DECENTRALIZATION AND GOVERNANCE: AN EMPIRICAL INVESTIGATION OF PUBLIC SERVICE
DELIVERY IN THE PHILIPPINES
Omar Azfar, Tugrul Gurgur, Satu Kähkönen,* Anthony Lanyi, and Patrick Meagher
IRIS Center, University of Maryland, College Park
*World Bank
December 19, 2000
This research was financed by a grant from the World Bank’s Netherlands Trust Fund to the IRIS Center, University of Maryland. This paper forms part of a larger study that includes a similar research effort in the Uganda. A synthesis paper will summarize findings and analyze data from the two countries, as well as the review of the theoretical and comparative literature on these questions.
1
Table of Contents
Tables and Figures
Abbreviations
Executive Summary
Chapter I: Introduction
Theoretical suppositions on decentralization
Summary of the paper
Chapter II: Structures of Decentralized Governance in the Philippines
Governmental Structure
Politics and Administration
Autonomous Regions
Health Services
Impact of Decentralization
Vertical Programs
Education
Governance Structure
Personnel and Resource Management
Chapter III: Data Sources
Chapter IV: Assessing Government Performance
Measuring Public Service Outputs and Outcomes
Allocative efficiency
Productive efficiency
Decentralization and corruption in the Philippines
Measures of corruption
Cost effectiveness
Cost recovery
Local Revenue Administration Problems
Survey Data on Tax Collection
Summary
Chapter V. Performance Factors: Civic Disciplines
Voice: Voting and Political Action
Voice: Media and Knowledge
Exit
Summary
Chapter VI: Performance Factors: Intergovernmental Disciplines
Fiscal Structure
Local Administration, Central Oversight
Adjustability
Summary
Chapter VII: Performance Factors: Public Sector Management
Meritocracy
Accountability
Capacity
Chapter VIII: Linkage to Outcomes
The Causes of Corruption
The Consequences of Corruption
Health
Education
Summary
Chapter IX: Conclusions
Bibliography
Annex
Tables and Figures
Table 1. Description of Surveys
Figure 1. Household Opinions on Municipality Priorities:
Table 2. Correlation Between Public Officials Preferences and Household Preferences for Funding Priorities
Table 3. Correlation Between Public Officials Preferences and Household Preferences for Funding Priorities
Table 4. What have you done to find out about local citizen’s preferences? (percentages)
Table 5. Correlation Between Public Officials Preferences and Ilocano Preferences for Funding Priorities
Figure 2. ICRG’s Corruption Rating for the Philippines
Table 6. Corruption Components and Index
Figure 3. Adjustability and Corruption Across Agencies
Table 7. Facilities’ Responses on Payment Delays (Index)
Table 8. Tax Collection Percentage by Province
Table 9. Why People Vote in Elections
Table 10. Sources of Information on Politics
Figure 4. Sources of Information on Local Politics and Sources of Information on Local Corruption
Table 11. Probit Estimates of Sources of Information about Local Politics on Whether Respondent Has Heard Any Reports of Local Corruption
Table 12. Probit Estimates of Sources of Information about National Politics on Whether Respondent Has Heard Any Reports of National Corruption
Table 13. Does Public Good Delivery Get Covered in the Media and Affect Voting
Table 14. Political Awareness Across Households of Different Ethnicities
Table 15. Mobility
Table 16. How easily can this facility adjust its plans to respond to suggestions of citizens?
Table 17. Adjustability Measures across Surveys
Table 18. Flexibility to People’s Suggestions and Authority to be Flexible
Figure 5. Officials’ Views of How Often the most Qualified Person Gets the Job (average)
Figure 6. Average Meritocracy per all public officials: Criteria used in promotions
Table 19. Meritocracy
Table 20. Accountability
Figure 7 Accountability Questions – Responses from Officials and Facilities
Table 21. Capacity
Table 22. The Causes of Corruption
Figure 8. Partial Plot on Causes of Corruption
Table 23. Explaining Waiting Time at and Satisfaction with Government Health Clinics
Table 24. Explaining Change in Occurrence of Diseases
Table 25. Effect of Corruption on Education Delivery
Figure 9. The Causes and Consequences of Corruption in the Philippines
Annex Table 1. Household Opinions on Municipality Priorities:
Annex Table 2. Preferences of Provincial, Municipal, and Household Preferences on Municipality Priorities:
Annex Table 3. Cleaning Corruption Perception of Households
Annex Table 4. Cleaning Corruption Perception of Municipal Health Officials
Annex Table 5. Probit Estimates of Determinants of Media Source for Local News
Annex Table 6. Probit Estimates of Determinants of Media Source for National News
Annex Table 7. Households Use of Health Facilities
Annex Table 8. Summary Means
Annex Table 9. The Causes of Corruption Robustness Check
Annex Table 10. Explaining Knowledge of Immunizations at Government Clinics
Annex Table 11. Cleaning Satisfaction with Government Clinic and Waiting Time Perceptions of Households
Annex Table 12. Explaining Waiting Time at, and Satisfaction with Government Health Clinics—Robustness checks
Annex Figure 1: Municipal Officials vs. Household Preferences (national avg.)
Annex Figure 2: Municipal Officials vs. Household Preferences (provincial avg.)
Annex Figure 3: Municipal Officials vs. Household Preferences (municipal avg.)
Annex Figure 4: Provincial Officials vs. Household Preferences (national avg.)
Annex Figure 5: Provincial Officials vs. Household Preferences (provincial avg.)
Annex Figure 6. Partial Plot Explaining Satisfaction and Waiting Time (composite index)
Abbreviations
ADBAsian Development Bank
AUSAIDAustralian Agency for International Development
ARMM Autonomous Region of Muslim Mindanao
BIRBureau of Internal Revenue
CARCordillera Autonomous Region
COACommission on Audit
DBMDepartment of Budget and Management
DECSDepartment of Education, Culture and Sports
DILGDepartment of Interior and Local Government
DOFDepartment of Finance
DOHDepartment of Health
DOJDepartment of Justice
DPWHDepartment of Public Works and Highways
EPIEnhanced Program on Immunization
IRAInternal revenue allotment
LabanLaban ng Demokratikong Pilipino
LakasLakas ng Edsa—National Union of Christian Democrats
LAMPLaban ng Masang Pilipino
LGCLocal Government Code
LGULocal government unit
MILFMoro Islamic Liberation Front
MNLFMoro National Liberation Front
NCRNational Capital Region
NDFNational Democratic Front
NGONon-governmental organization
NPCNationalist People’s Coalition
PCGCPolitical Commission on Graft and Corruption
PPPhilippine Peso
PTCAParent-Teacher Community Association
SEFSpecial Education Fund
SPCDSouthern Philippines Council for Peace and Development
SWSSocial Weather Stations
TBTuberculosis
TEEPThird Elementary Education Program
UNICEFUnited Nations International Children’s Emergency Fund
VATValue-Added Tax
WHO World Health Organization
1
Executive Summary
This paper—part of a larger study financed by the World Bank—focuses on the impact of selected institutional and social factors on the delivery of primary health services and primary education, under decentralized government arrangements, in the Philippines. The purpose of this research is to shed light on the conditions under which decentralization of governmental functions can improve public service delivery. Research was carried out through interviews and secondary sources, but most importantly through nine separate surveys of households, provincial and municipal officials, and employees of public health facilities, public schools and private schools, in 20 provinces, 80 municipalities, and 301 barangays.
Theory. The theory of fiscal federalism suggest that decentralization improves public service delivery because of greater allocative efficiency (matching public goods to local preferences), as well as improved productive efficiency (in part because of greater accountability, less bureaucracy, and better knowledge of local costs), and better cost recovery (greater willingness of citizens to pay fees and taxes for improved public services). Against this, some have argued that decentralization may lead to local decisions that have adverse inter-jurisdictional spillovers and undermine national policy objectives, and that the advantages of greater authority for local government could be offset by elite capture. The research team hypothesized that whether either of these supposed advantages or disadvantages of decentralization materialize depends on three institutional disciplines: civic disciplines, intergovernmental disciplines, and disciplines related to public sector management. These disciplines are defined, and relevant results are summarized, in the corresponding paragraphs below.
Structures of decentralized governance in the Philippines. The Local Government Code (LGC), enacted in 1991 and implemented since 1992, has increased responsibilities of all sub-national governments: 77 provinces, 69 cities, 1538 municipalities, and over 40,000 barangays. Regular elections are now held at each level for both executives and legislative bodies. Health expenditures were largely devolved to sub-national governments, although the central government retains responsibility for “vertical programs” like immunization and disease control. In education, the Department of Education, Culture and Sports (DECS) still maintains responsibility for budgeting and hiring for schools, but local governments now have responsibilities for operation and maintenance of schools, and de facto are often involved in hiring of teachers. Autonomous regions—notably the Autonomous Region of Muslim Mindanao (ARMM)—have more extensive responsibilities in education, but have, interestingly, performed quite poorly in this area; AARM has an especially bad reputation with respect to corruption.
Assessing government performance. Philippines is a middle-income country whose performance in health and education has shown substantial progress in some areas—such as immunization and primary school attendance—with some remaining weaknesses, such malnourishment of young children and quality of education. The research team looked at the areas in which decentralized government is supposed by theorists to perform well: allocative efficiency, productive efficiency, and cost recovery. Allocative efficiency was approached by examining how well household preferences were perceived and implemented by officials at provincial and municipal levels. It was found that household preferences—measured by responses to a survey question about preferred use of additional local funding—revealed that the main priority was roads and other infrastructure rather than health and education, apparently reflecting citizens’ perceptions with regard to adequacy of existing funding and the most urgent shortfalls. Municipal officials were to a substantial extent aware of these preferences, whereas provincial officials showed no evidence of knowledge about local demands. The main aspect of productive efficiency addressed wascorruption, which was widely perceived by officials themselves to be a problem, and these perceptions were in line with those of households. While corruption at the baranguay and municipal governments was seen as less pronounced than at the provincial and national governments, corruption at all levels of government was perceived as substantial. On another aspect of productive efficiency, delays in payments of salaries and allowances, and hiring delays, were a frequent problem. The one aspect of cost recovery examined in the study was tax collection, which showed wide variation among provinces. This is an area where local autonomy has been compromised by central government constraints; and disincentives to local tax collection have apparently been created by large transfers from the central government.
Civic disciplines are defined as the capacity of individual citizens, media and NGOs to make their views known to government—i.e., “voice”—as well as mechanisms to induce officials to act accordingly (voting and other forms of public action); also, civic disciplines include mechanisms of “exit” that involve citizens either switching from public to private services or moving to other jurisdictions. Thus voting participations, rationales for voting, other forms of public action, mobility, and choice of alternative service providers were surveyed. Moreover, the source of information of citizens about government and corruption—which is a necessary condition for the citizens’ positive input into local government—was also examined. The data suggest that civic disciplines on sub-national government in the Philippines may be substantially weaker than assumed by the pro-decentralization theorists. Voting patterns and rationales do not differ significantly between local and national elections. Popular knowledge of local government was inferior to that of national government, and there was less reliance on the media for information on local than on national politics. Indeed, citizens relied much more on local officials and personal contacts, rather than on the media, for information on local politics and corruption—in stark contrast with much heavier reliance on the media for information about national politics. This suggests the possibility of elite capture of local government and may explain weaknesses in local accountability. One positive sign was the large number of positive responses to questions about the existence of political action in local communities and the success of such action.
Intergovernmental disciplines are the mechanisms of oversight or budgetary constraints, typically imposed by the central (or provincial) governments on local authorities, affecting the lower levels’ adjustability, or discretion to adjust service delivery to local needs. The study shows that intergovernmental disciplines place tight constraints, both positive and negative, on the freedom of action of local governments. On the positive side, oversight of higher levels of government over lower ones imposes what are probably necessary constraints to misuse of funds; on the negative side, some of these constraints consist of political interference from the center in the allocation of investment funds at the local level. The adjustability index derived by the research team revealed somewhat more discretion at the municipal than at the provincial level on which services to offer, how many officials to hire, and how much to pay employees, but more authority for provincial officials to reallocate funds among different services.
Disciplines related to public sector management are those internal mechanisms and capacities at the local level—for instance, anti-corruption provisions, performance-based recruitment and promotion, and provision for periodic auditions—that affect competence and honesty of local public sector delivery. The study examined these factors under three headings: meritocracy, accountability, and capacity. With regard to meritocracy, officials interviewed claimed that merit and quality of service were the main criteria, although for health units personal connections were also regarded as important. On accountability, frequency of both internal and external audits is high, as are written targets. Most respondents, but to a lesser degree in DECS offices than in municipal administrations, said that an officer suspected of corruption faced social pressure to resign. Finally, capacity constraints do not appear severe in the Philippines, although resources of the organizations scored weaker in officials’ responses than training and education of employees.
Linkage of performance factors to outcomes: the causes and consequences of corruption. The research team found that statistical analysis of the linkage between the performance factors examined in the previous chapters and outcomes for health and education (both as objectively measured and as perceived by households) tended to be closely associated with the factor of corruption. The main linkages revealed by the statistical analysis were, first, that “adjustability” of municipal government has a significant positive effect on corruption, while reading national newspapers and voting local elections have significant negative effects. At the same time, corruption at the local level has significant negative effects on household satisfaction with health care services, waiting times at health facilities, and health officials’ knowledge of required immunizations. The health facilities’ performance (with regard to both household satisfaction and waiting times) is, in turn, strongly influenced (positively) by the officials’ knowledge of required immunizations. Corruption also had a significant negative impact on household satisfaction with public schools.
Broader implications. While decentralization in the Philippines has moved substantial authority and resources to sub-national governments, this decentralization has taken place within the context of a unitary governmental system where the hierarchy has much more discretion at upper than at lower levels, and stronger mechanisms for top-down discipline than any pressures in the opposite direction. Local governments do appear to be fairly aware of local preferences, but find it difficult to break out of the procedural, resource, and governance constraints that impede their response. The survey shows limitations to voter and media attention to local government, and this heightens the possibility of elite capture of local governmental institutions, including those responsible for public service delivery. These considerations illustrate the premise of the paper that decentralization initiatives should be guided less by simple pro and con arguments of theorists but by detailed understanding of the institutional setting in which the decentralized forms of governance are expected to operate. The study also suggests, however, that centralized administration—as in the education sector in the Philippines—exacts its own price in terms of efficiency and integrity, and is subject to local political pressures despite the attempt to insulate it from them.
Chapter I: Introduction
A major aspect of recent economic and political reforms throughout the world has been the devolution of power from central to local governments. The Philippines is an important example of countries that have decentralized in the 1990s. The reasons for this type of reform are both political and economic: they range from accommodation of local demands for greater autonomy, to the desire to increase democratic participation, and on to more economic or administrative aims, such as the improvement of public service delivery. This paper focuses on the latter aim, addressing the question of how well decentralized governance works in the Philippines with respect to specific types of public service delivery. But in attempting to answer this question, this study also examines the impact of political and bureaucratic arrangements, and the nature of civic participation, on the capacity of local government to deliver such services.
The Philippines seems a country designed for decentralized governance by sheer geography. Its 77 million people live on thousands of islands, the larger of which have mountain ranges and jungles that effectively separate large population groups. While the layout of the country has dictated some form of decentralized governance in the past, this was not always combined with devolution of political authority. In modern times, the revolution of 1986, when the Marcos regime was replaced by a democratic government led by Corazón Aquino, brought promises of decentralization from the new government. These promises were delivered with the Local Government Code (LGC), enacted in 1991 and implemented in 1992-3. The LGC significantly increased the responsibilities and resources of sub-national governments: 77 provinces, 69 cities, 1538 municipalities and over 40,000 barangays.[1] In addition, the LGC mandated the creation of local democracies at each level, with regular elections being held for both executives and legislative bodies. Expenditures devolved to sub-national governments covered a wide range of government activities, the most prominent of which was health. To defray the cost of devolved expenditures, the LGC provided for 40 percent of central government revenues collected three years before to be transferred back to sub-national governments as internal revenue allotments (IRAs). Provinces also received limited new taxing authority.