Introduction
A framework for poverty analysis must seek to reflect societal change and economic shocks, such as the current crisis, in distinctly human terms. For this, we need the right concepts and measures. In this paper, we discuss some problems associated with existing approaches to conceptualising poverty, social exclusion and deprivation, and discuss the contribution that the capability approach might offer in resolving them. It is argued that the capability approach can provide a framework that can reflect the many ways in which human lives can be blighted, and which thus offers some promise for poverty analysis.
This may come as some surprise to those familiar with the exchange between Sen and Peter Townsend in the 1980s (Sen, Reference Sen1983, Reference Sen1985; Townsend, Reference Townsend1985), which did little to endear the social policy community to Sen's approach. However, there are at least two reasons why it is timely to reconsider the potential of the capability approach. First, in the years since the Sen−Townsend debate, the capability approach has become much more prominent. It has provided the conceptual underpinning for the UN's Human Development Reports (UNDP, 2010), has influenced the understanding of well-being in the recent ‘Sarkozy Commission’ (Stiglitz et al., Reference Stiglitz, Sen and Fitoussi2009) and has been the basis for the Equality and Human Rights Commission's approach to monitoring equality in the UK (Burchardt and Vizard, Reference Burchardt and Vizard2011). Given this increased prominence, we might ask whether some advantages of the approach have been overlooked.
Second, there is, at present, an unresolved tension within poverty analysis between a desire to emphasise a broad measure of multidimensional poverty (e.g. Atkinson et al., Reference Atkinson, Cantillon, Marlier and Nolan2002: 79) and an insistence on conceptualising poverty in narrower terms around a core concept of resources (e.g. Nolan and Whelan, Reference Nolan and Whelan1996). One of the functions of the concept of social exclusion was to cover important additional terrain beyond the concept of poverty (e.g. Room, Reference Room and Room1995). However, the lack of progress in identifying what is meant by social exclusion not only raises questions about its suitability as a concept, but also places this additional terrain in jeopardy. It will be argued that the capability approach offers a way to reconcile this tension between narrow and broad conceptions of poverty, by respecting the former without losing sight of the latter.
The purpose of the paper is to examine the conceptual contribution the approach might make to the analysis of poverty in a developed nation, social policy context, and it is the literature in this field that the paper both draws on and seeks to contribute to. As will hopefully become clear, the approach is not a distinct field of study (‘capability studies’), but is, rather, a lens with which to view our existing concerns – in this case, the problem of poverty. This paper examines the original justifications for contemporary approaches to poverty analysis within social policy so that the distinctiveness of the capability approach, and the contribution it might make in this field, can be identified. A wider approach, drawing on literature from other disciplines, or contrasting concepts of poverty employed in developing and developed contexts, would also be of interest, but is beyond the scope of the present paper. While the focus of the paper is primary conceptual, in the penultimate section some empirical applications of the approach are discussed and some implications for adopting a capability framework are considered. The paper concludes by presenting six insights that the capability approach can provide for poverty analysis.
The capability approach and resource-based approaches to poverty analysis
The capability approach focusses on what people are able to do and be, as opposed to what they have, or how they feel. Sen argues that, in analysing well-being, we should shift our focus from ‘the means of living’, such as income, to the ‘actual opportunities a person has’, namely their functionings and capabilities (Sen, Reference Sen2009: 253). ‘Functionings’ refer to the various things a person succeeds in ‘doing or being’, such as participating in the life of society, being healthy and so forth, while ‘capabilities’ refer to a person's real or substantive freedom to achieve such functionings; for example, the ability to take part in the life of society (Sen, Reference Sen1999: 75). Of crucial importance is the emphasis on real or substantive – as opposed to formal – freedom, since capabilities are opportunities that one could exercise if so desired. The capability approach places particular emphasis on the capabilities a person has, irrespective of whether they choose to exercise these or not.
Sen's first exposition of the approach came in ‘Equality of what?’ (Reference Sen and Sen1982); a paper whose title reflects his emphasis on identifying on what ‘moral information’ considerations of equality, poverty, justice, etc. should be based. In the capability approach, the selection of functionings and capabilities as the ‘informational space’ of analysis is made prior to identifying the precise context of analysis (e.g. analysing poverty, justice, etc.). This priority is what we will call the ‘normative focus’ of the capability approach. It can be contrasted with what we can call a ‘conceptual focus’, which might first decide on the context, asking ‘what is poverty?’, prior to determining the informational space of analysis (e.g. whether to measure poverty using income, consumption, opportunities, etc.). Both questions must, of course, be addressed, but the order in which they are addressed may differ.
From a capability perspective, poverty is viewed as the deprivation of certain basic capabilities, and these can vary, as Sen has argued, ‘from such elementary physical ones as being well nourished, being adequately clothed and sheltered, avoiding preventable morbidity, and so forth, to more complex social achievements such as taking part in the life of the community, being able to appear in public without shame, and so on’ (Sen, Reference Sen, van de Walle and Nead1995: 15). There is no suggestion, as is sometimes claimed, that the capabilities concerned should be determined without reference to prevailing living standards; indeed, Sen has been clear that such standards will influence the selection of relevant capabilities (e.g. Sen, Reference Sen1984: 84–5).
The capability approach questions the central role often afforded to income in poverty measurement. Sen draws a distinction between the actual opportunities, or capabilities, a person has, which he argues are intrinsically important, and their income, which is merely a means to such opportunities, and whose importance is thus both instrumental and contingent (Sen, Reference Sen2009: 233). This relates to the distinction between direct and indirect concepts of poverty drawn by Ringen (Reference Ringen1988). Direct concepts of poverty focus on cases where living standards fall below a certain level, and typically assume that this is because of a lack of resources. Indirect concepts focus on cases where resources fall below a certain point, and typically assume that this results in a low standard of living (Ringen, Reference Ringen1988; Berthoud and Bryan, Reference Berthoud and Bryan2011).
Of course, such distinctions would be of little importance if low income were a good proxy for deprivation. But the capability approach holds that this is unlikely to be the case: people have varying needs and will thus require different levels of resources in order to achieve the same standard of living. For example, the additional costs associated with disability might mean that a disabled person requires a greater amount of resources to achieve the same standard of living as an able-bodied person (Sen, Reference Sen2009). Sen uses the term ‘conversion factors’ to refer to this variability in translating resources (or means) into capabilities (or ends). And while he discusses such conversion factors primarily in theoretical terms, the mismatch between indirect and direct measures of poverty has been one of the primary empirical findings from poverty analysis in the field of social policy in recent decades (see, inter alia, Gordon et al. Reference Gordon, Levitas, Pantazis, Patsios, Payne, Townsend, Adelman, Ashworth, Middleton, Bradshaw and Williams2000; Bradshaw and Finch, Reference Bradshaw and Finch2003).
In prioritising the intrinsic importance of what people can do or be over the resources they possess, the capability approach is unambiguous in favouring a direct approach to poverty analysis. However, there is a strong tradition of direct conceptualisation of poverty, social exclusion and deprivation in the field of social policy which can also claim to focus on what people are able to do and be; for example, whether they can participate in the life of society (e.g. Townsend, Reference Townsend1979). Is the capability approach effectively equivalent to these traditions and, if not, how does it differ? It is to these questions that we turn.
The direct tradition of poverty analysis
In Poverty in the United Kingdom, Peter Townsend pioneered the use of deprivation indicators as a method of tapping into the prevailing style of living in the UK. Townsend's influential definition of poverty was:
Individuals, families and groups in the population can be said to be in poverty when they lack the resources to obtain the types of diet, participate in the activities and have the living conditions and amenities which are customary, or at least widely encouraged or approved, in the societies to which they belong. Their resources are so seriously below those commanded by the average individual or family that they are, in effect, excluded from ordinary living patterns, customs and activities. (1979: 31)
Townsend provided two sets of deprivation indicators: a list of sixty indicators measuring ‘styles of living’ and a summary deprivation index of twelve items drawn from this longer list, which would be used to calibrate an income poverty line. The longer sixty-indicator list was intended to examine the full range of living conditions that were ‘customary, or are at least widely encouraged or approved, in the societies in which they belong’ (1979: 31). Indeed, this breadth was stressed by Townsend himself, who noted that ‘[i]n principle, such a list might be developed . . . from an exhaustive analysis of the amenities available to, and the customs or modes of living of, a majority of the population’ (1979: 251).
While an exhaustive list may always prove elusive, Townsend noted that he sought ‘to ensure that all the major areas of personal, household and social life were represented’ in the questionnaire (1979: 251). The breadth of the sixty indicators was indeed impressive: six items relating to dietary deprivation, four to clothing, four to fuel and light, nine relating to household facilities, four to household conditions and amenities, twelve to conditions at work, five to health, one to education, five relating to environmental deprivation, four relating to the family, two to recreational activities and four to social activities.
We are not told a great deal about the process of reducing these sixty indicators into a more manageable summary index of twelve items other than that there was a desire to select ‘those indicators which apply to the whole population’ (1979: 251) rather than to specific groups. However, this process is of some interest, because the resulting summary index was in no way representative of the breadth of the original list. Not a single item relating to conditions at work, health, education, environmental deprivation, fuel and light and clothing was included in the summary deprivation index, which was a subset of the longer list not only in length but also in scope and reflected a narrower concept of poverty than the full sixty-item set.
The most influential critique of Townsend's work came from Piachaud (Reference Piachaud1981) who questioned the implicit relationship between the absence of items on Townsend's deprivation index and poverty. Piachaud (Reference Piachaud1981: 420) noted that going without the items then might be ‘as much to do with tastes as with poverty’. Rather than assuming non-possession or non-consumption implied deprivation, Piachaud argued ‘[w]hat surely matters most is the choice a person has, and the constraints he or she faces. To choose not to go on holiday or eat meat it one thing: it may interest sociologists, but it is of no interest to those concerned with poverty. To have little or no opportunity to take a holiday or buy meat is entirely different’ (Piachaud, Reference Piachaud1981: 421). It is important to note at this point the close parallel between Piachaud's critique of Townsend and Sen's distinction between functionings and capabilities (and his priority of the latter). Both emphasise the importance of constraints for the conceptualisation of poverty.
Mack and Lansley's (Reference Mack and Lansley1985) subsequent Poor Britain survey, which built on the work of Townsend, defined poverty as an ‘enforced lack of socially perceived necessities’ and sought to establish the public's view on ‘what it is that people need for living in Britain in the 1980s’ (1985: 9). They accepted Piachaud's (Reference Piachaud1981) critique, and those who reported the absence of particular deprivation items were subsequently asked whether this was by choice or because they could not afford them, thus seeking to distinguish between choice and constraint. This second question not only shifted the focus from the absence of items to an enforced absence due to a lack of resources, however, but also influenced the selection of indicators:
The critical role of lack of resources to the concept of poverty also has wider implications, because it determines which aspects of our way of life should be included in a minimum standard of living aimed at measuring poverty. We decided that only those aspects of life facilitated by access to money should be tested in the Breadline Britain survey. (Mack and Lansley, Reference Mack and Lansley1985: 44)
They noted that health, education, conditions at work and environmental factors all affected an individual's quality of life, ‘but they are not in the main paid for’ and were thus excluded (Mack and Lansley (Reference Mack and Lansley1985: 44–5). Subsequent studies have continued broadly in this vein, focussing on this narrow concept of poverty (or ‘material deprivation’) (e.g. Gordon and Pantazis, Reference Gordon and Pantazis1997; Gordon et al., Reference Gordon, Levitas, Pantazis, Patsios, Payne, Townsend, Adelman, Ashworth, Middleton, Bradshaw and Williams2000), which bears a greater resemblance to Townsend's twelve-item summary index than his broader sixty-item list.
The direct tradition of poverty analysis and the capability approach
This narrow approach to conceptualising poverty – as ‘exclusion from the life of the society owing to a lack of resources’ (Nolan and Whelan, Reference Nolan and Whelan1996: 2) – has become enormously influential in recent decades and the indicators used in this approach, as we have seen, are typically justified by their responsiveness to resources. But from a normative perspective, this justification can be questioned, because the ability to participate in the life of society, for example, does not derive its importance from its relationship to resources.
There is no reason, from a normative perspective, why the important dimensions for analysis should be decided by their anticipated responsiveness to resources, and the capability approach, which adopts such a perspective, would offer a considerably broader focus than the direct tradition of poverty analysis. It allows us to consider dimensions such as health, housing deprivation, conditions at work and so forth, which often sit uneasily with existing concepts of poverty and may be relegated to ‘causes’ or ‘consequences’ of poverty or dropped from analysis entirely.
In terms of the shared emphasis on constraints, there is some resemblance between Sen's distinction between functionings and capabilities and Piachaud's insistence that the mere absence of deprivation items cannot straightforwardly be interpreted as implying poverty. For both, poverty must be distinguished from preference. Subsequent surveys have – to this day – responded to Piachaud's critique by asking not only whether deprivation items are absent, but, where they are, asking whether this is because of a lack of resources. However, this has created a false dichotomy whereby the absence of items for reasons other than a lack of resources is typically attributed to ‘choice’. Other constraints that might impede participating in the life of society, such as disability, discrimination or geographical isolation, inter alia, are simply overlooked.
It is incongruous, however, to argue that something is of serious concern if one is deprived of it because of a lack of resources but of little or no concern if as a result of other constraints. Seen in this light, Piachaud's critique raised the right question, but received the wrong answer. The concept of capabilities, on the other hand, considers all potential constraints to well-being achievement, whether this is a lack of resources, disability, discrimination, etc. The capability approach thus takes a broader approach to both the dimensions of interest and the constraints considered than direct approaches to poverty analysis.
The concept of social exclusion
The concept of social exclusion has received substantial attention in recent years, often as a concept allied to that of poverty. The distinction between the two is often said to lie in the multidimensional, dynamic nature of social exclusion, which is supposed to contrast with the static, unidimensional concept of poverty. One of the challenges in locating the social exclusion concept with respect to the capability approach, however, is the extremely contested nature of the social exclusion concept (de Haan, Reference de Haan1998; Levitas, Reference Levitas, Bradshaw, Gordon, Levitas, Middleton, Pantazis, Payne and Townsend1998). There might be widespread agreement on the need to fight exclusion, but ‘fighting exclusion means different things to different people’ (Silver, Reference Silver1994: 544).
The initial usage of the term ‘social exclusion’ is typically attributed to René Lenoir who, in the 1970s, employed the term ‘les exclus’ to refer to those who were falling through France's insurance-based social security programmes. This included ‘the mentally and physically handicapped, suicidal people, aged invalids, abused children, drug addicts, delinquents, single parents, multi-problem households, marginal, asocial persons, and other “social misfits”’ (Silver, Reference Silver1994: 532). While many of these groups may indeed be vulnerable, such a conceptualisation was derived from French Republican ideology which stressed the importance of social cohesion.
While this conceptualisation of social exclusion focussed on social rights (see also Room, Reference Room and Room1995: 243), an alternative conceptualisation focusses on multidimensional and dynamic exclusion from a certain standard of living in economic, political, cultural and social dimensions (Burchardt et al., Reference Burchardt, Le Grand, Piachaud, Hills, Grand and Piachaud2002; Stewart et al., Reference Stewart, Laderchi and Saith2007). In terms of the causes of exclusion, there is widespread agreement that social exclusion considers a broader range of constraints than a ‘lack of resources’ (Evans, Reference Evans1998; de Haan, Reference de Haan1998), and Stewart et al. (Reference Stewart, Laderchi and Saith2007) argue that the concept allows for a focus not only on individuals, but also on groups where group-membership itself is a cause of exclusion – for example, on the basis of ethnicity or caste.
Room (Reference Room and Room1995: 233–4) has suggested that the shift from income poverty to social exclusion entails a broadening of focus on three fronts – from income/expenditure to multidimensional disadvantage; from static outcomes to dynamic processes; and from a focus on the individual or household to the local community. However, while these may be features of a social exclusion approach, the extent to which they are distinctions has been questioned (Burchardt et al., Reference Burchardt, Le Grand, Piachaud, Hills, Grand and Piachaud2002; Gordon, Reference Gordon, Pantazis, Gordon and Levitas2006). Nonetheless, the concept of social exclusion is often presented as a companion to the concept of poverty in order to consider a broader conceptual terrain.
However, what remains contested is the very core of the concept – what it is we want the concept to do. The answers to two important questions lack sufficient clarity, and the implications of these answers are seldom considered. First, is social exclusion an outcome or a process? While the dynamic nature of social exclusion is often professed as an advantage, and, indeed, a feature which distinguishes the concept from the capability approach (Stewart et al., Reference Stewart, Laderchi and Saith2007), there is a crucial distinction between adopting a dynamic perspective and an approach focussing on a process. For example, Room (Reference Room and Room1995: 237) notes that ‘[i]t is not enough to count the numbers and describe the characteristics of the socially excluded; it is also necessary to understand and monitor the process of social exclusion and to identify the factors that can trigger entry or exit from situations of exclusion’. Here exclusion is presented as an outcome, although our focus may be on its causes. Elsewhere, however, he notes ‘[s]ocial exclusion is the process of becoming detached from the organisations and communities of which the society is composed and from the rights and obligations that they embody’ (1995: 243). Here, social exclusion is itself presented as a process.
However, these two conceptions are distinct. In the first – as outcome – we identify a group (the excluded) and, ex post, observe the risk factors that predict their exclusion. In the second – as process – it is the risk factors themselves that are of interest, irrespective of whether the related outcomes, in fact, occurred. As we will see, this lack of conceptual clarity is mirrored in terms of measurement.
The second question is: for whom is social exclusion bad? Is it bad for those afflicted by it, preventing them living as they themselves would like? If so, we may be interested solely in exclusions which are the product of constraints. Or it is bad for society as a whole due to, for example, a breakdown in social cohesion? If this is the case, then we may be justified in focussing on both voluntary and involuntary social exclusion. Barry (Reference Barry, Hills, Grand and Piachaud2002) argues that voluntary exclusion should be included in the concept, noting that seemingly voluntary withdrawal may obscure exclusionary processes, and that social solidarity may be required to promote individual well-being and achieve social justice.
Attempts to distinguish between choice and constraint are not always clear-cut and the concern that seemingly voluntary withdrawal may overlook exclusionary processes is a legitimate one. However, this does not challenge the primary theoretical focus on constraints. Furthermore, greater solidarity may indeed improve prospects for individual well-being and social justice, but to include solidarity in the concept of social exclusion is to confuse a particular strategy for preventing exclusion with its very nature. Voluntary exclusion may, thus, be bad for society, but it is not clear why we should unambiguously accept it is bad for individuals themselves.
Measuring social exclusion
That there is little agreement about how to measure social exclusion may come as no surprise given the varied interpretations of the concept. Indeed, the conceptual ambiguity relating to the two questions discussed above is reflected directly at the level of measurement. There is an important distinction between a measure or indicator and a risk factor, which is related to whether social exclusion is seen as an outcome or a process. Levitas (Reference Levitas, Gordon and Townsend2000: 369–73) has been critical of the use of certain indicators of social exclusion such as whether one smokes, or one's parents are divorced or, amongst working-age adults, whether one contributes to a private pension scheme, noting, as regards the latter, that while non-provision during one's working life may lead to future exclusion, it says nothing about one's current circumstances (2000: 376). The lack of clarity about whether social exclusion is an outcome or a process is thus reflected in the indicators used to measure the concept.
Furthermore, indicators of social exclusion often relate to (voluntary) disengagement and not (involuntary) exclusion per se, reflecting confusion about whom social exclusion is bad for. The influential Poverty and Social Exclusion survey noted that civic disengagement could be considered to be a component of social exclusion. The study included as indicators of social exclusion whether people voted in the last general election, helped on fundraising issues, urged others to vote, contacted a local councillor or urged some else to do so, acted as an officer of an organisation or club, made a speech before an organised group, had written a letter to an editor, actively participated in a political campaign or stood for public office (Gordon et al., Reference Gordon, Levitas, Pantazis, Patsios, Payne, Townsend, Adelman, Ashworth, Middleton, Bradshaw and Williams2000: 66). Others have used indicators such as how often respondents meet with friends or talk to neighbours, or whether they are members of a sports club or political party (Tsakloglou and Papadopoulos, Reference Tsakloglou and Papadopoulos2002). Not one of these indicators survives Piachaud's critique that our focus should be on constraints and not choices, a principle which is now part of the received wisdom when measuring poverty but which remains strangely neglected with respect to social exclusion.
Social exclusion and the capability approach
Perhaps the greatest obstacle to considering the relationship between the capability approach and the concept of social exclusion is the plurality of interpretations of the latter. It can, in fact, be questioned whether ‘social exclusion’ is really an academic concept at all, or whether it is merely a rhetorical device, laden with negative connotations but devoid of any shared meaning.
If social exclusion is primarily bad for society, then voluntary social exclusion may be evidence of a lack of social cohesion. But this would imply that the concepts of poverty and social exclusion are uneasy companions – one relating to persons and the other to society. If, on the other hand, social exclusion is bad for people themselves, then there is good reason to believe that voluntary social exclusion should not be included in the concept since there is nothing very ‘exclusionary’ about not having written a letter to an editor, for example.
If the focus of the concept is solely on constraints, then social exclusion would bear some resemblance to the concept of capabilities. Even then, one presumes that, while both are inherently multidimensional (Stewart et al., Reference Stewart, Laderchi and Saith2007), the concept of social exclusion must limit itself to dimensions of a ‘social’ nature (Sen, Reference Sen2000) and thus it is not clear whether and how the social exclusion concept could be used to tap into dimensions such as ill-health, time poverty and so forth. From a normative perspective, the importance of particular dimensions is not derived from, or limited by, the extent to which they are ‘social’.
Thus, the lack of clarity in the concept of social exclusion, despite its ubiquitous usage, poses serious problems for analysis. Without considerable theoretical advancement – and in a form that achieves at least some common acceptance – the concept of social exclusion would seem to be considerably less useful than it may, on first glance, appear. While the precise dimensionality of both social exclusion and the capability approach is ambiguous (Stewart et al., Reference Stewart, Laderchi and Saith2007), there is no reason why the capability approach must be limited to ‘social’ dimensions, and there are no ambiguities as regards whether capability deprivation is an outcome or a process or of whom capability deprivation is bad for. The capability approach thus provides a perspective which is both broader than social exclusion in terms of dimensionality, and which possesses a considerably firmer conceptual foundation.
The concept of deprivation
It is worth also considering the relationship between the capability approach and two contrasting approaches to the concept of deprivation. The first, and most common, approach is to use the concept of deprivation to refer to the direct tradition of poverty analysis involving deprivation indicators (for example, Saunders, Reference Saunders2004). The problem with this interpretation is that, on this view, the concept of deprivation adds nothing to the direct conception of poverty, (as material deprivation) which, as we have noted, is largely limited to commodities and activities that reflect variations in resources.
The second concept of deprivation has an older – and broader – conceptual pedigree. In Poverty in the United Kingdom (1979), Townsend offered a definition of deprivation distinct to that of poverty: ‘[p]eople can be said to be deprived if they lack the types of diet, clothing, housing, environmental, educational working and social conditions, activities and facilities which are customary, or at least widely encouraged or approved, in the societies to which they belong’ (1979: 413). The distinguishing features of Townsend's concept of deprivation are (i) the extension beyond resource-based dimensions and (ii) extension beyond a lack of resources as the sole cause of deprivation. In this conception, deprivation is viewed as being broader than poverty, and is conceptualised by Townsend as ‘a state of observable and demonstrable disadvantage relative to the local community or the wider society or nation to which an individual, family or group belongs’ (Townsend, Reference Townsend1987: 125), with a focus on those who are forced to lead ‘restricted or stunted social lives’ (1987: 128). There are, then, two distinct concepts of deprivation, namely material deprivation (i.e. focussing on resource-based dimensions) and multiple deprivation (i.e., across a wide range of dimensions), although this distinction is not always explicit within the literature.
Deprivation and the capability approach
There is some considerable similarity between the capability approach, with its focus on removing barriers which force people ‘to live less or be less’ and this second, broader conceptualisation of deprivation, focussing on ‘restricted or stunted social lives’ (Townsend, Reference Townsend1987: 128). One important similarity is the potential breadth of both approaches – a feature stressed by Townsend, who noted that ‘the scientist must consider deprivation as the darker side of the entire lifestyle of a people’ (Townsend, Reference Townsend1987: 129). And since a lack of resources is not the sole reason for being deprived, one is no longer required to focus only on deprivations that are responsive to variations in resources (Veit-Wilson, Reference Veit-Wilson1987).
The concept of deprivation is multidimensional; though, as with the capability approach, the exact dimensionality remains to be specified. The question of dimensionality is, we must note, an important one for all multidimensional approaches, whether their focus is poverty, social exclusion, deprivation or the capability approach. However, Nolan and Whelan (Reference Nolan and Whelan1996: 71) argued that the question of choice and constraint was overlooked by Townsend, and that the concept of deprivation should focus only on deprivations that are enforced. If this were conceded, then the concept of deprivation would bear important similarities to that of capability deprivation. It is not clear that this is universally accepted, however, and Townsend (Reference Townsend1987: 139) was not wrong when he suggested that the concept of deprivation was ‘not yet being treated very coherently’.
The capability approach offers a framework for poverty analysis which prioritises capabilities (ends) over resources (means), adopts a multidimensional perspective and takes a broad focus on the constraints that may restrict human lives. It is argued here that the capability approach can provide additional coherence to the concept of deprivation, not because the concept of capabilities should replace that of deprivation, but because the concept of deprivation should focus on people's capabilities.
Operationalising the capability approach
While the focus of this paper is primarily conceptual, in this section we focus on empirical applications of the capability approach – an important issue given some doubts about whether the approach can be operationalised at all. Questions regarding what the relevant dimensions should be, and by what process these should be decided have loomed large in the capability literature. Sen himself has famously refused to endorse a fixed list of capabilities, arguing for the importance of democratic debate in deciding the important dimensions (for example, Sen, Reference Sen2009).
In the UK, Burchardt and Vizard (Reference Burchardt and Vizard2011) have adopted such a deliberative approach in their capability-based framework for monitoring equality and human rights in England, Scotland and Wales. They drew on a ‘minimum core’ of dimensions from existing human rights frameworks, and subsequently engaged in deliberative consultation with the general public and with groups at risk of discrimination and disadvantage in order to refine and expand on this core. Their final list of dimensions covers life; physical security; health; education and learning; standard of living; productive and valued activities; participation, influence and voice; individual, family and social life; identity, expression, self-respect; and legal security. This is undoubtedly a broad focus and is an important example of a consultative process being used to determine a capability list, but their indicators are not taken from any single data-set – a normal requirement for poverty analysis in order to explore joint distributions and so forth. This points to the demanding nature of requiring information about multiple dimensions of deprivation to be collected from a single survey.
However, not all authors have pursued such a deliberative approach. Martha Nussbaum (Reference Nussbaum2000) has provided a list of ten, philosophically derived capabilities that, she argues, all citizens have a right to demand from their governments. These capabilities are life; bodily health; bodily integrity; senses, imagination and thought; emotions; practical reason; affiliation; other species; play; and control over one's environment (2000). Anand and colleagues have attempted to operationalise Nussbaum's list of capabilities by fielding their own UK-based survey (e.g. Anand et al., Reference Anand, Hunter, Carter, Dowding, Guala and van Hees2009), which, they suggest, demonstrates the ‘feasibility although non-triviality’ of using the capability approach to support empirical analysis.
Others have used the approach to motivate a concern with a range of non-resource based capabilities. For example, Brandolini and d'Alessio (Reference Brandolini and D'Alessio1998) used the approach to support a multidimensional poverty analysis focussing on health, education, employment, housing, social relationships and economic resources; Kuklys (Reference Kuklys2005) focussed on outcomes in health and housing, while Bonvin and Dif-Pradalier (Reference Bonvin and Dif-Pradalier2010) have emphasised the importance of the capability for work and the capability for voice. Taken together, these studies are welcome as they point both to the ability of the approach to support empirical analysis and to extend the focus of such analysis beyond resource-based dimensions such as material deprivation.
However, Robeyns (Reference Robeyns2005: 208, see also Robeyns and van der Veen, Reference Robeyns and van der Veen2007) has questioned whether it is useful to operationalise the approach by drawing on Nussbaum's capability list when engaging in comparative quality of life assessment, since Nussbaum's approach was intended to be a philosophical account of constitutional entitlements. And this returns us to an important point – for the capability approach, as we have noted, specifies the ‘informational space’ of analysis (functionings and capabilities) prior to identifying the specific context of study (e.g. poverty analysis). But the relevant dimensions for any particular analysis are likely to be highly context-dependent.
The capability approach is not a field of study (‘capability studies’), but is a framework that can provide a lens for poverty analysis which emphasises its normative or ethical dimension. It stresses the intrinsic importance of people's capabilities (as ends) as opposed to the instrumental importance of their incomes (as means); argues for the importance of multidimensional assessment in poverty analysis and adopts a broad perspective of the many kinds of constraints that can limit people's lives. This focus can help to ensure that the indicators employed ‘have a clear and accepted normative interpretation’ (Atkinson et al., Reference Atkinson, Cantillon, Marlier and Nolan2002: 21), a desirable feature of any analysis. Thus, in a capability framework for poverty analysis, the normative lens would come from the approach, while the precise dimensionality would be decided contextually, with reference, inter alia, to existing poverty research. Seen in this light, a capability framework can contribute – in a practical way – to the growing focus on multiple deprivation within social policy (e.g. Whelan et al., Reference Whelan, Lucchini, Pisati and Maître2010).
At present, such a framework might be only partially operational because of current limitations in existing datasets, and because the exact dimensionality of the approach is under-determined. But this is true of all multidimensional approaches, capability-inspired or otherwise, and unless more limited approaches, such as the low income or material deprivation approach, act as good proxies for multidimensional poverty and deprivation, then some important information is lost by the omission of wider dimensions. Indeed, establishing which important dimensions we cannot currently measure can be of some practical good – see, for example, the work of Alkire and colleagues to develop survey modules for ‘missing dimensions’ in multidimensional poverty analysis for quality of work, agency and empowerment, physical safety, the ability to avoid shame and psychological well-being (see Alkire, Reference Alkire2007 for an overview).
A more systematic approach to multidimensional analysis, drawing on the capability approach, can help to ensure that appropriate indicators are employed and can improve the coherence of analysis by minimising inclusion and exclusion errors (to borrow a concept from a rather different setting; see Cornia and Stewart, Reference Cornia and Stewart1993). Inclusion errors refer here to the inclusion of dimensions in empirical analysis that do not fall within our conceptual framework; it is argued that developing an explicit multidimensional framework can help to guard against inclusion errors. Exclusion errors, on the other hand, refer to the failure to include conceptually important dimensions in our empirical analysis, and may be at least partially unavoidable where one is relying on secondary datasets. Nonetheless, as we have noted, identifying which dimensions are ‘missing’ can be the first step towards rectifying their omission.
A capability framework for poverty analysis would provide the basis for assessing social and economic changes, both within and between countries. However, Dean (Reference Dean2009) has questioned whether the approach can provide an adequate assessment of capitalist societies because, he argues, it ignores systemic injustices created by capitalism itself, such as the exploitation of labour. Human capabilities are the product of social relations and may be structurally unequal. They are also in constant tension and conflict, since exercising my capabilities may serve to limit those of others (2009: 273).
As an assessment framework, the approach does not, in itself, provide a critique of capitalism, although a capability assessment should provide information for such a critique to be constructed. In focussing on what people can do and be (human ends) and not on their resources (means), and in focussing on both monetary and non-monetary dimensions and constraints, the approach can help to evaluate the outcomes of capitalist society in terms of the unequal capabilities that result.
However, Dean is right to point to the interdependency of human capabilities and to the questions of power and conflict that inevitably arise (see also Deneulin and McGregor, Reference Deneulin and McGregor2010). While a focus on such interdependency has perhaps not received the attention it deserves in the capability literature, it has not been entirely neglected either – for example, Sen has discussed the systematic deprivation of women's capabilities in male-dominated societies (for example, 2009: 244–5). The point, from a capability perspective, is that in assessing the extent to which capitalism is commensurate with human development (Dean, Reference Dean2009: 272), in terms of both its outcomes and its processes, our evaluation should focus on people's capabilities.
The capability approach: insights for poverty analysis
The capability approach can provide a framework for poverty analysis that overcomes some of the central problems with existing traditions of analysis. The direct measurement of poverty has increasingly focussed on a narrow concept of material poverty. This emphasis on resources as being the core of the poverty concept has limited not only the constraints that are considered (solely a lack of resources) but has also limited the indicators themselves to marketable items. We have argued that the subdivision of constraints is normatively arbitrary and the narrow measure that results ignores too much that should concern us about the problem of poverty. While social exclusion is often seen as complementing the concept of poverty, its lack of coherence renders it deeply problematic. It remains unclear whether social exclusion is an outcome or a process and for whom it is bad. This incoherence in terms of conceptualisation is mirrored in measurement, where indicators of disengagement are routinely interpreted as implying exclusion. Finally, the concept of deprivation, at least in its less popular but broader incarnation, can be seen to bear important similarities with the concept of capability deprivation. However, the importance of constraints to the concept of deprivation again lacks clarity, and, to be useful, the concept requires greater coherence. It is argued here that the capability approach can act as a framework for conceptualising poverty and deprivation.
The preceding discussion provides six important insights for poverty analysis. First, the capability approach questions whether we can be neutral between direct and indirect approaches to poverty analysis. In emphasising the intrinsic importance of people's capabilities over the merely instrumental importance of their resources, the approach focusses on those who have impoverished lives, and not just depleted wallets (Sen, Reference Sen2000: 3).
Second, the concepts we employ – and not just that of poverty – should reflect deprivations that are enforced and not voluntary non-participation. While on the surface this may seem self-evident, it would rule inadmissible many of the indicators used to measure social exclusion and would provide greater clarity to the concept of multiple deprivation.
Third, a lack of resources cannot be the only constraint of interest for poverty analysis. If we believe that the indicators we employ hold normative weight, then it is a nonsense to suggest that their absence is of serious concern if because of a lack of resources, but of no concern if as a result of other constraints (e.g. disability or discrimination). The ability to participate in the life of society, for example, does not derive its importance from its responsiveness to resources and nor, if we believe it to be important, can a lack of resources be the only constraint we are interested in, for there is a normative distinction between choice and constraint (Le Grand, Reference Le Grand1991) in a way that there is not between one type of constraint and another.
Fourth, the capability approach emphasises the multidimensional nature of poverty analysis, with a broader remit than focussing solely on ‘material’ poverty. There is, at present, a tension between a desire to reflect the multidimensionality of poverty (e.g. Atkinson et al., Reference Atkinson, Cantillon, Marlier and Nolan2002: 79) and an insistence on preserving the relationship between resources and the concept of poverty (e.g. Nolan and Whelan, Reference Nolan and Whelan1996). A broader focus is required because some of the most vulnerable members of society experience circumstances that may not immediately be due to a lack of resources: homeless people, drug and alcohol addicts, functional illiterates and those who have suffered physical or mental abuse (Volkert, Reference Volkert2006), for example.
It is sometimes suggested, however, that a resource-based approach can act as a proxy for multidimensional deprivation since resources are an all-purpose means (Rawls, Reference Rawls1971). But, as we have noted, research examining the relationship between indirect and direct measures of poverty has consistently emphasised that they identify different people as being in poverty (e.g. Bradshaw and Finch, Reference Bradshaw and Finch2003).
Fifth, the capability approach offers a way to reconcile the tension between narrow and broad approaches to poverty by respecting the former, while not losing sight of the latter. Lister (Reference Lister2004: 18) has argued that to conceive of poverty as capability-failure is to conflate poverty with an altogether broader notion such as quality of life or well-being, because not all ‘ill-being’ is related to a lack of resources.
The ‘normative focus’ of the capability approach seeks to identify the ‘informational space’ of analysis and is more concerned with delimiting this normative terrain than with deciding whether and/or how this terrain may be subdivided into constituent concepts. Thus, ‘poverty’ could retain its narrower meaning, with a lack of resources at its core, if the essential additional terrain stressed by the capability approach were taken up by the concept of (multiple) deprivation. Thus, we can concede Lister's critique, but in a way which emphasises the normative inseparability of the concepts of poverty and deprivation.
Sixth, a capability framework for poverty analysis would draw both on the capability approach and existing poverty analysis within social policy. The contribution of the capability approach is its normative focus – in prioritising capabilities (ends) over resources (means), in adopting a multidimensional perspective and taking a broad view of the constraints that restrict human lives. But the specific dimensionality of such a capability framework would be decided contextually with reference, inter alia, to the existing literature on poverty analysis within social policy.
The ways in which lives may be blighted by poverty and deprivation are many, and those who seek to tackle these twin evils require an analytic framework built on steady foundations. In focussing on intrinsically important capability deprivations rather than some convenient proxy such as income, the capability approach can provide such foundations, not only to understand poverty and deprivation – but also to combat them.
Acknowledgement
The author thanks Tania Burchardt, David Piachaud, Howard Glennerster, Eunica Aure, three anonymous referees, and the Editor for their helpful comments and suggestions. Thanks are also due to Julian Le Grand and Jane Lewis for providing feedback on an earlier iteration of some of the material presented here, and participants at the 2009 Social Policy Association conference in Edinburgh, at which a previous draft of the paper was presented. The errors that remain are my own. This work was supported by the Economic and Social Research Council [grant number ES/G01808X/1].
