## _wp0961

## Source details

**Canonical URL:** [_wp0961](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0961.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0961.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0961.pdf.json)

---

### Introduction and purpose
- Universal health care in the paper: population enjoys free of charge a specific list of health services, often linked to a list of diseases; financed by government via mandatory payroll/social security contributions, general taxation, or combination.
- Policy intent: assist policymakers in designing, financing and implementing universal health care based on experience of countries with such a system; primary focus on Eastern Caribbean Currency Union (ECCU) countries but policy implications could apply broadly.
- Key financing design criteria:
  - Financing option must be (i) domestically-based, (ii) sustainable, (iii) efficient, (iv) equitable, and (v) politically acceptable.
- Implementation preconditions emphasized:
  - Sound public administration, a small informal economy, transparent health financing system that builds social consensus; monitoring and incentives to minimize abuse and over-utilization.

### Fiscal constraints and recommended financing source
- Sources that can increase fiscal space: tax measures (tax increases or better tax administration); lower government spending; additional borrowing; external financing (grants).
- ECCU-specific constraints and observations:
  - ECCU countries are among the most highly indebted countries in international comparisons (IMF, 2008) — additional borrowing is a less desirable route.
  - Disbursements of official development assistance (ODA) have been disappointing; ODA for health is limited and mostly for low-income countries.
  - Donor assistance unpredictability reasons: (i) political and budgetary decisions by donors; (ii) administrative delays on donor side; (iii) substantial bureaucratic procurement and reporting requirements resulting in noncompliance with conditionalities.
  - Sustainability concern: sustaining services once donor funding stops is questionable.
- Recommendation:
  - Taxation should be the main source of financing because health spending involves recurrent expenses and should be financed by reliable, permanent sources.

### Demographic and epidemiological trends — projections and impacts
- Demographic projections and implications:
  - Working-age population (aged 15–59 years) is not projected to decrease in the next three decades, while the old-age population is expected to increase rapidly.
  - Number of people over the age of 60 projected to almost double by 2035, from 9.5 percent of the population in 2005 to 17 percent in 2035.
  - Demographic pressures alone could increase health spending (in percent of GDP) by 2035 by over two percentage points under feasible assumptions for most countries.
  - These projections assume no change in the organization and financing of the health system (no health system reform).
- Epidemiological transition specifics:
  - Shift from communicable causes to noncommunicable chronic diseases; leading causes include cardiovascular diseases, neoplasm malicious, and diabetes.
  - Diabetes specifics:
    - Described as the most significant “new disease” of the epidemiological transition (Caribbean Commission on Health and Development, 2005).
    - Requires lifelong, continuous medical care; economic impact expected to be large in next two decades.
    - One in ten of the adult population in the Caribbean region and one in five people over the age of 40 are currently living with diabetes.
    - Wild et al. (2004): diabetes estimated to affect 22,000 residents in the ECCU in 2000, with numbers projected to double by 2030.
- International evidence on age-related cost concentration:
  - Canadians consume more than 50 percent of their lifetime health expenditures after the age of 65 (Brimacombe et al., 2001).
  - Elderly patients (aged 65+ years) cost the U.K. health system five times more than those under 65 (Honjo, 2006).
  - Federal Statistical Office of Germany (2004): health costs for those aged 65-85 are at least twice as high as the German national average.
- World Bank finding:
  - Over the next 20 years, changes in population size and structure alone will increase total health spending needs by 47 percent in Latin America and the Caribbean (World Bank, 2006).

### Scenarios and methodologies for projecting long-term health-care costs
- Demographic-driven projection approach:
  - Projections derived by multiplying the last actual observation on health care cost per age group by projections for population by age group, so projected changes reflect only changes in demographics.
  - Health costs by age group are not available for ECCU; proxies from other countries and WHO data are used; population projections from each country’s latest Actuarial Report.
- Scenarios for elderly (aged 65+) per-capita cost relative to the rest of the population:
  - Scenario 1: health costs for those aged 65 and above is five times the cost of the rest of the population (same assumption as Honjo (2006)).
  - Scenario 2: health costs for those aged 65 and above is twice as high as that of the rest of the population.
  - Scenario 3 (most optimistic): health costs are equal across all population ages.
- Caveats and limitations:
  - Projections should be viewed with caution: demographic changes may not be the key driver of future health care expenditure.
  - Bjornerud and Martins (2005) show demographic changes accounted for only 10 percent of average growth in health spending across OECD countries over the last 30 years.
  - Other drivers to consider: changes in health status, technological and medical progress, economic growth, changes in organization and financing of health care systems.
- Age-adjustment methodology:
  - Uses Esmail and Walker (2005) approach with Canada as benchmark.
  - Assumes elasticity of health spending (percent of GDP) to the share of population over 65 of 0.698.

### Epidemiological profile and diabetes-specific findings (selected country highlights)
- Population health statistics (exact figures preserved):
  - 28.1 percent of the population has abnormal glucose levels and 8.1 percent of those tested were diabetic.
  - Almost half of St. Lucians have abnormal systolic blood pressure, described as the highest in the world.
  - Wild et al. (2004) projected diabetes incidence in St. Lucia rising from 5,200 in 2000 to 11,300 by 2030.
  - The Universal Health Care Diabetes and Hypertension (2007) pilot screening project recorded 13,000 diabetic cases as early as 2007.
  - Diabetic cases are more concentrated in those aged 45 and above.
  - Diabetes is higher among women due to higher obesity rates; increases in obesity trends expected to worsen diabetes outlook.
- Estimated treatment-cost assumptions and scenarios:
  - Annual cost assumption: EC$1,100 per diabetic case (based on Jamaican proxy including direct and indirect costs).
  - Jamaican cost composition (proxy): more than 50 percent of the total cost is explained by laboratory diagnostic tests; one-fourth of the cost is associated with doctor charges/visits.
  - Two scenarios for health-cost evolution:
    - Scenario 1: health costs rise in line with the average historical health inflation rate.
    - Scenario 2: health cost inflation is two percentage points above the historical average.
  - Projection outcome: Even under the scenario where health costs rise with their historical trend, the cost of treating diabetes could triple for Dominica and St. Lucia by 2030.
  - Note: Projections for diabetic cases use Wild et al. (2004); for St. Lucia, actual numbers are used up to 2007 and thereafter interpolated using Wild et al. (2004) trend.

### Provider arrangements, provider payment design, and efficiency
- Provider delivery and contracting:
  - Many countries use public provider networks while contracting non-governmental and private providers to expand choice, at the cost of higher administrative expenses.
  - Fewer providers have been more successful at containing health cost increases, possibly by negotiating lower doctors’ fees.
- Provider payment observations and recommendations:
  - Examples: Aruba experienced cost escalation under unregulated fee-for-service specialist payments; Canada contained costs using global budgets and monitoring service volumes per physician.
  - Optimal payment mix: mixed reimbursement systems (some prospective payment combined with some cost-based payment).
  - Suggested mix: a combination of fee-for-service and predetermined (salary or capitation) payments, possibly combined with positive patient cost sharing to reduce moral hazard and restrain supply.
  - Fee-for-service should only be applied where monitoring safeguards exist; otherwise introduce capitation payment to all providers.

### Benefit-package design, targeting, and role of private financing
- Benefit-package guidance:
  - Choice should reflect health needs, clinical priorities, and fiscal constraints.
  - Given ECCU budget constraints and high debt levels, recommend a universal, small package initially covering public goods, goods with externalities, and interventions with proven impact.
  - Clinical care and catastrophic expenditures could be left to targeted financing for the poor.
- Pharmaceuticals:
  - Drug benefits should apply to a defined and limited list of prescribed drugs.
  - Generic prescribing should be implemented when possible to contain costs.
- Targeting and safety-net mechanisms:
  - Streamlined universal package combined with targeted measures for the poor is advised.
  - Conditional cash transfers cited as successful with relatively minimal administrative costs.
- Role of private financing:
  - Private financing more suitable for private goods such as individual ambulatory and institutional care services, pharmaceuticals, and medical interventions.

### International targeting and implementation examples (Box 2 and Box 4 highlights)
- Targeting mechanisms:
  - Colombia: two kinds of basic coverage depending on income; payroll contributors access minimum defined coverage, non-payroll contributors access subsidized Mandatory Health Plan (World Bank, 2006).
  - Suriname: extremely poor individuals receive a health card permitting access to ambulatory, drugs and inpatient services at zero or minimal cost (Lalta et al., 2005).
  - Jamaica: PATH links cash benefits to school attendance and visits to health facilities; targets children up to 6 years, persons 65 and older, pregnant and lactating women, and poor adults (Lalta et al., 2005).
- Examples of gradual extension of coverage:
  - Republic of Korea: phased extension from 1977 (8 percent coverage) to universal health insurance legislated in 1989.
  - Taiwan Province of China: decade-plus preparation; pilot projects beginning 1987; legislation enacted in 1995; mandatory scheme fully operational by 2000.
- Taiwan POC outcomes (exact figures preserved):
  - Prior to reform (1995): 59 percent of the population was insured under ten plans; uninsured 41 percent.
  - By-end 1995 enrollment reached 92 percent.
  - Out-of-pocket spending dropped from 48 percent in 1993 to 30 percent in 2000.
  - Health spending rose from 5.3 percent in 1995 to 6 percent in 2000.
  - Average hospital admission rate increased from 110 per 1000 people in 1994 to 120 per 1000 in 1996.
  - Co-payment for outpatient visits ranged from $1.5 to clinics with referral to US$11 to academic medical centers without referral.
  - Inpatient coinsurance rate: 10–30 percent depending on length of stay.
  - Patient annual cap: total amount each patient pays capped to 10 percent of average per capita national income for a calendar year.
  - Poor households are exempt from all these costs.
  - Public satisfaction ratings reached 70 percent at some times (prior to 1995 ratings were around 30 percent).

### St. Lucia case: proposed universal health care (Appendix I) — costs, financing, and recommendations
- Objective:
  - Provide equal access to health care defined in a published package of limited services and benefits to all St. Lucians or non-nationals residing in the country for at least six months.
  - Emphasis on preventive services, comfort care and family planning services; inpatient, outpatient, limited overseas care, and inpatient pharmacy services primarily provided by two general hospitals and a new hospital under construction.
- Rationale:
  - Response to concerns about sustainability of current financing, poor quality and inequity of care, escalating health costs, and rising number of residents seeking medical treatment without paying.
- Estimated initial cost (exact figures preserved):
  - “The Taskforce… estimated that the initial cost of universal health care would be over US$20 million (over 3 percent of GDP).”
- Health Expenditure Before and After the Introduction of Universal Health Care (In Percent of 2003 GDP) — exact formatting and values preserved:
  - BeforeAfter
  - Public                                                                         3.34.6
  - Of which: Universal Health Care3.2
  - Private2.21.8
  - Total5.56.4
- Financing of Universal Health Care (million US dollars) — exact figures preserved:
  - Estimated Cost21.7
  - Financed by:
    - Government Transfer11.0
    - Envriromental Levy7.8
    - SalaryDeduction2.9
- Taskforce financing recommendations and policy guidance:
  - Use general taxation as the primary financing mechanism for universal health care.
  - Implement environmental levy and budgetary contributions as main initial financing sources.
  - Supplement main sources with a small salary deduction.
  - Consider VAT as a future supplement or replacement for other financing sources.
  - Defer reliance on salary deductions until the informal sector is effectively captured.
  - Do not rely on property taxation given current poor collection performance.

### Efficiency, public-expenditure management, and preconditions for reform
- Five criteria for judging taxes (IMF and World Bank, 2005):
  - Revenue adequacy and stability
  - Efficiency
  - Equity
  - Ease of collection
  - Political acceptability
- Comparisons of financing options:
  - Payroll/social security contributions:
    - Revenue potential depends on labor market conditions, existing payroll tax rates, and size of informal economy.
    - Examples of difficulty: Kazakhstan collected only 40 percent of expected revenue after introducing universal coverage financed via social security contributions.
    - Payroll-capped systems may struggle when health costs rise faster than per capita GDP (Taiwan POC example).
    - Payroll contributions may raise tax evasion, reduce formal labor market size, raise labor costs, and hinder growth.
  - General taxation:
    - VATs and sales taxes can tax a broader base including unemployed, self-employed, tourism sector, and large informal sector.
    - VATs are typically regressive but can be designed to be more progressive by exempting basic goods.
  - Ease of collection:
    - Payroll deductions are automatic where formal employment is large; ECCU’s large informal economy complicates payroll-based collection.
    - Spain and Iceland shifted from social security contributions to general taxation to avoid such problems.
  - Political acceptability:
    - Payroll contributions often more politically acceptable; general taxation acceptability can be enhanced if accompanied by a specific individual entitlement (a benefit tax), as in Ghana.
- Preconditions for successful implementation:
  - Tax administration should be sound.
  - Health care plans should be part of government medium- and long-term strategy and incorporated into medium-term expenditure framework.
  - Public financing of health care should be transparent to build consensus.
  - Informal sector should be tackled so beneficiaries pay according to ability.
  - Enhance public expenditure management and accountability; link budgetary allocations with health system performance.
  - Conduct solid actuarial studies and phase extension progressively.
  - Improve health system efficiency; significant waste-reduction opportunities exist among governments, hospitals and physicians (PriceWaterhouseCoopers, 2005).

### Key policy recommendations and operational measures (summary)
- Secure sustainable funding primarily from domestic revenues, notably taxes, given limitations and erratic nature of development assistance.
- Ensure tax administration, financing transparency, and measures to address the informal sector before expanding coverage.
- Start with a streamlined benefit package meeting basic needs; avoid including private goods that disproportionately benefit the better-off.
- Use mixed provider reimbursement (capitation plus fee-for-service) with monitoring safeguards to limit abuse and over-utilization.
- Conduct actuarial and feasibility studies and implement phased expansion to retain stakeholder support and ensure financial sustainability.
- Strengthen public expenditure management, monitoring and evaluation mechanisms to improve accountability and effectiveness of additional health spending.

*Source: _wp0961 (excerpts and appendices from the IMF working paper content provided).*

### References .............................................................................................................

### _wp0961 - References .............................................................................................................

### Introduction and purpose
- Universal health care in the paper: population enjoys free of charge a specific list of health services, often linked to a list of diseases; financed by government via mandatory payroll/social security contributions, general taxation, or combination.
- Examples of financing arrangements cited:
  - General tax-based funding: United Kingdom, Australia, Canada, Denmark, Sweden, Cuba, Italy, Brazil.
  - Mandatory payroll/social security contributions: Germany, Japan, France, Singapore, Costa Rica.
  - Most Caribbean countries: budgetary allocations from the central government are the dominant source of health financing (Table 1).
- Policy intent: assist policymakers in designing, financing and implementing universal health care based on experience of countries with such a system; primary focus on Eastern Caribbean Currency Union (ECCU) countries but policy implications could apply broadly.
- Key financing design criteria: financing option must be (i) domestically-based, (ii) sustainable, (iii) efficient, (iv) equitable, and (v) politically acceptable.
- Implementation preconditions emphasized: sound public administration, a small informal economy, transparent health financing system that builds social consensus; monitoring and incentives to minimize abuse and over-utilization.

### Financing options and fiscal constraints
- Sources that can increase fiscal space: tax measures (tax increases or better tax administration); lower government spending; additional borrowing; external financing (grants).
- ECCU-specific fiscal constraints:
  - ECCU countries are among the most highly indebted countries in international comparisons (IMF, 2008) — additional borrowing is a less desirable route.
  - Disbursements of official development assistance (ODA) have been disappointing in recent years to the region; ODA for health is limited and mostly for low-income countries (Table 3).
  - Donor assistance unpredictability reasons: (i) political and budgetary decisions by donors; (ii) administrative delays on donor side; (iii) substantial bureaucratic procurement and reporting requirements resulting in noncompliance with conditionalities.
  - Sustainability concern: even if donor assistance provided, sustaining services once donor funding stops is questionable.
- Recommendation: taxation should be the main source of financing because health spending involves recurrent expenses and should be financed by reliable, permanent sources (long-term resources) rather than debt or unpredictable external aid.

### Challenges: population aging and epidemiological transition
- Demographic trends and projections:
  - Working-age population (aged 15–59 years) is not projected to decrease in the next three decades, while the old-age population is expected to increase rapidly.
  - Number of people over the age of 60 projected to almost double by 2035, from 9.5 percent of the population in 2005 to 17 percent in 2035.
  - Population index chart reference: ECCU: Population by Age Group, 2005–35 (Index 2005=100).
- Projected impact on health spending:
  - World Bank (2006) finding: over the next 20 years, changes in population size and structure alone will increase total health spending needs by 47 percent in Latin America and the Caribbean.
  - International evidence cited:
    - Canadians consume more than 50 percent of their lifetime health expenditures after the age of 65 (Brimacombe et al., 2001).
    - Elderly patients (aged 65+ years) cost the U.K. health system five times more than those under 65 (Honjo, 2006).
    - Federal Statistical Office of Germany (2004): health costs for those aged 65-85 are at least twice as high as the German national average.
  - ECCU projection approach: estimates health spending (in percent of GDP) the ECCU could face in next three decades under three scenarios for per capita health cost of elderly (aged 65+ years):
    - (i) equals the national average;
    - (ii) is twice the national average;
    - (iii) is five times the national average.
  - Demographic pressures alone could increase health spending (in percent of GDP) by 2035 by over two percentage points under feasible assumptions for most countries (Figure 1).
  - With more than two-thirds of the health spending already accrued by the public sector in the ECCU, the projected impact on public finances due only to population aging is projected to be large.
  - Note: these demographic-impact projections assume the organization and financing of the health system remains unchanged (no health system reform).
- Epidemiological transition specifics:
  - Shift from communicable causes to noncommunicable chronic diseases is intensifying due to aging and economic development; chronic, age-related illnesses become more prevalent (World Bank, 2006).
  - Leading causes of death in most middle- and high-income countries (including ECCU region): cardiovascular diseases, neoplasm malicious, and diabetes.
  - Diabetes highlighted:
    - Described as the most significant “new disease” of the epidemiological transition (Caribbean Commission on Health and Development, 2005).
    - Requires lifelong, continuous medical care; economic impact expected to be large in next two decades.
    - One in ten of the adult population in the Caribbean region and one in five people over the age of 40 are currently living with diabetes.
    - Wild et al. (2004): diabetes estimated to affect 22,000 residents in the ECCU in 2000, with numbers projected to double by 2030.

### Design and implementation guidance
- Financing design guidance:
  - Best financing option is country-specific, depending on economic, cultural, institutional, demographic, epidemiological, and political-economy characteristics.
  - Regardless of option chosen, ensure financing is domestically based and sustainable; prefer taxation for recurrent costs.
- Benefit package design:
  - An appropriate and realistic benefit package should be designed with a progressive scenario for extension to ensure financial viability.
- Implementation guidance:
  - Preconditions for successful implementation include sound public administration, limited informal economy, and a transparent health financing system building social consensus.
  - Put in place monitoring processes and incentives to minimize abuse and over-utilization of services.

*Source: _wp0961 - References (excerpts from the IMF working paper content provided).*

### 28.1 percent of the population has abnormal glucose levels and 8.1 percent of those tested were diabetic. In

### _wp0961 - 28.1 percent of the population has abnormal glucose levels and 8.1 percent of those tested were diabetic. In

### Epidemiological profile and trends
- 28.1 percent of the population has abnormal glucose levels and 8.1 percent of those tested were diabetic.
- Almost half of St. Lucians have abnormal systolic blood pressure, described as the highest in the world.
- Wild et al. (2004) projected diabetes incidence in St. Lucia rising from 5,200 in 2000 to 11,300 by 2030.
- The Universal Health Care Diabetes and Hypertension (2007) pilot screening project recorded 13,000 diabetic cases as early as 2007.
- Diabetic cases are more concentrated in those aged 45 and above.
- Diabetes is also higher among women due to higher obesity rates.
- Increases in obesity trends are expected to further worsen the diabetes outlook.

### Estimated costs of treating diabetes — assumptions and scenarios
- Two scenarios for health cost evolution:
  - Scenario 1 assumes that health costs rise in line with the average historical health inflation rate.
  - Scenario 2 assumes that health cost inflation is two percentage points above the historical average.
- Annual cost assumption: EC$1,100 per diabetic case (based on Jamaican proxy including direct and indirect costs).
- Jamaican cost composition (proxy): more than 50 percent of the total cost is explained by laboratory diagnostic tests; one-fourth of the cost is associated with doctor charges/visits.
- Projection outcome: Even under the scenario where health costs rise with their historical trend, the cost of treating diabetes could triple for Dominica and St. Lucia by 2030.
- Note: Both scenarios project diabetic cases based on projections by Wild et al. (2004). For St. Lucia, actual numbers are used up to 2007; for later years data are interpolated using the trend by Wild et al. (2004).

### Implications for public finances and recommended financing source
- Population aging and the epidemiological transition will impose an increasingly large burden on public finances if the public sector continues to undertake the majority of spending on diabetes care.
- The study concludes that taxation should be the main source of universal health care financing, given projected increases in health spending amid population aging and the epidemiological transition.

### Comparison: General taxation versus payroll/social security contributions
- Five criteria for judging taxes (IMF and World Bank, 2005):
  - Revenue adequacy and stability
  - Efficiency
  - Equity
  - Ease of collection
  - Political acceptability
- Revenue adequacy and stability:
  - Payroll contribution revenue potential depends on labor market conditions, existing payroll tax rates, and the size of the informal economy.
  - Evidence cited: Kazakhstan collected only 40 percent of expected revenue after introducing universal coverage primarily financed via social security contributions.
  - Taiwan POC experienced difficulty when health costs rose faster than per capita GDP under a payroll-capped contribution system; the ceiling was raised in 2002.
  - General taxation (e.g., VATs and sales taxes) could better secure revenue adequacy for the ECCU by taxing a broader base including the unemployed, self-employed, tourism sector, and large informal sector.
- Efficiency:
  - General taxation’s broad base provides efficiency gains.
  - Payroll/social security contributions can raise tax evasion and reduce the size of the formal labor market; they also raise labor costs and may hinder economic growth, employment, and competitiveness.
- Equity:
  - VATs are typically regressive, but can be designed to be more progressive by exempting basic goods.
  - Payroll/social security contributions have been found to be progressive in some countries and regressive in others.
- Ease of collection:
  - Payroll contributions are automatically deducted, making collection easier where formal employment is large.
  - In the ECCU, a large informal economy and significant self-employed and agriculture sectors complicate payroll-based collection.
  - Spain and Iceland shifted from social security contributions to general taxation to avoid such problems.
- Political acceptability:
  - Payroll/social security contributions are often more politically acceptable than general taxation because the latter is typically regressive.
  - Political acceptability of general taxation can be enhanced if accompanied by a specific individual entitlement (a benefit tax), as in Ghana.

### Provider arrangements and provider payment mechanisms
- Many countries deliver universal coverage through a network of public providers while contracting non-governmental and private providers to expand choice, at the cost of higher administrative expenses.
- Fewer providers have been more successful at containing health cost increases, possibly by negotiating lower doctors’ fees.
- Examples of provider payment outcomes:
  - Aruba experienced escalating health costs when fee-for-service payments to specialists were applied without appropriate regulatory limits, leading to excessive service provision.
  - Canada contained health costs by establishing global budgets for hospitals and physicians and by monitoring service volumes per physician.
- Optimal provider payment design:
  - Literature suggests mixed reimbursement systems are optimal (some prospective payment combined with some cost-based payment).
  - A mix of fee-for-service and predetermined (salary or capitation) payments, possibly combined with positive patient cost sharing, can help avoid moral hazard and restrain supply.

### Designing the benefit package
- The benefit package choice should reflect the country’s health needs, medical and clinical priorities, and fiscal constraints.
- Given ECCU budget constraints and high debt levels, a universal, small package of services is recommended initially:
  - The package should encompass public goods, goods with externalities, and interventions with proven impact on health goals.
  - Clinical care and catastrophic expenditures could be left to targeted financing for the poor.
- Pharmaceuticals:
  - Drug benefits should apply to a defined and limited list of prescribed drugs.
  - Generic prescribing should be implemented when possible to contain costs.
- Targeting poor households:
  - A streamlined benefit package for the total population, combined with measures to target spending to the poor, is advised.
  - Conditional cash transfers have been used successfully to improve health and poverty outcomes with relatively minimal administrative costs.
- Role of private financing:
  - Private financing is suggested as more suitable for private goods such as individual ambulatory and institutional care services, pharmaceuticals, and medical interventions.

### Taiwan POC: illustrative national experience (Box 1 highlights)
- Prior to reform (1995): 59 percent of the population was insured under ten available insurance plans; uninsured 41 percent (mostly elderly and children below 14 years old).
- Post-reform features:
  - Comprehensive benefit package covering preventive and medical services, prescription drugs, dental services, Chinese medicine, AIDS/HIV treatment and home nurse visits.
  - Co-payment for outpatient visits ranged from $1.5 to clinics with referral to US$11 to academic medical centers without referral.
  - Inpatient coinsurance rate: 10–30 percent depending on length of stay.
  - Patient annual cap: total amount each patient pays capped to 10 percent of average per capita national income for a calendar year.
  - Poor households are exempt from all these costs.
- Outcomes:
  - By-end 1995 enrollment reached 92 percent.
  - Out-of-pocket spending dropped from 48 percent in 1993 to 30 percent in 2000.
  - Health spending rose from 5.3 percent in 1995 to 6 percent in 2000.
  - Average hospital admission rate increased from 110 per 1000 people in 1994 to 120 per 1000 in 1996.
  - Public satisfaction ratings reached 70 percent at some times (prior to 1995 ratings were around 30 percent).
  - The simple payer system enabled negotiation of lower fees and lower administrative and transaction costs; copayment, coinsurance, and global budgeting may have contributed to moderating health costs.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0961.pdf*

### Box 2. The Targeting of Health Services to the Poor: An International Perspective

### Box 2. The Targeting of Health Services to the Poor: An International Perspective

### International examples of targeting mechanisms
- Colombia
  - Established universal coverage under two kinds of basic coverage depending on the patient’s income.
  - Payroll-tax contributors have access to a minimum level of defined coverage and non-payroll tax contributors (i.e., the poor) have access to a subsidized Mandatory Health Plan that covers some additional services (World Bank, 2006).
- Suriname
  - Extremely poor individuals receive a health card permitting access to ambulatory, drugs and inpatient services from public and some private providers at zero or minimal cost.
  - Program is part of the Ministry of Social Affair’s social safety net; eligibility is based on field visits and applications (Lalta et al., 2005).
- Jamaica
  - The Program for Advancement through Health and Education (PATH) makes cash benefits available only when the targeted individual has met qualifications for school attendance and visits to health facilities (Lalta et al., 2005).
  - The program targets children up to 6 years, persons 65 years and older, pregnant and lactating women, and poor adults.

### Design features and benefit-package targeting
- Examples of cost-sharing and exemptions
  - France: patients pay 30 percent of the ambulatory care, 35–65 percent of the pharmaceuticals, and 40 percent of the laboratory costs; these costs are not payable for pregnant mothers, disabled children and people with specific diseases, such as diabetes and AIDS (Conference Board of Canada, 2004; UK National Audit Office, 2003).
  - United Kingdom: public financing decisions take into account cost effectiveness and efficiency allocation; Box 3 analyzes NICE’s role in prioritizing publicly-financed services.
- Recommendation on benefit-package scope
  - Initial package should be limited to ensure financial sustainability and include diseases policymakers rank high in priority.
  - Expanded packages that include private goods often benefit disproportionately the rich and less needy and could impair financial viability; downsizing later may be politically difficult.
- Institutional option
  - An independent institution could be set up to prioritize which services would be publicly financed based on clinical and cost-effectiveness.

### Preconditions for successfully implementing health-care reforms (financing and planning)
- Financing-side conditions
  - Tax administration should be sound, so that the tax effort is translated into strong tax revenues.
  - Health care plans should be part of the government’s medium- and long-term strategy; costs should be incorporated into the medium-term expenditure framework.
  - Public financing of health care should be transparent to build consensus and public support (Normand and Weber, 1994).
  - The informal sector should be tackled so that all health beneficiaries pay according to their ability.
- Planning and provision-side conditions
  - Ensure effectiveness of health spending by enhancing public expenditure management and accountability; link budgetary allocations with health system performance to generate accountability (example: Rwanda under its poverty reduction plan).
  - Wagstaff and Claeson (2004) finding: additional government spending on health has little impact on key Millennium Development Goals indicators in countries with poor governance as defined by the World Bank’s CPIA Index at or below 3.25 (scale 1 to 6).
    - At a CPIA score of 4 (one standard deviation above the mean), a 10 percent increase in the share of GDP devoted to government health spending results in a 7.2 percent decline in the maternal mortality ratio.
  - Establish appropriate provider payment arrangements to avoid over-utilization:
    - Fee-for-service should only be applied where monitoring safeguards exist.
    - If safeguards are not available, introduce capitation payment to all providers.
  - Design a realistic and progressive scenario for extending universal health care:
    - Conduct solid actuarial studies before implementation.
    - Phase extension progressively to ensure financial sustainability and allow administrative adjustment (Box 4 and Ron et al., 1990).
  - Improve health system efficiency; inefficiencies have important implications for long-term sustainability.
    - PriceWaterhouseCoopers (2005) finds governments, hospitals and physicians show greatest opportunity to eliminate wasteful spending in health care.

### Efficiency assessments and empirical findings for ECCU region
- Age-adjusted health-cost comparison method
  - Method accounts for different demographic structures; countries with older average populations have higher health-cost pressures.
  - Some ECCU countries (notably Grenada and Antigua and Barbuda) have very high age-adjusted health spending and thus appear to have high inefficiencies (Appendix III for methodology).
  - The age-adjusted hypothetical spending assumes an elasticity of health spending in percent of GDP to the share of population over 65 of 0.698.
- World Health Organization (2000) study findings
  - The ECCU region is relatively inefficient in terms of goal attainment and performance, with most ECCU countries (except Dominica) having relatively low rankings.
  - Examples: St. Kitts and Nevis ranked 98th and St. Vincent and the Grenadines ranked 92nd (out of a sample of 191 countries) on overall goal attainment.
  - WHO distinguishes overall health spending level from distribution across the population; tackling inefficiencies is pivotal before implementing universal health-care reforms.
  - Note: WHO ranking methodology has been criticized for reliance on imputed values, opinion-based indicator weights, and sensitivity of econometric specification.

### Examples of planning and gradual extension of universal coverage (Box 4)
- Republic of Korea
  - 1977: only 8 percent of the population was covered by social security insurance.
  - Reform process began in 1977 and included steady extension of social security-based health coverage: starting with workers of large enterprises (over 500 workers) and low-income households, then to smaller enterprises and civil servants; by 1987 medical insurance coverage was expanded to rural areas.
  - 1989: universal health insurance legislated.
- Taiwan Province of China
  - Preparations began more than a decade before legislation; first planning stage took two years of studies including progressive implementation.
  - 1987: first pilot project for expansion started with the farmer’s group.
  - 1995: legislation to create a mandatory health insurance scheme enacted.
  - 2000: mandatory scheme became fully operational (World Bank, 2006).

### Key policy recommendations and operational measures
- Secure sustainable funding primarily from domestic revenues, notably taxes, given limitations and erratic nature of development assistance.
- Ensure tax administration, financing transparency, and measures to address the informal sector before expanding coverage.
- Start with a streamlined benefit package meeting basic needs; avoid including private goods that disproportionately benefit the better-off.
- Use mixed provider reimbursement (capitation plus fee-for-service) with monitoring safeguards to limit abuse and over-utilization.
- Conduct actuarial and feasibility studies and implement phased expansion to retain stakeholder support and ensure financial sustainability.
- Strengthen public expenditure management, monitoring and evaluation mechanisms to improve accountability and effectiveness of additional health spending.

*Source: Box 2. The Targeting of Health Services to the Poor: An International Perspective*

### REFERENCES

### _wp0961 - REFERENCES

### Key findings and projections (selected from source content)
- ECCU: Long-term health care costs projected for 2005–35 are presented in percent of GDP with scenarios:
  - Elderly cost =5 times the average cost
  - Elderly cost =2 times the average cost
  - Same cost for all age groups
- Cost of treating diabetes, 2000–30, shown in percent of GDP for ECCU countries; note from source: “Cost of treating diabetes could increase significantly in some ECCU countries.”
- Diabetes prevalence projections by age group, 2000–30, presented in number of people for ECCU countries (age groups: 20–44, 45–64, 65+).

### St. Lucia: proposed universal health care — substantive analysis and projections (Appendix I)
- Objective:
  - Provide equal access to health care defined in a published package of limited services and benefits to all St. Lucians or non-nationals residing in the country for at least six months.
  - Emphasis on preventive services, comfort care and family planning services; inpatient, outpatient, limited overseas care, and inpatient pharmacy services primarily provided by two general hospitals and a new hospital under construction.
- Rationale:
  - Response to concerns about sustainability of current financing, poor quality and inequity of care, escalating health costs, and rising number of residents seeking medical treatment without paying.
- Estimated initial cost:
  - “The Taskforce… estimated that the initial cost of universal health care would be over US$20 million (over 3 percent of GDP).”
- Financing recommendations by the National Health Insurance Taskforce:
  - General taxation identified as the appropriate financing option.
  - Considered revenue sources: transaction-based taxes (environmental levy, “sin” taxes, VAT) and salary deduction.
  - Initial recommended financing mix: environmental levy and budgetary contribution as main sources, supplemented with a small salary deduction.
  - VAT could be considered in the future to “supplement or replace some or all of the sources of finance.”
  - Salary deductions appropriate only when the informal sector is captured; emphasis on broad-based taxation for flexibility to generate additional revenue if needed.
  - Property taxation rejected due to poor collection performance (only a relatively small percentage of property taxes are currently collected).

### St. Lucia: key numerical estimates (preserved exactly as in source)
- Estimated initial cost of universal health care: over US$20 million (over 3 percent of GDP).
- Health Expenditure Before and After the Introduction of Universal Health Care (In Percent of 2003 GDP) — figures presented in source (preserve exact formatting and values):
  - BeforeAfter
  - Public                                                                         3.34.6
  - Of which: Universal Health Care3.2
  - Private2.21.8
  - Total5.56.4
- Financing of Universal Health Care (million US dollars) — exact figures from source:
  - Estimated Cost21.7
  - Financed by:
    - Government Transfer11.0
    - Envriromental Levy7.8
    - SalaryDeduction2.9

### Policy recommendations (from Taskforce conclusions)
- Use general taxation as the primary financing mechanism for universal health care.
- Implement environmental levy and budgetary contributions as main initial financing sources.
- Supplement main sources with a small salary deduction.
- Consider VAT as a future supplement or replacement for other financing sources.
- Defer reliance on salary deductions until the informal sector is effectively captured.
- Do not rely on property taxation given current poor collection performance.

*Source: _wp0961 - REFERENCES (content and appendices provided in source PDF).*

### Appendix II. Projecting Long-term Health Care Costs Due to Population Aging

### Appendix II. Projecting Long-term Health Care Costs Due to Population Aging

### Methodology for demographic-driven projections
- Projections are derived by multiplying the last actual observation on health care cost per age group by projections for population by age group, so that projected changes in real health care spending reflect only changes in demographics.
- Health care costs by age group are not available for the ECCU region; existing data from other countries are used as a rough proxy to construct age-group specific health care costs for the ECCU.
- Health costs are based on WHO data; population projections are based on each country’s latest Actuarial Report.
- The study estimates health spending (in percent of GDP) that ECCU countries would face in the next three decades under three scenarios.

### Scenarios and assumptions
- Scenario 1: health costs for those aged 65 and above is five times the cost of the rest of the population (same assumption as Honjo (2006)).
- Scenario 2: health costs for those aged 65 and above is twice as high as that of the rest of the population.
- Scenario 3 (most optimistic): health costs are equal across all population ages.

### Caveats and limitations
- Projections should be viewed with caution: recent literature suggests demographic changes may not be the key driver of future health care expenditure.
- Bjornerud and Martins (2005) show that demographic changes accounted for only 10 percent of the average growth in health spending across OECD countries over the last 30 years.
- Other important drivers to consider if and when population projections materialize include:
  - changes in health status,
  - technological and medical progress,
  - economic growth,
  - changes in the organization and financing of health care systems.

---

### Appendix III. Estimating Age-adjusted Health Spending

### Purpose and approach
- Follows Esmail and Walker (2005) methodology to normalize health spending in a given nation to the level it would have incurred if its senior (over 65 years) population share equaled the level of a benchmark country.
- Canada is used as the benchmark case, following calculations used for The Bahamas (Esmail, 2006).
- Uses an elasticity of health spending (percent of GDP) to the share of population over 65 of 0.698, as estimated by Esmail and Walker (2005) for Canada.
- The elasticity implies that for every 1 percent increase in the share of population over 65, the increase in the share of GDP spent on health care will be around 0.7 percent.

### Age-adjusted Health Spending formula
- The formula is given by:
  Age-adjusted health spending in Estimated Country = 
  H
  E
  EB
  *)]698.0*1[
  −
  + 

### Variable definitions
- B: Senior’s proportion of population in benchmark country (in this case, Canada), based on U.S. Census Bureau International DataBase.
- E: Senior proportion of population in estimated country based on the country’s latest Actuarial Report.
- H: Health expenditure (in percent of GDP) in estimated country, based on WHO Health Statistics.

*Source: Appendix II and Appendix III, _wp0961 - Appendix II. Projecting Long-term Health Care Costs Due to Population Aging*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0961.pdf_