Dr Gemma Wright presented by David McLennan Second Peter Townsend Memorial Conference

1. Poverty measurement in South Africa: involving people in the definition process in a highly unequal society Dr Gemma Wright presented by David McLennan Second Peter Townsend Memorial Conference Measuring Poverty: The State of the Art 22nd January 2011

2. Introduction – involving people in defining poverty • The definition of poverty is not necessarily just the terrain of academics. • There are many ways in which people can inform the way in which poverty is defined and measured, which in turn can inform policy solutions, e.g. • democratically-derived income poverty threshold (e.g. minimum income question) • ‘Voices of the poor’ approach • democratically-derived set of multi-dimensional deprivation indicators (e.g. the consensual approach)

3. Introduction - South Africa • 9 provinces and 11 official languages • Population: 49.99 million people in 2010 (Stats SA, 2010) • Population groups: 79% Black African, 9% Coloured, 9% White, 3% Indian/Asian (using SA terminology) (Stats SA, 2010) • Life expectancy: 53 for men, 55 for women (Stats SA, 2010) • Headcount ratio: 0.70 (R949 per capita pcm) (Leibbrandt et al., 2010) • Gini coefficient: 0.70 in 2008 (Leibbrandt et al., 2010) • High levels of multiple deprivation, especially in former homeland areas (Noble et al., 2009)

4. Introduction – South Africa Central questions that arise when involving people in the poverty definition process in South Africa include: • In the context of such high levels of income poverty, will most people have minimalist views about necessities? • Given the legacies of colonialism, segregation and apartheid (i.e. stark racial, spatial and wealth inequalities) will we find different ideas of an acceptable standard of living amongst different groups? • How do definitions of necessities differ by place?

5. The Indicators of Poverty and Social Exclusion Project • Takes as its starting point what ordinary people regard as an acceptable standard of living that all South Africans should enjoy in the present day. • Follows the tradition of Townsend, Mack and Lansley, Gordon and Pantazis and the many subsequent studies using this approach in the UK and internationally. • Resonates with struggle history e.g. the collection of people’s political social and economic demands which were reported to the Congress of the People and which fed into the preparation of the 1955 Freedom Charter. • Explores the possibility of a relative, multi-dimensional definition of poverty which is consistent with the policy prescriptions in South Africa.

6. Methodology • 48 focus groups asking what people should be able to have/have access to/do in order to have an acceptable standard of living in present-day South Africa • Pilot survey of definitions in South African Social Attitudes Survey (SASAS) 2005 • Definition and measurement questions in SASAS 2006 (nationally representative data) • to develop a democratic definition of poverty • and to measure who has and does not have the socially perceived necessities

7. SASAS 2006 module • Module attempted to represent the breadth of issues raised in focus groups (across domains of deprivation) and a range of probable ‘basics’ through to probable ‘luxuries’. • Module did not attempt to address issues of quantity and quality and made no assumption about the provider of the necessities • Module had 50 definitional questions: 33 about possessions, 4 about activities, 8 about the neighbourhood, and 5 about relationships with friends and family

8. SASAS 2006 findings: the socially perceived necessities • 36 of the 50 items were defined as essential by more than half the population • For the 36 items that were defined as essential by 50% or more of the population, the scale reliability coefficient (alpha) was calculated to be 0.9201 • The percentage of the population defining each of the 49 common items as essential in 2005 and 2006 correlates 0.96 (Spearman’s rho)

10. Domains or Themes emerging • Using PCA a number of themes emerged - three were particularly prominent: • service provision/infrastructure-oriented (mains electricity, a weather-proof house, street lighting, tarred roads close to the house, and separate bedrooms for adults and children) • material possessions (TV, cell phone, radio, sofa/lounge suite, a special meal at Christmas or equivalent festival and some new clothes) • social networks (someone to talk to if you are feeling upset or depressed, someone to lend you money in an emergency, someone to look after you if you are very ill, having an adult from the household at home at all times when children under ten from the household are at home)

11. Was there a consensus about the necessities? A striking level of agreement between groups e.g. % defining each of the 50 items as essential correlate highly: • Women & men: 0.98 • Aged 16-24 & aged 25+: 0.97 • Aged 65+ & aged under 65: 0.95 • Equiv. hh income (R847 pcm): 0.92 • Urban & rural: 0.90

12. What are the patterns of possession of the socially perceived necessities? Number of items defined as essential is similar by population group, but possession of the necessities is a very different story

13. Median number of socially perceived necessities lacked by self-defined poverty status Source: SASAS 2006

14. Influence of context on preferences • This approach embraces the fact that: ‘the public’s perceptions of necessities reflect the conditions and dependencies of contemporary life – whether these are created by market availability or by developments in social structure and interaction’ (Pantazis et al., 2006: 119) • Nevertheless it is still useful to explore the extent to which people’s definitions of necessities are influenced by their circumstances (McKay, 2004; Hallerod, 2006). • We therefore wanted to explore issues such as whether: ‘Life on a low income can depress aspirations, leading some groups to express that they do not want items that, in other circumstances, they might well desire.’ (McKay, 2004: 218).

15. Preferences and poverty definition and measurement • The experience of poverty can, in itself, dampen expectations • This may manifest itself at both the definition and measurement stages • Definition – ‘out of my league’ (i.e. answering ‘not essential’ thereby deflating levels of % saying essential for items) • Measurement – ‘didn’t want it anyway’ (i.e. answering ‘don’t have, don’t want’ thereby deflating % with enforced lack) • Yet deciding that people need things other than what they feel (or are willing to say) that they need is problematic and in this context brings with it the danger of imposing ‘expert’ definitions of need

16. Adaptive preferences and time • People’s expectations about their standards of living may have lowered due to persistent poverty; or risen (e.g. due to political climate, change of reference group, change of economic status), or not changed at all over time. • Burchardt - a longitudinal study is needed to explore adaptive preferences • Yet it is possible to explore the issue by identifying situations where people’s circumstances may have had an impact on their definitions of socially perceived necessities and therefore where adaptive preferences may exist

17. Urban and rural definitions of necessities • Urban and rural responses to the 50 items correlate 0.9. • 31 items (of the 36 socially perceived necessities) were defined as essential by 50% or more of urban respondents and by 50% or more of rural respondents • There are instances where items are defined as ‘more essential’ by urban respondents than by rural respondents e.g. air quality and security.

18. ‘Bounded’ or ‘different’ realities • Conversely, some items were defined as more essential in rural areas than in urban areas. • In the SASAS 2005 pilot there was one item which rural respondents rated as a necessity but urban people did not, using the 50% threshold – the wheelbarrow (a transitional necessity…)

20. Necessities for poor people in informal settlements and former homelands • No significant discrepancies were found between people in informal urban areas and in former homeland areas for two of the three main emerging themes identified, i.e. items relating to material possessions (apart from the television) and social networks. • However, having controlled for income and possession, people in urban informal areas were more likely to define service delivery items as essential than people in other areas (urban formal, rural formal, former homeland): street lighting, tarred roads, flush toilet, bath or shower. • This does not mean that poor people in rural areas are satisfied with poor service delivery: SASAS shows high dissatisfaction in rural and urban areas.

21. Possible reasons for differences • Urban informal - influenced by the evidence of good service delivery in urban formal areas, e.g. over 100 service delivery protests in 2010 in deprived urban areas. • People in former homeland areas may be defining service related items pragmatically, dissatisfied with the status quo but not defining them as essential due to the unlikely probability of obtaining these items in the near future.

22. Concluding Remarks • People regard an acceptable standard of living in a multi-dimensional way and the ‘essentials’ go beyond mere subsistence • Striking levels of agreement across different groups about what constitutes an acceptable standard of living • Possible influence of place on definitions of necessities. Service delivery items are at the forefront for people in informal settlements – living at the cutting edge of an unequal society. • Amidst high levels of internal migration and urbanisation a longitudinal study would shed further light on adaptive preferences and place.

23. Acknowledgements • Grateful thanks to David McLennan for presenting the paper. • Members of the IPSE project team: Michael Noble (PI), Gemma Wright, Andrew Ratcliffe, Wiseman Magasela, Helen Barnes, Danielle Mason and Lucie Cluver (University of Oxford). • Collaborators: Jonathan Bradshaw (University of York), Sharon Zichawo and Richard Chigume (University of Fort Hare), Ben Roberts (Human Sciences Research Council). • Focus group facilitators: Wiseman Magasela, Andrew Ratcliffe, Esmerelda Alman, Tozama Diamond, Desiree Galeshewa, J Govender, Zola Hloma, Daryl Howard, Mbu Kunene, Xolani Kupiso, M Mogotsi, Khanyisa Ndzuta, Geoffrey Nkadimeng, BP Qwana, Siphiwo Tayo, Matimela Joyce Lerato Tsiti and Anna van den Berg • The IPSE project was undertaken between 2003 and 2008 for the Department of Social Development in the Republic of South Africa with funding from the UK Department for International Development Southern Africa, as part of the Strengthening Analytical Capacity for Evidence-based Decision-making (SACED) Programme.

24. Project outputs • Wright, G., Noble, M. and Magasela, W. (2010) ‘Towards a democratic definition of poverty: socially perceived necessities in South Africa’, in B. Roberts, M. wa Kivilu, and Y.D. Davids (eds.), South African Social Attitudes 2nd Report: Reflections on the age of hope, Cape Town: Human Sciences Research Council Press, pp. 143-166. • Noble, M., Wright, G., Magasela, W. and Ratcliffe, R. (2007) ‘Developing a Democratic Definition of Poverty in South Africa’, Journal of Poverty 11(4), pp.117-141. • Wright, G. (2008) Findings from the Indicators of Poverty and Social Exclusion Project: A Profile of Poverty using the Socially Perceived Necessities Approach: Key Report 7, Pretoria:  DSD, RSA.  • Barnes, H. and Wright, G. (2007) Findings from the Indicators of Poverty and Social Exclusion Project: Education,Key Report 6, Pretoria: DSD, RSA.   • Barnes, H., Cluver, L. and Wright, G. (2007) Findings from the Indicators of Poverty and Social Exclusion Project: Children,Key Report 5, Pretoria: DSD, RSA.   • Cluver, L., Magasela, W. and Wright, G. (2007) Findings from the Indicators of Poverty and Social Exclusion Project: Health and a Safe Environment, Key Report 4, Pretoria: DSD, RSA. • Magasela, W., Wright, G., Ratcliffe, A. and Noble, M. (2006) Findings from the Indicators of Poverty and Social Exclusion Project: Housing, Key Report 3, Pretoria: DSD, RSA. • Ratcliffe, A., Magasela, W., Noble, M., Wright, G., Zichawo, S. and Chigume, R. (2005) Methodology for the Qualitative Stage of the Indicators of Poverty and Social Exclusion Project, Key Report 2, Pretoria: DSD, RSA. • Noble, M., Ratcliffe, A., Magasela, W., Wright, G., Mason, D., Zichawo, S. and Chigume, R. (2004) Preliminary Findings from the Qualitative Stage of the Indicators of Poverty and Social Exclusion Project, Key Report 1, Pretoria: DSD, RSA.