Post-tax-and-transfer Gini Coefficient Calculation
Ministry of Trade and IndustrySpeakers
Summary
This question concerns the methodology for calculating the post-tax-and-transfer Gini coefficient, specifically whether corrections are applied uniquely to households or based on decile averages. Assoc Prof Jamus Jerome Lim asked if the data relies on actual income tax and Central Provident Fund statements or assumptions about average taxes and transfers. Minister Chan Chun Sing replied that the coefficient uses individual-level data from agencies such as IRAS and CPFB, supplemented by household-level estimations where necessary. For taxes like GST, estimations are derived from the Household Expenditure Survey using specific variables like dwelling type, income, and household size. These methods ensure that Government transfers and taxes are accounted for based on actual or characteristic-matched household data rather than general averages.
Transcript
24 Assoc Prof Jamus Jerome Lim asked the Minister for Trade and Industry (a) in the calculation of the post-tax-and-transfer Gini coefficient, whether the tax/transfer correction is a general formula or is applied uniquely to every household; and (b) whether the data used to calculate post-tax/transfer household income is based on actual income tax and CPF statements, or based on assumptions about the average taxes/transfers that will accrue to households in each decile of the income distribution.
Mr Chan Chun Sing: The Gini coefficient for each year is computed on the basis of a sample of households surveyed in the Comprehensive Labour Force Survey conducted by the Ministry of Manpower in that year, or the Censuses of Population and General Household Surveys conducted by the Department of Statistics in the years ending with '0' and '5' respectively.
The Gini coefficient that accounts for Government transfers and taxes is calculated using data on individual household members' income, transfers and taxes where possible. The individual-level data are based on data collected by agencies such as IRAS and CPFB where available.
For transfers and taxes where data at the individual-level are not available, they are estimated at the household-level. For instance, the Goods and Services Tax for each of the sampled households is estimated based on data from the Household Expenditure Survey, taking into account the characteristics of the household such as dwelling type, household income from work, age of household members and household size.