## Channing Arndt, Andy McKay, and Finn Tarp

Print publication date: 2016

Print ISBN-13: 9780198744795

Published to Oxford Scholarship Online: May 2016

DOI: 10.1093/acprof:oso/9780198744795.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy).date: 18 November 2017

# Poverty, Inequality, and Prices in Post-Apartheid South Africa

Chapter:
(p.393) 17 Poverty, Inequality, and Prices in Post-Apartheid South Africa
Source:
Growth and Poverty in Sub-Saharan Africa
Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780198744795.003.0017

# Abstract and Keywords

Post-apartheid poverty and inequality trends have been the subject of intensive analysis, yet relatively little attention has been devoted to the impact of differential price movements on the measurement of poverty and inequality. This chapter aims to tell the story of the evolution of both money-metric and non-money-metric poverty and inequality in post-apartheid South Africa, and to assess the effect of prices on this story. Without accounting for price changes, a consensus narrative emerges of falling income poverty and income inequality remaining stubbornly high. Asset and multidimensional poverty indices show stronger declines in poverty and a decline in inequality. Price changes are shown to be anti-poor and to worsen inequality in the latter half of the 2000s. Thus, these price changes dampen measured improvements in poverty and inequality. The prices of food, electricity, water, and transport are shown to be particularly influential in these results.

Keywords:   South Africa, poverty, inequality, post-apartheid, prices, poor

# 17.1 Introduction

The widespread poverty and extreme inequalities prevalent at the time of the democratic transition represented a key area of policy focus for the first democratic government, as well as one of the sets of outcomes against which its performance has often been judged. Poverty and inequality trends during the post-apartheid period have consequently been the subject of intensive analysis in South Africa. While little data on household incomes and expenditures existed prior to the transition, regular nationally representative household surveys undertaken since the early 1990s by Statistics South Africa and other institutions have filled this gap. The availability of these data provided the platform for the research effort.

That said, the effect of prices on purchasing power is typically given only passing attention in the South African literature on poverty and inequality. Typically, incomes or expenditures are deflated by a scalar derived from some version of the consumer price index (CPI) in order to make comparisons over time. This leaves the effect of differential price movements across the distribution as a key gap. It is hoped that this chapter will contribute towards filling this gap by more purposefully considering the impact of prices on estimates of poverty and inequality.

The chapter starts by reviewing the received wisdom on post-apartheid growth and poverty well-being. It summarizes the secondary literature along with some new descriptive work that uses income data and non-money-metric sources. Section 17.3 reviews the available national expenditure data covering the past almost twenty years, with a view to choosing appropriate datasets for the analysis of the role of prices. It also considers the relevant available price (p.394) data. The section goes on to assess in detail the sensitivity of poverty and inequality estimates to differential price movements over this period. The central task of section 17.4 of the chapter is to tease out which prices changes were particularly influential in driving these findings. Section 17.5 concludes by summarizing findings, paying particular attention to the implications for policies that have been put in place over the post-apartheid period to address the apartheid legacy of high poverty and inequality.

# 17.2 Existing Evidence on the Evolution of Post-Apartheid Well-being

## 17.2.1 The Narrative

South Africa’s economy has undergone substantial changes since the advent of democracy in 1994. Economic growth stagnated during apartheid due to sanctions on international trade and investment, uncompetitive local industries, rigid exchange controls, restricted skills development, and high levels of poverty and inequality (Aron et al. 2008). After the first democratic election, economic sanctions were dropped, labour restrictions were lifted, and policies were put in place to advance the interests of African workers. South Africa has since had stable macro management and, as shown in Table 17.1, the economy has grown steadily both in real and nominal terms.

Over the same period, the schooling system was transformed from one characterized by highly skewed spending across racial groups to one based on equitable government funding. School enrolment rates rose, though learning achievements remain very poor in previously disadvantaged schools (Van der Berg 2007). The new, young labour market participants have more education, on average, than their parents. Two in five young adults graduate with

Table 17.1. South African macroeconomic trends, 1993–2012

GDP (ZAR million)

GDP growth (%)

GDP per capita

GDP per capita growth (%)

1993

1,065,830

1.2

28,277

−0.9

1997

1,214,768

2.6

29,582

0.5

2001

1,337,382

2.7

30,024

0.8

2005

1,571,082

5.3

33,176

3.9

2008

1,814,594

3.6

36,392

2.3

2010

1,842,052

3.1

36,079

1.9

2012

1,954,303

2.5

37,476

1.5

Avg. 1993–2012

1,470,001

3.2

32,031

1.5

Source: Updated from Leibbrandt et al. (2010); South African Reserve Bank (2013)

(p.395) Matric certificates (the qualification awarded to those who pass nationally set, standardized exams at the end of secondary schooling).

Other countries, such as Brazil and India, have seen education gains translate into productivity and employment growth, and large decreases in poverty and inequality. Job creation in a dynamic labour market served as the key pathway through which these societies generated high social returns to improved education and indirect benefits to social transfers.

South Africa has not made similar gains. Over the post-apartheid period, poverty has fallen only sluggishly. Eighteen years after the first democratic election, the share of people living below a US\$2 per day poverty line has declined by no more than four percentage points from 34 per cent in 1993 to 30 per cent in 2008. These gains are often attributed to social policy reforms (a massive expansion of cash grant transfers) rather than economic development (Leibbrandt et al. 2010). Of equal concern is the fact that inequality has risen further from its very high levels under apartheid (Leibbrandt et al. 2010).

Just as the labour market was the key intermediary in the successes in Brazil and India, so the unsatisfactory performance of the labour market sits centre stage in South Africa’s disappointing development outcome. A total of 2.74 million jobs (net) were created between 1993 and 2008, of which 2.5 million were targeted at skilled labour, while unskilled workers lost a total of 770,000 jobs (net). Over the same period, unemployment rates more than doubled from 14 per cent in 1993 to a peak of 29 per cent in 2001, before declining to 23 per cent in 2008. By the time of the economic crisis in 2010, the unemployment rate had risen to 25 per cent, using the narrow definition of unemployment (National Treasury 2011). If discouraged workers—those who have stopped looking for work ‘because they do not anticipate finding any’—are included in this definition, the figure is substantially higher at about 32 per cent (Statistics South Africa 2012b). Of the total population of four million unemployed, 75 per cent are long-term unemployed and many young job seekers report having limited or no formal work experience, even at age 30 (National Treasury 2011). The informal sector is small, with only 6 per cent of South Africans in self-employment. The supply of labour is therefore primarily directed at jobs in the formal sector.

In general, this labour market situation has had a negative impact on poverty because of the failure to pull individuals from poor households into employment. Leibbrandt et al. (2010) show that, in 2008, 32 per cent of South Africa’s 13.2 million households were no-worker households. The same figure was 26 per cent in 1993, implying an increase over the last fifteen years in the number of households relying on assistance, especially child support grants, as their main form of income. Indeed, the improved aggregate poverty situation is due to increased support from social grants, and not from the labour market. Even in one-worker households, the poverty incidence remains high (p.396) as many workers are in low-paid employment. Clearly an employed person within a household does not a guarantee an escape from poverty.

The poverty impacts of pervasive unemployment are compounded by a social protection gap that exists for unemployed adults, as social cash grants target people who are not expected to be economically active: children, pensioners, and people with disabilities. This leaves unemployed adults deeply dependent on goodwill transfers from within their communities, placing a large care burden on communities and deepening poverty.

Leibbrandt et al. (2010) show that these same labour market dynamics have played a dominant role in driving inequality over the post-apartheid period. Even though the average share of wage income in total income has remained constant at around 70 per cent, wage income has contributed 85–90 per cent of the total inequality in household income. In contrast, state transfers are shown to be mildly redistributive or to have a neutral impact on the overall Gini coefficient.

## 17.2.2 Trends in Money-Metric Poverty and Inequality

Most of the analysis of poverty and inequality in post-apartheid South Africa has used income as the welfare measure. Leibbrandt et al. (2010) use household income per capita to track changes in inequality and poverty between 1993 and 2008, and include a short section on the comparability of income and expenditure in the datasets that were used. It is understood that the expenditure data in 1993 are not as reliable as the income data, thus motivating the focus on an income-based comparison.

We briefly present some of the quantitative analysis that has been undertaken in support of the above narrative using income data from national household surveys. Figure 17.1 shows three post-apartheid real income per capita densities as an example of extensive empirical work that has been undertaken on the distribution of income (Fedderke et al. 2003; Simkins 2004; Hoogeveen and Özler 2006; Van der Berg et al. 2006, 2008). It provides a representative snapshot of the weight of evidence.1 A poverty line is inserted on the graph as a reference point. It is a cost-of-basic-living poverty line developed by Hoogeveen and Özler (2006) and has a real value of ZAR573 per person per month.

The graph shows that the distribution of real income shifted rightwards at almost all points between 1993 and 2010. At the bottom of the distribution, the major shift took place between 1993 and 2000, with relatively little movement between 2000 and 2010. This pattern is reversed as we move up (p.397)

Figure 17.1. Distributions of income, 1993, 2000, and 2010

Source: Authors’ illustration based on data from PSLSD 1993, Income and Expenditure Survey (IES) 2000, and NIDS Wave 2 2010

the distribution (but remain below the poverty line) where we see that there was a significant rightward shift from 2000 to 2010.

There is evidence of a significant rightward shift at the very bottom of the distribution, and poverty dominance analysis confirms a reduction in poverty. However, this is not a dramatic decrease in poverty. According to the poverty head count ratio—simply the proportion of the population living below the poverty line—the poverty rate at the lower poverty line stood at 56 per cent in 1993 and remained steady at around 54 per cent for the later years in our analysis. The reduction in poverty incidence using the upper poverty line also stands at two percentage points—from 72 per cent in 1993 to 70 per cent in the late 2000s. The rightward shift at the bottom of the distribution is reflected by consistent decreases in the poverty gap rate, which gives us a broad measure of the depth of poverty in society. The main driver behind increasing incomes at the bottom of the distribution is the rapid expansion of the government social support programme. The importance of state grants in raising these incomes is highlighted in Leibbrandt et al. (2010), who note that in 1993 one-fifth of households were beneficiaries of state grants, while in 2008 this proportion had climbed to one-half, and Leibbrandt and Levinsohn (2011), Bhorat and Van der Westhuizen (2011), and Woolard and Leibbrandt (2011) show clearly that social grants reduced both poverty and inequality.

(p.398) The expansion of government grants was not complemented by a reduction in the unemployment rate. The labour market is by far the most important factor to consider when decomposing poverty (see Leibbrandt et al. 2010). While the expansion of state support has helped to lower poverty, persistently high levels of unemployment have prevented poverty reduction on a substantial scale. In 1993, almost 90 per cent of individuals living in a workless household were living below the poverty line. This reduced somewhat to around 80 per cent in the period under study, but it remains very high. In fact, almost half of the poor in the country live in a workless household. This is in contrast to the poverty share of those living in households with two or more workers, which stands at around 17 per cent.

Leibbrandt et al. (2010) decompose poverty by different groups and reveal that the decrease in poverty in post-apartheid South Africa is driven mainly by a fall in the poverty incidence among Africans, and particularly African males. Poverty rates for this group fell from 66 per cent to 60 per cent, while the corresponding figures for African females are 72 per cent and 68 per cent. Despite these changes, the African share of overall poverty remained constant at 93 per cent in 1993, 2000, and 2010. This far outweighs the African share in the overall population, which is close to 80 per cent.

A great deal of rural–urban migration took place in South Africa in the period under study. Our data reflect that the share of urban residents in the population rose from 49 per cent in 1993 to 60 per cent in the late 2000s. As a result of this movement, the urban share of total poverty rose from 30 per cent to about 43 per cent. That said, the poverty rate in rural areas was higher than in urban areas for any choice of poverty line.

We move now to a discussion of inequality. At the start of the post-apartheid period South Africa stood as one of the most unequal societies of all countries with reasonably good survey data. In panel A of Figure 17.2, we plot three Lorenz curves for 1993, 2000, and 2010.

All three Lorenz curves suggest the high level of inequality. The richest 20 per cent of people earn about 70 per cent of the total income, and the second richest about 20 per cent of total income. Thus the poorest 60 per cent together earn about 10 per cent of the total income in the population. This is approximately true regardless of which dataset is being used, and is exceptionally low by international standards. The primary observation is that the distributions do not vary much with time. In this case, the 2000 graph lies slightly below 1993, and the 2010 distribution almost perfectly overlaps with 1993. The big picture conclusion is that inequality has remained mostly stable and stubbornly high over the post-apartheid era (see Leibbrandt et al. 2010; Van der Berg 2011).

Whereas Lorenz curves are unaffected by the mean of the income distribution, the generalized Lorenz curves of panel B are shifted up by mean income. (p.399)

Figure 17.2. Lorenz curves, 1993, 2000, and 2010

Source: Authors’ illustration based on data from PSLSD 1993, IES 2000, and NIDS Wave 2 2010

While the Lorenz curves give a graphical measure of income inequality, the latter provide a graphical measure of social welfare through the inclusion of both inequality and mean income. If everyone in a society earned twice as much as they previously did, the new generalized Lorenz curve would rotate upwards, whereas the corresponding Lorenz curve would remain unchanged. (p.400) We observe from panel B that the 1993 distribution is always below the 2000 distribution, which in turn is always below the 2010 distribution. Thus, panels A and B together reflect a society with high but stable inequality and with rising mean incomes amounting to an improvement in aggregate welfare over this time period.

However, an increasingly pressing policy focus has developed as to why South Africa’s inequality seems to be so stubbornly persistent. Some of the evidence points to the emergence of a small but well-paid black professional class. Hoogeveen and Özler (2006) find increases in inequality between 1995 and 2000, and attribute this mostly to increases in inequality among the African subpopulation. Using decomposition work, Leibbrandt and Levinsohn (2011) have emphasized the importance of unemployment and earnings. Pushing further down this road, a literature has highlighted the high rates of return to tertiary qualifications in conjunction with wide variations in the quality of primary and secondary schooling (Van der Berg 2009; Pellicer and Ranchhod 2012; Branson and Leibbrandt 2013a, 2013b).

Figure 17.3 follows on to provide a representative snapshot of the empirical work that has been undertaken to understand the drivers of these changes. It shows the share of income sources in total household income by income quintile in 2008. The proportion of income derived from wages increases linearly by income quintile. If a person is a member of a household situated in the poorest five deciles, the person is likely to receive relatively little wage income and to depend quite heavily on government grants and subsidies.

Figure 17.3. Share of household income from various sources, 2008

Source: Authors’ illustration; see text

(p.401)

## 17.2.3 Trends in Non-Money-Metric Poverty

The concept of welfare extends beyond the simple flow of income or expenditure into and out of a household and includes, among other things, the various assets accumulated by households over time. Further, in the context of public policy, many government interventions involve the transfer of assets and provision of services to households that are not picked up in income measures and are not necessarily easily valued in currency terms. Such assets include, for example, the provision of sanitation services or housing.

This makes the non-money-metric welfare story of the post-apartheid era considerably more straightforward than the money-metric story. These measures of non-money-metric welfare include a large number of services and assets that are directly impacted by the state’s roll-out of services. The provision of low-cost housing, the provision of access to water, improved sanitation, and massive electrification of particularly poor township areas have been prioritized and have boosted access rates. As a result, all the studies point to declines in non-money-metric poverty and inequality irrespective of the period within the last twenty years since 1993.

Bhorat and Van der Westhuizen (2013), using factor analysis, construct an asset index and find significant declines in non-income poverty and inequality between 1993, 1999, and 2004. Non-income poverty rates and the non-income poverty gap declined across a number of demographic covariates, with the results robust to the choice of poverty line (Bhorat and Van der Westhuizen 2013: 18). The non-income measure included dwelling type (formal or not); construction materials for roofs and walls; water access; power sources for lighting and cooking; sanitation; access to telecommunications, to a vehicle, and to a television. While well-established South African patterns of deprivation are reaffirmed in the study—Africans, females, and rural dwellers are typically worst off—the authors find that improvements in asset poverty and inequality were concentrated in the immediate post-apartheid period, rather than in the latter half of the period.

Bhorat et al. (2007) construct a so-called Comprehensive Welfare Index, as well as separate private and public asset indices. The Comprehensive Welfare Index is constructed to include both private and public assets, wage and non-wage income, and education levels. The analysis reveals that while poverty across all three of these indices declined between 1993 and 2005, the decline was more rapid for the Comprehensive Welfare and Public Asset indices between 1993 and 1999, and more rapid for the Private Asset Index between 1999 and 2005 (Bhorat et al. 2007: 48). Given increasing fiscal allocations towards public services and assets, this suggests that earlier interventions may have targeted ‘low-hanging fruit’, meaning they were simpler, cheaper, and had higher numerical impact. The later rapid improvements in the Private (p.402) Asset Index were underpinned by the period’s relatively good economic performance.

Wittenberg and Leibbrandt (2014) add a detailed discussion of asset inequality to this corpus. They begin by flagging the fact that it is hard to make comparisons of asset indices over the post-apartheid period because a standard asset bundle changes substantially between the beginning and the end of the period. Nonetheless, after extensive sensitivity checks to settle on a defensible set of asset bundles, they affirm the fact that these bundles show real welfare gains for South Africans over the period. They then use these asset indices and recently derived methods to estimate changes in asset inequality over time. This shows a clear and marked fall in asset inequality over the post-apartheid period.

Finn et al. (2013) construct a multidimensional poverty index (MPI) and compare measures of multidimensional poverty from the 1993 PSLSD (SALDRU 1994a, 1994b) and the second wave of NIDS (SALDRU 2008; Brown et al. 2012) in 2010/11. The index comprises three dimensions—education, health, and living standards—which themselves contain nine indicators. Some examples of indicators include school attendance, child mortality, nutrition, access to water and electricity, as well as an asset index. Using an MPI poverty line of deprivation in at least one-third of weighted indicators, the authors calculate multidimensional headcount rates of 37 per cent in 1993 and 8 per cent in 2010. The proportion of the population in severe MPI poverty also dropped substantially from 17 per cent to 1 per cent. This strong decrease is reflected in the MPI measure itself (the MPI headcount multiplied by the intensity of poverty), which fell from 0.17 to 0.03 over the period. The largest drivers of the reduction in multidimensional poverty were access to water and electricity. The authors compare the drop in MPI poverty to money-metric poverty between 1993 and 2010, and demonstrate that MPI improvements were notably larger.

The consensus is that non-income measures of poverty and inequality tell a more positive story of the post-apartheid period, and one that does not appear to be materially impacted by the choice of base year for comparison. Asset poverty and inequality levels have declined as the state actively intervened to uplift poor and marginalized communities through the provision of basic services and improved housing. Importantly, the evidence suggests that demographic and locational markers of disadvantage are being eroded quite significantly over time, as within-group differences explain an increasing proportion of inequality.

# 17.3 Poverty, Inequality, and Prices

## 17.3.1 The Availability of Data to Assess the Role of Prices

The above story of slight declines in money-metric poverty, stronger declines in asset poverty, and persistently high inequality is consistent with a large (p.403) body of evidence and is pretty settled. However, expenditure patterns differ considerably across the distribution as do the price indices for different consumption components (Oosthuizen 2007; Finn, Leibbrandt, and Oosthuizen 2014). The impact of these relative price changes and relative inflation rates is a near unexplored aspect of these changes in well-being. This is despite the fact that the way in which price changes are accounted for may impact on the observed rankings themselves, as well as the rankings relative to the poverty line, distorting the chosen measures of poverty and inequality.

The analysis in this chapter thus far has used the headline CPI to deflate nominal household incomes and expenditures to 2008 prices. However, this implies that each expenditure item or category within the aggregate has experienced identical price changes. Even a cursory glance at the various published product indices will confirm that is contentious. In combination with differences in consumption bundles across households, this means that a single deflator is unable to adequately account for price changes for all households. Indeed, household-specific inflation rates can vary quite substantially in a given period. Between January 1998 and December 2008, for example, it is estimated that an average of one-third of urban South African households actually experienced rates of inflation within one percentage point of the overall urban inflation rate (Oosthuizen 2013).

Various household surveys collecting information on household expenditures have been conducted in the past twenty years in South Africa, varying in level of detail, geographical coverage, and representivity. Moreover, the expenditure data in these various datasets reveal some awkward anomalies that complicate comparisons over time (Finn, Leibbrandt, and Oosthuizen 2014). Taking these expenditure datasets at face value results in a volatile picture of poverty and inequality trends that lacks plausibility when benchmarked against the the extensive money-metric and non-money-metric literature. Given the similarities in methodology, Statistics South Africa’s Income and Expenditure Surveys (IES) of 2005/6 (Statistics South Africa 2008) and 2010/11 (Statistics South Africa 2012a) and Living Conditions Survey (LCS) of 2008/9 (Statistics South Africa 2011) seem to provide a coherent and plausible picture of the most contemporary period. Therefore we use them for the analysis that follows. Unfortunately, this analysis deems that there is not a 1990s dataset that is a plausible comparative baseline of the early post-apartheid period. The chosen period of analysis also coincides with a period in which the geographical coverage and classification system of the CPI is generally consistent. Going forward, we make use of official price indices for all urban areas.2

## (p.404) 17.3.2 Does Accounting for Price Changes Affect Poverty Trends?

Different households consume different baskets of goods and experience differing rates of inflation over time. How important were price changes for headcount poverty rates in South Africa between 2005 and 2010? In answering the question, we compare CPI-adjusted cumulative distribution functions (CDFs) to percentile-specific price inflation indices (PCPI) CDFs. In all cases we are converting nominal expenditures for 2005 and 2010 into 2008 real Rand equivalents. For the CPI CDFs we deflate (2010) or inflate (2005) each percentile of the entire distributions using the same headline CPI price index. For the PCPIs we deflate or inflate each percentile of the expenditure distribution for each period by its own price index.

Constructing the PCPI involves two steps. First, we calculate the share of each expenditure item in total consumption expenditure for each percentile in 2008, and then assign this share as the weight for each relevant item. This is done at the most detailed level of disaggregation for which Statistics South Africa publishes price indices. Second, we multiply the weight by the price change for each item and then sum across each item for each percentile to arrive at the PCPI. So, for example, the percentile-specific inflation faced by percentile x at time t across items i to k is:

$Display mathematics$
(17.1)

where w(i,x) is the weight of expenditure item i for percentile x and pi is the price of item i. For comparability with the CPI-adjusted CDFs and in order to focus on price changes, the weights were held constant across the years, and were based on the shares derived from the 2008 LCS dataset.

A quick way of graphically assessing the impact of percentile-specific price changes on poverty is to compare CDFs where expenditure is inflated or deflated by CPI in one case and by PCPI in the other. This is what is shown in Figure 17.4. In the upper panel representing the 2005 situation, the solid line inflates the 2005 expenditure distribution into a 2008 real equivalent, assuming that each good consumed by anyone anywhere along the distribution experienced the same price change between 2005 and 2008—namely, that represented by the increase in the CPI over the period. The PCPI adjusts expenditures using a percentile specific price index. The dashed line plots this CDF. It is nowhere below that of the solid line which corresponds to the CPI deflator. This implies higher proportions of the population have lower real expenditure levels when the PCPI is used compared to the CPI situation. The value of goods that 2005 expenditures can buy is less and headcount poverty rates are higher at the poverty line of ZAR6,084 or any other poverty line when percentile-specific price changes are taken into account. (p.405)

Figure 17.4. Impact of prices on poverty as demonstrated by CDFs

Source: Authors’ illustration based on data from IES 2005/6, LCS 2008/9, IES 2010/11, and published price indices

(p.406) The lower panel shows that the situation is different in the IES 2010 data. Now, the CDFs deflating by the CPI and the PCPI overlap almost perfectly over the range R0 to ZAR15,000 of household expenditure per capita per year. Between 2008 and 2010 the magnitude of each percentile specific price change was slightly less but very close to the price change reflected in the overall CPI. As a consequence, headcount poverty is very close too at the poverty line of ZAR6,084 or any other poverty line.

Given that the 2005 PCPI CDFs are notably above their respective CPI equivalents whereas the 2010 PCPI CDFs are very close to their CPI equivalents, the net effect of percentile-specific inflation over the period 2005 to 2010 worsens poverty at any poverty line. Finn, Leibbrandt, and Oosthuizen (2014) illustrate this point by decomposing poverty changes at one specific poverty line into a growth component, a redistribution component, and a price component, making use of the Günther and Grimm (2007) extension of the Datt and Ravallion (1992) poverty decomposition. This analysis shows that over the period 2005–10, price changes were anti-poor whether measurement is at the national level or separated into urban and rural components. Indeed, the rural poor were particularly hard-hit.

## 17.3.3 Does Accounting for Price Changes Affect Inequality Trends?

The relationship between prices and consumption inequality is complicated by the fact that higher expenditure households have a larger weighting within the calculation of consumer price indices. As a result, the expenditure weights underlying the official CPI are biased towards higher expenditure groups and the ‘representative household’ that the CPI is meant to track may not be representative of the broader population.

Exactly where in the distribution the representative household is located depends on the extent of inequality: higher levels of inequality are associated with a location further up the distribution. For example, the representative household in Spain in the 1990s was located at the 61st percentile of the distribution (Izquierdo et al. 2003: 149) and in the 75th percentile in the United States in 1990 (Deaton 1998: 43). In Brazil, Colombia, Mexico, and Peru, the representative household is located between the 80th and 90th percentiles based on data between the 1980s and the early 2000s (Goni et al. 2006: 7). In South Africa, the representative household was located in the 95th percentile in 2000 (Oosthuizen 2007: 20). Conventional CPIs, therefore, do not track the experiences of what would be considered the ‘average’ household and provide an imperfect measure of the inflation experienced by poor households.

Not surprisingly therefore, visual inspection of Lorenz curves is very useful. In Figure 17.5, we compare Lorenz curves for 2005 and 2010. In each year, the (p.407)

Figure 17.5. Impact of prices on inequality as demonstrated by Lorenz curves

Source: Authors’ illustration based on data from IES 2005/6, IES 2010/11, and published price indices

Lorenz curves are plotted using nominal expenditure and, in 2005, we plot the 2005 expenditures in 2010 prices. Comparing the nominal Lorenz curves in the two years suggests that inequality declined between 2005 and 2010—the 2010 Lorenz curve lies closer to the line of equality than does the 2005 curve.

When the 2005 basket is priced in 2010 prices it is marginally closer to the line of equality at the lower end of the distribution and slightly further away from the line of equality for the richest quarter of households. The gap between the own-priced Lorenz curves for 2005 and 2010 reflects the total change in nominal inequality and is equivalent to comparing the Gini coefficients of the two nominal distributions. The gap between the two Lorenz curves based on 2010 prices and in which only the baskets differ offers a measure of the change in real inequality. The gap between the two Lorenz curves based on the 2005 basket—in which only the prices differ—corresponds to the effect of differential inflation.

The effect of prices on inequality measures can be estimated more precisely. Following Ruiz-Castillo et al. (2002) and Goni et al. (2006), we assume an inequality index, ξ‎(xt), in period t to be an increasing function in the level of inequality. The vector xt measures nominal expenditures, i.e. xt = (x1t,…,xHt) = (pt′c1t,…,pt′cHt), where pt represents prices from period t and ct represents household consumption, for all households (denoted by the superscripts (p.408) 1,…,H). If xt,s = (pt′c1s,…,pt′cHs) is the vector of household consumptions in periods evaluated at the prices of period t, then:

$Display mathematics$
(17.2)
$Display mathematics$
(17.3)

The first pair of terms in Equation (17.2) refers to the difference in the inequality measure when the baskets from the two different periods (t and t−1) are priced in terms of the prices of a single period (period t)—the only difference between the two sets of baskets being the quantities consumed. The second pair of terms refers to the change in the inequality measure when the base period baskets (i.e. period t−1) are priced using the sets of prices from each period. In this case, the only difference between the two sets of baskets is the prices at which they have been valued. Changes in nominal inequality may therefore be thought of as consisting of a component that reflects the effects of changes in quantities consumed (∆‎ξ‎Q) and a component that reflects the effects of changing prices (∆‎ξ‎P). Goni et al. (2006: 5) describe ∆‎ξ‎Q as ‘changing real inequality’ and ∆ξ‎P as ‘inflation inequality’. In terms of the Lorenz curves presented in Figure 17.5, ∆ξ‎Q corresponds to the comparison of the two baskets in 2010 prices, while ∆ξ‎P corresponds to the comparison of the two Lorenz curves based on the 2005 basket.

Table 17.2 presents the decompositions for the 2005–10 period as a whole, by repricing individual households’ expenditure baskets using product category price indices. Estimates of inequality are estimates of nominal inequality. The percentage change in the estimates of inequality over the period is then decomposed into an inflation inequality component (P∆) and a real inequality component (Q∆). Since the chosen base period may impact on the results—the base period refers to the consumption basket used—the decomposition is performed using first the initial period and then the final period as the base period.

The data show that inequality declined in nominal terms between 2005 and 2010, with the Gini coefficient falling from 0.668 to 0.636. However, in real terms, the decline has been smaller, with inflation inequality estimated to account for between 35 per cent and 56 per cent of the observed changes in nominal inequality, depending on the base and the inequality measure. For example, price changes are responsible for 2.1 percentage points or 1.7 percentage points of the almost 4.8 per cent decline in the level of the Gini coefficient between 2005 and 2010, depending on the base period.

This is not a positive situation. As ∆ξ‎P is negative for the 2005–10 period, price changes have been against those at the bottom of the distribution. (p.409)

Table 17.2. Distributional effects of inflation inequality, 2005–10

Measure

Initial inequality

Final inequality

%∆

Base in initial period (percentage points)

Base in final period (percentage points)

P∆

Q∆

P∆

Q∆

National—2005 vs 2010

Gini coefficient

0.668

0.636

−4.77

−2.10

−2.67

−1.68

−3.09

Theil index

0.957

0.829

−13.37

−5.33

−8.05

−5.08

−8.29

Mean log deviation

0.838

0.756

−9.80

−5.46

−4.35

−4.09

−5.71

Note: (a) Estimates of nominal inequality use prices from March 2006 for the 2005/6 IES expenditures, from March 2009 for the 2008/9 LCS expenditures, and from March 2011 for the 2010/11 IES expenditures. These months are the same as those used by Statistics South Africa as the base month for IES and LCS data. (b) Inequality measures are calculated on the basis of per capita household expenditure (i.e. these are individual-level measures).

Source: Authors’ calculations

Differential price changes, therefore, created a ‘wedge’ between real inequality (∆ξ‎Q) and nominal inequality (∆ξ‎). For the period under review, the effect of prices has been to exaggerate the changes observed in terms of real inequality as both the real inequality and the inflation inequality effects were in the same direction.3 This finding is consistent with the finding of anti-poor price movements in terms of the above analysis of poverty.

# 17.4 Which Prices Drove these Price Impacts on Poverty and Inequality?

The poverty and inequality decompositions presented above confirm that differential inflation rates across the distribution matter for the measurement of poverty and inequality trends over time in South Africa. Both the poverty and the inequality decompositions characterize the 2005–10 period as one during which price changes were anti-poor, with inflation rates for poorer households typically exceeding those of better-off households. In this section we seek to understand these changes by probing the expenditure categories that have promoted a gap between the inflation rates of poor and non-poor households.

In order for poor households to experience higher average rates of inflation over a given period, poor households must be ‘overexposed’ to relatively high-inflation items and ‘underexposed’ to relatively low-inflation items, with (p.410)

Figure 17.6. Poor households’ exposure to high-inflation expenditure categories, 2005–10

Note: Four expenditure categories with negative overall price changes were omitted from the scatter plot. These are: postal services and telecommunication services (−11.4 per cent per annum; 0.7 relative weight); furnishings, floor coverings, and textiles (−2.1 per cent per annum; 1.1 relative weight); footwear (−0.5 per cent per annum; 2.0 relative weight); and purchase of vehicles (−0.5 per cent per annum; 0.01 relative weight).

Source: Authors’ illustration based on data from IES 2005/6, IES 2010/11, and published price indices

over- or underexposure defined in terms of these items’ weights among poor relative to non-poor households. Figure 17.6 illustrates this by plotting the relative weights of each expenditure category in the CPI—calculated as the ratio between the weight for the poorest 40 per cent of households and the weight for the richest 60 per cent of households—against each category’s average annual price change. The horizontal line indicates the average rate of inflation for the period and the vertical line shows the categories in which poor households are underexposed (to the left of the line) or overexposed (to the right of the line).

Expenditure categories with both high relative weights and above-inflation price increases over the period contribute to widening the gap in the price indices of poor and non-poor households. Ten of these have been highlighted. Of these, six are food categories and a seventh is a beverage category. The remaining three categories are electricity (a price regulated by government), tobacco (a price that is strongly affected by government, through taxation), and household supplies and services. At two extremes are electricity and bread and cereals. Bread and cereals have a very high relative weight—poor households spend 3.5 times more than non-poor on this expenditure category relative to (p.411) their total expenditures—so that almost any rate of price increase above inflation would make this category a substantial contributor to the gap between poor and non-poor price indices. Electricity, on the other hand, has a moderately high relative weight—the weight for poor households is 2.8 times that for non-poor—but its rate of price increase has been particularly strong. In fact, electricity is by some margin the expenditure category with the highest rate of price increase, having increased by 15.3 per cent per annum over the five-year period.

While the figure highlights expenditure categories characterized by rapid price increases and relatively high expenditure weights for poor households, it does not show that the higher average rate of inflation experienced by poor households is also caused by poor households’ underexposure to low-inflation items. The key items in this regard are purchase of vehicles; domestic worker wages; tertiary education fees; insurance; postal services and telecommunication services; and recreational and cultural services. Using more detailed price data for the 1997–2006 period and expenditure data for urban areas only, earlier work finds that purchases of vehicles, insurance of buildings, and computers and telecommunication equipment widen the gap between decile one and all urban households due to poor households’ underexposure to low-inflation items (Oosthuizen 2007: 54). Paraffin, other tobacco products, matches, and candles were found to widen the gap through poor households’ overexposure to these high-inflation items.

In Table 17.3 we investigate whether our analysis of the 2005–10 period is likely to apply to the whole decade. We do this by determining if the inflation rate for poor households is higher than the overall inflation rate over the 2000s. If so, inflation can be said to be anti-poor, and vice versa. The table presents a matrix of income and/or expenditure datasets from the 2000s, which yield a number of periods based on comparisons between pairs of

Table 17.3. Patterns of inflation, 2000–13

IES 2000

IES 2005/6

LCS 2008/9

IES 2010/11

Census 2011

IES 2005/6

Anti-poor

LCS 2008/9

Anti-poor +

Anti-poor

IES 2010/11

Anti-poor +

Anti-poor

Pro-poor −

Census 2011

Anti-poor +

Anti-poor

Neutral

Anti-poor −

Oct. 2013

Anti-poor +

Anti-poor

Anti-poor −

Anti-poor

Neutral

Notes: (a) Base periods for prices are as follows: IES 2000–2000 (since total country price indices are not available prior to 2002); IES 2005/6–March 2006; LCS 2008/9–March 2009; IES 2010/11–March 2011; and Census 2011–October 2011. (b) ‘Neutral’ means that the gap between the All Items price index and the All Items index for the poorest 20 per cent of households is less than one point. ‘Anti-poor +’ means that the gap between the All Items price index and the All Items index for the poorest 20 per cent of households is greater than twenty points. A minus sign (‘−’) indicates that inflation is weakly pro- or anti-poor (i.e. the difference in indices is more than one but less than three index points).

Source: Authors’ calculations

(p.412) datasets. Based on the datasets’ respective base periods in terms of prices, the pattern of inflation in each period can be determined. We have chosen to compare quintile one households (i.e. the poorest 20 per cent of households) with the national average. Patterns of inflation are described as pro- or anti-poor, or neutral. Inflation patterns are described as neutral when the difference between the two national indices and the quintile index is less than one point.

The table confirms that for most two-way comparisons between the five datasets (or points in time they represent), inflation was anti-poor. This means that conventional estimates of changes in poverty that deflate expenditures using the headline inflation rate would upwardly bias improvements in real incomes or expenditures for the bottom quintile. Similarly, not deflating expenditures or incomes (or using headline inflation to deflate) would overstate improvements in inequality or understate the extent of worsening real inequality. This effect would be quite strong when making comparisons between the IES 2000 and any of the later datasets. However, the effect would be much weaker when comparing any dataset from the LCS 2008/9 onwards with another later dataset. The effects here are typically weak and effectively neutral.

# 17.5 Conclusion

This chapter had two specific objectives. The first was to tell the story of the evolution of poverty and inequality in post-apartheid South Africa in a way that covers both money-metric and non-money-metric dimensions of well-being and teases out the drivers of these changes. For the most part, this was based on a review of an existing corpus. The second goal of the chapter was to assess the impact of prices on estimates of poverty and inequality, an issue that has received very little attention.

The post-apartheid narrative has to start with the inherited legacy of very high inequality and high poverty. Most obviously, the markers were the strong racial and spatial (rural) disadvantages associated with apartheid. Using income data from 1993 and an extensive literature that draws on these data, we detail the unequal human capital, asset, and labour market situation at the start of the post-apartheid period. At the broadest level, the specific intent of post-apartheid policymaking has been to confront and overturn this inherited situation. With the benefit of hindsight and our current knowledge about poverty traps, inequality persistence, and socioeconomic marginalization, we can understand how the pernicious correlation of these racial and spatial legacies with poorer human capital accumulation and poorer household and community assets presented the post-apartheid government with a most daunting set of constraints.

(p.413) Jumping ahead to the contemporary situation, we used comparable, national 2010 income data to show a somewhat improved money-metric poverty situation and an income inequality situation that is, at best, as unequal as the 1994 situation. This shows just how hard it has been to overturn the embedded apartheid dynamics of persistence poverty and inequality. That said, there are signs of progress. A comparison of multidimensional poverty indices shows that there have been improvements in human capital. Access to education and average years of schooling have expanded, and on the health side, there have been improvements in child mortality and nutrition. Plus there have been notable improvements in access to assets. Work on multidimensional poverty and the derivation of asset indices collectively show notable improvements in access to water and electricity, even for the poor. An analysis of asset indices shows improvements in access to sanitation and housing too. It is clear that asset inequality has fallen over the post-apartheid period.

These improvements in multidimensional poverty represent notable policy achievements. However, these achievements are not mirrored in equivalent money-metric improvements and this raises a fundamental question in understanding South Africa’s post-apartheid performance. Why has South Africa not generated social returns from its investments in human capital and assets? We have some but not all of the pieces of this puzzle in place. Drawing on a decomposition literature we show that South Africa’s failure to generate more jobs has been central to the inability to see improved human capital realized in more earners and in higher earnings for more productive workers. A dynamic labour market has to be the central mechanism through which a society transforms itself, through which assets embodied in people become livelihoods and income. The South African labour market has been very static and has not done this.

We reference a literature that situates some of this failure within the schooling system itself. This literature argues that the quality of schooling has declined. The apparent schooling improvements are a cruel mirage. This view needs to be balanced against the fact that the post-apartheid economy has operated in the global economic environment and a skills twist in the demand for labour. South Africa—along with all developed economies and many middle-income developing economies—has seen an increasing demand for skilled and semi-skilled labour and very flat demand for unskilled labour. What makes this skills’ twist so pernicious in the South African context is that the improvements in years of schooling (from seven to ten years) lie in the zone of falling returns. We have yet to see large increases in complete secondary (twelve years) or tertiary education. The market has strongly dampened the returns to the improved educational access that has been achieved.

(p.414) Another dimension of the labour market failure is a sluggish employment response to economic growth. Partly this failure is about the levels of growth. Our growth rates have hovered just below 3 per cent for most of the post-apartheid period. Nearly all planning models of the economy have been clear that we require growth rates in excess of 5 per cent to kickstart robust job creation. In the mid-2000s South Africa’s growth rates inched upwards towards these levels only for the financial crisis to cut them back to close to zero in 2008. We show that growth has been pro-poor. However, the key mechanism has not been strongly inclusive employment creation. Rather, increased tax revenues have been used to finance significantly increased state expenditures on social grants. However, as with education and health, without a dynamic labour market, these grants remain remedial. They are not generating second-round effects or pathways out of poverty through financing labour market entry or through financing the creation of small enterprises.

Thus, there have been significant successes but the need to generate labour-absorbing growth is highlighted as a fundamental challenge to further reductions in poverty and to reducing inequality. Does this picture require significant adjustments when giving account to different consumption bundles across the distribution and the different prices and price changes of these bundles?

We explored the role of prices in driving expenditure-based poverty and inequality over the 2005–10 period using price/poverty decompositions and price/inequality decompositions. In line with our earlier analysis, we showed that both growth and redistribution were pro-poor. However, percentile-specific price indices dampen improvements in poverty. In addition, while measured inequality declined somewhat over this period, this decline is exaggerated by the fact that inflation was anti-poor. In other words, giving specific attention to what the poor and those at the bottom of the income distribution consume, and pricing this bundle as accurately as possible, we showed that these groups have to spend more than an average CPI adjustment would reveal on their consumption bundles. At the end of the day it is the consumption bundle that is the real measure of well-being and expenditure is merely a proxy measure. Indeed, the implication of our decomposition work is that some of the increase in the expenditures of the poor do not signal an increase in consumption and therefore in real well-being but rather an increase in the cost of the same consumption bundle.

Adding prices into the post-apartheid narrative does not make things look any better. Rather, it cautions that, over the 2000s, the real support for those at the bottom of the distribution has not been as substantial as usually indicated. Our detailed analysis signals the exposure that those at the bottom of the expenditure distribution have had and continue to have to food price (p.415) movements. This would be true anywhere in the world; although the particular vulnerability of rural communities in South Africa flags the absence of support from subsistence food production. More novel is our finding that those at the bottom of the distribution have been relatively overexposed to high-inflation items such as electricity and food, and underexposed to lower-inflation items such as services generally and transport. Our review of non-money-metric well-being has shown evidence of the successful rollout of such services in the post-apartheid period. Success in the fight against poverty can be undermined by sharp increases in the pricing of such services. Moreover, basic foodstuffs and these services are more important components in the consumption bundles of those at the bottom of the distribution than those at the top. In the 2000s their prices rose more sharply than those goods consumed by the better-off, and this worsened inequality.

References

Bibliography references:

Aron, J., B. Kahn, and G. Kingdon (2008). ‘South African Economic Policy under Democracy: Overviews and Prospects’, in J. Aron, B. Kahn, and G. Kingdon (eds), South African Economic Policy under Democracy. Oxford: Oxford University Press, 1–27.

Bhorat, H. and C. Van der Westhuizen (2011). ‘Pro-Poor Growth and Social Protection in South Africa: Exploring the Interactions’. Unpublished submission to the National Planning Commission. Cape Town: Development Policy Research Unit, University of Cape Town.

Bhorat, H. and C. Van der Westhuizen (2013). ‘Non-monetary Dimensions of Wellbeing in South Africa, 1993–2004: A Post-Apartheid Dividend?’ Development Southern Africa, 30(3): 295–314.

Bhorat, H., C. Van der Westhuizen, and S. Goga (2007). ‘Welfare Shifts in the Post-Apartheid South Africa: A Comprehensive Measurement of Changes’, DPRU Working Paper 07/128. Cape Town: Development Policy Research Unit, University of Cape Town.

Branson, N. and M. Leibbrandt (2013a). ‘Education Quality and Labour Market Outcomes in South Africa’, OECD Economics Department Working Paper 1021. Paris: OECD Publishing.

Branson, N. and M. Leibbrandt (2013b). ‘Educational Attainment and Labour Market Outcomes in South Africa, 1994–2010’, OECD Economics Department Working Paper 1022. Paris: OECD Publishing.

Brown, M., R. C. Daniels, L. De Villiers, M. Leibbrandt, and I. Woolard (2012). National Income Dynamics Study Wave 2 Beta Release User Manual. Cape Town: Southern Africa Labour and Development Research Unit, University of Cape Town.

Datt, G. and M. Ravallion (1992). ‘Growth and Redistribution Components of Changes in Poverty Measures: A Decomposition with Applications to Brazil and India in the 1980s’, Journal of Development Economics, 38(2): 275–95.

(p.416) Deaton, A. (1998). ‘Getting Prices Right: What Should Be Done?’, Journal of Economic Perspectives, 12(1): 37–46.

Fedderke, J., J. Manga, and F. Pirouz (2003). ‘Challenging Cassandra: Household and Per Capita Household Income Distribution in the October Household Surveys 1995–1999’. Unpublished mimeograph, University of Cape Town.

Finn, A., M. Leibbrandt, and M. Oosthuizen (2014). ‘Poverty, Inequality, and Prices in Post-Apartheid South Africa’, WIDER Working Paper 2014/127, October. Helsinki: World Institute for Development Economics Research.

Finn, A., M. Leibbrandt, and V. Ranchhod (2014). ‘Post-Apartheid Poverty and Inequality Trends’, in R. Kanbur (ed.), The Oxford Companion to the Economics of South Africa. Oxford: Oxford University Press, 291–7.

Finn, A., M. Leibbrandt, and I. Woolard (2013). ‘What Happened to Multidimensional Poverty in South Africa between 1993 and 2010?’, SALDRU Working Paper Series 99. Cape Town: Southern Africa Labour and Development Research Unit, University of Cape Town.

Goni, E., H. Lopez, and L. Serven (2006). ‘Getting Real about Inequality: Evidence from Brazil, Colombia, Mexico, and Peru’, World Bank Policy Research Working Paper 3815. Available at <http://documents.worldbank.org/curated/en/2006/01/6525971/getting-real-inequality-evidence-brazil-colombia-mexico-peru>, accessed January 2013.

Günther, I. and M. Grimm (2007). ‘Measuring Pro-Poor Growth When Relative Prices Shift’, Journal of Development Economics, 82(1): 245–56.

Hoogeveen, J. and B. Özler (2006). ‘Poverty and Inequality in Post-Apartheid South Africa: 1995–2000’, in H. Bhorat and R. Kanbur (eds), Poverty and Policy in Post-Apartheid South Africa. Cape Town: HSRC Press, 59–94.

Izquierdo, M., E. Ley, and J. Ruiz-Castillo (2003). ‘The Plutocratic Gap in the CPI: Evidence from Spain’, IMF Staff Papers, 50(1): 136–55.

Leibbrandt, M. and J. Levinsohn (2011). ‘Fifteen Years On: Household Incomes in South Africa’, NBER Working Paper 16661. Cambridge, MA: National Bureau of Economic Research.

Leibbrandt, M., I. Woolard, A. Finn, and J. Argent (2010). ‘Trends in South African Income Distribution and Poverty since the Fall of Apartheid’, OECD Social, Employment and Migration Working Papers 101. Paris: OECD.

National Treasury (2011). ‘National Budget Speech’. Available at <http://www.treasury.gov.za/documents/national%20budget/2011/speech/speech2011.pdf>, accessed January 2013.

Oosthuizen, M. (2007). ‘Consumer Price Inflation across the Income Distribution in South Africa’, DPRU Working Paper 07/129. Cape Town: Development Policy Research Unit, University of Cape Town.

Oosthuizen, M. (2013). ‘Inflation Inequality in South Africa’, DPRU Working Paper 13/158. Cape Town: Development Policy Research Unit, University of Cape Town.

Pellicer, M. and V. Ranchhod (2012). ‘Inequality Traps and Human Capital Accumulation in South Africa’, SALDRU Working Paper Series 86. Cape Town: Southern Africa Labour and Development Research Unit, University of Cape Town.

Ruiz-Castillo, J., E. Ley, and M. Izquierdo (2002). ‘Distributional Aspects of the Quality Change Bias in the CPI: Evidence from Spain’, Economics Letters, 76(1): 137–44.

(p.417) SALDRU (Southern Africa Labour and Development Research Unit) (1994a). ‘Project for Statistics on Living Standards and Development, 1994. South Africans Rich and Poor: Baseline Household Statistics’. Cape Town: SALDRU.

SALDRU (Southern Africa Labour and Development Research Unit) (1994b). ‘Project for Statistics on Living Standards and Development’. Cape Town: SALDRU.

SALDRU (Southern Africa Labour and Development Research Unit) (2008). ‘National Income Dynamics Study’. Cape Town: SALDRU.

Simkins, C. (2004). ‘What Has Happened to the Distribution of Income in South Africa between 1995 and 2001?’. Mimeo, South African Regional Poverty Network (SARPN), Pretoria.

South African Reserve Bank (2013). ‘Quarterly Bulletin Online Statistical Query’. Online database. Available at <http://www.reservebank.co.za/>, accessed June 2013.

Statistics South Africa (2008). ‘Income and Expenditure of Households 2005/2006’, Statistical Release P0100. Pretoria: Statistics South Africa. Available at <http://www.statssa.gov.za/>, accessed January 2013.

Statistics South Africa (2009a). ‘Consumer Price Index Collection Frequency’. Available at <http://www.statssa.gov.za/>, accessed January 2013.

Statistics South Africa (2009b). ‘The South African CPI Sources and Methods Manual’. Pretoria: Statistics South Africa. Available at <http://www.statssa.gov.za/>, accessed January 2013.

Statistics South Africa (2011). ‘Living Conditions of Households in SA: 2008/2009’, Statistical Release P0310. Pretoria: Statistics South Africa.

Statistics South Africa (2012a). ‘Income and Expenditure of Households 2010/2011’, Statistical Release P0100. Pretoria: Statistics South Africa.

Statistics South Africa (2012b). ‘Quarterly Labour Force Survey (2012Q1)’, Statistical Release P0211. Pretoria: Statistics South Africa.

Van der Berg, S. (2007). ‘Apartheid’s Enduring Legacy: Inequalities in Education’, Journal of African Economies, 16(5): 849–80.

Van der Berg, S. (2009). ‘The Persistence of Inequalities in Education’. In J. Aron, B. Kahn, and G. Kingdon (eds), South African Economic Policy under Democracy. Oxford: Oxford University Press, 327–54.

Van der Berg, S. (2011). ‘Current Poverty and Income Distribution in the Context of South African History’, Economic History of Developing Regions, 26(1): 120–40.

Van der Berg, S., R. Burger, R. Burger, M. Louw, and D. Yu (2006). ‘Trends in Poverty and Inequality since the Political Transition’, DPRU Working Paper 06/104. Cape Town: Development Policy Research Unit, University of Cape Town.

Van der Berg, S., M. Louw, and D. Yu (2008). ‘Post-Transition Poverty Trends Based on an Alternative Data Source’, South African Journal of Economics, 76(1): 58–76.

Wittenberg, M. and M. Leibbrandt (2014). ‘Measuring Inequality by Asset Indices: The Case of South Africa’. Paper delivered at the UNU-WIDER conference ‘Inequality—Measurement, Trends, Impacts, and Policies’, 5–6 September, Helsinki.

Woolard, I. and M. Leibbrandt (2011). ‘The Role of Cash Transfers in Reducing Poverty and Inequality in South Africa over the Post-Apartheid Period’. Unpublished submission to the National Planning Commission. Cape Town: Southern Africa Labour and Development Research Unit, University of Cape Town. (p.418)

## Notes:

(1) This section is based on Finn et al. (2014).

(2) Since price data at the product level are not publicly available, the approach taken here is to use price indices at the most detailed level of disaggregation available (typically at a product group level).

(3) Although not presented here, separate urban and rural decompositions are broadly consistent with these aggregate results. Results for urban and rural areas separately can be found in Finn, Leibbrandt, and Oosthuizen (2014).