2.6 Distribution of income or consumption See Table 2.6 here

About the data
Definitions
Data sources

Back to Contents

Inequality in the distribution of income is reflected in the percentage share of income or consumption accruing to segments of the population ranked by income or consumption levels. The segments ranked lowest by personal or family income typically receive the smallest share of total income. The Gini index provides a convenient summary measure of the degree of inequality.

Data on personal or household income or consumption come from nationally representative household surveys. The data sets refer to different years between 1985 and 1994. Footnotes to the survey year indicate whether the rankings are based on per capita income or consumption, or, in the case of high-income economies, household income. Where the original data from the household survey were available, they have been used to directly calculate the income (or consumption) shares by quintile. Otherwise, shares have been estimated from the best available grouped data.

The distribution indicators for low- and middle-income economies have been adjusted for household size, providing a more consistent measure of per capita income or consumption. No adjustment has been made for spatial differences in cost of living within countries, because the data needed for such calculations are generally unavailable. For further details on the estimation method for low- and middle-income economies, see Ravallion and Chen (1996).

Because the underlying household surveys differ in method and in the type of data collected, the distribution indicators are not strictly comparable across countries. These problems are diminishing as survey methods improve and become more standardized, but achieving strict comparability is still impossible (see the notes to table 2.5).

The following sources of noncomparability should be noted. First, the surveys differ in whether they use income or consumption expenditure as the living standard indicator. For 37 of the 66 low- and middle-income economies the data refer to consumption expenditure. Income is typically more unequally distributed than consumption. In addition, the definitions of income used in surveys are usually very different from the economic definition of income (the maximum level of consumption consistent with keeping productive capacity unchanged). For these reasons, consumption is usually a much better measure. Second, the surveys differ in whether they use the household or the individual as their unit of observation. Further, household units differ in size (number of members) and in extent of income sharing among members. Individuals differ in age and consumption needs. Where households are used as the observation unit, the deciles or quintiles refer to the percentage of households rather than of population. Third, the surveys differ according to whether they rank the units of observation by household or per capita income (or consumption).

World Bank staff have made an effort to assure that the data for low- and middle-income economies are as comparable as possible. Whenever possible, consumption has been used rather than income. Households have been ranked by consumption or income per capita in forming the percentiles, and the percentiles are of population, not households. The comparability of the data for high-income economies is more limited, because the observation unit is usually a household unadjusted for size and households are ranked according to total household income rather than income per household member. These data are presented pending the publication of improved data from the Luxembourg Income Study, which ranks households by the average disposable income per adult equivalent. The estimates in the table should therefore be treated with considerable caution.

Table 2.6a Income shares of lowest and highest quintiles, 1960s-1990s
percent

Source: Deininger and Squire 1996.

Back to top
Back to Contents

• Survey year is the year in which the underlying data were collected.

• Gini index measures the extent to which the distribution of income (or, in some cases, consumption expenditures) among individuals or households within an economy deviates from a perfectly equal distribution. A Lorenz curve plots the cumulative percentages of total income received against the cumulative number of recipients, starting with the poorest individual or household. The Gini index measures the area between the Lorenz curve and a hypothetical line of absolute equality, expressed as a percentage of the maximum area under the line. Thus a Gini index of zero represents perfect equality while an index of 100 percent implies perfect inequality.

• Percentage share of income or consumption is the share that accrues to subgroups of population indicated by deciles or quintiles. Percentage shares by quintiles may not add up to 100 because of rounding.

Data on distribution for low- and middle-income economies are compiled by the Poverty and Human Resources Division of the World Bank's Policy Research Department, using primary household survey data obtained from government statistical agencies and World Bank country departments. Data for high-income economies are from national sources, supplemented by:

Luxembourg Income Study database, 1990. Eurostat, Statistical Yearbook. United Nations, National Accounts Statistics: Compendium of Income Distribution Statistics (1985).

Back to top
Back to Contents