Kya Sands: Earnings inequality has widened since apartheid. Picture: Getty Images/Per-Anders Pettersson
Kya Sands: Earnings inequality has widened since apartheid. Picture: Getty Images/Per-Anders Pettersson

It is widely accepted that SA’s persistently high levels of inequality are not only unsustainable, they also curb the country’s growth potential.

Yet despite well-targeted government policies, inequality has remained sky-high. Over the past decade, progress to reverse inequality has, it would seem, effectively stalled.

Now the latest academic research shows that one of the chief reasons for the persistence of high inequality in SA is that top income earners have flourished while everyone else has struggled. The researchers have, for the first time, used SA Revenue Service (Sars) income tax data, which is better able to measure the income of high earners.

University of Cape Town (UCT) economics professor Martin Wittenberg estimates that not only has earnings inequality widened since apartheid, but that the Gini co-efficient inequality rating (for earnings), when measured by the tax data, could be higher than previously thought, at 0.599 not 0.567.

"Such high levels of inequality threaten the social fabric, increase the risks of political and economic upheaval and prevent the majority from living up to their full potential. All of these are likely to harm the country’s long-term developmental prospects," says Prof Murray Leibbrandt, who heads the Southern Africa Labour & Development Research Unit (Saldru) at UCT.

Leibbrandt and researchers Vimal Ranchhod and Pippa Green have co-authored a new SA-TIED working paper, "Taking Stock of SA Income Inequality", which summarises the latest research in the field produced by Saldru researchers as part of a large UN University World Institute for Development Economics Research (UNU-WIDER) project investigating inequality in Brazil, India, China, Mexico and SA.

SA-TIED is a collaboration between UNU-WIDER, the National Treasury and other governmental and research organisations in SA.

Such high levels of inequality threaten the social fabric, increase the risks of … upheaval and prevent the majority from living up to their full potential
Murray Leibbrandt

Leibbrandt says the Sars data has been critical in understanding income inequality in SA, particularly since the growth of top incomes was previously an underexplored area. Besides the tax data, the latest papers use new empirical methods to generate new insights about what is really driving inequality in the country.

A key finding by Stellenbosch University professor of economics Ingrid Woolard and UCT researcher Ihsaan Bassier is that the biggest driver of inequality since 2012 has been educational attainment — especially tertiary education. These gains, they found, have accrued disproportionately to the top earners.

They confirm Wittenberg’s finding that earnings at the top end rose much faster than for other groups — faster even than SA’s national income growth, which means that top earners are receiving an ever greater "share of the pie".

Between 2003 and 2015/2016, the real incomes of SA’s top 1% of income earners almost doubled. By contrast, the incomes of 95% of the population stagnated, or for those at the bottom showed only slight growth, in their case mainly because of social grants. In fact, nearly 60% of the population earned no taxable income at all during this period.

"This has put upward pressure on SA’s inequality levels, which partly explains why inequality has remained so persistent," the researchers explain.

In addition, the top percentile, which starts at a taxable income of R800,000 a year, has a much higher wealth-to-income ratio than the rest of the population. Income from sources other than salaries increases rapidly in the top two percentiles, allowing these people to accrue wealth faster and do better in periods of low economic growth, when wage growth typically slows overall.

In short, the wealthiest groups are able to draw on a far broader array of income sources and assets (physical, financial and human), and this has enabled them to flourish even as economic growth has slowed.

In another paper, Arden Finn and Leibbrandt show that wage earners stuck in the middle of the earnings spectrum, which has not moved since the end of apartheid, are predominantly African, male and in their 30s. They work in a mix of occupations including those involving craft and services, and in clerical posts.

While the average age of the median band hasn’t changed markedly, it has risen in the top half — which suggests that younger people are finding it harder to move into higher-paying jobs. The fact that it has not changed much in the bottom half either indicates that young people are finding it difficult to enter the labour force at all.

Equally alarming is some of the most recent work on social mobility.

A fascinating new paper, "Snakes and ladders and loaded dice — poverty dynamics and inequality in SA, 2008-2017", by Saldru researchers Rocco Zizzamia and Leibbrandt with Simone Schotte, a research associate at UNU-WIDER in Helsinki, shows that when looked at over time, poverty in SA is more pervasive than previously thought.

The researchers distinguish five major socioeconomic classes in SA: the chronically poor, the transient poor, the vulnerable middle class, the stable middle class, and the elite. They conclude that the stable part of the middle class is much smaller than previously thought, at only 21% of the population. This is because a large portion of society has, for the past decade, been engaged in a game of "snakes and ladders", slipping in and out of poverty in response to positive and negative shocks.

According to Stats SA, 55% of the SA population lived in poverty in 2015.

However, this snapshot of data is confined to a point in time. It understates the extent of the poverty problem, as the researchers estimate that while about 49% of the population live in chronic, persistent poverty, another 11.4% can be classified as "transient poor", and a further 19% constitute the "vulnerable middle class".

Both the latter groups are at risk of falling back into poverty at one time or another, when they lose a job or a family member. And though finding a job is important for lifting people out of poverty, it is not sufficient, say researchers.

Those with unstable jobs, without contracts or union protection, are more vulnerable to falling back into poverty than those with permanent, formal employment.

In the stable middle class, two-thirds of household heads have a matric or higher qualification and three-quarters are employed, typically in the formal sector. They also earn, on average, twice as much from their jobs as households in the vulnerable middle class do (roughly R13,127 a month compared with R5,366 a month).

The elite class is almost three-quarters white and predominantly urban based, and earn about R38,223 a month on average from the labour market. They also get a significant amount of their income from capital investments. Their expenditure level is, as you’d expect, much higher than that of the stable middle class at R25,659 a month, compared with R4,536 a month on average per capita.

Chronic poverty is associated with poor education (less than a matric), larger households, female-headed households (which also tend to be single-parent households), unemployment, geographic location and race.

"The legacies of apartheid, which forced many African people to live in poverty-stricken rural homelands far from economic opportunities, are still deeply felt," say the researchers.

What it means

Access to economic opportunities, markets and services that are currently at present unavailable in impoverished areas should be improved

For instance, they found that across four or five waves of data, nearly 83% of the rural poor remained so between 2008 and 2017. The urban African population is, however, more affected by transient poverty, with many more moving in and out of poverty over the same period.

"The other striking feature of poverty is its female and youthful face," the study says. Nearly 72% of those in female-headed households remained in poverty across four or five waves, compared with only 29% of those in male-headed households.

Race is another persistent legacy. Though the sample of white people was relatively small (just 274 individuals), 93% remained consistently non-poor between 2008 and 2017. By contrast, about 63% of Africans were poor most of the time, with only about 9% remaining non-poor the entire time.

On the other hand, there has been a significant shift in the racial composition of both the middle class and the elite. The African component of the middle class grew from 47% to 64% between 2008 and 2017, and that of the elite class from 14% to 22%. Despite this, Africans are still underrepresented in the middle class, while white people are overrepresented, compared with their share in the overall population.

The policy implications are clear: closing the skills gap and increasing the quantity and quality of jobs is the government’s central challenge, the goal being to lift larger parts of the population into the middle class and to prevent them sliding back into poverty.

The job prospects of the transient poor and vulnerable middle class could be improved by strengthening support for small enterprises, while providing insurance mechanisms could help buffer them against earnings shocks.

But for the large chunk of the population in persistent poverty, social grants and basic social services are indispensable. However, as these grants are too small to allow this group to break out of poverty, the focus should be on improving access to economic opportunities, markets and services that are currently unavailable in impoverished areas.

The efficiency of social spending must be raised to make existing government budgets go further.