Southern African Regional Poverty Network (SARPN) SARPN thematic photo
Country analysis > South Africa Last update: 2020-11-27  

 Related documents

The measurement of poverty in South Africa Project: Key issues

Studies in Poverty and Inequality Institute (SPII)

27 February 2007

SARPN acknowledges the Treasury Department as a source of this document:
[Download complete version - 301Kb ~ 2 min (60 pages)]     [ Share with a friend  ]


‘Society means a shared life. If some and not others are poor, then the principles on which life is shared are at issue: society itself is in question.’ (Halsey, 1985: xxiii).

Poverty is a contested concept; and it is contested with good reason. Arguments over how poverty should be conceptualised, defined and measured go beyond semantics and academic hair-splitting. The conceptualisation, definition and measurement of poverty in a society is like a mirror-image of the ideals of that society: in conceptualising, defining and measuring what is unacceptable in a society we are also saying a great deal about the way we would like things to be. It is therefore vital that the concepts, definitions and measurements of poverty, as well as being theoretically robust, are appropriate to the society in which they are applied.

Poverty is also political because it relates to the allocation or distribution of resources, and reflects the impact of past and present policy choices (Meth, 2006). The ways in which politicians, citizens and experts use the concept of poverty have very divergent and diverse roots in social, political and philosophical discourses. Present day poverty discourse draws on complex and sometimes contradictory underlying assumptions about what people are supposed to need in order to live a minimally human life; about the obligations between individuals and society, about the relation between have and lack, ill-being, well-being and suffering; and about social life and individual agency. These underlying discourses and narratives are not neatly aligned — and this means that the concept of poverty as it exists in ordinary language has an inherent ‘messiness’ about it. It can (and is) often used in divergent ways, to highlight different phenomena, and to serve a wide range of purposes. This is not necessarily a bad thing. In fact, its protean breadth and inner diversity is part of what makes the concept of poverty so important and powerful in the debates by which government, social arrangements, institutions and policies are legitimised. It allows the concept to be used in very nuanced, complex and responsive ways (du Toit 2005). Given this ’messiness’, it is not possible to refer to any single ‘scientific’ understanding of poverty (Alcock, 1993, quoted in Magasela, 2005a).

Enquiry into levels of poverty amongst South Africans is not new. Levels of poverty amongst the white population formed the basis of the first Carnegie Commission Inquiry into Poverty during the Great Depression in an attempt by the government to address the “poor white problem” in 1928 (Magasela, 2005). In the early 1980’s, as a result of concerns by progressive forces within the country about growing levels of poverty amongst the population as a whole, the Second Carnegie Commission Inquiry1 into Poverty in South Africa was held, and this was followed by the 1993 “Project for Statistics on Living Standards and Development” undertaken by the World Bank and South African researchers for the African National Congress who wanted a definitive assessment of the extent of poverty within the country prior to taking office (Magasela, 2005). This work built on a strong tradition of documentation, research and analysis into income levels and causes of poverty by researchers such as Simkins and McGrath and Pillay, and research units including the Economic Research Unit at the former University of Natal, and other research initiatives such as the project into the gap between white and black incomes funded by the Anglo American Chairman’s Fund.

After 1994, a number of seminal reports on poverty were commissioned by various bodies, including the office of the Deputy President. These included the 1995 Key Indicators of Poverty in South Africa, the Participative Poverty Assessment – South Africa report and the Poverty and Inequality Report, both published in 1998. These studies and reports were part of a national commitment to eradicating poverty that was embodied in the “War on Poverty” declared by then President Nelson Mandela on behalf of the state, together with broader civil society organisations in 1996. Following on the footsteps of these works was the very comprehensive work, including useful analysis of the state of poverty in South Africa, undertaken by the Committee of Inquiry into a Comprehensive System of Social Security for South Africa (the ‘Taylor Committee’ whose work was published in March 2002 in a report entitled Transforming the Present – Protecting the Future. The committee recommended the adoption of a five pronged social protection system to address not just widespread income poverty, but also a lack of access to assets and basic needs, as well as widespread capabilities poverty2.

In terms of the 1995 Copenhagen Declaration, South Africa has a commitment to adopt an official measure of poverty. It has not yet done so. Instead, different measures have been developed and used by different researchers as well as different government departments and agencies.3 This lack of consensus on poverty measures has in truth had both positive and negative consequences. While the use of different measures has led to dissent — and sometimes confusion — about some of the findings of research on poverty, it is interesting to note that certain government departments have begun to conceptualise and define poverty in ways that reflect the different dimensions of the manifestation of poverty, with specific reference to their constitutional mandate. This has enabled these departments to develop more comprehensive applications of anti-poverty policies than those they might have devised had they simply ‘defaulted’ to an already established poverty line. Many of these dimensions of poverty correspond with the socio-economic rights guaranteed in the Constitution of South Africa. Because of this fundamental link, Magasela (2005a) argues that there is a need for poverty research to focus more keenly on the use of indicators of multiple deprivation in South Africa rather than absolute or minimalist income based poverty lines, to enable government departments to be true to the realisation of their constitutional mandates.

At the same time, the lack of official measures has also sometimes led to confusion, and has certainly fed into to the development of differences and contestations around actual levels of poverty in the country. According to Everatt (2003, 77), “The impact of definitional imprecision has been and remains considerable, affecting the development programmes while fuelling illtempered, if ultimately rather hollow, debate”. Meth (2006) argues that there are two main aspects to the dissent in South Africa - firstly around the actual conceptualisation of poverty, and secondly relating to both the execution and the interpretation of the surveys that provide the data for studies into the incidence of poverty. Sometimes this has led to great uncertainty about poverty levels and the changes in the extent and nature of poverty in South Africa.

Thus, according to the 2003 UNDP South Africa Human Development Report, income poverty and inequality were found to have increased during recent years. Despite this, the report also found that using a national poverty line of R354 per month per adult equivalent based on 1995 values, the total percentage of people living in poverty had fallen from 51.1% in 1995 to 48.5% in 2002, likewise the total number of people living below the World Bank line of $2 per day had fallen from 24.2% in 1995 to 23.8% in 2002. The total number of people living below $1 per day (in other words, in destitution) however was found to have risen from 9.4% to 19.5% between 1995 and 2002, and the study also found that despite a slight drop in the rate of people living in poverty, the total number of poor people had actually risen from 20.2 million to 21.9 million people between 1995 and 2002 (UNDP, 2003, page 41).

Again, according to the Towards a Ten Year Review Discussion Document released by PCAS, since 1994, South Africans have grown wealthier at slightly more than 1% per annum (PCAS, 2003, page 35). However, according to a subsequent discussion document, A Nation in the Making. A Discussion Document on Macro-Social Trends in South Africa, 2006, the number of households living below an estimated poverty line of R322, per month rose from 28% in 1995 to 33% in 2000 (PCAS, 2006, page 12). Yet the document also claims that since 2000, the total number of people living in poverty has fallen from approximately 18.5 million poor people to 15.4 million (ibid). This claim, which has its origins in Van der Berg at al (2005) however has been vehemently rejected by economists such as Charles Meth (2006a; 2006b). The fact that such heated argument was generated by minute percentile differences demonstrates both the height of the political stakes and the lack of clarity about the meaning of these statistics. Constructive debate needs to be much better informed, with more clarity as to both the nature of the underlying concepts and the meanings of the statistical data and findings themselves.

Moves are currently afoot to dispel some of this confusion and lack of a shared approach. In his 2005 Budget Speech, the Minister of Finance announced government’s intention to adopt an official measure of poverty. It is hoped that this initiative will aid in the development of appropriate policy interventions to reduce the impact of poverty and ultimately, work towards its eradication. In addition it may significantly increase the public debates on the various measures of poverty, the causes and implications of poverty in South Africa. This represents an important political opportunity in the development of poverty-related policy in South Africa. More often than not internationally, the task of defining poverty falls to those in power, and hence the conceptualisation as well as the definition is reflective of the dominant ideologies (Magasela, 2005a). The task of defining poverty is often given the appearance of technical neutrality, which masks the political and ideological assumptions which are made. Embarking on a transparent consultative process in finalising the state’s approach to defining poverty will go far in ensuring that those who support the objectives of government, but not necessarily the dominant political and economic ideologies, will not view the result of such enquiries as the exclusive domain of the state.

The new emphasis on finding ‘official’ measures of poverty in South Africa therefore can have a positive effect, but could also bring its own dangers. On the one hand, measures and indicators are vital: they can help take poverty debates beyond rhetoric, and can bring a great deal of concreteness and specificity into discussions that could otherwise be rather ungrounded. On the other, concrete measurement is only one of the ways in which poverty should be understood, there is a degree of inherent complexity that measurement cannot (and should not) dissipate. As this document will argue in the pages below, this means inter alia that South Africans should beware of thinking that there is one single ‘objective’ measure or indicator that can satisfy all the quantitative needs involved in engaging with the reality of poverty in South Africa. Given the fact that a significant proportion of the population lack the means to meet basic needs in South Africa, including food, housing, education and health provision, there is no doubt that there is a need for a measurement of poverty in South Africa that is based on meeting these basic needs, many of which are guaranteed in the South African Constitution. However – even if it were possible to reach agreement on just what those ‘basic needs’ are – there is also a need to take into account more than just the meeting of basic needs. This is because in some important respects stark economic divisions could continue to exist even once everyone’s’ most basic needs have successfully been met. Thus it is important to acknowledge the impact of equality and inequality of access to resources in addition to the study of the actual state of impoverishment or enrichment of access to these same resources.

Neither is this a process that can simply be left to experts. For post-apartheid South Africa the definition and selection of a set of appropriate poverty measures presents an opportunity for a participative national expression of the “values, ideals and aspirations” which informed the hard fought battle against the exploitation and oppression of the majority of South Africans for so long (Magasela, 2005). Though expertise is important, as important is a process of engaging with the underlying issues involved in this exercise, and developing a shared approach to the broader problems of poverty and inequality in South Africa.

There are a number of different ways of conceptualising poverty which we unpack in the body of this paper. These range from the historical narrow or subsistence view that equates poverty with the lack of resources necessary for the basic survival of a person, to a broader view that recognises that poverty is multidimensional and that the state of well-being or social inclusion for any person requires access to a far wider range of resources than those necessary to ensure mere survival. This second, more recent conceptualisation of poverty seeks to define the resources that any person living in a community or society requires in order to promote social cohesion and inclusiveness, which are inherent in the right to human dignity. This diversity in the conceptualisations of poverty affects both how poverty is defined and measured and the tools that are used to operationalise these measures.

One key theme in this paper is that is important to pay attention to the underlying assumptions that shape the form and content of a particular poverty measure. The outcome of any measurement exercise is determined by the initial parameters, or put more starkly, “the measure determines the result” (Bradshaw (2000), quoted in Magasela, 2005(a)). The more simplistic the measure, the less ably can the result reflect comprehensively the nature or extent of the state of impoverishment being studied (Meth, 2006). One of the purposes of this paper is to make these underlying assumptions and trade-offs explicit, and to render them more available for debate.

  1. These Inquiries were so called because they were funded by the American philanthropic Corporation of Andrew Carnegie.
  2. There has been much political interest nationally and internationally regarding the extent to which poverty has increased or decreased since 1994 (Bhorat and Kanbur, 2005: 3). There have been numerous studies on the extent to which income poverty levels have changed overall and for different subgroups of the population during the first years of democracy (e.g. Adato et al., 2004; Agьero et al., 2006a; Bhorat et al., 2004; Bhorat and Cassim, 2004; Bhorat and Kanbur, 2005; Hoogeveen and Цzler, 2005; Leibbrandt et al., 2004; Leibbrandt et al., 2005; Leite et al., 2006; May and Meth, 2005; Meth, 2006; Meth and Dias, 2004; Roberts, 2005; Simkins, 2004; UNDP, 2003; Van der Berg et al., 2006; Woolard and Klasen, 2005). There have also been several studies on the extent to which different poverty alleviation strategies have reduced poverty levels, particularly regarding social security transfers (Agьero et al., 2005; Barrientos, 2005; Case and Deaton, 1998; EPRI, 2004). This political interest has sharpened the focus on poverty measurement and data sources over the past few years.
  3. The Key Indicators of Poverty in South Africa used five different types of poverty lines in its study on levels of poverty (Magasela, 2005), while the Poverty and Inequality Report (PIR), considered both money-metric measures of poverty as a standard proxy for poverty, but it also used a “broader, composite indicator of deprivation “ (May, 1998) to deepen understandings about the comprehensive manifestations of poverty.

Octoplus Information Solutions Top of page | Home | Contact SARPN | Disclaimer