CHAPTER 6 Inequality Measures

Summary

Inequality is a broader concept that poverty in that it is defined over the entire
population, and does not only focus on the poor.

The simplest measurement of inequality sorts population from poorest to richest and
shows the percentage of expenditure (or income) attributable to each fifth (quintile) of
the population the poorest quintile typically accounts for 6 10 % of all expenditure, the
top quintile for 35-50%.

A popular measure of inequality is the Gini coefficient, which ranges from 0 (perfect
equality) to (perfect inequality), but is is typically in the range of 0.3 - 05 for per capita
expenditures. The Gini is derived from Lorenz curve which sorts the population from
poorest to richest, and shows the cumulative proportion on the population on the
horizontal axis and the cumulative proportion of the properties mean independence,
population size independence, symmetry, ans Pigou Dalton transfer sensitivity it cannon
easily be decomposed to show the sources of inequality.
The best know entropy measures are Theil`s and Theil`s L, both of which allow one to
decompose inequality into the part that is due to inequality within areas (e.g urban rural)
and the part that is due to differences between areas (e.g urban rural income gap).
Typically at least three quarters of inequality in a country is due to within group
enequality, and the remaining quarter to between group differences.

Atkinson´s class of inequality measures is quite general and is sometimes used. The
decide dispersion ratio defined as the expenditure ( or income) of the richest decile
divided by that of the poorest decile is a popular but a very crude measure of inequality.

It is often helpful to decompose inequality by occupational group or by source of

6.1 definition of inequality

the main focus of this manual is on poverty which looks at the situation of individuals
of hause holds who find themselves at the bottom of the income distribution; typically
this requires information both about the mean level of of (say) expenditure per capita as
well as its distribution at the lower end. But sometimes we are more interested in
measuring inequality than poverty per se, and for that reason we have included this
relatively brief chapter inequality.

Inequality is a broader than poverty in that it is defined over the entire population, and
not just for the population below a certain poverty line. Most inequality measures do not
depend on the mean of the distribution, and this property of mean independence is
considered to be desirable property of an inequality measure. Of course, inequality are
often calculated for distributions other than expenditure – for instance, for income, land,
assets, tax payments, and many other continuous and cardinal variables.
The simplest way to measure inequality is by diving the population into fifths (quintiles)
from poorest to richest, and reporting the levels or proportions of income (or
expenditure) that accrue to each level.
Table 6.1 shows the level of expenditure per capita, in ‘000 dong per year, for Vietnam
in 1993, based in data from Vietnam living standards survey. A fifth of the individuals (
not household) included in the survey were allocated to each expenditure quintile. The
figures show that 8.4% of the all expenditures were made by the poorest fifth of
household and 41.4% by top fifth. Quintile information is easy to understand, although
sometimes one wants a summary measure rather than a whole table of figures.
Table 6.1 breakdown of expenditure per capita by quintile Vietnam 1993

6.2.1 Gini coefficient of inequality:

The most widely used single measure of inequality is the gini coefficient. It is based on
the Lorenz curve, a cumulative frequency curve that compares the distribution of a
specifie variable (e.g income) with the uniform distribution that represents equality. To
construct the Gini coefficient, graph the cumulative percentage of household ( from
poor to rich) on the horizontal axis and the cumulative percentage of expenditure (or
income) on the vertical axis. The Lorenz curve shown in figure 1 is based on the
Vietnamese data in table 6.1. The diagonal line represents perfect equality perfect
equality. The gini coefficient is defined as A/(A+B), where A and B are areas shown on
the graph is A= 0 the Gini coefficient becomes 0 which means perfect equality whereas
if B=0 the gini coefficient becomes 2 which means complete inequality, In this
example the Gini coefficient becomes 1 which means complete inequality. In this the
Gini coefficient is about 0.35.

Lorenz Curve.

6.1--------

6.2 For users of Stata, there is a gini command that may be downloaded and used
directly (see appendix 3) this command also has the advantage that it lows one to use
weights, which are not incorporated into two equations shown above.

The gini coefficient is not entirely satisfactory. To see consider the criteria that make a
good measure of income inequality, namely:

Mean independence: This means that if incomes were doubled the measure would not
change. The gini satisfies, namely.
Population size independence: If the population were to change the measure of
inequality should no change ceteris paribus. The Gini satisfies this too.
Symmetry; If you and I swap incomes, there should be no change in the measure of
inequality the gini satisfies this.
Piguo Dalton Transfer sensitivity: Under this criterion, the transfer of income from rich
to poor reduces measured inequality. The gini satisfies this too.

It is also desirable to have

****Statistical testability: One should be able to test for significance of changes in the
index over time. This is less than it used to be confidence intervals con typically be
generated using bootstrap techniques.

6.2.2 Generalized entropy measures:

There are a number of measures of inequality that satisfy all six criteria, among the most
widely used are the Theil indexes and the mean log deviation measure. Both belong to
the family of generalized entropy inequality measures. The general formula is given by:

6.3

Where y is the mean income ( or expenditure per capita ) the values of GE measures
vary between o and OO, with zero representing an equal distribution and higher level of
inequality. The parameter a in the GE class represents the weight given to distances
between incomes at different parts of the income distribution, and can take any real
value. For lower values of a, GE is more sensitive to changes in the lower tail of the
distribution, and for higher values GE is more sensitive to changes that affect the upper
tail. The commonest values of a used are 0,1 and 2 GE (1) is theil´s T index, which may
be written as

6.4. GE(o) also know as theil`s L, and sometimes referred to as the mean log deviation,
is given by.

6.5 Once again, users of Stata do not need to program the computation of such measures
from scratch; the GE command, explained 3, allows one to get these measures when
weights need to be used the data.

6.2.3 Atkinson`s inequality measures

Atkinson has proposed another class of inequality measures that are used from time to
time. This class also has a weighting parameter e (which measures aversion to
inequality ) and some of its theoretical properties are similar to those of the extended
Gini index. The Atkinson class, which may be computed in Stata using the Atkinson
command is defined as:

6.6
Table 6.2 sets out in some detail the computations involved in the computation of the
Generalized entropy and Atkinson measures of inequality. The first row of numbers
gives the incomes of the ten individuals who live in a country, in regions 1 and 2. the
mean income is 33. To compute Theil`s T, one first computes y/ybar, where ybar is a
mean income level; then compute in (y/ybar), take the product, add up the row, divide
by the number of people. Similar procedures yield other generalized entropy measures,
and also the Atkinson measures.