In a comment the other day I alluded to the impossibility of having a society that’s both a perfect meritocracy and highly unequal. I hadn’t actually realised how strong the relationship between these two aspects was, until seeing Paul Krugman’s post “The Great Gatsby Curve” today. Here’s the chart that sums it up:
Note that “elasticity” here refers to the elasticity of a child’s earnings where the controlling variable is the parents’ earnings. The relationship is both stronger and more straightforwardly linear than I had expected.
As always with these things, the points that don’t fall right on the line are interesting to consider: particularly the UK, France & Japan, which had very similar Gini coefficients in 1985, but a fairly wide range of intergenerational earnings elasticity. I have no particular insight into why this would be so, and I’m interested in any ideas you might have.
No firm answer, just a comment that these are only a handful of datapoints, at a fairly high level of analysis. It might be interesting to think whether there is another level at which the data could be analyzed, to see if the same general pattern still holds. Longitudinal data on countries might be one idea, to see what happens as one variable shifts over time. Or smaller geographic units, perhaps?
More generally, I think meritocracy is a problematic idea to start with. Andrew Gelman has said something I really like on the subject: http://andrewgelman.com/2011/12/lamentably-common-misunderstanding-of-meritocracy/
You’ve got me thinking. One thing I do know a priori is that Japan urbanised very much later than Britain or France and then did so really quickly – I wonder if that upheaval has an impact on social mobility, in which case we should be able to find it by comparing time periods. And the chances are that at least Scotland compiles similar statistics for itself as the UK as a whole, so we could probably compare like with like for Scotland vs rest-of-UK, which would be worth a look.
I agree with you about meritocracy as a problematic concept, and I think James Flynn’s point about those with status using it to benefit their offspring is the [or at least, a major] causal factor behind this relationship. I seem to recall first encountering that idea through Gelman’s blog because you linked to it, so: thank you.