Inequality, disadvantage and the
Bridging conceptual framework and empirical analysis
Rod Hick, Cardiff University
The aim of this short paper is to examine the prospects and problems of using the capability approach as a framework for understanding disadvantage and inequality. In preparing the paper, I was asked to focus on these issues specifically as they relate to empirical analysis. This, as it turns out, was a welcome steer, for it touches on what I view as one of the central issues for the capability approach. That is, the capability approach is a conceptual framework, not an empirical one, and some of the central questions that arise for the capability approach relate to the challenges of bridging from this conceptual framework to empirical analysis in a given context.
In this paper, I make three key points:
The capability approach provides a useful conceptual framework for understanding poverty and inequality, but it is even more underdeveloped than typically assumed;
The multidimensional perspective provided for by CA provides distinctive answers to some, but not all, questions;
Applying CA to study inequality (and not just poverty) opens up potentially interesting avenues of enquiry, but raises both old and new challenges in terms of operationalisation.
The paper discusses these issues in turn before reflecting briefly on what the capability approach might bring to the analysis of the ‘welfare services state’.
2 What kind of approach is the capability approach?
In this first section, I want to consider what kind of approach the capability approach is. In doing so, I draw on a major new text in the field – namely, Ingrid Robeyns’ recent book Wellbeing, Freedom and Social Justice: The capability approach re-examined (Robeyns 2018).
One reason why Robeyns’ book is significant is that presents us with a distinctive genealogy of the capability approach. She argues:
‘It is a mistake to understand the capability literature as a field with two main thinkers [Sen and Nussbaum] who have each proposed one version of the capability approach, which have then inspired the work of many other scholars. Rather, there is only one capability approach which is a generalisation of the work of Sen together with further developments by many others’ (Robeyns 2017, p. 79-80, emphases in original).
Elsewhere, Robeyns notes that she ‘will propose a definition and an account of the capability approach that does not exactly equal Sen’s but rather can be interpreted as a generalisation of Sen’s definition’ (2018, p. 7-8, emphasis in original). This emphasis on ‘generalisation’ matters, I believe, because it implies that we cannot even take all elements of Sen’s articulation of CA as being necessary. An example of this can be observed in the emphasis by Sen on those capabilities that one ‘has reason to value’. Robeyns tells us that this clause is not essential, but rather that it is ‘a special case of the general definition of the capability approach’ (p8 fn1).
Robeyns presents a modular approach to understanding CA, containing essential and optional components (modules) but the point in relation to the ‘reason to value’ clause is significant because it means that even within Sen’s own work there is a mixture of essential and optional components, and these need to be parsed in order to identify the essential core of the capability approach.
A second point relates to the modular approach itself. I am sympathetic to the suggestions that there are core and optional elements within the capability approach, though not always to the division that Robeyns proposes between these elements. For example, Robeyns places the account of agency in Module B (“Non-optional modules with optional content”). She tells us that ‘this doesn’t mean that “anything goes” in terms of the choice of the content, but it does mean that within each module, there is a range of options to choose from’ (2018, p. 59). This makes the account of agency sound as if it were fundamental to the approach. But she then goes on to tell us that some quantitative applications, ‘for example, when one wants to investigate the correlation between an income metric and some achieved functioning’ (2018, p. 64), might not need such an account and that there may be ‘good reasons’ why an account of agency might be absent. This suggests to me that the truly necessary components of the approach might be fewer than she claims.
Furthermore, in previous work, I have argued that empirical demands must be taken more seriously by scholars working with CA, and that pragmatic, empirical considerations should not be viewed as deviations from the approach. For example, in relation to my own field [poverty analysis], Robeyns argues that:
‘One could aim to work on multidimensional poverty analysis and highlight the fact that we should be interested in the combination of achievements that people are able to have. This would point at two important insights in the capability approach – namely its multidimensional character, as well as focussing on opportunity sets rather than on outcomes. But if the opportunities one focusses on are not capabilities, but rather opportunities to access certain bundles of commodities, then it would be an unjustified inflation to call this a capability application; rather, it would be another opportunity-based multidimensional poverty measure’ (Robeyns 2018, p. 78, emphasis in original; subscript is mine).
I do not believe this is correct, and moreover I argue that the nature of the error says something of significance about the capability approach. As argued elsewhere (Burchardt and Hick 2018; see also Hick 2016c, p. 52), capability scholars have been prone to using the verb “to focus” to elide the distinction between conceptualisation and measurement. And as noted above, the capability approach is a conceptual framework, not an empirical one. I agree that, in conceptual terms, capability scholars will understand wellbeing in terms of capabilities and functionings. But empirical application must be guided by measurement considerations as well as theory, so if resource-based measures are better proxies for capabilities than the ‘direct’ measures that may be at our disposal, then it is not at variance to the capability approach to select these as the basis for empirical analysis.
This has two implications which are, I believe, are significant. The first is that we need to take care that the theorising of the capability approach does not infringe to too great an extent on the practical business of measurement. There is indeed some distance between conceptualisation and measurement in all approaches; this must be recognised, and bridging this gap may require compromises. The second is that the additional information that is required to supplement the capability approach may not only come from additional theories, but also from empirical considerations.
Third, the deliberately incomplete nature of the capability approach leads to a problem of authorship. One issue for the capability approach is that it is not always clear where the balance lies between that which comes from Sen and that which comes from the author(s) of a particular application.
The capability approach, then, risks becoming a game of “Sen says”, for two reasons. The first is that Sen has outlined many ideas and concepts. As Robeyns explains (rightly, in my view), while some of these are core principles and are thus to the capability approach, others are optional extras or even just examples used to make a particular point. For example, in the field of poverty analysis Sen’s supposed emphasis on shame is sometimes claimed to be central feature of the approach (e.g. Walker et al. 2013). In my view, the discussion of shame in Poor, Relatively Speaking (Sen 1983) was just an example, to try to support the claim that relativity in terms of resource requirements might be consistent with absolute performance in terms of capabilities. A focus on shame as being one of the functionings of interest is in my view is consistent with, but not required by, the capability approach. In short, not everything Sen says is fundamental to the capability approach.
The second reason is that, because of the indeterminacy of the approach, we, the ones who apply the capability approach in any given context, must make – and justify – numerous decisions. As Goerne argues (2010, p. 10):
‘A common rhetoric is to present results or recommendations “from a capability perspective”, suggesting that the use of the CA (i.e. the terminology of the five building blocks [see below]) itself would make certain conclusions necessary – although the findings depend as much on the chosen normative reference point…. To consider fulfilment through work, for example, as a valuable functioning may make perfect sense. It is the author's choice, however, to do so – and not an obvious choice from the perspective of the CA’.
In short, Sen does not say all that is attributed to him.
3 How should we understand the essence of the capability approach?
Illustration 1 outlines what Alexander Goerne (2010, p. 7) describes as the five building blocks of the capability approach. These, to my mind, represent the essential core of the capability approach. Why? Because there is a clear normative linkage between the building blocks. Why do we evaluate capabilities or functionings as the ends of justice instead of resources or commodities? Because of ‘conversion factors’, which capture variations in people’s needs. Why not utility? Because of adaptive preferences, which can lead us down the road of ‘physical condition neglect’.
Source: Goerne (2010: 7)
Unlike these building blocks, many of the other concepts found within the capability literature are not justified or motivated in the same way. Why, for example, must we focus on the capabilities we “have reason to value”? This is, to my mind, not similarly justified. Thus, the core components of the capability approach are the claim that capabilities and functionings should the core concepts when evaluating well-being, and the critique of income/commodities and utilitarian measures. Many of the other concepts: positional objectivity, well-being and agency freedom, ‘reason to value’ are not similarly justified, though they may prove useful in particular applications of the approach.
From this discussion, I take the following points: (i) that the essential core of the capability approach is, in fact, thinner than is often claimed; (ii) that this implies that there is substantial supplementation needed to take concepts provided by the capability approach and to apply them empirically in any given context; and (iii) that this creates to a problem of authorship, where we are at risk of blurring the lines about what comes from CA and what comes from us, as analysts.
4 Poverty as Capability Deprivation
One of the main strands of work within the capability literature has been to apply the approach to conceptualise and measure poverty multidimensionally. In one influential, recent text, Alkire et al. (2015, p. 1) has presented a CA-inspired definition of poverty as being ‘a condition in which people are exposes to multiple disadvantages – actual and potential’.
In previous work, I have sought to articulate an alternative, capability-based approach to understanding poverty which I label Poverty as Capability Deprivation (see Hick 2014). Rather than viewing poverty as simply the accumulation of disadvantages, I have defended the idea of two concepts of material poverty and multiple deprivation, on the grounds that the questions that the capability approach and poverty analysts typically ask are distinct. While those who conceptualise poverty ask ‘what is poverty?’, the question the capability approach sets itself is closer to ‘what matters?’ or ‘what is of value?’. By insisting on analysis of two concepts of material poverty and multiple deprivation, we can respect the resource-centric nature of the concept of poverty without losing sight of the wider questions of deprivation in non-economic domains (for a longer discussion, see Hick 2012, p. 306).
In this framework, material poverty is defined as ‘inadequate material living standards arising from a lack of resources’ (Hick 2014, p. 307) and is intended to capture economic deprivation. The concept of multiple deprivation is defined as ‘the enforced experience of low living standards’ (Hick 2014, p. 310), and this is intended to capture multiple non-economic dimensions of disadvantage. Both of these substantive concepts, I argue, should be understood using the conceptual categories of capabilities and functionings.
But which substantive dimensions should we consider within these concepts of material poverty and multiple deprivation? In considering this, I return to the defence against the capability approach made by John Rawls’, who favoured ‘primary goods’ as his metric of justice, defined as “rights and liberties, opportunities and powers, income and wealth” (Rawls, 1971, p. 92).
Rawls’ asserted due to the varied conceptions of the good that people held, considerations of justice could only relate to the means by which such conceptions might be pursued; otherwise the state would be forced to prioritise one set of ends over another. Primary goods, then, are “things which it is supposed a rational man wants whatever else he wants” (1971, p. 92), because “whatever one’s system of ends, primary goods are necessary means” (Rawls 1971, p. 93). As I argued then (Hick 2014, p. 311):
‘To focus on ends themselves will only violate neutrality between the ends which people hold (the “fact of pluralism”, Rawls 1988, p. 259) if there do not exist “primary goals”—ends which each person shares, whatever their conception of the good and whatever else they value’.
The legitimate focus of multidimensional poverty analysis, then, is on these primary goals, or what might be labelled core capabilities, though we may investigate these using a range of approaches (including a theory-derived approach or by including some element of public participation).
What seems certain, however, that the multidimensional perspective provided for by the capability approach implies greater complexity than is required by an income-centric approach. One challenge to those who advocate such a perspective might be that this complexity is not required, and that the results provided for by income-centric and capability-inspired approaches are effectively equivalent. To examine this, I have conducted analysis of the relationship between two measures of material poverty, low income and material deprivation, and seven dimensions of multiple deprivation, (i) general health, (ii) mental health, (iii) housing deprivation, (iv) autonomy, (v) life satisfaction, (vi) financial stress and (vii) unemployment. The aim was to examine the extent these measures identified the same people as being poor/deprived, considering identification for (a) both individuals/households and groups, and (b) multiple dimensions individually and as their collective constructs of material poverty and multiple deprivation.
The results from this analysis showed that ‘although material poverty and multiple deprivation identify very different individuals, they display greater congruence in terms of identifying vulnerable groups, especially where aggregate measures are used (Hick, 2016: 277). Thus,
‘there are novel insights to be gained from the multidimensional perspective that cannot be obtained by relying on measures of material poverty alone. Adopting a multidimensional perspective does influence who we identify as being poor. However, the distinctiveness of the multidimensional perspective in empirical terms is not an all-or-nothing affair: it depends significantly on whether we analyse aggregate or disaggregate measures and on whether our interest lies in identifying a vulnerable individual or vulnerable groups – two distinct, but important, tasks for public policy’ (Hick 2016, p. 307, emphasis in original).
As we see in Figure 1 reproduced below, when we aggregate to ‘material poverty’ and ‘multiple deprivation’ more broadly, and when we focus on the experience of these for particular groups, we find that one predicts the other to a very high degree of accuracy.
Source: Hick (2016: 299)
What this implies more generally is that we should not expect the capability approach to lead to distinctive answers to each and every question we might seek to ask. This distinctiveness of the this multidimensional poverty analysis over and above a narrower measure based on material poverty alone was dependent on the type of identification considered. Understanding to a greater extent the areas where CA is likely to add value, and those where it does not affect the substantive conclusions, strikes me as representing important area of future work.
5 Inequality and the capability approach
In recent work with Tania Burchardt (Burchardt and Hick 2018), we sought to move from a concern solely with poverty to asking what the capability approach might be able to contribute to the study of inequality, and of advantage in particular and, correspondingly, what lessons might be learned for the capability approach by seeking to apply it in this new terrain.
We started by noting that arguments which motivate CA in terms of the study of poverty also apply when we are to consider poverty. For example, we similarly expect different needs (conversion factors) to apply in terms of converting resources into functionings and capabilities. Thus, the claim that we should focus on ends and not means in terms of evaluating inequality still seem applicable in this context. A second advantage of drawing on the capability approach when considering inequality is in terms of its focus on multidimensionality. Much of the recent literature on inequality is limited to the study of economic inequality, and especially to inequalities of income (and, to a lesser extent, wealth). Adopting a capability perspective would enable analysts to ‘bring in’ non-economic dimensions into the analysis and, moreover, to examine the extent to which advantages in one domain correlate with those in another (corresponding to the literature examining the coupling of disadvantages within poverty analysis, see Hick 2016b). Finally, we argued that the freedom focus of the capability approach or, in weaker terms, the distinction between functionings and capabilities would also be of use when considering advantages. One reason for this is that some advantages are better thought of as possibilities or options rather than observable outcomes. We argued that the nature of advantage might be found ‘in political influence, in geographical mobility, their room for legal manoeuvre, in security and in access to knowledge and information. Crucially, they do not necessarily need to actualise these freedoms in order to secure advantage – the capability is often sufficient (Burchardt and Hick 2018: 44, emphasis in original).
But applying the capability approach in terms of the field of inequality brings new challenges too. These relate primary to measurement and to the challenge of operationalising the approach, which we argued was more severe when examining advantage as opposed to disadvantage. But our claim is that these empirical challenges say something about the nature of the approach itself.
We also identified three issues that arise when trying to apply the capability approach for this new purpose of studying inequality, and advantage in particular. The first of these is the greater complexity when trying to apply the approach in this context. Without “basic capabilities” to cling to, the dilemma about how capabilities should selected and prioritised was, to our mind, more serious, since the ends that advantages individuals would pursue might be extremely heterogeneous indeed. The second is that this might lead to a retreat from functionings and capabilities in terms of measuring advantage, since capturing top-end achievement in capability space is, at the present moment, rather lacking. This may require focussing on resource metrics if our interest is in what is happening in the top end of the distribution of advantage, even if we believe advantage should be understood in terms of what people can do and be (that is, in terms of functionings and capabilities). And if this presented a ‘defensive’ reason to continue to place some emphasis on income and wealth in terms of measurement, then our third argument is that some forms of advantage are hard to understand without reference to resources. If we are concerned about some of the attributes of advantage – say, inheritance or tax evasion, then it matters how much we are talking about. While our ultimate concern with these is on the capabilities they may (unjustly) enable, they may be difficult to measure in the space of capabilities or functionings, not least because wealth is sometimes stored precisely to enhance one’s functionings in the future and thus may not be observable in the present moment (see Burchardt and Hick 2018 for a longer discussion).
Our conclusion, then, is that some the core components of the capability approach present the opportunity to understanding in terms of inequality. But, at the same time, the empirical challenges were arguably more significant than in terms of the study of poverty. Importantly, and contrary to Robeyns’ line of argument, we do not believe that pragmatic choices should be viewed as inconsistent with the capability approach. Rather, our understanding of what the capability approach is must allow space for empirical and measurement considerations to be taken seriously.
6 Applying the capability approach to consider the significance of the ‘welfare service state’
Thought it is not the primary subject of this paper, we might also consider what contribution the capability approach might make to the study of the ‘welfare service state’. In doing so, I want to take Jane Lewis’ (1992) in her article Gender and the development of welfare regimes as a reference point, and consider whether the conclusion she reached in that analysis would be altered by adopting a capability perspective.
In that article, Lewis examined policy logics for women in each of Esping-Andersen’s three worlds of welfare capitalism, selecting Ireland and the UK, France, and Sweden as exemplars for the Liberal, Corporatist and Socialist worlds, respectively. When women entered the paid labour market in the UK and Ireland they did so on the same terms as men, with publicly-provided care for dependent relatives absent. This in turn led to a preponderance of part-time working and to women typically acting as secondary earners. In Sweden, women were not only facilitated to enter the paid labour market – the linkage between social entitlements and labour market participation assumed this would happen: women were ‘”forced’ into the labour market”, in Lewis’ words, and occupied a better position within the labour market than their British counterparts. France, by contrast, seemed to offer the greatest freedom of choice on behalf of women to prioritise roles as workers or mothers.
Lewis’ argument not only questioned decommodification as the dependent variable when considering the welfare state, relying on the male breadwinner model as an alternative criterion, but this was seen to lead to a new conclusion, one which questioned the Swedocentrism of welfare state research. But Lewis’ analysis can also be read through a capability lens. Her concern is with the (constrained) choice that women to prioritise their roles as workers and/or mothers. Her analysis is not only interested in this binary, or indeed solely on the extensiveness of the choice they had, but on the quality of the options at hand to women.
Lewis’ conclusion is, however, indeterminate (or, perhaps, partial, to use Sen’s term): ‘while it is impossible to come to any definitive conclusion as to where women “do best”, both France and Sweden would seem to offer women more than Britain’ (Lewis 1992: 162). Why does Lewis not find that France is preferable to Sweden? After all, it appears to offer greater freedom of choice between these roles. This is not articulated explicitly, but it appears to be that Sweden does not concentrate women in precarious employment (thus, relating to the quality of the capability in relation to work), whereas she notes in relation to the French labour market that ‘the percentage of French women working part time has increased significantly due to conscious efforts to restructure the labour market’ (Lewis 1992, p. 167).
Might the capability approach provide the basis for greater certainty? Might it favour the greater freedom of choice afforded to women in France? It is tempting to believe that this might be the case, but I believe the answer to be “no”, for the analysts working within a capability approach would equally have to consider the quality of the capabilities on offer.
My conclusion, then, is also indeterminate. Adopting a capability perspective might not lead to a novel conclusion here. And, and equally of importance, if one did arrive at a distinctive conclusion starting from a capability perspective, it might well be because of the supplementary considerations added by us – the analysts – rather than anything that Sen says.
We should not be blinded by the deployment of the terminology of functionings and capabilities, and the many concepts invoked by Sen, to see that, without supplementation, the capability approach is a “thin” framework. Moreover, just as in the study of multidimensional poverty, the capability approach would not I believe result in distinctive answers to all questions and different solutions to all problems. Key questions, then, for those who seek to understand the emerging welfare service state, I suggest, are: what, if anything, are the distinctive aspects of a capability-inspired understanding of the welfare service state? And what possibilities are opened up by drawing from the capability toolbox to examine contemporary welfare state developments which, while perhaps not entirely distinct to other approaches, might nonetheless serve to shift attention, priorities and emphasise different values?
7 Conclusions: Bridging conceptual framework and empirical analysis
There are, then, three conclusions. The first is that while the ‘deliberately incomplete’ nature of the capability approach is widely known, I have argued that is even more underdeveloped than is often assumed. While Robeyns’ distinction between the essential and non-essential components of the approach is, in my view, welcome, and may contribute towards ironing out some of the many confusions that surround the approach, I have argued that the essential core may be even narrower than she suggests. This is in part because I would categorise some of specific elements of her modular approach differently. But also, and more fundamentally, a consequence of understanding the capability approach as a conceptual, and not an empirical, framework, is that in seeking to apply the approach – that is, in bridging from conceptual framework to empirical analysis – many decisions are required, which entails supplementation. It is we, the analysts, who are often building these bridges, and this ought to be explicit.
The second is that the capability approach has, I have argued, contributed to the understanding of inequality and poverty in important ways, opening up windows onto aspects of analysis that have sometimes been neglected by other approaches. Two important ways it has done this is by emphasising multidimensionality and, relatedly, by bringing in non-economic domains, which are fundamental in terms of understanding our quality of life, but are sometimes neglected in the literature on poverty and inequality.
The final conclusion, then, relates to the prospects of drawing on the capability approach for understanding what has been labelled the emerging welfare service state. On the one hand, the capability literature provides a rich array of conceptual tools. While I have argued that not all of these are essential components of the approach, they may nonetheless be very helpful in terms of seeking to understand and critique contemporary policy developments. But, on the other, existing work in this area is not limited to economic domains and already considers impacts across multiple dimensions, as our discussion of Lewis’ work has shown. Thus, two significant “advantages” that the approach has brought to the study of poverty and inequality might not be so clearly apparent in this new policy terrain. Negotiating the valued ends to be the subject of analysis will continue to be crucial. And this, in turn, raises questions about whose valuation we are relying upon, and on what concepts we want to draw upon in constructing our analysis. As Sen says little about these matters, it is up to us to take the building blocks provided by the capability approach and to construct a framework to analyse and critique the emerging welfare service state.
Alkire, S., Foster, J., Seth, S. Santos, M. E., Roche, J. M., & Ballon, P. (2015). Multidimensional poverty measurement and analysis. Oxford: Oxford University Press.
Burchardt, T., & Hick, R. (2018). Inequality, advantage and the capability approach. Journal of Human Development and Capabilities, 19(1), 38 – 52.
Goerne, A. (2010). The capability approach in social policy analysis: yet another concept? (REC-WP Working Paper 2010/3). Edinburgh: RECWOWE Publication, Dissemination and Dialogue Centre.
Hick, R. (2016). Material poverty and multiple deprivation in Britain: the distinctiveness of multidimensional assessment. Journal of Public Policy, 36(2), 277-308.
Hick, R. (2016b). The coupling of disadvantages: material poverty and multiple deprivation in Europe before and after the Great Recession. European Journal of Social Security, 18(1), 2-29.
Hick, R. (2016c). Between income and material deprivation in the UK: In search of conversion factors. Journal of Human Development and Capabilities, 17(1), 35-54.
Hick, R. (2014). Poverty as Capability Deprivation: Conceptualising and measuring poverty in contemporary Europe. European Journal of Sociology, 55(3), 295-323.
Hick, R. (2012). The capability approach: Insights for a new poverty focus. Journal of Social Policy, 41(2), 291-308.
Lewis, J. (1992). Gender and the development of welfare regimes. Journal of European Social Policy, 2(3), 159-173.
Rawls, J. (1971). A theory of Justice. Cambridge, Massachusetts: Harvard University Press.
Robeyns, I. (2018). Wellbeing, freedom and social justice: The capability approach re-examined. Cambridge, UK: Open Book Publishers.
Sen, A. (1983). Poor, relatively speaking. Oxford economic papers, 35(2), 153-169.
Walker, R., Kyomuhendo, G. B., Chase, E., Choudhry, S., Gubrium, E. K., Nicola, J. Y., Lodemel, I., Mathew, L., Mwiine, A., Pellissery, S., & Ming, Y. (2013). Poverty in global perspective: is shame a common denominator? Journal of Social Policy, 42(2), 215-233.
Rod Hick, PhD.
Cardiff University, United Kingdom
School of Social Sciences