Inequality of Capability?

by Will Wilkinson on May 21, 2008

Lane Kenworthy kindly responds:

We should care about inequality of income not simply because it contributes to inequality of well-being, but also because it contributes to inequality of capability.

Even if consumption inequality has increased only a little, the rise in income inequality has produced a noteworthy increase in inequality of capability. The rich aren’t forced to purchase goods and services whose prices have increased more rapidly; they could switch to the same consumption bundle as the poor if they wished.

I feel like maybe we’re speaking a different language or something here. I’m not sure what Kenworthy means by “inequality of capability.” It sounds to me like he is using a currently popular term to repeat the not-contested because self-evident fact that people with a bigger budget have a bigger budget. And I’m not clear on the source of Kenworthy’s preference for thinking about what people can do with their money (capability) over what they do do with it (consumption), since it seems to me to come to pretty much the same thing, unless you think people really do systematically tend to act against their interests. No one disputes the fact that wealthier people have more options. That’s basically what it means to be wealthier. But isn’t the issue at hand the size of the difference in real quality of life once the wealthy and the poor have taken their best options? Anyway, that’s what I think the question is. If the best a poor person’s budget can buy improves faster than the best a rich person’s budget can buy, then that means the poor person’s quality of life has become more like the rich person’s even if the rich person’s budget expanded at a faster rate than the poor person’s. The analysis applies mutatis mutandis to capabilities. Isn’t this absolutely central to the question of economic inequality? Isn’t this in fact the central question of economic inequality — the gap in the quality of the lived experience of the rich and poor?

Kenworthy goes on:

In my view the Broda and Romalis analysis is important for our understanding of (absolute) poverty, rather than inequality. They find that the prices of goods poor Americans tend to purchase have risen less rapidly than the overall inflation rate. I can’t assess whether they’ve accurately analyzed the data and how much measurement error the data contain. But if the finding is correct, it suggests that the trend in living standards for America’s poor was more favorable (or less unfavorable) between 1994 and 2005 than income data imply.

I don’t get it. If living standards for the poor are better than we had thought, doesn’t that mean that the difference in living standards between the poor and the rich is smaller than we thought?

  • John Thacker
    The fact that these things were not in your poor person’s consumption bundle to begin with, does not mean that you are indifferent to them becoming even more unobtainable.

    Luis Enrique, IMO, the relevant question is what sort of sacrifices and changes one would have to make in order to obtain these things, whether that means changing jobs or getting education, or skimping on other things in order to splurge on the occasional nice item (or nice private education, or whatever.) Depending on the situation, they're not necessarily even more unobtainable.

    For someone who is poor, since such a large percentage of the consumption basket is cheaper goods, then these sorts of changes make make the one or two expensive items more affordable, not less. Save enough at Wal-Mart and even if that high class item gets more expensive quickly, you still have enough extra left over in your budget to afford. The greater the proportion of one's income that's already spent on the more expensive goods, the more it becomes harder to afford those goods or to buy more.

    Thus, the complaint to me seems to be more that it's harder for the middle and upper middle class to keep up with the upper class by buying many expensive goods, rather than a complaint that it's harder for those of modest means to continue buying a small number of expensive goods. It's related to the old "are America's rich falling behind the super-rich" complaint.
  • John Thacker
    Yeah, one problem with rampant inflation for the rich, is that it sharply reduces the probability that poor people can partake in new products geared towards the rich.

    Perhaps, but not necessarily. It's often more complicated. If I'm poor, then an iPod-like product moving from $200 to $300 makes it harder for me to buy it, but if I save over $100 on cheaper clothes from China, then it's easier, not harder to partake in the iPod.

    There are a lot of people of modest means who save by buying lots of cheaper products buy who splurge on the occasional expensive item, whether it's an iPod, a nice car stereo and speakers, or nice athletic shoes. If expensive products all rise in price simultaneously faster than cheaper products, it might be easier to buy one or two expensive products but harder to buy them all-- and that can promote consumption equality, not inequality.
  • Luis Enrique
    But mk, if you are poor, you do know what you are missing out on. You know what houses in nice neighbourhoods are, you know what an expensive education is, and what comprehensive health insurance is (not to mention yachts and German kitchenware). The fact that these things were not in your poor person's consumption bundle to begin with, does not mean that you are indifferent to them becoming even more unobtainable.

    What's the measure of inequality that matters? The fact that the living standards of the poor have moved closer to the rich, in so far as they can now afford more cheap food & cheap imports ...or the fact that the rich person's consumption basket has moved further beyond their reach. I lean toward the latter - I think this research shows inequality has increased.
  • mk
    Conchis,

    I'll concede it's possible that the size of the set means something for my well-being. It can feel good to have options (as you say, it can also lead to stress).

    Actually it seems clear that it must mean something, and the question is to quantify the effect.

    I think things come into play here which can be tough to analyze economically, e.g. "awareness". If I don't know or haven't seen what I'm missing out on, it probably matters less to me.
  • conchis
    mk, I half agree with your analysis of the fasting example (which is why I mentioned that I wasn't convinced it demonstrates Sen's point).

    Nonetheless, it still seems plausible to me (and consistent with what limited evidence there is) that the size of individuals' choice sets has an effect on "utility", independently of the specific choice made. Being constrained in our choice sets is typically unpleasant and often quite stressful, even if we are constrained only to pick the option that we would otherwise have picked anyway.*

    Imagine that the would be faster were actually unable to feed herself. She might be able to rationalise her not eating as politically righteous in exactly the same way as if she had chosen to fast (at least, let us suppose she can) - and yet she might still feel worse off because she also values the size of her choice set and resents not having the ability to choose. It's then no longer the case that her utility is determined solely by her chosen bundle of goods: it's determined both by the chosen bundle, and the choice set from which that bundle was chosen.

    * Of course, sometimes being constrained in our choice sets can reduce decision stress, but that doesn't detract from the point that unchosen options may be important in themselves.
  • mk
    conchis -- I would object to the fasting example. The decision to fast here is made for some compensating benefit (feeling politically righteous, let's say) so the person's overall utility is pretty high.

    That, to me, is the deciding factor. The fellow's well-being is still being determined by the utility he gets from his chosen bundle of goods (where "goods" I will concede needs to be expanded to include more intangible things). His well-being is not a matter of the size or nature of his feasible set, only of his choice within the set.
  • conchis
    whoops, should have read, "chooses to fast, though they have the means to eat".
  • conchis
    "I’m not clear on the source of Kenworthy’s preference for thinking about what people can do with their money (capability) over what they do do with it (consumption), since it seems to me to come to pretty much the same thing."

    Sen argues extensively to the contrary in his work developing the capability approach. Pretty much anything of his in the area is likely to give the flavor of the argument. His most commonly-used example is the intuitive difference between someone who is starving because they can't afford to eat, and someone who chooses to fast, though they have the means to do so. While the two have the same consumption, we are unlikely to consider them equally well-off, and Sen argues that the reason for this is that the faster has greater capabilities. (FWIW, I'm not sure this particular example necessarily demonstrates the point, but I nonetheless find it fairly plausible more generally.)
  • Luis Enrique
    mk

    What worries me is that the goods in the rich consumption bundle may include a good education, health care and housing in a safe neighborhood - rather than just yachts and designer German kitchenware.

    Goods like these (education etc.) probably include feedbacks to income, so if they are getting further beyond the reach of poor people, that looks like pretty bad news to me - news which is only partially mitigated by the fact that cheap food, clothes and goods in the poor consumption bundle have gotten cheaper.
  • mk
    Luis - Yeah, one problem with rampant inflation for the rich, is that it sharply reduces the probability that poor people can partake in new products geared towards the rich.

    Ten years from now, if some new IPod-like thing comes out, it might be so ludicrously expensive that poor people will basically never own it. If inflation were stamped out, the introductory price would be lower, and more poor people could buy the product. So the differential inflation does have long-term utility impact for the poor.

    One could imagine that an increasing differential in nominal wages increases the time lag between something being a "vanity" product for the rich, and full market penetration, merely because incentives become skewed for companies selling the product.

    On the other hand, maybe the existence of the super-rich creates more incentive to develop nifty vanity products, which leads to further innovation and some beneficial trickle-down. Maybe the next big thing won't be developed unless there are some super rich people willing to pay ridiculous amounts for it.

    It's a tough question.
  • Luis Enrique
    Will,

    I think you can interpret 'capabilities' as consumption possibility frontier, feasible consumption set, something like that.

    mk, I'm going to try to get you to bounce back again. I don't think it's right that "we only care about price changes that impact the utility of my optimal bundle." I think that with the relative income and price movements we are discussing, some of the things that the rich enjoy have gotten even further beyond the reach of the poor, even though these things are not in the poor's optimal consumption bundle, and that matters when we're thinking about how unequal society is.
  • Changes in relative *current* incomes says nothing about the change in relative "capabilities".

    Capabilities or opportunities or the size of the budget set or whatever are determined by lifetime (not current) incomes. Current income has a tenuous relationship to lifetime income. (see the Ambrosini Critique link above for citations)

    This exposes a potential problem in the B/R paper. People use their current incomes to purchase current consumption and to finance past or future consumption. Deflating income by the cost of current consumption baskets involves deflating some amount of future and past consumption by current cost of living. In other words, just as current income is bad proxy for lifetime income, current cost of living could potentially be a bad proxy for lifetime cost of living.
  • An aside: interpersonal utility comparisons being inherently difficult (if not impossible) was an insight of Pareto. Yet we make working assumptions like most people's utility functions being pretty similar absent revealed preferences to the contrary. If people did not really think that such assumptions were fundamentally justified, we wouldn't give gifts because to make sense that practice requires such assumption. If we truly thought interpersonal utility comparisons were impossible, we wouldn't.
  • mk
    I am bouncing back and forth, and I think Lane is wrong again.


    It is not true that a mere restriction in my consumption possibility frontier lowers my utility. As Will says, it is only true if it shifts my purchased basket to a lower-utility one.

    Of course a mere increase in IPod price, all else being equal, reduces the summed utility of all poor people. Because some poor people buy IPods.

    But all else is not equal. Crappy imports are declining in price really fast, and that increases my options as a poor person, in a way that increases the utility of my optimal bundle.

    It also increases the options of a rich person, in a way that does not increase the utility of their optimal bundle.

    As Will is saying, we only care about price changes that impact the utility of my optimal bundle.



    To add a caveat to Will's analysis, the study does not purport to measure the relative utility of rich vs. poor. Instead it measures relative bundle-purchasing power.

    This is not the same thing: If 1) I can afford twice as many poor bundles as last year, and 2) a rich person can afford twice as many rich bundles as last year, we don't know what happens to relative utility.


    In other news, we badly need an economist to help us understand the methodology of the CPI, not to mention this study. Tyler?
blog comments powered by Disqus

Previous post:

Next post: