Indicator 10.4.2 Redistributive impact of fiscal policy

Indicator 10.4.2: Redistributive impact of fiscal policy

Target 10.4: Adopt policies, especially fiscal, wage and social protection policies, and progressively achieve greater equality

The World Bank Group, Washington DC, USA; henceforth WBG.

Definitions:

The Redistributive Impact of Fiscal Policy indicator is defined as the Gini Index of pre-fiscal per capita (or equivalized) income less the Gini Index of post-fiscal per capita (or equivalized) income. These terms are elaborated below and can be calculated with some different variations.

Concepts:

-Gini Index: a commonly used measure of inequality capturing the statistical dispersion in the distribution of income over a population (Gini, 1936). A Gini Index of zero expresses perfect equality: that is, every individual in the population has the same income. A Gini Index of 100 expresses maximum inequality: that is, all income accrues to a single individual, and all other individuals have zero income.^[1]

Household income: this can be calculated: (i) in per capita terms (household income divided by the number of household members); or (ii) in equivalized terms (household income divided by the square root of the number of household members).^[2] If a different definition is used, it should be noted in the reporting document.

-Pre-fiscal income: the cumulative income accruing to an individual (or a household) from market and private sources only. The Redistributive Impact of Fiscal Policy indicator can be estimated with reference to two different pre-fiscal income concepts depending on assumptions regarding the nature of the public, contributory old-age pension system (please also see the figure below, adapted from Lustig (2018) and in Lustig chapter 1, Section 2.2, pp. 20-29):

1. Pre-fiscal income 1 - under the “pensions as deferred income” scenario: When incomes from public contributory old-age pension-system are counted as deferred market income and old-age pension-system contributions are counted as savings from current income (that is, the old-age pension system is treated as the equivalent of a mandatory savings program), pre-fiscal income is defined as an individual’s earned and unearned incomes from market and other private sources: wages, interest and dividend income; imputed income from owner-occupied housing and from consumption of own production;^[3] remittances; private transfers; old-age pension income from the public contributory pension system; and, less any contributions to the public old-age contributory pension system. In this case, the pre-fiscal income concept is called Market income plus pensions.
2. Pre-fiscal income 2 - under the “pensions as government transfer” scenario: When incomes from current pension system are counted as a government transfer and old age pension system contributions are counted as a tax on current income, pre-fiscal income is defined as: wages, interest and dividend income; imputed income from owner-occupied housing and from consumption of own production; remittances; and private transfers only. In this case, the pre-fiscal income concept is called Market income.

When pensions are treated as pure government transfers, the redistributive effect of pensions may be exaggerated as retirees with zero or near zero pre-fiscal incomes will receive pension income that is – at least in part – income deferred when the individual was working. It is important to note that deferral of own income from one’s working years to one’s retired self is possible regardless of whether the pension system is actuarially fair and in both defined-contribution and defined-benefit pension plans. Treating the public contributory pension system income as pure deferred income, however, does not allow us to capture any portion of the redistributive effect of pensions which may in effect exist. Therefore, we view the pensions as government transfer and pensions as deferred income scenarios as imperfect upper and lower bound estimates (respectively) of the true redistributive effect of contributory pensions. Rather than generating estimates of the redistributive effect of fiscal policy under specific assumptions about public contributory pension system income, the OECD instead reports estimates of the redistributive effect for the population under 65 years of age (while treating contributions to the public contributory pension system as a tax). This is most comparable to the “pensions as deferred income” scenario, although not exactly the same.

-Post-fiscal income: The Redistributive Impact of Fiscal Policy indicator can be estimated with reference to two different post-fiscal income concepts, Disposable Income and Consumable Income. The most comprehensive concept is that of Consumable Income, which incorporates not only the impact of direct taxes and transfers but also of indirect taxes and price subsidies.

Disposable and Consumable Income are equal in value under the “pensions as deferred income” and “pensions as government transfer” scenarios. However, they are derived from pre-fiscal income 1 and pre-fiscal income 2 differently; please see the figure below, adapted from Lustig (2018):

1. Post-fiscal incomes under the “pensions as deferred income” scenario:

Post-fiscal Income A - Disposable Income: pre-fiscal income less direct taxes paid and less social insurance contributions made to the public fiscal authority plus direct cash transfers and the monetary value of benefits (measured at what governments spend) received by households in the form of near-cash transfers (e.g., food stamps, school breakfasts, school uniforms).

Post-fiscal Income B - Consumable Income: pre-fiscal income less direct and indirect taxes paid and less social insurance contributions other than for old-age pensions made to the public fiscal authority plus direct cash transfers and the monetary value of benefits (measured at what governments spend) received by households in the form of near-cash transfers (e.g., food stamps, school breakfasts, school and indirect price subsidies.

1. Post-fiscal incomes under the “pensions as government transfer” scenario:

Post-fiscal Income A - Disposable Income: pre-fiscal income less direct taxes paid and less social insurance contributions and less contributory old-age pension contributions made to the public fiscal authority plus direct cash transfers and the monetary value of benefits (measured at what governments spend) received by households in the form of near-cash transfers (e.g., food stamps, school breakfasts, school uniforms).

Post-fiscal Income B - Consumable Income: pre-fiscal income less direct and indirect taxes paid and less social insurance contributions and less contributory old-age pension contributions made to the public fiscal authority plus direct cash transfers and the monetary value of benefits (measured at what governments spend) received by households in the form of near-cash transfers (e.g., food stamps, school breakfasts, school uniforms), and plus indirect price subsidies.

CEQ Income Concepts

Source: adapted from Lustig (2018).

- Gini Index points: The Redistributive Impact of Fiscal Policy indicator is the difference between pre-fiscal Gini Index and the post-fiscal Gini Index. Thus, if a simple difference is applied the measure is the change in Gini Index points.

The Redistributive Impact of Fiscal Policy indicator is constructed from a range of data sources using a standardized methodology as outlined in Lustig, 2018. To construct this indicator requires a nationally representative micro-data set (a Household Budget Survey, for example, or an Income and Expenditure Survey) and fiscal or budgetary or administrative data on revenue collections, social expenditures, and expenditures on consumption subsidies. The data sources employed at the country level are detailed in the country-specific footnotes.

Ultimately the data providers are national-level statistical agencies for the micro-data sets and national-level fiscal agencies and bodies for budgetary and administrative data. Most OECD countries also calculate their own pre- and post-fiscal Ginis. That is, they directly calculate the 10.4.2 indicator. These are collated by the OECD and will be sent directly to the World Bank as custodians.

Where a country produces its own 10.4.2 indicator it will take precedence over estimates produced by other institutions, subject to meeting the reporting requirements below. For all other countries, estimates and indicators produced by the WBG and/or the Commitment to Equity Institute will be considered.

Reporting on assumptions: The choice of whether to report the Redistributive Impact of Fiscal Policy indicator under the pensions as deferred income or pensions as transfers scenario will be left to the country authority or international agency in charge of submitting this indicator, but the choice must be clearly indicated in the reporting document. For countries for which the data exist, pre-fiscal and post-fiscal inequality should be calculated for both pension scenarios, and the default included in the SDGs database is pension as deferred income. If only data treating pensions as transfers are available, it is recommended to report them only for the working age population (under 65 years of age). Some authorities may also choose to use equivalized income instead of per capita income as the welfare indicator. This too should be clearly indicated in the reporting document. Last, some authorities may report these data based on a micro-data set using income or expenditure as the relevant welfare concept. Once these decisions are taken, they should be maintained in subsequent years in order to assure comparability, except that all countries are encouraged to provide data with pension as deferred income. The data reported in the UN Global Database try, to the extent possible, to distinguish between the different concepts used for different countries.

Feasibility: The Redistributive Impact of Fiscal Policy indicator can be estimated for any country with a micro-data set detailing incomes or expenditures (or both) at the household or individual level and with a set of fiscal, administrative, or budgetary records detailing public expenditures at the program level and revenue collections at the revenue-collection instrument level.

Suitability/Relevance: The Redistributive Impact of Fiscal Policy indicator provides a direct estimate of the current impact of fiscal policy on redistribution (of incomes). It therefore provides a direct estimate of progress on SDG Target 10.4: “Adopt policies, especially fiscal, wage and social protection policies, and progressively achieve greater equality.”

Limitations: The Redistributive Impact of Fiscal Policy indicator does not address wage policy. It does not include the benefits of public provision of in-kind benefits, such as health, education, sanitation and housing services, which may have both present-day and longer-term impacts on present-day and future inequality.

Pre-fiscal income can be derived from a nationally-representative micro-data set (an Income and Expenditure Survey, for example). Post-fiscal income is estimated via the allocation of the tax burdens and the expenditure-based benefits that stem from fiscal policy (direct and indirect taxes, social contributions, direct cash and near-cash transfers, subsidies, et cetera). Procedures for constructing pre-fiscal and post-fiscal income concepts and estimating their distribution from an underlying microdata set are detailed comprehensively in Lustig (2018) (Chapters 1, 6, and 7).

The Gini Index is calculated rescaling the Gini Coefficient by a factor of 100. The Gini Coefficient is calculated according to standard formulas for a (generalized) Gini Coefficient. See, for example, Duclos and Araar (2006):

$\mathrm{G}\mathrm{I}\mathrm{N}\mathrm{I}\mathrm{\ }\mathrm{I}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{x}=100\mathrm{\ }\mathrm{x}\mathrm{\ }\mathrm{G}\mathrm{I}\mathrm{N}\mathrm{I}\mathrm{\ }\left(\mathrm{{\rm X}};\mathrm{\upsilon }\right)$

where X is a random variable of interest with mean μ(X), F(X) is its cumulative distribution function, υ is a parameter tuning the degree of ‘aversion to inequality’. The standard Gini corresponds to υ = 2. Cov is a Covariance estimate.

2021-02-05

Institutional information: The World Bank Group is the official custodian for this indicator. This metadata documentation was developed and agreed by the three institutional data providers, CEQ Institute, OECD and The World Bank.

The Impact of Fiscal Policy on Poverty (see Lustig, 2018, chapter 6).