Appended Note 3-3 Overview of Generational Accounting
- Japanese version
- English version
We adopted the same analytical method for generation accounting used in "Annual Report on Japan's Economy and Public Finance 2000-2001" (Figure 3-3-3, Figure 3-3-5, Appended Note 3-6) and latest incorporated data.
1. Classification of generation
For distributing government outlays and revenues to income and spending of individuals in each generation, we use generation-by-generation income and spending data in the "Family Income and Expenditure Survey" and the "National Consumption Situation Survey" by the Ministry of Public Management, Home Affairs, Posts and Telecommunications. "Individuals" in each generation here mean "households" in each generation.
In consideration of constraints on past data in the "Population Census" by the Ministry of Public Management, Home Affairs, Posts and Telecommunications, we have divided present people into five generations-the 20's, 30's, 40's, 50's groups, and those in their 60's and older. The life expectancy for householders is set at 80 years. Every householder is presumed to end his or her life at the age of 80.
2. Estimation of benefits and contributions for present generations
(1) We have calculated generation-by-generation benefits and burdens for each year after 1955 for which comparable data are available in the "National Accounts (SNA)."
(2) Government receipts and payments in the "National Accounts (SNA) Institutional Sector-Wise Income and Spending Data (General Government)" have been distributed to the generations in accordance with the breakdown of wage-earning households, and generation-wise householder income and expenditure in the "Family Income and Expenditure Survey" and the "National Consumption Situation Survey." We have then estimated benefits and burdens per household for each generation, based on the number of households in the "Population Census."
(3) As for the "National Accounts (SNA) Institutional Sector-Wise Income and Spending Data (General Government)," we have used 93SNA-based real data for years from 1990. For years before 1990, we modified the 68SNA-based data into 93SNA-based data (using the ratio of the 93SNA to 68SNA in the data-overlapping period for each receipt and payment).
(4) Benefits from public fixed capital have been distributed evenly to all households. We have used fixed capital depreciation in the "National Accounts (93SNA)" for calculating macroeconomic benefits for years from 1990. For years before 1990, we have used net government fixed assets, etc. in the "National Accounts (68SNA)."
(5) Fixing Government Receipts and Payments
1） Government receipts
Taxes on production and imports, taxes on income and assets, social security contributions, other contributions
2） Government payments
Actual final consumption + goods and services transfers, subsidies, social security benefits, other benefits, savings
3） Actual final consumption + goods and services transfers have been divided into "education expenditure" and "others." We have estimated "education expenditure" based on education expenditure breakdowns in actual social transfers and actual final consumption of purpose-wise general government outlays (Appended Table 7) in the "National Accounts (SNA)." Even with the 68SNA, data have been available since 1970. Therefore, we have used 1970 data for years before 1970.
4） Social security benefits have been divided into "pensions," "elderly health care benefits," "nursing care insurance benefits" and "other benefits." "Pensions," "elderly health care benefits" and "nursing care insurance benefits" have been based on a table (Appended Table 9) of specifics of transfers from the general government to households in the "National Accounts (SNA)."
"Pensions" have been available since 1970 even with the 68SNA of the "National Accounts." For years before 1970, therefore, we have used data for each pension scheme in the "Annual Report on Social Security Statistics" to calculate pensions.
5） Social security contributions have been divided into "pension premiums," "elderly health care contributions," "nursing care insurance premium" and "others." We have based the calculation of "pension premiums" and "nursing care insurance premium" on the table of social security contribution specifics (Appended Table 10) in the "National Accounts (SNA)."
(6) Proportional distribution to generations
1） Taxes imposed on production and imports have been distributed to households in proportion to consumption in the "Family Income and Expenditure Survey."
2） Of taxes on income and assets, those on households have been distributed to households in proportion to direct tax in the "Family Income and Expenditure Survey."
3） Of taxes on income and assets, those on corporations are deemed to be passed on to individuals through wages, dividends, product prices, etc. and have been distributed to households in the following way:
- Half passed on to factor income of the supply side (A)
- The other half passed on to product prices of the demand side (B)
In accordance with the latest macroeconomic labour share, we have treated 75% of (A) as passed on to wages and distributed that portion to households in proportion to household wage income and the wage income as in the "Family Income and Expenditure Survey." We have treated the remaining 25% as passed on to capital income and distributed that portion to households in proportion to percentages of household-owned financial assets and outstanding savings in the "National Consumption Situation Survey."
Since outstanding savings in the "National Consumption Situation Survey" have been available only since 1974, we have used the 1974 data for years before 1974.
The taxes passed on to product prices (A) have been distributed to households in proportion to household consumption shares and consumption in the "Family Income and Expenditure Survey."
4） We have distributed social security contributions to households in proportion to social security premiums in the "Family Income and Expenditure Survey." Of them, pension contributions have been distributed to generations in proportion to household pensions and pension premiums in the "National Consumption Situation Survey."
Since pension premiums in the "National Consumption Situation Survey" have been available only since 1974, we have used 1974 data for years before 1974.
Of them, elderly health care benefits have been considered to be going to householders aged 60 or more.
5） Subsidies (subsidies and social assistance benefits in the "National Accounts (SNA)") have been distributed to households in proportion to weighted averages based on the number of households for each generation in the "Population Census."
6） Education spending in actual final consumption + goods and service transfers have been distributed to generations in proportion to household education spending and education spending in the "Family Income and Expenditure Survey."
7） Other benefits, other burdens and savings have been distributed evenly to households.
(7) Proportional distribution of benefits from public investment
For years from 1990, fixed capital depreciation in the "National Accounts (93SNA)" has been distributed in accordance with the number of households in each generation. For years before 1990, net fixed assets in the 68SNA-based year-end balance sheet account (general government) have been multiplied by a certain rate of return (the ratio of fixed asset depreciation in the "National Accounts (93SNA)" to net fixed assets in the "National Accounts (68SNA)" for 1990-98) for distribution according to the number of households in each generation.
(8) Modification of wage-earning households' tax burdens
Elderly people's share of wage-earning householders is very low. If only wage-earning household data alone are taken as representing the national average, therefore, tax contributions may be overestimated. In order to avoid this bias, we have multiplied tax burdens (direct tax) of those aged 60 or above by wage-earning households' share of total households to calculate tax burdens (direct tax) in the "Family Income and Expenditure Survey."
(9) Modification of generation-wise breakdown of households
A breakdown of households extracted from the "Family Income and Expenditure Survey" and the "National Consumption Situation Survey" tends to indicate the number of elderly households as less than in the "Population Census." In order to ease such bias, we have adopted not a sample breakdown in the "National Consumption Situation Survey" or the "Family Income and Expenditure Survey," but a generation-wise breakdown of two-or-more-member households in the "Population Census" to distribute total households for our generation-wise breakdown of households.
(10) Realizing generation-wise benefits and burdens in each year
We have used the GDP deflator (100 for the base year of 2001) in the "National Accounts (SNA)" to realize generation-wise benefits and burdens in each year.
(11) Present value of past benefits and burdens for present generations
We have inflated benefits and burdens in and after 1955 by each year' s real interest rate (one-year time deposit rate--CPI deflator growth rate) to determine values in current (2001) prices.
3. Estimation of future benefits and burdens for each generation
(1) We have assumed that the generation-wise benefit and burden structure for present generations in the latest year (2001) will remain unchanged in the future.
(2) We have assumed that benefits per household from public fixed capital will remain constant in and after 2030.
(3) As for public pensions, the government's share of basic pension costs is presumed to rise from the current one-third to 50% in and after 2004. Regarding the current plan to gradually lift the pension-starting age, we have assumed that the age will be lifted by one every three years for the fixed portion of pensions from 2001 to 2013 and for the other portion proportionate to income from 2013 to 2025.
(4) As for the nursing care insurance, we have assumed that nursing care insurance benefits will all go to the 60-and-older generation. Incidentally, the increase in the nursing care insurance premium in fiscal 2003 has been taken into consideration.
(5) For the years up to 2010, per-household economic growth and expected interest rate are presumed to move as projected in the "Structural Reform and Medium-Term Economic and Fiscal Perspective" and for the year after 2010, per-household economic growth is presumed at 2% and the expected interest rate at 4%. These rates have been used to adjust future benefits and burdens, as estimated in (1) 
(4), to the present value for the latest year (2001).