decentr alizing health: lessons from indonesia, the philippines, and

This chapter examines the decentralization experi-ence of three East Asian countries from the per-spective of how well they have addressed the specialfeatures and requirements of the health sector.These features include the substantial role of exter-nalities, the high degree of specialization, the criti-cal role of quality and timeliness, and the high levelof knowledge required to participate in the healthcare system at all levels. These characteristics haveimportant implications for the design of healthpolicy in general, and especially for a decentralizedsystem of service delivery and sector management.This chapter outlines the decentralization healthpolicies and programs of Indonesia, the Philippines,and Vietnam, focusing on the period 1985–2003,spanning the years before and after significantdecentralization began in these countries. Thechapter also points to areas where reforms mayfacilitate more effective health care delivery.

The Health Care Context of Decentralization

Experience with decentralizing health in develop-ing economies is limited, and the literature reveals

155

8

DecentralizingHealth: Lessons

from Indonesia,the Philippines,

and VietnamSamuel S. Lieberman, Joseph J. Capuno, and Hoang Van Minh

no widely accepted, best-practice health policyframework. This is partly because decentralizationin many low-income countries is a recent develop-ment, while affluent countries historically movedin a centralizing direction as constituent statescame together to form federal unions.

Another constraining factor has been the top-down, centrist bias in the influential Health for All(HFA) paradigm, used to build dominating butdifficult-to-manage and ineffective health ministriesin many countries. HFA’s main sponsor, the WorldHealth Organization (WHO), has been traditionallyuncomfortable with decentralization. The literatureon government roles in health systems is relevant tothe extent that privatization is a form of decentral-ization, but this literature misses the critical decen-tralization issue: the allocation of roles among levelsof government. The literature also lacks a connectionbetween options for decentralization and healthfinancing. Meanwhile, most approaches that focuson health challenges—including the MillenniumDevelopment Goals of the United Nations and theWorld Bank–sponsored Poverty Reduction StrategyPapers in developing countries—assume a strongcentral role for ministries of health.

east_155-178.qxd 5/17/05 11:42 AM Page 155

156 East Asia Decentralizes

Nevertheless, a debate has developed arounddecentralization design issues, with contributorsdividing into two camps. Proponents see decen-tralization, if handled well, leading to systematiccitizen involvement in setting the goals, design,and financing for health policy, and in monitoringservice provision and other functions. In this view,decentralization can also spur providers to obtainthe skills, material support, and authority theyneed to offer high-quality services. Decentraliza-tion can further enable clients to secure informa-tion, financing, and bargaining power, and offerhealth ministries a chance to jettison impracticalobligations and carve out a new role and image.

Detractors warn, however, that (badly designed)decentralization heightens vulnerability to near-term crises and longer-term risks. Typical start-upproblems include staff opposition, leading tobreakdown of deployment and other personnelmechanisms; mismatches between health carefunding and spending requirements; ambiguity inresponsibilities and premature delegation of func-tions, leading to deteriorating service quality; anddisruptions in reporting, accountability, and qual-ity control. Medium-term concerns include risingsystem costs. Specifically, downsizing administra-tive units may yield designs for key health func-tions that are neither technically efficient nor cost-effective because of diseconomies of scale. Such“transitional” problems may be difficult to correct.

Because of these risks, public health commenta-tors have called for careful introduction and man-agement of decentralization. Most analysts supportWHO’s recommendation that countries phase indevolution under central guidance, subject to strin-gent criteria, with health ministries continuing totake responsibility for specialized services, medicalsupplies, basic education and training, and otherkey functions (WHO 1995). This advice illustratestwo recurring themes: that the overriding rationalefor health decentralization is improved effective-ness and efficiency, and that the timing of theprocess is subject to ex ante design.

Such premises are usually not valid, since theimpetus for decentralization is generally political.Improved health is only a second-order objective,with imperatives such as preserving national unityusually driving the process and shaping the deci-sion to devolve to particular levels of government.

That was arguably the case in the Philippines,Vietnam, and Indonesia, which decentralized their

health services starting in 1992, 1996, and 2001,respectively.1 Evidence suggests that health ministriesin these countries initially were not prepared toarticulate and assume a new specialized role of sys-tem manager rather than main provider. Inconsis-tent policies further indicated that expediencyrather than strategy guided official responses totransition problems.

At the same time, the faltering performance ofthe health systems in these countries beforedecentralization signaled a need for significantchanges in health policy. In the Philippines,improvements in infant survival rates and otherhealth status indicators in the 1980s were begin-ning to plateau, indicating decreasing returnsfrom health expenditures that were higher than inother developing countries in terms of grossdomestic product (GDP) (Solon et al. 1992).

In Vietnam, the collapse of the agriculturalfinancing system and economic reform in the 1980sundermined funding for primary health care serv-ices and produced shortages of drugs and skills,deteriorating quality of care, and a decline by a halfor more in use of government facilities. Fundinggaps also led to higher user fees, which became afinancial barrier and reduced access to care by thepoor.

In Indonesia, the 1997 financial crisis broughtfunding cuts that confirmed the susceptibility ofthe government network to drug shortages andother breakdowns. But performance problems hadexisted earlier. The country did not sustain favor-able trends in survival and nutrition rates from the1980s in the 1990s, despite large-scale intervention.Use of public services also faltered. After rising tonearly a third, the share of people who soughtoutpatient care from public providers fell below30 percent by 1995, and below 20 percent by 1998.Meanwhile, households in the top expenditurequintile were far more likely than the poor to usepublic facilities as inpatients, and nearly as likely asto use such facilities as outpatients.

Objectives of the Chapter

This chapter examines decentralization experiencesin Indonesia, the Philippines, and Vietnam with aneye toward three sets of questions. First, how candeveloping countries design decentralization toprovide an appropriate framework for a publichealth system, and what policies and instruments

east_155-178.qxd 5/17/05 11:42 AM Page 156

promise to be effective in improving the efficiencyand equity of a decentralized health system? Second,how should countries handle transition problemsand other risks? Third, what lessons do the experi-ences of these three countries reveal?

In addressing these questions, the chapter exam-ines the emerging role of central health ministries.Critical functions include monitoring and control-ling communicable diseases, setting standards andassuring quality for devolved health services andpharmaceuticals, ensuring access of the poor tohealth services, and sustaining health financing.

The three countries have broad features thatfacilitate comparison. All have tropical or semi-tropical climates, and all are highly populateddeveloping countries composed mostly of rural-based agricultural households, with significantnumbers of poor. Each country also has a colonialhistory in which the struggle for independence ledto a unitary form of government with a strongcenter. Communicable diseases are the main causeof morbidity and mortality in all these countries,although each is now experiencing an epidemio-logical transition that brings growth of non-communicable, lifestyle-related diseases. Yet eachcountry also possesses unique features that provideinteresting contrasts to the other two. Unlike thePhilippines and Indonesia, for example, Vietnamhas adopted a market-oriented economic policyonly recently while retaining socialist features in itsgovernment structure. Unlike Vietnam, Indonesiaand the Philippines are archipelagos that are insu-lated to a degree from disease transmission acrossland borders.

The available data limit comparison betweenthese countries. Official statistics on health, demo-graphics, government finances, and other socio-economic indicators vary in scope, detail, and qual-ity. The same applies to secondary sources ofinformation. Data constraints also add to themethodological challenges of tracing the impact ofdecentralization on health amid other socioeco-nomic factors, external conditions, and policy inter-ventions. Thus, the analysis draws only broad con-clusions and policy guidelines.

The analysis suggests that decentralization divi-dends so far have proved modest and concentratein some areas of each country. Decentralizationmay have helped sustain overall improvements inhealth status and spurred local initiatives in healthplanning, service delivery, and financing. However,

decentralized arrangements have not worked aswell as hoped, especially regarding access to high-quality health services for the poor.

The Origins of Decentralization

The fact that these countries have experienced onlymodest health gains from decentralization reflectscircumstances outside the control of policy makers.In particular, these countries introduced decentral-ization in less than favorable economic and politi-cal environments. For example, the 1997 Asianfinancial crisis underscored the direct link betweenmacroeconomic performance and health expendi-tures. Before 1998, the Philippine economy—muchlike that of Indonesia and Vietnam—was growingsteadily. Per capita income in the Philippines rosefrom US$2,310 in 1985 to US$3,870 in 1997. Fol-lowing the outbreak of the crisis in late 1997, percapita income fell to US$3,730. The impact of thecrisis in Indonesia was graver. Its per capita incomedeclined from US$3,030 in 1997 to US$2,580in 1998; by 2001 per capita income was US$2,900,still lower than before the crisis. Vietnam was lessaffected by the crisis, although per capita govern-ment health spending leveled off and may havefallen after the crisis.

In the Philippines, the crisis prompted thenational government to invoke “an unmanageablepublic sector deficit”—a provision under the LocalGovernment Code of 1991 that allows a 10 percentcut in the Internal Revenue Allotment (IRA) dis-tributed to lower levels of government.2 Becausemost local governments depended heavily on theIRA, the reduction further reduced local healthspending, especially among provinces and munici-palities, which had absorbed the bulk of devolvedhealth functions.

Uneven regional growth aggravated the situa-tion. In the Philippines, provinces in the EasternVisayas and Northern Mindanao continued to lagbehind other provinces, especially those in Metro-politan Manila and surrounding provinces. InVietnam, the cities of Hanoi and Ho Chi Minh weredeveloping faster than other areas, and similarunevenness existed in Indonesia. Utilization ratesand other indicators of health access therefore var-ied widely across regions in all three countries.

Indonesia and the Philippines also implementeddecentralization amid considerable uncertainty fol-lowing political crises. After the fall of the Marcos

Decentralizing Health: Lessons from Indonesia, the Philippines, and Vietnam 157

east_155-178.qxd 5/17/05 11:42 AM Page 157

regime, the Philippines ratified a new Constitutionin 1987 and further articulated strong decentralistprovisions in the Local Government Code of 1991.In Indonesia, the overthrow of the Soeharto regimein 1998, and then de facto secession of the erstwhileprovince of East Timor in 2000, contributed to theclamor for decentralization. The Philippines expe-rienced several military uprisings after 1986, themost recent in late 2003, and has had four presi-dents and nine secretaries of the Ministry of Healthunder the present Constitution. These frequentmusical chairs in the health ministry have disruptedpolicy priorities and the ministry’s momentum inadapting to a decentralized setting. Persistent ruralinsurgencies and kidnappings in areas such asSouthern Mindanao have made it difficult for boththe private sector to pursue investments and thepublic sector to reach out to the poor.

Weak governance in the Philippines, includingcorruption in key branches of government, has alsoled to loss of revenues and waste of limitedresources. Mechanisms like Health Boards andother local consultative bodies have seldom beenconvened for counsel or feedback, contrary to theintent of the Local Government Code (World Bank2000a). However, the proliferation of nongovern-mental organizations (NGOs) and other civil soci-ety groups has been a major positive development.Many such organizations now work side by sidewith key national agencies in agrarian reform,health advocacy, local capacity building, livelihoodprojects, community mobilization, and governancereform.

A second set of reasons for the modest gainsfrom decentralization in these three countriesrelates to weaknesses in policy itself. While externalfactors limited the potential benefits, better man-agement by the central ministry of health, espe-cially in critical health functions, would havehelped. Experiences in the three countries suggesthow to define and pursue an effective role for thecentral health ministry.

The next section analyzes the features andimplementation of each country’s decentralizationpolicy. The following section examines the impactof decentralized health services on health status,service coverage, overall efficiency, and equity inthese three countries. Ensuing sections examineintergovernmental fiscal challenges, personnelmanagement, and service delivery under the

decentralized health care systems in these threecountries. The final section summarizes findingsand draws lessons regarding the role of the centralhealth ministry in managing the health sector.

Health Policy under Decentralization

Consistent with their respective constitutions, thethree countries passed legislation that enabled, if notmandated, the decentralization of health services.Besides added administrative powers and responsi-bilities, local governments attained greater fiscalautonomy through higher shares of national gov-ernment revenues and expanded taxation powers.

In Indonesia, the principal enabling legislativeacts were Law 22 and Law 25 of 1999, while Regula-tion 25 of 1999 facilitated implementation. ThePhilippines promulgated decentralization throughthe Local Government Code of 1991, implementedthe following year.

In Vietnam, the doi moi economic reforms thatbegan in 1986 and the Public AdministrationReform of 1995 shaped health decentralization,with implementation based largely on the 1996 and2002 State Budget laws. The latter two measuresbrought fundamental changes in the preparation,approval, and execution of budgets for all govern-ment agencies, from the central to local levels. Since2004, province-level People’s Councils have hadmore authority to prioritize expenditures anddetermine sectoral allocations and transfers tolower tiers, and stronger means of mobilizingresources. Transfers from the center for stable peri-ods of three years will promote local planning,while provinces must produce forward-lookingexpenditure plans in return.

Decree 10, another element in Vietnam’s legalunderpinning for decentralization, took effect inJuly 2002. When fully implemented, this decreewill give managers of facilities much greater con-trol over their budgets, and more (though still lim-ited) discretion regarding pay and employment,user charges for nonbasic services, and domesticborrowing.

In each country, later laws further articulated,directly supported, or affected the decentralizationof health services. In the Philippines, for example,these laws included the Magna Carta for PublicHealth Workers of 1992, the Barangay Health


east_155-178.qxd 5/17/05 11:42 AM Page 158

Workers’ Benefit and Incentives Act of 1995, andthe National Health Insurance Act of 1995. InVietnam, the Seventh Communist Party Congresspassed a resolution to broaden the “scope of respon-sibilities and power of the sectors and localities,”and passed the Grassroots Democracy Decree in1999 (Communist Party of Vietnam 1993; Govern-ment of Vietnam 1999).

Main Design Features and Implementation

At first glance, the division of responsibility forcritical health functions between the national andlocal governments in these countries broadlyreflects efficiency principles. That is, local govern-ments have assumed responsibility for health func-tions that are simple to administer or confer local-ized benefits. The central government or higherlocal governments have assumed responsibility forhealth functions with significant economies ofscale or interjurisdictional spillovers. For example,basic, primary health care services are assigned tocommunes in Vietnam, including the network ofvillage health workers, to villages in Indonesia, andto barangays (villages) in the Philippines. Primary-level health facilities are assigned to cities andmunicipalities in the Philippines and to districts inVietnam. Secondary-level hospitals are assignedto provinces in the Philippines and Vietnam.Tertiary-level and specialty hospitals, on the otherhand, are mainly the responsibility of the centralgovernment—that is, the central heath ministry—in all three countries.

Central governments continue to provide cer-tain public goods such as health research and devel-opment, and merit goods such as maternal, childcare, and family planning services. Local govern-ments are often involved in and sometimes cofi-nance these programs. However, overall, the devo-lution of health functions and corrective measuresreveal flaws.

Decentralization occurred gradually in Vietnamand not without setbacks. Local mobilization wasseen as a key element in the country’s impressiveachievements by the mid-1980s in delivering pri-mary health care. As mentioned, the combinedprovince and commune share of governmenthealth outlays was already significant in the early1990s. Thus, local officers had experience withdecentralization when the 1996 State Budget Law

assigned additional health tasks to provinces anddistricts (Fritzen 1999). The law established finan-cial links underpinning a unitary system in whichnational authority is delegated to lower levels. Ateach level, budget preparation and implementationare the responsibility of the People’s Council.

In contrast, implementation in the Philippinesand Indonesia occurred in Big Bang fashion. Theformer completed the transfer of 45,896 healthpersonnel, along with hospitals, clinics, and otherfacilities, in 1993, two years after passing the LocalGovernment Code. Indonesia completed a similartransfer in 2001, less than two years after enactingLaws 22 and 25.

The Big Bang approach has its merits, but expe-riences in Indonesia and the Philippines reveal itsdisadvantages. In Indonesia, decentralization lawsand rules and regulations do not provide enoughdetail on functional and operational responsibili-ties, resulting in confusion and divergence betweenprovinces and districts. For instance, provinces aresupposed to handle cross-district tasks, but nodefinitive finding tells them how to apply that rule.The laws and regulations governing decentraliza-tion are also often inconsistent with other laws,especially civil service rules. This inconsistency haslimited the ability of local governments to right-size inherited health bureaucracies and anticipatepersonnel matters.

Moreover, administrative preparation was inad-equate. For example, many local officials in thePhilippines were unaware of the precise nature andextent of their new expenditure responsibilities andpowers, and the central Department of Health(DOH) was slow to transform itself structurallyand operationally.3 Lack of personnel severelyhampered the Local Government Assistance andMonitoring Service, created to troubleshoot transi-tion problems, and the service lacked clout, as dif-ferent DOH divisions managed public health pro-grams as before. With DOH looking uncertain,many local governments seemed to adopt a wait-and-see strategy, apparently hoping that the agencywould be blamed for the breakdown in the publichealth system and be forced to recentralize healthfunctions.

In different degrees, these three countries alsointroduced local governance mechanisms to pro-mote transparency, accountability, and participa-tion as they devolved health services. This, of


east_155-178.qxd 5/17/05 11:42 AM Page 159

course, complicated the transition, as local govern-ments initially had to adopt these mechanisms ontheir own without much guidance or experience,leading to delays, perfunctory compliance, or fail-ure to convene the mandated consultative bodies.

Health Dividends under Decentralization

On the whole, each country sustained favorabletrends in overall health status after decentralization(see table 8.1). In the Philippines, gaps in healthstatus across regions continued to close during the1990s. Measured as the difference in infant mortal-ity rates (IMR) between the poorest region (EasternVisayas) and the richest region (MetropolitanManila), the gap narrowed from 15 in 1980 to 9.8 in1985. By 1990 the gap was almost zero, with theIMRs of Metropolitan Manila and Eastern Visayas27.4 and 27.1, respectively. In 1995, the gapremained near zero, although the IMRs of Metro-politan Manila and Eastern Visayas improved to21.3 and 21.6, respectively. Following the 1997Asian financial crisis, however, the disparitywidened. By 2000 the IMR of Metropolitan Manilawas 19.4—worse than the 10.7 of Eastern Visayas.4

Each study country also experienced an epi-demiological transition in the 1990s, in which theincidence of chronic, lifestyle-related diseases like

cancer and heart diseases began to match—if notovertake—that of communicable diseases such astuberculosis and malaria (Solon et al. 1999).

In Vietnam, disparities in survival rates betweenregions appear to have widened in the late 1990s,captured in the rising ratio of highest to lowestregional IMRs by region. After growing from 1.7 in1989 to 2.3 in 1994, this ratio rose to 3.6 in 2002.This is not to imply that rates and underlying con-ditions were static. On the contrary, IMRs them-selves fell by at least half in every region betweenroughly the early 1990s and 2003. However, thedecline in these changes was extraordinary, drop-ping to a third or less of the early 1990s figure in theMekong, Central Highlands, Southeast, and Cen-tral Coast regions.

Some Progress in Health Outputs and Access

The favorable trend in overall health status wasarguably due partly to progress in health outputsand service coverage. In the Philippines, for exam-ple, the proportion of births attended by trainedhealth workers, and of the population with accessto clean water source and sanitation services, rosein the 1985–2000 period.

Similar developments in health status, outputs,and access indicators occurred in Indonesia overthe same period. The 2002 Demographic Survey


TABLE 8.1 Selected Health Status Indicators

Year

Indicators 1980 1985 1990 1995 2000 2001 2002

Infant mortality rate(per 1,000 live births)Indonesia 79 70 60 46 35 33 1.9Philippines 65 55 45 36 30 29 1.9Vietnam 50 43 36 32/30 28/18 30/18 1.9Under-5 mortality rate(per 1,000 live births)Indonesia 125 108 91 66 48 45 1.9Philippines 81 74 66 51 40 38 1.9Vietnam 70 60 50 43 34 38 1.9Life expectancy at birthIndonesia 55 59 62 64 66 66.3 66.7Philippines 61 63 66 68 69 69.5 69.8Vietnam 60 63 65 67 69 69.4 69.7

Sources: World Bank 2002; WHO 2002.

east_155-178.qxd 5/17/05 11:42 AM Page 160

and Health Survey pointed to a continuation oreven an acceleration of favorable trends in fertility,contraceptive use, malnutrition, and trained mater-nal care. Some indicators worsened: immunizationrates fell between 1997 and 2002–3 for childrenunder age two, for example, while the prevalenceof childhood illness remained the same as in 1997.In Vietnam, on the other hand, output and accessmeasures all pointed in a positive directionbetween the mid-1990s and 2002. For example, thecountry reported a significant increase in child-hood vaccination coverage, and in the proportionsof women receiving prenatal care and giving birthattended by skilled health personnel (Committeefor Population, Family, and Children 2003).

Health Expenditures

Decentralization may have more than sustainedmomentum in improving health status and evenreversed worsening trends. Unfortunately, availabledata do not allow us to verify these two supposi-tions, nor do input measures such as health expen-ditures enable definitive conclusions.

According to the World Development Report2004, annual health expenditures remained moreor less a constant proportion of GDP throughout1997–2001 in the three countries. The averageannual proportion was 2.5 percent in Indonesia,3.5 percent in the Philippines, and 4.9 percent inVietnam. In per capita terms, however, total healthspending fell in Indonesia from US$26 in 1997 toUS$16 in 2001, and in the Philippines from US$41in 1997 to US$30 in 2001 (World Bank 2004).5

Asia’s financial crisis led to a steep decline in 1998in health spending in these two countries: 50 percentin Indonesia and 24 percent in the Philippines.Seemingly immune to the financial crisis, Vietnam’sper capita health spending rose from US$16 in 1997to US$21 in 2001.6 In general, health expendituresas a percentage of GDP in these three countrieswere similar to those of most of their neighbors.For example, the average percentage share of healthexpenditures in GDP in Thailand and Malaysia was3.7 and 3.2, respectively.

Meanwhile, the public sector share of totalhealth expenditures in each of the three countriesdid not change much between 1997 and 2001.Indonesia’s public sector accounted for roughly afourth of total health expenditures. In Vietnam, the

public sector share fell slightly from 31.5 percent in1997 to 28.5 percent in 2001. In the Philippines,the public share rose from 43 percent in 1997 to45 percent in 2001.

A closer look at public sector outlays reveals ashift in the financing burden from the central tolocal governments. Most local governmentsdevoted health spending to hospital and personalcare services, much like the pattern before devolu-tion. This is understandable, as local governmentsabsorbed many hospitals under decentralization.However, this orientation may be inappropriategiven the high prevalence of communicablediseases and high relative cost of hospital-basedinterventions.

The Philippines certainly saw such a shift infinancial burden. The annual share of local govern-ments in public health expenditures climbed upfrom less than 5 percent before 1992 to 12.5 percentin 1993. By 2001, the local share reached 20.9 per-cent, exceeding the 16.6 percent share of thenational government. Moreover, personal care serv-ices constitute the bulk of public expenditures forhealth in the Philippines, and, ominously, a grow-ing portion of the health outlays of local govern-ments as well (Solon et al. 1992).

In Indonesia, regional governments now accountfor most routine spending, while developmentspending at the regional level grew fourfold. How-ever, central development outlays rose almostthreefold, and nearly half of development expendi-tures still come from the central budget. Under-standably, local governments continue to regard theMinistry of Health as a key supplier of financialresources as well as personnel, equipment, drugs,and vaccines.7

In Vietnam in the early 1990s, subnational gov-ernments, including those at the commune level,were already spending more on health than thecentral government (Knowles et al. 2003). Provin-cial spending accounted for 68 percent of govern-ment health expenditures in 1991 (not includingrevenues from user fees and donor support), whilecentral spending accounted for 13 percent. How-ever, five years later, after the country passed theLaw on State Budget, provinces accounted for53 percent and central units 26 percent of allgovernment outlays. By 2000 the latter figure hadfallen to 17 percent, while the province-level share(narrowly construed) had dropped to 44 percent.


east_155-178.qxd 5/17/05 11:42 AM Page 161

However, the thrust of the 1996 legislation wouldseem to suggest including revenue from healthinsurance and user fees in the provincial total. Inthat case,provincially“controlled”outlays accountedfor 76 percent of total government health spend-ing, up from 70 percent in 1996. (Donor outlaysare treated as a separate category influenced byparticular agendas and criteria.)

In light of these financing and spending pat-terns, it is doubtful that decentralization haswidened access by the poor to quality health care. Anational client survey confirmed that Filipinos ingeneral were more satisfied with private hospitalsand clinics than with government health facilities.Filipinos also tended to rate traditional healers asmore satisfactory than any public providers (WorldBank 2001a). The low regard for public health serv-ices prevailed even among the poor, an indicationthat the public health system does not serve its tar-get clients well.

A World Bank study of socioeconomic differ-ences in health, nutrition, and population inselected developing countries corroborated theseobservations (see table 8.2). In the Philippines,children born in 1998 to the poorest families weretwice as likely to die within a year as children bornto the richest families. The infant mortality rate forthe poorest families (48.8) was 1.7 times that of therichest families (28.8). The life chances of theseunfortunate Filipinos did not seem to improve withage: under-five mortality rates were 79.8 and 29.2for the poorest and richest families, respectively.

Vietnam also had disparities in health statusacross economic groups, resembling those in

Indonesia before decentralization. Like their coun-terparts in the Philippines, the poorest familieshad infant and under-five mortality rates severaltimes higher than those of the richest families.Discrepancies in health status between thepoorest and richest households appeared to beworse in Indonesia than in Bangladesh, India,the Philippines, and Vietnam.

Perhaps because of the inferior quality of publichealth services, the poor—like well-off fellowFilipinos—continue to self-finance their access toprivate health services. Private sources, includingdirect out-of-pocket payments, accounted for anannual average share of 57 percent of total healthexpenditures in the Philippines from 1991–2001.

Local Initiatives in Health Services and Financing

Decentralization has given local authorities andother stakeholders greater leeway to adapt or evenreplace once-standard methods for delivering andfinancing health services. And these greater discre-tionary powers have led to numerous local innova-tions in health planning, service delivery, andfinancing. Most notably in the Philippines, thereare the provincial health insurance programs ofBukidnon and Guimaras, as well as other “text-book” cases, such as the health card system ofParanaque City, the City in the Pink of Healthprogram of Marikina City, and the CommunityPrimary Hospital Program of Negros Oriental(Pineda 1998; Bautista et al. 1999; Quimpo 1996;and Legaspi 2001). Several of these programs havereceived formal recognition from government


TABLE 8.2 Health Status of the Poorest and Richest Population Groups in Selected East Asian Countriesa

Infant mortality rate Under-5 mortality rate(per 1,000 live births) (per 1,000 births)

Country Poorest/ Poorest/(year) Poorest Richest richest Poorest Richest richest

Bangladesh (1996–7) 96.3 56.6 1.701 141.1 76.0 1.857India (1992–3) 109.2 44.0 2.482 154.7 54.3 2.849Indonesia (1997) 78.1 23.3 3.352 109.0 29.2 3.733Nepal (1996) 96.3 63.9 1.507 156.3 82.7 1.890Philippines (1998) 48.8 20.9 2.335 79.8 29.2 2.733Vietnam (1997) 42.8 16.9 2.533 63.3 23.0 2.752

Source: www.worldbank.org/hnp. a. Economic groups are based on asset (wealth) quintiles.

east_155-178.qxd 5/17/05 11:42 AM Page 162

agencies and private bodies such as the GalingPook Foundation (see box 8.1) and the PhilippineHuman Development Network.8

Vietnam has seen numerous instances ofsponsored and spontaneous innovation at theprovince level. An example of the former was thehealth ministry’s effort to encourage local responsesto childhood diseases, including community-determined indicators (Fritzen 1999). Reactions toHIV/AIDS illustrate the spontaneous case. As inseveral other provinces, the epidemic spurred theThanh Hoa government to pursue preventiveactivities such as harm reduction and 100 percentcondom use. These initiatives resulted from strongcommitment by the People’s Committee. Besidesensuring the participation of the police (Depart-ment of Public Security), the committee allocatedan annual budget to fight HIV/AIDS. The com-mittee also organized a provincial Steering Com-mittee on HIV/AIDS headed by the Departmentof Health, under an umbrella Steering Committeeon HIV/AIDS, Drugs, and Prostitution Control

chaired by the vice-chair of the People’s Commit-tee. District and communes adopted the samestructures.

In Indonesia, Yogyakarta province showed howto use the country’s still immature decentralizedframework to introduce health sector reforms andelements of a health insurance system. Theprovince used donor funds to secure technicalassistance and conduct assessments, trials, bench-marking, workshops, training, and coordinationmeetings with districts, as well as advocacy events.The province established a board of trustees andnew fund-holder institutions, as well as a benefitpackage and an insurance premium.

The province has created a plan for a qualitycouncil to accredit facilities and license practition-ers based on local standards. Meanwhile, theprovince increased user fees under local control toreflect the actual unit costs of providing services.Task forces developed strategies for improvingservice quality based on consumer surveys; pur-chasing equipment; developing accountability


BOX 8.1 Local Innovations in Health Service Delivery in the Philippines

Charging User Fees for Health Services in Malalag, Davao del Sur

In December 1993, the local Sangguniang Bayan(municipal council) of Malalag, in the provinceof Davao del Sur, enacted the Malalag RevenueCode. This code established a socialized feeschedule for health services, among other provi-sions. The graduated payment scheme reflectsusers’ annual family income: those earningP=15,000, from P=15,000 to P=50,000, and morethan P=50,000 pay 25 percent, 50 percent, and100 percent, respectively, of fixed servicecharges. The code also gave low-income familiespriority in receiving health services. Public con-sultations, hearings, and an information andeducation campaign overcame initial resistanceto the scheme. Partly as a result, the local gov-ernment earned about P=1 million worth of feeson an outlay of P=688,888. This enabled the gov-ernment to provide additional health services,including surgical, medical, and dental services.With these improvements, the local clientelehave become more demanding of the quality ofhealth services and the performance of healthpersonnel.

Transforming a Rural Health Center into aCommunity Clinic in Sebaste, Antique

Under Mayor Juanita de la Cruz, Sebaste inAntique—a remote sixth-class municipality—became a prime example of how to transform abasic rural health center into a full-service com-munity clinic despite limited resources. With onlyP=800,000 in IRA funds from the center, the gov-ernment tapped foreign donors, local people, andformer residents living abroad for support for itshealth goals while also appealing to the sense ofmission of health personnel. After creating a trustfund, the government infused P=3.085 million intothe project from 1994 to 1998. By 1997, thecommunity clinic employed 16 people, includingtwo physicians, and remained open 24 hours aday, providing primary health care, laboratoryand pharmacy services, and minor surgery. Theclinic has reduced the cost of these services to thelocal clientele while also serving the medicalneeds of residents of neighboring municipalities.

Source: Galing 2001.

east_155-178.qxd 5/17/05 11:42 AM Page 163

mechanisms based on focus groups and thecomplaint resolution system established duringthe financial crisis; and improving health work-force management, and submitted a tighter organi-zation structure to the provincial government forconsideration.

The province launched the new system in 2003by paying the premium to enable the poor to usepublic facilities. The program became available tothe nonpoor in 2004, competing with privateproviders in providing a benefit package. Theapproach in Yogyakarta gives districts a key role, asspecified in Law 22, but also responds to the loss ofeconomies of scale that may make health servicesineffective and inefficient. The province encouragescross-district collaboration, especially in upgradingtechnical support, sharing medical and technicalspecialists and trainers, and organizing communi-cable disease control, quality assurance, and healtheducation and advocacy. The Joint Health Councilfacilitates such activities, with task forces makingrecommendations. But the provincially staffedTechnical Review Team plays the largest role byreviewing district proposals and providing feed-back and guidance.

A number of other provinces are closely watch-ing and applying this approach. Central agencieshave not been deeply involved, although Yogyakartasought their guidance on establishing standards forits regulatory framework.

Overall, the health impacts of decentralizationare not easy to estimate. Few data indicate signifi-cant windfalls in health benefits linked to decen-tralization. The early phases were not incompatiblewith sustaining impressive overall improvements inhealth status, and decentralized governance openedthe way to promising local initiatives in healthplanning, service delivery, and financing. However,much better results would seem to be within reachthrough policy adjustments.

Identifying appropriate policies—though noteasy—is critical. In the Philippines, central agencies,often in partnership with NGOs, have documented,disseminated, and advocated best practices in localpublic services through the media, educational tripsfor local officials, and various training programs.Despite these initiatives, however, the speed ofinnovative practices has been limited, and the over-all level and quality of local health services havebarely improved. A lack of incentives rather than

missing models—including political dividends andother signals—appears to be holding back neededpolicy interventions by local decision makers.

Health Care and IntergovernmentalFiscal Challenges

A country’s intergovernmental fiscal system shouldmeet the complex goals of its health system, as withother public services. The key to an effective fiscalsystem is “finance follows function.” The intergov-ernmental fiscal system must also usually addresshorizontal as well as vertical equity, key relation-ships between levels of government and jurisdic-tions, and incentives for collaboration. Weaknessesin the design of each country’s fiscal system havehad important consequences for the delivery ofhealth services.

In Vietnam, the budgeting and financing systemformalized in the 1996 Budget Law revealed suchchallenges:

• Norms emerged during the 1990s to determinealmost every kind of input into the system. Forexample, gaps between provinces in per capitahealth spending reflect a complex fund alloca-tion based on population norms and allocationsto and within sectors. These take into accountdiffering geographical conditions betweenprovinces but do not offset revenue and cost dis-advantages and variations in need. The Ministryof Finance may also rely on other criteria duringthe budget negotiating process.

• Provinces have substantial discretion in allocat-ing resources to districts and communes, andthe methods they use vary considerably. Dis-tricts have little autonomy, and interdistrictvariations are significantly greater than at theprovince level.

• The norm-based system relies on flows of accu-rate data. For example, user fees and insurancereimbursement rates reflect costs and expendi-tures in different settings. Besides the high costof regularly updating such information, thesefigures are unlikely to capture variations linkedto scale and quality as well as discretionary ele-ments. What’s more, because the steps to createfee schedules are cumbersome, time consuming,and costly, the schedules remain in effect foryears; that in use today dates from 1995.


east_155-178.qxd 5/17/05 11:42 AM Page 164

• Despite formal autonomy at lower levels of gov-ernment, the norm-based system constrainsflexibility. Civil service salaries have first call onfunds and absorb most expenditures. Provinceswith fewer local revenues have less flexibility.

• The norm-based process limits sectoral inter-ventions. The Ministry of Health did not fullyparticipate in budget discussions and lackeddetailed information on expenditures byprovinces and lower levels of government. Theministry also could not assess whether actualspending by lower levels was consistent with sec-toral policies.

• Limited investment in local facilities, such asupgrades to community health centers, also af-fected quality of care, and more patients bypassedsuch facilities as a result.

• The Ministry of Health focused on allocatingfunds to national programs aiming at combat-ing high-priority diseases such as tuberculosis.Some of these programs were particularlyimportant to disadvantaged groups and imple-mented mainly through local governments.However, no mechanisms ensure that oncenational goals are met, those programs arediscontinued.

• Reliance on norms has also discouraged creationof medium-term planning frameworks thatfacilitate recognition of tradeoffs and set priori-ties between and within sectors.

The December 2002 budget law, which tookeffect in January 2004, gives more discretion to sub-national governments. Province-level People’sCouncils have more power and a greater obligationto prioritize health spending, determine allocationsand transfers to lower tiers, implement policy, andmobilize resources. The fact that the central gov-ernment establishes three-year transfers once itreaches agreement with provinces on expenditureplans may allow the Ministry of Health to influenceallocations across functions and service levels.

In the Philippines, the primary fiscal vehiclesupporting decentralization is the Internal RevenueAllotment (IRA), which transfers funds to localgovernments. As noted, most local governmentsdepend heavily on this source, as do devolvedhealth services. The central Department of Health(DOH) also created the Local Government Assis-tance and Monitoring Service to manage transition

problems, and to provide financial assistance tolocal governments unable to maintain health serv-ices or meet their Magna Carta obligations becauseof inadequate resources. DOH also implemented aconditional matching grant program, the Compre-hensive Health Care Agreements, intended tosecure local funding for devolved functions andcore public health programs. This reflected animportant part of the country’s strategy of usingincentives and disincentives to achieve nationalobjectives in a decentralized system.

However, the relationship between service deliv-ery and financing arrangements entailed significantweaknesses. For example, devolution of publicfacilities led to fragmentation of the hospital refer-ral system. Under the new regime, each hospital orclinic primarily serves the constituency of a localgovernment. Several provinces therefore reducedbudget appropriations to urban hospitals andchanneled resources to less-well-off municipalities,in the process raising the average cost of urbanservices. Instead of cofinancing these facilities withthe provinces, many cities opted to refurbish theirown clinics or build enclave hospitals. Further,weak monitoring of local compliance with Com-prehensive Health Care Agreements did not helpensure financing of the devolved services.

Ensuring Equity

In Vietnam, two factors have undermined the dis-tribution of heath services to the poor. First, thecentral government has not targeted resource flowsto poorer provinces, concentrating instead on thesupply side by improving multitiered service deliv-ery. The government has taken demand for serviceslargely for granted and has not weighed it heavily inpolicy making, at least until recently.

Second, longstanding funding shortfalls contin-ued through the decentralization process. Startingin 1989, hospitals in Vietnam were allowed to col-lect user fees and mark up drug prices, and theresulting revenues became, and still are, a sizablesource of health financing. However, user fees werea disincentive to enhanced utilization by the poor.And with user fees only partially offsetting fundinggaps, lower quality followed. All this led to reduceduse of health services from the late 1980s, withdemand often shifting to “private” providers rang-ing from retired government doctors to informal


east_155-178.qxd 5/17/05 11:42 AM Page 165

drug vendors. These developments likely exacer-bated variations in health indicators by region andincome group, with poorer areas such as the north-ern uplands recently falling further behind.

Some cities and provinces reportedly reduceduser fees charged to the poor and other groups.Recent findings show that distribution of centraland local budget funding, official developmentassistance (ODA), and health insurance reimburse-ments among provinces benefit the poor dispro-portionately more than do other sources ofprovince-level funding. However, only the ODA isstrongly propoor. Neither central and local budgetfunding nor ODA relates significantly to province-level measures of health needs, household povertyrates, or the percentage of minorities. The distribu-tion of public health expenditures among provincesis weakly propoor, thanks largely to the state budgetand ODA (Knowles et al. 2003).

Decision 139, issued in October 2002, furtherrequires each province to set up a Health Care Fundfor the Poor to finance free health care for disad-vantaged groups, with budgetary support from thecentral government. Decision 139 entails a majorincrease in health spending in Vietnam amountingto D 700 billion (some US$0.5 billion) per year.This program is starting slowly to allow the countryto overcome difficulties in identifying the poor andchanneling funds to poorer provinces. The pro-gram may improve health access in remote areas asit does not cover the indirect costs of care. Alongwith Decree 10, which gives hospital managersmuch greater control over their pay and employ-ment, user charges, and use of surplus funds, Deci-sion 139 represents a shift in that the Ministry ofHealth is moving from direct service provider tosectoral steward, directing central resources to thepoor and other vulnerable groups based on cleardefinitions of eligibility. The directives also implydifferentiation of government roles, with provincialhealth departments organizing delivery of care andVietnam’s insurance agency responsible for collect-ing contributions and purchasing services.

This change is important because while themain clients of devolved health services in all thesecountries are the rural poor, their access to qualityhealth services is highly uneven owing to wide vari-ation in local revenues and the flawed design offiscal transfer programs. In the Philippines, forexample, economic growth remains uneven across

regions, and only cities have generally robusteconomies. Most provinces and municipalities relyheavily on fiscal transfers, principally the IRA.However, the IRA formula favors highly populatedlocal governments and those with large land areas,and so does not ensure an overall propoor bias inhealth services. Studies have also shown that otherfiscal transfers, including those administered by thecentral Department of Health, correlate onlyweakly with poverty, with poor regions appearingto have received lower DOH budget allocations in1994 and 1997. On the other hand, one of the rich-est regions in the country receives a disproportion-ate amount of the DOH budget (Mercado 1999;Capuno 2002).

As in Vietnam, government hospitals in thePhilippines may collect user fees and impose up toa 30 percent markup on drugs. However, cost-recovery rates remain low because of the inordinatevolume of charity and subsidized patients. In thecase of provinces, for instance, the combined shareof hospital fees never reached 13 percent of totalhospital outlays from 1992 to 2000. Municipalitiesfared better, with the share of fees in total hospitaloutlays rising from 9 percent in 1992 to 29 percentin 2000 (Capuno 2002). But because most hospitalswere devolved to provinces, the unintended resultof low cost-recovery rates is that many hospitals arepoorly maintained, understaffed, and ill-equipped.

Vietnam and the Philippines instituted healthinsurance schemes in 1993 and 1995 that target thepoor. Health insurance in Vietnam has become asignificant financing source, more important thanin the Philippines (Knowles et al. 2003). Neverthe-less, in Vietnam coverage is still low and mainlyincludes civil servants and others employed in theformal sector. Decision 139 represents a potentiallysignificant scaling up of the number of people withinsurance.

Records from the Philippine Health InsuranceCorporation (PHIC) show that among thoseenrolled in the national health insurance program,the number of paying members, from both thepublic and private sectors but excluding the insuredindigent families, grew from about 5.57 million in1999 to 7.62 million in 2001. This suggests thatnearly four in eight Filipinos have social insurancecoverage, but that the program is still far fromachieving its target of universal coverage. However,since 2000, the PHIC has been aggressive in


east_155-178.qxd 5/17/05 11:42 AM Page 166

enrolling indigent families under its Medicare parasa Masa (indigent program). As a consequence, thetotal number of indigent families enrolled hasgrown from 2,904 in 1997 to 1,762,116 in 2003. ByJune 2004, the total ballooned to 6,175,651 indigentfamilies. Whereas in 2001 about 37.8 percent ofthese indigent members were concentrated in therichest regions, by 2003 the same regions accountedfor only about 19.91 percent of the total member-ship. The fact that other regions have gained signif-icance suggests that a wider set of poor householdsnow enjoys coverage. However, this trend is likely toslow as more local governments must copay withthe national government the insurance premium ofpoor constituents. Most local governments see thiscontribution as another unfunded mandate.

Health Care Personnel and CivilService Management

Many local governments find it difficult to hirephysicians, nurses, and medical technicians, whoare in great demand in foreign markets. In thePhilippines, for example, local governments inmany areas where tuberculosis is epidemic havefound it difficult to hire medical technologists andrural physicians. Indeed, staff anxiety and opposi-tion were major problems during the transition todecentralization in Indonesia and the Philippines.Though usually temporary, staff discontent canaffect the quality and quantity of personnel avail-able under decentralization.

In the Philippines, health workers were perhapsthe largest group opposing decentralization. Manyinitially feared loss of job security, “politicization”of their functions and positions, limited careerprospects, and lower pay.9

To appease devolved workers, the centralDepartment of Health pushed for the Magna Cartafor Health Workers in 1992. Among other features,this law provides for higher compensation andextra benefits and allowances to all health workers,including those devolved to local governments, andrequires the latter to pay the additional compensa-tion. This unfunded mandate would have demoral-ized other staff and made some rural physicians thehighest-paid local public employees, earning morethan mayors, which they considered unacceptable.

As a stopgap measure, the Department of Healthinstituted the Doctor to the Barrios Program,

which supplied temporary, contractual, and better-paid doctors to remote areas. In May 1993 theprogram began to deploy physicians to 271 munic-ipalities lacking doctors, and by December 2003,198 of these municipalities had received doctors.They receive an attractive package of salary andbenefits for serving two years, and some also receivehonoraria and material support, such as free boardand lodging, from local governments. However,only about a third remain after their two-year tourof duty, discouraged by the lower pay and fewerprivileges that accompany local employment. Thenumber of applicants to the program is dwindlingowing to a surge in foreign demand, and conflictareas remain underserved because of a lack ofincentives.

To supplement the local health force, theBarangay Health Workers’ Benefit and IncentivesAct of 1995 provided for training volunteer work-ers and providing minimal incentives to convincethem to join barangay health stations. These volun-teers assist in clerical tasks and minor health proce-dures, such as weighing and measuring patients.However, these workers do not effectively cater tothe health needs of the population.

In Indonesia, the central government estab-lished the contract doctor (PTT) scheme in theearly 1990s to ensure a flow of doctors to remotelocations. Doctors hired after completing their ini-tial medical degree received substantial monetaryincentives for practicing in more remote areas forthree years, as a condition of advancement. Special-ists also had to complete compulsory assignmentsfor one to four years or two to three years as PTTstaff.

Discontent grew over the obligatory nature ofassignments, relatively low salaries, and pooradministration of program benefits. In 1999, regu-lations were eased to permit alternatives such asteaching in a medical school, working as a PNS(civil servant) in designated areas, or working inprivate practice as a clinic employee in remoteareas. Service requirements for very remote areaswere reduced to two years, and new graduatescould postpone mandatory service if they wantedto start specialist training. These changes did notsatisfy the PTT lobby, and doctors continued topress to scrap the regulations.

PTT issues remained unresolved as decisionmakers launched decentralization. The Ministry of


east_155-178.qxd 5/17/05 11:42 AM Page 167

Health has been exploring new ideas such as allow-ing medical personnel to serve in the military andthe police. Districts, meanwhile, have stayed withthe residual national system despite its flaws, aswithout central funds and guidance they might nothave been able to integrate the numerous centrallyassigned, locally based staff transferred overnightvia Law 22. Moreover, few districts can turn downcentral offers to recruit and assign PTT doctorsusing central funds. Still, district officials are con-cerned that staffing policies that reflect local priori-ties and conditions have not been established,including options to “right-size” staffs within eachdistrict. This issue arises especially in districts obli-gated to handle staff oversupply left behind byflawed centralized-era policies.

Strong political and administrative leaders insome provinces have created master plans toreshape the bureaucracy to fit local conditions.These include using downsizing mechanisms suchas redeployment of staff, early retirement, volun-tary resignation with severance payments, andretraining to encourage entrepreneurship. How-ever, implementation of these plans awaits fullpolitical commitment, facilitating legal steps, andan injection of cash. The inability to proceed high-lights concerns voiced by district and provincedecision makers about dependence on centralgovernment for salary payments and methods for“right-sizing.”

Vietnam confronted personnel issues underdecentralization as well, as the distribution ofhealth personnel did not occur exactly as planned.Enough doctors and other higher-level staff aregenerally available in cities, but numbers are inade-quate in rural areas. Provinces with medical schoolshave about the right number of staff, but poorerprovinces do not, especially newer ones with nosecondary medical schools. One study showed only1.7 doctors per commune in the North Highlandsand the North Central Coast, while a commune inthe Southeast Region averages 6.8 doctors (WorldBank 2001b).

The number of health workers at the provinciallevel is generally adequate (in relation to Ministryof Health guidelines), although some provinces donot have enough specialists. But districts generallylack enough doctors who specialize in priorityareas, such as obstetrics and gynecology and emer-gency surgery. Communes do not have enough

doctors (nearly all at commune health centers areupgraded former assistant doctors), and often lackenough staff with other training as well, except indensely populated delta areas and near cities. Forexample, in 1997, 26 percent of communes lackedan obstetric-pediatric assistant doctor or a midwife.The Ministry of Health requires all communes toretain such an employee, reflecting the high priorityaccorded to local maternal and child care.

Average monthly salaries of health staff haveremained essentially unchanged in real terms since1994. In 1998, the average monthly salary of a gov-ernment health worker was merely US$29, eventhough user fees supplement salaries somewhat.This low pay, compared with the education sector,has induced many health staff to seek additionalsources of income, reducing the time, attention,and dedication they devote to their work (Dunget al. 2001).

Service Delivery Mechanisms

Health programs are prime examples of the need todesign institutional arrangements carefully toensure that parties in the service delivery processhave the understanding, ability, and incentives tofulfill their roles. Coordination is invariably a cru-cial requirement of system effectiveness. While acountry’s central health ministry should take keyresponsibility for controlling communicable dis-eases, it cannot do this efficiently and effectivelywithout the cooperation of local governments, asthe latter are at the forefront of service delivery.

In the Philippines, programs to combat commu-nicable disease depend on devolved health person-nel and local counterpart funds, which are in shortsupply. Local governments see their participation asanother unfunded mandate, and program coordi-nation has suffered as a result. To elicit local sup-port, the central Department of Health under theComprehensive Health Care Agreements (CHCAs)matches each peso that a local government com-mits with a higher amount. However, the local gov-ernment must first commit a minimum amountfor its devolved health functions. This requirementhas proved stringent, as many local governmentsinitially lacked the resources to finance these func-tions, much less meet their contractual obligationsfor vertical programs. DOH also did not develop amonitoring and enforcement mechanism to track


east_155-178.qxd 5/17/05 11:42 AM Page 168

compliance. Worse, many local officials believedthat strict compliance with the program was notnecessary, as DOH would always take the blame forpublic health failures (Esguerra 1997; Medalla1996). Hence, only after two rounds of implemen-tation, CHCAs were discontinued in 1997.

In the three study countries, integration betweenprograms within provinces as well as acrossprovinces remains poor. Subnational governmentsimplement national programs separately, leading tooverlap and overload of grassroots health facilities.

Vietnam partly solved this problem by givingprovinces a greater role in setting goals, developingplans, and using funds for national targeted pro-grams. This is appropriate given variations indisease profiles across regions, and is said to haveraised immunization rates and lowered fatalityrates.

Low Quality and Unsteady Supply of Drugs at the Local Level

The supply and quality of drugs at the local levelhave become a concern owing to limited funds,deficient drug management systems, and loopholesin procurement rules. In Indonesia, provinces havenot been aware of or prone to intervene in drugsupply, stocks, and use at the district level underdecentralization. Districts have been able to planfor and purchase their own drugs based on stan-dard procurement practice. However, compliancewith quality assurance procedures has been poor,partly because responsibilities have not been clearand districts do not have the technical capacity tohandle the task.

In the Philippines, each local government simi-larly manages its own system of drug procurement,inventory, dispensing, and financing. The quality oflocally procured drugs is generally poor, the pur-chase price is often higher than in private pharma-cies, stock shortages are frequent, and irrationaldrug use occurs. A principal reason is that localtherapeutic committees are not constituted, notfunctioning, or not well trained in modern drugmanagement. Local drug procurement is alsocorrupt in many places: bids are rigged, qualifiedbidders are “preidentified,” and bidders connive.Moreover, the supply chain extends only to urbancenters; poor outlying municipalities rely on itiner-ant drug peddlers who arrive infrequently.

To ensure drug quality in all public health facili-ties, the central health ministry in all three coun-tries has adopted drug formularies and drawn upan essential drug list. In Indonesia and the Philip-pines, the central ministry even advocates and pro-motes generic drugs. However, these regulatorymeasures have not ensured the overall quality ofdrugs, owing to weak information campaigns andenforcement mechanisms. In the Philippines, forexample, many local governments, with supporteven from their own health officials, routinely buybranded drugs because of their supposed provenefficacy. Further, the Bureau of Food and Drugs,which lacks laboratory and regulatory capacity, hasnot convinced doctors of the supposed equivalenceof generic drugs (Lim and Pascual 2003). InIndonesia, hospitals buy drugs and unbrandedproducts outside the essential drug list.

Unlike in the Philippines, in Vietnam andIndonesia state-owned enterprises dominate thedrug supply, as they can assure quality more easilythan a private drug market. In Vietnam, the state-owned VINAPHARM, which includes central andprovincial trading and manufacturing enterprises,is responsible for supplying drugs countrywide.The Drug Administration Department within theMinistry of Health is responsible for overall drugmanagement, supported by the Drug Quality Con-trol Institute and the Drug Inspectorate. In eachprovince a Drug Quality Control Department fallsunder the Provincial Health Bureau, while a DrugTesting Center and Inspection Department moni-tor drug quality in the local market.

In Indonesia, four state-owned enterprises pro-duce generic drugs and vaccines. Regulatory func-tions, including enforcement, are the responsibilityof the Directorate General of Food and Drug Con-trol, a unit of the Ministry of Health. Quality assur-ance efforts include establishing the essential druglist; enforcing standards in the development, test-ing, registration, manufacture, and distribution ofdrugs; and overseeing health professionals. Work-ing through 26 province-level branches, the direc-torate monitors drug quality and safety throughfollow-up visits and testing programs. The direc-torate bases inspection of manufacturers on criteriafor good practice adopted in 1971.

State-owned enterprises impose their own inef-ficiencies on the market. In Indonesia, these unitsare protected by tariffs and limits on final-product


east_155-178.qxd 5/17/05 11:42 AM Page 169

imports, constraints on foreign investment, andrestrictions on registering new drugs, opening newpharmacies, and the nonpharmaceutical activitiesof retailers. Reforms adopted in the 1990s relaxedsome restrictions on foreign drug companies,encouraged generic drug prescriptions in publichealth centers, and enforced good manufacturingpractice. Hospitals could also keep drug revenues tosecure supplies at the facility level.

However, inconsistencies and missteps weak-ened or negated these pre-2001 reforms, and theoutcomes were unsatisfactory. Moreover, districtsinherited a flawed and incomplete reform agenda,with impacts on government stewardship obliga-tions. Since Law 22 has taken effect, deviation fromestablished standards, patterns, and procedures hasgrown. For example, complaints about the physicalappearance or expiration date of drugs are wide-spread, suggesting that longstanding quality assur-ance procedures are not being observed. Reportingof drug quality problems is not formalized, andprocedures for addressing quality concerns areunclear and complicated by multiple sources ofdrugs and funding. Limited skills surely play a roleas well, as many service units and district warehousesdepend on unqualified staff. Nor can provinces stepin, as they lack the authority to monitor, yet alonesupervise, district drug procurement.

More generally, laws and regulations providelittle detail on operating responsibilities andhave brought confusion and divergence betweenprovinces and districts. For instance, some pro-curement procedures have spurred small purchasesfrom 15 or more suppliers. Districts usually rejectpooled procurement despite possible cost savings.

Nor is there a definitive view on which drugsbelong in categories defined by the Ministry ofHealth in pre-2001 preparations. Central involve-ment appears to be limited, and provinces reveal nocommon pattern of procurement. Some are notsupplying any drugs, and plan to reduce future drugsupply. Some still buy drugs to cover emergenciesand temporary district shortfalls. Districts are usingtheir own funds to buy drugs from all three classes ofthe essential drug list. Meanwhile, the drug supplyand regulatory system in hospitals is different fromthat at the primary care level. Hospitals, which havelong been allowed to procure and dispense drugsoutside the essential drug list, are buying mainlybranded drugs, funded through self-financingrevolving funds and using spot buying methods.

The change from central to district procurementmay have also increased drug prices because of lowerprocurement volumes. This would likely widen vari-ations in drug prices, reducing equity and loweringthe availability of orphan drugs.

Vietnam represents an interesting comparator.As in Indonesia, deregulation of pharmaceuticalproduction and distribution brought heightenedactivity among informal drug vendors and phar-macy shops and greater availability of drugsthroughout the country. Consumer purchases ofdrugs, especially for self-medication, grew as well,from 2.1 annual contacts per capita with drug ven-dors and pharmacy shops in 1993 to 6.8 by 1998.However, the two countries differ in their experi-ences with drug prices. In Vietnam, deregulationbrought a 30 percent fall in the real price of medi-cines in the 1993–98 period, while in Indonesiaprice reductions do not seem to have followed pol-icy reforms.

Moreover, the risks facing Indonesia are compa-rable to those of Vietnam, where drug vendorsaccount for roughly two-thirds of health servicecontacts, and antibiotic resistance has reached epi-demic levels thanks to excessive and otherwiseinappropriate use.10 The resistance problem iscompounded by the limited training of pharma-cists and the low average education level of drugvendors and the public. Even when doctorsprescribe drugs, compliance with appropriatetreatment guidelines is low. Oversight of healthproviders is weak as well. Enforcement of the manyregulations and decrees governing minimum qual-ity standards and the protocols expected of healthproviders through regular inspections of healthfacilities is less than satisfactory.

In the Philippines, on the other hand, a few drugmanufacturers and importers, which are mostlymultinationals, dominate the upstream segment,while a single drugstore chain effectively controlsthe retail segment of the domestic drug industry. Aparallel drug importation policy has not helpedbring down the overall price of drugs because thegovernment chose to maintain an import monop-oly with capital of just P=50 million—not enough toaffect the multibillion-peso domestic drug trade.

Health Information Systems

Decentralization in Indonesia, the Philippines, andVietnam has fragmented the health information


east_155-178.qxd 5/17/05 11:42 AM Page 170

system and undermined coordination among vari-ous sectors, and thus effective and efficient controlof communicable diseases. Subnational govern-ments are unaware of their roles, and, more criti-cally, lack the incentives and technical capacity toassume those roles. Subnational governments needsupport for activities from collecting health infor-mation to providing further inputs to performingoverall health planning to actually implementingprograms.

From 1992 to 1995, the Philippine Departmentof Health implemented the German-funded Healthand Management Information System, whose mainobjective was to institutionalize a “need-responsiveand cost-effective health information system” at thenational and local levels. The initiative introducedsoftware modules and processes to fortify the pro-duction and use of information. Besides developingdistrict-level health indicators, the system sup-ported innovations in community health carefinancing and service delivery. However, the initia-tive stopped short of a nationally integrated butlocally operated health information system and wasnot sustained.

In Indonesia and the Philippines, the centralhealth ministry relies on local governments toreport information voluntarily. This has resulted inerratic or delayed submission and poor-qualitydata. The devolved staff members who wereresponsible for such data under the old regime nowsupply information on health expenditures andinput indicators to provincial and lower-levelelected officials, who are less concerned with out-puts and outcome indicators.

In Indonesia, only 36 percent of health centersreported infectious disease surveillance data in2002.11 Those that do report do so irregularly orlate. Thus, the limited data that flow through thesystem may not be reliable enough for use in plan-ning, policy analysis, or evaluation.

In these countries, the central health ministry,lacking information, is less able to monitor thequality of laboratory services, hospitals, and otherdevolved services. This is worrisome as, for exam-ple, local health centers in the Philippines areresponsible for both finding cases of tuberculosis,which requires sputum examination, and monitor-ing cases. Indonesia discontinued some programssuch as those tracking leprosy because districts didnot monitor the number of cases. Quality assur-ance systems for provincial hospitals in both the

Philippines and Indonesia continue to rely heavilyon input indicators such as the number of beds,floor area, and medical instruments, with onlyinfrequent verification of such information by localgovernments.

In the Philippines, although the Department ofHealth deploys its own representative to provinces,cities, and municipalities to help monitor diseaseoutbreaks and coordinate vertical programs, theflow of health information remains slow. DOHrepresentatives must often double-up as serviceproviders, as many local governments lack theneeded personnel.

Performance Standards and Incentives

In the Philippines, efforts to improve the quality ofhealth services have relied on both incentives,including awards and accreditation measures, anddisincentives. One example of an approach thatencourages local governments to upgrade theirhealth services is the Sentrong Sigla accreditation.Aside from providing a mark of quality, this accred-itation originally conveyed a P=1 million grant tolocal health centers. Based mostly on input indica-tors of a health facility’s “readiness to provide serv-ices,” the program had certified some 48 percent ofhealth centers, 14 percent of district and provincialhospitals, and 3 percent of barangay health stationsby October 2003. Though these numbers areencouraging, they represent only a minority of thecountry’s facilities. Moreover, most of the cashawardees are better-off municipalities that do notneed the funds as urgently as localities that do notqualify (Lamberte 2003).

In the Philippines, the Department of Healthissued Administrative Order no.100 in 2003, whichestablished new guidelines to improve the SentrongSigla program. Instead of cash awards, the newguidelines specify a matching grant for new quali-fiers, and also make certification a prerequisite forother DOH grants and the Capitation Fund pro-gram, recently introduced by the Philippine HealthInsurance Corp. Under the latter program, anaccredited local government can claim reimburse-ment for services extended to PHIC-insured indi-gent families in their localities. However, theseaccreditation schemes are voluntary.

Vietnam and Indonesia, in contrast, haveadopted norm setting as their approach. Vietnam’sMinistry of Health sets province-level norms, but


east_155-178.qxd 5/17/05 11:42 AM Page 171

quality remains uneven because province and dis-trict health officials introduce norms and guide-lines as well. The Ministry of Health has useddecrees and circulars to define the quality ofhuman resources and equipment and performancestandards. The ministry also has issued more than100 treatment guidelines, though a survey con-ducted in 20 district hospitals in 2000 on acute res-piratory infections showed that compliance was 25to 40 percent, probably reflecting weak support andsupervision at the local level. Meanwhile, overuseand overprescription of injection drugs were com-mon, with representatives of drug companies influ-encing physicians’ prescriptions (Dung et al. 2001).

Indonesia established minimum health servicestandards by ministerial decree in 2003. Districtsmust deliver services according to local needs in32 areas, including immunization, nutrition serv-ices, prevention of communicable diseases, andcurative care. Such standards will help define theservice levels that districts are accountable fordelivering. Whether these standards will be require-ments or targets needs further consideration, alongwith measures for dealing with districts that do notmeet the standards.

Toward a New Role for Ministriesof Health

Decentralization in Indonesia, the Philippines, andVietnam may help sustain overall improvements inhealth that have occurred during the last twodecades. Decentralization has appeared to spurlocal initiative in planning, delivering, and financ-ing services. Users are now participating in plan-ning in many regions, leading to more appropriateand better-targeted health services. Volunteerssupplement limited local financial and technicalresources. More important, perhaps, citizenshipand trust in local government have deepened. Theresulting efficiency gains and social capital supportthe decentralization of health services.

Still, experience in these three countries revealsthat decentralization dividends have been modestfor two reasons. First, these countries decentralizedhealth services in less-than-favorable environ-ments. Inequitable economic growth, populationpressures that brought epidemiological changes,and political uncertainties have limited the poten-tial gains from decentralization. Improvements in

health status are therefore greater in well-offprovinces, and service innovations have failed tospread beyond areas where the local economy isrobust and the political situation relatively stable.In these provinces, local governments have hadthe resources to meet growing demand for healthservices.

Weakness in decentralization policy also con-tributed to lower-than-expected health payoffs.These include ambiguities in goals, lack of detaileddesign, inconsistency with other policies, andpoorly thought-out implementation strategy.These follow from the fact that health was not themain—much less the sole—driver of decentraliza-tion. The Philippines, for example, included healthservices in its decentralization strategy only whenresistance from the education lobby forced legisla-tors to look at other national expenditures.

Inconsistent priorities have translated intoinconsistencies in policies and poor design of pol-icy instruments, especially the intergovernmentalfiscal transfer system. Local governments are alsotypically unaware of the types and timing ofnational interventions—information that is crucialto their own budget and investment planning. Withprior knowledge of available grants, technical assis-tance, and other support from national agencies,local governments can use information on localneeds, and proximity and direct accountability tobeneficiaries, more effectively.

In Indonesia, the slow and arduous emergenceof a consensus on a health decentralization frame-work partly reflects a governmentwide determina-tion to avoid service interruptions. But this focushas also allowed the central government to post-pone difficult decisions over the role and scale ofkey central ministries.

The Indonesian Ministry of Health also tends toview the public as passive service recipients ratherthan discriminating customers, owners, and poten-tial allies, and to present itself as a policing andstandards-upholding authority rather than a tech-nical agency. What’s more, the government hasnot yet developed a clear concept of the role ofprovinces in the health system. Decision makersknow that districts are typically too small to sup-port cost-effective programs, but they have notdesigned institutional solutions to encourage jointservice areas, or to make provinces agents of publichealth and related programs.


east_155-178.qxd 5/17/05 11:43 AM Page 172

Policy weaknesses also stem from laws and regu-lations, introduced in Big Bang fashion, that lackeddetail on functional and operational responsibili-ties and brought confusion and divergence betweenprovinces and other local governments. The Min-istry of Health also failed to coordinate with localgovernments and other actors in performing criti-cal health functions.

In these countries, reactive responses to transi-tion problems did not necessarily resolve fundamen-tal design issues. For instance, in the Philippines, theMagna Carta for Public Health Workers—whichprovided supplemental funds to only a few localgovernments—temporarily appeased disgruntledhealth workers but upset local governments byimposing unfunded mandates. Similarly, the coun-try “resolved” the mismatch between the distribu-tion of the IRA and devolved expenditures after 1992by providing grants to cities for hospitals they werealready financing before 1991. Thus, compensatingcities for their supposed “losses” due to the adjust-ments in the IRA was a politically necessary butcostly way of ensuring adequate funding for thehealth functions devolved to provinces and munici-palities (Capuno 2001).

However, efforts to correct these weaknessescould expand the gains from decentralizing health,even within a less-than-favorable environment.Toward that end, central health ministries mustfocus on specific tasks such as setting up qualityassurance mechanisms for drug supplies, safe-guarding access to medicines by the poor, and dis-mantling state monopolies on drugs. Responsibili-ties for communicable diseases include monitoringnational and regional trends, supporting laboratorycapacity and quality control and assurance, alertingprovinces to outbreaks elsewhere, and advocatingfor emergency financing.

With the Stewardship of the Ministry of Health

These examples illustrate the contributions neededfrom central government during decentralizationof health services. Above all, central agenciesshould concentrate on activities that go beyond thedirect provision of preventive and curative services,focusing on core public health functions, respond-ing to overall imperatives, and preventing potentialfailures.12 These efforts include not only tasksrelated to pharmaceuticals and communicable

diseases but also workforce training, recruitment,pay and benefits, and supervision.

Other core public health functions includeensuring that the poor have access to affordablecare, overcoming micronutrient shortfalls, creatingsustainable funding arrangements, acting as asource of ideas and best practices from theprovinces, and providing technical assistance on aselective basis. As the steward, a Ministry of Healthwould build consensus on national health objec-tives and standards, and coordinate rather thanrequire local governments and civil society groupsto meet these goals. Instead of relying on some-times heavy-handed regulation, the ministryshould align incentives to elicit the cooperation andparticipation of all sectors. Rather than imposehigh standards, the ministry should perhaps pro-mote them through advocacy and by strengtheninglocal governance mechanisms.

The stewardship role also entails pushing forgreater consistency among goals, programs, andpolicies of different national agencies to supportlocal governments. Finally, stewardship includesmore than content: it clearly entails leadership anda flexible, opportunistic mode of building partner-ships and exploiting opportunities.

Central interventions are warranted partlybecause these functions may not convey theurgency or tangible appeal of disease-specific pro-grams, and thus districts may neglect them. Subna-tional governments have little incentive to pursuecore public health functions because they cannotfully capture the returns, and because some func-tions are difficult to perform well because oflimited resources or lack of scale economies. Theimpacts of core public health functions are alsohard to measure: gauging the effects of a strong dis-ease surveillance and reporting system is difficult,while the direct distribution and use of drugs byinfected patients has obvious benefits.

Has the Philippines Turned the Corner?

The central health ministries in all three countrieshave taken steps to fulfill this new role, but nodefinitive transformation into stewardship has yetoccurred. Furthest along is the Philippines, whoseDepartment of Health has examined whetherdecentralization has paid off as a health reformvehicle.


east_155-178.qxd 5/17/05 11:43 AM Page 173

In 1999, under the Estrada administration,DOH formulated a comprehensive decentralizationstrategy called the Health Sector Reform Agenda(HSRA). The HSRA noted a slight resurgence ofcertain diseases and persistent inequities in serviceaccess. To counter these problems, DOH posi-tioned itself as a health leader, enabler, and capac-ity builder, administering only certain services(Department of Health of the Philippines 1999b).As a leader, it would primarily be responsible forsetting national health policy and regulations andstrengthening regulatory agencies. As an enablerand a capacity builder, it would seek to promoteinnovations and standards in health services, espe-cially at the local level. And as an administrator, itwould confine itself to pushing hospitals towardfiscal autonomy, securing funds for priority publichealth programs, and promoting universal cover-age under the National Health Insurance Program.

The novel aspects of this strategy are DOHreengineering and the convergence of all DOHinterventions in each province under the HSRAframework. DOH reengineering meant streamlin-ing operations, finances, and bureaucracy, anddeploying 1,638 personnel from the central officeto regional health offices, retained hospitals, andother DOH agencies. Under the convergence strat-egy, some provincial officials have drawn up healthdevelopment plans and interlocal health zones,with the DOH providing technical input and otherassistance. These zones bring together contiguouslocal governments around a district hospital to findways of improving the hospital referral system,exploiting economies of scale, and containingspillovers.

In December 2004, for example, Capiz provincedevised a five-year development plan for enrollingindigent families in the National Health InsuranceProgram, upgrading selected hospitals, and adopt-ing revolving drug funds and new drug manage-ment systems, with specific targets and activitiesat provincial and zone levels. With initial DOHsupport of P=10 million, the Capiz plan is expectedto yield gains from economies of scale in hospitaloperations, pooled procurement of drugs, andcontrol of epidemics. Similar arrangements areexpected in 2005 in other sites such as Pangasinan,Agusan del Sur, and Misamis Occidental.

As a dynamic process, decentralization in thePhilippines will continue to require adjustment

guided by HSRA. For example, the country cando more to prepare the intended beneficiaries, suchas local governments and health care users, anddiffuse political resistance. At the same time, thecentral Department of Health needs to extend itspartnership with health NGOs and civil societyorganizations with whom it is already working.Local health finances must rest on a firmer footing,including through greater reliance on local funds.Carefully designed user charges would not onlymake service delivery more efficient but also makelocal health programs sustainable and help subsi-dize the health needs of the poor. But to justifyhigher user fees, local governments must improveservice quality and require up-front financingfor facility improvements, personnel training andhiring, and drugs and medical equipment. DOHmatching grants could support enhanced services iflocal governments introduced new fee schedules.Grants would also make local public employmentmore attractive to health workers.

Waiting for Indonesia...

Indonesia has not clarified the health roles andresponsibilities of central and lower governmentsafter three years of decentralization. Nor has thecountry moved to emphasize core public healthfunctions, or seen marked improvements in spe-cific areas such as infectious disease control, phar-maceuticals, and human resources. Sectors besideshealth also have indeterminate policies, promptingadvice to clarify assignments across levels of gov-ernment and sectors (World Bank 2003a).

Two strands of thinking on decentralization areevident within the Ministry of Health. The first isdistilled in a 2003 decree that lists 29 strategic issuesrelated to core public health functions and adds keysteps to address them, such as minimum servicestandards (MSS), partnerships with NGOs, andservices for the poor. The decree points to account-ability mechanisms and traditional command-and-control instruments to limit the risks of servicedisruption. The former include the use of MSS toelicit district commitment, including assistance infunding core public health functions. The Ministryof Health has been relatively assertive in exertingits authority in responding to infectious diseaseoutbreaks such as severe acute respiratory sys-tem (SARS), and overseeing surveillance of and


east_155-178.qxd 5/17/05 11:43 AM Page 174

programs to combat diseases of national impor-tance and involving international obligations, suchas tuberculosis and HIV. The Ministry of Healthdepends on central and donor funding to achievethese ends, though each is unreliable, and haslooked for district support, seeing MSS as targetsfor district spending.13 The decree assigns keyresponsibility to district chief executives, and statesthat efforts to attain MSS should rely entirely ondistrict budgets, with central and provincial gov-ernments providing technical help, supervision,and oversight.

This approach is risky. Detailed, extensive MSScould undermine decentralization, and poorer dis-tricts could reject them because of limited fiscalcapacity. Most MSS have been set at high levels,imposed on rather than owned by local govern-ments, with the means of enforcement and penal-ties for noncompliance undisclosed. MSS would bebetter seen as medium-term goals rather than per-formance requirements that trigger funding andrequire enforcement. The Ministry of Health needsto develop ways of boosting district ownership ofefforts to prevent and control infectious diseases.

The second strand of thinking within theIndonesian Ministry of Health takes a more benignand constructive view of decentralization. Thisapproach is embodied in initiatives under way inYogyakarta and three other provinces (Lampung,North Sumatera, and West Java), with twenty-oneother provinces due to come on stream later. Min-istry officials who support this approach are tryingto use the momentum of decentralization as a cata-lyst for sector reform, with provinces playing animportant mid-level role. However, the above-citeddecree limits provinces to backstopping central anddistrict-level initiatives.

The province-based approach remains new andunder trial. It has already survived early bureau-cratic and other challenges, but the Ministry ofHealth needs to carefully assess experiences, impli-cations, and lessons emerging from the province-based framework and disseminate them to keystakeholders. The ministry can also help implementnew and existing provincial programs, includinginterventions that widen and deepen the decentral-ized approach. The ministry can also support pilotwork and research aimed at helping provincesrespond to the diverse challenges of managing anddeveloping the health workforce, such as by helping

provinces rationalize staffing numbers. The min-istry could also sponsor trials of approaches toattract doctors, especially specialists, to remote andundesirable locations. The ministry also needs todevelop standards that provinces and districts canuse to license service providers; work with profes-sional associations to strengthen quality improve-ment efforts and establish partnerships for profes-sional development; consult with consumer groupsand hospitals on workforce quality; focus on trendsin medical education; and fund and deploy special-purpose health teams.

Opportunities, imperatives, and stakeholderpressures could support the Ministry of Health’simpetus toward devolution and health reform.Budget constraints may force the ministry to lookto districts and provinces as sources of funds andjobs for the health workforce. Demands for betterservice quality and other public pressures havebegun to register with local political leaders andwithin the national ministry. A medium-term sce-nario resembles that in the Philippines: contesteddecentralization followed by a faster pace andmajor adjustments, leading to consolidation andmid-course correction. The Filipino experiencemost relevant for Indonesia is arguably the changein outlook within the Department of Health regard-ing its role in a decentralized system.

Stewardship by Vietnam’s Ministry of Health

The tempo of change in Vietnam is quickening,with several distinct drivers governing the pace.The first is the reappearance of deadly public healththreats at the top of the policy agenda. Vietnam isclearly vulnerable to new and more virulent strainsof diseases such as SARS, HIV/AIDS, and influenza,and reemerging diseases such as tuberculosis anddengue fever. Malaria remains a major publichealth problem in mountainous and ethnic minor-ity areas.

Successive crises have spawned rapid-reactionstructures and shown the importance of timely andwell-targeted responses guided by updated diseasesurveillance data. The Ministry of Health appearsto be developing expertise in explaining diseasechallenges and engaging the public and politicalleaders while soliciting various sorts of assistance.The ministry has also gained credibility and builtstronger ties to decision makers at the provincial


east_155-178.qxd 5/17/05 11:43 AM Page 175

level and in key central ministries, including theMinistry of Finance, the Ministry of Public Infor-mation, and the Ministry of Labor, Invalids andSocial Affairs. These drivers of change could inter-sect if, as seems likely, the revised agenda on com-municable diseases leads to requirements for morespending. This effort should involve a review ofarrangements for funding disease control andthe possibility of consolidating such outlays. Arelated issue is the need to avoid substituting forlocal expenditures; the ministry could make acase for requiring matching contributions fromlocal governments.

The 2002 law requires some acceleration inefforts to adjust government roles. These effortswill entail significant shifts, and the experience andcredibility accumulated in fighting SARS and otherdiseases could prove helpful.

In particular, the advent of provincially man-aged service delivery suggests the need for formalrecognition backed by real authority and resources,with the Ministry of Health focusing on key stew-ardship functions. Implementation of the 2002 lawwill thus enable the ministry to get out from undersecond- and third-best aspects of the de factohealth decentralization system that took hold in the1990s:

• Provinces are supposed to provide updates onhow they are allocating their recurrent budget.However, this requirement appears to belargely a formality. The Ministry of Health haslittle information on health budgets, and it isnot clear to what extent, if at all, it can influ-ence provincial spending of budgets alreadyapproved.

• The ministry lacks a clear role in formulatingand assessing policy and determining centralallocations to health; the ministries of Financeand Public Information are the key agencies inthis process. Central recurrent health spendingreflects projected revenue growth and recurrentexpenditures.

• Such incremental budgeting is not sensitive tothe goals and priorities set by the Ministry ofHealth.

The ministry can respond to the 2002 law partlyby strengthening budgeting procedures as well asimproving allocation. These efforts may include

replacing allocation norms and hospital paymentmechanisms with instruments based on the price ofhealth care services. The ministry would like to pre-pare expenditure norms to support management,monitoring, supervision, and control functions,and explore the use of norms that reflect populationneeds and improve equity in service access and use.

The Ministry of Health also recognizes that itneeds other policies with significant near-termimpacts to address disparities in health outcomesand per capita health expenditures across provinces.Per capita spending in the richest seven provincesis over three times that in the poorest sevenprovinces. Central and donor transfers do not pro-vide a counterweight, as the richest provincesreceive the largest per capita amount, and becausethe resources involved are relatively small.

Endnotes

1. No consensus has developed on the starting point forhealth decentralization in Vietnam. This chapter treats the1996 Law on State Budget as a path-breaking measure.

2. On December 27, 1997, then-President Ramos issuedAdministrative Order 372, which effected the withholdingof the “amount equivalent to 10 percent of the IRA.” Localgovernments challenged this order before the SupremeCourt and won in June 2000.

3. This is a common observation by local officials interviewedfor the Rapid Field Appraisal of Decentralization (Associatesin Rural Development 1993a, 1993b, 1994).

4. The standard deviation in regional IMR also fell from 8.51in 1980 to 4.84 in 1990, and then to 3.34 in 2000.

5. Figures from the Philippine National Health Accounts alsoshow that total per capita health spending rose fromUS$20.82 in 1991 to US$29.79 in 2001. In real terms,however, the country recorded a per capita decline fromUS$12.15 in 1991 to US$8.84 in 2001.

6. Vietnam also achieved a hefty threefold nominal rise intotal health expenditures over a five-year period, fromUS$0.68 billion in 1993 to US$2.17 billion in 1998(Knowles et al. 2003).

7. This occurs through the JPS-BK (the health componentof the social safety net); the fuel subsidy; and the DAK (aspecial-purpose grant) channels.

8. The Galing Pook Foundation presents awards to local gov-ernments for the best innovations in public service deliv-ery. The Philippine Human Development Network uses theHuman Development Index to identify and honorprovinces that made the greatest strides in promotinghealth, wealth, and education.

9. Moreover, the Department of Health itself did not antici-pate these issues, as the initial plan was to decentralize edu-cation first. Opposition from that sector led the govern-ment to consider decentralizing health instead (Diokno2003).

10. Data from the 1998 Vietnam Living Standards Survey indi-cate that 93 percent of all drug vendor contacts entailedefforts to obtain medicines without a prescription, withlittle variation across economic groups.


east_155-178.qxd 5/17/05 11:43 AM Page 176

11. A survey conducted by the National Family Planning Coor-dination Board in 2002 found that less than 10 percent ofhealth centers followed the manual on preventing infec-tions that may result from the use of contraceptives. Thesurvey also discovered that counseling in the family plan-ning program was poor, and that 20 percent of public facil-ities had never been supervised.

12. The Pan American Health Organization established 11essential public health functions through internationalconsensus. These have been field-tested and implementedin 43 countries of Latin America, the Caribbean, andEurope.

13. Central spending is limited and subject to strong competi-tion from outside and inside the Ministry of Health. Forexample, during 2001 and 2002, the central budget washighly constrained, and the ministry’s main funding initia-tive aimed not to control disease per se but to hire doctorsto fill vacancies. The ministry allocated some funds forcombating tuberculosis and HIV and a few other pro-grams, leaving insufficient central resources for tacklingother diseases.

Bibliography

Associates in Rural Development. 1993a. Synopsis of Findings:Second Rapid Field Appraisal of Decentralization. Reportssubmitted to the USAID-Governance and Local DemocracyProject. Manila: Associates in Rural Development.

———. 1993b. Synopsis of Findings: Third Rapid Field Appraisalof Decentralization. Reports submitted to the USAID-Governance and Local Democracy Project. Manila: Associatesin Rural Development.

———. 1994. Synopsis of Findings: Fourth Rapid Field Appraisalof Decentralization. Reports submitted to the USAID-Governance and Local Democracy Project. Manila: Associ-ates in Rural Development.

———. 1995. Synopsis of Findings: Fifth Rapid Field Appraisalof Decentralization. Reports submitted to the USAID-Governance and Local Democracy Project. Manila: Associatesin Rural Development.

———. 1996. Synopsis of Findings: Sixth Rapid Field Appraisalof Decentralization. Reports submitted to the USAID-Governance and Local Democracy Project. Manila: Associatesin Rural Development.

———. 1998. Synopsis of Findings: Eighth Rapid Field Appraisalof Decentralization. Reports submitted to the USAID-Governance and Local Democracy Project. Manila: Associatesin Rural Development.

Bautista, M. C. G., M. E. C. Yap, and E. S. Soriano. 1999. “LocalGovernment’s Health Financing Initiatives: Evaluation,Synthesis, and Prospects for the National Health Program inthe Philippines.” Bethesda, MD: Abt Associates.

Brinkerhoff, D., and C. Leighton. 2002. “Insights for Imple-menters: Decentralization and Health System Reform.” Part-ners for Health Reform, No. 1. Bethesda, MD: Abt Associates.

Capuno, J. J. 2001. “Policy Reform under Decentralization:Financing of Health Services in the Philippines.” RegionalDevelopment Studies 7: 109–28.

———. 2002. “An Analysis of Health Care Financing and Ser-vice Delivery in the Philippines under Decentralization.”Paper prepared for the Philippine Public Expenditure,Procurement, and Financial Management Review. Manila:World Bank.

Chen, Lincoln, and Linda Hiebert. 1994. “From Socialism toPrivate Markets: Vietnam’s Health in Rapid Transition.”

Working Paper No. 94.11. Cambridge, MA: HarvardUniversity, Center for Population and Development Studies.

Committee for Population, Family, and Children. 2003. “Demo-graphic and Health Survey 2002.” Hanoi: Population andFamily Health Project.

Communist Party of Vietnam. 1993. 7th Congress’s Resolutionof Vietnam’s Communist Party. Hanoi: Communist Party ofVietnam.

Department of Health of the Philippines. 1997. “An Assessmentof the Devolved Health System in the Philippines, 1996.”Manila: Department of Health, Local Government Assis-tance and Monitoring Service.

———. 1999a. National Objectives for Health, 1999–2004. HSRAMonograph Series No. 1. Manila: Department of Health.

———. 1999b. Health Sector Reform Agenda, 1999–2004. HSRAMonograph Series No. 2. Manila: Department of Health.

———. 2000. Department of Health Strategic Plan 2000–2010.Manila: Department of Health.

Diokno, B. E. 2003. “Decentralization in the Philippines afterTen Years: What Have We Learned? What Have I Learned?”Discussion Paper No. 0308. Quezon City: University of thePhilippines, School of Economics.

Dollar, D., P. Glewwe, and J. Litvack, eds. 1998. HouseholdWelfare and Vietnam’s Transition. Washington, DC: WorldBank.

Dung, P. H., et al. 2001. “Improving Physicians’ Prescription atOutpatient Department of District Hospitals.” Hanoi:Health Strategy and Policy Institute.

Esguerra, E. F. 1997. “Pursuing a National Health Strategy in aDecentralized Fiscal Regime.” Philippine Review of Econom-ics and Business 24 (2).

Fritzen, S. 1999. “Decentralization, Disparities, and Innovationin Vietnam’s Health Sector.”

———. 2002. “The ‘Foundation of Public Administration’?Decentralization and Its Discontents in TransitionalVietnam.” Paper presented at the Asia Conference onGovernance in Asia: Culture, Ethics, Institutional Reform,and Policy Change, City University of Hong Kong.

Galing, Kaban. 2001. The Philippine Case Bank on Innovationand Exemplary Practices in Local Governance. Galing PookFoundation.

Government of Socialist Republic of Vietnam. 1999. Decree29/CP.

Gwatkin, D., et al. 2000. Socioeconomic Differences in Health,Nutrition, and Population in the Philippines. Washington,DC: World Bank.

Knowles, J. C., N. T. Hong Hoa, D. B. Huong, N. Khang, T. T. MaiOanh, N. K. Phuong, and V. Ngoc Uyen. 2003. MakingHealth Care More Affordable for the Poor: Health Financing inVietnam. Hanoi: Vietnam Ministry of Health and AsianDevelopment Bank.

Lamberte, E. E. 2003. Sentrong Sigla: The Challenge Ahead andWay Forward—Formative Assessment and ImplementationReview Study. Manila: Philippines Department of Health,Bureau of Local Health Development.

Legaspi, P. E. 2001. The Changing Role of the Local Governmentsunder a Decentralized State: Some Cases in Philippine LocalGovernance. Quezon City: University of the Philippines,National College of Public Administration and Governance.

Lieberman, S. 2001. “Decentralization and Health in thePhilippines and Indonesia: An Interim Report.”www.worldbank.org/eap/eap.nsf.

Lieberman, S., and P. Marzoeki. 2000. Indonesia Health Strategyin a Post-Crisis: Decentralizing Indonesia. Washington, DC:World Bank.


east_155-178.qxd 5/17/05 11:43 AM Page 177

Lim, J., and C. Pascual. 2003. Toward a Pro-Poor Drug Policy inthe Philippines. Manila: World Bank.

Litvack, J., and D. Rondinelli, eds. 1999. “Fiscal Decentralization,Disparities, and Innovation in Viet Nam’s Health Sector.” InMarket Reform in Vietnam: Building Institutions for Develop-ment. Westport, CT: Quorum Books.

Manasan, R. G. 1997. “Local Government Financing of SocialService Sectors in a Decentralized Regime: Special Focus onProvincial Governments in 1993 and 1994.” DiscussionPaper No. 97-04. Makati City: Philippine Institute for Devel-opment Studies.

Medalla, F. 1996. “Co-Financing of Devolved Health Services:Lessons from the CHCA Experience.” Paper submitted to theHealth Policy Development Project. Quezon City: Universityof the Philippines, School of Economics.

Mercado, R. G. 1999. “Regional Budget Allocation: A PolicyRevisit.” Discussion Paper No. 99-29. Makati City: Philip-pine Institute for Development Studies.

Ministry of Health of Vietnam. 1999. Health Statistic Year Book1999. Hanoi: Ministry of Health.

———. 2002. Annual Activity Report. Hanoi: Ministry ofHealth.

Pham Tat Dong, Pham Trong Thanh, Dam Viet Cuong, DuongHuy Lieu, and Nguyen Hoang Long. 2002. “Impact of UserFees and Health Insurance on Equity in Financing, Accessto, and Utilization of Health Services.” Hanoi: Ministry ofHealth.

Philippine Health Insurance Corp. 1998. Annual Report 1998.Quezon City: Philippine Health Insurance Corp.

Pineda, V. S. 1998. “Health Management: Strategies fromSelected Cities.” Discussion Paper No. 98-36. Makati City:Philippine Institute for Development Studies.

Quimpo, B. A. 1996. “Devolution Matters. A Documentationof Post-Devolution Experiences in the Delivery of HealthServices.” Manila: Philippines Department of Health,Local Government Assistance and Monitoring Service.

Rood, S. 1998. “An Assessment of the State of Knowledge Con-cerning Decentralized Governance under the Philippines’1991 Local Government Code.” Philippine Journal of PublicAdministration 42 (1, 2): 58–82.

Smithson, P. 1993. “Health Financing and Sustainability inVietnam.” Hanoi: Save the Children.

Solon, O., R. M. Gamboa, Brad Schwartz, and A. N. Herrin.1992. “Health Sector Financing in the Philippines.” Health

Finance Development Project, Monograph No. 2. QuezonCity: University of Philippines School of Economics.

Solon, O., A. N. Herrin, R. H. Racelis, M. G. Manalo, V. N. Ganac,and G. V. Amoranto. 1999. “Health Care ExpenditurePatterns in the Philippines: Analysis of National HealthAccounts, 1991–1997.” Discussion Paper No. 9908. QuezonCity: University of Philippines, School of Economics.Ter-Minassian, T., ed. 1997. Fiscal Federalism in Theory andPractice. Washington, DC: International Monetary Fund.

Tuan Tran. 1995. “Historical Development of Primary HealthCare in Vietnam: Lessons for the Future.” Takemi Program inInternational Health, Research Paper No. 100. Cambridge,MA: Harvard University, School of Public Health.

WHO (World Health Organization). 1995. Decentralization andHealth System Change: A Framework for Analysis. Geneva:WHO, Division of Strengthening of Health Services,National Health System and Policies Unit.

———. 2002. World Health Report: Reducing Risks, PromotingHealthy Life. Geneva: WHO.

World Bank. 1992. “Vietnam: Population, Health, and NutritionSector Review.” Washington, DC: World Bank.

———. 1995. “Vietnam: Poverty Assessment and Strategy.”Washington, DC: World Bank.

———. 2000a. “Combating Corruption in the Philippines.”Washington, DC: World Bank.

———. 2000b. “Vietnam: Public Expenditure Review.”Washington, DC: World Bank.

———. 2001. “Vietnam Growing Healthy: A Review ofVietnam’s Health Sector.” Hanoi: World Bank.

———. 2001a. “Philippines Filipino Report Card on Pro-PoorServices.” Washington, DC: World Bank.

———. 2002. “Philippines: Public Expenditure, Procurement,and Financial Management Review.” Washington, DC:World Bank.

———. 2003a. “Decentralizing Indonesia: A Regional PublicExpenditure Review Overview Report.” Washington, DC:World Bank.

———. 2003b. “Poverty.” Vietnam Development Report 2004.Joint Donor Report to the Vietnam Consultative GroupMeeting, Hanoi.

———. 2004. World Development Report 2004. Washington,DC: World Bank.


east_155-178.qxd 5/17/05 11:43 AM Page 178

Documents

decentr alizing health: lessons from indonesia, the philippines, and