Cambridge University Press
9780521862875 - The Capability Approach - Concepts, Measures and Applications - Edited by Flavio Comim, Mozaffar Qizilbash and Sabina Alkire
Excerpt
SABINA ALKIRE, MOZAFFAR QIZILBASH AND FLAVIOCOMIM
Amartya Sen’s capability approach has generated remarkable interest in recent years. This volume brings together a selection of papers initially presented at an international conference on the capability approach (CA) held at St Edmund’s College, Cambridge in 2001. This conference marked an important turning point in research on the capability approach. It brought together many young scholars who were interested in the approach as well as others who had been working on it for some time. The conference was initially motivated by issues relating to the usefulness of the approach in the particular contexts of poverty and injustice. However, conference papers covered a wide range of topics relating to concepts, measurement and other applications. In this volume, the papers are categorised in terms of these broad and overlapping areas. In 2002 a follow-up conference explored Martha Nussbaum’s version of the approach, and annual conferences have been held in subsequent years.1 Numerous initiatives have since emerged, including the Human Development and Capability Association (www.hd-ca.org). In part as a result of these initiatives, but also quite independently of them, a large literature on the capability approach has emerged.
Amartya Sen’s 1980 Tanner lecture, ‘Equality of What?’, set out a broad agenda for debate and further research. While the approach has been extensively discussed, Sabina Alkire suggests in Chapter 1 that work in this area is still at a relatively early stage. The drawing on the cover of this volume – a version of Jean-François Millet’s ‘Les Premiers Pas de l’Enfance’ (‘The First Steps of Childhood’) – shows a child taking its first tentative steps, supported by her mother. Only time will tell whether this image provides an appropriate metaphor for this early phase of work on the capability approach. Part of the value of bringing together a set of papers in a volume of this sort is that these papers allow us to assess how far the approach has gone and to define – however tentatively – potential directions for work on the approach. The volume brings together a diverse set of voices, each of which engages with the approach in its distinct manner. However, we emphasise that many of the chapters engage critically with different aspects of the approach, freely questioning and wrestling with it. Indeed, such critical engagement is a common theme of this volume. We hope to bring out the flavour and nature of this engagement in what follows through reference to relevant chapters in this introduction.
At this stage, it is not entirely foreseeable which directions will be pursued in future work on the capability approach and how fruitful they will turn out to be. If we return to the Millet crayon drawing, part of what engages our attention is the unpredictability of the child’s first steps and the hope – and anxiety – that unpredictability generates. The steps of a child are powered by its unique curiosity, temperament and circumstances. Similar unpredictability is evident in the emerging literature on the capability approach. It is part of what makes this literature both intriguing and exciting. We hope that this volume will convey some of that excitement.
The central concepts involved in the capability approach are capability and functioning. Functionings are what Sen (1999: 75) calls ‘the various things a person may value being and doing’. Examples include being adequately nourished, being in good health, avoiding escapable morbidity, being happy, having self-respect, and taking part in the life of the community (Sen 1992: 39). There is no definitive list of basic functionings because different sets will be relevant to different groups and in distinct settings (Sen 2005: 157–160). A person’s capability ‘represents the various combinations of functionings (beings and doings) that the person can achieve’ (Sen 1992: 40). To this degree, the person’s capability reflects her freedom or (real) opportunities. Sen has used these concepts to analyse the quality of life, egalitarian justice and poverty inter alia. He has demonstrated the insights which arise from a capability or functioning-based analysis in comparison with analyses which exclusively use information on resources, or income, or ‘utility’ (when this is understood as happiness or desire satisfaction). The capability approach thus broadens the informational basis used in normative evaluations.
To illustrate some of these ideas, consider the quality of life of the painter Vincent Van Gogh, in the winter of 1889. At that time Van Gogh painted an interpretation of Millet’s ‘The First Steps of Childhood’.2 It is certainly true that Van Gogh had little income and that he was heavily dependent on his brother for financial support. However, if we considered his position only as regards income or resources we would have a very limited understanding of the quality of his life. In the months when he was working on this painting – as well as other paintings based on Millet’s work – he was extremely unwell and had recurrent fits. To this degree, he was clearly deprived in terms of Sen’s functioning ‘being in good health’. In addition, these paintings were created in the asylum of Saint-Rémy de Provence where he did not have people who could sit for portraits. As a consequence, his brother Theo sent him some black and white reproductions of works by Millet and Eugène Delacroix, which he worked from. Van Gogh’s choice of ‘The First Steps of Childhood’ as a subject reflected the limited opportunities or capability he had. His limited opportunities involved a form of disadvantage which may not be adequately captured through an analysis which merely checked his level of ‘utility’ (in terms of happiness or desire satisfaction), partly because he may have learned to adjust to the circumstances he found himself in.
Capability and functioning remain intimately connected but independently useful concepts in Sen’s writings. Because capability is a collection of functionings a person can achieve, capability is evaluated in the ‘space’ of functionings, thus functionings are integral elements of capabilities. However, the focus on capability directs our attention to freedom and opportunity – which functionings cannot do. Sen does not claim that capability is all that matters; functionings retain ongoing value in themselves. He also leaves open the relative importance of capability as opposed to functionings as well as the relative weights to be given to different capabilities or functionings (Sen 1992: 49–53 and 1999: 76–77). These are some of a range of ways in which the approach is intentionally open-ended and incomplete.
In addition to capability and functioning, Sen defines a third core concept, agency. On his account, an agent is ‘someone who acts and brings about change’ (Sen 1999: 19). The agency aspect is important in assessing ‘what a person is free to do and achieve in pursuit of whatever goals or values he or she regards as important’ (Sen 1985: 203). In some writings, agency – as well as capability – figures centrally. For example, the approach adopted in Drèze and Sen’s book India: Development and Participation as well as many of Sen’s single-authored writings ‘puts human agency (rather than organisations such as markets or governments) at the centre of the stage’ (2002: 6). When Roland de Leeuw notes that Van Gogh initially had as a ‘social objective’ that his Millet paintings would be presented to a local school – presumably to expose young people to Millet’s work – it is agency which is relevant (de Leeuw 1996: 466). Of course, Van Gogh’s aim if realised would also mark an achievement in terms of functionings, as regards ‘taking part in the life of the community’.
Sen uses a range of distinctions in his writings on freedom and development. To clarify his conceptual framework and to avoid potential confusion, we introduce two further terms: ‘opportunity freedom’ and ‘process freedom’ (Sen 1999; 2002: chapters 19–21). While ‘opportunity freedom’ refers to what people have opportunity or ability to achieve, ‘process freedom’ refers to ‘the process through which things happen’ (Sen 2002: 585). Clearly capability is closely related to opportunity freedom; agency relates to personal process freedoms.
As might be expected given the richness of foundational concepts, several interpretations of the scope of the capability approach are used in the wider literature and indeed in this book. These can be charted between two poles: one narrow and one broad, with the broad subsuming the narrow. The capability approach proposes that the comparison or evaluation of advantage or deprivation (whether or not through measurement) should occur in the space of capabilities inter alia (rather than simply utility or commodities), or in some sensible approximation of capabilities such as a vector of achieved functionings. The narrow interpretation sees the approach primarily as identifying capability and functionings as the primary informational space for certain exercises. The broad interpretation views the capability approach as providing a more extensive and demanding evaluative framework, for example by introducing human rights or plural principles beyond the expansion of capabilities – principles which embody other values or concerns such as equity, sustainability or responsibility.
Both interpretations can be found in Sen’s writings. Like the narrow interpretation, the broad interpretation argues that the quality of life should be evaluated primarily in the space of capabilities. However, information on capabilities alone is not sufficient. Other considerations (such as rights, process or agency) would enter the overall evaluation of states of affairs in this framework. To illustrate, consider an example which Sen has used recently. The example starts from the well-known claim that in similar conditions women live longer than men. It might be possible, Sen suggests, to equalise people’s capability as regards their life chances. However, pursuing such equality, perhaps by discriminating against women in the distribution of health care, would violate process freedom (Sen 2002: 660–661 and 2005: 156; see also Tsuchiya and Williams 2005). On a narrow interpretation, this example can be used to illustrate the limits of the capability approach. By contrast, on a broad interpretation, the very same example might be used to show how the capability approach introduces additional distributional considerations (see also Sen 1985 and 2000). In both the narrow and broad interpretations, the capability approach is viewed as a tool for evaluation – comparing situations with respect to the real opportunities they offer, among other things.
Sen (1984, 1990 and 1999) also frames the objective of development as an ‘expansion of capabilities’. This has led to an interest in identifying courses of action or policies that would further this objective. So going beyond the capability approach as an evaluative space or framework, we can identify a third preoccupation in the literature on the capability approach and, relatedly, human development which focuses on generating prospective policies, activities and recommendations. This preoccupation is central to the discussion in the section on measurement and other applications later in this introduction. The chapters in this volume, nonetheless, span all three aspects of the literature.
Much of the philosophical literature is concerned with debates relating to the capability approach as an evaluative space and its relationship to, and perceived merits and weaknesses in comparison with, other approaches. Contributions have included a wide range of papers on justice, happiness, needs and opportunities.3 Chapters by Alkire, Mozaffar Qizilbash and Ingrid Robeyns engage critically with these issues. Alkire traces the boundaries of the approach and distinguishes evaluative and prospective aspects. Robeyns investigates the ability of the capability approach to address feminist concerns and shows that it can be seen as a ‘gender-sensitive evaluative framework’. She expresses a worry raised elsewhere in the literature about the ‘under-specified’ nature of the approach.4 Qizilbash considers the extent to which Sen’s approach contrasts with the views of happiness, poverty and gender justice in John Stuart Mill’s writings. He finds the two approaches remarkably similar in spite of the fact that one is a leading critic, while the other is one of the founders, of utilitarianism.
Another theme in the debate is the relationship between the individual and society in Sen’s writings on capability.5 Chapters by Alkire, Séverine Deneulin, Robeyns and Miriam Teschl and Laurent Derobert engage critically with this debate at the conceptual level. Alkire argues that many criticisms of the so-called ‘individualism’ of the capability approach arise when the capability approach is drawn upon to generate ‘prospective’ recommendations (rather than evaluations in the broad or narrow sense). She clarifies that prospective recommendations generated in the capability literature inevitably draw upon institutions and intermediary processes and do not posit Robeyns’ methodological individualism, so the criticisms, while accurate in substance, misattribute an individualism that the capability approach lacks.
Deneulin is unconvinced that Sen’s capability approach can give sufficient importance to what Charles Taylor has called ‘irreducibly plural goods’. She puts forward the notion of ‘socio-historical agency’ as central in the promotion of capabilities, bringing into perspective an empirical illustration of capability expansion in Costa Rica. Her chapter can be read as making the case for a further broadening of the informational basis of the capability approach – when this is used as the basis for prescriptions – to include Paul Ricoeur’s notion of ‘structures of living together’ which belong to a particular historical community but are irreducible to individual relations. Deneulin’s argument suggests that in its current form the approach is not just incomplete but potentially misleading.
Robeyns distinguishes between ethical individualism – where the ultimate unit of concern is the individual – and methodological and ontological individualism – which hold that social phenomena can be explained by reference to individuals alone, and that society is merely a sum of its individual parts. She defends ethical individualism, arguing that it is necessary for an adequate account of the wellbeing of women and children. Teschl and Derobert explore how a person’s agency and identity influences their choice of functionings from their ‘capability set’ – the set of vectors of functioning from which they choose. They note the powerful role that a person’s diverse social identities can have in influencing their choices. In spite of the apparent contrast between Sen’s alleged ‘individualism’ and the focus on community in the ‘communitarian’ literature, Teschl and Derobert find that Sen’s position is closer to that of one leading figure in that literature – Michael Sandel – than either Sen or Sandel might acknowledge.
Given that evaluation of capability raises a challenging array of issues of measurement, aggregation, comparison, vagueness, etc., it is with good reason that a growing literature explores these issues. Sen has distinguished three ways in which the capability perspective can inform empirical and quantitative measurement work: the ‘direct approach’ – which ‘takes the form of directly examining and comparing vectors of functionings or capabilities’; the ‘supplementary approach’ – which involves ‘use of traditional procedures of interpersonal comparisons in income spaces but supplements them with capability considerations’; and the ‘indirect approach’ – which ‘remains focussed on the familiar space of incomes, appropriately adjusted’ (Sen 1999: 82–3). Each of these approaches is seen as a way of giving ‘practical shape to the foundational concern’ (Sen 1999: 81).
In this introduction, we interpret the notion of ‘application’ broadly so that it covers the various ways in which a conceptual approach can be given a practical shape or value. Applications matter, not only because intellectual effort can contribute to practical change and inform policy-making but also because they can reshape understanding and contribute towards better conceptualisations of social phenomena and assessment procedures. Some applications involve measurement, but measurability is not a necessary condition for giving practical shape or value to a conceptual approach. The wide range of capability applications described in this book may contribute to shaping and illuminating the insights of the capability approach and can provide further refinements of its conceptual foundations.
The measurement literature includes examples of the direct, indirect and supplementary approaches at work. The direct approach is the most ambitious way of applying the capability approach. Attempts to pursue it typically address the multi-dimensional nature of wellbeing, inequality or poverty when these are understood in terms of capability or functionings. For this reason, some applications of the capability approach are close relatives of other approaches to multi-dimensional measurement. A large literature on such multi-dimensional measurement of wellbeing, poverty and inequality has emerged.6
Some of the issues which arise for multi-dimensional measurement are illustrated in Figure 0.1 with respect to poverty. The vertical axis represents achievement in terms of some indicator(s) for some domains. The horizontal axis shows the time across which achievement is measured, which may include future as well as present poverty. A specific level, or range of levels, of achievement constitutes a poverty threshold, or fuzzy poverty band, for each domain which may change
| Image not available in HTML version |
Figure 0.1 Schematic overview of multi-dimensional poverty for individual i
over time. This is represented by the broken arrow in the diagram. If a person or group falls within the fuzzy poverty band it is ambiguous whether they are poor. Multidimensional measurement would include information of this sort for each dimension.
Recurrent questions in this literature comprise the following. Which are the domains or dimensions that will be included, and on what basis?7 Which indicator(s) best represent each domain or functioning, and on what grounds will these be selected? What is the poverty threshold for each indicator, or, if a fuzzy threshold is defined, what are the upper and lower boundaries of the fuzzy poverty band? How does one represent the interaction between different indicators and the interactions between dimensions of poverty and identify substitutes and complements? In those cases in which it is necessary to aggregate across domains, how is this achieved and what relative weights are set for various domains? And how does one aggregate across individuals? Various approaches to multi-dimensional poverty measurement propose clear answers to these questions. A multi-dimensional measure of poverty – the human poverty index – which Sen developed with Sudhir Anand (Anand and Sen 1997) is an example of such a particular measure which is inspired by the capability approach. Decisions about the selection of dimensions, indicators and weights are made in all the multi-dimensional measures of human development – most obviously in the Human Development Index (or HDI), developed by the United Nations Development Programme (UNDP) – and related measures of gender inequality (McGillivray and White 1993; Anand et al. 1994; Streeten 1994; Anand et al. 1995; Anand and Sen 1997; Anand and Sen 2000; Anand and Sen 2000; Sen 2000; World Bank 2000; Ogwang and Abdou 2003; McGillivray 2005).
While there is a significant overlap between applications of the capability approach and other approaches to multi-dimensional measurement, the capability approach is distinctive inasmuch as it stresses that capabilities and functionings have value in themselves: ‘intrinsic value’. Income, by contrast, is seen as having ‘instrumental value’ – value as a means to the realisation of other ends. While some ‘indirect’ applications of the capability approach use income as a proxy measure for certain capabilities (see Anand and Sen 2000 and Klasen 2000), income is not usually seen as a dimension of wellbeing itself. Furthermore, the fact that income has an instrumental rather than intrinsic value can influence the form in which income enters into a multi-dimensional measure.8 This is one among a number of instances where the capability approach as a conceptual framework has implications for measurement. Tracing out such implications is a central theme of Flavio Comim’s chapter. Drawing on the writings of both Nussbaum and Sen, he shows the relevance of the approach – understood broadly as an extensive evaluative framework – to measurement issues. Comim also illustrates his claims in various concrete contexts, discussing empirical work carried out in research projects aiming to measure capabilities.
Figure 0.1 also allows us to address a question which has been neglected in the literature on capability: how to handle time? It locates any individual’s or group’s achievement in a relevant dimension in time. If relevant information is available across time, we would then be able to judge whether a person’s failure to achieve a minimally adequate level in some dimension is merely temporary or ‘chronic’. This would be one way to link work on capability to work on ‘chronic poverty’ (Hulme and Shepherd 2003). By locating people or groups in time, Figure 0.1 illustrates how one might study capability dynamics. It also allows one to consider whether or not a person or group situated at
© Cambridge University Press