population policy implementation - The Population Policy of South

TOWARDS A 10-YEAR REVIEW OF THE POPULATION POLICY IMPLEMENTATION
IN SOUTH AFRICA (1998-2008)
Poverty and inequality
Niël Roux
Whynie Adams
MARCH 2009
© DEPARTMENT OF SOCIAL DEVELOPMENT, 2009
Table of Contents
Executive Summary .................................................................................................. i
1. Introduction ........................................................................................................ 1
1.1
Population and poverty ......................................................................................... 2
1.2
Economic Context ................................................................................................. 3
1.3
Definitions .............................................................................................................. 4
1.4
Data Sources .......................................................................................................... 5
1.4.1
Population censuses and the Community Survey .............................................. 6
1.4.2
Income and Expenditure Surveys (IES) ............................................................. 6
2. Trends in income monetary poverty and inequality........................................ 8
2.1
Trends in income poverty ..................................................................................... 8
2.1.1
Spatial dimensions of poverty .......................................................................... 13
2.1.2
Poverty shifts by gender of household head .................................................... 15
2.2
Trends in income inequality ................................................................................ 16
2.2.1
Share of consumption expenditure by population group................................... 17
2.2.2
Income and inequality ...................................................................................... 17
2.2.3
Gini coefficient ................................................................................................. 18
2.3
Discussion............................................................................................................ 20
3. Service poverty and inequality, and the ‘social wage’ .................................. 22
3.1
Access to services............................................................................................... 23
3.1.1
Water provision ................................................................................................ 23
3.1.2
Electricity provision .......................................................................................... 24
3.1.3
Sanitation ........................................................................................................ 25
3.1.4
Refuse removal ............................................................................................... 25
3.1.5
Telephone facilities .......................................................................................... 26
3.1.6
Education ........................................................................................................ 26
3.1.7
Healthcare ....................................................................................................... 27
3.1.8
Service delivery and the spatial development programmes ............................. 28
3.1.9
Service delivery and social unrest .................................................................... 29
3.2
Estimating the ‘social wage’ ............................................................................... 31
4. Poverty reduction ............................................................................................. 33
4.1
Overview ............................................................................................................... 33
4.2
Examples of major and interventions................................................................. 37
4.2.1
Public works .................................................................................................... 37
4.2.2
Integrated Food Security Strategy ................................................................... 38
4.2.3
Early childhood development programmes ...................................................... 39
5. CONCLUSION ................................................................................................... 40
Endnotes ................................................................................................................. 43
Executive Summary
More than a decade after the promulgation of the South African population policy, poverty
remains the single most pressing socio-economic challenge facing South Africa. By one
measure, in 2005/06 no less than 48% of the South African population, and 56% of the
African population, was living in poverty.
The purpose of this paper is to present an overview of what is known regarding trends in the
nature and extent of poverty and inequality since 1994. The paper also examines the impact of
the state’s attempts to alleviate both poverty and inequality through its various mechanisms.
Poverty is a complex and multidimensional phenomenon that encompasses many aspects,
including the occurrence of hunger, unemployment, exploitation, poor education and limited
access to essential services such as clean water, sanitation and healthcare. For many, the
experience of poverty also includes homelessness and extreme vulnerability to crises. For the
purpose of this paper, poverty and inequality are understood both narrowly – that is, in terms
of household-based monetary measures – and more broadly, whereby access to services and
the social dimensions of poverty are also taken into account. Moreover, attention is given to
significant sub-populations, including the population group and gender dimension. Due to the
country’s history, Africans account for a disproportionate share of the poor, an African, in
particular an African woman, is more likely to be poor than a member of any other population
group.
The key message emanating from the paper is that in the post-apartheid period – roughly
coinciding with the period since the introduction of the Population Policy for South Africa –
South Africa has made significant inroads into the various dimensions of poverty. This is
despite two major handicaps: first, enormous infrastructure and service delivery backlogs
inherited from the apartheid regime; and second, adverse economic trends that began during
the apartheid era and have carried on well into the post-apartheid period. While some of these
economic conditions have improved over the past few years, the success achieved in terms of
poverty reduction has been largely related to the scale and effectiveness of compensatory
measures – not least social grants – and state investments to address the infrastructure and
service delivery backlogs.
i
Various studies seem to indicate that poverty, when measured in terms of income, has
decreased steadily since 2001. This is corroborated by improvements in nutritional status after
2001, shown by a 15% decline in the proportion of households reporting that a child went
hungry from 31% in 2002 to 16% in 2006. Real income for all race groups also increased
since the mid-1990s, especially since 2002. For Coloureds and Africans in particular, this
increase in income has largely been the product of the rapid expansion of the country’s social
safety net system and can be ascribed to the doubling of social grant payments between the
financial years of 2000/01 and 2005/06.
Positive and impressive improvements also occurred as far as service and infrastructure
delivery were concerned. The share of households with access to basic services, such as piped
water, decent sanitation and electricity for lighting, increased and, in some cases significantly
so, between the 1996 and 2001 censuses, indicative of the remarkable shift of fiscal resources
toward poor households. No doubt, for many people, these improvements have been slower
and patchier than they would have liked, but an objective view suggests that the achievements
have been appreciable.
The manner in which the state has addressed income poverty is, perhaps not how it would
have wished to; that is, in 1994, the new government did not set out to create such a large
social safety net. Nor does it relish the idea that this might remain one of the main tools to
keep poverty at bay on a large scale. Unfortunately, making the labour market more inclusive
is a medium-to-long-term process over which government has very incomplete control.
Moreover, its efforts to engineer employment creation through developmental initiatives have
not touched huge numbers of people. In order to use such initiatives to address poverty on a
more significant scale, it will have to improve its performance in respect of these initiatives,
before or while it scales them up.
Possibly, the biggest disappointment of the post-apartheid era is the persistence of inequality.
South Africa remains a highly unequal society, with the wealthiest 10% of the population
earning 50% of the total household income; the poorest 40% of the population accounts for
less than 7% of the household income and the poorest 20% of the population accounts for less
than 1.5% of the total income. While the trends are more positive if one takes social transfers
and tax incidence into account, the reality is that the economic change that South Africa is
experiencing has the tendency to sharpen rather than dull inequality. In light of the
ii
stubbornness of inequality, the importance of improved service delivery and compensatory
measures such as public works is all the greater.
Generally, the recommendation is to build on existing trajectories, that is, in terms of
improving the reach of initiatives already in place and, beyond this, to ensure that they are
optimally focused on vulnerable groups, such as women, youth and rural dwellers. Whether
developmental initiatives can take a larger place adjacent to social grants and pro-poor service
delivery is perhaps the biggest challenge. The Accelerated Shared Growth Initiative for South
Africa (ASGISA), and the Second Economy Strategy that falls under it, are key in this regard;
the current thinking is that more attention needs to be placed on scalable, programmatic
interventions rather than project-based initiatives, and this is almost certainly correct.
Whether this host of interventions is sustainable is not really the question; fiscally there is
little doubt that they are, albeit as a second preference relative to rapidly increasing
employment via a growing formal economy. The real question is whether South Africa can
afford not to pursue these measures more aggressively and to greater effect. Given the stakes,
it is clear that government must carry on its present course, but given the costs involved and
the scourge of inequality, proper targeting becomes all the more critical.
iii
1.
Introduction
More than a decade after the promulgation of the South African population policy, poverty
remains the single most pressing socio-economic challenge facing South Africa. Using a
poverty line of R322 (in 2000 prices), at least 48% of the South African population, and 56%
of the African population, was living in poverty in 2005/06.
1
Moreover, various measures
reveal that, in terms of income, South Africa is still one of the most unequal countries in the
world. The new democratic order established in 1994 inherited vast inequalities in access to
healthcare, education and basic services such as safe water, sanitation and housing. Despite
the state’s attempts to address inequality, its prevalence is still marked and largely defined
along gender, age, race and spatial dimensions. For example, in as late as 2005 poverty was
still virtually non-existent amongst Whites. By contrast, poor South Africans tend to be
female, African and rural.
Poverty alleviation was at the core of the post-apartheid’s government’s priorities and the
initial planning framework through which this was to be achieved was the Reconstruction and
Development Programme (RDP). The central goal of the RDP was to improve the quality of
life of all South Africans while prioritising the poor. Over the past fifteen years, government
has implemented several poverty alleviation measures aimed at achieving a better quality and
dignified life for all South Africans. Government’s recognition of poverty as its foremost
concern is reflected in the ever increasing prominence of social spending in national budgets,
emphasis on poverty relief programmes and in the focusing of many development initiatives
on those deemed most vulnerable. 2 Expenditure on key social services has remained high and
increased even after the adoption of the Growth, Employment and Redistribution (GEAR)
strategy. 3
The purpose of this paper is to explore the changes in the nature and extent of poverty and
inequality since the Population Policy was first promulgated in 1998. It also examines the
impact of the state’s attempts to alleviate the devastating impact of both poverty and
inequality through its various mechanisms. As far as possible, this first section draws on
official statistical data to describe trends in poverty and inequality, as well as changes in
access to education, healthcare and other services.
1
The structure of the paper is as follows. Section 2 presents recent evidence on trends in
monetary measures of poverty and inequality, to some extent drawing out distinctions based
on population group, gender and geography. Section 3 delves into the question of service
delivery, seeking to trace trends in service poverty and inequality over time, concluding with
a discussion of the ‘social wage’, through which the aggregate value of these services is
measured and implications for poverty and inequality discussed. Section 4 presents an
overview of government’s anti-poverty measures, and then picks out a few key interventions
for particular attention. Section 5 concludes. The balance of this introductory section attempts
to ‘set the scene’ by means of identifying the overall philosophy of the Department of Social
Development in respect of the nexus between population policy and poverty reduction. This is
demonstrated by giving an overview of the recent economic trends that form and affect
poverty and inequality and the context in which government policy operates. Then some
definitions of key concepts used in the paper are provided. Finally, the sum of the main data
sources drawn upon for the analysis that follows are identified.
1.1
Population and poverty
The current development paradigm places the population at the centre of development, and
people as the driving force and ultimate beneficiaries of development. This focus is reflected
in the Population Policy for South Africa, 4 for which the core concept is sustainable human
development. By means of a process of development that is participatory and responsive,
sustainable human development seeks to create a society characterised by a high and
equitable quality of life for all South Africans. The focus on poverty and development is
therefore aimed at enriching people’s lives by providing them with increased options and
enhanced choices.
Low levels of socio-economic development are typically associated with high rates of
fertility, mortality and population growth. Much evidence suggests that development leads to
lower fertility and mortality rates, lower rates of population growth and, ultimately, to people
living longer, healthier lives. It is also generally agreed that improved education,
empowerment of women, and greater wealth are also associated with better health and lower
fertility. However, despite the acknowledged need to improve education, reduce gender
inequality and effect other positive changes, the Population Policy recognises that South
Africa is one of the few countries in the world where widespread poverty persists, despite
fertility rates already having been reduced. This implies a failure to reap the potential benefits
of the ‘demographic dividend’, whereby a country takes advantage of the change in age2
structure that immediately follows a decline in fertility, at which point the working-age
population grows relative to younger and older dependents. 5 The ability to reap the benefits
of the demographic transition is mediated by society’s ability to harness its potential through
requisite social investments like education, health, but above all the capacity to be absorbed
into gainful employment. This fact underlines the central tenet of the Population Policy,
which is that promoting sustainable human development must take precedence over attempts
to manipulate mortality, fertility or migration, but that efforts must be sharpened to take
advantage of demographic changes that occur spontaneously.
A number of strategies have been proposed to promote the objectives of the Population
Policy. Two refer directly to poverty and employment, namely:
3.5.10
Reducing poverty and socio-economic inequalities through meeting people’s
basic needs for social security, employment, education, training and housings, as
well as the provision of infrastructure and social facilities and services;
3.5.18
Creating employment-generating growth with a focus on economic opportunities
for young people and women.
1.2
Economic Context
In the final years of the apartheid era, economic growth was weak and failed to match
population growth. During the period 1994 to 2000, the economy grew at an annual rate of
3%, while annual population growth averaged approximately 2.4%. The economic growth
rate subsequently improved, reaching an average of 4.1% per year during the period 2000 to
2006. The period immediately after 2008 is almost certain to be marked by modest rates of
growth. It is important to note that these later periods were accompanied by a continued slow
down in the rate of population growth leading to an increase in average per capita GDP. 6
These figures, however, hide the far-reaching impact of shifts in the structure of the economy
and in the labour market. These structural changes had a peculiar impact on poverty rates and
income distribution. During the past two to three decades, the economy has moved from being
heavily dominated by the primary sector (agriculture and mining) to one with a stronger
service sector (tourism, banking, ICT, etc.). The economy also diversified from a strong focus
on exporting primary products to a more varied economy requiring labour that is more skilled.
The decline in the demand for unskilled workers was rapid. The proportion of the labour force
classified as unskilled fell from 31% in 1995 to 27% in 2002. The structural changes in the
economy initially led to rising unemployment rates and rapidly increasing income among
3
those employed in the upper end of the formal sector. In 2002, the unemployment rate peaked
at 41% of the population of economically active age, or 26% of the economically active
population defined by the ‘narrow’ definition of unemployment. In the absence of a vibrant
informal sector, widespread peasant agricultural production or massive unemployment
benefits, high unemployment rates have meant deepening poverty for many.
As wages make the largest contribution to income (and thus also to income inequality), and
because wages are derived from employment, an improved labour market constitutes one of
the most important vehicles for the reduction of poverty and inequality in the long run.
Although the rise in the unemployment rate until 2002 shows that employment growth did not
keep up with growth in the labour force, the situation has subsequently improved. The official
Labour Force Survey (LFS) shows that approximately 1.7 million jobs were created in the
seven years between 1995 and 2002. In the next four years (between 2002 and 2006) an
additional 1.2 million jobs were created. These statistics belie the claim that the posttransition period has been one of “jobless growth”. 7 However, a large proportion of the jobs
that have been created are in the informal sector or are in the secondary labour market, and as
such tend to be less remunerative and secure than permanent jobs in the formal sector. Since
obtaining formal sector employment largely depends on having skills and education, the
provision of social welfare has become increasingly important to those unable to re-enter the
formal sector labour market. 8
1.3
Definitions
Poverty is a complex and multidimensional phenomenon that can be viewed from a number of
different perspectives. It is associated with hunger, unemployment, exploitation, poor
education and limited access to essential services such as clean water, sanitation and
healthcare. For many, the experience of poverty includes homelessness and extreme
vulnerability to crises. While clearly many of these issues are intimately related to not having
enough money, it is simplistic to ignore the non-income and sometime non-material aspects of
poverty. The poor are not concerned exclusively with adequate incomes and the implied
ability to consume. For the poor, achieving other non-monetary goals such as security,
independence and self-respect may be just as important as being able to buy or access basic
goods and services. 9 Despite this, most studies on poverty tend to focus on money-metric
measures of welfare such as income and expenditure. One of the main reasons for the
emphasis on money metrics is that income and expenditure are easier to measure. However, a
better understanding may be gained when the money metrics are combined with other
4
measures like access to services. Households with superior access to healthcare, education
and social services are, ceteris paribus, better off than households with similar incomes and
more limited access to these resources. The “social wage” approach is to place a monetary
value on such services and to treat access to these services as income. The efficacy of this
approach ultimately depends on the value placed on such services, despite the fact that the
value of such services is contestable.
Although poverty can also be defined as the inability to attain a minimum standard of living
(measured in terms of basic consumption needs or the income required to satisfy these needs),
it is not usually a static condition. Non-poor individuals, households and communities may be
vulnerable to falling into poverty as the result of shocks, crises or other structural constraints.
Ideally, the poverty line – the income level that separates the poor from the non-poor – should
be set at a level of income that enables individuals to achieve certain objectives, such as a
healthy and active life and participation in wider society. 10 The poverty rate is the proportion
of the population that fall below the poverty line. One limitation of the poverty rate as a
measure of the extent of poverty is that it fails to take into account how far below the line the
poor find themselves – someone just shy of the specified level is treated the same as someone
living in absolute poverty. Thus another useful concept is the poverty gap. It is defined as the
amount of money that would be necessary to raise each poor individual to the level of the
poverty line. If these poverty gaps were summed across all poor individuals, and divided by
the sum of money that would be consistent with this number of people being exactly at the
poverty line, one would have the poverty gap ratio or poverty gap index.
If at all possible, society should give all its members equal access to resources and
opportunities. 11 The extent to which this idealised situation is not achieved presents a measure
of inequality in society. Due to measurement issues,
12
we focus primarily on monetary
measures when dealing with inequality, although some attention is also devoted to inequality
in access to services. In the rest of this document comparisons of income and expenditure are
converted into real values, i.e. the value is adjusted for the effects of inflation and pegged to a
given year. When time comparisons are not made, nominal values (i.e. not adjusted for
inflation) are used and are stated as such.
1.4 Data Sources
Data quality and comparability of datasets has been a constant issue when attempting to speak
to the economic changes at household level in post-apartheid South Africa. Currently, in so
5
far as poverty issues are concerned, a range of different official datasets is used to measure
changes and the status quo. Prominent resources include the October Household Surveys
(1995-1999), population censuses (1996 and 2001), the Community Survey of 2007, the
Income and Expenditure Surveys (IES: 1995, 2000 and 2005/06), the bi-annual Labour Force
Surveys (LFS), and the General Household Surveys (GHS: 2002 onwards). All of these
studies are nationally representative and run under the auspices of Statistics South Africa.
For our purposes, one limitation of these official datasets is that no one instrument coincides
exactly with the period during which the Population Policy has been in effect (1999 to 2008).
Neither the official Statistics South Africa surveys, nor the census and census replacement
survey, nor the Income and Expenditure Surveys, coincide precisely with the beginning and
end of this period. Moreover, for various reasons, using mixes of the surveys and censuses
does not prove to be a viable option. At best, making time comparisons reflects the
importance of trading off imperfect data against the prevalent policy imperatives. 13
1.4.1
Population censuses and the Community Survey
Population censuses provide researchers with rich data sets that do not suffer from sampling
error and ancillary sampling problems. However, they do not collect income data in a
particularly reliable manner. Instead of obtaining an exact measure of income, they opt to
classify income into a small number of income bands, typically based on a single question put
to the household respondent. Analysts are then required to derive means, distribution curves,
etc., based on inferences or assumptions. These assumptions can affect the reliability of the
derived poverty and inequality trends.
Two other problems with income data from the census is that it is characterised by a high
number of households reporting zero income or failing to report their income at all. In the
2001 census, almost one quarter of households (23%) reported having “zero income”,
whereas in the 1996 census the figure was 13%. As a result of the high non-response rate and
the high proportion of zero-income households reported in 2001, Statistics South Africa
imputed income values.
1.4.2
Income and Expenditure Surveys (IES)
The Income and Expenditure Surveys have features that make them particularly suitable for
asking questions about trends in poverty. Their samples are large enough to reduce the impact
of sampling error to a minimum, and the fine detail in respect of expenditure on goods and
6
services allows for a precise accounting of total household expenditure. However, questions
have been raised as to the comparability of the 1995 and 2000 data sets. After publishing a
report contrasting the results of these surveys, Statistics South Africa admitted that the two
surveys were not directly comparable. 14 The quality of data is also brought into question by
an implied drop of almost 40% in real household income earnings between the two surveys.
This drop is enormous (it is of similar magnitude to that suffered by the South African
economy during the Great Depression of 1930) but is also at odds with the national account
figures, 15 which point to an increase in real earnings (see Figure 1 below). State revenue in the
form of value added tax, income tax and company tax, provides further evidence that incomes
have increased and that the 2000 IES in particular might not be an accurate reflection of the
South African economic reality.
FIGURE 1: Trend in per capita Gross National Income, 1980 – 2007 (constant 2000 Rand)
30000
25000
20000
15000
Rand per annum
10000
5000
81
83
85
87
89
91
93
95
97
99
01
03
05
07
1980 82
84
86
88 1990 92
94
96
98 2000 02
04
06
Source: South African Reserve Bank, 2008, series KBP6271Y accessed from www.reservebank.co.za.
Where possible, the section dealing with poverty trends will therefore compare the results of
the various IES surveys with support, if feasible, from an unofficial dataset, the All Media and
Products Survey (AMPS). As many of the anomalies arise from the way in which the data
were collected (use of point versus category estimates, focus on households versus household
members, the level of detail solicited from respondents, etc.), time trend comparisons should
be restricted to similar instruments. In other words, estimates of inequality derived from LFS
data should not be compared to estimates derived from the IES or censuses. Fortunately, the
general trends within each stable of instruments are generally consistent.
7
2.
Trends in income monetary poverty and inequality
2.1
Trends in income poverty
Due to data problems discussed above, parallel, yet broken, time periods covering the period
1995-2005/06 (using the IES) and 1996 to 2007 (using census and Community Survey data)
are used to establish poverty trends based on official data from Stats SA. However, these are
then supplemented with evidence from other data sources that allow a more continuous
picture of changes over time. Table 1 (below) presents the changes in the headcount rate and
the poverty gap ratio derived from the IES of 1995 and 2005/06.
TABLE 1: Poverty shifts by population group, 1995 and 2005/06
Population
Group
Headcount Rate
1995
2005/06
Poverty Gap Ratio
1995
2005/06
R322 a month poverty line
African
63.0%
56.3%
31.9%
24.4%
Coloured
39.0%
34.2%
14.7%
13.0%
Asian
4.7%
8.4%
1.0%
2.2%
White
0.5%
0.4%
0.2%
0.1%
Total
52.5%
48.0%
26.0%
20.6%
R174 a month poverty line
African
38.2%
27.1%
14.7%
8.6%
Coloured
14.6%
12.3%
4.1%
3.9%
Asian
0.8%
1.6%
0.1%
1.1%
White
0.2%
0.0%
0.1%
0.0%
Total
30.9%
22.7%
11.8%
7.1%
Source: H. Bhorat and C. van der Westhuizen, 2008, p.3
NOTES:
1. Poverty lines are in 2000 prices.
2. All changes in the values of the headcount rates and the poverty gap ratios between 1995
and 2005/06 are statistically significant at the 95% level.
3. The population in 1995 has been weighted by population weights according to the 1996
Census. Population weights are not available for the 2005/06 dataset and the population has
been weighted by the household weight multiplied by the household size. The 2005/06
weights are based on the 2001 Census.
It is clear from Table 1 above that household poverty, as measured by the proportion of
households falling below a poverty line of R322 per month (in 2000 prices) declined by five
percentage points between 1995 and 2005/06. Similarly, the headcount rate at the lower
poverty line of R174 per month decreased from about 31% to just below 23%. A similar
national trend is indicated by the poverty gap: when using the R174 a month criterion, the
income gap ratio declined from 12% in 1995 to 7% in 2005/06. In this period, not only had
the proportion of the population living in poverty declined by approximately five percent, the
8
average income of the poor had increased by a similar amount. Africans experienced the
biggest decrease in the headcount rate. Using the R322 month criterion, the proportion of
Africans living in poverty dropped by seven percent from 63% to 56%. The Coloured
headcount rate dropped from 39% to 34%, while the proportion of Asians living in poverty
increased from 4% to 8%. A similar trend is evident when using the lower poverty line of
R174 per month.
Despite the relatively large decline in the proportion of Africans living in poverty, the total
number of Africans living in poverty remains high. Furthermore, Africans account for a
disproportionate share of the poor, i.e. an African is more likely to be poor than a member of
any other population group. The breakdown is presented in Table 2. Once again, this data is
based on the IES.
TABLE 2: Shares in population and poverty, by population group
1995
Population
Group
2005/06
Share of poor
Share of poor
Population
share
R322
poverty
line
R174
poverty
line
Population
share
R322
poverty
line
R174
poverty line
African
77.28%
92.74%
95.45%
79.40%
93.22%
95.05%
Coloured
9.31%
6.92%
4.40%
8.79%
6.26%
4.77%
Asian
2.61%
0.23%
0.07%
2.48%
0.44%
0.18%
White
10.79%
0.11%
0.08%
9.23%
0.07%
0.00%
Source: H. Bhorat and C. van der Westhuizen, 2008, p.35
NOTES:
1. Poverty lines are in 2000 prices.
2. The population in 1995 has been weighted by population weights according to the 1996
Census. Population weights are not available for the 2005/06 dataset and the population has
been weighted by the household weight multiplied by the household size. The 2005/06
weights are based on the 2001 Census.
In 1995 and 2005/06 Africans comprised 77% and 79% of the population respectively.
However, for both years, 93% of South Africans who lived below the R322 poverty line were
African as were 95% of South Africans living under the R174 poverty line. By contrast,
Coloureds made up about 9% of the population in 1995 and 2005/06. However, only 4% to
6% (depending on the poverty line used) of the poor were Coloured. Although they
collectively represent about 12% of the country’s population, Asians and Whites account for
less than 1% of the poor in South Africa in either year.
9
There is widespread consensus amongst analysts that poverty actually increased during the
period immediately following the advent of democracy in 1994. Using IES data and the $2 a
day criterion the headcount rate between 1995 and 2000 increased from 32% to 34%. 16 Using
the same criterion, census data also shows a 2% increase in poverty between 1996 and 2001 –
from 26% to 28%. Hence, over the 1995-2000 period, we find that significant increases in the
headcount rate and the poverty gap are evident. Another study focusing on the income of
individuals over the age of 18, found that individual incomes declined by as much as 40%
between 1995 and 2000.
17
Poverty therefore also increased sharply. Different studies of
census data not only identified an increase in income poverty between 1996 and 2001, but
also established that the trend was a continuation of what had been happening since 1991. 18
Using the AMPS data, Figure 2 below represents the poverty headcount rates based on a
poverty line of R250 per month or R3 000 per year in 2000 Rand.
FIGURE 2: Poverty headcount rates based on AMPS data, 1993 to 2006
Source: S. van der Berg et al., 2007, p.20
Note: The poverty line is R3000 per capita per year in 2000 Rand values.
It is clear from Figure 2 (above) that poverty continued to increase until the mid-1990s.
Between the mid-1990s and the turn of the century, poverty rates remained stable prior to a
decline after 2001.
The initial increase in poverty was probably due to a combination of sluggish economic
growth and poor labour market prospects during the second half of the 1990s. However,
the decline in poverty since 2002 is striking. The finding that poverty, when measured in
terms of income, had decreased steadily since 2001 is corroborated by improvements in
nutritional status after that year. A comparison of GHS surveys shows that nutritional
10
outcomes have improved greatly since 2002. The proportion of households reporting
that a child went hungry declined from 31% in 2002 to 16% in 2006. This trend is
illustrated in Figure 3.
FIGURE 3: Percentage of households with children reporting that children went hungry
in the past year, 2002 - 2006
Source: S. van der Berg et al., 2007, p.25
The drop in poverty, particularly since the turn of the century, is driven by upward income
mobility among the poor and among Africans in particular. The improvements in income are,
in turn, the result of a combination of faster economic growth, improving labour market
prospects and increased state spending on social grants. 19
As shown in Figure 4 below, the real income for all race groups has increased since the mid1990s, especially from 2002 onwards. 20 The robust growth in the income of Whites can be
attributed to members of this group maintaining a constant share of wage income despite their
declining population numbers as well as increases in their income from property. For similar
reasons, the income of Indians climbed steadily throughout the post-transition period. Given
their relatively high education levels, the modest acceleration of economic growth rates in the
decade immediately after the transition coupled to the increased demand for skilled workers,
was of particular benefit to these groups. While the situation of African and Coloured people
also improved in this period, these improvements have largely been the product of the rapid
expansion of South Africa’s social safety net system. Much of the increase in real income
among these groups can be ascribed to the doubling of social grant payments between the
financial years of 2000/01 and 2005/06. 21
11
FIGURE 4: Per capita income by population group from AMPS, 1993 to 2006
Source: S. van der Berg et al., 2007, p.18
12
2.1.1
Spatial dimensions of poverty
Table 3 (below) presents the headcount rate and poverty gap rate for rural and urban areas in
1995 and 2000. Using the $2 per day criterion, the table indicates that the proportion of urban
households classified as poor increased from 13% to 16% between 1996 and 2001, and from
15% to 18% between 1995 and 2000.
TABLE 3: Changes in poverty between 1995 and 2000 22
Poverty line
Urban
Rural
Statistic
Headcount index
Poverty gap ratio
Headcount index
Poverty gap ratio
$2/day poverty line
(R174)
Lower bound
poverty line
(R322)
1995
2000
1995
2000
15%
5%
45%
16%
18%
5%
55%
22%
36%
14%
75%
37%
40%
16%
80%
44%
Source: J. Hoogeveen and B. Ozler, 2006, p.65
The proportion of the rural population living in poverty increased from 45% to 46% on a $2
per day poverty line between 1996 and 2001. The IES shows an even more marked increase in
poverty (from 45% to 55% of the rural population) between 1995 and 2000.
It is important to note that census data show that the proportion of rural poor in overall
poverty is declining. This decline is largely due to rapid urbanisation. While the rural poor
accounted for 62% of all poor households in 1996, five years later this had declined to 56%.
This decline (when read in conjunction with the increase in the proportion of poor living in
poverty) indicates the rapidity of migration to urban areas – a process that will continue to
shape the spatial nature of poverty in South Africa. The process of urbanisation and the
parallel movement of people from poorer to richer provinces present challenges for wealthier
provinces and their cities. 23 The metropolitan areas in particular are confronted with having to
cater for a massive influx of migrants while attempting to reduce poverty levels and improve
services to large resident populations.
Table 4 below presents the headcount rate and poverty gap for each of the nine provinces by
both poverty lines in 1995 and 2005/06, based on data drawn from the IESs. During this
period, the provincial headcount rate of poverty decreased by between 4% and 6%, depending
on which threshold is used. However, the provinces display vastly different experiences. The
two most urbanised provinces, Gauteng and the Western Cape, have headcount rates and
13
poverty gaps well below the national average. Furthermore, despite heavy in-migration,
neither province exhibited a marked shift in poverty levels during the decade in question. For
the R322 per month poverty line the headcount rate for the Western Cape declined by 0.7%,
whereas it actually increased by 6% for Gauteng. Using the R174 per month poverty line, the
headcount rate for both provinces increased, albeit marginally. Despite these increases, the
two provinces still have the lowest poverty rates. The slight increase in poverty in these
provinces may be ascribed to an influx of poor into these provinces.
Alongside Gauteng and the Western Cape, the Free State is the only province to have a lower
poverty headcount rate than the national average. The Eastern Cape, North West and the Free
State provinces experienced large decreases in the headcount rate. By contrast both Limpopo
and KwaZulu-Natal experienced small increases in the headcount rates during the same
period.
TABLE 4: Poverty shifts by province, 1995 and 2005/06
Province
Headcount Rate
1995
2005/06
Poverty Gap Ratio
1995
2005/06
10.1%
39.9%
27.7%
36.4%
27.3%
33.4%
6.8%
28.9%
33.5%
26.0%
10.1%
26.2%
23.7%
15.6%
28.4%
21.7%
8.2%
24.7%
29.3%
20.6%
2.4%
19.9%
11.3%
19.2%
11.7%
15.3%
2.0%
11.7%
16.4%
11.8%
2.8%
8.5%
8.5%
4.3%
11.4%
7.8%
1.7%
9.2%
10.6%
7.1%
R322 a month poverty line
Western Cape
Eastern Cape
Northern Cape
Free State
Kwazulu-Natal
North West
Gauteng
Mpumalanga
Limpopo
Total
29.1%
72.9%
58.3%
65.0%
56.8%
65.0%
19.1%
61.0%
63.9%
52.5%
28.5%
58.9%
53.3%
41.6%
60.1%
51.0%
25.6%
55.0%
64.7%
48.0%
R174 a month poverty line
Western Cape
Eastern Cape
Northern Cape
Free State
Kwazulu-Natal
North West
Gauteng
Mpumalanga
Limpopo
Total
9.4%
49.8%
33.9%
45.3%
31.5%
40.5%
6.5%
34.2%
40.5%
30.9%
9.5%
29.0%
28.2%
15.6%
33.1%
23.8%
6.7%
28.1%
34.0%
22.7%
Source: H. Bhorat and C. van der Westhuizen, 2008, p.35
The changes in the provincial headcount rates using the R174 per month criterion are similar
to those observed for the R322 line. Particularly large decreases in poverty headcount rates
are observed in the Free State, and to a lesser extent the Eastern Cape and North West. In
14
contrast to the R322 line, poverty in Limpopo did decline, albeit slightly. With the exception
of the Western Cape at both poverty lines, and Gauteng at the R322 line, all provinces
experienced a decline in the severity of poverty as measured by the poverty gap ratio. The
lack of improvement in Western Cape and Gauteng almost certainly is due to the fact that they
were relatively non-poor in 1995, but the failure to improve further could be due to influxes
of relatively poor people over this period.
The provincial proportion of population is presented in Table 5. Not surprisingly, both
Gauteng and the Western Cape display large discrepancies between their shares of population
and the number of residents living in poverty. It is noticeable that close to two thirds of all the
poor, using either poverty line, are located in the three provinces with large rural populations,
namely KwaZulu-Natal, Eastern Cape and Limpopo.
TABLE 5: Shares in population and poverty, by province
1995
2005/06
Share of poor
Province
Western Cape
Eastern Cape
Northern Cape
Free State
Kwazulu-Natal
North West
Gauteng
Mpumalanga
Limpopo
Share of poor
Population
share
R322
poverty
line
R174
poverty
line
Population
share
R322
poverty
line
R174
poverty
line
10.1%
15.9%
2.1%
6.4%
20.4%
8.6%
17.1%
6.9%
12.5%
5.6%
22.0%
2.4%
7.9%
22.0%
10.6%
6.2%
8.1%
15.2%
3.1%
25.6%
2.3%
9.4%
20.7%
11.2%
3.6%
7.7%
16.4%
10.0%
14.5%
2.4%
6.2%
20.9%
7.0%
20.2%
7.4%
11.3%
5.9%
17.7%
2.7%
5.4%
26.5%
7.4%
10.8%
8.5%
15.3%
4.2%
18.5%
3.0%
4.3%
30.6%
7.3%
5.9%
9.2%
17.0%
Source: H. Bhorat and C. van der Westhuizen, 2008, p.35
2.1.2
Poverty shifts by gender of household head
Table 6 (below) shows that both male and female-headed households experienced a decline in
poverty levels as measured by both poverty lines. Using the upper poverty line, the decline in
the headcount rate was larger for male-headed than for female-headed households. The
situation is, however, reversed when using the lower poverty line. Nevertheless, poverty
remains a disproportionately female phenomenon as can be seen from the consistently higher
headcount rates for female-headed households.
15
TABLE 6: Poverty shifts by gender of household head, 1995 and 2005/06
Headcount Rate
Poverty Gap Ratio
Household Head
1995
2005/06
1995
2005/06
22.2%
33.5%
31.3%
16.0%
26.6%
20.6%
9.8%
15.6%
16.1%
5.4%
9.5%
7.1%
R322 a month poverty line
Male-headed
Female-headed
Total
45.8%
65.6%
58.0%
38.3%
60.6%
48.0%
R174 a month poverty line
Male-headed
Female-headed
Total
26.1%
40.3%
37.9%
17.1%
29.9%
22.7%
Source: H. Bhorat and C. van der Westhuizen, 2008, p.6
This is supported by the observation from Table 7 below, which shows that individuals living
in female-headed households continue to account for a disproportionately large share of the
poor. Despite the fact that only 43% of households were headed by women in 2005/06,
between 55% and 57% of the poor lived in female-headed households. Unfortunately, the
concentration of poverty in female-headed households was also accompanied by an increase
in the proportion of households that were headed by women. Between 1995 and 2005/06, the
proportion of households headed by women increased from 34% to 43%. This underlines the
importance of the Child Support Grant, which tends to benefit female-headed households
because women household heads are often primary caregivers for dependent children.
TABLE 7: Shares in population and poverty, by gender of household head, 1995 and 2005/06
1995
Household
Head
Male-headed
Female-headed
2005/06
Share of poor
Share of poor
Population
share
R322
poverty
line
R174
poverty
line
Population
share
R322
poverty
line
R174
poverty
line
66.2%
33.8%
57.7%
42.3%
55.9%
44.1%
56.5%
43.4%
45.1%
54.8%
42.6%
57.3%
Source: H. Bhorat and C. van der Westhuizen, 2008, p.35
2.2
Trends in income inequality
The recent high rates of economic growth have had an ambiguous effect on income
inequality. On the one hand, job creation has started to keep apace with growth in the
population of economically active age. While economic growth undermines income
inequality by ensuring that some people excluded from the economy can now earn wages and
salaries, the relationship between growth and inequality is complex. For example, greater
16
competition for skilled workers contributes to upward pressure on the wages of skilled
workers can demand, which aggravates existing levels of inequality. While economic growth
may have a net effect of increasing income inequality, it has the ancillary effect of increasing
tax revenue accruing to the state. This, in turn, equips the state to increase grants to the poor
while expanding services to them. Economic growth thus has a complex correlation to
poverty, especially when the latter is understood in broad terms.
2.2.1
Share of consumption expenditure by population group
Between 2000 and 2005/06, 24 black African households’ share of consumption expenditure
rose slightly from 42.9% to 44.3% of total consumption – an increase of 1.4%. At the same
time Africans’ share of the population rose by 1.1% (from 78.3% to 79.4%), indicating that
Africans, on average, had become slightly more affluent over the period. White households’
share of consumption expenditure fell by a similar amount, i.e. 44.1% in 2000 to 42.9% in
2005/06. This drop was slightly faster than that of their population share, which declined from
10.1% to 9.2%. Thus, Africans are growing relatively more affluent, and Whites poorer, while
the share of expenditure by both Coloured and Indian/Asian population groups has remained
static. 25 These relative changes in aggregate expenditure by population group, however, are
silent as to how that consumption is distributed within each of the groups in question.
2.2.2
Income and inequality
South Africa continues to be a highly unequal society, with the wealthiest 10% of the
population earning more that 50% of total household income. The poorest 40% of the
population accounts for less than 7% of household income, while the poorest 20% of the
population accounts for less than 1.5% of total income. 26 As indicated above, mean real per
capita income increased between 2000 and 2005/06. How this increase affected different
economic and social groups can be seen by examining each income decile. An income decile
represents one of ten (increasingly wealthy) tranches of the population, with each tranche
containing 10% of the population. During this period, above-average increases in income are
evident in deciles 1, 2, 3 and 10 (i.e. among the poorest 30% and the most affluent 10% of the
population). 27 Below-average increases occurred in deciles 4 to 9. 28 The income profiles are
heavily dominated by income from work, including employment, self-employment and
business income. However, the importance of social grants as a source of income is
increasingly evident among lower-income households. These grants are central to the increase
in incomes shown among the poorest 10% of the population. The differing population
composition of the deciles illustrates the extent to which racial disparities in income persist.
17
While virtually all members of the poorest decile (93%) are African, Whites and Africans,
respectively, accounted for 73% and 17% of the wealthiest decile. The deciles per population
group are presented in Table 8 below.
TABLE 8: Population group shares of total household income per decile
Decile
African
Coloured
Indian /
Asian
White
Total
1
93.2%
3.2%
0.5%
3.0%
100%
2
94.2%
4.0%
0.8%
1.0%
100%
3
93.0%
5.4%
0.4%
1.1%
100%
4
90.3%
7.9%
0.8%
1.0%
100%
5
83.6%
12.0%
2.6%
1.7%
100%
6
78.7%
16.0%
2.7%
2.6%
100%
7
78.7%
13.6%
2.4%
5.0%
100%
8
63.7%
12.9%
7.0%
16.1%
100%
9
47.8%
11.4%
6.8%
33.8%
100%
10
17.0%
5.5%
4.7%
72.7%
100%
Total
41.2%
8.6%
4.8%
45.3%
100%
Share of all
households
79.4%
8.8%
2.5%
9.2%
100%
Source: Statistics South Africa, 2008a
2.2.3
Gini coefficient
However, expressing the level of inequality in terms of deciles in not conducive to making
sectoral or even international comparisons. A more convenient measure is income inequality
is the widely used Gini coefficient. The lower the value of the Gini coefficient, the more
equally household income is distributed. A Gini of 0 denotes perfect equality (all individuals
in the population receive the same income), while a Gini of 1 denotes perfect inequality (one
individual in the population has everything). Unfortunately, the various datasets point to
different income profiles and thus several estimates of the Gini coefficient are presented.
Nonetheless, as a rule, the trends within individual datasets (e.g. comparing census
information, IES datasets, etc.) are consistent, although the absolute levels may differ.
Despite changes in terms of increasing income, particularly among the poorest deciles, the
available data suggest that South Africa experienced a sharp rise in income inequality
between 1995 and 2005/06. During this period, the Gini coefficient (drawn from IES data)
rose from 0.64 in 1995 to 0.69 in 2005/06 (Table 9). 29 This increase is of particular concern
given that South Africa was already one of the most unequal societies in the world to start off
with. The latest Human Development Report from the UNDP indicates that only nine other
countries are known to have more severe income inequality. 30 However, the UNDP report
18
bases its Gini estimate of 0.58 on information from 2000 data (presumably on information
derived from the IES). Using census data for 1996 and 2001 suggests that the Gini coefficient
rose markedly from 0.68 to 0.73, 31 indicating an increase in inequality. However, one must
bear in mind that a large number of households also refused to state their income, or indicated
that they had zero income. Despite this, the censuses show that the income inequality rose
across all population groups. By contrast, the IES data indicate that income inequality
increased markedly for all population groups other than Africans. The coefficients are
presented in Table 9 below. Urban and rural inequality rose in tandem. 32
TABLE 9: Inequality shifts by population group: Gini coefficients for 1995 and 2005/06
Category
African
Coloured
Asian
White
Total
1995
2005/06
0.55
0.49
0.45
0.39
0.64
0.56
0.58
0.53
0.45
0.69
Source: H. Bhorat and C. van der Westhuizen, 2008, p.11
Given South Africa's political heritage, inequality between population groups has historically
been a significant contributor to overall inequality. Some sense of this has already been
conveyed by means of the income and consumption shares data presented above. However,
the Theil Index enables a decomposition of overall inequality into its ‘within-group’ and
‘between-group’ constituents, which provides a more nuanced picture of the manner in which
inequality has been changing over time.
In the 1970s, the contribution of between-group inequality to the Theil Index stood at about
60%, 33 which subsequently declined to 43% in 1996 and 40% in 2001. 34
According to the 1998 Income and Inequality Report, between-race inequality accounted for
37% of total inequality in 1995. 35 While the contribution of between-group inequality has
been declining, it has been replaced by significant growth in inequality within the population
groups. Over the past three decades, the within-group inequality has increased by 20%. This
change is primarily driven by rising inequality amongst African households. Inequality
amongst African households accounts for between 29% and 49% of total inequality,
depending on the measure chosen. This is borne out by the high (and rising) Gini coefficient
amongst African households of 0.54.
36
More recently, however, within-group inequality
appears to have declined relative to between-group inequality. By 2005/06 (using IES data)
19
they contributed in almost equal measures to total inequality (50.3% compared to 49.7%). The
data are presented in Table 10 below.
TABLE 10: Inequality within and between groups, 1995 and 2005/06
1995
Within-group component
Between-group component
Total inequality (Theil-T)
2005/06
Value
Share
Value
Share
0.43
0.38
0.81
53.1%
46.9%
100.0%
0.51
0.50
1.01
50.3%
49.7%
100.0%
Source: H. Bhorat and C. van der Westhuizen, 2008
2.3
Discussion
Changes in the labour market remain central to understanding the shifts in income poverty
and inequality. Central to changes in the labour market are the shifts in the types of skills
demanded and a massive increase in unemployment between 1995 and 2000. During this
period, the “broad” unemployment rate increased from 31% to 42%. The “narrow”
unemployment rates (that include only that proportion of the economically active population
who recently sought employment) rose from 18% to 31%. 37 Hence, the data for South Africa
on employment and unemployment trends strongly reinforce the income trends observed
above. During the period 1995-2002, aggregate employment grew by some 1.5 million jobs,
at an average rate of 2.3% per annum. This remains slightly below the economic growth rate
over the period. In turn, however, the labour force grew by some 5.2 million individuals,
resulting in a massive rise in the national unemployment levels from 4.2 million in 1995, to
close to 8 million in 2002. 38 Thus, while the economy did not experience ‘jobless growth’ in
the post-apartheid period, employment absorption was sufficiently poor to result in rising
unemployment rates for all races and both sexes.
During this period, the economy was undergoing dramatic changes – changes that spurned
unskilled labour in favour of certain categories of better skilled workers. Although demand for
skilled workers has been increasing rapidly, many of the skills provided on the market place
have failed to meet employers’ demands. Some studies have reported the worrying spectre of
sharply-rising graduate unemployment, driven primarily by the low demand for the types of
tertiary qualifications on offer. 39
In the post-apartheid period, the proportion of skilled and semi-skilled workers in the
workforce has continued to rise. Unemployment remains concentrated among historically
disadvantaged groups and is particularly high among the rural, female, uneducated, and
youthful of the population. This can be seen in Tables 11 and 12.
20
While it is easy to see how employment offers a bridge out of poverty, it should be realised
that virtually all of the additional jobs created in this period were ‘informal’. As such they are
associated with much lower earnings (and security) than jobs in the formal sector. During the
period 1994–2002, formal employment contracted steadily, falling by an average of 1.25%
each year. This trend was finally broken in late 2003, when growth in formal employment
became positive.
The unemployment rate varies greatly by province. In September 2007, the unemployment
rate was highest in KwaZulu-Natal (30.0%) and lowest in Western Cape (15.7%). In all
provinces there has been a downward trend in unemployment rates since 2003.
TABLE 11: Unemployment rate by province 40
Sept
2001
Sept
2002
Sept
2003
Sept
2004
Sept
2005
Sept
2006
Sept
2007
Western Cape
17.7%
19.6%
19.5%
18.6%
18.9%
15.0%
15.7%
Eastern Cape
31.4%
32.7%
31.7%
29.6%
29.9%
32.0%
26.1%
Northern Cape
25.0%
24.9%
26.4%
24.5%
24.7%
28.7%
26.0%
Free State
27.0%
29.1%
28.0%
28.6%
30.2%
26.5%
25.2%
KwaZulu-Natal
33.8%
35.0%
31.6%
28.7%
32.8%
26.6%
30.0%
North West
28.6%
30.6%
28.4%
28.0%
27.4%
29.7%
24.6%
Gauteng
30.4%
30.5%
27.6%
25.7%
22.8%
23.2%
17.4%
Mpumalanga
29.2%
30.1%
25.3%
24.8%
26.9%
28.0%
22.0%
Limpopo
34.6%
34.1%
31.1%
27.8%
30.1%
32.0%
27.3%
RSA average
29.4%
30.4%
28.0%
26.2%
26.7%
25.5%
22.7%
Province
Source: Statistics South Africa, 2008b
As expected from the income data, there are strong racial dimensions to unemployment. In
September 2007, the unemployment rate among black African men was 23.3% compared to
3.5% among White men (Table 12). Similarly, the unemployment rate among African women
was 30.9%, relative to a figure of 4.2% for White women. Also worth noting is the relatively
high unemployment rate among women compared to men.
21
TABLE 12: Unemployment rate by gender and population group
Sept
2001
Sept
2002
Sept
2003
Sept
2004
Sept
2005
Sept
2006
Sept
2007
Male
Black African
31.5%
31.5%
30.0%
27.6%
26.6%
25.3%
23.3%
Coloured
19.5%
19.9%
18.8%
19.7%
20.6%
16.6%
20.0%
Indian/Asian
15.7%
15.6%
15.5%
12.4%
14.0%
6.6%
7.4%
White
4.7%
5.0%
4.0%
5.1%
3.6%
4.6%
3.5%
Average
25.8%
25.9%
24.7%
23.1%
22.6%
21.2%
19.8%
Female
Black African
40.7%
42.3%
38.7%
36.0%
37.1%
36.4%
30.9%
Coloured
23.1%
26.6%
23.6%
24.1%
24.6%
22.6%
21.3%
Indian/Asian
23.5%
27.1%
18.4%
15.4%
18.6%
14.3%
10.2%
White
7.4%
7.4%
6.2%
5.8%
6.9%
4.4%
4.2%
Average
33.8%
35.9%
32.0%
30.2%
31.7%
30.7%
26.1%
Source: Statistics South Africa, 2008b
While high unemployment is a determinant of poverty in itself, the fact that it is occurring
now is, in part, a result of the fact that the majority African population has now reached the
beginning of the demographic transition. The decline in fertility rates is leading to a relative
increase in the economically active population, i.e. as the sub-population of working-age
adults grows relative to the number of children and elderly dependents, and fewer childrearing responsibilities facilitates women entering or staying in the labour force. In principle,
this represents an historic opportunity – the ‘demographic dividend’ – but in practice, the
opportunity is being lost precisely because only a small share of the youth joining the labour
force in any given year are finding gainful employment. Thus, the urgency of addressing
unemployment and youth unemployment in particular, arises, first for its own sake, but
second, in order to take advantage of the historic opportunity that the current phase of
demographic change has to offer.
3.
Service poverty and inequality, and the ‘social wage’
“Overcoming poverty is not a gesture of charity. It is an act of justice. It is the
protection of a fundamental human right, the right to dignity and a decent life.”
– Former President Nelson Mandela, 2006 41
As has been stated, poverty alleviation, with its central goal of improving the quality of life
for all South Africans, is at the core of the post-apartheid government’s priorities. This has
22
been reflected in budgetary shifts to social spending, in poverty relief programmes and in
development programmes that target those deemed vulnerable, such as women.
42
In
particular, the post-1994 era is notable for the rapid reallocation of resources through fiscal
channels, from wealthy, White households to poor, African households. Broadly, while
approximately 40% of aggregate social spending was directed to Whites and 43% to Africans
in the mid-1980s, by the late 1990s fully 80% of total social spending was assigned to the
African populace and less than 10% to Whites. The results of this fiscal switch are evident in
the data on assets and services. Hence, the share of households with access to basic services
such as piped water, decent sanitation and electricity for lighting increased, and in some cases
increased significantly between the 1996 and 2001 censuses, indicative of the remarkable
shift of fiscal resources toward poor households.
3.1
3.1.1
Access to services
Water provision
Provision of clean water is a major tool for improving the quality of life among the poor. The
1996 census showed that fewer than half (45%) of South African households had a tap inside
their dwelling. In 1998, conditions were still dire, especially for rural households, where only
19% had access to clean piped water. Without access to a source of clean piped water, rural
residents (usually women and children) could be forced to spend three or more hours a day
fetching water. 43
However, by 2001, water provision had improved considerably and the proportion of
households with access to clean water (on their property or from a communal tap) had
increased to 85%, while the proportion of households with piped water inside their dwelling
or in the yard had increased to 61%. 44 Much of the increase in access to piped water was the
result of the mass installation of public taps. This reduced household reliance on watercarriers, tankers, boreholes, springs and rivers. 45
In 2006, three quarters of all households received piped water from their local municipality.
Over one third of those who did so, did not pay for the service and relied entirely on free
water provision. Seventy three per cent (73%) of households met the slightly more rigorous
RDP criterion that the water must be piped to within 200 meters of their dwelling. 46
23
3.1.2
Electricity provision
Wood is used to meet the heating needs of one in five South African households, while more
than one-third of the country’s population uses wood as an energy source. Poor rural
households make extensive use of wood particularly for cooking and heating. Although wood
is often free to rural households, a significant cost is incurred in terms of time and effort spent
collecting and transporting it. Furthermore, when compared to other fuels, wood is
significantly unhealthier for household use in terms of the chemicals released on burning. As
with fetching water, one finds that the burden of collecting and carrying wood falls
particularly on women and children. 47
Over the past decade and a half, however, there has been a marked increase in the number as
well as the proportion of households with access to electricity. The proportion of households
with electricity increased from under one third of households before 1994, to 70% of
households in 2001. In 2006 almost 80% of households had access to electricity. The increase
in access to mains electricity implies a significant improvement in the living conditions of
poor households. It certainly indicates that less effort is spent on household chores and
exposure to more hazardous energy sources is reduced. The same percentage of households
used paraffin for cooking in 2001 as did in 1996 (about 21%). However, this proportion
dropped to approximately 16% in 2006. Households that used gas for cooking also decreased
from 3.2% in 1996 to 2.5% in 2001 and 2.2% in 2006.
TABLE 13: Main household energy sources for cooking, heating and lighting, 2006
Source
Cooking
Heating
Lighting
Electricity
63.5%
49.9%
81.2%
Gas
2.2%
0.9%
0.1%
Paraffin
16.1%
13.7%
3.5%
Wood
15.5%
20.3%
NA
Coal
2.1%
4.6%
NA
NA
NA
14.8%
Animal dung
0.0%
0.4%
NA
Solar energy
0.0%
0.1%
0.2%
Other
0.3%
10.1%
0.2%
Candles
Source: Own calculation from GHS 2006
Although the free basic allocation of electricity is insufficient to meet all the demands of the
typical household, about 20% of households who now receive electricity rely primarily on this
24
allowance. Consequently, the major impact of the free basic allocation is in terms of improved
lighting rather than in terms of cooking and heating. In 2001, only half of households
dependent on the free basic allocation used electricity for cooking or heating.
48
In 2006,
virtually all households that received electricity relied on this energy source for lighting.
However, only 80% of those with access to main electricity used the source as their main
energy source for cooking. Sixty percent of households with mains electricity used it as their
main means of heating.
3.1.3
Sanitation
Providing decent sanitation for households has long been, and remains, a key developmental
challenge because of the obvious health implications.
49
In 1998, as many as 32% of
households used pit latrines as a toilet, while 12% had no toilet facility at all. 50 During this
same period, only 11% of rural households had access to a flush or improved latrine. 51 The
vast majority of rural households then relied on pit latrines, of which only 20% had been
improved to an acceptable, hygienic standard. Nineteen percent of rural households had no
toilet at all. 52
In 2006, two thirds of the country’s households had a sanitation/toilet service that met RDP
standards – a VIP, flush toilet or chemical toilet. 53 Almost a quarter of all households (23%)
use a non-improved pit latrine.
The percentage of households that used bucket latrines declined slightly from 4.6% in 1996 to
4.1% in 2001. 54 Again, the provision of adequate sanitation in urban areas far exceeds that in
rural areas. The percentage of households that used flush or chemical toilets in urban areas
increased slightly to 54% in 2001 while the percentage of households that used pit latrines
decreased from 32% in 1996 to 29% in 2001. 55
3.1.4
Refuse removal
Refuse removal poses special challenges in rural and semi-rural areas. In general, there was
some improvement in service levels after 1996. The percentage of households that reported
that their refuse was removed on a weekly basis by local authorities increased from 51% in
1996 to 55% in 2001. 56 In 2006, 59% of households had their refuse removed at least once a
week by the local authority. 57
25
3.1.5
Telephone facilities
The RDP initially proposed that all households would have access to affordable
communication services. An indicator of this service level can be taken as the presence of a
fixed line telephone in one’s dwelling. Obviously, access to cellular phones increased across
all income groups since their introduction in the late 1990s. However, this service is
significantly more costly than a fixed line service and may not meet the standard set by the
RDP. 58 In both 1996 and 2001, the percentage of households that had no access to a fixed line
telephone remained unchanged at 19%. 59 However, the percentage of households that had a
cellular phone or a fixed telephone in their dwelling increased from 29% in 1996 to 42% in
2001. 60 By 2006 almost two-thirds of the country’s households had access to a cellular phone,
which is three times more than the number of those who have access to a fixed line phone.
Conversely, the percentage of households that relied on public telephones, increased slightly
from 36% in 1996 to 39% in 2001. 61
3.1.6
Education
Between 1995 and 2000, net enrolment in primary education in South Africa remained high at
95%, comparable to many developed countries, e.g. the United States (95%) and Switzerland
(94%). However, in 1996, almost 80% of adult South Africans had not matriculated. For the
African population group, the figure was 87%, while for Coloureds, Indians/Asian, and
Whites, the figures were 85%, 65%, and 44%, respectively. 62 Approximately half of all South
Africans that had studied beyond the matriculation level were from the White population
group. The great disparity in the quality of education for the different population groups
ensured that government policies after 1994 focused on transforming the education system, in
fact, making expenditure on education the nation’s largest budgetary item.
Various pieces of legislation and policy documents in the field of education have appeared,
for example, the SA Schools Act of 1996 that provides for the development of national norms
and standards. A National Qualifications Framework has been established. There is an
expansion of the Adult Basic Education Programme; a Further Education and Training Act; a
Higher Education Act; and a Skills Development Act. The White Paper on Inclusive
Education (July 2001) provides for a national policy on education, including norms for early
childhood development and for learners with special needs. New policies on school
assessment, and Outcomes Based Education (OBE Curriculum 2005), and a New Academic
Policy for Programmes and Qualifications in Higher Education have been introduced.
Transformation of the higher education system has also been embarked upon.
26
The School Quality Improvement and Development Strategy aims to provide the poorest
quintile of primary schools with resources for effective teaching. The programme is in its first
year but more than 5000 schools have already been provided with books. Another 6000
schools will receive indigenous language materials to improve literacy in the foundation
phase of schooling. The programme also targets teacher training and the development of
mathematics, science and technology skills. Not all provinces have allocated the necessary
funding for the 2007 school year. This is being addressed by the Department of Education.
The Department of Education has also identified the Further Education and Training (FET)
colleges sector to expand educational opportunities, preparing young people for the working
world in response to intermediate and higher-level skills requirements. A R1.9 billion
programme will recapitalise the sector, tackle vital and long-neglected curriculum reform, and
invest in staff training and development, infrastructure and equipment. Partnerships are
actively encouraged to ensure that colleges are responsive to social and industry demand.
Modernised college programmes must encompass industry-based training, and this is being
addressed by the Departments of Education and of Labour.
3.1.7
Healthcare
The main thrust of the Department of Health has been to improve access to healthcare for
those who could not afford it, and it did so through the primary healthcare approach (PHC).
This resulted in an increase in public healthcare expenditure in the period 1995-2003,
although real per capita (uninsured) expenditure has remained between R967 and R907.
Primary healthcare services are available to poor people through the district healthcare
service. Major programmes include the free healthcare for pregnant and lactating women and
children under the age of five as well as for people with disabilities. Although the
implementation of the PHC programme has resulted in increased utilisation rates, indicating
increased access, 63 the per capita PHC visits remained between 1.3 and 2.7, which is still
slightly under that of 3 to 3.5 visits per capita per annum recommended by the World Health
Organisation (WHO) and the Department of Health.
Clinic building and upgrading in previously underserved areas made healthcare a reality for
many previously deprived of these services. In 2003 there were over 4 350 PHC access points
available to the population. In terms of clinics alone, this represents an increase of 701
additional clinics nationally. With regard to the provision of new healthcare infrastructure,
27
1600 clinics have either been built or upgraded since 1995 and 11 new hospitals have been
built since 1998. 64
Expenditure on healthcare and associated infrastructure in all provinces continues to improve,
but under-spending persists in Limpopo (R310 million in 2005/06) even though this was a
14.9% increase on the previous period. There were significant improvements in Mpumalanga
and North West. Year-on-year expenditure increases were strong across the board, with a
16.1% increase to R47.1 billion in 2005/06. Higher spending appears to be related to
improved management capacity. The Hospital Revitalisation Grant to upgrade or replace
infrastructure at government’s 386 hospitals will be R2 billion by 2008/09, comprising around
40% of capital expenditure.
Other healthcare services comprising initiatives to provide access to all South Africans
include programmes such as the Integrated Nutrition Programme to address malnutrition, the
expanded immunisation programme and the Tobacco Control Act. Intensive work is also in
progress on the development of an AIDS vaccine and in the treatment of people living with
HIV/AIDS.
3.1.8
Service delivery and the spatial development programmes
One of the challenges identified several years ago in improving the quality of life of those in
poor communities was the large number of simultaneous activities and different role-players
associated with them. Two programmes were launched in order to improve co-ordination
among these role-players (not at least different spheres of government), so as to properly
focus development efforts particularly in those areas that were most deprived. These two
programmes are the Integrated Sustainable Rural Development Programme (ISRDP) and the
Urban Renewal Programme (URP). The remits of the ISRDP and URP are broadly similar –
both cover service delivery, local economic development and general improvements in
governance. The ISRDP encompasses 13 nodes in all, 12 of which are district municipalities,
and one of which is a local municipality. The URP meanwhile covers eight nodes, most of
which are urban townships.
A baseline survey undertaken in 2006 by the Department of Social Development in the
ISRDP and URP nodes notes some areas of improvement, particularly in terms of service
delivery, but also poverty reduction.
65
The extent to which these improvements can be
attributed to the ISRDP is not altogether clear. However, there is some suggestion that in
28
urban areas, the URP has indeed facilitated the rapid change observed, while for rural areas
the performance of the ISRDP varies:
“The performance at nodal level is very uneven. At one level, there is a discernible
rural/urban difference, where urban municipalities are outperforming their rural
counterparts in providing infrastructure and services to citizens. This should be
addressed by interventions such as Project Consolidate, but the survey found little
evidence of a consistent improvement across rural nodes. At this level, the drop in
poverty is very evident in urban nodes – where poverty levels dropped from an
average of 27.1% in 2001 to 18.2% in 2006. In rural areas, the drop was
considerably less marked, falling from 53.7% in 2001 to 47.8% in 2006. These are
major achievements for which government should be commended, but with a clear
need to bolster delivery of services – infrastructure, grants and so on – in rural
nodes. The true challenge of co-ordination and integration – of government planning
together and providing an integrated set of services to citizens – is in rural areas,
where spatial challenges, the small local tax base and limited economic
opportunities make the situation more urgent and more complex.” 66
If nothing else, there is evidence that the ISRDP is addressing a very real challenge, albeit a
stubborn one.
One of the findings from Department of Social Development’s more recent longitudinal
review of the nodes is that, notwithstanding service delivery challenges in urban areas, those
in rural areas are far more severe; coupled with the relative difficulty of finding employment
in rural areas, the pull of urban areas for rural dwellers is often irresistible. 67 Thus, while
boosting efforts to improve service delivery and spark economic revival in rural areas remains
critical, rural-to-urban migration must be accepted as feature that will be with us for the
foreseeable future, and indeed, given the relatively high costs of servicing rural communities,
it represents a strategic opportunity to provide efficient services to more people as quickly as
possible.
3.1.9
Service delivery and social unrest
How does one reconcile the observations made above about the positive and sometimes
impressive improvements in service provision, with the large number of community protests
over service delivery? Poor service delivery in particular, and specifically the weak
performance of municipal government, was the main focus of the political campaigns leading
29
up to the March 2006 local government elections. However, since then, protests have
continued to sporadically flare up with the common feature that the anger is mostly directed at
municipalities for their ‘failure to deliver’. Ominous predictions that such protests could
spread through some kind of spontaneous contagion have become common. 68 Such a situation
is, in itself, of concern in terms of possible implications for maintaining safety and security
and avoiding the fiscal burden of vandalised municipal offices, but more so for what it implies
about the ‘subjective welfare’ of the nation, the possibility of building social cohesion and
avoiding mass disaffection.
There is reason to doubt that the protests over poor service delivery can be traced in some
clear, direct manner to poor service delivery as such, as something more complex regarding
perceptions and expectations could underlie these outbursts of protestation. Most simply, it
could be that, however good the record of service delivery, it is still short of people’s
expectations. Alternatively, there is some evidence that unhappiness among residents of one
place is a function of the fact that neighbouring communities are benefiting more rapidly.
This implies that the problem is not that service delivery is poor, but that it is uneven in its
success. The evidence from the HSRC’s social attitude survey is that, speaking generally
about discontent with government performance, relative deprivation is more significant than
absolute deprivation (whether in terms of income or access to services). This point
underscores the importance of inequality, in and of itself. Moreover, it is not necessarily those
who are most deprived who are most likely to be acutely dissatisfied. 69
Arguably, the rage over poor service delivery is a misnomer. Also based on the HSRC’s
social attitudes survey data, it appears that, while a high proportion of poor black South
Africans are dissatisfied with their access to services, this is mild in comparison to their
frustration over other things – first and foremost, job creation, second, in respect of crime
prevention, and third, housing. 70 One interpretation of the poor service delivery backlash is
that it is channelling the anguish about unemployment and poor employment prospects. It is
just that the party responsible for poor service delivery is accessible and makes a good target
for public displays of discontent. However, this is not to minimise the importance of
accelerating and improving service delivery, particularly as these are ways of addressing
inequality.
30
3.2
Estimating the ‘social wage’
State policies are geared towards assisting the poor in various ways including direct financial
transfers, education and other grants and the provision of free essential services. While some
‘free’ services may be provided to all households, other support systems tend to be based on
households passing a means test. These means tests are designed to exclude wealthy
individuals and households from benefits. Direct payments like state pensions, childcare and
disability grants are rigorously based on such tests to ensure that only individuals and
households falling below a stipulated income threshold qualify for state support. Housing
subsidies, like pensions and childcare grants, are similarly based on households’ income.
Some state initiatives, although they are intended to benefit the poor, are not subject to a
means test. As a consequence, even wealthy households may receive, at least nominally, the
benefit. An example is the provision of free basic water. While affluent households are often
allocated a lifeline quota of free water, their high consumption levels (in conjunction with
block tariffs) ensure that they pay enough to subsidise their free allocation and often the
allocation made to poorer households. Although providing free services benefits the poor
disproportionately, the net effect of such transfers is to reduce income inequality. However,
demonstrating this effect is often beyond the scope of surveys like the IES.
The 2005/06 IES includes among sources of household income wages, pensions, earnings
from capital and ‘payments-in-kind’. ‘Payments-in-kind’ include imputed values of free water
and electricity, education grants and estimates of the rental value of housing. By incorporating
the value of grants, free services and imputed rent, the IES in essence reflects many
components of the ‘social wage’. By doing this it can be seen what the impact of state
transfers is on poverty and inequality.
The impact of state grants and in-kind payments is particularly evident among poor
households. For example, the poorest 20% of households earned, on average, R4 666 a year in
2005/06. When imputed payments and state transfers are incorporated into earnings (i.e. the
values for grants, housing, free services, etc.) the average household income in this quintile
rises to R7 231. This represents an increase of 55% on what they earned from wages and
economic activity. Over 80% of this increase can be attributed to state transfers in the form of
pensions, childcare and similar grants. The balance is attributable to free and subsidised
services.
31
Among the next highest quintile (i.e. households that are richer than the first quintile but still
part of the poorest 40% of the population) the impact of imputed payments and state transfers
is also pronounced. The average income of households in this quintile is 65% greater than the
average earnings from employment. Whereas the value of state transfers to the poorest
quintile was an average of R2 282 transfers, to the households in the second quintile they
were almost twice as large, R5 630 per annum, despite higher incomes from economic
activities such as employment. Among households in the second quintile, earnings from
wages were almost double that of households in the poorest quintile.
These figures indicate that, among the poorest half of the population, state transfers are
regressive – in other words, the state transfers more to those households that are somewhat
wealthier in terms of earnings from employment and self-employment. Among the wealthier
half of the population, state transfers are both smaller and make up a decreasing proportion of
total household income. The extent to which the state transfers to the richer half of the
population can be considered progressive.
When state pensions, disability and other grants and the value of free services and imputed
rent for subsidised state housing are excluded from consideration, the Gini coefficient of
income in South Africa is 0.72. This represents a situation of profound inequality.
When the in-kind payments for free basic services and state education and housing grants are
incorporated with income (by imputing values), the Gini coefficient should drop as these
services are intended to benefit the poor. In fact, the inclusion of these items has an
insignificant effect on inequality levels. The difference lies only in the third decimal point of
the Gini coefficient. 71
There are three primary reasons for the negligible impact of these transfers on income
inequality. The first is the differential impact of education grants. On the one hand, education
grants and subsidies tend to accrue to households with more members at school and thus
benefit larger (and younger) families. On the other hand, the value of the grant/subsidies is
greatest among those who stay at school longer and further their studies most. While the first
dimension tends to benefit the poor, the second dimension tends to benefit the wealthy. The
net effect is to ensure that education grants have a neutral impact on income inequality. This
contributes to the fact that ‘social wage-adjusted’ Gini differs little from the normal Gini
because the education budget comprises such a large share of the total social wage package.
32
The second reason is due to the fact that when free basic services are provided by a local
authority, they tend to provide the minimum free amount to all households. Thus, both
wealthy and poor households nominally receive the free basic allocation. However, while the
poor may avoid consuming more than the free allocation, wealthier households exceed the
free allocation to the extent that the service provider recoups the cost of their free basic
allocation. The latter dimension is not captured by the IES and wealthy and poor consumers
appear to have similar benefits in terms of free services. Again, the impact of free services as
‘payments-in-kind’ on income inequality is neutral.
Finally, the negligible impact of RDP housing on income distribution lies in the small
proportion of respondents stating they received a state grant for their property (less than 7%
of respondents 72) and the rental imputed for these properties. The average imputed rental on
properties where the household had received state support was only R250 a month. These low
values are the primary reasons why housing and land subsidies do not contribute to a
significant reduction in income inequality.
A greater impact on inequality can be seen in the role played by state pensions, disability
grants and “family and other allowances”. When these are included as income, the Gini
coefficient drops to the more familiar 0.68. A decline of 7% in the level of income inequality
can be directly attributed to these payments and thus to the state. 73 The relatively large impact
of these payments lies in the fact that the benefits are significant and do not accrue to
individuals and households that are not poor. 74
4.
Poverty reduction
4.1
Overview
In 2003 Rand, in 1983 per capita social expenditure (including water service provision) stood
at R2170 per person, whereas by 2003 this figure was R3 451 per person, which represents
close to a 60% increase in total social spending. It should be clear then that a dramatic fiscal
reallocation process took place in the post-apartheid era. Driven by a need to overcome severe
backlogs in these different spheres, there was a sharp refocusing of expenditure towards
potential poverty-alleviating assets and services. 75
33
Analysis suggests that expenditure is well directed in that it reaches its intended beneficiaries.
Between 1995 and 2000, per capita social spending increased by between 21% and 38% for
the 1st and 2nd deciles. In turn, social spending declined on a per capita basis by 9% (6%) for
the 9th (10th) decile. By population group, per capital social spending on Africans increased by
20% between 1995 and 2000, while it declined for all other population groups. Allocation to
rural areas increased by over 30%, while that for urban areas declined. Summary measures of
expenditure indicate that spending on education (particularly school and less so for tertiary
education), healthcare, housing and water are equity-enhancing, while spending on social
security is strongly equity-enhancing. On the latter, for example, the introduction of a new
state transfer, the Child Support Grant, together with very high take-up rates in existing social
grants – most notably the old age pension – has made the state’s social security provision the
most effective anti-poverty intervention. Of total social security expenditure, in 2000, 61%
was allocated to individuals in the 1st and 2nd deciles.
All of these trends suggest that government’s commitment to poverty reduction is very real.
The purpose of this section is to describe various other aspects of government’s strategy to
promote poverty reduction, in particular focusing on those interventions that have not yet
been mentioned, that is, other than improved service delivery and an expanded social grant
system and budget. Our brief exposé of so-called ‘developmental interventions’ does not do
justice to the true scope of these activities, but seeks to identify some of the more significant
ones, as well as to note some of the challenges faced in terms of developmental interventions.
Figure 5 seeks to put these developmental interventions into perspective at least insofar as
expenditure is concerned. While figures that are more current might show a slightly different
picture, in essence, there is an abiding chasm between the big-spending items such as
education, health, social security and infrastructure on the one hand, and the employment and
developmental interventions such as public works, land redistribution, and so forth on the
other. Even if one were to attempt to separate out more carefully the share of education,
health and infrastructure expenditure that targeted the poor, the chasm would remain large.
Very likely too, the relatively low levels of spending on employment and developmental
interventions would reflect the fact that creating an administrative infrastructure for such
activities takes a great deal of time, and also, perhaps, that these activities are intrinsically
complex and difficult.
FIGURE 5: Approximate budgets for poverty reduction, 2004/05 76
34
Source: DBSA, HSRC & UNDP, 2005, Development Report 2005, p.37
NOTES: The figure for land redistribution refers only to expenditure on grants, i.e. it excludes
administration and management; expenditure on land restitution is not reflected because its aims are not
primarily developmental. The figure for poverty alleviation refers only to expenditures budgeted by the
Department of Social Development (designated as ‘Poverty Alleviation and Food Security’), and thus may
be a slight under-statement. ‘CASP’ stands for the Comprehensive Agricultural Support Programme
financed by the Department of Agriculture.
The table that follows (Table 14) is based on a survey of the interventions commissioned by
the Public Service Commission, bearing in mind that actual administrative data on these is
sparse. The table furthermore attempts to summarise beneficiaries in terms of ‘full-time
equivalents’, as to control for the fact that some beneficiaries (for example, the Expanded
Public Works Programme), benefit for less than 12 months.
TABLE 14: Estimated employment creation through poverty reduction projects 77
Approx net annual full-time
equivalents, 2006/07
Programme type
Public works*
Land redistribution
Income generating projects
Total
110 000**
80 000
100 000
290 000
Source: Adapted from Public Service Commission, 2007, p.89
NOTES: *Subsumes EPWP-supported early childhood development as well as home/community-based
care programmes. ** Net of employment that would have been realised through the application of
conventional construction methods, based on evidence from case studies.
A major limitation of the table is that it does not take ‘indirect beneficiaries’ into account, for
example, those who benefit from the infrastructure created, or those who benefit from the
35
services made available, for example, at early childhood development centres supported
through the EPWP. 78 Furthermore, it must be acknowledged that the figures are merely orderof-magnitude estimates. Even so, what is clear is that the overall figure is small: 290 000
people, relative to around 4 million unemployed people (in terms of the narrow definition), or
more than 5 million households living in poverty.
The Public Service Commission study also examined the performance of the projects on the
ground, and interviewed participants and beneficiaries to seek to understand their view on the
significance of these projects. The study concludes as follows:
“The purpose of this study was to provide an objective, comparative assessment of
the different main components of the South African government’s project-based
poverty reduction initiatives. The focus was accordingly on four main programme
types, including public works, land redistribution, income generating projects, and
individual services. The overall finding is that, notwithstanding conspicuous
problems across all of the programme types, the achievements of some of the
constituent projects are notable and the projects have individually, and collectively,
made a modest but meaningful contribution to poverty reduction in the country. The
true significance of these projects is obscured in several ways, not least the
inadequacy of the data about them, but possibly more so the fact that, as observers,
we cherish differing ideas about what constitutes a good project. One conclusion is
that there could be significant gains to be reaped if implementers aligned their
expectations and standards more closely to those of the beneficiaries they are
seeking to help, which among other things means taking on board the rich diversity
of abilities and aspirations. There is every reason to believe that government could
seize on these modest successes and improve upon them and amplify them as a short
and medium-term means of mitigating poverty. …[T]he general finding of this
exercise is that the ‘objective’ circumstances of these types of projects – in
particular their tendency to perform poorly relative to their business plans, should
not eclipse their significance to their beneficiaries, who, the data show, often regard
their participation in these projects as more important than the social grants they
receive or than other benefits received from government. In essence, to the extent
there appears to be a general mood swing in government away from these projects,
the suggestion of this study is that they still have a role to play, arguably even a
larger and more focused role than at present.” 79
36
4.2
Examples of major and interventions
Government has in place a number of major interventions to address poverty and
unemployment in the short-term. Most of these interventions have been developed and refined
over the years, and plans are being put in place to scale most of them up in order to reach
larger numbers of people. One advantage of these particular interventions is their ability to
target those who are most vulnerable, namely, women, youth and rural dwellers.
4.2.1
Public works
The magnitude of our country’s structural unemployment crisis is such that in September
2003, 4.6 million South Africans were unemployed in terms of the strict definition and 8.3
million in terms of the broad definition. In the 16 to 34 age group, 70% of the unemployed
had never worked while 59% of all unemployed people had never worked. It was estimated
that, to reach government’s target of halving unemployment by 2014, 546 000 new jobs
would have to be created each year. 80
Public works programmes can be seen as initiatives that were created in a context of high
unemployment, low skills and a large backlog of public services. Expenditure on public
works programmes has increased almost tenfold since 1998. In 2003, the various separate
public works initiatives were unified under the common umbrella of the Expanded Public
Works Programme (EPWP). The EPWP is a cross-cutting programme, involving all spheres
of government and state-owned enterprises. The idea is that the EPWP will draw significant
numbers of the unemployed into productive work, so that workers gain skills while they work
and increase their capacity to earn an income. The EPWP comprises four main components,
or ‘sectors’:
 Infrastructure – increasing the labour intensity of government-funded infrastructure
projects;
 Environment – creating work opportunities in public environmental improvement
programmes, e.g. the Working for Water Programme;
 Social – creating work opportunities in public social programmes, e.g. Home/
Community Based Care workers and Early Childhood Development centres; and
 Economic
–
including
income
generating
learnership/incubation programmes. 81
37
projects
and
small
enterprise
The EPWP does not have its own special budget for projects, but is funded by earmarking
funds on the budgets of line function departments, provinces and municipalities. In the
infrastructure sector, R15 billion of the conditional infrastructure grants to provinces and
municipalities has been earmarked for the 2004–2009 period. 82 Meanwhile, R4 billion has
been set aside for the environmental sector, and at least R600 million reserved for the social
EPWP programmes. 83
Research indicates that public works programmes vary in their efficiency of transferring
income to the poor, with the average expenditure per worker varying from between R27 242
in Limpopo to R6 515 in the Eastern Cape. In these terms, public works programmes are not
as efficient as social grants in alleviating income poverty. However, unlike the latter, part of
their rationale is to create infrastructure, provide services and impart skills.
4.2.2
Integrated Food Security Strategy
Food insecurity is one of the major indicators linked to poverty and vulnerability. The South
African Constitution (Chapter 2, Section 27.1b) states that every citizen has the right to have
access to sufficient food and water and that the state must take reasonable legislative and
other measures, within its available resources, to ensure that this happens. In response to this
declaration, the National Department of Agriculture’s (DoA) Integrated Food Security
Strategy (IFSS) adopted as its guiding vision the attainment of universal physical, social and
economic access to sufficient, safe and nutritious food for all South Africans to meet their
dietary requirements. In accordance with the Millennium Development Goals (MDGs), the
overarching goal of the strategy is to eradicate hunger, malnutrition and food insecurity by
2015. 84
Even though South Africa is a self-sufficient food producer, and despite the fact that food
security has become a vital government priority, food insecurity remains a substantive
development challenge. As discussed above, national surveys suggest that the incidence of
hunger has declined markedly since around 2001. Unfortunately, recent anthropometric
measures of malnutrition are hard to come by, though the National Income Dynamics Study
should soon correct this.
Food security, however, should not only be seen as a failure of agriculture to produce
sufficient food at the national level but also to consider that it refers to a failure of livelihoods
to guarantee access to sufficient food at the household level.
38
85
Food shortage is often
associated with low income levels or the inability to generate an income. Statistics South
Africa’s 2000 Income and Expenditure Survey shows that 57% of households derive their
main source of income from wages/salaries, followed by social grants (14%) and remittances
(10%). A mere 4% of households reported agriculture as their primary source of income.
Most poor people therefore use whatever cash income they have, such as social grants, to buy
food to improve their food security. 86
Two basic intervention strategies are usually employed to deal with food insecurity. First,
emphasis is placed on stimulating urban and rural household food production, usually in the
form of food gardens and livestock in the case of rural households. This venture more often
than not entails co-operation between different sector departments, with the Departments of
Agriculture and Social Development playing the leading roles. 87 Second, the national school
nutrition programme aims to ensure that school-going children obtain a minimum amount of
nutrition through their schools, while aid is given in the form of food relief, for example, food
parcels, targeting impoverished households and communities. Since the majority of poor
people are particularly dependent on their physical strength as a source of livelihood, the
effects of food insecurity on poor health can be devastating. Again, the long-term effects of
not acting in situations like these can have severe and adverse consequences. Large-scale food
relief can effectively prevent further decline into destitution and chronic poverty. 88
By 2004, the Department of Agriculture’s sub-directorate of Household Food Security and
Poverty Alleviation managed about 60 urban agriculture projects (e.g. food gardens), of
which about one third have been outsourced to non-government organisations (NGOs).
Support, over a maximum of three years, was provided in the form of extension services as
well as funding for some implements and annual inputs. A prerequisite for support was access
to secure tenure. In addition, 416 food security projects were funded in the 2002/03 financial
year as part of the Department of Social Development’s Poverty Relief Programme.
4.2.3
Early childhood development programmes
The early childhood phase from birth to nine years is widely recognised as the most important
phase for every human being. Research shows that the early years are critical for physical,
mental, emotional, social and moral growth and development. This is the time in which a child
acquires concepts, skills and attitudes (e.g. the acquisition of language, perceptual motor skills
required for learning to read and write, basic numeracy concepts and skills, problem solving
39
skills, a love for learning and the establishment and maintenance of relationships) that lay the
foundation for lifelong learning. 89
Giving children the best start in life means ensuring them good health, proper nutrition and
effective early learning. From an environmental perspective, this means safe water, basic
sanitation and protection from violence, abuse, exploitation and discrimination. The National
Programme of Action for Children (2000 and beyond) sets Early Childhood Development
(ECD) as one of the country’s major priorities. Money invested in ensuring children the best
start in life through ECD services yields a meaningful return for children, their families and
taxpayers. 90
The core departments providing or facilitating the provision of ECD services are the
Departments of Health, Education and Social Development. ECD services are holistic,
attending to the child’s health, nutrition, educational and psycho-social development and other
needs. Local municipalities, parents, civil society, ECD practitioners and other stakeholders are
also important participants in providing an integrated service to children. In fact, local
municipalities have a clear constitutional and legislative mandate towards service provision of
ECD services. The Regulations to the Child Care Act, 1983, requires local municipalities to be
involved in ECD facilities, and to give approval of the establishment or continuation of such
facilities as a condition of registration of said facility. Many local municipalities also have bylaws that regulate and monitor ECD centres and other childcare facilities. 91
Government is planning a fourfold expansion of educational and care services for children
under the age of six from 2008 to 2011, and has put an unprecedented R9.7 billion towards
this, plus a R7.9 billion cash injection for Grade R, the reception year for primary school.
5.
CONCLUSION
The key message emanating from this contribution is that in the post-apartheid period –
roughly coinciding with the period since the introduction of the Population Policy for South
Africa – South Africa has made significant inroads into the various dimensions of poverty.
This is despite two major handicaps: first, enormous infrastructure and service delivery
backlogs inherited from the apartheid era, and second, adverse economic trends that began
during the apartheid era but which have carried on well into the post-apartheid period. While
some of these economic conditions have improved over the past few years, the success
achieved in terms of poverty reduction has been largely related to the scale and effectiveness
40
of compensatory measures – not least social grants – and state investments to address the
infrastructure and service delivery backlogs. No doubt, for many people, these improvements
have been slower and patchier than they would have liked, but an objective view suggests that
the achievements have been significant. In terms of service delivery, the biggest
improvements have been in access to water and electricity. Less dramatic improvements are
evident with regard to sanitation, refuse removal and telecommunications. Service delivery
has generally improved more in urban than rural areas, owing to a combination of the higher
costs associated to providing services to rural areas, and the generally weaker state of local
government in predominantly rural areas.
The manner in which the state has addressed income poverty is possibly not how it would
have wished; that is, in 1994, the new government did not set out to create such a large social
safety net. Nor does it relish the idea that this might remain one of the main tools to keep
large-scale poverty at bay. Unfortunately, making the labour market more inclusive is a
medium-to-long-term process over which government has very incomplete control. Moreover,
its efforts to engineer employment creation through developmental initiatives have not
touched huge numbers of people. In order to use such initiatives to address poverty on a more
significant scale, it will have to improve its performance in respect of these initiatives, before
or while it scales them up.
Arguably, the biggest disappointment of the post-apartheid era is the persistence of inequality.
While the trends are more positive if one takes social transfers and tax incidence into account,
the reality is that the economic change that South Africa is experiencing has the tendency to
sharpen rather than dull inequality.
In light of the stubbornness of inequality, the importance of improved service delivery and
compensatory measures, such as public works, is all the greater. Generally, the
recommendation is to build on existing trajectories, that is, in terms of improving the reach of
initiatives already in place and, beyond this, to ensure that they are optimally focused on
vulnerable groups such as women, youth and rural dwellers. Whether developmental
initiatives can take a larger place, adjacent to social grants and pro-poor service delivery, is
perhaps the biggest challenge. The Accelerated Shared Growth Initiative for South Africa
(ASGISA), and the Second Economy Strategy that falls under it, are key in this regard; the
current thinking is that more attention needs to be placed on scalable, programmatic
interventions rather than on project-based interventions, and this is almost certainly correct.
41
Whether this host of interventions is sustainable is not really the question; fiscally there is
little doubt that they are, albeit as a second preference relative to rapidly increasing
employment via a growing formal economy. The real question is whether South Africa can
afford to not pursue these measures more aggressively and to greater effect. Given the stakes,
it is clear that government must carry on its present course, but, given the costs involved and
the scourge of inequality, proper targeting becomes all the more critical.
“Endemic and widespread poverty continues to disfigure the face of our
country. It will always be impossible for us to say that we have fully restored
the dignity of all our people as long as this situation persists. For this reason
the struggle to eradicate poverty has been and will continue to be a
cornerstone of the national effort to build the new South Africa”
– President Thabo Mbeki, 2004 92
42
Endnotes
1
H. Bhorat and C. van der Westhuizen, 2008, “Economic Growth, Poverty and Inequality in
South Africa: The First Decade of Democracy,” paper commissioned by the Presidency.
2
Republic of South Africa, 2004, “Country Report on the International Conference on
Population and Development + 10”, discussion document.
3
Department of Social Development, 2006a, Annual Report for the Department of Social
Development 2005/06 Financial Year, Pretoria: Republic of South Africa.
4
Department of Welfare, 1998, Population Policy for South Africa, Pretoria: Republic of
South Africa.
5
G. McNicoll, 2007, Population and Poverty, Special Issue.
6
S. van der Berg, M. Louw and L. du Toit, 2007, Poverty Trends Since the Transition: What
We Know, Stellenbosch: University of Stellenbosch.
7
Ibid.
8
Ibid.
9
I. Woolard, 2002, “An Overview of Poverty and Inequality in South Africa,” report
prepared for DFID-SA.
10
S. van der Berg et al., 2007.
11
J. May, 2000, “Introduction,” in J. May (ed.), Poverty and Inequality in South Africa:
Meeting the Challenge, London: Zed Books.
12
S. van der Berg et al., 2007.
13
H. Bhorat and R. Kanbur, 2006, “Introduction: Poverty and Well-being in Post-apartheid
South Africa”, in H. Bhorat and R. Kanbur (eds.), Poverty and Policy in Post-apartheid
South Africa, Cape Town: HSRC Press.
14
S. van der Berg et al., 2007.
15
Ibid.
16
H. Bhorat and R. Kanbur, 2006.
17
M. Leibbrandt, J. Levinsohn and J. McCrary, 2005, Incomes in South Africa since the Fall
of Apartheid, NBER Working Paper No. 11384, Cambridge: National Bureau of Economic
Research.
18
Leibbrandt et al. found that income poverty increased between 1996 and 2001, continuing
an earlier trend noted by Whiteford and van Seventer for the period 1991 to 1996. M.
Leibbrandt, L. Poswell, P. Naidoo, and M. Welsh, 2006, “Measuring Recent Changes in
South African Inequality and Poverty Using 1996 and 2001 Census Data,” in H. Bhorat
and R. Kanbur (eds.), Poverty and Policy in Post-apartheid South Africa, Cape Town:
HSRC Press; A. Whiteford and D. van Seventer, 2000, “South Africa’s Changing Income
Distribution in the 1990s,” Journal of Studies in Economics and Econometrics, 24(3): 730.
43
19
S. van der Berg et al., 2007, p.21. Prior to 1994, social grants were allocated on a racial
basis. Since then, government, through the Department of Social Development, has
equalised the Old Age Pensions (OAP), and spread the Child Support Grant (CSG) among
all eligible children. The CSG has now been extended to include all eligible children
younger than 14 years old. Other grants include the war veterans, foster care and disability
grants.
20
H. Bhorat and R. Kanbur, 2006.
21
S. van der Berg et al., 2007, p.17.
22
J. Hoogeveen and B. Ozler, 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.
23
Ibid.
24
The Income and Expenditure Survey of 2005/06 is designated as such because the survey
period lasted from September 2005 through August 2006.
25
Statistics South Africa, 2008a, Income and Expenditure of Households 2005/2006,
Statistical Release P0100, Pretoria: Statistics South Africa.
26
Ibid.
27
Based on income excluding imputed rent.
28
Ibid.
29
These estimates are derived from the various IES surveys.
30
United Nations Development Programme, 2008, Human Development Report 2007/08,
New York: UNDP. Curiously, three of these nine countries border South Africa, namely
Botswana, Lesotho and Namibia.
31
Leibbrandt et al., 2006.
32
H. Bhorat and R. Kanbur, 2006.
33
A. Whiteford & D. van Seventer, 2000.
34
H. Bhorat and R. Kanbur, 2006.
35
J. May, 1998, p.25.
36
Ibid., p.25.
37
In terms of the official definition, a person is regarded as unemployed if they did not work
in the previous week, wants to work, is available to begin work within a week and has
taken active steps to look for employment (or create self-employment) in the previous four
weeks. The official definition is thus synonymous with the ‘narrow’ definition of
unemployment. The ‘expanded’ or ‘broad’ definition is the same as above, except that the
person is not required to have actively sought work in the previous four weeks. The
expanded definition can thus be thought to include ‘discouraged’ work seekers.
38
During this period the population grew at approximately 2.4% per annum. However, the
age structure is such that the population of economically active age is growing at
approximately 4% per annum, significantly faster than the population as a whole.
39
H. Bhorat and R. Kanbur, 2006.
44
40
Statistics South Africa, 2008b, Labour Force Survey, September 2007, Statistical Release
P0210, Pretoria: Statistics South Africa.
41
Nelson Mandela’s speech on receiving the Ambassador of Conscience Award,
Johannesburg, 1 November 2006.
42
Republic of South Africa, 2004.
43
Department of Welfare, 1998, p.12.
44
The Presidency, 2003, Towards a Ten Year Review: Synthesis Report on Implementation of
Government Programmes, Pretoria: Government Communication and Information
Services; p.24.
45
Statistics South Africa and Human Sciences Research Council, 2007, Using the 2001
Census: Approaches to Analysing Data, Pretoria: Statistics South Africa; p.182.
46
Own calculation from GHS 2006.
47
Department of Welfare, 1998, p.12.
48
The Presidency, 2003, p.25.
49
Department of Social Development, 2006b, p.12.
50
Statistics South Africa, 2000, Measuring Poverty in South Africa, Pretoria: Statistics South
Africa; p.66.
51
This refers primarily to ‘ventilation-improved pit’ latrines.
52
Department of Welfare, 1998, p.12.
53
Own calculations based on GHS 2006.
54
Statistics South Africa and Human Sciences Research Council, 2007, p.186.
55
Figures for 2001 are provided because a consistent categorisation of “urban” is not
available for later years.
56
Statistics South Africa and Human Sciences Research Council, 2007, p.187.
57
Own calculations based on GHS 2006.
58
A local call on a cellular phone may cost five times the price of a call of the same duration
made on a fixed line. However, access to a fixed line service usually requires a significant
monthly service fee that makes this option unattractive to the poor.
59
Own calculations based on GHS 2006.
60
Statistics South Africa and Human Sciences Research Council, 2007, p.188.
61
Ibid., p.188.
62
Statistics South Africa, 2001, Education in South Africa: Selected Findings from Census
’96, Pretoria: Statistics South Africa.
63
The Presidency, 2003, p.21.
64
Ibid., 2003, p.21.
65
Department of Social Development, 2006b, “Baseline Survey of the 21 ISRDP and URP
Nodes: Topline Report & Data Tables,” unpublished report.
66
Department of Social Development, 2006b, p.13.
45
67
Department of Social Development, 2008, ‘Building Sustainable Livelihoods’ … and
overview, Pretoria: Republic of South Africa
68
E.g. see the Mail & Guardian, 6-12 July 2007.
69
D. Hemson, 2007, “Winters of Discontent? Attitudes Towards Service Delivery”, draft
book chapter based on the Social Attitudes Survey of South Africa (SASAS), HSRC.
70
B. Roberts, 2005, “The Happy Transition? Attitudes to Poverty and Inequality after a
Decade of Democracy,” in Pillay, U., Roberts, B. and S. Rule, South African Social
Attitudes: Changing Times, Diverse Voices. HSRC Press, Cape Town.
71
Items included in addition to earnings are free water, free sanitation, free electricity and
grants for educational purposes.
72
A more realistic estimate of the number of households benefiting from state aid for
purchasing land and housing would be more than double the observed 7%.
73
Items now included in income estimates are old age pensions (excluding pensions from
previous employment), disability grants, and family and other allowances.
74
The Gini coefficient would no doubt also improve if one were to take tax incidence into
account.
75
H. Bhorat and R. Kanbur, 2006.
76
DBSA, HSRC & UNDP, 2005, Development Report 2005, Midrand, Development Bank of
Southern Africa.
77
Public Service Commission, 2007, Report on the Evaluation of Government’s Poverty
Reduction Programme, Pretoria: Public Service Commission.
78
Rather, in terms of this example, what is counted is the number of employment
opportunities created at the centres.
79
Public Service Commission, 2007, pp.90-91.
80
S. Phillips, 2004, “The Expanded Public Works Programme,” paper presented at the
conference on Overcoming Underdevelopment in South Africa’s Second Economy,
Pretoria, 28 & 29 October 2004.
81
Ibid., p.7.
82
Ibid., p.8.
83
Ibid., p.8.
84
Human Sciences Research Council, 2004, “Food Security in South Africa: Key Policy
Issues for the Medium Term, paper commissioned by the National Treasury; p.25.
85
Ibid. p.28.
86
Department of Social Development, 2001, Business Plan 2001/2003: Poverty Relief
Programme. (Revised), Pretoria: Department of Social Development; p.5.
87
Ironically, however, the Labour Force Survey reveals that between 2000 and 2001, there
was a large shift away from engaging in gardening/farming to provide a main source of
food, in favour of providing an extra source of food. The likely explanation is that, through
better access to social grants and to some extent due to improvements in the labour market,
more and more households are able to rely more on the food purchases than on own
production, almost certainly to the benefit of their perceived sense of food security.
46
88
Human Sciences Research Council, 2004.
89
Department of Social Development and UNICEF, 2006, Guidelines for Early Childhood
Development Services, Pretoria: Department of Social Development; p.2.
90
Ibid.
91
Ibid.
92
Address by the President of South Africa, Thabo Mbeki, on the occasion of his
inauguration and the 10th Anniversary of Freedom, Pretoria, 27 April 2004.
47