for NG to channel donor (loan) resources to devolved activities; - 44 - * Devolution is likely to have reduced local resource mobilization efforts; and * Devolution is likely to make the fiscal system more inequitable. 3.12 Fiscal Resource Gap at Macro Level At the aggregate level, the reform in the IRA transfer system devolved more revenues than the cost of the devolved functions. By 1994, when the devolution was fully implemented, the national government transfers to LGUs had grown to P 46.7 billion from about P 14-16 billion in 1990-91 immediately before inmplementation of devolution (see Background Paper Section 1), i.e., an increase of about P 30 billion (1.7 percent of GNP). Savings of the National Government from devolving functions to local governments has been estimated by the DBM to be P 6.3 billion in 1992 prices or about P 7.4 billion in 1994 prices. Net resource transfers, therefore, in 1994 were as much as P 22 billion ( 1.4 percent of GNP). Furthermore, NG is gradually shifting its tax structure away from international trade taxes (lowering of import tariffs) and rmloving towards increasing internal taxes through, for example, an expanded VAT. Such a move will naturally raise the share of internal revenue, automatically pushing up the total amount of the IRA allocation in the future relative to the other expenditure items. 3.13 IRA Allocation Imbalances at Individual LGUs The IRA allocation formula has no direct linkage to the amount of incremental fiscal resources LGUs need to finance newly devolved functions. LGUs have Mhus not necessarily received additional IRA transfers that are commensurate with their increased responsibilities--some have received less resources than they need to perform the functions assigned to them, and others have received much more than they need. The new IRA formula is highly generous to cities (Table 3-1): 60 cities share 23 percent of the total transfer whereas a similar amount is shared by 77 provinces and 34 percent by some 1,500 municipalities. In addition, because the cities already provided a high level of services on their own before devolution, the cost of functions additionally devolved to themii was only P 500 million as opposed to provinces whose additional responsibilities cost them P 3 billion. The transfer of resources that are not tied to devolved functions is, on average, P 100 million for a city, while an average municipality receives P 4 million, and a province receives P 39 million. There is thus no net resource loser among cities. But more than 100 municipalities and 27 provinces need to pay more for devolved functions than they receive additional IRA allocations. Provinces are particularly disadvantaged because of the narrow tax bases and the high recurrent costs imposed on them as a result of higher than average facilities and personnel devolved. Thlese are issues because there are no mechanisms that would allow the NG to compensate these LGUs who incurred net losses from devolution. In the short run, residents of these LGUs may have to suffer less adequate 3 levels of public services 2 Since other NALGU funds were allocated arbitrarily, only IRA allocation is compared here. 3 But. various adjustments are possible in the medium run; there are anecdotal evidences that some cities are already providing health services for residents in the surrounding municipalitics. - 45 - TABLE 3-1: NET TRANSFERS TO LGUS IN 1993 (NOMINAL MILLION PESOS) Incremental IRA Cost of devolved Net transfers of Net Transfers per functions untied resources LGU Provlcc, 6046 3008 3037 39 Citie" 6446 494 5952 99 Municil"ditics 8885 3 124 5761 4 5710 0 5710 0.1 Source: DBM 3. 14 Sector Spending Total government (NG and LGUs) expenditures on health, educationl, agriculture, road transport, the sectors expected to have been most affected by devolutioll, in fact did not change significantly between 1991 and 1993 (Table 3-2). At the local governmiilenit levels, social sector expenditures increased significantly from 0.29 percent of GNP in 199 1 to 0.71 percent in 1993. But economic services remain rather stagnant (Table 3-3). The same table shows a decline in the NG spending in social and economic sectors in 1993, and the decline was larger than implied by devolution; the cost of devolving some agriculture and natural resource management functions was about P 1.2 billion or 0. 1 percent of GNP; on social services, the cost of devolving health and social welfare functions was estimated at e 4.9 billion or 0.3 percent of GNP. The decline in NG spending appear to involve more general expenditure cuts and a shift towards expenditures on general services 'able 3-2: Consolidated Expenditures of National and Local Governmients: Selected Items (percent of GNP) 1991 1992 1993 Education 2.67 2.44 2.60 Agriculture 0.45 0.55 0.47 Health 0.79 0.74 0.77 Roads and Trausportation 1.76 1.90 1.93 Source: DBM, Commission on Audit. mission estimates - 46 - Table 3-3: Sectoral Allocation of National and Local Government Expenditures (percent of GNP) 1986 1990 1991 1992 1993 1. National governimient Economic services 4.71 4.79 4.66 4.34 3.40 Social services 3.52 4.54 5.14 5.03 3.38 General public services 1.12 1.28 0.86 2.0(2 2.51 2. Local governments Economic services 0.41 0.54 0.67 0.48 0.67 Social services 0.42 0.27 0.29 0.38 0.71 General public services 0.51 0.78 0.82 0.92 1.08 3. Consolidated: 1 +2 Economic services 5.12 5.33 5.33 4.83 4.07 Social services 3.94 4.80 5.43 54(1 4.09 General public seivices 1.64 2.06 1 68 2 94 2.91 Source: Commission on Audit, Department of Budget and Management, and the National Statistical Coordination Board; compiled by mission. 3.15 At all levels of local governments, social expenditures increased impressively with devolution (Table 3-4). As expected, health expenditures, which have been devolved, increased dramatically with higher local governmnent resources. While the rise in healtil spending for provinces and municipalities may be viewed as imposed because so much of additional health responsibilities have been devolved to them, at least during the early years of the devolution, cities are voluntarily setting aside more resources for health services. More surprisingly, at all levels of LGUs, education expenditures (which have not been devolved) not only increased as a percent of GNP but also as a share of the higher LGU spending. Without being forced to do so, the LGUs are choosing to spend a substantial amount of additional resources on education whenever they can--over and above the amount earmarked under the Special Education Fund (SEF). The underlying simple elasticity of education expenditures is 9 for provinces and about 3 for cities and municipalities. This means for every I percent increase in total resources available for a LGU (in real terms), provinces tend to spend 9 percent more on education, and cities and - 47 - munuicipalities spend 3 percent higher real resources on education. In the education sector where the role of the center is well-defined and its financial limitation well-known. LGUs are willing to provide additional resources to complement NG-provided services. 3. 16 Economic services (including rural roads, water, irrigatioin and agricultural extensioin) are taking the brunt of adjustment. Provincial spending has actually declined; when real resource availability goes up by I percent, provinces reduce real spending on economic services by I percent because so much additional spending is needed for health. For municipalities, the elasticity is very close to zero. Spending on general services seem to be the most stable or inflexible--with elasticity at around 0.7 for all LGU levels. 3.17 The shrinking share of spending on economic services in the LGU expenditures can be attributed, at least in part, to the willingness of the central governmllelnt to provide resources for devolved functions. Some central departments in the Executive contilue to fund devolved functions, and Congress still votes for appropriations of funds for capital projects that fall within the responsibility of the local governments such as local roads and irrigation. These latter expenditures are not without future costs to the LGUs' finances and accountability. Maintenance and operation will become recurrent local obligations competilg with other demands on LGU budgets. Such central interventions in devolved activities make it difficult to hold the local authorities accountable for their actions with respect to devolved activities. It may also distort the investmenit decisions of local authorities as they tend to wait for Congress to act. While data to support this statement is still limited, interviews with a large number of LGU officials have repeatedly made it clear that they would not provide services that central officials are willing to provide--especially services that substitute local provision--be it irrigation, forestry, or other infrastructure facilities. - 48 - Table 3-4: Expenditures by Sector, All Local Government Units: 1991, 1993 (percent of GNP) Particulars 1991 1993 Provinces Municipalities Cities Provinces Municipalities Cities General public services 0.19 0.43 0.22 0.21 0.59 1t.30 Social seivices 0.10 0.07 (0 11 0.23 0.26 0.23 Healiti 0.02 0.01 0.05 0.13 0.10 0.06 Education 0.01 0.03 0.03 0.03 (.(9 (0.07 Ecoonomic services 0.24 0.23 0.21 ().18 0.24 0.2 5 OtLlieis 0.02 0.03 0.03 0.03 0.05 0.08 Source: Commission of Audit, National Statistics Coordination Board 3. 18 Project Implementation Another serious problem is that project implementation has become difficult in devolved areas. This is partially because LGUs caninot beniefit from borrowing from outside sources as the NG cannot guarantee LGUs' external borrowing. This also constrains LGUs' access to the expertise and institutional building such projects often provide. At the same time, due to the unconditional nature of the block grant, the NG has not heeln able to provide local governments with effective incentives to implement activities of national priorities, such as poverty alleviation or environmental protection. An associated problemii is that it has become difficult to channel donors' loan resources to local governments in the devolved sectors. The NG is often willing to be a conduit and "on-lend" these resources to local governimienits, hut local governments consider donor projects, especially in areas where full cost recovery is difficult, as primiiarily NG responsibilities and demand that the NG on-grant these resources. The NG on the other hand is averse to on-granting donor loans to local governments claiming that NG has already provided them with more than adequate financial resources throughi the IRA allocationi. The resulting gridlock has jeopardized preparation and implementation of maniy projects linanced by loans from donors. 3.19 Impact of Devolution on Local Revenue Mobilization Local revenue efforts appear to have declined with higher IRA resources. Tables 3-5 confirmis tllat in 1992. when the IRA allocation started to increase significantly, the collection rate (collection as a share of what was due) of the real property tax came down about 10 percentage points for all LGUs, and all provinces and cities. In addition, a detailed econometric study discussed in the Background Paper Section 11 confirms that the increased IRA allocations have had negative and significant effects on local government revenues across all levels of LGUs in 1992 and 1993. - 49 - 3.20 Disturbing as it may be, this is rather expected from the unconditional grant system that the Philippines opted to implement. Residents of LGUs, when their income is augmented by the additional transfers from the NG, naturally desire to enjoy both increased public service provision and increased consumption of private goods--the latter comes from paying lower taxes (i.e., it may be that the tax collection efforts become lax and default rates rise, rather than tax rates goling down). The NG needs to decide if it should give more incentives for further local resource mobilization through, for example, an IRA transfer system that is conditioned on the local revenue effort. Table 3-5: Collection Rates of Real Property Tax (percent) All LGUs Provinces Cities 1986 51.3 49.6 53.4 1987 52.8 49.5 56.7 1988 54.3 49.4 60.7 1989 58.0 55.6 61.0 1990 57.8 53.6 63.3 1991 58.9 54.1 65.1 1992 49.7 44.3 56.4 Source: Background Paper Section 11 4 3.21 Devolution and Equitv According to an earlier study, new IRA allocations have not contributed to any of the equity objectives. These objectives include: provision of more resources to the poor or poorer LGUs; provision of resources according to the fiscal needs; or provision of resources to equalize the level of supply of critical services across the country. The only significant coincidental relationship found in this study is that the IRA allocation increases with a LGU's tax base per capita or its recurrent expenditures. IRA resources are thus being channeled primarily to those LGUs who already have more fiscal resources than others. Efforts to shift the IRA allocation weights away from the population and more towards land area has yet to show significant effects on poverty alleviation. 3.22 Detailed data on LGU spending that could help analyze the impact of devolution on equity are still sketchy. Conjectures, however, can be made that devolution as currently implemented may increase spending in more affluent areas. On health, NG allocations to regions were cut severely from an average of P 98 per capita in 1991 to P 25 in 1993 and to P 22 per capita in 1994 (all in 1991 prices). This reduction in NG expenditures appears to be accompanied by a shlift of the allocation pattern from progressive (poorer region received more resources per 4 Internal Revenue Allotment System of the Philippines: Trends. Patterns, and Effects, Local Development Assistance Program (GOP/USAID), 26 July, 1994. - 50 - capita) to regressive. At the same time, because the IRA allocation was not designed to provide more funds to poorer jurisdictions, the spending was each according to its means and preferences. Table 3-6 summarizes the result of a survey of 172 LGUs. The increase in health expenditures per capita in 1991 prices ranges from e 70 in a city in a high income province to P 50 in a municipality in a low income province. The average real decline in NG's per capita health expenditures can be offset by the increase in LGU spending in a more affluent area, but is not enough for poorer areas. If the reduction in NG spending was more for poorer regions, the situation would be worse. A similar situation appears to be taking place in education--although education has not been devolved, under-funding of the public school system from the NG is leading to selective defacto decentralization of the sector. Again, winners among LGUs have resources available for extra support of public schools, and voluntarily supplement teacher salaries, hire extra teachers and provide additional supplies. Given the importance of equity objectives for the GOP, it is critical that more detailed information is compiled to make further analysis of the equity impact of devolution. Table 3-6: Total, Social and Health Expenditures Per Capita by LGUs 1991 and 1993, in nominal pesos LGU (Number of obs.) 1991 1993 Total Social Health Total Social llealth Increase in health spending in 1991 prices 1. Provinces (36) 96 6 3 133 40 32 24 HiglI lncome(12) 90 7 4 139 53 43 31 Middle Income(12) 109 6 3 115 33 26 9 Low Income (12) 81 4 2 168 40 34 26 2. Cities (19) 483 105 15 708 186 63 39 3. Municipalities 6 (1 17) 170 I5 I 287 54 25 20 Source: Background Paper Section Table Al-12 Thcrc is no income or expenditurc data by city or municipality. Therc are income classcs (from first to sixth) of citics and municipalities but thcy are based on the size of tax base rather than incomc per capita, and do not rcflcci accurately how rich or poor residents of a particular jurisdiction are. - 51 - Principles for Improvements 3.23 Improving the way devolution works should be based on the following three principles: 3.24 Define the Roles of Local Governments and the NG. The general rule is that local authorities should be able to provide the quantities and qualities of public goods and services that their residents prefer and are willing to pay for. This is the principal rationale for the devolution of fiscal power to LGUs. 3.25 The NG role is to be instrumental in the provision of services that are socially necessary according to the national priority, but are under-funded by LGUs. For example residents in a local jurisdiction can be unwilling to fully fund programs, either because benefits and costs of certain services will be shared by residents of other jurisdictions (e.g., environmental protection, infant inoculation), or because residents are not aware of the full extent of benefits (e.g., early childhood care). The NG may also decide that every citizen should have an access to a miniiimum level of certain essential services. Disparities in tax capacity among local authorities result in large variations in the tax costs of providing local services. Given that per capita cost of services is likely to be similar in both poorer and richer areas, taxes per capita would be higher in the former than the latter for the same level of services. 3.26 Define NG Incentive for LGUs Where Necessary Among the services that are of national priority, the NG can delegate the service provision to LGUs when it is more efficient to organiize the service provision at the local govermnent levels. The NG can provide LGUs with additional incentives for increased provision of these services. Alternatively, given that the size of the IRA allocation exceeds the cost of devolved functions by a large margin, minimum provision of these critical services could be mandated without further incentives as long as the NG ensures that LGUs with weaker fiscal capability can provide these minimum services. Thus, the NG needs to make a number of fundamental decisions: 1. agree on precisely defined services that, in the national interest, the NG needs to insist that local governments provide; 2. identify the size of the fiscal gap between the expenditures of LGUs necessary to fulfill their local and national responsibilities and their own potential revenues; 3. decide whether and how the NG should help meet, at least in part, the fiscal gaps. The NG could aim to compensate, in part, for the differences in fiscal capacity (per capita tax base) and/or tax efforts (tax collection applied to the tax base) to meet their expenditure responisibilities. Alternatively, the NG could share the LGUs' costs, at least in part, only when they perform certain designated national interest functions ; and 4. assuming appropriate conditions attached to NG transfers to LGUs have been designed, decide if the projected amounts of transfers are likely to be within the fiscal competence of the NG. 3.27 Establish Hard Budget Constraints Once the respective roles of the NG and LGUs have been established, and changes in the IRA transfer mechanism are, as needed, designed, there - 52 - should not be any exceptions to the rule. The NG should not carry out activities chosen as LGU functions, nor should the NG provide more resources than agreed in the design of the resource transfer mechanism. In other words, LGUs should face a strictly hard budget constraint and be held accountable for their action. Similarly, LGUs can borrow from domestic financial institutions (so long as their debt service payments are below 20 percent of the IRA allocation they receive). It should be clear that, while the NG monitors such borrowing carefully, it should be a commercial transaction between LGUs and banks based on LGUs' creditworthiness Policy Direction and Recommendations 3.28 Since the devolution process has already started, it will be more difficult to address these questions. In this section, we discuss two sets of improvements the NG could make to the process of devolution given what has already been accomplished. First, the management of the existing system can be improved by inter alia making LGUs fully accountable for devolved activities, encouraging cost recovery, and providing management assistance to LGUs. Second, if the NG decides to reform the LGU transfer system, it can consider a set of mieasures by moving away from the concept of totally unconditional IRA grants. 3.29 Accountability It is first recommended that transfer payments to LGUs be placed on a statutory basis that would make the payments independent of the anniual appropriations of the C(ongress. This is because the NG must minimize the uncertainty associated with the amounts and timing of IRA grant allocation to each LGU to make LGUs more accountable for devolved functions. 3.30 For the same reason of increasing LGUs' accountability, the Executive branch and the Congress should not be permitted to propose and appropriate funds for projects and programs that fall within the responsibilities of the LGUs except for areas where NG provides incentives to LOIUs for selected service provision (e.g., matching grants) or other mutually agreed programs. Despite budgetary constraints, Congress and some executive officials refuse to let go of the activities devolved to the LGUs. The end result is that the national governmllent continues to appropriate funds for activities that have been devolved to local governmenits despite the mounting evidence that LGUs in aggregate may have more resources at their commllalnd than the national government. Some local government officials have expressed the view that for as long as the national government continues to appropriate funds for these activities, they will continue to allocate local resources away from them. 3.31 Finally, the NG needs to make clear to all LGUs that they cannot expect a bail-out when they get into debt service difficulties from voluntary borrowing. The borrowing by LGUs should be monitored carefully by the NG, and the mandated cap on debt service (20 percent of total resources of each LGU) should be strictly enforced. 3.32 Cost Recovery The rule should be adopted that those who enjoy the benefits flowing from public investments should pay for them. NG capital funding of local government utilities and other infrastructure projects from which it is feasible to exclude those who do not pay for their services, should be contingent upon the adoption of full cost pricing. NG's cost sharing of activities when direct cost recovery is difficult (for example, local roads) should be contingent - 53 - on improved property tax administration so that increased property values due to infrastructure investmenits, be captured by tax net. 3.33 Technical Assistance At their request, the NG should give management/technical assistance to LGUs related to their problems in qualifying for NG grants. There should be no implication that the "advice" is. in some sense, binding on the LGUs. However, it would be prudent for the NG to retain responsibility for the definition of each acceptable local tax base, including the real property tax base and the business tax base, and the structure of rates that can be applied to those bases. Setting the tax rates applied to those bases should, however, be a local prerogative. Retaining NG control of the definitions of tax bases and the structures of rates is necessary to ensure that the fiscal capacity estimates are comparable among LGUs of the same type and to prevent the adoption of beggar my neighbor revenue raising practices by LGUs. 3.34 One way to improve the implementation of the existing transfer system is to improve population statistics within LGU jurisdiction. The current estimates of the size of population are extrapolated from the last census data. These projections tend to suffer from significant errors in a dynamic economy in which the amount of inter-regional migration is non- trivial. Improved population statistics at local government levels is thus critical for better functioning of the existing system. 3.35 Reform of Intergoveriimenltal Transfers If the govermnent is to modify the transfer system, significant benefits would be realized, by replacing the present unconiditional IRA grant system., with a grant system that combined unconditional revenue deficiency grants, (sometimes known as equalization grants) and conditional cost sharing programs for national priority activities. 3.36 The unconditionial revenue deficiency grants would be made only to LGUs with per capita tax bases below the national average with respect to a group of standardized taxes. These LGUs would receive grants equal to the revenues they would have collected had they applied their cffective tax rates to the national average per capita tax base rather than to their actual per capita tax bases multiplied by their populations. Note that these revenue deficiency grants are unconiditionial in the sense that the NG does not direct the uses of the funds received by the LGUs. The NG. in consultation with the LGUIs, would define a standard bundle of local taxes to be included in the calculations. 3.37 With revenue deficiency grants in place, LGUs that set low nominal tax rates or had poor collection procedures would receive low per capita revenue deficiency grants. For poor [-GUs (i.e. those with below national average per capita tax bases), the exercise of greater fiscal effort [defined as the nominal tax rate applied to each of the tax bases included in the standard] and of greater diligenice in tax adminiistration would be rewarded by larger revenue deficiency grants, and conversely. 3.38 The cost sharing programs would encompass a number of activities that are the responsibility of the LGUs to which the NG attaches a high priority (e.g. preventative health care, basic education, basic sanitation). One possible method of proceeding can be described as follows: Idirst. carefully define the eligible programs in terms of the nature and quality of the outputs and of the conditions of access of local residents to these outputs. Donor-assisted loan projects should be evaluaLed with the same criteria that are applied to any other project. Second, only the necessary - 54 - expenses incurred by LGUs in producing and distribution these outputs would be subject to cost sharing. Third, the NG would monitor the performance of LGUs in meeting the output and access requirements at minimum cost, and LGU programs that failed to meet these requirements would not be eligible for future cost sharing. Fourth, the actual per capita expenditure of each LGU on each shared cost program in the base period (an average over a few recent years) would be determined, and the NG would grant to each of the l GUs with respect to each of its shared cost programs an amount equal to a fraction of the increase in its per capita expenditures on that programs relative to the base period multiplied by its population. Finally, the grant fraction would be highest for the LGUs with the lowest per capita revenue raising capacity (i.e. lowest per capita standard tax bases) and the grant fraction would be lowest for the LGUs with the highest per capita revenue raising capacity. 3.39 The Department of Health has a head start in this effort of cost sharing/matching grants and some lessons may be learned from their initial health care agreements with selected local communities. The ability of the DOH to initiate such a program was due to the additional budget given to the department in 1993 and through higher than expected grants from external sources. In this respect, it may not be a good model since it is based on unstable sources of financing. A desirable characteristic of a sound matching grant mechanism is that it should have a predictable and reliable financial source. If one such mechanism is to be designed, one likely source would be a fixed proportion of the IRA grant to LGUs, say 10 percent of the IRA base or 25 percent of the IRA. In the design of the matching grant mechanism, it should take into account the following considerations: (a) provinces have relatively less resources for counterpart funding than cities and municipalities; and (b) poorer communities are less able to raise the necessary counterpart fund partly because of their narrow tax bases and the general unwillingness of local officials to impose added burden on their depressed communities. 3.40 The total cost of these two programs should not exceed the amounts to which LGU are entitled by devolution. Perhaps, for example, the total cost of revenue deficiency grants should not exceed, say 50 percent of the present IRA, and a similar limitation should be placed on the total cost of the shared cost grants. With the improvement in national tax adminiistration, the NG might reduce its shares of some national priority shared cost programs and offer to collect a personal income tax for each province defined as a surcharge on the NG's personal income tax. 3.41 Alternatives--Unconditional Grant and Unfunded mandate Should these changes be judged to be too drastic for political reasons, the NG should at least mandate that entitlement to IRA grants would be conditional on the adequate supply by each LGU of a small set of essential service programs (immunization, basic health care, safe water, etc.). Compliance by LGUs would be monitored by the NG and inadequate performance would be penalized hy a reduction in IRA grant allocation in accordance with a prescribed formula. 4. CIVIL SERVICE REFORM Introduction 4.1 The public sector's role in the Philippines is undergoing substantial change because of the government's recent reforms in the direction of liberalization, privatization and devolution. This change provides a unique opportunity for carefully redefining the roles and priorities of the civil service. The government has seized this opportunity by starting a process of civil service reform called "streamlining the bureaucracy" in order to improve the performance of the central administration. The purpose of this chapter is to help the government formulate its reform strategy in a way which reflects the lessons of other countries and of the World Bank in this difficult area of civil service reform. The chapter begins with a brief summary of the Philippine civil service and current reform initiatives. With this as background, it goes on to examine the rationale, objectives and strategies for civil service reform. A detailed description of the Philippine civil service and the ongoing Government's efforts to streamline the bureaucracy is in Background Paper Section IV. 4.1 It has been found that the existing GOP approach is focused on "reorganizing" the GOP structure and does not emphasize: reform of financial incentives, desired skill mix, and personnel management reforms. That the Department of Budget and Management (DBM) is leading the GOP efforts but Civil Service Commission (CSC) has a separate mandate complicates the process. The chapter thus recommends: (a) coordination of the GOP efforts should be made with strong political commitment from the highest level of leadership; (b) a reform of incentive system should be implemented to achieve and maintain the desired skills mix, which will probably include a higher proportion of professional staff. Such a reform would allow higher-level civil servants to be remunerated at rates more compatible with their private sector counterparts; (c) to achieve the desired skill mix efficiently, a combination of voluntary retirement and targeted retrenchment of surplus staff would complement the reform of the incentive; (d) a personllel management reform based on a review of the current allocation of management tasks amonig central agencies would also improve the civil service. Background analysis (a) Size, salaries and cost 4.2 Definition and Size According to the most recent count, there are 1.447 million government employees in the Philippines. Of these, 20.3 percent work for local government, and 8.8 percent for government owned and controlled corporations. The remaining 70.9 percent constitute the national government civil service. Of these 1.025 million, some 25 percent are uniformed (either military or police). The analytical focus of this chapter is on the non-uniformed national government civil service, a group of .773 million. Information on civil service size trends in the Philippines is available only for the number of permanent, authorized positions, rather than for filled positions (whether permanent or not). Using this as the basis for measurement, one observes a massive increase under Marcos from .43 million in 1971 to .84 in 1980 and then to almost I millioni in 1985. There was a further increase under Aquino frorn .95 million in 1985 to 1.12 million in 1992. Since then the size of the civil service has fallen somewhat to its current level - 56 - of 1.01 million. However, some two-thirds of this fall is directly attributable to decentralization and so will have been matched by a concurrent rise in local government personnel. Another one-quarter of the fall is due to cuts in the Department of National Defense. Hence one should view the Ramos period to date as, apart from the military, one of civil service containmenit rather than contraction. 4.3 Major Government Departments The civil service of the Philippines is very concentrated in four agencies: 43.7 percent of all civil servants, mainly teachers, are in the Department of Education, Culture and Sports; 13.7 percent are in the Department of National Defense (the military); 12.5 percent in the Department of the Interior and Local Governimlenit (DILG, mainly police); and 4.6 percent in the State Universities and Colleges (tertiary educators). 4.4 Salaries The total salary paid to civil servants is made up of the basic salary and various allowances. On average, the basic salary makes up about 70 percent of the total salary. However, at the lower grades, this falls to 60 percent, in(dicating that the distribution of allowances favors the lower-paid. This is because most, though not all, allowances are of a lump-sum nature. Excluding the President and Vice-President, the civilian civil service is divided into 33 salary grades. There is a heavy concentration of posts (55.5 percent of the total) in the three grades 10- 12. Grade 10, which includes many teachers, by itself contains 31.8 percent of all positionls. Grade 11 contains 13.1 and Grade 12 10.5 percent. 4.5 Changes to the Salary Structure The salary structure is currently undergoing major change as the result of the ongoing implementation of the Salary Standardization Law 11 (SSL 2), which re-sets basic salary levels for all grades. The law authorizing SSL2. passed in 1993, gives 1997 as the date by which SSL2 is to be fully implemented. If and wheni it is fully implemented, SSL2 will have increased the average basic salary by 158 percent (relative to 1993, the last year prior to implementation, and using a weighted average based on 1993 positions). The uniformed sector is treated even more generously under SSL2, with an average salary increase of 189 percent. The ambitiousness of the SSL2 targets is confirmed by simulations conducted for this report. These indicate that even with high rates of growth (7 percent per annum), witlh no staff cuts, SSL2 targets could onily be met by 2000 given the Government's wage bill target of 5.6 percent of GNP. Moreover, if SSL2 were to be met on time (1997), a massive 26 percent cut in staffing would be required if the wage bill constraint was to be adhered to, and that is assuminig rapid growth. 4.6 SSL2 will also greatly change the salary structure, by increasing the salaries of the lower-levels by much more than those of the higher-levels. Although the SSL2 target remains a long way off, the mode of its partial implementation to date, which is one of equal absolute increments to all grades, has already made a big difference to the salary structure. The flattening this is causing can be seen in the summary statistics of dispersion, such as the compression ratio, which gives the ratio of the highest-paid to the lowest-paid. The compression ratio for total salaries fell from 6.9 in 1993 to 5.7 in 1994 and will fall again to 4.7 in 1995. That for basic salaries fell from 8.8 in 1993 to 6.7 in 1994 and will fall again to 5.2 next year. Fully implementing SSL2 would halve the compression ratio for the basic salary from 8.8 in 1993 to 4.4 in the year of implemenitation; the compression ratio for the total salary would also fall. from 6.9 to 4.6. 4.7 Public vs. Private Sector Salaries This flattening of the compression ratio comes at a time when available evidence suggests higher-level staff in the civil service are significantly - 57 - underpaid relative to their private-sector counterparts, whereas in most cases the lower grades in the civil service receive salaries roughly equivalent to those of their private-sector counterparts. 4.8 The Wage Bill The wage bill is a large component of government expenditure. It currently stands at 31 percent of total expenditure. Moreover, the wage bill has been, until recently, on an upward trend. As a percentage of GNP, it rose from 3.8 percent in 1983 to 4.1 percent in 1985 to 5.8 percent in 1990. It has since declined slightly to 5.4 percent. Overall (end- to-end) growth in the personnel bill between 1983 and 1993 in the real salary bill was 58 percent, compared to real GNP growth over the these years of only 16 percent. This rapid growth in the salary bill is projected to continue. Despite the much healthier GNP growth predicted (around 7 percent a year), the salary bill is still projected to grow 50 percent faster than GNP, a result of the SSL2 targets. 4.9 International Comparisons The size of the civil service in the Philippines is average for the Asia region. However, international comparisons of size, while useful in providing context, cannot be used as a basis for policy, mainly because they tell us very little about civil service efficiency. To illustrate, note that the Singaporean and Philippine civil services appear to be quite similar on the basis of the sort of figures used in international comparisons. Both countries have a ratio of civil servants to total population of around 2.1-2.2 percent and both have a wage bill which costs 30 percent of total government expenditure and around 5.5 percent of GDP. No-one could sensibly argue on this basis, however, that the two civil services were similar in terms of performance and the need for reform. (b) Management issues 4.10 Management Agencies There are two main central personnel agencies. The Civil Service Commission (CSC), a constitutional body, has wide-ranging powers, exceeding those of most CSCs around the world. Its role in the Philippines has expanded in recent years, and the CSC now not only ensures adherence to CS laws and regulations, but also controls recruitment and promotes efficiency, performance evaluation and the professionalization of the civil service. The other central personnel agency is the Department of Budget and Management (DBM), the executive agency responsible for the budget. It controls the number of positions available, and also grades positions and advises on salaries and allowances. This division of responsibilities between CSC and DBM gives rise to several difficulties. For one, it can be a source of delay. Say an agency wants to create and then fill a new position. First it must clear the creation of the nrw position with DBM. Then, if external recruitment is sought, this must be cleared with the CSC. Finally, assuming appointment goes ahead, the agency must once again clear its intended appointee with CSC. Clearance with both agencies can be a slow process and may take up to six months or longer to complete. l. i I Promotion and Performance Review Advancement within grades is based purely on seniority. There are eight steps within each grade and movement to the next step is automatic after three years. Promotion to a higher grade is, at least in principle, based on merit. Recent reforms have given more emphasis to performance. The CSC has recently introduced a six-monthly performance evaluation system. The precise manner in which the evaluation is carried out is left up to each agency to decide. These evaluations are intended to provide a basis for promotion and dismissal decisions. (For example, two verdicts of unsatisfactory performance will result in - 58 - dismissal.) Performance Incentive Benefits have also been introduced to reward top performers. Agency heads are to nominate annually their best performers to share in a fund worth some 2% of the total personnel budget. 4.12 Exit Although attrition rates have only become available since the passing of the Attrition Law, they are widely held to be low and stable in the aggregate. The recent figures certainly confirm this. At the national government civilian level, 1993 figures give an attrition rate of just 1.3 percent. The most common cause of exit is retirement or death. Retrenchment is rare. Entrance to the public sector from graduation followed by mid-career transfer to the private sector is a well-known career path in the Philippines, as it is in many other countries. Demands for the skills of professionals from the private sector will rise with liberalization, and exit can be expected to increase. - 59 - (c) Reform initiatives 4.13 A historical Perspective Civil service reform in the Philippines has a long but unhappy history. The Marcos era, after a promising beginning, left an legacy of corruption, politicization and over-staffing, which still plagues the public sector. The Aquino Government set out on a reform program to fundamentally alter the nature of the civil service. However, while it was successful in establishing institutions to guard against the future abuse of government, such as the Office of the Ombudsman, its reform efforts achieved little systemic change to the way in whichi the public sector operated. 4. 14 Reform under Ramos President Ramos has made civil service reform a key objective of his administration: according to his speech at the 1992 Economic Summit, one of eight goals which would guide his presidency. The most important current initiative is the Governrnent's "Streamlining of Bureaucracy" plan. A Reorganization Bill has been submitted to Congress. If passed, it would give the President authority to re-organize the civil service via a series of Executive Orders which would reduce the overlap between different spheres of government, and attempt to check the ad hoc way in which government has grown. At the time of writing, the passage of the Act seems unlikely before the adjournmuent of Congress before elections scheduled for May. However, the administration is confident that the new Congress will act quickly to pass the Act. In the meantime, DBM, the agency with responsibility for this initiative, is undertaking preparatory studies. It is also encouraging individual agencies to go ahead with restructuring initiatives on their own rather than wait for legislation to be passed. Box I Civil Service reform in Individual Agencies: the Case of the Central Bank Following the act reconstituting the Bank in 1992, BSP Management appointed first a Task Force and then a Study Team to address questions of intemal management. Though much remains to be improved, there are positive signs. Eight departments have been abolished. And the total number of layers in the hierarchy has been reduced from eleven to eight. Staff size has also fallen: from 6,170 in 1992 to 5,570 in 1993 to 4,520 in 1994. This was achieved by a generous voluntary retirement program (on which, see the text). The BSP has also been given greater control over its salary structure, though the initial plan to give it full control was watered down so that it now only sets the salaries of grades 19 and above (the grading system has not been chlaniged and the Bank still adheres to CSC qualification requirements). Salaries for these relatively senior positions ,have been upgraded on the basis of a comparison with private-sector banking salaries. Though BSP salaries are still below private sector levels, the gap has been substantially reduced. Since the gap was found to be the highest at the execultive level, salaries at this level have been increased the most: by between 120 and 220 percent. The difference between the BSP and the rest of the civil service is most vividly illustrated by consideration of the Governor's salary. The Governor of the Central Bank receives iR 150,000 a month, about half what a secretary of a department earns in a year. Despite these pay rises, total personnel costs have fallen as a result of the massive staff cuts: by 12 percent in nominal terrns between 1992 and 1993. - 60 - 4.15 Civil service reform has made the most progress in two government agencies regarded as pivotal to economic recovery and therefore given increased autonomy under special legislation: the new Central Bank of the Philippines (BSP) and the National Power Corporation. The BSP's use of this greater independence is detailed in Box 1 and is illuminating as an example of wlhat could be achieved in relation to the civil service as a whole. 4.16 The other major area of reform activity has been efforts to check civil service growth. The Attrition Law, passed in 1992, forbids external filling of vacated CS positions, except in exempt sectors, such as education and local government. Alternative, more flexible ways of controlling civil service size are also being explored. DBM has introduced wage bill targets (5.6 percent of GNP), and early retirement is also being considered. Ratioiiale and objectives 4.17 Lessons of Experience During the Marcos and Aquino eras, the size of the public sector grew dramatically, yet its effectiveness and credibility were severely undermined. The present civil service, while not growing, still suffers, at a time when the Government is launching into an economic transition based on outwardness and competition, from the endemic problems of the past. Drawing both on the (largely unsuccessful) experience of civil service reform in the Philippines and also on international experience, we would suggest that the following elements are needed for a successful CS reform: * a precise and convincing rationale; * clear objectives; * a strong and continuing political commitment to reform on the part of the President, manifest in both the institutionalization of the reform process and his continuing personal involvement; and * a coherent and phased strategy which begins with the objectives and considers tradeoffs between theml. 4. 18 In this section, we consider the rationale for civil service reform and the objectives of reform. In the next section, we consider the questions of political commitment and strategy. (a) Rationale 4.19 There are three fundamental reasons for undertaking a reform of the Philippine civil service: efficiency; the changing role of government; and macroeconomic constraints. - 61 - Box 2 An Audit of DECS In 1993, a privately-funded study was undertaken by the Centre for Corporate Citizenship into management practices of the Department of Education, Culture and Sport, the department with the largest budget and the most employees. The study investigated a large number of business processes undertaken within DECS. It found(I problems across the board, a sample of which are presented below. .. Technology and skills of manpower in information gathering, storage, processing and retrieval are inadequate ... One explanation for this is the inability of DECS to recruit and retain the most qualified personnel for various reasons, [such as] low compensation and the outdated information network. (p.7) ... [Tihe Planning Service of Central Office receives only an average of 60 percent of enrollment data required. [The result is] ... some areas of the country have sufficient or even excess classrooms while many others suffer chronic shortages. (p.8) ... Some schools being approved for construction are not in the priority list submitted by the elementary schools. This happens because of "congressional insertions", while the local government officials also "participate" and influence the final selection of school beneficiaries. This results in having some schools with totally dilapidated classrooms while other schools have relatively new classrooms. (p. 10) ... [Diuring the field work, all turned-over new classrooms and buildings supervised by DPWH in the schools visited were incomplete ... The quality of work done by contractors ... supervised by DPWH was inferior ... if compared to the work done by the contractors supervised by DECS/School Heads. (p. 11) ... [The allocation of additional positions] required 14 signatories from the school level up to the regional level (Signatory School -2, District Office -2, Division Office -4, and Regional Office -6). (p. 12) ... JAIpproval of [authorized positions] by the Central Offices (DECS and DBM) takes about 6 to 12 months. (p. 15) ..1 Illiplementing] salary increases resulting from promotions usually [takes] ... several years. (p. 15) ... Budget implementation is too centralized at the Regional Office. [Tihey decide how much to give and to whom to give as they please ... This creates opportunities for misappropriating the budget. [Lless priority iieeds are allocated funds and ... funds are sometimes diverted for other purposes. ... Once the budget is approved, lit] loses the purpose for which it was prepared and becomes an exercise in futility. (p.20) ... [Tlwo of the three sample divisions were not able to deliver the textbooks before the opening of classes for [thel school year. (p.34) Source: Diagnostic Review and Systems Audit of the Department of Education, Culture and Sports, a confidenltial study by the Centre for Corporate Citizenship and SGV & Co. for DECS, 1994. 4.20 Improving Efficiency There is widespread agreement that the public sector is inefficiently managed. Three key sources of inefficiency can be identified, namely an unwieldy structure, red-tape and graft, and demoralization. For insight into how these manifest themselves in particular agencies, see Box 2, which extracts findings from a recent management audit of DECS. 4.21 The Changing Role of Government Economic reform, in the form of devolution, privatization and liberalization, is already well underway in the Philippines. All these imply a - 62 - different, narrower, though crucially important role for the civil service, one of management rather than control. The changes may also require adjustments to the skills mix and remuneration system. A deregulated, open and outward-oriented economy needs a professional corps of technicians to provide the public sector with the required highly specialized skills. It lias already become evident that certain economic, financial and legal skills are difficult to retain within tihe civil service because of the present incentive structures. If the economy really does begini to grow at a high rate anid in a sustained fashion, then how will the talent be retained or recruited to undertake, for examiple, highily technical supervisory tasks like the oversight of government corporations or of the finanicial and bankiig sector? In addition, if economic reforms are to be sustained and deepened and dlevelopment priorities articulated and met, then the civil service will need to be transformed into a higilly motivated professional force for change and development - as it is elsewhere in Asia, see Box 3 for examples. The momiientum for economic reform will not be sustained unless a comipetent, well motivated and professional civil service emerges to create the institutions and organizations which are necessary for steering a liberal economy. Box 3: Competent Bureaucracies and Rapid Economic Growth In Chapter 4 of The East Asian Miracle study by the World Bank, the authors focus on the specific institutional mechanisms which governments have used to foster shared economic growth and at the same time enhance their own legitimacy and effectiveness. The main mechanisms arc: wealth-sharing programs designed to include non-elites in economic growth; a cadic of cconomic technocrats insulated from narrow political pressures; institutions and mechanisms to share information and win the support of business elites. In discussing the characteristics of successful insulation of the technocrats, the following lessons are partiCularly pertinent to the Philippine situation: - the primacy given to the bureaucracy over the legislature in the drafting of laws and the considerable levcrage available to bureaucrats for minimizing political pressure on the approval process, particularly in the casc of thc annual budget. - the development of honest and competent civil services based on: competitive entry; promotion by merit; total compensation competitive with the private sector; ample rewards for the top echelons. The East Asian experience also shows that only a competent and honest civil service can create a credible and enforceable legal and regulatory environment conducive to rapid private sector investment and development. An important aspect of the relationship between the public and private sectors in the high growth East Asian economies is the development of formal institutions in order to ensurc that the rules of the game are ma(le transparent to all potential investors on an equal and shared footing. In many cases deliberation councils have served to create a forum to discuss policy, exchange information and arrive at a consensus on critical policy dircctions. Research on the determinants of effective councils clearly shows that they are not able to function without an efficient, competent and honest bureaucracy which can establish a level playing field for all lirilis active in the market. This in tnLrn ensures that profits and rents are made through open market competition hased on information-sharing and clear rules and not on the basis of special favors from I,O goTllnlelllt. 4.22 Macroeconomic Constraints Another key element of any rationale is the importance of the wage-bill as a macroeconomnic variable. The government faces a tight budgetary situation. - 63 - Civil service management needs to be reformed so that costs are contained without sacrificing the flexibility required to meet future needs. A central objective in this regard is the setting of firm targets as part of overall fiscal and financial management. 4.23 The present streamlining effort finds its rationale mainly in response to the first two factors discussed above, namely inefficiency and the changing role of government. More recently, DBM has begun to set wage-bill targets and it plans to undertake its salary and impact mitigation studies within the framework of the budgetary targets it has set. The aim ultimately is "to achieve and maintain a lean, efficient and effective national government bureaucracy" (Re-engineering proposal). This is a laudable, but also, in the light of past experience, difficult goal. Thle government will need to take every opportunity to link the rationale for CS reform to a vision of an economy which is undertaking a fundamental transition away from government control and protection towards competition, liberalization and outward-orientation. (b) Objectives 4.24 Juxtaposing this rationale against the CS status quo in the Philippines implies five key objectives for civil service reform: i) oni the basis of a redefinition of the role of the public sector and the civil service, restructure and reorganize departments and agencies; ii) identify and reconfigure as required the civil service skills mix; iii) improve incentives to attract and maintain the desired skills; iv) establish controls over the size and cost of the public service; and v) develop a more integrated system for managing personnel budgets, positions and individuals. 4.25 In the current streamlining initiative, some of these objectives have received more attention than others. The first has been key to the government's efforts. More recently, DBM has broadened its agenda to cover, to varying degrees, all five of the above objectives. This broadening of the government's agenda is a positive development, since all five objectives are not only important hut also interrelated. While it may not be possible to pursue all aspects of reform simultaneously, it is important that those responsible for the reform are able to articulate their overall goals and how these will be realized over time. Strategic issues 4.26 In this final section, we outline, in sub-sections (a) to (e) respectively, five key strategic issues which will arise in the pursuit of the objectives outlined in the previous section. The first is the basic question of how best to institutionalize the reform process. The second is the need for consistency and transparency in the way in which the goal of streamlining is pursued across the government. The third is the tradeoff between improving salaries and skills-mix, on the one hand, and the need to contain the wage bill on the other. The fourth is the question of controlling civil - 64 - service size while allowing for a change in the skills mix. The fifth and final issue is that of personnel management. (a) Political commitment and institutional arrangements 4.27 Any civil service reform effort will inevitably face a large numllber of political, management and technical challenges. For it to be successful, it is essential that the reform effort be backed at the highest political levels and that this backing be effectively institutionalized. Countries which have successfully carried out civil service reform have usually achieved it through the establishment of a strong, central unit equipped with the authority, independence and requisite technical expertise to plan and implement the process. This unit is usually located in the office of the Prime Minister or President and is seen to be operating under his authority. Malaysia's Administrative Modernization and Manpower Planning Unit (MAMPU). established to spearhead CS reform in the late seventies, provides a good example of what is required. 4.28 In the Philippines, a different route has been followed. As we have already detailed, the reform process is a presidential initiative (para. 14) and a Presidential Committee on1 Streamlining the Bureaucracy, whichi has representatives from a number of government agencies. has been established. An Interagenicy Technical Committee on1 Streamlining has also been established, which meets regularly to review the streamlining program. In addition, the Cabinet Committee on Development Administration - with representation from DBM, NEDA, the Presidenit's Office, CSC and the Commission on Audit - has an oversight role. However, the actual preparation, execution and implementation for the streamlining program remain the exclusive responsibility of the DBM. The choice of DBM as sole executing body does carry with it certain risks. DBM itself is an integral part of the bureaucracy, and some of its own procedures and arrangemilenits, for example in the area of personnel management, themselves need to be reformed. Nev. r theless, at this stage, drastic, mid-stream change to the institutional arrangements for the r "oi mn process would be counter-productive. However, it must be acknowledged that the reform goals DBM has been charged with implementing are demanding ones. If the delays already encountered in the course of preparatory work are not to be repeated, senior DBM staff responsible for the exercise should be provided with adequate resources and be supported as necessary by senior national and international consultants. Moreover, to minimize the risks implicit in exclusive reliance on DBM as the sole executing agency, it is important that the Presidential Committee on Streamlininig, and its technical counterpart, be as active and broadly representative as possible, both to hielp bring about consensus and to send out the strongest signal possible that the reform process has the full backing of the President. It is also essential that the Civil Service Commission, which has such important personnel responsibilities, actively involve itself in the streamliniiiig process. Without CSC's involvement, it will not be possible for the reformz process to address the realignmiiienit of the skills Imix and the rationalization of persoinnel management. (b) Streamlining 4.29 As mentioned earlier, it is unlikely that the Reorganization Act will be passed in this session ot Congress, though it may be passed soon under existing after the May elections. However. even if the Act is not passed, it should still be possible under existing legislation to reallocate resources based on a redefinition of priorities among existing departments and agencies. The reorganization of the Bureau of Internal Revenue is indicative of just how much can be - 65 - achieved at the agency level. For reform to be achieved without the passing of legislation, however, it is necessary that the legal situation be clarified so that legal challenges can be avoided. Managers are reluctant to implement the powers they have under the law for fear of legal reprisals or congressional or other political interference. Whether or not the Reorganization Act is passed, the Presidential Committee should review all the existing statutes, such as the "Magna Carta" for primary teachers, which impede effective personnel management with a view to setting itself an explicit legislative agenda to remove these anomalies and constraints. 4.30 Whatever the legislative outcome, DBM needs to continue the work of setting down clear guidelines or rules of the game for the redefinition, restructuring and/or reorganization. The aim is to identify and eliminate overlapping or redundant responsibilities and bottlenecks, simplify procedures and improve both productivity and effectiveness. It is essential to require that this process be applied evenly across departments and with objectivity, consistency and transparency. The best way of ensuring such fairness is to insist that departments and the central body responsible be helped in undertaking reviews of their mandates, organization and staffing by outside teams with international experience in this area. There is a need for the center also to provide guidelines for the nomenclature and organizational structure of departments and agencies, while recognizing that the actual internal organization will vary from departnent to department according to sectoral needs. DBM's work is already well advanced in this regard. (c) Salary Strategies 4.31 In addition to the department-level analysis of objectives, costs and staffing requirements, it is important that a government-wide analysis of different staffing options be undertaken to ensure consistency with macroeconomic targets. Even in a civil service in which more r esponsibility is delegated to departments, some critical parameters will remain under central control, the most important of these being salaries, the overall CS size and the overall wage bill. Salaries should be set in relation to the experience of the government in being able to recruit and retain the staff it needs. Private sector salaries obviously contain a lot of information in this regard. Determining desired staffing levels is difficult and to some extent arbitrary, but there are standard tools available. Ratio analysis for support positions should be developed by means of a survey of the prevailing situation and reference to international comparators. Similarly for population-based programs, such as health and education, the adequacy of present ratios may be assessed in the light of a determination of sectoral priorities. 4.32 Cost Simulations We briefly indicate by illustration what such an analysis of staffing options would involve. We cannot be more than illustrative because of the limited information we have, especially about staffing composition (by skill and occupation). What we hope to show by our simple example is that a more detailed salary simulation exercise would be both feasible and worthwhile. The reform program simulated takes the year 2000 as its target date, and aims to improve both the skills profile and the salary structure of the civil service by this date (the latter by moving towards pay-comparability with the private sector) within the constraints imposed by the overall wage bill target of 5.6 percent of GNP. Since the wage differentia! is greater at higher levels, the simulation aims for a 180 percent increase in real terms in the salaries of grade 19 and above, and of 40 percent in the salaries of those below grade 19. In terms of the skills mix, the numbers of support staff (grades 1-9) are reduced to bring the ratio of support to total staff to the international average of approximately 15 percent (a per annum reduction of 5.1 percent turns out - 66 - to be required). Teachers increase on account of population growth by 1.5 percent, while the number of high-level professionals is increased by 10 percent over the period. The simulations indicate that such changes are affordable if the high growth rate (7 percent per annum) the Philippines is aiming for is achieved. If only a lower growth rate is thought possible, more modest targets should be aimed for. These alternative reforms would also leave CS size almost unchanged in the aggregate. 4.33 Salary Reform Strategies Unlike the reform package illustrated above, the SSL2 program, as shown by our earlier simulations, is unrealistic if macro constraints are to be recognized. The SSL2 target is also unrealistic for another reason: it will become impossible to go from giving all grades the same absolute increase to giving some grades (those which will still be a long way from their SSL2 target) a large increase and other grades (those which will soon have reached their SSL2 targets) no increase at all. Assuming that the government is bound by SSL2, it should move away from its policy of giving all civil servants the same absolute increment. At the very least, it should give equal percentage increases. But, if possible, SSL2 should be replaced by a salary policy which takes greater cognizance of private-public sector wage differentials. If this is not possible, the government should view SSL2 only as a staging-post and should begin plans for reform-ls to widen the compression ratio once SSL2 has been attained. To this end, it is essential that DBM embark as soon as possible on its planned studies on remuneration. A comprehensive survey of private-sector salaries will be needed to provide the basis upon which a rescheduling of public-sector Falaries can be undertaken. 4.34 Rapid growth, and the expected increase in fiscal revenues as a consequence, may prevent the GOP from suffering from a wage-led blowout of the government deficit even if nominal wages are increased significantly. Of course, an alternative strategy would be to employ a more coniservative wage policy and let the share of GNP which salaries consume decline as the economy grows. This is a choice for government. In our opinion, however, without increased funding, successful reform of the civil service will be very difficult, and without successful civil service reform growth will be more difficult to maintain. Wage targets should, however, be contingent on satisfactory attainment of GNP and revenue targets. (d) Ways of changing the civil service size and skills mix 4.35 Attrition The current strategy of reliance on attrition to downsize is unattractive. It is highly likely that the present low attrition rate will not be able to bring about either the number or ret,nisite profile of reductions in staff which will be indicated by the reorganization of '-p r,inef Its. In addition, the notion that the problem of surplus staff can be resolved by attrition is in conitradictioni with the essential purpose of the streamlining: rationalizilg the structure and organizationl and matching people and jobs. For higher level positions at least, recruitment will be needed in order to fill new positions and bring in new skills. Through overall controls are required, better tools than the Attrition Law are available. In addition to budgetary caps, absolute caps can also he placed on civil service numbers (a practice used in Japan and many other countries). 4.36 Retrenchment Compared to attrition, retrenchment is a much more purposive means of down-sizing. Combined with recruitment, it also enables the skills mIlix to be improved. However, the legal situation as to how easily civil servants can be retrenched is unclear, at least in the minlds of the key players, and the political difficulties are obvious. Ongoing retrenchment can - 67 - also cause serious morale problems. There are various ways, however, in which some of the sting can be taken out of retrenchment. Handsome redundancy packages can be crafted, and long lead-up times to redundancy can be introduced so that staff have time to search outside, for example. 4.37 Voluntary Retirement Even with sweeteners added, retrenchment may prove neither sufficient or politically feasible. The more that this is so, the more voluntary departure should be considered. Voluntary retirement is not a popular notion within either DBM or CSC, although DBM has commissioned an "Impact Mitigation Strategy" study as part of its background studies for streamlining (para. 14). Officials of both agencies feel that in the past voluntary retirement has resulted in both little movement and the departure of some of the best staff. However, voluntary retirement can be a very effective transition tool. This is borne out both by international experience and by the case of the Central Bank in the Philippines (see Box 2), where staff size was reduced from 6,170 in 1992 to 4,520 in 1994 through use of a generous early retirement program. Nor need its cost be prohibitive. Analysis carried out for this study shows that the reductions in staff souglht in the simulation presented in the previous sub-section (40,000 departees) could be achieved at a cost of less than 10% of the 1994 wage bill. 4.38 Experience does point, however, to the importance of design in determining the success or otherwise of a voluntary retirement package. In particular, it should have the following features: (i) Any early retirement scheme should, unlike the one tried in the Philippines in 1988, be part of a reorganization: no voluntary retirement program should be initiated without the executive also deciding to act decisively on streamlining. This makes it possible to signal to well-motivated workers with good skills that they have a good future in the organization and to undesired workers that their prospects are not bright. (ii) It should be presented as one-off, both to make it effective and so that it does not harm incentives (by being seen as a "reward for failure"). (iii) The offer to take early retirement should have a closing date attached to it. Additional incentives for early signing-up should also be considered. This makes coordinating with a reorganization easier. It will encourage people to leave quickly, rather than waiting until their salary is raised or they find a private sectnr job. Having a closing date removes the particular problem in the Philippines that people will po"tpone retirement until the implementation of SSL2, with the huge wage rises it will bring. (iv) It should be generous relative to existing schemes. (v) It should be targeted. Exclusion of particular skill categories, age groups and salary grades should be considered. Only those considered surplus to requirements - for example, whose positions have been abolished - should be allowed to leave under the scheme. Some degree of central oversight will be required to enforce this. 4.39 Retraining In addition to encouraging exit and entry, the skills mix can also be influenced, while retaining control on the total number of civil servants, by retraining. This is obviously a politically attractive option. There is a great deal of talent already in the CS and recenitly training has received more emphasis as a result of initiatives of the CSC. However, for - 68 - some specialized tasks, the execution of which requires long-term training or private-sector experience, retraining will not suffice. The precise mix between retraining and hiring will vary from agency to agency. 4.40 In summary, a more creative approach to the problem of controlling civil service size and skill composition is warranted than the current exclusive reliance on the Attrition Law allows. Thae Attrition Law should be abandoned or at least allowed to expire in 1997. Overall civil service size and cost should be controlled through the budget, with additional reliance on absolute size limits if required. To the extent possible, retraining should be used to improve the skills mix. Exit, whether for downsizing or to make room for recruitment of needed skills, can be encouraged via a spectrum of options ranging from voluntary retirement to retrenchment. (e) Personnel management 4.41 Progress has been made in improving personnel management. For example, more emphasis is now being given to performance evaluation (see para. 11). This is a welcome development, which should itself be closely monitored, fine-tuned and further emphasized. However, much remains to be done. Personnel policy reform should be guided by two overall objectives. First, there is a need to review and rationalize the sharing of management responisibilities between the central agencies and line departments. Personnel and related budgetary resources should be managed in a flexible way which is responsive to the needs of departments, enhlances the accountability of managers and contributes to an improved incentive structure for managers and staff. At the same time effective running controls over posts and costs must be implemented. 4.42 Second, the current system of uncoordinated central personnel management needs to becomae m1ore effective and coordinated. There are a variety of ways in which this can be done and differenit countries handle the problem in different ways. The key requirements of any successful systemii are a clear demarcation of responsibilities among and good coordination between the central or core malnagement agencies. Current arrangements should be reviewed in the light of this goal. DBM and CSC are the two key central personnel agencies, and better coordination between the two is a sine qua non for rationalizing the personnel management system. The overlapping legal iiiandates of DBM and CSC in relation to personnel management notwithstanding, a more coordiniated approach from the two agencies should be possible given a constructive attitude on both sides. 4.43 In addition to these generic problems, there are a number of specific issues which require government attention. One is the paucity of career streams for professionals within the civil service. A second is the nexus of problems affecting senior management. In addition to improving salaries and other incentives, the current lack of security faced by this class must be addressed. Finally, as pointed out earlier, there is little information available on the composition of the civil service by variables such as age, sex or education. It is crucial to redress this deficiency by the end of the reform period to ensure that over the medium term an human-resource planning, monitoring and control capacity can be developed. In this regard, plans of the central personnel agencies for a conmmiloni data base are a welcome development. 5. BUDGET PLANNING AND MANAGEMENT 5.1 Current budget processes impact the GOP in providing and sustaining needed projects programs effectively and efficiently. They appear to often: hinder coordination between budget planning, development programming, and project and program implementation; facilitate the selection of low priority projects over higher priority projects; and contribute to delays in implementing high priority projects. Furthermore, they sustain a persistent bias against funding for adequate maintenance of existing facilities, with a consequent decline in some public services. This chapter (1) briefly describes that process; (2) analyzes the characteristics of the process that gives rise to these problems; and (3)concludes with recommendations that could ameliorate them. This chapter also has an Annex on "Recognition and Management of Contingent and Other Liabilities", which recommends that the cost of these liabilities be recognized explicitly in the budgetary framework when liabilities are created (i.e., do not wait until the contingency is called). 5.2 The expenditure budget process is complex, especially in a democracy, because it entails the reconciliation and compromises of a multitude of conflicting interests, each seeking to maximize its net benefits from the public sector. For the Philippines that has a congressional form of government, there are the inhlerent constitutionally designed balance of power among the three branches of government: the Executive, Congress and the Judiciary. 5.3 The prevailing structures and processes within which the expenditure budget is formulated each year either affect or are perceived to affect budget outcomes, and hence the relative satisfaction of the multitude of interests competing for advantages. Attempting to change the institutional structures and procedures is tantamount to attempting to change the rules of the budgetary game and, as such, will generate support from those who perceive themselves to be potential winners and attract opposition from those who perceive themselves to be potential losers. Althouglh it has what might be termed technical elements. reforming the expeniditure budget process is largely a political matter. Indeed, the recommendations in this Chapter may not change the "proceclures" so much as how the procedures are implemented. The Formal Budgetary Cycle 5.4 The budget process consists of four phases: development, appropriation and authorization, implementation and monitoring , and accountability. Details are described in Background Paper Section 111. The process of preparing the 1996 budget is described as an example below. Phase One: Development of President's Proposals for Presentation to Congress. 5.5 Process Broad parameters for the budget year (e.g. interest rates, exchange rates, sectoral growth rates and unemployment rates) are set at 13-14 months prior to the actual start of the fiscal year (i.e., October-November 1994 for preparing the 1996 Budget). These assumptions will allow derivation of the projected rates of growth national government expenditures and revenues. The "Budget Call: FY 1996" is issued toward the end of 1994 to all departments and agencies, and in January, 1995, the DBM issues to line agencies indicative budget ceilings for 1996 - 70 - Cabinet. The DBM uses the baseline budget approach in setting these ceilings. This is defined as the minimum level of expenditures required to continue the basic functions of the department or agency adjusted for increases in population and productivity improvements. The departments and agencies submit their 1996 budget proposals to the DBM in March 1995, and after intense consultations and reviews, the DBM prepares the preliminary budget allocation. In June and July the final budgetary figures are confirmed in meetings between DBM, and agency heads and the Chairmen of the Regional Development Councils before being printed as the Budget of Expenditures and Sources of Financing (BESF) document that is placed before Congress by the President. 5.6 Analysis It is noteworthy that only foreign-assisted capital projects are scrutinized as to their costs and benefits in the process of preparing the budget. These projects typically originate in operating departments, and are evaluated by the NEDA Secretariat in terms of certain plausible criteria. Final selection is made by the Investment Coordinating Committee consisting of the Executive Secretary to the President, the Secretaries of Economic Planning, Finance, Budget and Management, Agriculture, Energy, the Governor of the Central Bank and the Chairman of the Coordinating Committee on Philippine Assisted Projects (CCPAP). Locally funded projects do not get as Imluch scrutiny on project viability that is necessary. Furthermore, the recurrent budgets of departments and agencies are carried forward from year to year virtually without analysis, except for adjustments related to population changes. 5.7 The problems of project selection within the executive, arise not from any lack of coordiniation among the major organizational players but rather from the uneven quality of the iiformiiation on which they base decisions. Departments are, of course, unlikely to submit complete and unbiased information in support of the projects they wish to pursue. The NEDA secretariat can hardly be expected to ascertain all of the relevant, but nevertheless obscure, facts about all of the projects. Phase Two: Authorization by the Congress 5.8 Process In July, the President's BESF is submitted. Both the House and the Senate simultaneously conduct hearing. Conflicts between the two bodies are reconciled by joint conference committees. 5.9 Analysis In this process, the Congress, the President and the Executive (DBM) have different but strong powers. The Congress can make amendments to the BESF as long as the total budget size remains equal to or less than the BESF. To counter these changes Congress may introduce, the President can exercise line-item veto. The DBM exercise somewhat different power later on; they can maneuver cash releases during implementation of the budget, especially when cash rationing must be introduced due for example to overestimated revenues, to alter priorities of projects from the congressionally approved budget. 5.110 Notable in this appropriation process is the power of the Congress to insert appropriations for items that have lower priority, in the view of the executive, than the items eliminated. Projects may be approved by the Congress that have not been evaluated by the Executive and for which no essential preliminary steps have been taken. such as land acquisition for capital projects. Congress may also appropriate funds for projects that clearly fall within the - 71 - mandate of the local authorities. As a consequence, local authorities may reduce their funding of these types of projects in the expectation that Congress will appropriate additional funds for items that should be funded from their own appropriations. When the congress appropriates funds for capital projects that have a low priority in the eyes of the local authority, the Congress, in effect, distorts the expenditures of the local authorities that are required to use their own funds to operate and maintain them when they are in place. Phase Three: Implementation 5.11 Process The DBM prepares the National Expenditures Program (NEP) based on the approved budget. The NEP delineates the funds that are "expected to be available" to each government agency to accomplish their assigned tasks. Because the funds thus allocated may likely exceed the funds actually available (actual revenues plus net borrowing), and because of Congressional changes have been made in the BESF, several important steps are introduced to reconcile the expenditures that have been formally approved and what can actually be implemented by departments and agencies. 2 5.12 Each line department first prepares an Agency Budget Matrix (ABM) . ABMs typically envisaged expenditures that were less than the funds stated in the NEP. As an interim measure pending approval of their ABMs, departments and agencies are permitted to spend amounts based on their last year's recurrent budget. As a consequence, virtually no capital 3 spending takes place in the first quarter of the fiscal year. 5.13 After reviewing WFP/ABM of any line agency, DBM issues allotment release orders (AROs). Until 1994, advice of allotment (AA) was used instead of AROs. These documiients authorize agencies to enter into obligations or commitments for specific activities. tJnlike AA which was the same for all expenditures, allotment release orders are classified into two kinds depending on the nature of payments: general allotment release orders (GARO) and special allotment release orders (SARO). GAROs apply to almost all routine recurrent expenditures and to some routine capital expenditures (such as school buildings), whereas SARO applies to most capital projects and other outlays that require additional clearances. AROs can be issued at any time and are meant to increase flexibility of budget management. Later, the DBM then issues to each agency, on a quarterly basis, a Notice of Cash Allocation (NCA) to authorize the department or agency to pay obligations up to an approved amount on any of the items approved in the GAROS/SAROs. This process might best be characterized as an attempt to ration cash disbursement among departments and agencies to the cash available to the central government on a quarterly basis, as determined by the Cash Programming Committee. The Committee considers the inflow of revenues, the proceeds of net borrowings, the proceeds of asset sales and the caslh balances carried forward in the accounts of the Central Government and each of its departmiienits and agencies. This process is changing rapidly with the introduction from FY1995 of the simplified fund release system. 2 ABMs wcrc called Work and Financial Plan (WFP) until 1994. 3 The samie happens when budget is not approved by the Congress prior to the start of the fiscal year. - 72 - 5.14 Analysis The current system with two-layer budget release mechanism (allotment release orders--and NCAs) has been designed to safeguard thie targets for fiscal deficits committed by the Government. It should also be noted that the Executive Branch (DBM, in particular) maximizes its control over the expenditure budget by increasing the appropriations (by overestimating future revenues) relative to the actual cash available because, while it forces the rationing of cash among departments and agencies over the course of the year, it also affords the Executive an opportunity to ration cash in accordance with its own priorities rather than those of the Congress. The overstating the available revenues also has the advantage that it gives the appearance of greater availability of counterpart funds for foreign-assisted projects (FAPs). FAPs increases the inflow of foreign exchange which can substitute for more foreign borrowing. Although it has not happened every year, the shortfall of the actual flow of revenues can arise for many reasons. Among them are: a) errors in the basic assumptions concerning the major statistical indicators of the economy; b) inadequacies of the models used to derive revenue forecasts from the projected level of econcmic activity; and c) inclusion in the BESF of future revenues based on proposed structural changes in the revenue system that are not approved by Congress or are not implemented to the degree anticipated. 5.15 It should also be noted, that, because departments and agencies generally do not know the annual quantum of funds that will be made available to them with which to meet their current obligations until the year is virtually at an end, and that quantum is almost certainly less than the AA/ARO, they have learned through painful experience to "stockpile" enough idle balances to meet their annual wage and salary obligations before entering into commitments for contract maintenance or for capital projects. The inevitable consequence is a deterioration in the existing stock of physical capital as a result of inadequate maintenance and inordinate delays in undertaking approved capital projects, even those with a high priority. 5.16 Finally, implementation of the budget is monitored by a number of agencies. In particular, foreign-assisted capital projects are monitored carefully by both NEDA and CCPAP. But, locally funded projects are not included among those to be monitored oni a timely basis. The status of these projects is reported only annually. Phase Four: Budget Accountability 5. 17 In the course of any fiscal year each department is required to report to the Congress on their cumulative allotments, obligations incurred/liquidated, total disbursements, unliquidated obligations, unobligated and unexpended balances and the results of expended appropriations. The general consensus appears to be that Congress makes little use of these data, perhaps because it is so voluminous while not disclosing much about the actual performance of departments and agencies in delivering services to their clients. 5.18 The Commission of Audit is mandated by the Constitution to oversee the integrity of the Government's financial transactions. In other words it conducts a system wide audit to ensure that funds are expended only on congressionally approved objects within the authorized time periods. The Government also needs what is now commonly termed a management audit that seeks to assess whether government received good value for the money expended and has achieved reasonable cost effectiveness at the program level by improving program1 design and by attaining - 73 - maximum operational efficiency. The COA is authorized to implement such audits but so far has not had adequate capacity for management audits. Recent Budgetary Initiatives: Palliative or Cure? 5.19 Recently, there have been a number of attempts to ensure timely implementation of priority projects: Pump-Priming Program, more aptly entitled the "Earlier Release of Funds Program", introduced in 1993; Resource Based Budgeting Strategy introduced in 1994, with the purpose of accelerating the implementation of high priority capital projects by assigning priority to certain projects should cash rationing be necessary; Core Public Investment Program; and Flagship Programs/Projects, which might be considered as an extension of the CP1P to include not only projects but policies, strategies and programs that have been accorded high Presidential priority. 5.20 These attempts have met at best with only limited success. The problem can be attributed to several causes: (i) The Congress often accords higher priority to some of its own projects than some of those favored by the Executive, and inserts them into the budget. Thus, when new projects are introduced by the Congress into the budget, departments and agencies have to adjust their budgets to accommodate these changes. It is only after these adjustment has been made and their WFP/ABM approved by the DBM that the resource releases can be started. This normally takes about two to three months losing the crucial dry season suitable for construction; (ii) The departments may stockpile cash releases at first to ensure adequate payments of wages and salaries; (iii) In the course of the year departments and agencies may also submit to the DBM revised WFPs for the current year. These entail new approvals, revised AAs and revised NCAs. For example, project appraisals undertaken by the Executive Branch are sometimes inadequate. Although approved in the General Appropriations Act, as a result of more detailed assessments, some projects of executive origin may be subsequently modified or abandoned; (iv) AAs are only approval in principle, and since the total amount of AAs is usually not released, agencies do not enter into contracts until they receive NCAs; (v) D)epartments and agencies may deliberately delay priority projects as a strategy of directing work to favored contractors, or to substitute them with their favored, and in the national point of view, lower priority projects; (vi) Many construction projects cannot be undertaken in the rainy season. Since the dry season is the first three months of the year, delays introduced by any of the foregoing factors may be compounded by the fact that when all of the approvals have been obtained and contracts made, it may not be feasible to undertake the work, or work in progress may be halted for several months, for climatic reasons; and - 74 - (vii) Insufficient funds may be released to departments and agencies to meet their current obligations. Issues 5.21 A number of issues can be identified from the situation described above. The first issue is whether the expenditure priority is established rationally taking into account the goals of the Government, the role of the private and the public sectors, and the need for maintenance. The second issue is that if an adequate amount of funds reach line agencies in a timely manner for implementation of the priority activities. The third, inadequate project preparation or short-term considerations appear to be affecting implementation capacity of line agencies. Finally, measures need to be taken so accountability of line agencies be enhanced and timely corrective action can be taken if and when priority activities are under-funded. 5.22 Priority Setting There is an inherent tension in reconciling the conflict between the spending priorities of the executive and those of the elected representatives in the Congress. The Congress can impose their own priority through amendments to the President's budget proposal. In the Philippine's Constitution this power is restricted in two important ways: Congressional additions to appropriations must be matched in Peso terms by Congressional deletions so that total spendinig remains the same as that proposed by the President; and the President has a line item veto over Congressional additions to appropriations, although this power has been rarely used except for vetoes against debt service cap. 5.23 The Congressional additions are usually on locally-funded activities most notably represented by three sectors: local roads; state colleges and universities; and irrigation. As discussed in Chapter II, these are either of low return (state colleges and universities) or devolved to local governments. From economic points of view, therefore, continued central funding is inefficient, although the political returns to these activities are presumably quite different. Future O&M requirements of these projects are also not evaluated adequately straining local government budgets or putting more burden on the scarce education resources that should be better used for basic education. Furthermore, congressional deletions often include a number of projects, foreign assisted or locally funded, which have been well prepared. 5.24 It might be argued in this environment that a threat of revenue shortfalls during the budget implementation, and resulting cash rationing represent a conscious strategy by the executive designed to reassert its spending priorities. By overestimating revenues, total appropriations approved by Congress exceed those that can be financed. This, or more precisely, the threat of revenue shortfalls necessitates the introduction of cash rationing by the executive during the fiscal year. Cash rationing makes it possible for the executive to make funds available to finance the projects to which it accords the highest priorities and to withhold the funding for projects introduced by the Congress. Note that the overestimates of future revenues do not have to be actual every year as long as there has been enough past occasions to make this threat credible and cash rationing possible. - 75 - 5.25 Another factor is that the Congress has not approved the Medium-Term Philippine Investment Plan (MTPIP) developed by the NEDA, and thus the Congress may not feel constrained to restrict its project insertions to those that are consistent with the Plan or to treat Presidential proposals that flow from the Plan as any more deserving than any others. One merit of a multi- year plan is to have both Congress and the Executive agree on the medium-term goals and priority of the country, and to have direct links between program planning and program budgeting. 5.26 Finally, the allocation of responsibility for judging project priority within the Executive also poses problems. Operating departments and agencies are supposed to originate project proposals that are then to be evaluated by the NEDA secretariat and approved by a hierarchy of interagency coordinating committees. The originating departments and agencies are under no incentive to critically assess their own proposals and NEDA does not have enough human resources for assessing all of them in depth. In particular, the technical expertise to assess the managerial feasibility of so many diverse projects (e.g., location, land assembly, costing, future net operating costs and contract delivery dates) is in short supply. With weak planning it is not surprising that implementation delays frequently occur because the decision to proceed is little more than a decision to begin the planning that should have taken place before the decision to proceed was made. 5.27 Fund Releases to Priority Activities Cash rationing also has the effect of delaying the implemnentation of the executive's high priority projects because departments and agencies face grcat uncertainty concerning the flow of funds that will be forthcoming. They are required to spend much staff time in meeting the reporting and other bureaucratic requirements of the central agencies in order to secure even a fraction of the cash allocations necessary to finance the projects approved by Congress. In addition, given the uncertainty, some agencies try to safeguard their montlh by month cash allocations for the payment of wages and salaries first before starting spending on more discretionary budget items. 5.28 Issues at Line Agencies There are, however, factors other than cash release meclhaniisnm delaying the implementation of priority activities as can be seen from the limited success of the several ad hoc attempts of recent years (briefly described in the previous Section) to avoid the delays of cash rationing on the executives highest priority projects. Among these additional deterrents are: inadequate project appraisals, inadequate trained manpower in depar-tments and agencies, the time consuming task of keeping track of appropriations, allotments, comllmitment authorities and monthly cash allocations, the diffusion of responsibility among the central agencies, and conflicts between the priorities assigned to projects by the central agencies and the priorities favored by the operating departments and agencies thus subverting the priorities established by the President Policy Recommendations 5.29 The recommendations set forth below seek to ameliorate these problems, and thus accelerate the implementation of projects, by shifting the conflict resolution process from the "behind closed doors" forum where cash rationing takes place, to Congressional debates about the appropriate size of the public sector and the appropriate allocation of spendiig among the broad functional areas of national interest. - 76 - 5.30 Under the present expenditure budget system, there is an undue emphasis on the limited flow of funds to government in the short-term and inadequate weight given to the medium term economic outlook and the policy options available to government to modify cash flow under alternative scenarios. This myopia tends to obscure the important choices that have to be made between achieving in the medium term the appropriate mix among taxing, borrowing and spending. On the other hand, by placing these issues before Congress squarely and unequivocally it may be possible to secure a higher degree of public and Congressional commitment to the spending priorities and the macroeconomic parameters within which annual expenditure budgeting takes place. 5.31 Taking these into account, recommendations are made here establish a three-year fiscal framework and a resource allocation consistent with the framework (called a development program), which would be tabled to the Congress. Reforms in budgetary operations management and organizational arrangements are also discussed and recommended. Fiscal Framework 5.32 It is recommended that the National Government adopt a tlhree-year Fiscal Framework, which incorporates quarterly projections of: fiscal revenues, expenditures (capital and current), deficit, net loans and advances and net cash requirements. The Fiscal Framework should be updated every quarter, probably every time new national accounts quarterly data is available. (Perhaps the same projections should also be provided on a national accounts basis in addition to a government accounting basis to avoid debate about alternative statistical concepts). The net financing requirement projections should not simply be derived arithmetically from projections of the other items: the Executive must assess the appropriate and sustainable amount of central government borrowing in the light of emerging economic circumstances, and prudent debt mianagement. Guarantees and other contingent liabilities also have potential real costs attached to them. The Fiscal Framework and annual budgets need to incorporate these costs explicitly (see Annex to this Chapter). 5.33 The Fiscal Framework should be the subject of a Joint Resolution of the Congress at the beginning of the term of every Congress and prior to each tabling of the President's Expenditure Budget. The Executive should give the reasons for past deviations from the Framework and for modifications in the future. These modifications should be voted upon annually. 5.34 The executive should establish assumptions underlying the Framework in consultation with the BSP. These parameters include: inflation rate, foreign exchange reserve levels, exchange rate, government cash balances, unemployment rate, and income growth rate. The Executive and BSP should carry out sufficient macroeconomic analysis (probably within a framework of a consistent economy-wide model) to be able to specify these assumptions as objectively and reliably as the state of our understanding permits. It would be prudent to publish thret alternative scenarios together with assumptions when the Framework is tabled at the Congress: a high scenario, a low scenario and the most likely scenario. 5.35 One or more independent institutes and business associations should be encouraged to make public forecasts of these conditions on a regular basis. This will serve to educate the - 77 - commiiunity concerning the factors that affect the mid-term economic outlook. Moreover, the availability of this information will provide an imiportant check on the Executives assumptions. 5.36 The Fiscal Framework (in particular, the most likely scenario) should be based on existing tax rates and structures, and the Executive should not raise the allowable expenditures in a currenit Framework on the assumption that their proposed tax measures will be approved by the Congress. However, a proposal for a tax reform should be accompanied by a revised Fiscal Framework that would take into account the changes in revenue that are expected to result from the proposed reform. By the saute token, measures introduced in the Congress that would alter the projected flow of revenues relative to the existing system should be debated in the light of their probable eftects on the Fiscal Framework. 5.37 The Executive would be encouraged to be prudent and should allow sufficient cash reserves on hiand to finance all approved expenditure appropriations when the projected budgetary revenues fall short of the amount projected in the most likely case Framework. Should actual revenues exceed those of the worst case scenario, supplementary appropriations could then be introduced in the course of the fiscal year. 5.38 If the concept of the Fiscal Framework were implemented as described above, the possibility of revenue shortfalls and subsequent cash rationing should be substantially reduced. As discussed above, rationing of cash has been the "real" expenditure budget system. This method of budgeting is perverse from the point of view of facility maintenance and project implementation. Operational departmiients adopt a defensive strategy of reducing maintenance and delaying projects to ensure that they have the cash to pay future wages and salaries. On the other hand, cash rationinlg increased the autlhority of the Executive to deny funding for projects and programs that it did not approve. The Executive will be forced to be more forthright in resisting expenditures for projects that it does not want to proceed. The Presidential veto is available for just that purpose. Development Programming 5.39 Based on the available funds for capital and recurrent expenditures within the Fiscal Framework, responsible central department(s) should carry out development programming. This activity will determinie resource allocation so as to maximize the rate of growth of per capita incomiie of Philippine residents while taking into account the needs to alleviate poverty and to progress towards equality of opportunity. 5.4(0 This activity comprise: (i) evaluating and proposing modification in the proposals of new and existing projects and programs prepared by line departments; and (ii) proposing to Congress the allocation of funds among major functional categories including operations and maintenianice, within the constraints specified in the Fiscal Framework. Congress would be asked to approve this allocation, perhaps after modification, and accept the constraint that additional appropriations made by Congress must be financed by reductions in appropriations within the same functionial category. This would serve to reduce the circumvenltion of Presidential priorities by departmiienlts and agencies appealing directly to members of Congress. 5.41 For capital spending, a distinction needs to be made between routine, relatively low cost, pr-ojects and unique and larger projects. The former would include, say, school buildings and - 78 - health clinics. Here Development Programming would be concerned with establishing, for each type of investment common formats for proposals, determining evaluation criteria and establishing rules of thumb in the use of these criteria. For unique projects over a certain cost it would be necessary to proceed on a case by case basis: each requires its own cost-benefit analysis conducted according to the best international practice. The Executive needs the capacity to critically evaluate the proposals submitted by operating departments, taking into account potential economic returns, the relationship with private sector activities (does it complement or substitute the private sector?), maintenance expenditure implications, and operating departments implemelntation capacity. 5.42 Development programming needs to go beyond proposing the allocation of available funds among broad government functional categories, for example, education, health, transportation. The appropriate allocation depends upon the cost effectiveness built into the design of the programs and the weight to be attached to competing programs withill the same functional category, e.g. preventative health care vs. curative care, and primary education vs. secondary and higher education. 5.43 Congressional Insertions Congress will be asked to discuss and approve the Fiscal Framework and the Development Program, at the beginning of each congress (i.e. every three years) and as updated annually before the budget thereafter. It should also be known that the congressional changes that are inconsistent with the Framework and the Program will be vetoed. The responsibilities assigned to local authorities should be clarified so there is no ambiguity concerning them. The president should also announce his intention to veto all insertions of project expenditures that relate to local authority responsibilities. It is imperative that these vetoes be exercised with an even hand without regard to political affiliation. 5.44 The criteria established by the Executive (with Congressional approval, preferably) for thc assessment of small, routine projects should be promulgated and the President should make it cleaf to the Congress in advance that the same criteria will be applied to Congressional insertions as to those originating in the executive. Furthermore, Congressional insertions involving major projeuts would each require an in-depth cost benefit analysis, perhaps carried out by an independent consultant(s), and the resulting report would be subject to Congressional hearings. Operational Management 5.45 Once the Fiscal Framework and Development Programming are in place, annual hudgets should be prepared and implemented efficiently--a task of Budget Operational Management Group. Operational Management is concerned with the acquisition of inputs at the lowest possible cost, maximizing the operational efficiency with which these inputs are converted into efficient delivery ol desired government services to the targeted clientele. 5.46 Operational management is therefore responsible for: (i) preparation of budget proposals to implement the Development Program; (ii) establishing rules for personnel management and incentives together with the Civil Service Commission, and for procurement of materials and supplies and government contracting; and (iii) monitoring the operational performance of other departments and ensuring that efficiency in production and distribution is achieved. In this last regard, the GOP needs to decide which of COA or the Operations Management should have the core capacity to conduct management audits. - 79 - 5.47 The use of the baseline budget concept for programs should be continued: this is the real cost (i.e. net of changes in the level of wages/salaries and prices) of providing the same quality and quantity of outputs to the same types of clients. In short, the only adjustments in the baseline expenditures from year to year are the change in the size of the clientele as a result of changes in population, internal migration etc. as offset, in whole or in part, by planned changes in operating efficiency. There should be prior agreement between the operating department and the Operational Manager concerning the projected growth rate of the operating departments baseline budget expenditures over future years. With this information on ltnd the Operations Manager can easily project the future cost of programs assuming no policy changes. By aggregating these projections across all departments the relatively fixed portion of the expenditure budget is known. When necessary to accommodate new programs or enriched existing programs, it may be necessary to introduce across the board or preferably selective cuts in baseline budgets through the constraints specified in the call letters to operating departments. 5.48 Budget proposals submitted by operating departments should state the agreed baseline budget for the year and the proposed changes from that amount and the reasons for the changes. A distinction should be made between proposed policy changes that involve changing the type or quality of the outputs or changes in the type of clients from operational changes that will change the method of production or distribution of outputs. The former changes must be assessed in terms of the Development Programming; the latter changes must be assessed against conventional business criteria: what cost savings are to be effected over what period of time. 5.49 Capital expenditure proposals should be accompanied by estimates of the future recurrent costs involved. Before capital projects are approved, agreements should be reached concerning the financing of these recurrent costs: are they to be added to the baseline budget in whole or can some be absorbed, i.e. offset by cuts in other programs or by increasing operating efficiency. 5.50 Operational Management need also to monitor implementation of activities, as part of preparation of periodical management audits. As such, all major projects should be monitored, not only the foreign assisted ones as is currently done. When it is found line departments are substituting projects agreed under Fiscal Framework and Development Programming with their favored projects, thereby subverting the President's priority, discipline on the department heads and those responsible should be swift and substantial. 5.51 Accounting system There is a need for a computerized accounting system defined in a manual, and operating departments must be staffed with qualified accountants capable of implementing that system. The same system should be prescribed for local authorities to permit consolidation of data. The coding system must retain enough detail to allow classifications by activity, program, object of expenditure, responsibility center and geographic location of the inputs . id outputs. It is important that the coding system be flexible enough to accommodate changes in program design and organizations yet yield consistent data over time at a fine level of desegregation. Expenditures should at least be classified as proposed, recommended, appropriated, authorized. comrmitted and actual. The computer system must be robust with much backup and accessible to poorly trained personnel. - 80 - 5.52 In accounting, the Philippine budget is carried out in nominal terms. This could cause some confusion especially in a multi-year activities. It would be better to request departments and agencies their budget proposals expressed in real terms in the Budget Call Letter. Inflation adjustments can be made later by the Operational Management. Budgetary Control " 5.53 Once this framework is established, there would seem to be no need for further central control of budgetary resource releases. That is, if the availability of cash is not a constraint; an agreement has been reached between DBM and each operating department with respect to the future changes in its base budget; and procurement is governed by acceptable rules and procedures, the auditor is probably the only protection against the use of funds for purposes other those designated in the appropriation. Thus the whole WFP-AA-NCA (or ABM- GARO/SARO-NCA) procedure would seem to be unnecessary for recurrent spending. All appropriations included in the General Appropriations Act should be guaranteed funding for recurrent expenditures as required by the departments and agencies. 5.54 For capital spending, funds may be appropriated in advance of design and assessment. There would appear to be a continuing need for WFPs for each of these projects, at least those projects above a certain Peso value, and for control of commitments tied to the completion of stages of plan development and construction. The Government is addressing this issue through issuance of SAROs for these projects. 5 55 This may provide a method of achieving co-ordination among agencies. The form of the WFP could be specified so that certain stages are explicit. Thus funds are released for design with responsibility assigned to the originating department; the second stage would require operational management checks on costing and responsibility would be assigned to Operational Management; the next stage would be economic and financial evaluation and this be the responsibility of development programming. If the proposal passes these stages then the onus is on the originating department to proceed. The originating department and the Operational Management devise, on a case by case basis, the series of benchmarks that would trigger payments to the contractor as work proceeds or investigations when progress is behind schedule. Organizational Arrangements 5.56 There are three tasks to be performed: fiscal management including the preparation of the Fiscal Framework; economic policy formulation, including the preparation of the Development Program; and Operational Management of the Budget. Although outside the scope of this paper, to these might be added legislative management and intergovernmental relations (consisting of local governments on the one hand and foreign governments on the other). All of these tasks might be called staff/headquarters functions, as distinct from line/operational functions. All are involved in the control/command/direction of operating departments or in negotiating arrangements with the legislature or other governments. Generally speaking their relationships are with other government officials (elected or appointed) rather than with the general public. - 81 - 5.57 These headquarters functions are highly interdependent. Yet each calls for a high degree of specialized knowledge. How then should responsibilities be allocated among specialized organizations while maintaining consistency and coherence? The method of achieving co- ordination often employed in the Philippines is through the establishment of committees. This, however, is not very efficient unless a unanimity rule is always enforced. Otherwise, each member can deny responsibility for any cormmittee decision. It is thus better that responsibilities be assigned to the individual officers in charge of separate organizational units than to committees, although consultation with those carrying out related responsibilitieg should be mandatory. 5.58 Among the three headquarter functions, those who establish the Fiscal Framework and Development Programming should work very closely together due to the nature of the tasks. If these two functions should have to be divided organizationally, it needs to be made clear that the financial constraints, which are dominated by macro-economic considerations, were absolutely binding on development programming. Operational management, together with project monitoring and management audit capacity, should probably be separate from the other two organizationally. ANNEX TO CHAPTER 5 RECOGNITION AND MANAGEMENT OF CONTINGENT AND OTHER LIABILITIES Introduction 1. In recent years, many governments throughout the world have suffered from fiscal restraint. As a consequence, these governments have turned to new financial instruments and more creative financial mechanisms to pursue their policy objectives and to undertake their programs with the limited resources available to them. Amongst these new instruments are loan and performance guarantees offered to the private sector to encourage public-private partnerships. Since encountering an energy crises in 1992 and 1993, the Government of the Philippines, through the National Power Corporation, has used these types of guarantees to encourage infrastructure development, particularly power generation facilities under "Build-Operate-Transfer" (BOT) and similar arrangements, and there is a likelihood that the use of these guarantees will be extended in future to other infrastructure projects such as ports, toll roads, and gas pipelines. To the end of 1994, the government's exposure through these guarantees may have amounted to near $10 billion. 2. Because loan and performance guarantees do not require an immediate cash outlay by a government, there is a tendency to view them as an inexpensive way to provide financial assistance to the private sector to encourage that sector to undertake activity in keeping with government policy objectives. However, it must be clearly recognized that loan guarantees have a cost, and experience of some countries has shown that, in the long run, loan guarantees can be very expensive. As a consequence, it is essential that appropriate management practices and controls be imposed on these instruments and that sufficient safeguards be built into the budgetary process to avoid large future losses which could prevent the government from reaching its fiscal targets. 3. In addition to direct loan and performance guarantees, the Government of the Philippines, either directly or through its agencies and state owned corporations, has incurred a number of other contingent and direct liabilities which do not require any immediate cash outlay. Examples include liabilities associated with social security institutions (SSS and GSIS) and liabilities associated with agrarian land reform. As is the case for loan guarantee liabilities, these liabilities need to be properly managed, accounted for and reported in order to avoid future shocks to the government's fiscal plan or framework. The importance of good management and control of these liabilities is underscored by the fact that the two social security institutions are found to be financially unviable in the long term under the current terms of contributions and benefits, and with current operating practices . Also, P 2 billion of guarantees related to agrarian land reform is expected to increase rapidly and could reach as high as P 10 billion in the near term. 4. This annex reviews the definitions of these instruments, examines methods to manage and control their costs in the context of the government's fiscal plan and budgetary targets, and recommends accounting for and reporting of these liabilities in order to complete the accountability loop. "Philippincs--An Agenda for the Rcform of Social Security Institutions" World Bank. October 12, 1994 - 83 - Definitions 5. To assure full comprehension of the issues discussed, there must be a common understanding of the words and terms used in this paper. Abbreviated definitions of some of these terms follow. The attachment at the end of this annex has a more extensive description of these terms. * Liabilities - obligations to outside parties resulting from past txzansactions. * Contingent liability - a liability that is dependent on some future event that may or may not occur. * Actuarially determined liabilities - liabilities related to future events that are virtually certain - not a contingency. * Disbursements or outlays - amounts of money actually paid out. * Expcnditures - costs incurred that have been paid for, or will be paid for in the future. * Costs - both expenditures and non-cash expenses incurred to undertake an activity. * Reserves - allowance and provisions - amounts set aside to meet future payment requirements. Usually recorded as a liability rather than money placed in a separate bank account. Management and Control of Contingent Liabilities (a) Objectives 6. Loan guarantees have a cost. Therefore, an objective of any government should be to impose appropriate management practices and controls on these instruments to control these costs, and to ensure that sufficient safeguards are built into the budgetary and expenditure process to avoid large future losses which could prevent the government from reaching its fiscal targets as set out in its fiscal plan. 7. A process for management and control of loan guarantees and other contingent liabilities should incorporate a control mechanism which will influence decision-making. Decisions to incur costs, such as entering into loan guarantee agreements, must have immediate fiscal consequences, otherwise governments will have little incentive to use guarantees judiciously and effective controls over their use will be weak. A proper fiscal management process, which incorporates agreed upon fiscal parameters or spending limits established at the beginning of the year, must be in place to govern spending decisions. Such a fiscal management process has been recommended in Chapter 5 and includes appropriate operational planning, budgeting, accounting and reporting functions. 8. As explained in the definition of a liability, expenditures for loan or performance guarantees should only be recorded in the year in which the cost is identified. For purposes of this discussion paper, it is assumed the cost is identified at the time the guarantee is issued. - 84 - (b) Budget and expenditure management process (i) Existing process 9. As noted in Chapter 5, under the existing budget and expenditure management process used by the Government of the Philippines, rationing of cash has been the "real" budget and expenditure management process. Appropriations, while representing the legal limits to government spending as enacted by Congress, significantly exceed cash availability and Therefore have ceased to be effective in limiting government spending. The only control on spending is the availability of cash. 10. As explained above, the issuing of a loan or performance guarantee incurs a cost but that cost has no immediate cash requirement. Due to this absence of a cash requirement, control over the guarantee falls outside the purview of the "availability of cash" control mechanism and, therefore, guarantee decisions are essentially uncontrolled by the budgetary process. (ii) Requirements of a revisedprocess 11. To be effectively controlled, loan and performance guarantee costs must be treated as expenditures in the year guarantee arrangements are entered into and these expenditures must fall within congressionally approved spending limits or appropriations. To properly control all spending, the Government of the Philippines must implement an effective budgetary and expenditure management process as outlined in Chapter 5, incorporating key fiscal targets, such as revenues, expenditures, deficit and borrowings. These targets should be agreed to by those authorities involved in the spending process and enacted into law by Congress. 12. As noted in Chapter 5, revenues should be a realistic estimate of actual collections, and the executive should be required to operate within expenditure limits in order to achieve the targeted deficit. Since Congress authorizes the executive to incur expenditures by appropriating funds, the appropriation limits allocated to Ministries should become the key control mechanism for spending decisions. 13. As stated in the definitions above, the issuing of a loan or performance guarantee incurs an expenditure which should form part of the annual expenditures charged to an appropriation. In this way guarantee costs, or expenditures, incurred by issuing a guarantee use up available appropriations and are limited by the authority to spend money as vested by Congress. (iii) Types of appropriations 14. While the foregoing enunciates the basic principles of congressional control over spending, including spending arising from the issue of guarantees, in reality governments use more detailed mechanisms to impose authorities and limits. Appropriations are frequently divided into various types so that Congress may direct expenditures more specifically. For this reason, separate appropriations for operating expenditures, for capital expenditures, for grants, for specific projects, etc. are created. If Congress wishes to impose limits on annual loan or performance guarantee costs, a separate appropriation within the fiscal expenditure plan can be established for this purpose; or a number of appropriations distributed across different ministries or different sectors of the economy could be created. Nevertheless, no matter how appropriations are structured, issuing - 85 - a guarantee incurs a cost which must be covered by an appropriation. This basic principle should remain rigidly in force. (iv) Managing contingent liabilities through the budget 15. Instead of providing for the cost of guarantees as an expenditure against an appropriation when the guarantee is issued, or when a change in provision is required, an alternative method of managing these costs is through the budget process. This method calls for a reduction in the spending authority or appropriation limit to offset the estimated eventual cost of the guarantee. Reducing spending authority forces the ministry responsible for administering the appropriation to spend less on other activities. The costs of the guarantee would be recorded by a central agency, such as the Department of Finance, rather than by the ministry directly involved. 16. The advantages of this process are two-fold. First, should higher costs need to be recorded in a year subsequent to the year of issue of the guarantee due to failure of the debtor, the increase in costs would be absorbed by the central agency rather than by the ministry involved. This would obviate the need to disrupt the financial plans of the ministry in that year due to an action that took place in an earlier fiscal year. 17. The second advantage to managing through the budget is that it allows the central agency to monitor and manage all of the government's contingent liabilities in a uniform and consistent manner by specialists. This should ensure that objective and high standards of management are applied to all guarantees. 18. The disadvantage of this process is that the government ministry responsible for the creation of the guarantee is never accountable for its disposition. 19. Managing through expenditures or managing through the budget are both acceptable processes and both will be successful if the government has the will and discipline to apply them vigorously. (c) Valuation 20. Attributing a cost to a loan or performance guarantee may be a difficult task, but nevertheless it is an important function in the management and control of these instruments. The amount of this cost, which serves to use up funds available for spending in an appropriation, will influence decision-making of managers who are responsible and accountable for appropriations. 21. Various techniques and methods may be used to estimate expected losses on guarantees. The parameters of the guarantee will indicate, to some extent, the amount of the needed provision. Estimations of losses should take into account the principal amount of the loan guaranteed or the maximum exposure to the government under a performance guarantee. Accrued and unpaid interest, amounts recoverable from the borrower, and value of assets pledged as security are amongst other factors be considered. In some cases, market information may suggest a valuation. Market valuations are discussed in Mody and Patro, (World Bank, March 1995). In other situations, loss estimates may be made by reference to - 86 - (a) recent financial performance of the borrower as disclosed in financial statements, (b) economic conditions which may affect the borrower, (c) independent credit reports, (d) adequacy of security pledgedfor loans which have been guaranteed, and (e) recent collection experience on the loans and related interest. 22. Measurement techniques can include statistical models, historical rates and trends, and specific information on the borrower or the project. Notwithstanding the techniques used, the final assessment will depend on judgment of government officials to a considerable degree. (d) Timing of Expenditure Recognition 23. One key issue in accounting for loan and performance guarantees is assessing when a provision for losses should be recognized as an expenditure and a liability. Proper accounting treatment requires recognition of losses when it is determined that they are likely. In virtually all cases, it is known that a loss is likely at sometime before the government must make payment on a guarantee and, in some instances, it is known at the time the guarantee is issued. Frequently, the likelihood of payment becomes evident during the life of the guarantee. 24. Notwithstanding when an initial estimate is recognized, estimates require reappraisal as new events occur, as more experience is acquired, or as additional information is obtained. Changes in the circumstances of the borrower may require changes to the provision for losses. Changes in the provision may be either increases or decreases. As a general principle, the provision for losses on loan guarantees should be reviewed on an ongoing basis and any changes should be charged or credited to expenditures of the year in which the change occurs. (e) Minimizing losses 25. For any government that issues loan or performance guarantees, reducing risks and minimizing potential losses is a prime concern. An important way to reduce risk is to share it with the party being guaranteed. A government can achieve this by guaranteeing only part of a loan, say 75 percent, forcing the lender to bear the risk on the remaining 25 percent. Besides reducing any future losses by 25 percent, this sharing forces the lender to thoroughly assess potential leans and to not lend money for excessively high risk cases. This enables the government to indirectly use the credit assessment expertise of a lender to its advantage. 26. Other risk-reducing techniques include obtaining project assets as security in the event of guarantee payout, and charging loan guarantee fees. - 87 - Management and Control of Actuarially Determined Liabilities (a) Objectives 27. Actuarially determined costs and liabilities relate to insurance and pension plans where cash payments are frequently made many years after the costs and liabilities are incurred. A government should aim to impose management practices and contrqls on these costs and liabilities, and to build sufficient safeguards into the budgetary process to ensure that these costs are accounted for in the multi-year fiscal plan. 28. As is the case with loan guarantees and other contingencies, a process for management and control of pension and insurance plan costs should incorporate a control mechanism which will impact on decision-making. Costs must be accounted for when incurred and expenditures must be recorded against appropriations at that time. (b) Budget and Expenditure Management Process 29. The principles for budget and expenditure management for pensions and insurance costs are the same as those outlined in section (6) on page 86. These include establishment of a fiscal plan, legislative authorization of spending (appropriations) within that plan, and recording of expenditures as costs are incurred leaving less spending available for other activities. 30. Annual costs related to pension and insurance plans represent the value of benefits earned by employees and other plan members during a fiscal year. Since plan members often contribute to the costs of the plans, government costs equal the total value of benefits earned less contributions of the plan members. Generally, actuarial assessments of these plans are made on an annual or triennial basis to establish the amount of the government's annual contribution, and this amount is usually recorded as the government's cost. This amount should be charged as an expenditure against appropriations. 31. Actuarially determined costs and liabilities related to pensions require two additional considerations: (i) plan amendment costs, and (ii) fluctuations in liabilities caused by changing economic and demographic assumptions used in the actuarial estimates. These can result in experience gains and loses. 32. A plan amendment occurs when a plan is changed and retroactive credit is given to plan members for services rendered in the past. For example, a benefit formula may be improved for a pension plan or insurance coverage may be increased resulting in higher claims. Plan amendments may have a cost. This cost together with the resulting increased liability should be accounted for by the governnent and charged against an appropriation in the year the amendment is made. The higher costs arc the result of a decision to improve benefits and a full accounting of the cost should be made in the fiscal year in which the decision is made. - 88 - 33. The periodic actuarial valuation of a government's pension liability will usually determine that estimation adjustments are required due to experience gains and losses or to changes in actuarial assumptions. Estimation adjustments are inevitable because the actuarial valuation requires ongoing forecasts of uncertain future events. These adjustments can be sufficiently large to materially affect the achievement of targets set out in the fiscal plan. Therefore, adjustrnents should be amortized over a reasonable future period because of their tentative nature and because further adjustments will likely be required in the future. 34. For amortization of estimation adjustments, a reasonable future period could be related to the number of years which members of the pension plan have remaining as contributors. Spreading these costs over a number of years should smooth out the annual fluctuations of costs and diminish distortions in the fiscal results. Accountability (i) Reporting and confidentiality 35. A govermnent demonstrates its financial accountability to elected representatives and the public through financial and performance reporting. Generally, this reporting should disclose sufficient information on outstanding loan guarantees and their related provisions, on other contingent liabilities and on actuarially determined liabilities to allow Congress to judge the government's management of these obligations. The costs should also be reported relative to congressionally authorized appropriations. 36. However, disclosing provisions for losses on loan guarantees and other contingencies may result in commercially sensitive information entering the public domain. This could be detrimental to the financial position of both a borrower and the government. It is therefore recommended that the government give careful consideration to the extent of information made public and, if practical, present only highly aggregated figures. (ii) Audit 37. Because the government may disclose only summarized information on loan and performance guarantees, it is essential that detailed information be subject to independent audit scrutiny. Provisions are frequently based on subjective judgments and may be influenced by factors unrelated to expected losses on outstanding guarantees. Independent audit can provide Congress with assurance as to the credibility of guarantee costs and provisions or, if ihe auditor is not satisfied with them, he can report accordingly to Congress. The audit function completes the accountability process. Conclusion 38. Loan and performance guarantees have become significant tools that governments use to achieve policy objectives. These guarantees have a cost which should be recognized, accounted for and reported as part of the fiscal management and control regime of a government. The same applies to actuarially determined costs and liabilities where cash payments may be made many years after the costs are incurred. Pension and insurance plans are examples. - 89 - 39. To properly develop control mechanisms over these types of costs, a stable fiscal management process incorporating the budgetary and expenditure cycles and fiscal targets should be in place. The elected Congress should authorize realistic spending ceilings through appropriations. Guarantee, pension and insurance costs should form part of spending limited by appropriations so that issuing guarantees or amending pension or insurance plans will represent a spending decision. A government manager must judge these spending decisions relative to available resources and other spending options. The government must impose a high level of discipline over this process to ensure it functions as designed. - 90 - ATTACHMIENT DEFINITIONS (a) Liabilities In accounting terms, liabilities are financial obligations of a government to outside organizations and individuals as a result of transactions and events occurring on or before the year- end. They are the result of contracts, agreements and legislation in force at the year end which require a government to make future payments related to transactions or events which have already occurred. It should be noted that accounting for a liability effectively recognizes that an expenditure has been incurred, an expenditure which must be included in the determination of a government's annual deficit. In relation to guarantees, the transaction or event giving rise to a future payment is the act of entering into a guarantee arrangement. If it is determined that this guarantee has a cost (see Section 6 below), that cost in the form of an expenditure should be recorded in the accounts of the government in the year the cost is identified. This may be the year in which the arrangement is entered into or a subsequent year. (b) Contingent liabilities A contingent liability is a liability that is dependent on some future event that may or may not occur. Any future payment on a contingent liability is uncertain as at the reporting date and costs should only be recorded if there is a likelihood that a payment will eventually be made. Of necessity, assessment of amounts to be recorded under contingencies requires judgment and estimations. However, the mere fact that an estimate is involved does not preclude the recognition that a cost and liability have been incurred nor the reporting of this cost and liability in financial statements. A government guarantee that the "rules of the game" will not change during the life cycle of a project is a form of assurance to a project sponsor and is entirely under the control of the government. In this regard, changing the "rules" and any costs related thereto are not actions that are "contingent" or "dependent" on some uncontrolled future event. Therefore any costs associated with changing the rules should only be recognized in the year when the change takes place and no contingent liability is created by simply issuing assurance. (c) Actuarially determined liabilities such as insurance and pensions While future payments under some contractual arrangements depend on future events, there is virtual certainty that these events will take place, particularly if these arrangements involve a large number of individual contracts as is the case for life insurance and pension plans. These - 91 - future payments are not considered to be "contingent liabilities; they are generally treated as definitive obligations. Because future payments are dependent to some extent on mortality rates, demographics and economic forecasts, estimating these liabilities requires an actuarial valuation, and therefore they may be referred to as actuarially determined liabilities. They should not be considered as contingencies. (d) Government budget and reporting entity Many governments operate a general fund into which all revenues are credited and against which all expenditures are charged, and for which the government's budget is prepared. In addition, governments frequently create a number of agencies or state-owned corporations to which funds are transferred from the general fund but whose operating transactions are not recorded in the general fund. This structure gives rise to "off-budget" transactions. An example would be loan guarantees issued by the Land Bank of the Philippines or performance guarantees of the National Power Corporation. While the legal structure of these entities clearly distinguishes them from the core government (as defined by the general fund), many of them are financially dependent on the government and, in substance, form part of government activity. To properly manage and control the financial transaction of these organizations which substantively form part of government, and to ensure they are adequately accountable to the country's elected body, the government's budget, which is approved by the elected legislature, should encompass these entities. This is not to mean all government owned or controlled entities should form part of the budget and reporting entity, but those which are financially dependent should be fully consolidated with the government for budget and reporting purposes. State enterprises which are financially self-sufficient may be excluded. However, as a rule of thumb, those organizations which derive more than 20 percent of their operating revenue (or cost recovery) from the government should be treated as "government" for budgetary and financial reporting purposes. In this way, financial activities of these organizations which impact the fiscal plan and future fiscal targets will be subject to the same congressional and executive control as "core" government. Furthermore, for those organizations which appear to be financially self-sufficient but whose obligations are guaranteed by government, the government should recognize as expenditures any of their liabilities which are expected to be repaid by the government as soon as repayments are identified as likely. (e) Disbursements (outlays), expenditures and costs These terms are frequently used interchangeably to describe government spending. However, when elements of accrual accounting are introduced into government financial reporting ( -S -f;. bg K ig. i= : .g M48;; a\ga : - .Hii ti it$' } \ 082 f'X m0'|'0 $ X ;i; i iEis '4$ 0 a::: ::Et U . .... iD * l 0 ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~l 000; 00000'~~ g agga:i g awt Q d 5 i; aa*z a;,;; ................................................................................. ;g-ga, ; i, a; ,i, , ;; ;f; ff 0 ; ;;; ; ..............................