## 1. Japan: Decomposition of Changes in Health Spending, 1970–2011

## Source details

**Canonical URL:** [1. Japan: Decomposition of Changes in Health Spending, 1970–2011](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14142.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14142.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14142.pdf.json)

---

### Introduction and context
- Japan achieved universal health insurance coverage in 1961 and started to provide long-term care insurance in 2000.
- Public and private spending on health care and long-term care more than doubled from 4½ percent of GDP in 1990 to 10 percent of GDP in 2011.
- The corresponding increase in public spending on health care and long-term care was 4¾ percentage points of GDP during 1990–2011.
- The increase in health and long-term care spending, together with rising public pension spending, accounts for a large portion of the increase in the general government deficit of 11½ percentage points of GDP during 1990–2011.
- Japan’s old-age dependency ratio is the highest among advanced economies; nevertheless, public health spending as a percent of GDP is only marginally higher than the OECD average.

### Scope, data, and definitions
- “Health spending” encompasses total (public and private) spending on health care (HC) and long-term care (LTC) covered by the public insurance system, using data published by the Ministry of Health, Labor, and Welfare (MHLW).
- Private spending equals patient copayments; public spending equals total spending minus private spending.
- Health spending equals HC spending for 1970–2000 and the sum of HC and LTC spending for 2001–11. A structural break occurs around 2000 due to the launch of long-term care insurance.
- The chosen data differ from OECD System of Health Account (SHA) data in specific spending item coverage; Appendix I notes these differences and reports results are not sensitive to data source choice.
- HC spending is broken down into 18 age cohorts (age 0–4 to age 85+). For years before 1998, normalized per capita spending by age cohort is proxied by per-capita spending for the working age population (age 20–59). For LTC, per-capita spending uses cohort age 65–69.

### Methodology overview
- Focus: ratio of health spending to GDP and decomposition of the growth rate of this ratio into three components:
  - Ageing: changes in population shares by age cohort (older cohorts spend more).
  - Excess cost growth: benchmark per capita health spending rising faster than GDP per capita (captures non-demographic effects such as technological advances, Baumol effect, income elasticity, policy/institutional factors).
  - Spending profile shift: changes in per-capita spending by age cohort relative to the benchmark (detects “healthy ageing” if elderly per-capita spending at a given age declines).
- Excess cost growth is further decomposed to separate the effect of the output gap on excess cost growth (benchmark per capita spending in real terms and potential GDP per capita / GDP gap).

### Decomposition results: past spending increases (1970–2011)
- Overall change:
  - The ratio of health care and long-term care spending to GDP increased from 4.4 percent in 1990 to 9.8 percent in 2011.
  - This increase is equivalent to an average annual growth rate of 3.1 percent for the spending-to-GDP ratio over the last two decades.
  - The ratio had decreased during the 1980s at an annual rate of 0.2 percent, implying a discontinuous pickup in the early 1990s.
  - The growth rate of the spending ratio accelerated to 3.5 percent annually during 2003–11.
- Contribution breakdown (key quantitative findings):
  - Ageing effect:
    - Accounted for about two-thirds of the spending ratio’s increase during the last two decades.
    - Grew by 1.8 percent annually (mean 1.8; St. dev. 0.2).
    - Health care component: 1.6 percent annually (mean 1.6; St. dev. 0.2).
    - Long-term care component: 4.4 percent annually (mean 4.4; St. dev. 0.6).
    - Ageing accelerated from an annual growth rate of 1 percent during the 1970s to 1.9 percent during the last decade.
  - Excess cost growth:
    - Contributed on average 0.9 percent annually to the spending ratio (mean 0.9; St. dev. 2.9).
    - Benchmark per capita spending (real) grew on average 1.8 percent annually (mean 1.8; St. dev. 2.1).
    - Potential GDP per capita growth averaged 0.9 percent annually (mean 0.9; St. dev. 0.9).
    - The contribution of the GDP gap was negligible (mean 0.0; St. dev. 1.7).
    - The pickup in excess cost growth from minus 1.5 percent during the 1980s to 1.3 percent during 2003–11 reflects a decline in per-capita GDP growth during the “lost decades,” not widening of the output gap.
  - Spending profile shift:
    - Contributed 0.3 percent annually since 1998 (mean 0.3; St. dev. 1.2).
    - Health care spending profile shift: 0.4 percent annually.
    - By cohort: age 0–19: 0.1 percent; age 20–59: 0.0 percent; age 60 and over: 0.2 percent.
    - Long-term care spending profile shift: 0.2 percent.
    - The positive spending profile shift indicates that healthy ageing did not occur in Japan during the last decade (elderly per-capita spending at given ages did not decline).
- Memorandum items:
  - Ratio of health care spending to GDP: mean 3.0; St. dev. 2.7.
  - Ratio of long-term care spending to GDP: mean 5.5; St. dev. 4.5.
- Interpretation:
  - Ageing is the dominant driver of the post-1990 increase in health spending-to-GDP ratio in Japan because the per-capita spending difference between old and young is more pronounced in Japan than in other advanced countries.
  - Excess cost growth in Japan over the last two decades was positive and at a rate comparable to other advanced economies, accounting for roughly one-third of the spending increase.
  - Healthy ageing (lower per-capita spending at given ages as life expectancy rises) has not materialized in Japan during the last decade.

### Aging and excess cost growth: empirical findings
- The ageing component for Japan grew at 1.8 percent annually; this was the highest among advanced economies (the median increase was 0.5 percent).
- Japan’s excess cost growth (inclusive of spending profile shift) was 1.3 percent annually; the median among peers was 1.1 percent annually.
- Per-capita health spending by age in Japan:
  - A person age 85 years and more spends about 7 times as much on health care as a person age 40–44 years.
  - A person age 85 years and more spends about 30 times as much on long-term care as a person age 65–69 years.
- These age-ratios are the highest among OECD countries and are also pronounced for cohorts age 70 years and older.

### Spending projection scenarios and key projected outcomes (2010–2030 and beyond)
- Framework: projections done separately for health care (HC) and long-term care (LTC), then aggregated; scenarios consider annual excess cost growth of 0 percent, 1 percent, and 2 percent.
- Scenario outcomes (health spending as percent of GDP):
  - Excess cost growth = 0 percent per year:
    - Health spending rises from 9½ percent of GDP in 2010 to 13 percent of GDP in 2030.
  - Excess cost growth = 1 percent per year:
    - Health spending rises to 15½ percent of GDP in 2030 (a 6 percentage point increase between 2010 and 2030).
  - Excess cost growth = 2 percent per year:
    - Health spending rises to 19 percent of GDP by 2030.
- Longer-horizon implications:
  - If 1 percent annual excess cost growth continues through 2060, spending will exceed a quarter of GDP.
  - Uncertainty surrounding the spending projection grows dramatically by 2060.
- Statistical fan-chart assessment (through 2030):
  - Assumes annual excess cost growth follows a white noise process with the mean and standard deviation observed during 1990–2011 (excluding 2000–02) and population under the medium variant.
  - In 2030, the lower bound of the spending ratio’s 90 percent confidence interval roughly coincides with the projection under no excess cost growth; the upper bound roughly coincides with the projection under 2 percent excess cost growth.
- Sensitivity to fertility and mortality assumptions:
  - Projected spending ratio in 2030 under low-fertility-and-low-mortality differs by only 1 percentage point from that under high-fertility-and-high-mortality.

### Healthy ageing and its potential impact
- Population projection (IPSS, medium variant): life expectancy is expected to increase by 2.3 years between 2010 and 2030.
- Healthy ageing scenario:
  - If per-capita spending by age shifts downward in line with improved health status accompanying higher life expectancy, healthy ageing could reduce the projected increase in health spending over 2010–30 by about 1 percentage point of GDP.
- Supporting literature note:
  - Tajika and Kikuchi (2014) estimate that a decline in the mortality rate and improvements in health status of the elderly would reduce per-capita health spending in 2060 by 3.7 percent and 13.9 percent, respectively.

### Interaction with economic growth (decoupling scenarios)
- Government-produced illustrative GDP growth scenarios for 2013–2022:
  - Reform scenario: real GDP growth about 2 percent per year.
  - Pessimistic scenario: real GDP growth about 1 percent per year.
- Illustration assuming benchmark per-capita health spending growth at 2 percent annually and average population growth of -0.5 percent per year during 2010–30 (mid-fertility-mid-mortality):
  - Reform scenario (GDP growth ≈ 2 percent) implies excess cost growth of -0.5 percent per year, corresponding to an increase in health spending by 2.2 percentage points of GDP over 2010–30.
  - Pessimistic scenario (GDP growth ≈ 1 percent) implies excess cost growth of 0.5 percent per year, corresponding to an increase in health spending by 4.7 percentage points of GDP over 2010–30.

### Financing mix in 2010 and projected evolution to 2030
- Financing shares in 2010:
  - Premium contributions: 48 percent of health spending (48 percent for HC and 45 percent for LTC).
  - Patient copayments (out-of-pocket): 13 percent of health spending (13 percent for HC and 10 percent for LTC).
  - Government transfers (central and local): 39 percent of health spending (residual).
- Age-profile of financing burdens (2010):
  - Working-age groups (age 20–55): age-specific premium contributions exceed their HC spending.
  - Contribution-to-spending ratio declines with age; it falls below 10 percent for those aged 75 years and older.
  - Copayment-to-spending ratio is above 20 percent for people age 5–69 years and below 10 percent for age 70 years and more.
- Projection approach for financing mix (scenario with excess cost growth = 1 percent per year):
  - HC: apply 2010 age-specific premium contribution rates (percent of spending) and copayment rates by age cohort to projected HC spending.
  - LTC: project financing mix using statutory splitting rule (45 percent by premium contributions, 10 percent by copayments, and 45 percent by government transfers).
  - Government transfers computed as residual (total HC and LTC spending minus sum of premium contributions and copayments).
- Projected changes (2010–2030) under the 1 percent excess cost growth scenario:
  - Copayments share of total health spending is projected to decline marginally from 13 percent in 2010 to 12 percent in 2030, reflecting the rising share of elderly whose effective copayment rate is lower.
  - As a percent of GDP, copayments will increase by 0.6 percentage points over 2010–30 because assumed excess cost growth raises the spending ratio to GDP.
  - Premium contributions are projected to increase as a percent of GDP.
- Premium contributions projections:
  - Premium contributions as a percent of health spending are projected to decrease from 48 percent in 2010 to 43 percent in 2030.
  - Premium contributions as a percent of GDP rise from 4.4 percent in 2010 to 6.6 percent in 2030.
  - For LTC, premium contributions will increase in parallel with total spending owing to their fixed financing share (45 percent).
  - For HC, excess cost growth raises HC spending regardless of age cohort, bringing up premium contributions relative to GDP.
  - The increase in the contribution rate for HC is disproportionately higher for the working-age population:
    - Contributions for the working-age population increase from 3¾ percent of labor income in 2010 to 12 percent in 2030.
  - Government transfers are projected to increase from 3.6 percent of GDP in 2010 to 7.1 percent of GDP in 2030.
  - The “financing gap” of 3.5 percentage points of GDP by 2030 is equivalent to raising the consumption tax rate by 7 percentage points.

### Projections and sensitivity scenarios (summary table notes)
- IMF staff projections highlight the role of excess cost growth and healthy ageing assumptions:
  - HC + LTC, public + private, excess cost growth = 0%: Total 3.4; Ageing/demographic 3.4; Other 0.0; Combined 6.8 (2010–30).
  - HC + LTC, public + private, excess cost growth = 1%: Total 6.2; Ageing/demographic 3.4; Other 2.8.
  - With healthy ageing and excess cost growth = 0%: HC + LTC, public + private: Total 2.5; Ageing/demographic 2.5; Other 0.0.
- Comparison with other projections:
  - Latest government projection envisages an increase of 4–5 percentage points of GDP for public health spending over the next two decades.
  - IMF Fiscal Monitor (IMF, 2014) projects public health spending to rise by about 2 percentage points of GDP over 2010–30 (assumes zero excess cost growth and healthy ageing).
  - OECD envisages public health spending to rise by 2–3 percentage points of GDP for 2010–30 and 3–8 percentage points for 2010–60.
  - Long-run projections by the Fiscal System Council (2014a) and Ueda et al. (2011, 2014) are more pessimistic than the IMF projection with no excess cost growth, but more optimistic than the IMF projection with 1 percent excess cost growth.

### Policy options to contain health spending growth
- Categories of policy options:
  - Macro-level controls to cap spending or regulate price and quantity of services.
  - Micro-level reforms to improve the functioning of the health system and increase spending efficiency.
  - Demand-side reforms to curb health care demand by increasing cost sharing by patients.
- Micro-level reforms (examples):
  - Introduce gatekeepers for secondary and tertiary care.
  - Improve public management of health care resources and strengthen prefecture coordination.
  - Reform provider payment arrangements to reduce supplier-induced demand (shift from fee-for-service toward DRG-type arrangements).
  - Encourage use of generic drugs (Japan’s utilization was 24 percent in 2009).
  - Promote prevention of lifestyle related diseases (target high blood pressure, smoking, high blood glucose).
- Macro-level controls (design caveats):
  - Introduce budget caps and fine-tune supply constraints and price controls.
  - Risks: budget caps can limit access and lengthen waiting times; supply constraints must be targeted (beds, equipment).
  - Japan’s OECD indicator for the stringency of the budget constraint is 0 (indicator ranges between 0 and 6; OECD average is 2.0).
  - Japan’s OECD indicator for regulation of health workforce and equipment is 1.5 (indicator ranges between 0 and 6; OECD average is 2.9).
- Demand-side and financing reforms:
  - Raise patient copayments to limit increases in public health spending and reduce excess demand; implement progressive copayment rates to preserve equity.
  - Raise premium contributions while correcting inequalities among insurers and winding down preferential treatments (for example, exempting dependent spouses under employment-based insurance from contribution payments).
  - Address exemption where dependent spouses with annual earnings below 1.3 million yen are exempted from premiums while similar dependents under NHIP pay contributions.
  - Consider progressive contribution schedules to reduce tax-wedge effects on low-skilled workers.

### Price controls, provider incentives, and gatekeeping
- Price controls:
  - Prices of services covered by health and long-term care insurance are determined by the government and adjusted every two years for health care and every three years for long-term care.
  - The government restrained price increases during 2003–08 and succeeded in reining in health spending especially in 2006.
  - Price control measures in 2006 and 2008 were followed by a pickup in spending in the subsequent year, consistent with providers responding by increasing service volumes or directing patients to higher-cost services.
- Provider payment arrangements:
  - Fee-for-service is widely used; limited use of DRG-type arrangements for hospitals.
  - OECD indicator of strength of providers’ incentives to increase volume of care: Japan’s score is 5.7; OECD average is 3.1.
  - Evidence: very high doctor consultations per capita and average length of hospital stay compared with other advanced economies.
  - Positive correlation observed between average length of hospital stay and number of beds per capita at prefecture levels and between average length of stay and per-capita health spending of the elderly.
- Gatekeeping and public management:
  - Japan has no gatekeeping arrangement; patients have “free access” to choose hospitals and providers without referrals.
  - 20 out of 29 OECD countries use gatekeeping arrangements at least to some extent.
  - Share of hospitals offering highly acute care was about 40 percent in 2010, higher than government’s ideal (under 20 percent).
  - Policy option: strengthen regulatory power of prefecture governments to monitor and coordinate resource allocations.

### Health system indicators (selected comparisons)
- Physicians per 1,000 population (head counts): Japan 2.2; OECD median 3.3.
- Total hospital beds per 1,000 population: Japan 13.4; OECD median 4.2.
- Curative (acute) care beds per 1,000 population: Japan 8.0; OECD median 3.0.
- MRI units, total, per million population: Japan 46.9; OECD median 10.7.
- CT scanners, total, per million population: Japan 101.3; OECD median 16.1.
- Doctors consultations, number per capita: Japan 13.1; OECD median 6.6.
- Average length of stay at hospitals, days: Japan 18.2; OECD median 7.2.

### Options to raise copayments and premium contributions
- Raising patient copayments:
  - Copayments in Japan are relatively low in levels and as percent of total spending compared with other OECD countries.
  - Shigeoka (2013) finds out-of-pocket medical spending drops by about 70 percent when a patient turns 70 years old (copayment rate decreases from 30 percent to 10 percent); this reduction drives up outpatient visits and inpatient admissions by 10 percent and 8 percent, respectively, with no improvement in mortality.
  - Recommendation: increase copayments with progressive design to protect the poor.
- Raising premium contributions:
  - Increase ceilings on individual premium contributions (proposed by the 2013 National Council Report).
  - Risk: higher premium contributions could increase the labor tax wedge and reduce labor supply and demand.
  - Mitigation: introduce a progressive contribution schedule to reduce tax wedges for low-skilled workers.
  - Address inequality: average contribution rate for the NHIP is three times as high as that for employment-based Kumiai Kempo insurance programs.

### Estimated fiscal savings (illustrative estimates for 2030, percentage points of GDP)
- Options to contain health spending growth:
  - Introduce budget caps: 0.2
  - Strengthen supply constraints: 0.2
  - Introduce gatekeeping: 0.3
  - Incentivize use of generic drugs: 0.3
  - Promote prevention of lifestyle related diseases: 0.1
- Options to raise copayments and premium contributions:
  - Increase copayment rates by 5 percent for elderly and LTC users: 1/0.6 (interpreted as illustrative savings of 1.0 and 0.6 in alternative assumptions)
  - Collect premiums from dependent spouses: 2/0.3 (illustrative)
- Total illustrative savings: 1.9 percentage points of GDP (aggregate of listed items).
- Caveats:
  - Estimates are illustrative and based on econometric models and government estimates; they are subject to estimation error and possible unintended effects on utilization and outcomes.

### Concluding remarks (key findings)
- Two-thirds of the spending increase over the last two decades resulted from population ageing; the rest from excess cost growth.
- Ageing effect alone will raise health spending from 9½ percent of GDP in 2010 to 13 percent of GDP in 2030.
- If excess cost growth continues at the trend rate observed during the last two decades, health spending will reach 15½ percent of GDP in 2030.
- Worst-case: if excess cost growth is twice the trend rate, health spending reaches 19 percent of GDP in 2030.
- Healthy ageing in line with expected life expectancy increases could reduce health spending in 2030 by about 1 percentage point of GDP relative to the baseline.
- There is scope to introduce micro-level and macro-level reforms to potentially contain spending; on financing, raising copayments and premium contributions can contain government transfers to the health system but should be designed to preserve equity.
- Research needs:
  - Large uncertainty surrounds projections and the sources of excess cost growth.
  - Key policy question: How to curtail per-capita spending by the elderly.
  - Suggested micro-level research: examine factors contributing to disparities in per-capita spending by the elderly across regions.

### Box 1. Japan’s Health System (key features)
- Coverage:
  - Universal HC and LTC.
  - HC system virtually covers the entire population.
  - LTC covers people age 65 and older and those between ages 40–64 who meet eligibility criteria.
- Fragmentation:
  - Over 3,000 plans by residence and employment; insured enjoy “free access” to any providers.
- Hospitals:
  - Privately owned hospitals account for a majority of total hospital beds; operate generally on a not-for-profit principle.
- Insurers:
  - People aged 75 and above covered by insurer operated by prefecture government.
  - For those aged 74 and below: Kyokai Kempo, Kumiai Kempo, Mutual Aid Associations, National Health Insurance Program (NHIP).
  - Around 1,500 Kumiai Kenpo insurers; around 1,800 NHIPs.
- Contributions:
  - Dependent spouses with annual income below 1.3 million yen are exempted from paying contributions.
  - Contribution rates for NHIP vary across municipalities and are based on family unit.
- Patient copayments:
  - HC copayment rate uniform at 30 percent, except pre-school children (20 percent) and the elderly (10 percent).
  - LTC copayment uniform at 10 percent.
  - Copayments are subject to monthly cap based on age and income.
  - People on welfare receive free care (such families account for about 3 percent of total households).
- Government controls:
  - Central government sets unit prices of all medical procedures, drugs, and devices every two years; sets LTC prices every three years.
  - Government regulates health care workforce through enrollment quotas for medical schools.

### Appendix I: data sources and methodological notes
- Two data sources for HC and LTC spending:
  - Government data: spending covered by public health and long-term care insurance systems (Ministry of Health, Labor, and Welfare).
  - OECD data: System of Health Account.
- Data coverage differences:
  - OECD includes items excluded from government data (e.g., procedures not covered by public health insurance such as expenses for normal childbirth and comprehensive medical checkups, and capital expenditure by hospitals).
  - Government data includes items excluded from OECD data (e.g., personal care services covered by public LTC insurance).
- Observed divergence from 2000 onwards reflects large LTC spending excluded from OECD data and other coverage differences.
- Projection methodology highlights:
  - Population: IPSS 2012 medium variant for five-year age groups.
  - Labor force participation: age-specific LFP from Statistics Bureau; aggregate LFP declines from 57½ percent in 2010 to 55½ percent in 2030 and 50 percent in 2060.
  - Employment and earnings: unemployment assumed to decline to 4 percent by 2018 and remain; nominal wage growth 2¼ percent during 2018–22 and 2½ percent from 2023 onward; inflation fixed at 1 percent in the long term.
  - Pension benefits: follow baseline scenario used by Kashiwase, Nozaki, and Tokuoka (2012).

*Source — _wp14142 - 1. Japan: Decomposition of Changes in Health Spending, 1970–2011 (chapter content provided from the source PDF).*

### 1. Japan: Decomposition of Changes in Health Spending, 1970–2011 ...................................10

### 1. Japan: Decomposition of Changes in Health Spending, 1970–2011

### Introduction and context
- Japan achieved universal health insurance coverage in 1961 and started to provide long-term care insurance in 2000.
- Public and private spending on health care and long-term care more than doubled from 4½ percent of GDP in 1990 to 10 percent of GDP in 2011.
- The corresponding increase in public spending on health care and long-term care was 4¾ percentage points of GDP during 1990–2011.
- The increase in health and long-term care spending, together with rising public pension spending, accounts for a large portion of the increase in the general government deficit of 11½ percentage points of GDP during 1990–2011.
- Japan’s old-age dependency ratio is the highest among advanced economies; nevertheless, public health spending as a percent of GDP is only marginally higher than the OECD average.

### Scope, data, and definitions
- “Health spending” in this analysis encompasses total (public and private) spending on health care (HC) and long-term care (LTC) covered by the public insurance system, using data published by the Ministry of Health, Labor, and Welfare (MHLW).
- Private spending equals patient copayments; public spending equals total spending minus private spending.
- Health spending equals HC spending for 1970–2000 and the sum of HC and LTC spending for 2001–11. A structural break occurs around 2000 due to the launch of long-term care insurance.
- The chosen data differ from OECD System of Health Account (SHA) data in specific spending item coverage; Appendix I notes these differences and reports results are not sensitive to data source choice.
- HC spending is broken down into 18 age cohorts (age 0–4 to age 85+). For years before 1998, normalized per capita spending by age cohort is proxied by per-capita spending for the working age population (age 20–59). For LTC, per-capita spending uses cohort age 65–69.

### Methodology overview
- Focus: ratio of health spending to GDP and decomposition of the growth rate of this ratio into three components:
  - Ageing: changes in population shares by age cohort (older cohorts spend more).
  - Excess cost growth: benchmark per capita health spending rising faster than GDP per capita (captures non-demographic effects such as technological advances, Baumol effect, income elasticity, policy/institutional factors).
  - Spending profile shift: changes in per-capita spending by age cohort relative to the benchmark (detects “healthy ageing” if elderly per-capita spending at a given age declines).
- Excess cost growth is further decomposed to separate the effect of the output gap on excess cost growth (benchmark per capita spending in real terms and potential GDP per capita / GDP gap).

### Decomposition results: past spending increases (1970–2011)
- Overall change:
  - The ratio of health care and long-term care spending to GDP increased from 4.4 percent in 1990 to 9.8 percent in 2011.
  - This increase is equivalent to an average annual growth rate of 3.1 percent for the spending-to-GDP ratio over the last two decades.
  - The ratio had decreased during the 1980s at an annual rate of 0.2 percent, implying a discontinuous pickup in the early 1990s.
  - The growth rate of the spending ratio accelerated to 3.5 percent annually during 2003–11.
- Contribution breakdown (key quantitative findings):
  - Ageing effect:
    - Accounted for about two-thirds of the spending ratio’s increase during the last two decades.
    - Grew by 1.8 percent annually (mean 1.8; St. dev. 0.2).
    - Health care component: 1.6 percent annually (mean 1.6; St. dev. 0.2).
    - Long-term care component: 4.4 percent annually (mean 4.4; St. dev. 0.6).
    - Ageing accelerated from an annual growth rate of 1 percent during the 1970s to 1.9 percent during the last decade.
  - Excess cost growth:
    - Contributed on average 0.9 percent annually to the spending ratio (mean 0.9; St. dev. 2.9).
    - Benchmark per capita spending (real) grew on average 1.8 percent annually (mean 1.8; St. dev. 2.1).
    - Potential GDP per capita growth averaged 0.9 percent annually (mean 0.9; St. dev. 0.9).
    - The contribution of the GDP gap was negligible (mean 0.0; St. dev. 1.7).
    - The pickup in excess cost growth from minus 1.5 percent during the 1980s to 1.3 percent during 2003–11 reflects a decline in per-capita GDP growth during the “lost decades,” not widening of the output gap.
  - Spending profile shift:
    - Contributed 0.3 percent annually since 1998 (mean 0.3; St. dev. 1.2).
    - Health care spending profile shift: 0.4 percent annually.
    - By cohort: age 0–19: 0.1 percent; age 20–59: 0.0 percent; age 60 and over: 0.2 percent.
    - Long-term care spending profile shift: 0.2 percent.
    - The positive spending profile shift indicates that healthy ageing did not occur in Japan during the last decade (elderly per-capita spending at given ages did not decline).
- Memorandum items (from Table 1):
  - Ratio of health care spending to GDP: mean 3.0; St. dev. 2.7.
  - Ratio of long-term care spending to GDP: mean 5.5; St. dev. 4.5.
- Interpretation:
  - Ageing is the dominant driver of the post-1990 increase in health spending-to-GDP ratio in Japan because the per-capita spending difference between old and young is more pronounced in Japan than in other advanced countries.
  - Excess cost growth in Japan over the last two decades was positive and at a rate comparable to other advanced economies, accounting for roughly one-third of the spending increase.
  - Healthy ageing (lower per-capita spending at given ages as life expectancy rises) has not materialized in Japan during the last decade.

### Projections and scenarios (summary from introductory findings)
- Base demographic effect projection:
  - Ageing alone will raise health spending from 9½ percent of GDP in 2010 to 13 percent of GDP in 2030.
- If excess cost growth continues at the trend rate observed during the last two decades:
  - Health spending projected to reach 15½ percent of GDP in 2030.
  - Projected financing of the increase by 2030: premium contributions (2 percentage points of GDP), government transfers (3½ percentage points of GDP), and patient copayments (½ percentage points of GDP).
  - The estimated increase in government transfers is equivalent to raising the consumption tax rate by 7 percentage points.
  - The increase in premium contributions will need to be achieved by higher payroll tax rates amid a shrinking working-age population, with adverse labor supply and demand effects.
- Worst-case scenario:
  - If excess cost growth is twice the trend rate, health spending reaches 19 percent of GDP in 2030.
- Healthy ageing scenario:
  - If healthy ageing occurs in line with expected life expectancy increases, health spending in 2030 could be reduced by 1 percentage point of GDP relative to the baseline projection.

### Reform options and potential fiscal impact
- Micro-level reforms to contain spending growth without harming health outcomes:
  - Introduce gatekeepers for secondary and tertiary care.
  - Improve public management of health care resources.
  - Reform provider payment arrangements to reduce supplier-induced demand.
  - Encourage use of generic drugs.
- Macro-level controls (to be designed carefully):
  - Introduce budget caps.
  - Fine-tune supply constraints and price controls.
- Financing-side reforms:
  - Raise patient copayments to limit increases in public health spending and reduce excess demand; implement progressive copayment rates (higher rates for the rich than for the poor) to preserve equity and insurance protection.
  - Raise premium contributions while correcting inequalities among insurers and winding down preferential treatments (for example, exempting dependent spouses under employment-based insurance from contribution payments).
- Overall: Reforms can materially contain health spending growth; a mix of micro-level efficiency improvements, carefully designed macro controls, and equitable financing adjustments is advocated.

### Organization of the rest of the study (as described)
- Section II: Decomposition of past increase, cross-country context, projections on HC and LTC spending and financing mix, comparison with other studies.
- Section III: Discussion of reform options and estimates of fiscal savings.
- Section IV: Concluding remarks.

*Italic: Source — _wp14142 - 1. Japan: Decomposition of Changes in Health Spending, 1970–2011 (chapter content provided from the source PDF).*

### 1.8 percent annually). This is a result of the significant ageing effect for Japan—the

### _wp14142 - 1.8 percent annually). This is a result of the significant ageing effect for Japan—the

### Aging and excess cost growth: empirical findings
- The ageing component for Japan grew at 1.8 percent annually; this was the highest among advanced economies (the median increase was 0.5 percent).
- Japan’s excess cost growth (inclusive of spending profile shift) was 1.3 percent annually; the median among peers was 1.1 percent annually.
- Per-capita health spending by age in Japan:
  - A person age 85 years and more spends about 7 times as much on health care as a person age 40–44 years.
  - A person age 85 years and more spends about 30 times as much on long-term care as a person age 65–69 years.
- These age-ratios are the highest among OECD countries and are also pronounced for cohorts age 70 years and older.
- For non-Japan countries in the analysis, the profile of per-capita health spending by age cohort is derived from De La Maisonneuve and Oliveira Martins (2013); that profile varies by country but does not change over time for the exercise, so spending profile shift is regarded as part of excess cost growth.

### Spending projection scenarios and key projected outcomes (2010–2030 and beyond)
- Framework: projections done separately for health care (HC) and long-term care (LTC), then aggregated; scenarios consider annual excess cost growth of 0 percent, 1 percent, and 2 percent.
- Scenario outcomes (health spending as percent of GDP):
  - With excess cost growth = 0 percent per year:
    - Health spending rises from 9½ percent of GDP in 2010 to 13 percent of GDP in 2030.
  - With excess cost growth = 1 percent per year:
    - Health spending rises to 15½ percent of GDP in 2030 (a 6 percentage point increase between 2010 and 2030).
  - With excess cost growth = 2 percent per year:
    - Health spending rises to 19 percent of GDP by 2030 (the ratio will double by 2030 relative to lower baselines).
- Longer-horizon implications:
  - If 1 percent annual excess cost growth continues through 2060, spending will exceed a quarter of GDP.
  - The uncertainty surrounding the spending projection grows dramatically by 2060.
- Statistical fan-chart assessment (through 2030):
  - Assumes annual excess cost growth follows a white noise process with the mean and standard deviation observed during 1990–2011 (excluding 2000–02) and population under the medium variant.
  - In 2030, the lower bound of the spending ratio’s 90 percent confidence interval roughly coincides with the projection under no excess cost growth; the upper bound roughly coincides with the projection under 2 percent excess cost growth.
- Sensitivity to fertility and mortality assumptions:
  - Projected spending ratio in 2030 under low-fertility-and-low-mortality differs by only 1 percentage point from that under high-fertility-and-high-mortality.

### Healthy ageing and its potential impact
- Population projection (IPSS, medium variant): Japanese life expectancy is expected to increase by 2.3 years between 2010 and 2030.
- Healthy ageing scenario:
  - If per-capita spending by age shifts downward in line with improved health status accompanying higher life expectancy, healthy ageing could reduce the projected increase in health spending over 2010–30 by about 1 percentage point of GDP.
- Supporting literature note:
  - Tajika and Kikuchi (2014) estimate that a decline in the mortality rate and improvements in health status of the elderly would reduce per-capita health spending in 2060 by 3.7 percent and 13.9 percent, respectively.

### Interaction with economic growth (decoupling scenarios)
- Government-produced illustrative GDP growth scenarios for 2013–2022:
  - Reform scenario: real GDP growth about 2 percent per year.
  - Pessimistic scenario: real GDP growth about 1 percent per year.
- Illustration assuming benchmark per-capita health spending growth at 2 percent annually (observed in last decades) and average population growth of -0.5 percent per year during 2010–30 (mid-fertility-mid-mortality):
  - Reform scenario (GDP growth ≈ 2 percent) would imply excess cost growth of -0.5 percent per year, corresponding to an increase in health spending by 2.2 percentage points of GDP over 2010–30.
  - Pessimistic scenario (GDP growth ≈ 1 percent) would imply excess cost growth of 0.5 percent per year, corresponding to an increase in health spending by 4.7 percentage points of GDP over 2010–30.

### Financing mix in 2010 and projected evolution to 2030
- Financing shares in 2010:
  - Premium contributions: 48 percent of health spending (48 percent for HC and 45 percent for LTC).
  - Patient copayments (out-of-pocket): 13 percent of health spending (13 percent for HC and 10 percent for LTC).
  - Government transfers (central and local): 39 percent of health spending (residual).
- Age-profile of financing burdens (2010):
  - Working-age groups (age 20–55): age-specific premium contributions exceed their HC spending.
  - Contribution-to-spending ratio declines with age; it falls below 10 percent for those aged 75 years and older.
  - Copayment-to-spending ratio is above 20 percent for people age 5–69 years and below 10 percent for age 70 years and more.
- Projection approach for financing mix (scenario with excess cost growth = 1 percent per year):
  - HC: apply 2010 age-specific premium contribution rates (percent of spending) and copayment rates by age cohort to projected HC spending.
  - LTC: project financing mix using statutory splitting rule (45 percent by premium contributions, 10 percent by copayments, and 45 percent by government transfers).
  - Government transfers computed as residual (total HC and LTC spending minus sum of premium contributions and copayments).
- Projected changes (2010–2030) under the 1 percent excess cost growth scenario:
  - Copayments share of total health spending is projected to decline marginally from 13 percent in 2010 to 12 percent in 2030, reflecting the rising share of elderly whose effective copayment rate is lower.
  - As a percent of GDP, copayments will increase by 0.6 percentage points over 2010–30 because assumed excess cost growth raises the spending ratio to GDP.
  - Premium contributions are projected to increase as a percent of GDP (projection continues beyond excerpt).

*Source: IMF staff estimates and cited Japanese sources as presented in the supplied content unit.*

### 4.4 percent in 2010 to 6.6 percent in 2030.

### _wp14142 - 4.4 percent in 2010 to 6.6 percent in 2030.

### Financing mix and premium contributions
- Premium contributions as a percent of health spending are projected to decrease from 48 percent in 2010 to 43 percent in 2030.
- Premium contributions as a percent of GDP rise from 4.4 percent in 2010 to 6.6 percent in 2030.
- For LTC, premium contributions will increase in parallel with total spending owing to their fixed financing share (45 percent).
- For HC, excess cost growth raises HC spending regardless of age cohort, bringing up premium contributions relative to GDP.
- The increase in the contribution rate for HC is disproportionately higher for the working-age population:
  - Contributions for the working-age population increase from 3¾ percent of labor income in 2010 to 12 percent in 2030.
- Government transfers are projected to increase from 3.6 percent of GDP in 2010 to 7.1 percent of GDP in 2030.
- The “financing gap” of 3.5 percentage points of GDP by 2030 is equivalent to raising the consumption tax rate by 7 percentage points.

### Projections and sensitivity scenarios
- IMF staff projections highlight the role of excess cost growth and healthy ageing assumptions:
  - Our projection, HC + LTC, public + private, excess cost growth = 0%: Total 3.4; Ageing/demographic 3.4; Other 0.0; Combined 6.8 (2010–30).
  - Our projection, HC + LTC, public + private, excess cost growth = 1%: Total 6.2; Ageing/demographic 3.4; Other 2.8; Combined 17.2? (table format retained as in source).
  - With healthy ageing and excess cost growth = 0%: HC + LTC, public + private: Total 2.5; Ageing/demographic 2.5; Other 0.0.
- Comparison with other projections (key observations):
  - The latest government projection envisages an increase of 4–5 percentage points of GDP for public health spending over the next two decades; the reform scenario projects higher spending than the no-reform scenario.
  - The IMF Fiscal Monitor (IMF, 2014) projects public health spending to rise by about 2 percentage points of GDP over 2010–30; this reflects an assumption of zero excess cost growth and healthy ageing.
  - The OECD (De La Maisonneuve and Oliveira Martins, 2013) envisages public health spending to rise by 2–3 percentage points of GDP for 2010–30 and 3–8 percentage points for 2010–60.
  - The long-run projections by the Fiscal System Council (2014a) and Ueda et al. (2011, 2014) are more pessimistic than our projection with no excess cost growth, but more optimistic than our projection with 1 percent excess cost growth.

### Policy options to contain health spending growth
- Categories of policy options:
  - Macro-level controls to cap spending or regulate price and quantity of services.
  - Micro-level reforms to improve the functioning of the health system and increase spending efficiency.
  - Demand-side reforms to curb health care demand by increasing cost sharing by patients.
- Budget caps (macro-level):
  - Budget caps limit overall health care spending for hospitals and practitioners and have been used in many successful reform episodes.
  - Risks: blunt instrument that can limit access and lengthen waiting times for elective surgery (evidence from Canada, Sweden, United Kingdom).
  - Current indicators: Japan’s OECD indicator for the stringency of the budget constraint is 0 (indicator ranges between 0 and 6; OECD average is 2.0).
  - Implementation challenges in Japan: public hospitals account for only about 30 percent of total hospital beds, making strict caps more challenging.
- Supply constraints (macro-level):
  - Japan currently imposes supply constraints on workforce and facilities, but they are less effective for facilities:
    - Government controls health care workforce with enrollment quotas for medical schools (quota reduced from 8,280 in 1981 to 7,625 by 2003 and frozen thereafter).
    - Japan regulates the number of hospital beds by restricting installation of new units beyond regional benchmarks, yet Japan has the highest number of beds per capita among OECD countries.
    - High availability of high-tech medical equipment per capita, since hospitals have discretion over equipment purchases.
  - OECD indicator for regulation of health workforce and equipment: Japan’s score is 1.5 (indicator ranges between 0 and 6; OECD average is 2.9).
  - Policy implication: redesigning supply constraints—targeting the number of beds per capita and rationalizing medical equipment—may help contain health spending; evidence from Canada and Finland shows bed reductions helped contain spending.
- Micro-level reforms and other options:
  - Introduce gatekeepers and reform provider payment arrangements to prevent inefficient use of resources.
  - Expand use of generic medicines.
  - Promote prevention of lifestyle related diseases to generate savings.
- Design caveats:
  - Budget caps and supply constraints must be carefully designed to avoid unintended consequences such as reduced access to care.
  - Lessons can be drawn from experiences in other advanced countries where health services are mainly provided privately and budget targets trigger containment measures when overshooting occurs.

*Italic line: Source: IMF staff estimates and referenced Japanese Ministry of Health, Labor, and Welfare and National Institute of Population and Social Security Research, as presented in the source PDF.*

### 2007. It has been gradually raised since 2008.

### _wp14142 - 2007. It has been gradually raised since 2008.

### Health system indicators and utilization (excerpt)
- JapanOECD median
- Japan's 
  ranking
- No. of 
  countries
- Physicians per 1,000 population (head counts)2.23.329th highest34
- Total hospital beds per 1,000 population13.44.21st highest34
- Curative (acute) care beds per 1,000 population8.03.01st highest34
- MRI units, total, per million population46.910.71st highest32
- CT scanners, total, per million population101.316.11st highest32
- Doctors consultations, number per capita13.16.62nd highest34
- Average length of stay at hospitals, days18.27.21st highest34

### Price controls
- Price controls are a primary lever for the Japanese government to control health spending; prices of services covered by health and long-term care insurance are determined by the government and adjusted every two years for health care and every three years for long-term care.
- The government restrained price increases during 2003–08 and succeeded in reining in health spending especially in 2006.
- Price control measures in 2006 and 2008 were followed by a pickup in spending in the subsequent year, consistent with providers responding by increasing service volumes or directing patients to higher-cost services.
- At times, price adjustments across HC services resulted in unintended consequences and exacerbated distortions in the health system (the 2013 National Council Report).
- Conclusion: Price controls should be carefully designed and can be complemented by other options.

### Provider payment arrangements (micro-level reform)
- Shifting from fee-for-service to a Diagnostic-Related Group (DRG) arrangement would:
  - Reduce supplier-induced health care demand.
  - Help contain spending growth by paying providers a fixed amount based on diagnostics rather than service volume.
- Current system features and evidence:
  - Fee-for-service used for compensating doctors; limited use of DRG-type arrangements for hospitals.
  - OECD indicator of strength of providers’ incentives to increase volume of care: Japan’s score is 5.7, OECD average is 3.1.
  - Japan has very high doctor consultations per capita and average length of hospital stay compared with other advanced economies (see Table 3 in source).
  - Positive correlation observed between average length of hospital stay and number of beds per capita at prefecture levels (Figure 13, left panel) and between average length of hospital stay and per-capita health spending of the elderly (Figure 13, right panel).
- Implication: Mitigating incentive problems through payment reform could be effective in reining in health spending growth.

### Gatekeeping and public management
- Gatekeeping (patients register with a general practitioner and/or require referral to access specialist care) can prevent unnecessary treatments and save costs.
  - 20 out of 29 OECD countries use gatekeeping arrangements at least to some extent.
  - Japan has no gatekeeping arrangement; patients have “free access” to choose hospitals and providers without referrals.
  - The 2013 National Council Report emphasizes the need for gatekeeping.
- Public management and coordination issues:
  - High share of private hospitals and providers, relatively lax supply controls, and limited effectiveness of price controls have led to suboptimal allocations.
  - Share of hospitals offering highly acute care was about 40 percent in 2010, higher than government’s ideal (under 20 percent).
  - Large number of long-term care beds remain in hospitals rather than in long-term care facilities.
  - Supply shortages of in-home care induce provision of long-term care at hospitals.
  - Supply and demand mismatches across doctor specialties and regions persist despite medium-term planning by prefecture governments started in 2006.
- Policy option: Strengthen regulatory power of prefecture governments to monitor and coordinate resource allocations (as proposed by the 2013 National Council Report and Shibuya and others (2011)).

### Other cost-saving options
- Promote use of generic medicines:
  - Japan’s utilization of generic medicines was 24 percent in 2009.
  - Comparative utilizations: United States 89 percent, Germany 75 percent, United Kingdom 71 percent (Sheppard, 2010).
  - Recommendation: Further incentivize and advertise use of generic drugs.
- Prevention of lifestyle related diseases:
  - Major risk factors (high blood pressure, smoking, high concentrations of blood glucose) can be further reduced for the Japanese (Ikeda and others, 2011).
  - Ongoing public campaign can be supplemented by greater use of monetary incentives, such as higher tobacco taxes (OECD, 2009).

### Options to raise copayments and premium contributions
- Context: Raising patient copayments and premium contributions are considered as options to finance higher health spending, with attention to economic efficiency and equity.

Raising patient copayments
- Evidence and rationale:
  - Copayments in Japan are relatively low in levels and as percent of total spending compared with other OECD countries (Figure 14).
  - Higher copayments are expected to discourage unnecessary demand and help contain total health spending.
  - Shigeoka (2013) estimates that out-of-pocket medical spending drops by about 70 percent when a patient turns 70 years old (copayment rate decreases from 30 percent to 10 percent), and this reduction drives up outpatient visits and inpatient admissions by 10 percent and 8 percent, respectively; no improvement in health outcomes such as mortality was found.
- Equity considerations:
  - Increasing copayments should be designed carefully to preserve public insurance protection for the poor.
  - Option: Differentiate copayment rates based on income levels.

Raising premium contributions
- Benefits and costs:
  - Benefit: Tighten link between contributions and income to ameliorate negative effects on income inequality.
  - Scope: Increase ceilings on individual premium contributions (proposed by the 2013 National Council Report).
  - Cost: Higher premium contributions could increase the labor tax wedge and reduce labor supply and demand.
  - Mitigation: Introduce a progressive contribution (payroll tax) schedule (rate rises as income rises) to reduce tax wedges for low-skilled workers.
- Inequality in contribution payments:
  - Current exemption: Dependent spouses of employees covered by employment-based HC insurance programs are exempted from paying health care insurance premiums if annual earnings are below 1.3 million yen (approximately US$13,000), while dependent spouses of self-employed workers (covered by the NHIP) are required to pay contributions.
  - Consequence: Exemption creates strong incentive for exempted spouses to work part-time and keep labor earnings below the threshold.
  - Average contribution rate for the NHIP is three times as high as that for employment-based Kumiai Kempo insurance programs (Ikegami and others, 2011).
  - Recommendation: Reduce inequality in contribution rates across health care insurers and consider collecting premiums from currently exempt dependent spouses.

### Estimated fiscal savings (summary and projections for 2030)
- Aggregate note: Table 4 summarizes estimates of fiscal savings from reform options; estimates are illustrative and based on various models and government estimates.
- Estimated savings in 2030 (In percentage points of GDP):
  - Options to contain health spending growth
    - Introduce budget caps0.2
    - Strengthen supply constraints0.2
    - Introduce gatekeeping0.3
    - Incentivize use of generic drugs0.3
    - Promote prevention of lifestyle related diseases0.1
  - Options to raise copayments and premium contributions
    - Increase copayment rates by 5 percent for elderly 1/0.6
    - Collect premiums from depndent spouses 2/0.3
  - Total 1.9
- Underlying assumptions and caveats:
  - Introducing budget caps, strengthening supply controls, and introducing gatekeepers could contain health spending by about 0.7 percentage points of GDP by 2030 (based on a cross-country econometric model by Clements and others (2012)).
  - Incentivizing generic drugs and promoting prevention estimated to contain about 0.4 percentage points of GDP by 2030 (based on government estimates).
  - Raising copayment rates for patients aged 70 and older and LTC users by 5 percentage points in effective terms estimated to generate fiscal savings of 0.6 percentage points of GDP by 2030 (estimate does not factor in possible dampening effects on service utilization).
  - Collecting premium contributions from dependent spouses currently exempted could add revenue of 0.3 percent of GDP by 2030.
  - Estimates should be regarded as highly illustrative due to possible estimation errors; some coefficients had unexpected signs (e.g., coefficient for reforming provider payment arrangements).

### Concluding remarks (key findings)
- Findings summarized:
  - Two-thirds of the spending increase over the last two decades resulted from population ageing; the rest from excess cost growth.
  - Ageing effect alone will raise health spending from 9½ percent of GDP in 2010 to 13 percent of GDP in 2030.
  - If excess cost growth continues at the trend rate observed during the last two decades, health spending will reach 15½ percent of GDP in 2030.
  - There is scope to introduce micro-level and macro-level reforms to potentially contain spending.
  - On financing, raising copayments and premium contributions can contain government transfers to the health system but should be designed to preserve equity.
- Research needs:
  - Large uncertainty surrounds projections and the sources of excess cost growth.
  - Key policy question: How to curtail per-capita spending by the elderly.
  - Suggested micro-level research: Examine factors contributing to disparities in per-capita spending by the elderly across regions.

### Box 1. Japan’s Health System (key features)
- Coverage:
  - Universal health care (HC) and long-term care (LTC).
  - HC system virtually covers the entire population.
  - LTC covers people age 65 and older and those between ages 40–64 who meet eligibility criteria.
- Fragmentation:
  - Japan’s HC system is highly fragmented: over 3,000 plans by residence and employment.
  - Insured enjoy “free access” to any providers regardless of residence or insurer.
- Hospitals:
  - Privately owned hospitals account for a majority of total hospital beds and operate generally on a not-for-profit principle.
- Insurers:
  - Age and employment determine HC insurer.
  - People aged 75 and above covered by insurer operated by prefecture government.
  - For those aged 74 and below: four kinds of insurance programs exist (Kyokai Kempo for small and medium enterprise employees; Kumiai Kempo for large firm employees; Mutual Aid Associations for public sector employees and teachers at private schools; National Health Insurance Program (NHIP) for self-employed, unemployed, pensioners).
  - Around 1,500 Kumiai Kenpo insurers; around 1,800 NHIPs.
  - LTC insurance administered by municipal governments.
- Contributions:
  - Contributions for HC insurance adjusted to income and differ across insurers; for employment-based programs contributions are proportional to income but become flat beyond a threshold; employers provide matching contributions.
  - Dependent spouses with annual income below 1.3 million yen (about $13,000) are exempted from paying contributions.
  - Contribution rates for NHIP enrollees based on family unit and vary across municipalities.
  - Contribution rates for LTC insurance adjusted to income and differ across municipalities.
- Patient copayments:
  - HC copayment rate uniform at 30 percent, except pre-school children (20 percent) and the elderly (10 percent).
  - The 30 percent rate is applied to participants at age 70 and older who earn income comparable to working-age population.
  - LTC copayment uniform at 10 percent.
  - Copayments are subject to monthly cap based on age and income.
  - People on welfare receive free care (such families account for about 3 percent of total households).
- Government controls:
  - Central government sets unit prices of all medical procedures, drugs, and devices every two years, applied uniformly to all physicians and hospitals.
  - Government determines prices of LTC services every three years.
  - Government regulates health care workforce through enrollment quotas for medical schools.

### Appendix I. Two sources for health spending data (notes)
- Two data sources for spending on HC and LTC in Japan:
  - Government data: spending covered by public health and long-term care insurance systems (Ministry of Health, Labor, and Welfare).
  - OECD data: spending based on the System of Health Account.
- Data coverage differences:
  - Some items included in OECD data are excluded from government data (e.g., procedures not covered by public health insurance such as expenses for normal childbirth and comprehensive medical checkups, and capital expenditure by hospitals).
  - Some items included in government data are excluded from OECD data (e.g., personal care services covered by public LTC insurance).
- Observed divergence:
  - Government data diverges from OECD data from 2000 onwards, reflecting large long-term care spending excluded from OECD data and other coverage differences.

*Source: IMF staff excerpt from _wp14142 - 2007. It has been gradually raised since 2008.*

### introduction of long-term care insurance. Second, capital expenditure, which is covered by

### _wp14142 - introduction of long-term care insurance. Second, capital expenditure, which is covered by

### Public health care and long-term care spending: decomposition findings
- Capital expenditure, which is covered by the OECD data but not by the government data, has decreased gradually as a percent of GDP.
- Government data adjusted for the impact of the long-term care spending (line B) moves closely in parallel with the OECD data adjusted for capital expenditure (line C) in the last two decades.
- The spending-decomposition results are not sensitive to the choice of the two series provided the structural break for the introduction of long-term care insurance is taken into account.
- Using the OECD data, the annual average increase of the public spending-to-GDP ratio over 1990–2010 is 2.8 percent.
- In the main text (Table 1), the annual average increase of the ratio is 3.1 percent over 1990–2011, excluding observations for 2000–2002.

### Methodology to project contribution rates for health care (HC) and long-term care (LTC) insurance programs
- Population
  - Age-specific population projections for five-year groups come from the National Institute of Population and Social Security Research (IPSS 2012).
  - Birth and death rates are based on the medium variant.
- Labor force participation (LFP)
  - Age-specific LFP rates for 5-year groups come from the Statistics Bureau of the Ministry of Internal Affairs and Communications.
  - The 2012 data is used to calculate the age-specific LFP rates during 2013–18 such that the aggregate LFP rate matches the projection by the IMF (2013) in each year.
  - For 2019 onward, age-specific LFP rates remain the same as in 2018.
  - The aggregate LFP rate is computed based on the age-specific population and LFP rates. It declines from 57 ½ percent in 2010 to 55½ percent in 2030 and 50 percent in 2060, largely due to the aging of the population.
- Employment and earnings
  - A similar methodology is applied for projections of age-specific employment rates and average earnings for 5-year groups.
  - Outturn data comes from the Statistics Bureau.
  - Projections incorporate data from the World Economic Outlook (IMF 2013), and assume that the unemployment rate will come down to 4 percent and remain at this rate from 2018 onwards.
  - Age-specific employment and the corresponding average earnings provide information on aggregate labor earnings in the economy, which also matches the projection by IMF (2013) in each year during 2012–2018.
  - The nominal wage is assumed to grow by 2¼ percent on average during 2018–22 and by 2½ percent from 2023 onward.
  - Inflation is fixed at 1 percent in the long-term.
  - The assumptions of nominal wage growth and CPI inflation follow closely the long-term projection of the conservative growth scenario (Cabinet Secretariat 2011).
- Pension benefits
  - Projections for both aggregate and age-specific pension benefits follow the baseline scenario used by Kashiwase, Nozaki, and Tokuoka (2012).

### Institutional indicators of health care systems (selected indices for Japan vs. OECD median)
- Budget caps / central government oversight: Japan 4.3; OECD median 4.7
- Supply constraints / regulation of workforce and equipment: Japan 1.5; OECD median 2.9
- Priority setting: Japan 4.1; OECD median 2.9
- Price controls / regulation of providers' prices: Japan 5.0; OECD median 4.6
- Price controls / regulation of prices paid by third-party payers: Japan 5.0; OECD median 4.5
- Public management and coordination / gatekeeping: Japan 0.0; OECD median 3.0
- Public management and coordination / subnational government involvement: Japan 2.1; OECD median 1.8
- Contracting methods / volume incentives: Japan 5.7; OECD median 3.2
- Market mechanisms / choice of insurers: Japan 2.0; OECD median 1.0
- Market mechanisms / insurer levers: Japan 1.8; OECD median 0.0
- Market mechanisms / user information: Japan 0.0; OECD median 0.9
- Market mechanisms / private provision: Japan 4.4; OECD median 2.9
- Market mechanisms / choice among providers: Japan 6.0; OECD median 5.3
- Demand-side reforms / over-the-basic coverage: Japan 0.5; OECD median 0.8
- Demand-side reforms / price signals on users: Japan 0.9; OECD median 1.0

*Source: IMF staff summary of the supplied document content.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14142.pdf_