[CA 01]: Functionings and capabilities

by Ingrid Robeyns on March 6, 2014

There are two notions in the CA that are key – the notions of ‘functioning’ and ‘capability’. Since most of the discussions on the CA are about human beings, I will restrict the discussion now to human functionings and capabilities, and devote a separate post later to nonhuman capabilities. So unless specified otherwise, all references in what follows [in this and future posts] will be to human capabilities.

Capabilities are what a person is able to do or to be; functionings are those capabilities that are realized. Thus, functionings are ‘beings’ and ‘doings’. Examples of the former (the ‘beings’) are being well-nourished, being undernourished, being housed in a pleasantly warm but not excessively hot house, being educated, being illiterate, being part of a supportive social network, being part of a criminal network, and being depressed. Examples of the ‘doings’ are travelling, working, taking part in social events, caring for a child, voting in an election, taking part in a public debate, taking drugs, killing animals, eating animals, donating money to charity, consuming lots of fuel to heat one’s house.

Capabilities are a person’s real freedoms or opportunities to achieve functionings. For example, while travelling is a functioning, the real opportunity to travel is the corresponding capability. The distinction between functionings and capabilities is between the realized and the effectively possible, in other words, between achievements, on the one hand, and freedoms or opportunities, on the other.

From these examples we can draw a couple of observations. First, these examples indicate that many features of a person could be described either as a being or as a doing: we can say that a person is housed in a pleasantly warm house, or that this person does consume lots of energy to keep her house warm. Yet other functionings are much more straightforwardly described as either a being or a doing, for example ‘being healthy’ or ‘killing animals’.

Second, some functionings are very closely related to resource-holding or consumption. One main difference would be that in the case of functionings we have some information on the reason for the consumption, information on why consumption takes place. Consuming fuel comes very close to simply a category of spending, and hence a regular notion in consumer theory; what is added by considering this act from a capability lens is that we know why the fuel is consumed – namely to keep one’s house warm.

The third observation is that the notion of ‘functionings’ is a conceptual category that is in itself morally neutral. Functionings can be univocally good (e.g. being in good health) or univocally bad (e.g. being raped). But the goodness or badness of various other functionings may not be so straightforwardly determined, but rather depend on the context and/or the normative theory which we endorse. For example, is the child care of a mother who is caring full-time for her child a valuable functioning or not? A conservative-communitarian view [accounts which endorse ideals of ‘traditional’ motherhood] will most likely mark this as a valuable functioning, whereas a feminist-liberal theory [accounts stressing individual freedom and the unfairness of traditional gender norms] will only do so if the care work is the result of an autonomous choice made against a background of equal opportunities and fair support for those who have duties to care for dependents.

Basic capabilities
The terminology in the literature on the capability approach has changed over time, and this has led to some confusing use of certain terms, and also to difficulties in properly interpreting the earlier contributions to this field. This is importantly exemplified by the different interpretations of the term ‘basic capabilities’. Martha Nussbaum (2000: 84) uses the term ‘basic capabilities’ to refer to “the innate equipment of individuals that is necessary for developing the more advanced capabilities”, such as the capability of speech and language, which is present in a newborn but needs to be fostered. Amartya Sen (1980) mentioned the term ‘basic capability’ as his first rough attempt to answer the ‘equality of what?’ question, but changed his terminology in subsequent work (what he called ‘basic capability’ would later become ‘capability’). In later work, Sen reserved the term ‘basic capabilities’ to refer to a threshold level for the relevant capabilities. A basic capability is “the ability to satisfy certain elementary and crucially important functionings up to certain levels” (Sen 1992: 45 n. 19). Basic capabilities refer to the freedom to do some basic things considered necessary for survival and to avoid or escape poverty or other serious deprivations. The relevance of basic capabilities is “not so much in ranking living standards, but in deciding on a cut-off point for the purpose of assessing poverty and deprivation” (Sen 1987: 109).

Hence, while the notion of capabilities refers to a very broad range of opportunities, among development scholars and economists basic capabilities refer to the real opportunity to avoid poverty or to meet or exceed a minimal threshold of well-being. Basic capabilities will thus be crucial for poverty analysis and in general for studying the well-being of the majority of people in poor countries, or for theories of justice that endorse sufficiency as their distributive rule. In affluent countries, by contrast, well-being analysis would often focus on capabilities that are less necessary for survival. It is important to acknowledge that the capability approach is not restricted to poverty and deprivation analysis but can also serve as a framework for, say, policy evaluations or inequality measurement in non-poor communities. Sen’s and Nussbaum’s extensive writings on the capability approach may mislead us into thinking that the capability approach is about poverty and development issues only, but there is no reason to restrict its scope in this way, neither on conceptual nor on normative grounds.

Note also that some philosophers, especially those interested in notions of basic rights, would use the term ‘basic’ in another sense, namely as ‘central’ or ‘most worthwhile protecting’. While in terms of implications those two usages of ‘basic’ may coincide, that need not be the case. For example, in most accounts of human rights, the right to have an equal say in political decision making is seen as ‘basic’, but this is seldom discussed by development economists who discuss ‘basic capabilities’ in the context of determining who counts as poor; there the focus is on socio-economic or material dimensions of well-being, and not on political and civic liberties. In our exploration of the CA we will come across many instances of one word or notion being differently used in different disciplines, often causing confusion: this is a first example.

Capability or capabilities?
Another confusion in the literature is the difference between ‘capability’ and ‘capabilities’. In Sen’s earliest work, every person has only one capability, which consists of a combination of potential functionings. Functionings could therefore be either potential or achieved. In this earliest use of the capability language, ‘capability’ was synonymous for ‘capability set’. This kind of language is especially familiar to social choice scholars or theorists working on formal accounts of freedom, where the focus of the analysis is often the opportunity set. A person’s capability is best thought to be the equivalent of a person’s opportunity set, where the opportunity set refers to the set of achievable combinations of beings and doings, hence combinations of potential functionings. But many other scholars working in the capability paradigm, including Martha Nussbaum, have labelled these potential functionings ‘capabilities’. In their terminology the capability set consists of a number of capabilities, in the same way as a person’s overall freedom is made up by a number of more specific freedoms. In Sen’s earlier work, one does not find this plural usage of capabilities, and in his later writings he employs both uses of the word ‘capability’ interchangeably. Yet the plural use of ‘capabilities’ is widespread in Nussbaum’s work, in the work of Sen’s commentators and in the work of most other capabilitarian scholars. In 2010, at a conference I organized on Sen’s philosophy, he was pressed on this distinction, and he responded something to the effect that sometimes one just has to go with the flow and adapt to what has become the dominant way of using a set of notions. Whereas the ‘older’ terminology that Sen used is today still used by a small minority of scholars, it is clear that the fact that the definitions shifted over time, without clear highlighting by either Sen or Nussbaum, doesn’t make it easier to understand the CA for those who also make the effort to read the first papers by Sen, or who are going back and forth between different disciplines.

Questions/themes postponed for later:
1. Martha Nussbaum’s distinction between different types of capabilities.
2. Des Gasper’s proposal to distinguish between S- and O-capabilities.
3. Other ways to distinguish between different capabilities, e.g. general vs. specific.
4. Do functionings and capabilities have to be individual, or can there be ‘collective capabilities’?
5. Non-human capabilities: animals, organizations?
6. Discussion of Nussbaum’s list of capabilities: are they all capabilities, and should they all be on that list?
[feel free to add themes/questions!]

Nussbaum (2000) Women and Human Development, CUP.
Sen (1980) ‘Equality of What?’ Tanner Lectures on Human Values.
Sen (1987) The Standard of Living, CUP.
Sen (1992) Inequality Reexamined, Oxford: Clarendon Press.



Matt 03.06.14 at 2:55 pm

I’m very excited to see this start, Ingrid. You pick up on something right away that has always worried me about this approach when you say, the notion of ‘functionings’ is a conceptual category that is in itself morally neutral. Functionings can be univocally good (e.g. being in good health) or univocally bad (e.g. being raped)

I’m no expert on the idea of ‘functions’, either in, say, the philosophy of biology/mind literature, or in the line of work that goes back to Aristotle (these overlap somewhat, but seem to go in importantly different ways), but all of them seem, if I recall correctly, to take the idea of “proper function” as basic. That is, you can’t know what the function of something is unless you know what it should be doing, in some sense of “should be”, spelled out in different ways by different people. For the virtue theorists, at least, this won’t be “morally neutral”, unless we think that virtue theory is just independent from morality- a different sort of game all together. But, at the least, it seems that the idea of function involves an _evaluative_ element- there’s no such thing, say, as being a good broken chair, or functioning well as a broken chair should- a broken chair is just failing to function as a good chair in some way or another. Or so I have always understood the idea of functioning, in the philosophical literature. (Again, I’m no expert on any of it, though I’ve read a fair amount.)

What sort of worries this raises for the capabilities approach I’m not completely sure, but I think it’s connected heavily to the charge, especially raised [rightly, I think] against Nussbaum, that the approach has a sort of illiberal perfectionism built into it from the start.

Anyway, even if I’m not getting the issue all wrong, there is a lot of heavily debated stuff here, and I, at least, don’t feel that confident about the answers. I do tend to think that “functioning” essentially involves an evaluative element, even if not a “moral” one, strictly speaking, and at least tend to worry that this is going to lead to some smuggling in of various values at different points.


Ingrid Robeyns 03.06.14 at 7:37 pm

Matt, this is very interesting, thanks. Let me first try to get to the easy case, which is what I think we *should* avoid, and that is to define functionings as those beings and doings we have reason to value. Quite a large number of (generally non-philosophical) scholars within the capability literature define functionings and capabilities in such a way. But this is presupposing that we know what functionings and capabilities are valuable – whereas that is precisely a major part in which the broad approach becomes more specific when one develops better-worked-out capabilitarian theories. I think we should not assume, from the start, that all functionings are things worth wanting, for at least two reasons: first, what I just mentioned (the selection is normative in itself), second, we may want to compare people not only in how they are doing in terms of positively valuable capabilities, but also how they are doing in ‘negative functionings’. For example, if two people A and B have the same well-being level (understood in terms of functionings), but B in addition suffers from major bouts of depression, then I think it is very plausible to say that B has lower all-things-considered wellbeing than A.

But the issue you point at is deeper. Do I understand you correctly that you are questioning whether the idea of functionings can be *fully* non-normative since there may be an idea of human nature underlying the idea of functioning, which is not entirely value-neutral or non normative? If so, I don’t have an answer to this worry.

Perhaps we could illuminate this issue more if we could single out a functioning that could be contested? I’ve once mentioned that ‘flying like a bird’ is not a functioning, and a colleague objected that it is not evident at all that humans will never be able to fly like a bird, and hence why should we conclude that this is not a functioning? …. something like this?

On final thought: perhaps ‘functioning’ is also a somewhat misleading term since within the CA it is used in a ‘technical’ sense? It may suggest that it is similar to or directly related to the commonsense term ‘to function’ – which it may not quite be to the extent that we are believed to think it is.


Matt 03.08.14 at 12:50 am

Thanks for this, Ingrid. I’ll admit that the idea of a function is one where I feel like (and not just I) start to lose a grip quickly once I move beyond a fairly surface level. I _expect_ that for most uses of the CA, a surface level is really all that’s needed or should be wanted. I _think_ that Nussbaum wants to push beyond that (invoking Aristotelian and Marxist ideas) and that this is one of the things that’s problematic about her approach- it depends on a view of human nature at the core that’s hardly neutral or obvious, among other things- but I don’t feel very confident here. These are really “worries” on my part at times rather than objections- perhaps areas where I can imagine things going wrong if people are not careful, rather than strong objections to the approach.


Val 03.10.14 at 6:43 am

interested in this discussion but I commented on John Quiggin´s post yesterday and my comment has disappeared, I don’t know why.


John Quiggin 03.10.14 at 7:15 am

@Val I can see your comment, and my two replies, so I don’t know why you’re seeing something different. Note that you may have been bumped down because comments held in moderation are added in order of original posting.


Val 03.10.14 at 7:43 am

sorry please ignore last comment – the earlier comment is visible on the mobile device but not the laptop (I don’t know why again)


Sam Clark 03.10.14 at 11:23 am


I _think_ that Nussbaum wants to push beyond that (invoking Aristotelian and Marxist ideas) and that this is one of the things that’s problematic about her approach- it depends on a view of human nature at the core that’s hardly neutral or obvious, among other things- but I don’t feel very confident here.

You’re quite right that Nussbaum appeals to a rich and non-neutral account of human nature. She defends that approach e.g. in her ‘Human Functioning and Social Justice: In Defense of Aristotelian Essentialism’ in Political Theory 20.


Matt 03.10.14 at 1:15 pm

Thanks Sam- I’ve read lots of Nussbaum on this topic but not this paper. I’m pretty skeptical of the views about human nature in her other work (even skeptical that they are coherent with her other views) but I’ll have to try to check this out.


Cap 03.10.14 at 7:38 pm

Ingrid Robeyns @2:

For example, if two people A and B have the same well-being level (understood in terms of functionings), but B in addition suffers from major bouts of depression, then I think it is very plausible to say that B has lower all-things-considered wellbeing than A.

This example is a bit difficult to undestand for a newcomer to CA. I got the impression that being depressed is also a functioning, ‘a being’. Should one read the example as A and B having the same well-being level, when taking into account all functiongs except being depressed? Then B would have all functionings considered lower well-being than A.


ingrid robeyns 03.10.14 at 8:08 pm

The problem with Nussbaum is that there is the Early Nussbaum and the Later Nussbaum who defend exactly the same list of capabilities, but the Early Nussbaum grounds them in Aristotelian Essentialism (and is happy to defend a perfectionist account of wellbeing), whereas the Later Nussbaum vigorously claims to be a political liberal and to present a list of capabilities based on a Rawlsian overlapping consensus. The Political Theory paper (which incidentally I am teaching to a group of PhD students tomorrow) is the Early Nussbaum. The later Nussbaum denies to be perfectionist. There are several philosophers who have asked into question whether that project really is and can be political liberal – I suppose this warrants a separate post, so that we can discuss this in some detail. The main thing we should remember for the time being is that there is not one Nussbaum, but two. I will also try to dig up the paper where she first announced the shift.


Ingrid Robeyns 03.10.14 at 8:20 pm

Cap @ 9 – yes, you’re right – I meant wellbeing in all other dimensions except being depressed. The point of the example is to illustrate that a functionings should also be allowed to be negative (and hence wellbeing lowering), since otherwise we would have to come to the implausible conclusion that A and B had the same wellbeing.


John Quiggin 03.12.14 at 2:49 am

Ingrid, a question came up in my discussion thread, which I thought I would raise with you. Commenter Sam Clark says (of my paper)

There’s something strange, on the face of it, about basing what looks like a way to do cost/benefit analysis on the CA: I took it that the CA is value pluralist in the sense that it claims that there are a number of distinct, incommensurable values, and therefore resistant to any such move to arithmetic measurements and trade-offs.

I replied that, since I want answers to policy questions like “should we fund an oncology ward or a palliative care centre”, I definitely want (local) commensurability and trade-offs, whether or not this is consistent with CA.

But it seems pretty clear that a number of commenters share concerns like Sam’s, so I thought I would ask you whether there is an established view, or views, in the CA literature on whether capabilities are commensurable/subject to trade-offs.


John Quiggin 03.12.14 at 2:58 am

Following up, Google suggests to me the following
1. Both Sen and Nussbaum assert that capabilities are incommensurable: but
2. Both advocate (as an imperfect measurement tool) the Human Development Index, which is an arithmetic measurement involving trade-offs between capabilities.

My interpretation is that the claim of incommensurability applies to the procedures of benefit-cost analysis which purport to generate cardinal dollar values, but not to ordinal rankings over capability vectors, which is what my paper (with Han) is about. But, I’d be very grateful for your thoughts on this.


Cap 03.13.14 at 10:30 am

There are many interesting discussions in these blog posts and in their comments. However, I think that one issue is not, at least explicitly, discussed before, and so I’ll take the question up here. Does CA allow ‘feedback loops’ between capabilities and functionings? Or does the set of capabilities stay unaffected from use of functionings? I’m thinking here that the learning of the more advanced skills needs the basis of basic skills. E.g., it is very hard or even impossible to study CA, if one has not learned to read and write fluently. The main question might be that is the set of capabilities determined outside of the use of functionings? I.e., capabilities would be only those that we would agree on to be the ‘basic capabilities of humans’?


Cap 03.13.14 at 10:41 am

Also, related to my previous comment/questions. Are functionings only beings/doings that are ‘activated’ at the time of ‘inspection’? E.g., is being depressed a functioning only at the time when one is depressed? Or is it also a functioning, when one is not depressed, but has a ‘tendency/propensity’ to get periods of depression from time to time?

Comments on this entry are closed.