ConditionsforEffectiveDecentralizedGovernance

《ConditionsforEffectiveDecentralizedGovernance》由会员分享，可在线阅读，更多相关《ConditionsforEffectiveDecentralizedGovernance（101页珍藏版）》请在装配图网上搜索。

1、CONDITIONS FOR EFFECTIVE DECENTRALIZED GOVERNANCE:A SYNTHESIS OF RESEARCH FINDINGSOmar Azfar, Satu Khknen*, and Patrick MeagherIRIS Center, University of Maryland* The World BankMarch 14, 2001This research was financed by a grant from the World Banks Netherlands Trust Fund to the IRIS Center, Univer

2、sity of Maryland. This paper forms part of a larger study that includes companion papers on the Philippines and Uganda. The authors express gratitude to Tugrul Gurgur, Anthony Lanyi, Jeff Livingston, and Diana Rutherford for their assistance. Special thanks are due to Social Weather Stations in the

3、Philippines and Makerere Institute of Social Research at the Makerere University in Uganda for their implementation of the surveys.TABLE OF CONTENTSTABLE OF CONTENTSiFIGURESiiTABLESiiiAbbreviationsivExchange RatesivExecutive Summary1Chapter I: Introduction5Chapter II: Conceptual and Research Framewo

4、rk6Conceptual Framework6Arguments in Favor of Decentralization6The Case Against Decentralization7When Does Decentralization Produce Benefits in Practice?8Research Design9Data Collection11Chapter III: The Institutional Settings13Structural Overview13The Philippines13Uganda15Sectoral Issues: Health an

5、d Education16Health16Education18Chapter IV: Measuring Performance22Education and Health Care: Access, Service Quality, Outcomes22Allocative Efficiency: Preference-matching24Productive Efficiency Issues: Corruption29Overview of Corruption Trends29Survey Data on Corruption31Chapter V: Performance Fact

6、orsCivic Disciplines33Assessing Voice and Exit33Voice33Exit/Mobility37Citizen Information Sources37Chapter VI: Performance FactorsPublic Sector Institutional Disciplines37Intergovernmental Disciplines37Experiences with Hierarchical Constraints37De Facto LGU Discretion: Adjustability37Public Sector M

7、anagement Disciplines37Structures and Capacities in Local Administration37Indicators of Effective Public Sector Management37Chapter VII: Linking Disciplines and Performance to Outcomes37Uganda: Disciplines and Outcomes37The Philippines: Causes of Corruption37The Philippines: Consequences of Corrupti

8、on37Chapter VIII: Conclusions37Bibliography37FIGURESFigure 1. Research Logic10Figure 2. Household Satisfaction with Government Health Units23Figure 3. Household Satisfaction with Primary School (percentage of households)24Figure 4. Corruption in Uganda and the Philippines29Figure 5. Prevalence of Co

9、rruptionResponses from Officials (means)32Figure 6. Uganda: Source of Information on Local Politics37Figure 7. Philippines: Source of Information on Local Politics37Figure 8. How often does the most qualified person get the job (means)37Figure 9. Average Meritocracy in Health Sector: Promotion Crite

10、ria37Figure 10. Average Meritocracy in Education Sector: Promotion Criteria37Figure 11. Uganda Accountability QuestionsResponses from Officials and Facilities37Figure 12. Philippines Accountability QuestionsResponses from Officials and Facilities37TABLESTable 1: Structural Comparison of the Extent o

11、f Decentralization21Table 2: Uganda and the PhilippinesComparative Welfare Indicators*22Table 3. Communicable Diseases23Table 4. Uganda: Household Opinions on Sub-County Government Funding Priorities:25Table 5. Philippines: Household Opinions on Municipality Priorities26Table 6. Correlation between

12、public officials preferences and household preferences for funding priorities128Table 7. Percentage of Households Reporting Corruption in Government31Table 8. Percentage of Sub-County/Municipal Officials Acknowledging Type of Corruption31Table 9. Percentage of People Voting for Each Reason1, 2, 336T

13、able 10. Type of Primary School used by Surveyed37Table 11. Uganda: Household Use of Health Facilities (percentage of households)137Table 12. Philippines: Household Use of Health Facilities (percentage of households) 137Table 13. Sources of Information about Politics in Uganda and Philippines37Table

14、 14. Sources of Information about Corruption in Uganda and Philippines37Table 15. Uganda: Political Awareness Across Households of Different Ethnicities137Table 16. Philippines: Political Awareness Across Households of Different Ethnicities137Table 17. Summary Table: Fiscal Federalist Arguments for

15、Decentralization37Table 18. Discretion: How easily can you respond to suggestions of local people.37Table 19. Uganda: Adjustability Measures across Surveys (means)37Table 20. Philippines: Adjustability Measures across Surveys (means)37Table 21. Respondents and co-workers need for second jobs37Table

16、22. Facilities Responses on Payment Delays (Index)37Table 23. Ugandan Staff Competence Across Surveys (aggregate indices )37Table 24. Filipino Staff Competence Across Surveys (aggregate indices)37Table 25. Uganda: Civic Disciplines and the Quality of Education37Table 26. The Philippines: Causes of C

17、orruption37Table 27. The Philippines: Explaining Knowledge of Immunizations at Government Clinics37Table 28. The Philippines: Explaining Waiting Time at and Satisfaction with Government Health Clinics37Table 29. The Philippines: Explaining Changes in Occurrence of Diseases37Table 30. The Philippines

18、: Explaining Education Outcomes37Annex Table 1. Issue Areas Covered in Surveys37AbbreviationsADBAsian Development BankARMMAutonomous Region of Muslim MindanaoAUSAIDAustralian Agency for International DevelopmentBIRBureau of Internal RevenueCAOChief Administrative OfficerCARCordillera Autonomous Regi

19、onCOACommission on AuditDBMDepartment of Budget and ManagementDECSDepartment of Education, Culture and SportsDEODistrict Education OfficerDHODistrict Health OfficerDILGDepartment of Interior and Local GovernmentDOFDepartment of FinanceDOHDepartment of HealthDPWHDepartment of Public Works and Highway

20、sDSCDistrict Service CommissionEPIEnhanced Program on ImmunizationHUMCHealth Unit Management CommitteeICRGInternational Country Risk GuideIGGInspector General of GovernmentIRAInternal revenue allotmentLabanLaban ng Demokratikong PilipinoLakasLakas ng Edsa-National Union of Christian DemocratsLAMPLab

21、an ng Masang PilipinoLGALocal Governments ActLGCLocal Government CodeLGULocal Government UnitLTBLocal Tender BoardMFPEDMinistry of Finance, Planning and Economic DevelopmentMILFMoro Islamic Liberation FrontMNLFMoro National Liberation FrontMOEMinistry of EducationMOHMinistry of HealthMOLGMinistry of

22、 Local GovernmentNCRNational Capital RegionNGONon-Governmental OrganizationNPCNationalist Peoples CoalitionNRMNational Resistance MovementOLSOrdinary Least SquaresPAFPoverty Action FundPCGCPolitical Commission on Graft and CorruptionPPPhilippine PesoPTAParent-Teacher AssociationPTCAParent-Teacher Co

23、mmunity AssociationRDCResident District CommissionerSEFSpecial Education FundSMCSchool Management CommitteeSPCDSouthern Philippines Council for Peace and DevelopmentSWSSocial Weather StationsTBTuberculosisTEEPThird Elementary Education ProgramUNICEFUnited Nations International Childrens Emergency Fu

24、ndUPEUniversal Primary EducationUshUgandan ShillingVATValue-Added TaxVHCVillage Health CommitteeWHOWorld Health OrganizationExchange RatesUS$1 = 50.15 Philippine Peso (PP)US$1 = 1830 Uganda Shillings (Ush)Executive SummaryUnder what conditions does decentralized governance prove most effective? This

25、 paper starts unraveling this question by presenting the main results from a field study of decentralization in two countries: the Philippines and Uganda. Specifically, the paper analyzes the impact of selected institutional and social factors on decentralized public health and education service del

26、ivery and governance quality. The theoretical arguments in favor of decentralization stress allocative efficiency (matching of local preferences for public goods), increased productive efficiency and integrity, and better cost-recovery. The possible disadvantages of decentralization noted in the lit

27、erature include potential failures of policy coordination, tendencies toward “elite capture” of local governments, and failures due to inadequate capacities in local government. It should be noted, however, that favorable outcomes from decentralization depend on placing responsibility for different

28、types of public goods at appropriate levelse.g., goods where local features dominate at the local level, and those with strong spillovers at higher levels.The actual outcome of decentralization, to a significant degree, depends on certain conditions that are explicitly or implicitly assumed in tradi

29、tional analyses, notably the presence of key institutional disciplines. The paper focuses on three kinds of institutional disciplines: Civic disciplines are those associated with the capacity of citizens, media, and non-governmental organizations to make their views known to the government (“voice”)

30、, and to switch to other localities or suppliers of services (“exit”). Intergovernmental disciplines are those exerted between different levels of government. Examples include central government oversight of local government operations, or budgetary constraints (or limits to taxing authority) impose

31、d by the center on lower levels of government. Public sector management disciplines are the ways in which each government body regulates and constrains the behavior of its own officials. Examples include anti-corruption provisions, performance-based recruitment and promotion, and provision for perio

32、dic audits. Focusing on these factors makes it possible to evaluate the conditions for successful decentralized governance against a backdrop of largely uniform formal institutions defining a countrys political and administrative hierarchy. In other words, the formal structures of decentralized gove

33、rnance define incentives and resource allocations across the public sector, but they are far from being the whole story. The Philippines and Uganda have decentralized within unitary structures that (largely) do not vary across regions. While moving authority and resources to lower levels, this strat

34、egy has put in place a system of policy development, fiscal allocation, and governance monitoring that is mainly top-down in nature. Such an outcome, however, appears to be typical, especially in the absence of federal arrangements. Despite this, decentralization in the two countries is substantial

35、and has encouraged local political mobilization and policy initiative, although in both countries, a complex mixture of formal checks, centripetal politics, and administrative rigidities make local governments much less responsive than they could be.The studies presented here entailed the administra

36、tion of nine survey instruments in each countryto households, officials at two sub-national levels, as well as schools and health facilities. The subject matter and frequently the questions were repeated across surveys in an effort to collect comparative data that could also be cross-checked and ana

37、lyzed with some confidence. The research also included key informant interviews and the collection of documents and secondary data. The paper provides results in three main areas: measuring government performance (i.e., the extent to which the advantageous aspects of decentralizationallocative effic

38、iency, increased integrity and reduced corruptionwere realized); assessing the impact on government performance of three sets of determinants (i.e., civic, intergovernmental, and administrative disciplines); and evaluating the effects of performance (and its determinants to a lesser extent) on the o

39、utcomes of public services (e.g., immunization and primary school enrollment rates, household satisfaction with public services, and incidence of childhood diseases). The main results of the paper are as follows:Civic disciplines on sub-national governments in Uganda and the Philippines may be subst

40、antially weaker than anticipated by the theories of decentralization, including fiscal federalism. Voting patterns and rationales do not differ significantly for local and national elections, although there is some evidence that government officials knowledge of local conditions is better at the loc

41、al than at the national level. The study revealed important constraints on information flow, which can be expected to exert a major influence on the quality of governance and of service delivery. Citizens in both countries (Uganda more so than the Philippines) rely substantially on community leaders

42、 rather than the media for information on local politics and corruption. This raises the potential for state “capture” of local government by the elite and may explain the apparent weakness of local accountability in practice. In Uganda, moreover, there is credible survey evidence linking citizen in

43、formation access to the quality of education. Information flow in the opposite directioni.e., conveying preferences of the local population to officials (“voice”)appears less constrained in both countries than access of households to information.The theoretical concern about locating authority for p

44、ublic goods at appropriate levels gains support in the evidence from both countries. In Uganda, while immunization programs are “vertical” initiatives of the central government, they rely substantially on local support. Data on preferences of households suggest that further investment in improving i

45、mmunization delivery does not occupy a high priority in most communities. Local governments appear to have grasped this, and many have failed to invest scarce resources effectively in the necessary personnel, storage systems, and equipment. While this choice might be wise in light of competing prior

46、ities, it does suggest at least an important tension between the central governments commitment to childhood immunization and its devolution of important aspects of the delivery system (a public good with “spillovers”). In the Philippines, the reverse seems to be true of primary education. There, ce

47、ntralization appears to impose tangible costs in terms of governance, efficiency, and responsiveness.Corruption, as expected, remains an important concern in both countries. As for the causes of corruption, the paper produced some evidence on the effects of discretion, voting patterns, and media acc

48、ess. Households in both countries generally perceived there to be more corruption in higher levels of government, and officials at higher levels usually reported a greater scope of discretion. Data from the Philippines showed a clear association of discretion with corruption, and suggested a negativ

49、e impact of voting participation and media access on corruption.The research also demonstrated some significant deleterious consequences of corruption. Most notable was the adverse effect of corruption on health care services and health outcomes in the Philippines. The research aimed to uncover evid

50、ence of the interaction between decentralized government and social status differences such as ethno-linguistic and religious identity. On the whole, these did not prove highly significant as determinants of access to public services or governance quality, although they were cited as problems in the

51、 primary schools and with respect to standardized test scores, and they did have an effect on information access and political participation. Moreover, the data from both countries suggest that non-meritocratic criteria, which include political and kinship relations, intrude significantly into perso

52、nnel management decisions.Hierarchical constraints were evident. Local officials reported substantial restrictions on their ability to adjust funding and service delivery to local demand. Higher-level governments (at the provincial/district level) reported significantly more discretion, either acros

53、s the board (Uganda) or with respect to funding allocation (the Philippines). Governments at this level also reported stricter accountability, for example in the form of audits. This is consistent with the more general picture that discipline in these two systems usually runs from the top down. More

54、over, the pattern of increasing discretion at higher levels broadly corresponds to reports of greater corruption at higher levels.In short, while decentralization in both countries has moved authority and resources to sub-national governments, the results do not match the most optimistic theoretical

55、 expectations. Local governments in the Philippines and Uganda are not consistently responsive to local preferences, although they appear to be aware of local preferences. In most cases they cannot break out of the procedural, resource, and governance constraints that prevent them from responding. P

56、erhaps the most notable concern regarding decentralized governance in these contexts is the flow of information from governments to their constituents. Here, particularly outside major urban centers, there arise the possibilities of government capture by local elites, with potentially harmful conseq

57、uences for governance and public service delivery. Thus, the cases of Uganda and the Philippines both suggest caution in the planning of decentralization processes. This concern goes beyond the possibility that the theoretical benefits may have been diluted due to incomplete decentralization. First,

58、 the theoretical prerequisites for fiscal federalism (or full political decentralization) are sufficiently broadly defined that their existence in any case lies in the eye of the beholder. Second, it would be difficult to find another poor developing country that has pursued decentralization as rigo

59、rously over a period of years as the Philippines and Uganda. Third, there can, of course, be too much of a good thing. Effective decentralization implies restraints from above and below, hence limits beyond which it becomes dysfunctional. The research presented here addresses itself less to the opti

60、mal extent of decentralization than to conditions that appear to produce the best results within the range of decentralized arrangements practicable in developing countries.Chapter I: IntroductionEvents around the globe clearly show that decentralization has long since “arrived” as both a public sec

61、tor reform model and a development strategy. More recently the debate on decentralized governance appears to be taking a healthy turn from theoretical arguments to the marshalling of empirical evidence, and from the sweeping to the particular. The question, “Is decentralization a good or a bad idea?

62、” is gradually yielding its status as the central preoccupation in this area. That there are no generic answers is increasingly recognized, and political developments have in many cases made the question moot. This paper summarizes the results of research that analyzed the conditions for effective d

63、ecentralization in the Philippines and Uganda.1 This paper forms part of a series produced under a grant from the World Banks Netherlands Trust Fund. The other papers include a literature review as well as empirical studies undertaken in the Philippines and Uganda.1 Hence, this paper proposes a particular way of framing the decentralization question: under what conditions does decentralized governance prove most effective? This way of framing the issue places the emphasis not on the merits of decentralization (i.e., as compared to central