[CA 02] The helicopter view: What is the capability approach?

by Ingrid Robeyns on March 6, 2014

Assuming we now have a basic understanding of the notions of ‘functioning’ and ‘capability’, we can ask what the capability approach is. The best way to answer this question is by first taking a helicopter view, and having a not-too-detailed look at the entire terrain we will be covering. Perhaps an outsider would expect that this is an easy question, but alas it is not. In my view, it is poorly analyzed in the literature, sometimes misleadingly discussed, and also the source of many confusions and possible flawed arguments [arguments for this view will be provided in future posts, not now!].

Here’s how Amartya Sen described the CA in a paper devoted to clarifying the approach:

“[The capability approach] is an intellectual discipline that gives a central role to the evaluation of a person’s achievements and freedoms in terms of his or her actual ability to do the different things a person has reason to value doing or being.” (Sen 2009: 16)

Sen clearly opts for a general description the CA, that doesn’t tie it to one particular scholarly discipline or debate. I agree with the general trust of Sen’s description. Yet let’s try to get this a bit more specific.

Here’s how I see things. In its most general description, the capability approach is a flexible and multi-purpose normative framework, rather than a precise theory of well-being, freedom, ‘advantage’, justice, or anything else. At its core are the notions of ‘functionings’ and ‘capabilities’ and two normative claims: first, that the freedom to achieve well-being is of central moral importance, and second, that freedom to achieve well-being is to be understood in terms of people’s valuable capabilities, that is, their real opportunities to do and be what they have reason to value. This framework can be used for a range of evaluative exercises, including most prominently the following: (1) the assessment of individual well-being; (2) the evaluation and assessment of social arrangements, including assessments of social and distributive justice; and (3) the design of policies and proposals about social change in society, which is at the core of social ethics. In all these normative endeavors, the capability approach prioritizes (a selection of) peoples’ beings and doings and their opportunities to realize those beings and doings, for example their genuine opportunities to be educated, their ability to move around or to enjoy supportive social relationships. This stands in contrast to normative frameworks which endorse other accounts of value, like mental states or which focus on instrumental values (e.g. resources). All capabilitarian theories focus on what a person is able to be and to do (her capabilities) and/or those capabilities that she has realized (her functionings).

This first description raises a host of questions. What does it mean to say ‘those capabilities people have reason to value’? Who are those ‘people’ – are that individuals, or is this a collective process? How does this approach exactly differ from other established approaches, such as recourcism, utilitarianism, etc.? Can this approach at all be developed into a plausible theory of justice? Is the CA not informationally overdemanding? Can functionings and capabilities at all be measured? Is this a viable approach for the further development of welfare economics? Can the CA be helpful for those who believe that economics should again self-identify as a moral science? Is the CA not simply a sociological turn in anglo-american political philosophy and indeed also in economics? etc etc etc. [yes: zillions of questions].

I hope we can take up many of those questions over the next weeks and months. For the moment, I want to make one single point, which however I think is hugely important, and therefore it deserves a post of its own. The CA is not merely a theory of justice. The CA is not merely part of the foundations of a new variant of nonwelfarist welfare economics. The CA is not the just [part of] a theory of wellbeing. The CA is not merely the foundation of an normative account of freedom. The CA is, instead, a much vaguer thing, and at the same time the CA can be further developed in all of the above.

So this implies the following: the CA can be developed into any of the above more specific theories [whether the capabilitarian theory that then emerges is coherent and plausible, is a separate question]. But one cannot, without additional arguments, generalize from something that one thinks holds for the CA in one of those domains to all those domains. Take the following hypothetical example. Suppose we have good arguments for why in capabilitarian theories of justice the selection of capabilities should be done by method X which relies on a justificatory process Y. It does not follow that method X and justification Y are thereby the right method and justification across the board within the CA; rather, it may well be that in developing a capabilitarian pedagogy, or in developing a capabilitarian welfare economics, we need different methods and different modes of justification.

This also implies that every time someone says something about the CA, we need to ask ourselves immediately: is this a statement about the CA as a general framework, or the CA within welfare economics, or within development studies, or a capabilitarian theory of justice, or another area in which the CA is used?

Questions/themes:
1. All questions mentioned in this post.
2. (How) does this differ from Nussbaum’s account of the CA?

References:
Sen (2009) ‘Capability: Reach and Limit’, in: Enrica Chiappero-Martinetti (ed.), Debating Global Society. Reach and Limit of the Capability Approach, Milan: Feltrinelli, pp. 15-28

{ 74 comments }

1

Sam Clark 03.06.14 at 12:59 pm

As someone centrally interested in theories of well-being, I have a tendency to see the CA as just such a theory, so this is a really useful corrective for me.

Actually, that point and this post seem a better place to start than [CA 01], which left me asking ‘but what question do you think the CA is an answer to?’. This makes it clear that it’s not an answer to one question, it’s a way of trying to answer various questions – about well-being, justice, development, poverty, education etc. – by focussing on what people are really able to do, and what they actually do and become (aside: is that an adequate summary of the capability/functioning distinction?).

Looking forward to more of this series.

2

Straightwood 03.06.14 at 3:32 pm

There seems to be an important false premise here – that there is some sort of theoretical vector to an end state of fulfilling “capability.” The selfish gene has wired us to be restless, questing creatures. Thus, there is no durable environmental optimum. We will be cranky, cantankerous, and rebellious, even after the conquest of all diseases, universal guaranteed basic incomes, free education, and unlimited access to art and entertainment have been attained. Ennui, envy, resentment, and regret will be with us forever, no matter how perfectly we engineer our institutions.

CA theory exports analytical difficulty into the conveniently vague domain of “things a person has reason to value doing or being.” But these things are constantly shifting and morphing. Thus the great weakness of a theory of social values focused on enabling capability is that it is critically dependent on what we wish to enable. The pragmatic pillars of constructing an enabling society rest on the shifting sands of cultural and psychological preferences, and thus are structurally unsound.

3

Luis 03.06.14 at 4:15 pm

Straightwood: I’d love for Ingrid to correct me, but I think the Sen response to your query would be to say that all theories of rights/justice have similar problems, but most simply sweep them under the rug. (see, e.g., Rawls’ shifting positions on a variety of things over his lifetime, much less the disputes many Rawlsians have had with each other about what people in the original position would choose to do.) CA acknowledges the problem, and gives you a framework for labeling and discussing it.

I’m less familiar with Nussbaum (something I’m working on correcting through my reading of a CA textbook Ingrid contributed to) but my understanding is that she somewhat agrees and therefore proposes a fixed list of relevant capabilities.

4

Sam Clark 03.06.14 at 4:33 pm

I don’t think Sen or Nussbaum want to leave ‘things we have reason to value’ at that: it’s a promissory note (cashed out e.g. in chapter 2, ‘The Central Capabilities’, of Nussbaum’s Creating Capabilities).

I also don’t think they’d claim to have the finished answer to ‘what do we have reason to value?’, leaving nothing left to be done but working out tactics for achieving their list. Their own unfinished account of what the central human capabilities are is an intervention in an ongoing discussion, to be read, criticized, and superseded.

5

Ed Herdman 03.06.14 at 4:52 pm

I cringe when I see statements like “the selfish gene has wired us to be…” I don’t think even Dawkins will make that statement without qualification. However, I think we’re back on good ground with the following statement from Straightwood about there not being a durable environmental optimum – if we make the stipulation that there is a kind of chaotic dimension to social game-playing, which makes finding a good state of affairs quite different than, say, attempting to intelligently design a durable system that uses reversible computing, where there aren’t competing interests (genes / memes).

Apologies for being off-topic here, I’m very interested in the topical discussion but thought this could use a bit more fleshing out.

6

js. 03.06.14 at 5:24 pm

This is really helpful. Thanks. To kind of echo Sam Clark, I tend to think of CA almost exclusively in the context of theories of distributive justice, and this is a useful corrective.

7

Jameson Quinn 03.06.14 at 5:46 pm

One nice thing about utilitarianism is that it provides a set of mathematical ground rules which can protect you from getting tripped up by certain kinds of logical failures (such as utility pumps). In principle, there’s nothing to stop anyone from using the same or similar ground rules in a CA approach — that is, essentially do utilitarianism, but replace the actual substance of utility with “capability” rather than, as often tends to happen, with “a monotonic function of money per person”. But in practice, it seems as if this CA talk is leading us inexorably away from anything we could put numbers onto.

The reason I bring this up is that I’m starting a course with Sen in which my goal is to put the normative theory of single-winner voting systems on a more solid foundation. I think that that’s a reasonably important goal, from the point of view of CA. And yet in order to do that, I really need something with numbers in it. Even if I’m not entirely ready to fully embrace utilitarianism (and also, I don’t want to provoke Sen by appearing to do so), I see little choice; there’s not much in CA, as I understand it from these two posts, that could help me. What am I missing?

8

Ed Herdman 03.06.14 at 5:54 pm

That’s an interesting way of thinking about utilitarianism – as a kind of first sweep with a broom to sweep up the grit.

Sen has an argument somewhere (I think the Adler/Posner anthology on Cost-Benefit Analysis) which looks for all the world like utilitarianism, except urging that we should consider different costs and benefits than those traditionally used (which – I might be going beyond Sen’s argument here – come with a bias, i.e. considering costs in just financial terms seems to be biased towards concerns which seek to explain everything in those terms, so favors highly capitalized groups over wildlife and local citizens). I’m not sure it’s nearly correct enough to say “CBA = utilitarianism” but it is moving in that direction.

9

Bill Gardner 03.06.14 at 6:01 pm

So… it seems like the vector of capabilities that we use to characterize well-being | justice | educational goals | [what have you] will be specific to a specific society and historical era? It will be defined by the horizon of achievable well-being, as seen by that community?

10

Ed Herdman 03.06.14 at 6:14 pm

I don’t think that is necessarily against the possibility of revolutionary changes in the concept of “the meaning of life” or goals. Any society will necessarily have its predispositions towards a concept of well-being based on its own conceptions. Think of the supercomputer AI analogy: A dog thinking 100 times faster than a normal dog is still not going to achieve much, and the same for a human. To get to the next breakthrough (and whether you can actually achieve much more is also an unsettled question) takes at least a revolution in thought.

In any case, I have a hard time really believing that people have, or could have, a radically different concept of well-being across the ages, in terms of the meaning of life and things of that sort – now justice for minorities and many other issues obviously have a direct claim towards being viewed in a particular way so we don’t fall into the trap of moral relativism.

11

MPAVictoria 03.06.14 at 6:45 pm

Can anyone further expand on what “a new variant of nonwelfarist welfare economics” means?

/While I do not feel like I have the background to contribute much in the comment sections of these posts I will be closely following along. Thank you Ingrid.

12

Straightwood 03.06.14 at 6:57 pm

@10

In any case, I have a hard time really believing that people have, or could have, a radically different concept of well-being across the ages

The “capabilities” sought by medieval monks, Mongol warriors, and modern American consumers are wildly different.

13

Bill Gardner 03.06.14 at 7:02 pm

Yes, to Straightwood. And wouldn’t wellbeing in an ancient Christian or Buddhist society would differ from a contemporary market/secular society?

14

js. 03.06.14 at 7:11 pm

Can anyone further expand on what “a new variant of nonwelfarist welfare economics” means?

MPAV,

I just lost a 3-para response to the virtual wilderness. I’ll try to see if I can recreate a bit later. In the meantime, here are a couple of links from the SEP: welfarism, welfare economics.

15

Ed Herdman 03.06.14 at 7:20 pm

Right, and I’ve avoided using the categories of justice, educational goals, and “what have you” because it seems like you can sweep up whatever you want in order to make the CA seem like it can’t work. Maybe the CA has a somewhat limited scope here, but again, I don’t think that what’s good for an ancient monk in China is actually radically different from what’s good for me, or from an early hunter-gatherer, or even from a pre-human ancestor. Most of our psychological and physiological needs are precisely the same (or quite close). This just looks like a confusion of terms to me.

Also note Ingrid’s early line: “Sen clearly opts for a general description the CA, that doesn’t tie it to one particular scholarly discipline or debate.” It seems to me that maybe a bit of generality is acceptable here.

16

Straightwood 03.06.14 at 7:34 pm

@15

I don’t think that what’s good for an ancient monk in China is actually radically different from what’s good for me, or from an early hunter-gatherer, or even from a pre-human ancestor.

I don’t think the CA approach is limited to food, clothing, and shelter. Once past enabling access to the most rudimentary necessities, it becomes obvious that giving people the capability for killing and enslaving others conflicts with giving people the capability to live a tranquil existence.

For most of human history, violent conquest has been highly prized as a badge of merit for individuals, tribes, and nations. In many subcultures today, this remains a dominant ethos. How does maximizing the capability for violent predation lead us to a constructive social outcome?

17

Ed Herdman 03.06.14 at 7:49 pm

So is the argument that there never was a robust-enough (not necessarily ‘perfect’) conception of well-being that has been durable throughout the ages, or that all ethics were necessarily too temporally and regionally distinct, or something else?

We surely can’t directly poll every person in history about what well-being involves, but a quick search of the Chinese literature shows that warfare was often for dynastic (personally useful, though bad for others), or the result of something like a utility calculation. And very many people actually did appeal to some sentiment or ethics against warfare, like the great T’ang poets Li Shangyin or especially Du Fu. So I don’t think we fall totally into the trap of projecting our own particular ethics when we say that we can talk reasonably about there being a conception of wellbeing.

I think it’s probably critical to say that while a monk’s activities may be very different from mine, a dispassionate but enlightened (had to throw in some monkey wrench!) should be able to see good in both lifestyles – while recognizing that there may also be some bad (as there inevitably is). So my lifestyle is certainly more sedentary than the Chinese monk, but perhaps in some ways it is not so bad that I do not partake some particular monastic practice or other.

A great example of this is found in the smooth-worn kneecaps of many medieval Christian monk skeletons – I would not seriously hold this physical sacrifice as evidence that the whole monastic lifestyle was bad. Likewise, I assume that those medieval monks would not accuse me of having a totally unredeemed life, even though their idea of “capabilities” clearly clashes with ours in some ways – but not totally.

Another possible response, though it is against what Sen, myself, you, and I assume everybody here stands for, is that violence can’t be CA’d out of the system. It’s there, it’s a tactically valid game move. So, in fact, there’s something that truly is constant, and we don’t have to assume that we can imagine some morally privileged viewpoint.

18

MPAVictoria 03.06.14 at 7:51 pm

I appreciate the help and the links js.

19

Straightwood 03.06.14 at 8:29 pm

@17

If the presence of deeply conflicting culturally-rooted desired “capabilities” were not a sufficient obstacle to the utility of CA, the modern existential perspective of rebellion as a primary means of validating one’s humanity poses an insuperable difficulty. Just how does one enable a constructive capability for rebellion? There is more wisdom in Camus than in Sen.

20

Ingrid Robeyns 03.06.14 at 8:31 pm

Sam Clark @ 1: “what people are really able to do, and what they actually do and become (aside: is that an adequate summary of the capability/functioning distinction?).”

If you ask me, I’d say: Almost, and would suggest the following small modification:
“what people are really able to do and the person they are able to be, and what they actually do, are and become.”

Example: suppose you are homosexual. Are you able to live your life as a gay person, that is, not being killed/harmed because of your homosexual nature? That would be one element of ‘being’ that is relevant. The relevant capability would be: “living your life in accordance with your sexual orientation”.

This example also illustrates that the CA is indeed not only about food, clothing, shelter and medical care. You can have all of those in abundance, and still fail on an important dimension of wellbeing: there are plenty of people in history who had non-poor lives but couldn’t live a life according to their sexual orientation.

21

Bill Gardner 03.06.14 at 8:39 pm

Straightwood, your mileage will differ, but I don’t see why culturally-rooted capabilities poses an insuperable difficulty. A pragmatist democrat would say that the “correct” capability vector would be those capabilities that, in an ideal limit of endless democratic deliberation, that this community would converge upon. Rebellion is possible, because that community could very well see capabilities that are currently being denied to (some or all of) them.

22

Ed Herdman 03.06.14 at 8:39 pm

@ Straightwood: Rebellion in what context? If we mean just randomness, and rebellion against established order – as my earlier post attempted to argue, there are multiple acceptable statuses a person can occupy. If we mean rebellion against a particular order, then perhaps we can say that they should be able to appeal to some kind of reason. Maybe it’s floating out there somewhere as a Platonic entity!

Interesting comment about Camus vs. Sen…been so long since I read The Stranger, and I really don’t know what you mean though.

23

Ingrid Robeyns 03.06.14 at 8:45 pm

Jameson Quinn @ 7: “But in practice, it seems as if this CA talk is leading us inexorably away from anything we could put numbers onto.” it is definitely true that it is much easier to measure other ‘metrics’ than functionings, let alone capabilities. Capabilities are opportunities/opportunity sets: has anyone every measured those in a reliable way? not that I know. Does that mean they play no role in public debate, public policy, the way we organize our social and economic and political institutions? Quite the contrary: when we see a violation of equality of opportunity, we try to undo it. So the same with capabilities: if you endorse equality of capability (or some variant, such as prioritarianism of capabilities, or minimal levels of capabilities for all, etc.) and your empirical analysis indicates a violation of what you normatively defend, then you can fix it.

Still, there is more to be said on how functionings and capabilities have been and possibly can be measured, but that deserves a separate post. You’ll have to wait for that, though, since I’m first going to focus on well-being and justice.

You also write: “my goal is to put the normative theory of single-winner voting systems on a more solid foundation. I think that that’s a reasonably important goal, from the point of view of CA.” First, I should say that I am not sure that this is something the CA cares deeply about: other theories, perhaps particular theories about democratic legitimacy, may care more. The CA has little to say about process-freedoms, so the concerns of the CA may be rather independent of the concerns of those who want to put the normative theory of single-winner voting systems on more solid foundations.” But suppose you are right, and that the CA should care about this. In that case, it may be that you have philosophically sound reasons to replace ‘utility’ with ‘capabilities’, or ‘capabilities set’, but that it is either mathematically too demanding (when you want to model it) or empirically too demanding (when you want to measure it). Yet what is also possible is that the CA is not the best available account to replace utility, and that you need to look elsewhere.

24

Ingrid Robeyns 03.06.14 at 8:54 pm

MPAV: sorry for the discipline-specific jargon! I will have a post on capabilitarian welfare economics, but perhaps John Quiggin will be first since I want to turn to wellbeing and justice first. But, briefly: welfare economics is the subdiscipline within economics that is concerned with the evaluation of economic institutions and the distribution of welfare. Traditional welfare economics is ‘welfarist’, which means that it replaces ‘welfare’ with ‘utility’ and ‘utility’ is understood as ‘preference satisfaction’. This is a purely subjective approach: the more people get what they want, they higher their welfare. Non-welfarist welfare economists reject welfarism, and explore other notions of ‘welfare’ or that what is focussed on when evaluating: that could be something more objective, like functionings and capabilities, but could also be ‘the absence of envy’ or something else that is not just preferences.

25

dbk 03.06.14 at 9:01 pm

I look forward very much to this series of posts, Ingrid. Do you think you could provide links to accompany each post to open-access articles/journal entries that we could study in order to better follow, perhaps even participate in, the discussion?

26

Straightwood 03.06.14 at 9:03 pm

I think the problem here is that political philosophers don’t believe that Godel’s theorem applies to their systems. There are deal-breaking exceptions for frameworks like CA arising from cultural diversity, evolutionary dynamics, and philosophical challenges. Comparing CA to competing political science theories is a feeble test. The great challenges to CA are posed by theories of human behavior from other domains.

If political scientists read more literature, they would understand that unhappiness in advanced societies stems from confronting chance, self-ignorance, and death, not from a shortage of “capabilities.” Americans, among the most free and prosperous people on Earth, consume 67% of antidepressants sold worldwide. What Camus tells us is that tinkering with social structures is no more likely to vanquish despair than buying more stuff. Anxiety and sorrow are fundamental components of human existence. There is no conceivable set of “capabilities,” short of changing the biology of Homo Sapiens, that will make us substantially happier.

27

Bill Gardner 03.06.14 at 9:24 pm

Ingrid @23 (and Jameson @7),
Can’t some capabilities be quantified usefully? I don’t know how to measure an opportunity set. But I can measure literacy, or orthopedic mobility, or a lot of other things that seem to be useful proxies for certain dimensions of opportunity to achieve functioning. Or am I missing something critical?

28

MPAVictoria 03.06.14 at 9:33 pm

Not at all Ingrid! I understand that this is an academic blog and I appreciate the opportunity to participate when I can. Thank you for providing a bit of background on “welfarist”.

29

Ingrid Robeyns 03.06.14 at 9:40 pm

dbk: thanks for the suggestion. I will try, but fear that they will be biased to my own stuff since I know where I can find that online… but I promise I’ll search for other work too.

Bill @ 27: Yes, you are right. In fact, that was what I did in my PhD dissertation [at some point in history I was actually able to do empirical work!]. Yet I used one of the richest quantitative databases that exist, the British Household Panel Survey, and still there are so many things one would have wanted to have information about that one does not have. Also, I think it is not so hard to measure a couple of functionings, perhaps even measure a couple of capabilities – but the real normative concern should be with the set of capabilities that people have access to. If I can hold a job and work 60 hours and in such a way feed my kids, than I have the capability of employment and of meeting some essential material needs of my children; but I cannot at the same time supervise them and care for them and hence I am falling short on the capability of social relations/parenting. One may, very plausibly in my view, defend the view that as a matter of a decent quality of life we owe it to people to organize labor markets and other social institutions in such a way that people can enjoy both capabilities at least up to a minimal threshold, so that they do not have to sacrifice one capability in order to realize another. If one has a view (on wellbeing, justice, development, or something else) that endorses this kind of structure, then one needs to measures capability sets and that is really hard, perhaps almost impossible. But we’ll return to this issue of measurement, since they are interesting but also quite challenging.

30

Marcus Pivato 03.06.14 at 9:44 pm

Straightwood@26: I don’t think capabilities theorists in particular (or welfare economists in general) see it as their business to deal with upper middle-class ennui, self-ignorance, existential angst at the absurdity of the human condition, or any other so-called “first world problems”. They are mainly motivated by developing better analytical frameworks for evaluating and ameliorating human suffering caused by genuine poverty and deprivation. For example, much of Sen’s work in welfare economics (including his work on capabilities) was originally motivated by the problems of developing economies like India.

That having been said, there is still plenty of poverty, deprivation, and lack of opportunity even in rich countries like the U.S. So, while capabilities theorists (or welfare economists in general) might not have much to say about ennui in the top 1% of the income distribution in the U.S., they would certainly be very concerned about poor health, inadequate educational opportunities, and social breakdown in, say, the bottom 20% (at the very least).

Also, I don’t really see what this has to do with Godel’s theorem. Unless for you “Godel’s theorem” is just some placeholder for the vague idea that “all human philosophical systems are doomed to collapse in incoherence and failure.” In which case that isn’t called Godel’s theorem, because that’s not what Godel’s theorem says.

31

Sam Clark 03.06.14 at 9:47 pm

Straightwood @26 said:

If political scientists read more literature, they would understand that unhappiness in advanced societies stems from confronting chance, self-ignorance, and death, not from a shortage of “capabilities.”

But we’re not arguing about the conditions of (subjective mental-state) happiness, we’re arguing about (amongst other things) what the best life for a human being is (best for her, not morally best). That’s the question of well-being. To assume that well-being is a mental state is just to beg that question.

So, that people have had different ideas about how a life goes best, or (different claim) that people have enjoyed or been happy in different kinds of life, aren’t in themselves arguments against the CA or any account of well-being: maybe some people’s beliefs or enjoyments are wrong.

(I can’t speak for Prof Robeyns, but for whatever it’s worth: I’m a philosopher not a political scientist, and I read quite a lot of literature.)

(Ingrid Robeyns @20: thank-you, that’s a helpful clarification. It makes the CA sound quite definitely perfectionist or self-realizationist, in a way I find pretty appealing…)

32

John Quiggin 03.06.14 at 9:50 pm

I’m hoping to say something about quantitative representation and non-welfarist approaches very soon. Unfortunately, my functioning is currently dominated by a previously overlooked and now urgent grant application, but once that’s done I’ll have a bit more capability to talk about interesting and important things ;-)

33

Straightwood 03.06.14 at 9:52 pm

In Israel, all Haredi men desire the capability to devote their lives to Torah study. This is a capability which, if extended to all future male descendants of Ultra-Orthodox families, would undermine the economic and military integrity of the state of Israel. A life dedicated to Torah study cannot coexist with participation in a modern economy. How would CA theory guide the government of Israel in this matter?

34

Marcus Pivato 03.06.14 at 9:56 pm

Ingrid Robeyns @24: Just a slight (pedantic) correction: not all “welfarist” theories define “welfare” as “preference satisfaction”. There are “hedonic” theories which define welfare in terms of either “moment-to-moment happiness” or “overall life satisfaction” (which are not the same thing). One could also imagine a notion of “welfare” which includes some eudaimonic component.

My understanding of “welfarism” is that broadly, it is any theory which purports that it is possible to represent all ethically relevant information about a person’s well-being or flourishing in terms of a single quantitative index (call it “utility” or “welfare” or “ophelimity” whatever you want), such that the quantitative indices of different individuals can then be aggregated in via some “social welfare function” (including, but not restricted to, a utilitarian sum). In contrast, capabilities approaches are inherently multidimensional (because different kinds of functionings/capabilities may be fundamentally incommensurable) and perhaps essentially non-quantitative (if some functionings/capabilities cannot be quantified).

This, as I understand it, is the gist of Jameson Quinn’s question @7: in the end, if you are going to do some kind of welfare economics, it would seem that you need some sort of quantity that you are trying to maximize. This is perhaps the reason why many welfare economists are reluctant to engage with the capabilities approach.

35

Ingrid Robeyns 03.06.14 at 9:57 pm

Sam @ 31: perfectionism is something we should examine. The CA is often said to be perfectionist, and for Rawlsians and many other liberal political philosophers, this is a reason to reject the CA. For a discussion of well-being, the perfectionist/nonperfectionist distinctions may be most helpful, though I wonder whether for political morality the neutrality/nonneutrality/degrees of neutrality might not be more helpful. I am not entirely sure about these issues, but will give it a stab [next week], and then you can help :)

36

Ingrid Robeyns 03.06.14 at 9:59 pm

Marcus Pivato: thanks for the correction! I think your conclusion is right. If you don’t aggregate the capabilities, you are in a multidimensional framework, and that poses huge measurement issues. If you want to aggregate the capabilities, the question is how? The selection of relevant capabilities is discussed at length in the literature, but aggregation/weighing much less so [also material for another post].

37

Straightwood 03.06.14 at 10:05 pm

@31

But we’re not arguing about the conditions of (subjective mental-state) happiness, we’re arguing about (amongst other things) what the best life for a human being is (best for her, not morally best). That’s the question of well-being. To assume that well-being is a mental state is just to beg that question.

The notion that it is possible to separate the concept of “the best life for a human being” from the happiness of said being is very strange indeed. Why not just issue the rules for “well-being” by proclamation and free up the political philosophers for manual labor? The notion that “well-being” can be optimized irrespective of the happiness of the beings in question is something only a prisoner of a self-referential theoretical system could entertain.

38

Marcus Pivato 03.06.14 at 10:09 pm

Straightwood @33: All theories in welfare economics (or theories of justice more broadly) are fundamentally about how to reconcile the competing interests of different people in society to achieve some notion of “collective good”. So clearly, if pandering to the desires of one religious minority group would undermine the long-term viability of the entire society (and in the process, jeopardize everyone else’s well-being), then pretty much any sensible theory (including the capabilities approach) would say that it is not just to pander to the minority group.

39

Straightwood 03.06.14 at 10:20 pm

All theories in welfare economics (or theories of justice more broadly) are fundamentally about how to reconcile the competing interests of different people in society to achieve some notion of “collective good”.

So, if society must choose winners and losers among conflicting “capabilities,” how does CA help us make better choices than choosing them under prior theories of justice? Is the goal to find a capabilities function that allows us to calculate the “well-being” index of alternative sets of capabilities? How could such a calculation possibly weight the emotional components of various alternatives? What protects minorities in this calculus?

40

Ed Herdman 03.06.14 at 10:22 pm

Sen’s approach is going to be (from what little I know) related to bringing more considerations to the fore – if the argument is that the CA is aggressively reductionist, I don’t think that’s likely true at all. Certainly Sen isn’t reductionist (as I read him).

41

Marcus Pivato 03.06.14 at 10:30 pm

Straightwood @39: These are all excellent questions. Indeed, these are some of the central questions of the entire field. They do not have any easy answers. But the fact that they do not have any easy answers does not constitute some kind of reductio ad absurdum of the entire approach. It just means there is real work to be done.

42

LFC 03.06.14 at 11:46 pm

I have read Ingrid’s posts so far, though not all the comments. I am not familiar with the CA literature (except a little bit at second-hand).

My concern (or one of my concerns) is that the CA will perhaps tend to sound like, more or less, simply a lot of nice verbiage unless it is put into some intellectual-historical context. Ingrid made the point in post [CA 01] that the CA is not just about development and poverty; however, it did, I believe, originate in Sen’s interest in development. So it would be helpful, for me at least, to know what it was about existing approaches to conceptualizing development that Sen found wanting. For example, why didn’t he like the ‘basic needs’ approach, which was all the rage in some circles in the late 1970s/early ’80s and which always seemed to me to make a considerable amount of sense.

What’s wrong with saying “let’s ensure people have a certain level of resources (defined broadly to include eg access to education) and then what they do with those resources is up to them”? The emphasis on what people are actually able to do seems hard to operationalize or implement or measure, since two people with identical levels of resources will not be actually able to do the same things. Give an artistic genius three crayons and a piece of paper and she or he will produce something on the order of a Rembrandt or a Manet or a Mondrian. Give me three crayons and a piece of paper and I will produce a mess, or at any rate something way less good. At a general level we are both able to produce drawings, but then you would have known that already once you had given us the crayons, hence the “able to” or “capability” part doesn’t seem to add anything in this case. Trying to determine what each person is “actually able” to do seems v. difficult, whereas trying to ensure everyone has certain resources (or means) seems more feasible.

43

js. 03.07.14 at 12:05 am

The emphasis on what people are actually able to do seems hard to operationalize or implement or measure, since two people with identical levels of resources will not be actually able to do the same things.

I’m sympathetic to this thought, but strangely enough, I think the second bit here is exactly what leads Sen to propose CA as a more plausible alternative to the Rawlsian view (in “Equality of What”). So, e.g., he has what I think has become a famous example of a disabled person, and he argues that given the same absolute level of resources, a disabled person will be able to do significantly less than a relevantly able one. And I think the idea is that this seems intuitively unfair on broadly Rawslian grounds—or at least an extension of such. (Actual adherents of CA might not put the point this way.)

Anyway, if the above is right, it seems like a fundamental concern with fairness itself should lead us from focusing on the provision of resources to focusing on the role these resources can or can’t play in people’s lives, i.e., again what people can actually do with them. And though I may wrong about this, I think Sen would happily concede that this makes the measurement problem _much_ harder.

(Honestly, though, I don’t know tons about CA either, so others can maybe fill in/correct any mistakes, etc.)

44

LFC 03.07.14 at 1:08 am

js. @43:
I’d agree that in that example it makes sense to look at what the disabled person is able to do. I’d still like a better sense than I now have of what was the initial impetus for the whole approach, but perhaps Ingrid will be getting to that. (For several reasons I’m feeling none too coherent at the moment, so I think I’d better leave it at that.)

45

Ed Herdman 03.07.14 at 3:50 am

I was also interested – tried writing an abortive post earlier – about the implications of LFC’s budding artist thought experiment. Are stable institutions (schools, professions) enough to deal with some capabilities (concert pianist, astronomer, physicist) being very expensive to develop, without having the resources dedicated to an individual’s development going out of balance? The weird thing about this is that experience tells me that a lot more people are capable of filling many of those roles than actually do – many don’t get there due to accidents of birth or circumstance.

46

GiT 03.07.14 at 4:42 am

“What Camus tells us is that tinkering with social structures is no more likely to vanquish despair than buying more stuff. “

Perhaps the CA approach would take Fanon over Camus here.

“mais il ajoute aussitôt que « si ce nègre se trouve à ce point submergé par le désir d’être blanc, c’est qu’il vit dans une société qui rend possible son complexe d’infériorité, dans une société qui tire sa consistance du maintien de ce complexe, dans une société qui affirme la supériorité d’une race ; c’est dans l’exacte mesure où cette société lui fait des difficultés, qu’il se trouve placé dans une situation névrotique. Ce qui apparaît alors, c’est la nécessité d’une action couplée sur l’individu et sur le groupe. En tant que psychanalyste, je dois aider mon client à « conscienciser » son inconscient, à ne plus tenter une lactification hallucinatoire, mais bien à agir dans le sens d’un changement des structures sociales »”

47

Sam Clark 03.07.14 at 8:37 am

Ingrid: excellent, look forward to it.

Preliminary to that (and also, obviously, because someone is wrong on the internet) I think Straightwood is making an important mistake.

The notion that it is possible to separate the concept of “the best life for a human being” from the happiness of said being is very strange indeed.

Here are some cases where mental-state happiness seems to come apart from a life going well. Some of them are even literary:

1) Brave New World: people are happy (contended, emotionally stable, frequently pleased), but their lives are a centrally-managed round of pointless amusements (anyone for centrifugal bumble-puppy?). They have no emotional depth, no aesthetic or ethical development, no character, no real friendships.

2) The Truman Show: Truman is happy, but his life is a sham. The woman he believes is his devoted wife is an actress who despises him; his town is a filmset; his lunch is a product placement.

3) The Moon and Sixpence: Charles Strickland achieves his artistic goals. He does what is in him to do, completes his work. Is he happy? He’d think the question impertinent. Happiness is contemptible compared to art.

Note that I’m not (yet) defending an alternative view of well-being: I’m just pointing out that the view that well-being = happiness (or even that it requires happiness as one component) is not obvious, indeed is in as much need of defending as any other view about well-being.

Oh, and:

The notion that “well-being” can be optimized irrespective of the happiness of the beings in question is something only a prisoner of a self-referential theoretical system could entertain.

Here’s an idea: let’s have a discussion where you imagine that people can sincerely hold other views than your own without being idiots or living in bad faith. Thanks.

48

seeds 03.07.14 at 12:37 pm

Straightwood:

Or, to grasp at another example, your interest in trolling the thread conflicts with my interest in reading the thread without it being derailed by your contributions. As I understand it, we could try to use CA to assess your level of freedom to provoke other contributors / become the centre of attention or, alternatively, my level of freedom to read the thread / become better informed.

Who can say which of those is a more noble set of goals? Happily, that isn’t what the OP is about – rather, it’s an attempt to pin down CA and distinguish it from the fields in which it could be used. Your attempt to mire everyone in argument about specific contexts misses the point. But I’m sure that you know that already.

49

Straightwood 03.07.14 at 3:09 pm

In deference to those who wish to explore CA without interference from those who “derail” the discussion, I ask what is the gold at the end of this rainbow. If CA flourishes as a successful theory guiding human relations, how will it impinge on politics, governance, and the daily lives of the people of the world? What can it supply that has been lacking in the examination of how we should live since Plato?

50

Bruce Baugh 03.07.14 at 4:04 pm

Ingrid, like some others I’m not even at the level of good questions yet, but I’m fascinated by this series. Thanks for doing it.

51

W R Peterson 03.07.14 at 5:43 pm

After reading the OP 3 times and the comments twice I think I understand a little bit and know I don’t understand a lot.

How does CA evaluate being a non-poor homosexual in an intolerant society versus being a poor homosexual in a tolerant one. Or a semi-poor homosexual in a semi-tolerant society.
Is there something internal to CA that guides the choice among those options or must additional judgements be added from outside? If outside judgements are necessary, are they sufficient without CA?

How does CA handle capabilities like the capability to own slaves? This specific example may not be the best, maybe it is easily dismissable. My intent was to try to find a case where the first normative claim in the OP clashed with the second, given that the second seems to allow people to define well-being how they want.

Perhaps a better example would be something Old Testament like the capability to suppress other people’s beliefs or behavior when those beliefs or behaviors risk bring down the collective punishment of God on us all.

Rawls used the veil of ignorance to deal with what I am trying to get at here, how does CA deal with it?

52

W R Peterson 03.07.14 at 5:48 pm

Above I tried to blockquote the part of the OP that described the 2 normative claims so people would have to scroll up to see what I am referring to but I am obviously an HTML noob.

two normative claims: first, that the freedom to achieve well-being is of central moral importance, and second, that freedom to achieve well-being is to be understood in terms of people’s valuable capabilities, that is, their real opportunities to do and be what they have reason to value.

53

Sam Clark 03.07.14 at 6:07 pm

Straightwood: so, are you admitting that you were mistaken to claim that obviously well-being = happiness? Or are you just dodging the issue?

54

Straightwood 03.07.14 at 7:14 pm

@53

Since the lexical silly-putty of “well-being” can be deformed into shapes different from happiness, there is no possibility of perfect equivalence, but people do know happiness when they find it. Happiness is a self-validating phenomenon, unlike what you call well-being, which appears to be best observed and proclaimed from the vantage point of advanced study.

The instances of well-being divergent from happiness you cite are drawn from speculative fiction that is far removed from the general experience of mankind. Those who would prescribe and enforce conditions for well-being that conflict with human happiness will carry a heavy burden of justification, one which CA theory will presumably be able to support, if the preservation of life, liberty, and the pursuit of well-being become our new desiderata.

55

MPAVictoria 03.07.14 at 7:25 pm

“In deference to those who wish to explore CA without interference from those who “derail” the discussion, I ask what is the gold at the end of this rainbow. If CA flourishes as a successful theory guiding human relations, how will it impinge on politics, governance, and the daily lives of the people of the world? What can it supply that has been lacking in the examination of how we should live since Plato?”

Well why don’t we just carry on reading what Ingrid writes and see where it takes us? I mean sure the approach would be unorthodox but it just might work.

56

Sam Clark 03.07.14 at 8:19 pm

Straightwood (last reply before signing off for the weekend):

Happiness is a self-validating phenomenon, unlike what you call well-being, which appears to be best observed and proclaimed from the vantage point of advanced study.

So, wait, having spent a lot of time thinking about and studying this matter counts against my arguments? In any case, I haven’t asserted any account of well-being: all I’ve said, against your confident pronouncements, is that the answer to the question ‘what is well-being?’ isn’t obvious. The ‘self-validating’ argument you gesture at here is actually an interesting one. But to argue for your view of well-being, you’re going to have to accept that it isn’t obvious.

The instances of well-being divergent from happiness you cite are drawn from speculative fiction that is far removed from the general experience of mankind.

Oh for goodness sake: I used literary examples because one of your earlier interventions in this thread complained about political scientists who don’t read enough literature. I could just as easily have described the ordinary human experiences they’re dramatising: the hollow happiness of consumerism, ignorance as bliss, deliberately sacrificing happiness to something that matters more to you. All of these are plainly part of human life, and plainly raise questions about what it is for a life to go best.

57

Ingrid Robeyns 03.07.14 at 9:35 pm

LFC@ 42 and 44: “I’d still like a better sense than I now have of what was the initial impetus for the whole approach”.

My reconstruction/interpretation is the following:

Amartya Sen was, around 1978-1980, involved in the following three academic sub disciplines: (1) empirical development economics; (2) theoretical welfare economics; (3) anglo-american political philosophy, especially the debate prompted by the publication of Rawls’s book ‘A Theory of Justice’.

Rawls tried to offer an alternative for utilitarianism with his social primary goods accounts; and utilitarianism was also the ethical theory underpinning theoretical welfare economics, and in a derived form all of mainstream economics, including increasingly also mainstream development economics.

Sen had written already a number of papers trying to attack utilitarianism, and sought an alternative *for its use in all these three sub disciplines*. That explains part of his attraction to Rawls’s theory, since it provided a full-blown alternative to utilitarianism. But Sen felt that Rawls’s social primary goods was still too focussed on material goods, and didn’t properly account for the different material needs that people have if they are, for example, disabled. He thus started to work on the notion of capability around 1979, *and had three targets in mind*: (1) to provide an alternative for the commodity-focussed metrics in empirical development economics; (2) to provide a theoretical alternative for utilitarianism; and (3) to provide an alternative for Rawlsian social primary goods.

What evidence do I have for this interpretation? Well, apart from the fact that I have never heard him contradict this interpretation [I had the privilege to talk a lot with him about the CA since he was my PhD supervisor], we can also see it from looking at Sen’s first papers on the capability approach. The first is his 1979 Tanner lecture, which is free to download here: http://tannerlectures.utah.edu/_documents/a-to-z/s/sen80.pdf
But there are a few empirical papers of the early ’80s where he presents research that could, most likely, also have been ‘theoretically framed’ in another language – perhaps the language of basic needs or of social indicators, – yet Sen choose to frame it as an issue of basic capabilities deprivation. In particular, I recall these two papers:
Sen, Amartya and Sunil Sengupta (1983). ‘Malnutrition of Rural Children and the Sex
Bias’, Economic and Political Weekly, 18: 855-864.
and
Kynch, Jocelyn and Amartya Sen (1983). ‘Indian Women: Well-being and Survival’,
Cambridge Journal of Economics, 7: 363-380.
These studies were published in 1983, but probably data were collected a few years earlier.

Summing up, in my understanding, Sen’s CA has *dual origins, in both economics and philosophy* – or perhaps, more specifically, triple origins in three debates: in discussions about development and empirical research in development economics; in post-Rawlsian discussions; and in theoretical welfare economics. It may well be that acknowledging these plural origins is the only way to make sense of Sen’s CA. It can also help explain why he didn’t want to adopt a modified version of the basic needs approach, which was flourishing in development economics/studies, since that would not have been a helpful approach if one were to construct a theory of justice that Rawls had in mind. To meet all the targets that he had simultaneously, Sen had to construct something new. That’s what he did, even though the first couple of papers he published on the CA presented the ideas in a quite embryonic stage.

58

Luis 03.07.14 at 9:38 pm

That is brilliantly useful context, Ingrid. Thank you.

59

LFC 03.08.14 at 2:16 pm

Ingrid: Thank you, that’s very helpful.

60

js. 03.10.14 at 3:58 am

One more thank you for the post @57. Super helpful.

61

John Quiggin 03.10.14 at 5:15 am

Thanks also from me

62

Sridhar 03.10.14 at 2:35 pm

Ingrid: welldone on these posts. They are really a nice effort to present some of the main ideas given the multi-disciplinary literature, and secondary literature that is often misguiding. I just had a comment on the ‘initial impetus’ for the CA. While you are right that it had its origins in both economics and philosophy, I would more specifically identify the origins of the capability approach in Sen’s re-evaluation of famines, and the Malthusian theory. The shifting of the paradigm of the analysis of famine starvation from the availability of food (i.e. commodities) to the entitlement sets (what bundles of food people are actually able to acquire) and the different factors that affect the different amounts of nutrition they need is then abstracted in the philosophical discussions about recognition of human diversity and commodities being the wrong target. Also, it is in the famine analysis where he also presents the issue that asking people, particularly women, about how they are doing in the midst of a famine results in misleading subjective assessments of wellbeing (i.e. mental states and subjective assessments are not sufficient or right targets.)

Once people understand that wellbeing is best assessed by what people are actually able to be and do, taken into account how a particular context is determines their abilities, then entitlement or capability set can be discussed in philosophical terms of distributive, relational, or other forms of justice. The famine analyis has been really useful for me in teaching about the conceptual and philosophical framework of the capapbilities approach as well as making the move from nutrition to health and wellbeing.

looking forward to your future installments! S.

63

Alex 03.10.14 at 2:54 pm

Isn’t part of the point of the CA to get around the issue of whether we are prescribing certain value-laden goods or activities or beliefs as being the stuff of a decent life, by leaving what they might be open to the individual? Capability is the ability to achieve them, whatever they might be; the range of potential possibilities.

This does leave us with the problem that some possibilities are intolerable to others – being the world dictator might be awesome, but that doesn’t mean society should strive to maximise my capability to achieve world dictatorship.

64

Alex 03.10.14 at 2:55 pm

Or perhaps it does, in which case I demand my nuclear bombs!

65

Luis 03.10.14 at 3:01 pm

Sen does address that problem with a caveat- basically that they have to be good things. (I can’t find the exact phrasing in my notes.) But I admit I found that caveat unsatisfying.

66

LFC 03.10.14 at 3:30 pm

Alex @62
Isn’t part of the point of the CA to get around the issue of whether we are prescribing certain value-laden goods or activities or beliefs as being the stuff of a decent life, by leaving what they might be open to the individual?

If so, in this particular respect CA would not really differ, istm, from Rawls in TJ or from standard versions of liberalism. (In the perhaps-now-played-out communitarianism-liberalism debates, one of the former’s points against the latter was its supposed reluctance to make judgments about whether certain activities are ‘better’ than others.)

According to Ingrid @57, Sen did not want to think in terms of ‘goods’ and also thought that Rawls’s ‘primary goods’ were too materially focused (though R. said that “self-respect” is prob. the most important primary good). Anyway, I don’t want to turn this into a thread on Rawls (we’ve had those before here and there’s no particular need, istm, for another one). Sen’s reasons for being dissatisfied w R’s approach I’m sure wd become clearer to me on reading some of his work (I’ve read things that quote and cite him but not much of Sen himself).

67

LFC 03.10.14 at 3:41 pm

OTOH if, as Luis says, Sen says they have to be “good” goals or activities or “doings,” then that does differ from R. in TJ. Obvs. no one advocates plainly megalomaniac or illegal activities — if your desire is to be a world dictator or a serial murderer you’re out of luck — but the question, at a more mundane level, has to do w how you want to spend your time, and on that R. is reluctant to make judgments, as in the counting-blades-of-grass passage in TJ that got some of his critics all riled up. But I tend to think this is, in the context of these posts of Ingrid’s, not an esp. important issue.

68

js. 03.10.14 at 7:38 pm

I am also curious about where/how/when normative elements come into the CA framework—some of this also dovetails with Matt’s comments on the other thread (and apologies in advance if this is somewhat OT). Anyway here’s what I was thinking about:

There’s the bit in ‘Equality of What’ (linked @57), where it looks like Sen balks at Rawls’ view that satisfaction derived from morally reprehensible actions or practices—e.g. enslaving someone—shouldn’t count at all in a determination of social value. To quote (p. 211 via the @57 link):

Welfarism requires the endorsement not merely of the widely shared intuition that any pleasure has some value—and one would have to be a bit of a kill-joy to dissent from this—but also the much more dubious proposition that pleasures must be relatively weighed only according to their respective intensities, irrespective of the source of the pleasure and the nature of the activity that goes with it. [emph. added--js.]

This suggests, at least to me, that contrary to LFC’s suggestion (@67), Sen is more permissive than Rawls when it comes to determining which functionings are or can be valuable. In other words, as per Ingrid’s suggestion here and in the other threads, Sen wants to stay quite neutral regarding which capabilities and functionings might have value—even intuitively reprehensible ones may turn out to have some value. I guess I find this unpersuasive. In particular, I don’t see how one can build up to a more robust conception of justice unless one is already relying on some sort of a normative (or proto-normative?) framework that allows us to make normatively grounded distinctions between different sorts of capabilities and functionings. But perhaps Sen provides this elsewhere? Or maybe Nussbaum is a good person to go to for this sort of question?

[Brief excursus on Rawls deleted in order to stay vaguely on topic.]

69

ingrid robeyns 03.10.14 at 8:16 pm

Sridhar – thanks for the famine links! I haven’t read Sen’s famine stuff for ages, and am currently at a residential phD course on ethical theory, so can’t check – but I think you are quire right about this. Siddiq Osmani has written a few great papers about “the Sen system of social evaluation” where he links his work on entitlements and capabilities. I should reread that but this may take me a while.

On which capabilities count: I agree with js. @68 that Sen’s CA is more permissible than Rawlsian social primary goods; Sen says that the CA should focus on those capabilities “we have reason to value”. The question is what that “reasoning to value” precisely entails and who does this – and, importantly, whether these are collective or individual reasoning processes. Too much for now, I fear, and I think that one has to reconstruct Sen, since I can’t immediately think of a paper where he has given a succinct account of what “reasoning to value” precisely entails in the CA.

70

Luis 03.10.14 at 8:31 pm

I was just about to come over here and paste the “reason to value” quote; would love to hear more about that at some point, Ingrid.

71

LFC 03.11.14 at 12:14 am

js. @68:
I wasn’t suggesting anything independently at 67 (as I think you’ll see from a re-reading plus the surrounding context), but I’ll refrain from further comment at least until I’ve read the link @57.

72

js. 03.11.14 at 12:18 am

Right, yes. I made it sound like you did, but I didn’t mean to do that. Sorry.

73

LFC 03.11.14 at 12:29 am

js.:
No problem.
As I say, until I’ve informed myself a bit better I’m going to clam up for a while and let the philosophers (of whom you are one, I believe) do the talking. ;) [plus I happen to be totally zonked rt now]

74

Ed Herdman 03.11.14 at 5:37 pm

I am confused about the nature of the problem with the more permissive “reason to value” criteria from Sen, over a more-rigid system. I think the wording still implies that somebody out there gets a vote on what somebody else’s goods are – say a politician or bureaucrat – so the libertarian will not be pleased. Still this seems like an improvement over a more rigid system, while the system does not become too unwieldy to use.

Comments on this entry are closed.