Chapter 11, the first chapter in a part of the book entitled « The Materials of Justice”, presents us with Sen’s well-known theory of capabilities. The main focus of the chapter is to emphasize that the capability approach is essentially a theory about human freedom, or more precisely, a theory about how freedom should be factored into the assessment of advantage and disadvantage. As against welfarist construals, Sen points out that we care not just that we achieve what we want, but also how we achieve what we want. Whether what we achieve results from our own agency, and whether we were able to exercise our agency on a range of valuable “functionings”, matters to an assessment of how well we do just as much, if not more, as does the result of our activity. We should be interested in “comprehensive outcomes”, not just “culmination outcomes”.
Two aspects of the capability approach are emphasized by Sen. First, the theory has an “informational focus”. As opposed to the approach developed by Martha Nussbaum, Sen’s just tells us what we should be concerned with in the measurement of advantage and disadvantage. He does not tell us what we should do with that information once we achieve it. Nor does he fill in the detail about what, precisely, we have reason to care about. The theory is thus neutral, at least on its face, as between different approaches to distributive justice – egalitarian, sufficientarian, prioritarian, and so on, as it is between various ways of filling out the detail of what we have reason to care about.
Second, the theory is pluralistic. There are a range of things that we have reason to value, that cannot be reduced to one metric. Bundles of desirable functionings will reflect this.
Sen deflects two worries about the capability approach. The first comes from welfarists like Arneson and Cohen who argue, on Sen’s way of putting their arguments, that we should be concerned with what people actually achieve, rather than with what they can achieve. Sen’s response to this is that the capability approach includes the welfarist approach because the functionings realized by an individual are part of the set of functionings that he could achieve. Making this broader set the index of his advantage provides us with better information about advantage because it includes freedom to achieve a range of bundles as an ingredient.
The other worry has to do with commensurability. How can we evaluate options given the irreducible pluralism of reasons to value that Sen affirms? Sen responds by stating, in line with much of the recent literature on evaluation, that incommensurability makes evaluation harder, but not impossible.
The chapter ends with three elaborations. The first points to the important role of public debate and deliberation in the process of evaluation of diverse bundles of functionings, especially in the context of pluralism. The second emphasizes the importance of community to human flourishing while denying that communities can themselves be seen as subjects of capabilities. And the third provides us with a sketch of what an environmental ethic that took the capability approach seriously would look like. Basically, it would split the difference between a utilitarian approach (“the environment matters because of its contribution to utility”) and a non-anthropocentric approach (“the environment matters, period”). On Sen’s view, the environment should be taken to matter because of the way in which it contributes to the full range of things we have reason to care about, which range is not reducible to that which conduces directly to our welfare.
I have questions and concerns about all of these steps. A first question has to do with the measurement of the kind of freedom that the capability approach affords. Sen argues, plausibly, that it makes a difference to the measurement of my advantage whether I am constrained, by circumstance or human agency, to do what I happen to want to do anyway, or whether I have been able to choose what I do among a range of options that I have reason to value. But how precisely does the presence of more available bundles of functionings increase my advantage? Is the function that exists between increases in available bundles and increases in advantage linear, such that the 1001th bundle matters just as much as the 2nd? Or is there some kind of diminishing marginal return? And if so, at what threshold does the return begin to diminish sharply, such that the effort deployed in trying to provide more options dwarfs its value to the agent?
This is not a purely theoretical worry. One can imagine two societies each of which provides its citizens with means to access bundles of functionings, so that citizens in each case choose rather than being simply constrained, but where the second offers its citizens, say, ten times the access to desirable bundles as the first.
A possible answer would be to point out that Sen is not interested in choice for the sake of choice. Bundles of functionings that do not allow us to realize what we have reason to value are disqualified.
This leads me to my second question/worry. What constraints exist within the Senian framework on the reasons that apply to an agent? Remember that Sen claims to be neutral as between various ways of fleshing out the substance of the theory. What’s more, as we learn from previous chapters, we should avoid parochialism about these issues (a desideratum that drives Sen’s choice of a Smithian “impartial spectator” as opposed to a Rawlsian procedure that restricts moral deliberation to a specific political community). So presumably, we should think about advantage/disadvantage transnationally rather than on the basis of the values that are contained in the political culture to which we happen to belong. There is a tension between Sen’s Smithian proclivities and the importance he ascribes earlier in the book to “positional objectivity”. This tension can be summarized through the following questions: how far does where I am situated condition what I have reason to value? To what degree is my particular position to be perceived by a theory that wants to emphasize freedom to be, to do, and so on, as a freedom-inhibiting constraint, and to what degree is it an enabling constraint?
Answers to these questions will in turn have an impact on Sen’s ability to respond to a third question, which has to do with his agnosticism as between different theories of distributive justice. Egalitarianism is probably written out of the story if the answer to the first question is that all additional available bundles of options contribute equally, or significantly, to a person’s advantage. But if there is an identifiable drop-off point, presumably we can aim at equalizing all agents’ being able to access that point.
A fourth question has to do with Sen’s sanguine answer to the commensurability worry. Granted that the kinds of metrics that utilitarianism or “primary goods” afford are unavailable without doing violence to what it is that people actually value, unless we have some rough sense of what counts as better and worse, and of how to measure increments of betterment and worsening in rough and ready ways, the theory risks lapsing back into a kind of Ross-style intuitionism. Now this is not necessarily a bad thing, but presumably not what Sen had in mind.
Here is an example to illustrate my worry. Imagine a fishing community in Newfoundland whose traditional way of life has been decimated by a range of economic and environmental factors. There are two way forward: one is to subsidize that way of life in order to make it viable despite these factors. Another is to facilitate the entry of Newfoundlanders into a “modern” economy. The “positional objectivity” of Newfoundlanders might at first glance make the first option seem more attractive (though, note, it may not actually increase the range of available bundles of functionings available to them), but as they engage in exercises of Smithian enlarging of perspective, they may come to appreciate to a greater degree the range of options that a modern economy makes possible, and which such an enlarging now gives them reasons to value that they did not have antecedently.
How does the capability approach speak to a real-world case such as this one? An answer suggested by the section of the chapter entitled “Valuation and Public Reason” is that this is for Newfoundlanders to figure out as they publicly deliberate about the bundles of functionings at issue. But this would be bad news for a theory like Sen’s since it risks transforming it into a theory of deliberative democracy, where the work is done by deliberation rather than by capabilities and functionings. That is my fifth concern: how much of the filling out of the theory can be delegated to deliberation without emptying it of its distinctive content?
I have some other questions/comments, but I have nattered on enough for now.