/ 24 May 2013

Gini cannot grant wishes

Gini Cannot Grant Wishes

Rian Malan's letter, which suggests some form of disinformation conspiracy in the measurement of inequality in South Africa, is silly and possibly irresponsible ("Gini's really out the bottle", May 17). To my knowledge, and in my work as an economist for the Institute for Democracy in Africa, a Gini coefficient for South Africa of between 0.6 and 0.65 has always been the rule of thumb. In other words, grants have always been included as income for poorer households.

The Mail & Guardian's article the previous week, then, was also a bit of a storm in a teacup: I don't know any serious commentators on the issue who have suggested a Gini of 0.8 for South Africa.

Also, no serious commentator has claimed that we are the most unequal country in the world: we are, however, and remain, one of the most unequal countries that credibly estimates its own inequality. Various despotic oil regimes and so on don't bother.

Incidentally, it is worth noting that the National Planning Commission, in its developmental vision for the country, sets ambitious goals for poverty reduction, but only a modest decrease of the Gini, suggesting that in government, at any rate, current inequality levels are regarded as fairly intractable. It remains to be seen whether our other goals are achievable in the face of the social and political fractures that accompany inequality. – Len Verwey, Cape Town

I initially decided to ignore Lynley Donnelly's strange (in timing and content) article on South Africa's Gini coefficient ("Welfare could be the Gini in the bottle", Business, May 10), but with Malan's triumphalist letter of last week, and the approving circulation of that on Twitter by the Guardian's South Africa correspondent, David Smith, I felt obliged to respond.

The fact that taking into account various social transfers significantly reduces the Gini in South Africa is not a new observation: the South Africa Yearbook 2003/4, produced by the Government Communication and Information System, asserted: "In 2000, [the Gini] was 0.57 for income alone, but became 0.35 when social spending was included."

Accounting for redistributive policies through a "social wage" will, by definition, reduce the Gini. While a "social-wage Gini" is an interesting way to think about redistribution, a few points need to be made. First, the Gini is traditionally used to calculate income inequality prior to transfers, as a kind of measure of inherent economic inequality. It has limitations, but a social-wage version probably has many more. Second, I have not seen any rigorous research that claims South Africa's international inequality ranking changes to any significant degree when such adjustments are made.

Therefore, Smith was incorrect to refer to the assertion that South Africa is the most unequal society in the world as a "media shibboleth" – neither Donnelly nor Malan have referenced any research that substantiates that conclusion.

Finally, a lowered Gini accounting for redistributive spending does not provide a basis for the conclusion insinuated by Donnelly's article, and more bluntly drawn by Malan, that South African inequality is really not so bad. This contradicts the consensus in the academic literature, not to mention the daily experiences of many South Africans.

I would be thrilled to discover that we do not suffer from stratospheric inequality, but the evidence suggests otherwise. Your readers deserve better than wishful thinking and shoddy reporting (or, indeed, reporting of shoddy analysis). – Sean Muller, School of Economics, University of Cape Town