Saturday, January 13, 2007

Sociologists and Income Inequality

by Unknown

In the comments to the Book Meme II post, Tina and Tom ask, well, why *have* sociologists largely ignored the past 30 years' growth in income inequality?

The answer my co-authors and I gave in the offending article was that sociologists have a long-standing disciplinary commitment to measuring inequality with either socioeconomic scales (SEI) or social class categories. Neither of these reference the income distribution in any direct way. SEI is a weighted function of the average (sort-of) income and education in a person's occupation, not individual-level incomes. In sociology, social classes are defined by position in the labor and capital markets, and are most commonly measured by a combination of occupation and self-employment status. For the class analyst, earnings and income are but two of many "life chances" that derive from the more fundamental class position.

So, part of the answer lies in the assumptions sociology has made about how inequality should be measured. This answer allowed us to proceed on our merry way in the paper, which documents changes in the class-based "lumpiness" in the income distribution over the past 30 years. (Short answer: it's increasing.)

But I think there's more to it than that. Sociologists are interested in income inequality, but just in a very different way than the labor economists who dominate the rising income inequality literature. For at least the past 50 years (far before economists' current forays into the topic), sociologists have focused on understanding the social rules of allocation -- why certain types of people do well while others don't. This has generated literally thousands of articles that document and explain how family background, race, gender, education, and so forth affect where people end up in the inequality space, whether it is measured by SEI, class, income, earnings, wealth, or what-have-you.

For all their richness, most of these articles take the structure of the inequality space as given. In modeling terms, sociologists tend to be more fascinated by between-group differences in means (and, of course, endless permutations of interaction terms) than by the overall variance. The relative lack of interest sociologists have shown toward rising income income inequality is thus a manifestation of a more general tendency to focus on the rules of allocation rather than on the overall distribution of rewards.

There is much more to be said about the disciplinary divide, the extent the measurement of inequality on both sides relies on untested assumptions, how social scientists in general might move forward from here, and so forth. It probably goes without saying that I think contemporary sociologists should devote more attention to the overall distribution of rewards and the structure of the inequality space, asking questions that economists -- with the exception of some Sen-inspired development economists -- simply aren't asking. But, I'm sure I've already bored you all to tears.
Comments:
For all their richness, most of these articles take the structure of the inequality space as given. In modeling terms, sociologists tend to be more fascinated by between-group differences in means (and, of course, endless permutations of interaction terms) than by the overall variance. The relative lack of interest sociologists have shown toward rising income income inequality is thus a manifestation of a more general tendency to focus on the rules of allocation rather than on the overall distribution of rewards.

You join a distinguished line of sociologists on this point. On p25 of this paper (which I inflicted on you and others a while back), a parade of people going back to the '50s -- S.F. Nadel, Harrison White, James Coleman, Bruce Western & Martina Morris -- try to remind us of the distinction between analyzing the allocation individuals within a field of positions and analyzing the creation of positions in a field. Something similar can be seen in political philosophy since Rawls. Clearly we keep forgetting this. I think it has something to do with the difficulties of formally analyzing this kind of process in the first place.
 
Very interesting, Kim. Thanks so much for elaborating. I agree with Kieran that sometimes we just go the easy way instead of the way that tells us the most. Can another reason be the focus on the single case? It seems that comparative research might reveal some differences in structures.
 
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?