Social Health Insurance for the Poor Programs of the Philippines and Vietnam
Capuno, Joseph J
This paper compares the design features and implementation of the social health insurance (SHI) for the poor programs of the Philippines and Vietnam. The National Health Insurance Program-Sponsored Program of the Philippines and the Health Care Fund for the Poor Program of Vietnam both intend to improve the health status and, ultimately, the economic condition of the poor population. Following the framework of Carrin and James (2004), these programs are evaluated in terms of revenue collection, risk pooling, and purchasing for the period 1996-2005. Both programs are tax financed. Coverage is high but not yet universal because of limited administrative and financial capabilities of local government units (LGUs). Risk-pooling is nationwide in the Philippines and the poor enrollees get cross-subsidies from other insurance members. Meanwhile, risk pooling is provincewide in Vietnam and the poor do not get any subsidy from other health insurance programs. In addition to financing, the LGUs in the two countries also provide services to the poor. As the experiences of these countries show, the traditional approaches to provider payments and quality control have only weak incentive effects on performance under a decentralized health setting. It is argued that more attention should be given to such institutional context in the design of SHI programs.
This paper compares the design features and implementation of the social health insurance (SHI) for the poor programs of the Philippines and Vietnam. These programs are both designed to improve the health status and, ultimately, the economic condition of the poor population. While the social objectives of these programs are laudable, meeting them is difficult from an insurance point of view. In particular, the financial sustainability of the program is compromised by adverse selection. As target beneficiaries, the poor have low ability to pay their premium contributions yet they also have greater health needs than others because of their lower health status. If the poor, however, are required to pay their insurance premiums, then only a few of them will have insurance coverage. In comparison, the moral hazard problem is less serious because the poor cannot overuse their insurance benefits due to lack of funds for transportation, drugs, and other out-of-pocket expenses. Due to their budget constraints, they may not also seek health care even when they need it. If so, then the SHI program likewise fails on its objective.
It is therefore interesting to know how these insurance issues are addressed under the National Health Insurance Program-Sponsored Program (NHIP-SP) of the Philippines and the Health Care Fund for the Poor (HCFP) Program of Vietnam. These SHI programs were formally started in 1995 and 2002, respectively, although similar programs preceded them. The comparison of the two programs follows the framework of Carrin and James (2004) who propose to evaluate SHI programs in two levels.
The first level focuses on the health financing elements, namely, revenue collection, risk pooling, and purchasing. These elements jointly determine the adequacy and sustainability of the funds, the extent by which members gain financial accessibility to health services, and the overall optimality of resource use.
The second level links the SHI’s impact to health policy goals such as improved overall health status, greater equity in health access, and increased responsiveness of the health system to changing and varied needs of the population.
This paper evaluates the health financing elements of these two SHI for the poor programs for the period 1996-2005 for which data are available. By presenting “counterfactual experiences” with alternative design features, the comparative analysis offered in this paper complements the more in-depth analysis of the SHI programs previously conducted in the Philippines (e.g., Solon et al. 2003; Schneider and Racelis 2004; Jowett and Hsiao 2005; Obermann et al. 2006) and Vietnam (e.g., Knowles et al. 2003; Tran 2005; Wagstaff 2007). However, it should be noted that these programs were developed and implemented in dissimilar institutional contexts.
In particular, the role of the private sector in health sector is much wider in the Philippines than in Vietnam. The participation of the private sector in the SHI for the poor programs poses competition to the public sector, whose role, organization, and management changed with the adoption of the decentralized approach in the Philippines and Vietnam. Interestingly, the health decentralization also brings into focus the limitations of traditional design of SHI programs. Under decentralization, the local government units (LGUs) became the service provider. With the introduction of the SHI for the poor programs, they also became the financier of insurance in both countries. In Vietnam, the LGUs also assumed the functions of third-party payor. The multiple roles assigned to LGUs give them enormous powers and also conflicting incentives. As suggested by the experience of these countries, the incentives and bargaining power of LGUs relative to the SHI agency may be different from what has been usually assumed.
The rest of the paper proceeds with a discussion in the second section of the poverty situation in the Philippines and Vietnam, followed by a description of the key design features of the SHI programs in the third section. A comparison of the revenue generation, risk pooling, and purchasing aspects of the two programs is taken up in the fourth, fifth, and sixth sections, respectively. Some concluding remarks end the paper.
THE POOR IN CONTEXT
The Philippines and Vietnam are both developing countries and have similar poverty situation. The poor households in these countries remain a big portion of the total, although Vietnam’s rate of poverty reduction is relatively faster because of its impressive economic performance since the Doi Moi (market-oriented) reforms started in 1986. The poor in the two countries are also mostly concentrated in certain regions and have relatively poor health status.
The two countries have about the same population, which for over 15years since 1990, has grown to 83 million, from 61.1 million in the Philippines and 66.2 million in Vietnam (Table 1). About half (50.4%) of the Filipinos and a fifth (21.6%) of the Vietnamese live in urban areas in 2004. The Philippines is slightly richer but grows slower than Vietnam. From 1998 to 2005, its income per capita grew by 49.71 percent, from US$3,540 to US$5,300. In contrast, Vietnam’s income per capita expanded by 78.11 percent, from US$ 1,690 to US$3,010, as a result of the country’s shift from a planned economy to a market-oriented economy.
Other development indicators also reflect such trend of increasing economic payoffs. Since 1980, for example, the prevalence of child malnutrition, infant mortality, under-five mortality, and life expectancy at birth have steadily improved in both countries. Even the number of poor people has declined through the years although they have benefited less than the others from the economic progress. About 47.5 percent of Filipinos in 2000 and 58.5 percent of Vietnamese in 2002 were living on less than US$2 a day. This can be explained partly by the persistently skewed distribution of income in these countries. In 2002, the Gini indices were 46.1 for the Philippines in 2000 and 37.0 for Vietnam, both of which have not improved much since 1997.
According to the 2005 World Health Report, the Philippines’ total health expenditures on health was about 3.2 percent of the gross domestic product (GDP) in 2003, slightly lower than the 3.5 percent share registered in 1999. In 1999, Vietnam also had about the same share of total health expenditures in GDP (3.4%). This share shot up to 5.4 percent in 2003. In both countries, the private expenditures on health-comprising mostly out-of-pocket expenditures-accounted for the bulk of the total expenditures on health each year from 1999 to 2003. In 2003, for example, the percentage share of private expenditures in the total expenditures on health was 56.3 in the Philippines; it was 72.2 in Vietnam. The percentage share of private out-of-pocket expenditures in the total private expenditures on health in 2003 was 78.2 in the Philippines and 72.2 in Vietnam.
The poor in the two countries are also unevenly concentrated across regions. In the Philippines, about a third of all poor families in 1997,2000, and 2003 live in just three regions: Southern Tagalog, Bicol, and Western Visayas (Table 2). Poverty rate, however, is consistently highest in the Autonomous Region of Muslim Mindanao (ARMM), Central Mindanao, CARAGA, and Bicol region. In contrast, the National Capital Region has both the lowest poverty rate and the fewest number of poor households.
Similarly, based on the 2002 poverty thresholds set by the General Statistics Office of Vietnam, three in five poor Vietnamese live in just three regions: Northern Uplands’, Mekong River Delta, and North Central Coast (Table 3). Roughly the same proportion is found in 2003 based on a different poverty threshold set by the Ministry of Labor, Invalid and Social Affairs (MOLISA). The poverty rate is also very high in the Central Highland region although it is consistently lowest in South East and Red River Delta, the two principal political-administrative regions.
Vietnam also has a large and mostly poor ethnic population. According to the Vietnam Development Report 2004 (World Bank 2004b), the share of ethnic minorities in the poor population has increased from 20 percent in 1993 to 30 percent in 2002. The poorest ethnic minorities are found in the Central Highlands and the Northern Uplands regions. In 2002, even controlling for all other characteristics, a member of an ethnic minority household was likely to have expenditure levels lower by 14 percent than those coming from a Kinh (dominant ethnic group) or a Chinese household. In the Philippines, the high poverty rates in the ARMM also have an ethnic dimension but the continuing armed conflict in the region is considered the bigger reason why the region lags behind the others.
The relative depravity of the poor population in both countries is also reflected in their health status and access to health care. In the Philippines, a child born in 1997 to a poor family was more than twice as likely to die within a year or five years after birth than a child born to a rich family (Table 4). Moreover, a poor child is more susceptible to common illnesses such as fever, diarrhea, or acute respiratory infection. The poor children in Vietnam faced the same bleak chances of survival in 1998. Child mortality and the incidence of illness even worsened for poor families in 2000.
More children of poor families die or get sick because of less access to health care. In both countries, less than 60 percent of the children of poor families in 1997 and 1998 received a complete basic immunization package (i.e., shots for BCG measles, DPT). In Vietnam, at least 13 percent of the children of poor families did not get any of the basic immunization shots in 1988 and 2000. In the onset of fever, diarrhea, or acute respiratory infection, less than 30 percent of the poor children in the Philippines in 1997 were brought to either public or private health facilities for treatment. In Vietnam, relatively higher proportions of poor children with similar health conditions were brought to public facilities in 1998 and 2000.
The effective insurance coverage of the poor and their access to health care services can be better understood by looking at the overall structure of the two countries’ health sector. The Philippines has a larger private health sector compared to Vietnam. In 2005, of the total 1,838 hospitals in the Philippines, about 61.81 percent were privately owned. They also accounted for about half of the total number of hospital beds (87,739) in the same year. In contrast, there were only 17 private general and specialist hospitals in Vietnam in 2001, according to some estimates of Vietnam’s Ministry of Health. The private sector catered more to outpatient curative care services, with about 33,800 general and specialist clinics, and maternity homes in the country.
In both countries, the public health sector is also highly decentralized. In 1993, following the passage of the Local Government Code of 1991, the Philippine Department of Health devolved 46,080 health personnel out of 78,000, 595 government hospitals out of 639, and all 12,580 rural health units and other primary health care facilities to LGUs. Altogether, the devolved health personnel, services, and facilities accounted for about 41.21 percent of the total budget of the Department of Health in 1992. Gaining an important role in the health sector, the LGUs, by 2004, have contributed 14.4 percent of the funds for all health expenditures in the country, a significant input considering that the national government’s share was 15.9 percent.
In Vietnam, the decentralization of health services began with the enactment of the State Budget Law of 1986. This was further strengthened in 2001 with the issuance of the Prime Minister’s Decision 13/2001 promulgating the government’s Administration Reform Program, and, a year later, of the State Budget Law of 2002. Based on the Vietnam National Health Accounts, the provincial governments consistently outspent the central government in health throughout the period 19912000. In 2000, for example, health expenditures of the provincial governments and the national government totaled VND 3,631,754 and VND 1,451,725, respectively, or 41.86 and 17.44 percent of the total health expenditures (from all sources) in the same year (Knowles et al. 2003). Like in the Philippines, local governments have control over secondary, primary, and other lower level health facilities.
In a decentralized setting, with better information available to local officials about the needs of their constituents, especially the poor, it is expected that the level and quality of devolved health services, including health insurance coverage, would have improved. This will depend, however, on the accountability of the local officials to their constituents and service clients. All local political officials in the Philippines are elected for a fixed term of office. Those in Vietnam are likewise elected, but most candidates are first vetted by the Communist Party of Vietnam , whose members occupy most local offices. With the differences in political structures, the local residents or service clients in the Philippines in principle have more direct influence on their political leaders, and therefore on the availability and quality of health services than their counterparts in Vietnam. It is a common observation, however, that the service performances of Filipino politicians are only weakly correlated with their election performances, -which suggests that the poor, despite their huge number, may only have a weak influence on local services.
PRO-POOR SOCIALHEALTH INSURANCE PROGRAMS
The SHI for the poor programs of the Philippines and Vietnam were initiated in 1995 and 2002, respectively, in an attempt to consolidate and expand past initiatives. In the Philippines, the National Health Insurance Act was passed in 1995 (i.e., Republic Act 7875) to establish a nationwide program that integrates the Medicare program, then administered by the Government Service Insurance System for the state employees and by the Social security System for the private sector employees, and extends coverage to the poor and other sectors. In Vietnam, the Prime Minister’s Decision 139/2002/QD-TTg issued in 2002 created the Health Care Funds for the Poor (HCFP). The HCFP builds on the heath sector’s experience with previous health care-for-the-poor schemes.2
The general objective of the NHIA is to provide all Filipino citizens, especially the poor, with health insurance to gain access to health services. The Philippine Health Insurance Corporation (PhilHealth) was established to administer the National Health Insurance Program, which includes among its members all enrolled indigents under its so-called Sponsored Programs.
Meanwhile, the HCFP program of Vietnam, which is administered by the Fund Management Board (FMB) in each province, has the same objective of extending insurance coverage to the poor population as well as to residents of the socioeconomically disadvantaged communes (as defined in Prime Minister’s Decision 135/1998/QD-TTg) and members of ethnic minorities (as defined in Prime Minister’s Decisions 168/2001/QD-TTg and 186/2001/QD-TTg). The key members of the FMB are the Vice Chairman of the Provincial People’s Committee and officials from the province’s departments of health, finance, social welfare, and labor, as well as representatives from the local office of the Vietnam Social security (VSS).
The key design features of the two programs are summarized in Table 5. Unless indicated, these are the original features contained in the NHIA of 1995 and the Prime Minister’s Decision 139/2002/QD-TTg. The NHIP of 1995 was amended in 2002 with the passage of Republic Act 9241. The key amendments are the specification of the list of personal health services excluded from insurance benefits, the adjustments in the share of LGUs in the payment of indigent premium contributions, and in the provider accreditation rules.
Meanwhile, some of the HCFP features were modified starting mid-2005 following the issuance of Prime Minister’s Decision 170/2005/QD-TTg and Government’s Decree 63/2005/ND-CP. Specifically, Decision 170/2005/QD-TTg sets the new and higher poverty lines for 2006-2010, which effectively increased the number of target beneficiaries of the HCFP. Decree 63/2005/ND-CP enunciated regulatory reforms in medical insurance, including the re-classification of HCFP beneficiaries as members of the compulsory health insurance scheme with same entitlements as the other paying members of this scheme. The financial management and control of the HCFP was also transferred from the provincial FMB to the VSS. The VSS is tasked to administer and expand the existing compulsory and voluntary health insurance schemes, with the goal of achieving universal coverage.3
While these recent changes are important improvements in the HCFP, their implementation is not yet reflected in the data presented in this paper. Nonetheless, these changes are better understood in the light of the threeyears’ worth of experience with the original HCFP design, as will be discussed in the succeeding section.
According to Carrin and James (2004), one of the performance targets of a wellperforming health financing system is “to generate sufficient and sustainable resources for health” through careful design of population coverage and method of finance. The extent of population coverage will determine how much resources can be generated from the members, given their socioeconomic profile, and how much of these resources will be used to support their health expenditures, given their demographic and health characteristics. To ensure financial sustainability, therefore, the SHI program should have full flexibility in the design of enrollment strategies and setting of premium contribution to exploit the variations in health status and service utilization across socioeconomic groups. However, such flexibility is found neither in the NHIP-SP of the Philippines nor in the HCFP program of Vietnam. In both programs, the target coverage and methods of premium payments are set by law, constrained by fiscal resources and administrative procedures, and, therefore, could have less-than-secure actuarial basis. While the target coverage is nearly achieved, the financing of these SHI for the poor programs is likewise precarious.
Identification and financing
The target beneficiaries in the two countries are identified and screened using formal means. In the Philippines, the poor are identified based on a means test, roughly the same approach used in Vietnam. Once identified, the eligible beneficiaries are enrolled by their LGUs into the SHI program. Loopholes in the means test instrument (which checks for visible assets) and snags in the identification and verification procedures lead to errors of inclusion and exclusion or spotty coverage. Consequently, financing is also inequitable.
The means test in the Philippines is administered by the local Social Welfare Development Officer. The target beneficiaries are defined as the lowest 25 percent of the local population. These beneficiaries are individually identified through the use of the Community Based Information System-Minimum Basic Needs indicators and a uniform poverty thresholds, and then further verified through the family data survey of PhilHealth. The list of qualified beneficiaries-comprising all family members and legal dependents excluding children 21 years old and above-is submitted to the LGU for verification and decision.
Under the law, an LGU is required to enroll the eligible poor household in the NHIP-SP and subsidize partially or fully their premium contribution, which is pegged at PhP 1,200 per year. The LGU shares with the national government the premium subsidy, the amount of which is based on the LGU’s income classification. In the first year of enrollment, LGUs from the first to third income classes contribute 50 percent of the total premium contributions. Those in the lower fourth to sixth income classes contribute 90 percent during the first and second years of enrollment, and their share is raised progressively thereafter until it reaches 50 percent by the 10th year.4
In Vietnam, the same approach is used in identifying the target beneficiaries: the low-income residents, the members of ethnic minorities, and the people living in mountainous and economically depressed areas. The local labor official is tasked to identify the target beneficiaries through a household economic survey and the nationally set income poverty thresholds. The initial list of qualified beneficiaries5 is presented in a commune meeting for verification. A revised list is later drawn up and sent to the provincial FMB for decision.
The FMB decides on whether or not to extend the insurance cover to the listed beneficiaries and to provide them the statutory premium subsidy of up to VND70,000 per person per year using state budget and other funds from private or international donors. The state budget consists of VND52,500 per beneficiary per year that the national government provides and whatever additional amount the provincial government can voluntarily raise from its own sources for the HCFP. Since most provinces are unable or unwilling to raise funds, the national government transfer is effectively the only source of finance for the HCFP. Of the annual state budget per beneficiary, a minimum of VND50,000 is used to subsidize premium contributions; the remaining amount (VND2,500) is allotted for administrative expenses.
Target vs. actual enrollment
As LGUs in Vietnam only need to enlist their identified beneficiaries using the annual state budget allocation whereas those in the Philippines are mandated to raise their counterpart premium contributions as well, the enrolment of the poor into the SHI program should be easier under the HCFP than under the NHIP-SP. The expected effects of these design features are very well reflected in the actual enrollment patterns. In the Philippines, only two provinces-Abra and Camiguin-participated in the NHIP-SP program in 1997 (Table 6). Together, they accounted for all the 2,094 enrolled indigent families, which represented less than one percent of the target beneficiaries in that year. By 2000, 140 LGUs participated and accounted for 95 percent of the total number of enrolled families. The remaining five percent17,993 indigent families in all-was enrolled under Lingap ParaSa Mahihirap of the Estrada administration. Still, however, less than 10 percent of the poor families at that time were insured. Thus, realizing the financial difficulties of the LGUs, the national government began providing full subsidy to more indigent families. Under the Plan 5M, the national government enrolled about 68 percent of the 6.285 million indigent families covered in 2004 when the universal coverage of the poor was purportedly achieved.6 This target was achieved with only 78 percent of the LGUs (1,328) participating in the NHIP-SP, which enrolled only about 1.5 million poor families. The rest of the enrolled indigent families were subsidized by other donors.
Compared to the NHIP, the HCFP was able to cover a significantly greater number of target beneficiaries during its first three years of operation. Actual HCFP beneficiaries totaled 7.8 million in 2003,13.2 million in 2004, and 12.4 million in 2005 (Table 7). However, still less than 100 percent of the poor were covered under the program. In 2003, only about 9.5 percent of the population was insured under the HCFP when in fact about 28 percent of them were considered poor.7 Even in 2004 when the number of enrolled poor population nearly doubled, the coverage rate of about 16 percent was still less than the poverty rate of nearly 23.2 percent. This may seem surprising considering that the provincial governments in Vietnam do not face the same financial obligations as those in the Philippines. There are two reasons for this.
First, the income thresholds used in the HCFP program are lower than those set by the General Statistics Office (Table 7). Estimated by the Ministry of Labor, Invalid and Social Affairs, these income thresholds are also used in other entitlement programs for the poor. To meet these obligations, the Government of Vietnam understandably limits the coverage of these programs to the poorest of the poor. Second, the administrative capacity of the local implementers in identifying the target beneficiaries is weak. The procedures followed in drawing up, revising, finalizing, and approving the list of beneficiaries is tortuous, open to leakage and fraught with delay, and involve local political leaders and labor officials from the commune, to the district, and up to the provincial level (Capuno et al. 2006). Ethnic minorities were also excluded in some provinces like Cao Bang, which misinterpreted the title of Decision 139/2002/QD-TTg (On Medical Check-up and Treatment for Poor People) as a an exclusive program for the poor. These reasons explain why in 2004, when provinces presumably have already adjusted to the new program, enrollments rates in the regions of North Central Coast, South Central Coast, and Mekong River Delta were still significantly lower than their corresponding poverty rates.
Horizontal equity and sustainability
The NHIP-SP and HCFP are both tax-financed programs but differ in approaches in eliciting LGU participation. The national government premium subsidy under the NHIP-SP and HCFP programs may be considered as matching grant and block grant, respectively, and are both conditional on the enrollment of the eligible beneficiaries. Both transfer schemes help explain the patterns and equitableness of enrolment and the financial sustainability of the programs.
Though perhaps initially generous, the matching grant approach under the NHIP-SP may have not been very effective in encouraging and sustaining the participation of the low-income LGUs. Some of these LGUs are observed to discontinue the enrollment of their poor constituents after some years even before their share in the premium payments reaches the maximum 50 percent. Moreover, because many LGUs are unable to raise the full amount to enroll all eligible beneficiaries, they resort to additional and nontransparent criteria to prune down the list of beneficiaries. When this happens, LGUs often commit errors in excluding those who are more deserving and including those who are less deserving. It is often noticed, for example, that “political indigents”-i.e., actual beneficiaries who are close to the local officials but otherwise ineligiblemake it to the list. Hence, the effective co-financing shares between the LGUs and the national government could explain the patterns of new, renewed, and discontinued enrolments across LGUs. With the resulting unevenness in the coverage of the poor, they are in effect “un-equally treated” under the program despite their similarities in economic conditions. According to the fiscal decentralization literature, such horizontal inequity can be expected when the financing of a redistributive program-like the NHIP-SP-is delegated to the LGUs. While the problem of horizontal inequity is partly solved when the national government singularly sponsored the bulk of enrolled families since 2004, the change in the financing mode also raises the concern about sustainability, given the government’s other competing expenditure priorities.
Horizontal inequity is less of a problem in Vietnam because the national government guarantees the minimum premium subsidy for the poor.8 However, as the LGUs are dependent on the block grant, they have less incentive to mobilize additional resources or improve their administrative capacity to ensure proper coverage. The national government therefore shoulders the burden of ensuring the financial sustainability of the HCFP. Despite this, however, regional variations in enrollment rates are noted, which indicates that the poor are also unequally treated in Vietnam.
Another important design feature of SHI programs is risk pooling. If the pool of members is large, the financial risks due to adverse selection is spread among the members and therefore minimized. The concern with SHI for the poor programs is that the target population as a whole has the highest health risk yet it has the lowest ability to pay compared to other population groups. To reduce financial risks, a minimum amount of revenues from premium contributions should be attained, possibly through public subsidies financed through general taxes, as in the case of the NHIP-SP and HCFP. To ensure further its financial sustainability, the SHI for the poor program must be integrated with another program with a different or, ideally, a national risk pool. Once integrated with the wider risk pool, the program can possibly expand its benefits to members through the subsidy coming from the high-income, low-risk groups. In this aspect, the NHIP-SP and HCFP differs.
From its inception, the NHIP-SP is already part of a nationwide risk pool, in contrast to the previous Medicare Program that was concentrated on the employed population in the formal and informal sectors. In addition to the indigent and employed population groups9, the PhilHealth now targets and provides lifetime coverage to retirees and pensioners. As of June 2004, the members under the Sponsored Program accounted for about 47.8 percent of the total membership. What is perhaps more impressive is that high indigent coverage is achieved in all regions in the country (Table 8). As mentioned above, these enrolment trends were mainly the consequence of the national government’s Plan 5M, which essentially “absolved” the LGUs from their financial obligations and temporarily transferred this obligation to the Philippine Charity Sweepstakes Office. Without the Plan 5M, the risk pool of the NHIP-SP would have been more limited, considering that many LGU still failed to sponsor most of their poor constituents.
Even with a highly variable enrolment of the poor across regions, the PhilHealth itself is in a position to attenuate the resulting risks. By law, it is tasked to administer the National Health Insurance Fund (NHIF), which includes contributions from program members, balances from the Medicare Program, special government appropriations, donations and grants-in-aid, and accruals. Currently, the NHIF comprises the benefit fund and the reserve fund. The PhilHealth draws from the benefit fund to finance all benefit packages provided to different members.10 The PhilHealth keeps under the reserve fund all unexpended revenues in the current year. Thus, cross-subsidization among pools of members is implicit in the structure of the NHIF.
In contrast, the HCFP, as originally conceived, does not have the same flexibility. The provincial FMBs are each tasked to manage all HCFP funds and to provide coverage to all target beneficiaries within their respective jurisdictions, independent of other provinces and other health insurance programs. Thus, risk pooling across regions or population groups at the program level was not possible.
Also, the FMBs were not allowed to keep reserves. All unexpended fund balances for the current year becomes part of the beginning balance in the following year. Thus, provinces with positive beginning balances receive less state budget for the program than what is due them had they used up all previous appropriations. However, the actual budget transfers apparently were less than the required amounts in some regions, especially in the early years of HCFP implementation. This led some regions to provide inadequate premium subsidies to their poor constituents. In 2003, for example, the average budget provided in each of the seven regions in Vietnam was less than the minimum mandated amount of VND52,500 per person per year (Table 9). One reason for the budget shortfalls is the discrepancy in the estimated number of target population used by the Ministry of Finance that provided the state budget for HCFP and the identified beneficiaries constructed by the FMBs. The problem of budget shortfall, except possibly in the Northern Uplands, was already solved in 2004 when the nationwide average premium subsidy per person was increased to VND54,519 from VND39,669 in 2003. However, there was a large variation in the average premium subsidy in 2004. The average state budget in the two poorest areas, Northern Uplands and Central Highlands, were VND42,444 and VND54.941, respectively. In the same year, the average state budget in the South East Region was still below the mandated minimum although this was reached already in all the other regions. This new problem was due to weak monitoring of HCFP fund balances, which led to greater budget transfers to provinces than was warranted.
The HCFP problems concerning limited risk pool and fragmented funds are now addressed under Decree 63/2005/ND-CP (Promulgating the Medical Insurance Regulation). The new decree mandates the creation of a unified health insurance fund comprising the premium contributions of all members, state contributions, donations from private and donor agencies, and other revenues, to be administered by the VSS. Compared to the FMBs, the VSS is in a better position to spread risks among a larger pool of members. According to Tran (2005), the VSS had about 14.39 million members under the compulsory and voluntary schemes in 2004, up by 50 percent from 2000. The VSS also provided an increasing number of health insurance cards to the HCFP beneficiaries, from 0.84 million in 2000 to 3.89 million in 2004. While the HCFP beneficiaries technically were not VSS members prior to Decree 63/2005/ND-CP, they represent the incremental pool of members, which points to additional financial risks and organizational demands that VSS will face under the new decree.
The final point of comparison between the two SHI programs for the poor is the optimality in the use of resources. These resources are expected to be spent on the provision of benefits and the administration or operation of the program. The program can control moral hazard on the patient side by specifying the benefit package and can minimize fraud on the provider side through payment mechanisms. However, excessive control over the payment of claims could reduce the effective benefits to the members and build up reserves, which may encourage unnecessary operational expenses or inoptimal investments. Thus, safeguards for attaining administrative efficiency must be put in place as well.
By law, the PhilHealth is allowed to keep reserves equivalent to two years’ worth of projected benefit payments. Its administrative expenses also cannot exceed 12 percent of total contributions and not more than three percent of investment earnings from the immediately preceding year. In contrast, the FMBs in Vietnam are not allowed to keep reserves, and the provincial governments are expected to support the administrative costs of the HCFP program. Both programs, however, keep large unexpended fund balances. It is estimated that the PhilHealth’s reserves are equivalent to about 4.08 years in 2004 (Jowett and Hsiao 2005). In Vietnam, the unexpended HCFP funds are high even in the poorest regions where presumably there is high latent need for financial support among the target beneficiaries. These findings suggest that actual benefits to members may be less than what each program can support.
Benefit packages and expenditures
Beginning in 1999, all members of the NHIP, including the poor, enjoy a uniform set of benefit entitlements. Presently, the inpatient hospital care benefits include room and board charges, professional fees, diagnostic, laboratory and other medical examination charges, fees for the use of surgical or medical equipment and facilities, and prescription drugs. The same range of benefits, except for room and board, is included in the list of outpatient care services provided to NHIP-SP members under the capitation program. The benefit exclusions are expenses on the fifth and subsequent normal obstetrical deliveries, nonprescription drugs and devices, treatment for alcohol abuse or dependency, cosmetic surgery, Optometrie services, and cost-ineffective procedures as defined by PhilHealth. Like all other NHIP members, the indigent members have protection against catastrophic health expenditures. However, because of benefit ceilings, they effectively co-pay for these large unexpected health expenditures.
There is a ceiling on insurance payments for the included services although the support ceilings for some of them have been increased and additional benefit packages were introduced through the years. For example, the benefit ceilings for drugs and for x-ray and laboratory services were raised to 34 percent and 45 percent, respectively, since January 1,2002. In 2003, PhilHealth launched four new benefit packages: TB-DOTS, maternity, SARS, and dialysis treatments. Additional benefit packages were later introduced in 2006: outpatient malaria, outpatient HIV, newborn care, and third-normal spontaneous delivery. In consideration of the special circumstances of the indigent members, outpatient benefit package (OPB) was initiated in 2000 in four LGUs that received their insurance payments on capitation basis. Through a local ordinance, these LGUs opened trust funds, called capitation funds, to secure the insurance payments for OPB services. Since 2003, the so-called capitation program was adopted in more LGUs. Under this initiative, indigent members can avail of outpatient services from their accredited rural health units (RHUs), which PhilHealth pays on a capitation basis.
There are indications that the indigent members enjoy the expanded benefit packages and higher support ceilings. For instance, PhilHealth’s benefit payments for SP members have risen from PhP 1.2 million in 1999 to PhP614.4 million in 2003, or from less than 2 percent of total program benefit payments in 1999 to nearly 10 percent in 2004 (Table 10). In 2003, the total amount of claims under the NHIP-SP was 23 percent more than the PhP561 million collected for this particular program (Obermann et al. 2006), suggesting actual cross-subsidy from other component programs. It seems, however, that more benefits can be extended to the NHIP-SP members. In 2003, the NHIP-SP members had the lowest total number of claims ( 174,000) and average value per claim (PhP4,008) compared to the member groups (Obermann et al. 2006). Schneider and Racelis (2004) also found in a sample of 46 RHUs in four provinces that insurance coverage had a weak effect in encouraging greater use of RHU services. They also found, however, that the utilization of these services by the NHIP-SP members was relatively greater in areas with high NHIP-SP enrolment rates.
In Vietnam, HCFP members are entitled to the same benefit packages as the members of the compulsory health insurance schemes. The benefit packages include both inpatient hospital services and outpatient care services. Unlike the members of the compulsory health insurance scheme, however, HCFP beneficiaries are entitled to 100-percent support ceiling. In some provinces, the HCFP beneficiaries are also given transportation and food allowances whenever they are referred to higher level hospitals. Because of these generous benefit packages and the fact that FMBs have weak incentives to save on the HCFP funds, it is surprising to find regions having budget surpluses through the years. For example, the Northern Uplands and the Mekong River Delta-both relatively poor regions-consistently spent less than 70 percent of their annual HCFP budgets from 2003 to 2005 (Table 11). At the national level, only 58.32 percent in 2003, 66.04 percent in 2004, and 74.97 percent in 2005 of the total HCFP funds were used.
There are three reasons for the budget surpluses. First, the time spent in preparing the list of beneficiaries was lengthy resulting in delays in the issuance of the health insurance or health care cards. In 2006, the lag was three to four months in some provinces. Without these cards, the beneficiaries cannot claim insurance benefits.
Second, many provinces chose to provide insurance coverage through the direct reimbursement scheme rather than through the health insurance card scheme. In 2004, only about 30 percent of the total number of HCFP beneficiaries were enrolled under the health insurance card scheme. This rate dropped to about 25 percent in 2005. Under the direct reimbursement scheme, the health facility bills the FMB for the services provided to an HCFP beneficiary (who pays nothing). The provincial government pays the health facility directly although the release of the payment is often slow. In some provinces, facilities are provided quarterly allotments from the HCFP funds. These funds have to be liquidated before subsequent allocations can be provided. Some hospitals complain that the amount transferred are only for drugs and sometimes inadequate. In contrast, under the health insurance card scheme, the FMB pays the local VSS office the face value (VND50,000) of each card issued to the beneficiaries. Under this setup, the VSS pays the insurance claims directly to the health facilities. While both the provincial government and the VSS are not allowed to keep reserves, the VSS has less flexibility to use the funds for other purposes11 or to pool them with other insurance programs to spread risk. It also cannot claim more than VND2,500 per beneficiary per year for administration expenses. In some provinces, however; the VSS is also slow to disburse the funds because it accredits only district hospitals and higher-level facilities, and thus deprives lower-level facilities of possible insurance reimbursements.
Third, many of the beneficiaries were still not fully aware of their rights and entitlements, the high indirect costs of treatment, and the low quality of health facilities most accessible to them. Many of the HCFP beneficiaries that belong to ethnic minorities have special needs and sensibilities that are not always addressed by the program. Although some HCFP beneficiaries can avail of supplemental transportation benefits, the budget for these expenses is limited and the practice is not observed in all provinces. Without the additional support, access to health facilities may be difficult, especially in the Northern Uplands region where the overall terrain is rugged and mountainous (Capuno et al. 2006).
Both the NHIP-SP and HCFP programs employ a capped fee-for-service payment scheme to encourage the provision of service to insured patients and to discourage overprovision. The actual effect of this payment system on provider performance in the two countries has yet to be fully investigated. However, there is an institutional reason why this payment scheme has weaker incentive effect among public physicians than among private physicians. The reason is that public doctors are paid fixed monthly salaries and their individual shares in the revenues from insurance payments are not tightly linked to their respective performances. This is because the LGUs that own and operate these hospitals often control the distribution of the insurance payments to hospital staff.
In the Philippines, the distribution of insurance payments to public hospitals for professional services has been a contentious issue among LGU personnel. Based on Department of Health guidelines, insurance payments for professional fees are to be transmitted directly to the hospital chiefs. The revenues are to be divided evenly between the doctors, on the one hand, and the hospital staff, on the other. What happens in most places is that the insurance payments are received and recorded in a trust fund of the local treasury and can only be released to the doctors and staff with the requisite appropriations and approval of the local chief executive and budget officials. The distribution is done semiannually. Often, the doctors and hospital staff have no idea how much is actually due them.
Partly to solve this problem, PhilHealth now specifies in its standard agreement with LGUs participating under the capitation fund program that 20 percent of the funds will be used for administrative costs, i.e., as supplemental allowance to rural health unit (RHU) doctors and other staff. While this secured the direct payment to health professionals for outpatient services rendered to NHIP-SP members in the locality, it also provoked further resentment from nonhealth personnel who think that the health staff are unduly favored. In some places, part of the 20 percent is shared with other nonhealth staff (e.g., budget officer).
In Vietnam, doctors and other members of the medical staff get a fixed share of the user fee revenues starting in 1989 when user fees were adopted in the public health system after the issuance of Decree 45/HDBT. From 35 percent of the user fee revenues, their share was reduced to 15 percent in 1994 and later restored to 30 percent in 1995. These additional incentives, however, were not tied directly to the overall quality of staff’s services, especially those provided to the poor, the disabled, orphans, homeless elderly people, those living in high mountainous areas, and other groups exempted from user fees (Knowles et al. 2003). These exemptions were later removed in 2002 under Decision 139/QD-TTg, which still protected the poor from user fees but through insurance.
Whether doctor’s performance has improved, either in quantity or in quality, with the adoption of the HCFP is not clear from available data. What seems clear is that hospital revenues from insurance payments may have increased. In 2004, provincial hospitals received 39 percent of the total HCFP reimbursements, while commune health stations and district hospitals (including polyclinics) got about 21.4 percent and 34 percent, respectively. Under Decree 10/2002/ND-CP, doctors should get a share of the revenues from user fees (which include insurance payments). It appears, however, that the doctors in many district hospitals and lowerlevel facilities did not get much of the revenues from insurance payments since the HCFP advances to these health facilities were mostly drugs and medical supplies (Capuno et al. 2006).
The ability of the PhilHealth and the FMBs to deal with a decentralized public health system determines the overall administrative efficiency of the NHIP-SP and HCFP. The PhilHealth maintains 12 regional offices that take care of enrollment and claims verification. The head office performs quality control and payment functions. This setup has led to a growth in enrollment but resulted in delays in insurance payment and accreditation, and inconsistent quality assurance. While the PhilHealth promises to reimburse all approved claims within 30 days, the actual number of days spent from filing to payment varies. In 2005, for example, the Taft District Hospital waited nearly 53 days and the Pintuyan Community Hospital 49 days before all their claims were settled. Delays were due to incomplete supporting documents submitted by the hospitals and the lack of authority of local provincial PhilHealth officials to act on the deficient claims at once without the need for sending them first to the regional office (Capuno 2005). Another source of complaint is the delay in the accreditation of LGU hospitals and RHUs. With the accreditation, many LGUs see the claims reimbursements as a way of “recouping their investments” in indigent enrollments.12 Accreditation does not, however, ensure that all services actually provided consistently pass the minimum quality standards, especially in devolved health facilities. After all, the PhilHealth has to exercise greater caution in revoking the accreditation of these facilities. Otherwise, it may unduly provoke the concerned LGUs to withdraw its participation in the NHIP-SP. It is the dual role of the LGUs as financier and service provider that renders difficulty to PhilHealth in performing its regulatory function.
In contrast, the HCFP is administered in a highly decentralized manner. In principle, the FMBs can act faster and more decisively on issues concerning enrollment, payments, quality standards, and grievances and complaints. However, the structures and procedures for acting on these issues are not as well defined in Vietnam as in the Philippines. Beneficiaries are more inclined to vent their insurance complaints to commune leaders and other political officials rather than directly to service providers. Moreover, the service providers are represented in the FMBs, which pose a possible conflict of interest.
The FMBs are also very powerful. They discharge the financing, enrollment, service provision, and quality control functions. But while the FMBs can settle HCFP issues with finality, they are not subject to a system of checks and balances that could help improve their decisions. Even the Ministry of Health (MOH), which monitors all FMBs, cannot maintain local presence at all times. Officials from the central office of the MOH visit provinces irregularly and usually just require the FMBs to submit their annual HCFP reports.
Both the Philippines and Vietnam accomplished a lot in terms of extending insurance coverage to their poor population under their respective SHI programs. Their experiences can provide important policy insights to other developing countries. Their experiences have also brought into focus the importance of reviewing the traditional design features of SHI programs, particularly their relevance when executed in a decentralized public health setting.
Similar only in their target beneficiaries, the two programs are distinct in other key design features. As new policies are implemented, however, the HCFP of Vietnam will resemble more of the features of the NHIP-SP of the Philippines. In particular, the new regulatory policy in medical insurance will unify under the VSS all social health insurance programs in Vietnam, similar to the NHIP-SP under PhilHealth. Private providers will also now be able to participate in the HCFP. In these aspects, the HCFP can learn from the experiences of the more mature NHIP-SP. The integration of the HCFP with other SHI programs in Vietnam will enlarge the risk pool, from which the HCFP beneficiaries can potentially benefit from, similar to the experience of the enrolled indigents under the NHIP-SP. To realize these benefits, greater health service utilization must be encouraged through information and education campaign. However, in both countries, it is perhaps the improvement in the quality of local health facilities that will have a greater impact on utilization. Without upgrading the quality, the VSS will just suck up the HCFP funds from the poor regions (which are unable to meet accreditation requirements) and transfer them to richer regions (where quality is better and utilization is greater).
This predicted cross-subsidization is clearly regressive but making sure it will not happen is difficult and also potentially inefficient. Cross-subsidization is unwarranted when certain members are systematically restrained from their insurance entitlements, say, due to inadequate or substandard services. With the recent amendments in the HCFP, the purchasing and quality control functions are transferred from the FMBs to the VSS. Like the PhilHealth, the VSS will have greater incentives and bargaining power to impose quality standards on participating providers.
In both countries, however, the default choice of service provider of the indigent at the local level is the public health sector. In this sector, the doctors and other health staffare paid fixed monthly salaries and their performances are only indirectly linked to the insurance payments. Even when the medical staffare entitled to these payments, their individual shares are not necessarily based on performance but rather on equity or status. It is common for hospital chiefs in the Philippines to receive higher pay than other physicians who may have attended more patients. Moreover, these additional payments to the health staff become a source of envy and enmity for other LGU personnel. Thus, the traditional SHI approaches to provider payment do not have their expected effects on the performance of public sector physicians and staff. This is an issue that requires further investigation.
In addition to these personnel incentive issues, the PhilHealth and the VSS likewise have to contend with the LGUs that finance or facilitate the enrollment of the poor into the program and provide the services to these enrollees. With this dual role, the LGU can try to bargain for a lower standard of care in exchange for enrollment. This situation is not traditionally investigated on in the insurance literature, as the financing agent (usually the enrollee) is assumed to be different from the service provider. While these functions are separate in the private sector, they are often discharged jointly under a decentralized public health setting. The case of the Philippines and Vietnam show that this is also an important policy research issue.
1 The Northern Uplands comprise the North East and North West regions.
2 For a review of the social health insurance programs in the Philippines before 1995, see Gamboa et al. (1993). For a review of the health care-for-the-poor schemes in Vietnam prior to the HCFP, see Knowles et al. (2002). On the evolution of the SHI programs in the two countries, see Obermann et al. (2006) and Akal (2004).
3 The draft proposed revisions in Decision 139 (as of June 2006) also seeks to allow the accreditation of qualified private providers and to require all provinces to adopt the health insurance card scheme. If approved, this new law will increase the current state budget support of VND52.500 to VND60.000 per person per year. Partial insurance subsidy will also be extended to those whose income is between one and one and a half times greater than that of the poor.
4 An LGU can refer to a province, a municipality, a city, or a barangay where the member resides. In some places, the provincial government shoulders the entire LGU share. In most places, however, the provincial government splits it with the city or municipal government. The exact sharing between the provincial government and the city/municipal government varies, depending on the agreed terms. There are also some cities, municipalities, and barangays that fully sponsor their own poor residents.
5 The eligible HCFP beneficiaries exclude children who are under six years old. These children are provided health care services under a separate law.
6 There is no official estimate of poverty incidence in 2004. The incidence of poor families in 2003 was 4.02 million.
7 Although the HCFP specifically targets the ethnic minorities and residents of poor communities, most members of these population groups are also poor.
8 In 2003, some provinces failed to cover all eligible HCFP beneficiaries because either there were delays in the release of the state budget or the provincial estimates of the HCFP beneficiaries were different from those of the national government’s.
9 The employed population groups include employees in the government and the private sector, the selfemployed, overseas Filipino workers, and professionals in private practice such as doctors, lawyers, and dentists.
10 According to Republic Act 7875, contributions of members from the government and the private sector shall not exceed three percent of their monthly salaries; contributions of self-employed members shall not exceed three percent of their estimated actual net income for the preceding year; and contributions made in behalf of indigent members shall not exceed the minimum contributions set for employed members.
11 It is possible, for example, that the HCFP funds were used in some provinces to buy drugs that were normally financed by the regular provincial health budgets. These drugs are distributed to health facilities for dispensing to target beneficiaries.
12 In the first year of participation in the NHIP-SP and capitation fund program, a sixth class LGU will pay premium of up to PhP120 per indigent member and get in return P300 per indigent beneficiary in capitation payments, or a ‘ financial gain” of PhP180 per indigent enrollee.
Akal, A. 2004. Social health insurance development in Vietnam. Hanoi: World Health Organization.
Asian Development Bank (ADB). 2004. Key indicators 2004. Manila: Asian Development Bank.
Capuno, J.J. 2005. Maximizing the revenue collection of LGU hospitals/RHUs from PhilHealth reimbursement claims and equitable user fees. Manila: German Technical Cooperation.
Capuno, J. J., T.TM. Oanh, N.T. Thanh and T.V Tien. 2006. Health care for the poor: a situation analysis. Report submitted to the Ministry of Health, Government of Vietnam under the MOH-WB Technical Assistance Project for the Northern Uplands Project. Hanoi, Vietnam.
Carrin, G. and C. James. 2004. Reaching universal coverage via social health insurance: key design features in the transition period. Health financing policy issue paper. Geneva: World Health Organization.
Gamboa, R.M., M.C.G Bautista and M.L.L. Beringuela. 1993. Health insurance in the Philippines. Health Finance Development Project Monograph No. 6. Manila: Department of Health.
Jowett, M. and W. Hsiao. 2005. The Philippines: extending coverage beyond the formal sector. Cambridge, MA: Harvard School of Public Health.
Knowles, J.C, T.H.H. Nguyen, B.H. Dang, K. Nguyen, T.M.O. Tran, K.P. Nguyen and N.U. Vu. 2003. Making health care more affordable for the poor. In Health financing in Vietnam, vols. 1 and II. Hanoi: Ministry of Health and Asian Development Bank.
Obermann, K., M.R. Jowett, M.O.O. Alcantara, E.P. Banzon and C. Bodart. 2006. Social health insurance in a developing country: the case of the Philippines. Social Science and Medicine 62: 3177-3185.
Philippine Health Insurance Corporation. 2004. Implementing rules and regulations of the National Health Insurance Act of 1995 (Republic Act 7875 as Amended by Republic Act 9241) [online] http:// www.philhealth.gov.ph [Accessed 1 March 2007].
_____. 2005a. Stats and charts, 1st Semester 2004 [online] http:// www.philhealth.gov.ph. [Accessed 20 July 2005].
_____. 2005b. The PhilHealth chronicles. The journey towards universal social health insurance [online] http://www.philhealth.gov.ph [Accessed 20 July 2005].
Solon, O., S. Alabastro-Quimbo and C. Panelo. 2003. Financing health care for poor Filipinos. Human Development Unit, East Asia and Pacific Region. Washington, DC: World Bank.
Schneider, P. and R. Racelis. 2004. The impact of PhilHealth indigent insurance on utilization, cost, and finances in health facilities in the Philippines. Partners in Health Rzformplus. Bethesda: Abt Associates, Inc.
Tran, VT. 2005. The inclusion of the poor in social health insurance framework: the strategies applied in Vietnam [online] http://www.shi-conference.de/ contribut.html. [Accessed 15 September 2006.]
Wagstaff, A. 2007. Health insurance for the poor: initial impacts of Vietnam’s health care fund for the poor. World Bank Policy Research Working Paper 4143. Washington, DC: World Bank.
World Bank. 2000. Vietnam public expenditure review. Hanoi: World Bank.
_____. 2004. Vietnam development report 2004. Hanoi: World Bank.
_____. 2005. Socio-economic differences in health, nutrition and population in 56 developing countries: country tables. Washington, DC: World Bank.
World Health Organization. 2004. Regional overview of social health insurance in South-East Asia. New Delhi: WHO Regional Office for South-East Asia.
JOSEPH J. CAPUNO *
* Assistant Professor, School of Economics, University of the Philippines (UPSE), Diliman, Quezon City 1101, Philippines. For their helpful comments and suggestions, he would like to thank Samuel Lieberman, Eduardo Banzon, Mario Taguiwalo, Alejandro Merrin, Lorraine Hawkins, Aleli Kraft, Stella Alabastro-Quimbo, and the participants in seminars at the UPSE and at the Philippine Institute for Development Studies, as well as in conferences held in Mumbai, India, and Beijing, China, where earlier versions of this paper were presented. Also, he would like to acknowledge the financial support of the Philippine Center for Economic Development. All errors are solely the responsibility of the author. He may be contacted at firstname.lastname@example.org.
Copyright Philippine Institute for Development Studies 2006
Provided by ProQuest Information and Learning Company. All rights Reserved