This article was originally published on Asian Development Blog and is republished with permission.
Assessing the extent of inequality in a particular country, as well as in regional comparisons, is often based on Gini coefficients. They measure the income distribution of the country’s residents, with 0 representing perfect equality and 1 perfect inequality. While inequality—according to Gini indicators—has been declining in almost all Central Asian and South Caucasus countries, there is a widespread social perception of the widening gap between the haves and the have-nots, leading to growing social discontent.
What does this tell us about how we measure inequality?
Gini coefficients have serious shortcomings and need to be considered with a healthy degree of skepticism. Several deficiencies are particularly relevant to most of the former Soviet Union countries, including those in Central Asia and South Caucasus.
First, inequality is often measured by a Gini coefficient for income, without adequate consideration of wealth. So, a country with a low income Gini coefficient can still suffer extreme wealth inequality. This is especially true of transition economies, where most wealthy social groups often operate outside the formal taxation regime, and significant wealth is under-reported or not reported at all.
In addition, many better-off households refuse to participate in household surveys—the basis for computation of Gini coefficients—or significantly under-report in these polls. This leads to strong biases in Gini indexes toward less inequality.
Other Gini shortcomings include a downward bias for small populations, and for less diverse economies. In both cases they tend to report smaller Gini coefficients, and hence underestimate the true extent of inequality. Likewise, the presence of a large informal economy and inability to value benefits and income from it may significantly affect the accuracy of the Gini coefficients, depending on the sectors and activities in which the informality is concentrated.
Gini coefficients can understate inequality
All these factors indicate that official Gini coefficients often understate the degree of inequality in places such as Central Asia and South Caucasus. Rising inequality in both regions, persistent unemployment and poverty levels, and the growing social tension are somewhat mitigated by migration of low-skilled and skilled workers. Examples of this are Armenia, Kyrgyz Republic, and Tajikistan.
Other criticisms of the Gini index are that economies with similar incomes and Gini coefficients can have very different income distributions, and that Gini coefficients focus on relative income distributions rather than real levels of poverty and prosperity.
A Gini coefficient might decline while the poor get poorer, and rise while everyone is getting richer.
You might also be surprised to hear that Guinea and Canada share the same Gini coefficient, 0.34, suggesting similar levels of inequality. But their real levels of poverty are very different; Canada’s per capita gross national income (GNI) is 38 times that of Guinea (in 2011 PPP$). Bangladesh and Japan provide another example of countries with the same Gini coefficient, 0.32, even though Japan’s GNI per capita is almost 12 times higher than that in Bangladesh.
These factors suggest that Gini indicators should be taken with a grain of salt when assessing inequality in these regions. To form a more realistic picture, we should look, in addition to Gini, at other indicators that can reveal more about assets and wealth distribution in a country, and non-income inequality, such as access to quality education, health, and social security and protection measures.