1. Mixed methods in poverty measurement: determining the ‘necessities of life’ Eldin Fahmy ¹, Eileen Sutton, Simon Pemberton Presentation to UEA/IDS Workshop ‘Mixed Methods Research in Poverty and Vulnerability’, University of East Anglia, London Campus - 1st-2nd July 2013 ¹ Contact: Dr E Fahmy, School for Policy Studies, Univ. of Bristol E: eldin.fahmy@bris.ac.ukeldin.fahmy@bris.ac.uk www.poverty.ac.uk/content/PSE2011

2. Despite increasing rapprochement between qual. & quantitative methods the language of dichotomy remains pervasive Tendency to view research methods in terms of polarised opposites frustrates attempts to address substantive research problems The idea of a fundamental methodological divide exaggerates differences between approaches, and underplays the diversity of assumptions which underpin social research The distinction between qualitative and quantitative approaches does not capture the full range of options that we face; and...misrepresents the basis on which these decisions should be made...the prevalence of the distinction between qualitative and quantitative methods tends to obscure the complexity of the problems that face us and threatens to render our decisions less effective than they might otherwise be Hammersley (1992: 52) Mixed methods: bridging the methodological divide?

3. Concept of ‘triangulation’ has often dominated discussion of MMR strategies Implies combining methods to address same research problems thus enhancing the validity of inferences Different data sources are seen as commensurate and amenable to integration in terms of the truth claims they make BUT qual and quant methods reflect different concerns, and contrasting strengths and weaknesses: the resulting data cannot be ‘triangulated’ Others refer to complementarity of different approaches: ‘in order to study different levels of enquiry and in order to explore different aspects of the same problem’ (Brannen, 1992: 16) MMR produces a more multi- dimensional account not more accurate or objective measures: “We should combine theories and methods carefully and purposefully with the intention of adding breadth or depth to our analysis, but not for the purpose of pursuing ‘objective truth” (Fielding & Fielding, 1986: 33) What are the objectives of multi-method research?

4. Widely adopted in survey- based poverty research in the UK and internationally – incl 2012 PSE-UK study Rooted in Townsend’s relative deprivation theory: needs are understood as socially determined and relative to prevailing normative standards Response to longstanding critique of expert judgement in determining the ‘necessities of life’ Survey methods used to ascertain public’s views on contemporary necessities, and to incorporate these views in subsequent survey measurement of deprivation The existence of a widespread public consensus on the necessities of life is central to the this approach BUT the nature and meaning of ‘consensus’ here is not currently well understood What is the consensual approach to poverty measurement?

5. Same overall measurement approach & question wording were used in1999 and 2012 studies Module on public perceptions of necessities within: – Summer 2012 ONS Opinions (GB) – June 2012 NISRA Omnibus (NI) Stratified random sampling of: – N=1,447 (GB), 51% response – N=1,015 (NI), 53% response Selection of items for inclusion based upon: – Analysis of existing survey evidence (e.g. 1999 PSE-GB and 2002 PSE-NI studies) – Expert review of potential survey items by the PSE-UK team and Intl. Advisory Group – Series of 14 focus group discussions with different population groups in Eng,. Scot., Wales & NI How are ‘necessities of life’ measured in the 2012 PSE-UK?

6. 2012 ONS Opinions Survey Necessities module question format (e.g. adult items)

7. A public ‘consensus’ is said to exist where: A simple majority of respondents agree that specified items and activities are things which everyone ‘should be able to afford and which they should not have to do without’ There are no significant social differences in respondents’ perceptions of these items, i.e. consensus exists across social divisions (class, gender, ethnicity, age, etc.) The subset of items meeting majority threshold test can be subject to standard statistical tests to establish degree of between- group consensus These items are then included in 2012 PSE-UK main stage survey, to establish how many households cannot afford ‘necessities’ and to derived a consensual deprivation index How is ‘consensus’ measured in the 2012 PSE-UK?

8. Public Perceptions of necessities in Britain, 2012 (% agree)

9. Aims: To contribute to question development in the NatCen Omnibus and PSE Main Stage surveys To explore public perceptions of deprivation, living standards and social exclusion in the UK today To update and advance the ‘state of the art’ in the measurement of PSE Quota sample design: 14 focus group interviews with114 participants were conducted in: Bristol, Cardiff, London, Glasgow and Belfast Separate groups recruited by: – household income – household composition – ethnic origin (min ethnic booster) Within groups, respondents were then recruited by: gender; age; tenure; and; (where relevant) employment status and; age of oldest child Focus group design: aims and sampling

10. Format: Groups comprised 6-10 participants lasting 2.5 hours each. Research comprised two phases… Phase One: Unstructured approach using brain-storming methods Aims to elicit participant suggestions on basis of group consensus All participants’ asked to consider situation of hypothetical couple with two children Phase Two: Sought to ‘test’ items agreed by Phase 1 groups, and also explore wider indicators of living standards based on card sort methods Participants sorted items into three groups: – Necessities: essential things which everyone should be able to afford if they want them – Desirables: things which many/most people have but which are not essential – Luxuries: things which are costly and exclusive and which fewer people have Focus group design: instrumentation & conduct

11. Can we establish a ‘valid’ consensus on the basis of individualised survey responses? Respondents ‘asked to provide immediate responses to tightly worded questions about complex and sensitive issues to which few of them will previously have given much thought’ (Walker, 1987) How stable are respondents’ views, e.g. in view of empirical evidence or wider public discussion on these issues? Survey methods tell us little about: – Evaluative criteria respondents employ in determining whether specific items are ‘necessary’ – Whether the concept of ‘necessity’ is itself equated by respondents with those items which ‘no-one should have to do without’ Does existing interpretation of ‘consensus’ ignore the extent of inter- individual variations in perceptions of necessities? (McKay, 2004) How can qualitative methods inform understanding of public views on necessities?

12. Interpersonal dynamics of focus group interactions tend towards consensus: this may limit diversity and/or intensity of views Knowledge is assumed to be inter-subjective, negotiated, and contingent rather than individualised, external, and absolute Is knowledge commensurate with survey evidence? Is it possible to ‘triangulate’ methods based on very different epistemologies? Process of deliberation differs markedly from the response process in surveys (e.g. social conformity pressures; impact of dominant individuals and shared knowledge) Primary unit of analysis is group interaction: it is therefore important to consider both the extent of between- group and within-group differences in participants views and perspectives Methodological issues in determining consensus

13. Distinction between positive and normative judgements concerning the items and activities that ‘all people should not have to go without’ Some interpreted this to denote items that people in fact would be likely to prioritise if they were experiencing poverty, rather than things all people ought to be able to afford Participants’ interpretation of ‘necessity’ sometimes denoted items or activities that people simply cannot live without (rather than ought not): necessities are things ‘the poor’ have Interpretation of ‘necessities’ as items that are both affordable and widely enjoyed (and perhaps also impossible to do without in our society today) was one widely supported Participants’ accounts sometimes referred to estimations of how difficult it would be to do without items regardless of affordability: luxuries are not always expensive Cognition issues in determining ‘necessities’

14. Funded by the UK Economic and Social Research Council (Ref: RES-060-25-0052) Cognition issues: interpreting necessity Int: Do you think fresh fruit or vegetables daily are important? →BRS2 RM: Not a necessity, no...In an ideal world yeah, everyone loves a bit of meat...but surely if you’re on the poverty line a bowl of porridge would just see you through CDF3 RF: I know it’s not a necessity but I think all children should experience some sort of a holiday...I think every family should be able...to afford to do that. GLS1 RM: There’s a difference between what that family should be able to afford and what a necessity is... a TV is absolutely 100% this family should be able to afford, but it’s not a necessity so it’s difficult NI3 RF: The way I would have to look at necessity is, can you survive without it? “Necessities: Things which are essential and which everyone should be able to afford if they want them in our society today”

15. Many participants felt they lacked the necessary info to make a reasoned decision: participants were provided with a hypothetical scenario or ‘vignette’ to facilitate group decision-making Even so, participants had difficulty in making decisions in the absence of sufficient contextual info Relevant contextual factors: Household/personal circumstances and perceptions of need : ‘a hairdryer’ Affordability and availability: ‘local amenities and services’ Affordability and social resources: networks, support, care Assumptions about eligibility and moral judgement (i.e. the ‘deserving poor’) Judgement issues: need, entitlement & the abstract individual SCENARIO: Tom (aged 38) and Jenny (aged 35) are a married couple with two children, Jack (aged 12) and Lizzie (aged 8). They live in the suburbs of a large city. Tom works at a local hospital and is the sole wage earner within the household. Both parents are in good health.

16. Funded by the UK Economic and Social Research Council (Ref: RES-060-25-0052) GLS2RF: It depends how much he’s earning first and foremost…it just really depends LDN2 RM: What sort of accommodation would he able to afford? Are they social housing, are they private housing? BRS1 RM: I’d say car only if public transport not available. NI1 RM: I think it depends where you work and where your schools are LDN3 RM: If you’re saying there’s nobody working in the house then I’d say no way, but you’ve got a working household you would hope in this country that people could [go out for a meal] CDF2 RF: I’m not being horrible to poor people but why should they be allowed to have double glazing when people who are working can’t afford it Judgement issues: need, entitlement & the abstract individual

17. Conclusions: taking seriously participants’ perspectives To be true to the consensual approach, people must be given scope to express their views. They need time to find their own words, to reflect on their own experience, and to grapple with the complexities of the subject. Researchers must equally be prepared to listen to their respondents and to work with their 'real-world' concepts. Similarly they should be willing to enter into a dialogue with their respondents. Opinions grounded in ignorance, while interesting in themselves and sometimes valuable as predictors of behaviour, have little utility as a basis for policy not least because they are likely to be very unstable. Moreover they do not do justice to the intellect of the respondents or to their presumed commitment to the research exercise. Researchers are therefore obliged to provide respondents with the information which they need in order to make reasoned choices and, as far as possible, to provide feedback on the consequences of the choices made (Walker, 1987: 221)

18. Mixed methods in poverty measurement: determining the ‘necessities of life’ Eldin Fahmy ¹, Eileen Sutton, Simon Pemberton Presentation to UEA/IDS Workshop ‘Mixed Methods Research in Poverty and Vulnerability’, University of East Anglia, London Campus - 1st-2nd July 2013 ¹ Contact: Dr E Fahmy, School for Policy Studies, Univ. of Bristol E: eldin.fahmy@bris.ac.ukeldin.fahmy@bris.ac.uk www.poverty.ac.uk/content/PSE2011