The World Inequality Lab’s a wealth tax for South Africa: A critique (I) – The distribution of wealth

This is the first of a pair of Briefs developing the HSF’s work on wealth taxes.

INTRODUCTION

In 2018, the HSF published eight Briefs on wealth taxes[1]. This Brief updates the series by considering the World Inequality Lab’s A wealth tax for South Africa[2],published in collaboration with the Southern Centre for Inequality Studies at the University of the Witwatersrand[3]in January 2021.

ESTIMATING THE DISTRIBUTION OF WEALTH

The distribution of wealth is estimated in the companion technical paper. The aim of the paper is to provide an account for net wealth in all its forms. The distribution of wealth is across individuals rather than households. The estimates are built up from a number of sources[4], which have to be reconciled and assumptions are made in the process. This is the practice in other countries – a single data source is never to hand - and the paper is explicit about its methods. But the study concludes that the data sources available to measure wealth in South Africa remain largely unsatisfactory[5]. This means that there is considerable uncertainty about the final estimates, a fact which the authors acknowledge.

It is instructive to consider key findings of the analysis. Table 1 sets out the composition of aggregate household wealth in 2018.

Table 1

Source: Chatterjee et al, 2020: Table 1

Aggregate net wealth including offshore wealth was estimated at R 11.2 trillion in that year.

Table 2 sets out the estimated distribution of this wealth.

Table 2

Source: Chatterjee et al, 2020: Table 9

Most, but not all, of the bottom 50% are net debtors. The top 1% consisted of 354 000 individuals in 2018. The ownership of bonds and corporate shares is particularly concentrated, with the top 10% owning 99.8% of these forms of wealth, and the top 1% owning 95.2%.

Wealth varies over the life cycle in South Africa, just as it does in other countries. The study found that:

Average net worth rises significantly and linearly between ages 20 and 55: individuals aged 20–25 have an average net worth lower than 25 per cent of the national average, while those aged 50–55 are between 50 per cent and two times wealthier than the average adult. Average wealth then stabilizes between ages 50 and 65 and decreases slightly for older individuals, but still remains more than 50 per cent higher than the national average for individuals older than 75[6].

This point matters. Imagine a society in which people accumulated and then drew down wealth along the same path by age, and at the same level. Then from a lifetime perspective, there would be no inequality. But if one measured the wealth distribution at any point in time (as the companion piece does), then inequality would be found if ages differ. Of course, individuals in South Africa do not accumulate wealth at the same level. The studies point out that income inequality, debt, savings patterns and inequalities in access to inheritances also drive high levels of wealth concentration[7].

There is comparative information on levels of wealth inequality, as measured by the shares of net wealth held by the bottom 50%, the middle 40% and the top 10% and top 1%. There are seven other countries[8] for which the same data have been compiled by similar methods and reported in the World Inequality Database, and South Africa is most unequal of the eight.

It has been possible to estimate net wealth in South Africa back to 1970. Aggregate net wealth declined from about 270 per cent of national income in 1975 to 210 per cent in 1995, rising to 260 per cent in 2007 (the peak of the commodities boom) and dropping to 230 per cent in 2018[9].

Finally, it is worth noting how few rich people live in South Africa. The companion paper estimates that 355 600 people had wealth in excess of R 3 670 000 (the threshold for the top 1%) in 2018. This wealth threshold is not high. Owners of four bedroom houses in metros, unencumbered with mortgages, could easily belong to the top 1%, even if they have no other net wealth at all. Only 110 200 people were dollar millionaires, i.e. had wealth of above R 13 240 000[10].

CONCLUSION

Despite the uncertainties to which the companion paper is subject, it is the best in a very sparse literature on the distribution of wealth in South Africa. The next Brief will consider the proposals for a wealth tax which use the companion paper to make estimates of revenue under various assumptions.

Charles Simkins, Head of Research, Helen Suzman Foundation. 

Footnotes:

[1] They were all authored by Charles Collocott and comprise (1) Wealth taxes I: Conceptual framework, 4 April 2018, (2) Wealth taxes II: Rationales, 4 April 2018, (3) Wealth taxes III: Problems, 4 April 2018, (4) Wealth taxes IV: International experience (description), 5 April 2018, (5) Wealth taxes V: International experience, 5 April 2018, (6) Wealth taxes VI: Land tax, 5 April 2018, (7) Wealth taxes VII: Lessons for South Africa, 5 April 2018, (8) The Davis Tax Committee Report: An overview, 26 July 2018. The Briefs can be read on the HSF website www.hsf.org.za

[2] Aroop Chatterjee, Leo Czajka and Amory Gethin, A wealth tax for South Africa, World Inequality Lab Working Paper No 2021/02, January 2021, available at file:///C:/economy/WorldInequalityLab_WP2021_02_SouthAfrica_WealthTax_2b.pdf

[3] There is a companion technical study: Aroop Chatterjee, Leo Czajka and Amory Gethin. Estimating the distribution of wealth in South Africa, United Nations University World Institute for Development Economics Research, WIDER Working Paper 2020/45, April 2020, available at https://www.wider.unu.edu/sites/default/files/Publications/Working-paper/PDF/wp2020-45.pdf

[4] These sources include the national accounts as published by the SA Reserve Bank, personal income tax microdata from the SA Revenue Service, Statistics South Africa’s Income and Expenditure Surveys and Living Conditions Surveys, the National Income Dynamics Study surveys, and estimates of wealth held offshore.

[5] Chatterjee et al, 2020, p 25

[6] Chatterjee et al, 2020, p 21

[7] Chatterjee et al, 2020, p 22

[8] These are China, France, India, Korea, Russia, the United Kingdom and the United States

[9] Chatterjee et al, 2020, Figure 1

[10] These estimates are based on Chatterjee et al, 2021, Table 3. The SARS average exchange rate for the year ending December 2018 is used.