Why the Gini coefficient remains crucial for understanding inequality

IOL News

3 days ago

The Gini coefficient is not the full story of inequality in South Africa, but it remains an important chapter.
Image: Ron AI
THE rubric of robust statistical measures is essential for evaluating policies and plans within the context of democratic governance.
A RECENT Sunday Independent article questions the relevance of the Gini coefficient as a measure of inequality in South Africa, describing it as outdated, narrow, and even politically manipulative.
It argues that the Gini fails to account for social grants, informal economies, and the growing black middle class, concluding that we must retire it and replace it with a new, locally informed metric.
The critique is welcome and necessary. As Statistician-General, I support public scrutiny of the tools we use to measure our society. But I caution against discarding useful instruments because they are imperfect. The Gini coefficient is not the full story of inequality in South Africa, but it remains an important chapter.
Developed in the early 20th century, the Gini coefficient is a single statistic that indicates how evenly (or unevenly) income or wealth is distributed. It is widely used by national and international bodies, especially in relation to the Sustainable Development Goals (SDGs), particularly Goal 10, which focuses on reducing inequalities.
This commitment is reflected in Agenda 2063, the African Union's (AU's) socio-economic transformation plan. It is embedded in South Africa's National Development Plan (NDP), which aims to reduce our Gini from 0.69 to 0.60 by 2030.
The article makes a valid point: inequality is complex, and no single measure can capture it all. The Gini does not reflect the value of the 'social wage' — free education, healthcare, grants, housing subsidies — and may undercount informal economic activity. But it is not meant to measure everything.
It is one tool among many, and it tells us something important: South Africa remains one of the most unequal societies in the world, even if we have made real progress in reducing poverty. As a statistician, I use it as part of my statistical toolkit.
Statistical measures are essential for data analysis and informed decision-making, revealing patterns and trends. In his 2005 paper, Aziz Othman emphasises that effective policies rely on quality data. There is a growing shift among governments and organisations from opinion-based to evidence-based policy, underscoring the need for credible statistical analysis in policy formulation.
National statistical agencies in the United Kingdom and Australia also produce Gini coefficient statistics relevant to their contexts. This highlights the importance of continuous monitoring of income inequality and the integration of statistical methods into policymaking, as discussed in Othman's paper.
With more than 30 years of experience in producing official statistics at both national and continental levels, I have come to understand that poverty and inequality are complex issues that span social, economic, and political dimensions. This complexity shows that a single measure cannot fully capture these challenges.
Thus, using various statistical methods is essential. Statistics SA (Stats SA) employs three main approaches to assess poverty: traditional money-metric measures based on national poverty lines, multidimensional methods like the SA Multidimensional Poverty Index (Sampi) and Child Multiple Overlapping Deprivation Analysis (Moda), along with subjective assessments that reflect personal views of poverty.
Similarly, in analysing inequality, the Gini coefficient is but one of several metrics used by Stats SA to quantify economic disparities. Additional indicators include inequality experts Henri Theil's indices, Anthony Atkinson's indices, and Alex Sumner's Palma ratio.
Each of these measures possesses distinct strengths and weaknesses, yet all are widely recognised and used by National Statistical Offices (NSOs) and scholars globally to elucidate the structure and magnitude of inequality within a nation.
It is important to note that the Gini coefficient facilitates understanding income and expenditure distributions across households rather than functioning as an all-encompassing indicator of inequality, contrary to what the article may imply.
Furthermore, additional measures based on asset data, service delivery data, and labour market information produced by Stats SA are also useful for understanding the broader issue of inequality beyond economic indicators such as the Gini coefficient.
The simplest approach to measuring income inequality involves segmenting the population or households into quintiles, ranging from the poorest to the richest, and analysing the distribution of income or expenditure across these segments.
Recent Income and Expenditure Survey (IES) results indicate that about 75% of white-headed households are within the upper income quintile. Conversely, nearly half (45.1%) of black African-headed households fall within the lowest two quintiles in terms of income. Similar trends are observed in expenditure, where about 45.3% of black African-headed households are also categorised within the bottom two expenditure quintiles.
This data underscores the significant disparities in economic status between these demographic groups.
The findings illustrate a significant disparity in income and expenditure per capita, clearly highlighting the entrenched income inequality in South Africa, particularly affecting black African-headed households. Notably, nearly 57% of households within the lowest income quintile are female-headed. However, this proportion diminishes across the quintiles, with 49.5% of the second quintile, 42.9% in the third, 34.5% in the fourth, and only 33.5% in the upper quintile.
This decreasing representation of female-headed households in higher quintiles underscores the persistent issue of gender inequality within the socio-economic landscape. South Africa has extensive survey data on individual and household welfare from Stats SA, which offers various indicators of poverty and inequality.