Does the Gini index show that South Africa is the most unequal society in the world?

13 June 2019

INTRODUCTION

You would certainly have come across articles or presentations where it is stated that South Africa is the most economically unequal society in the world. Invariably, reliance for such statements is placed on the so-called Gini coefficient or index, produced by none other than the World Bank. 

Ever tried to dig a little deeper?

WHAT IS THE GINI INDEX?

The Gini index attempts to measure income inequality. Briefly put, it can range from 0 to 100, where 0 reflects an equal distribution of income (where everyone has the same income) and 100 reflects absolute inequality (where one person has all the income and no-one else has anything).[1]

The World Bank Gini table shows a range from a low of about 25 (reflected in some European countries) to a high of that accorded to South Africa, which has a score of 63.

WHAT ARE THE LIMITATIONS OF THE WORLD BANK’S GINI INDEX?

There are surprisingly many. Here are the most basic ones:

- of the 218 countries and territories listed by the World Bank only 149 have reported Gini coefficients in the ten years from 2009 to 2018. Many important countries do not appear on the list, presumably because of the unavailability of data. The excluded countries include the following: Saudi Arabia, the Gulf States (Qatar, Bahrain, Kuwait etc.), Cuba, Cambodia, Afghanistan, Libya and Equatorial Guinea (an oil-rich state, where one family and some hangers-on own all the state’s assets);

- the index is relative, not absolute. If everyone’s income were to double, the Gini coefficient would remain constant. On the other hand, poverty is measured in an absolute manner by reference to absolute poverty lines, so that if everyone’s income were to double, poverty would decrease;

- the World Bank’s notes on its methodology provide two important warnings:

1. Household survey questionnaires can differ widely, and similar surveys may not be strictly comparable because of differences in quality. These problems are diminishing as survey methods improve and become more standardized, but achieving strict comparability is still impossible. Under-reporting of income and selective compliances are other sources of measurement errors, and these problems are unlikely to be distribution-neutral.

2. The measurement system was developed for the sole purpose of public replication of the World Bank’s poverty measures for its widely used international poverty lines, including $1.90 a day and $3.20 a day in 2011 Purchasing Power Parity. The methods are considered reliable for that purpose. However, one cannot be confident that the methods work well for other purposes, including tracing out the entire distribution of income. We would especially warn that estimates of the densities near the bottom and top tails of the distribution could be quite unreliable, and no attempt has been made by the Bank’s staff to validate the tool for such purposes.

- household income surveys can be designed to measure household income in one of four ways:

Moreover, income reported by households may not always correspond with the survey design.

- because the income concept can be ambiguous, poverty and inequality measures are sometimes based on expenditure instead. This was the case in Statistics South Africa’s widely cited Poverty Trends in South Africa report which covered the period between 2006 and 2015.

- three types of income are likely to be captured poorly or not at all: income from the informal sector, income in kind and the ‘social wage’. In South Africa, this includes free primary health care; no-fee paying schools; social protection (most notably old-age grants and child support grants); RDP housing; and the provision of free basic services (namely water, electricity and sanitation) to poor households.

- inequality of income is measured by the index, not inequality of wealth. Inclusion of wealth in the calculation would probably increase the level of inequality in many countries.

CONCLUSION

In spite of the limitations set out above, it is surprising that ostensibly knowledgeable commentators and presenters make a habit of glibly stating that the World Bank’s Gini index shows South Africa to be the most unequal society in the world.[2] Even the World Bank itself is guilty of this.[3] No-one will deny that South Africa is a very unequal society, but relying on comparative calculations which have such obvious glaring gaps, does these commentators no favours. 

Instead, one cannot avoid the impression that none of them have bothered to check what they are relying on and that we are dealing with the well-known phenomenon that something is regarded as true if it is repeated often enough by a sufficient number of important sounding people or organisations. 

A little more care would be welcome. And if, for example, the various statements included the simple qualifications that the Gini index neither includes all the world’s countries, nor is completely comparable across all countries, nor constitutes all information about inequality, no-one would be able to quibble.

Anton van Dalsen, Legal Counsellor, and Charles Simkins, Head of Research, Helen Suzman Foundation. 

[1] For the mathematically gifted, see the technical explanation in Inequality Measurement, Development Issues no. 2, October 2015, published by the Development Strategy and Policy Analysis Unit in the UN’s Department of Economic and Social Affairs: https://www.un.org/development/desa/dpad/wp-content/uploads/sites/45/publication/dsp_policy_02.pdf

[2] Surprisingly little has been published on the limitations of the index. However, see Income inequality and limitations of the Gini index: the case of South Africa, HSRC Review, November 2014.

[3]Overcoming Poverty and Inequality in South Africa, 2018, published by the World Bank, states on page xv that “… the country (ie. South Africa) had a Gini coefficient of 0.63 in 2015, the highest in the world…”.