So these days we go data, representation, interpretation, wild speculation?

Why presume that differences in the impact of merit are at work here? Why not demographics (how much variation is there among peers of the same generation?), geography (are children more or less likely to move rural to urban or vice versa), education (are children more or less likely to match parents’ education levels), legal considerations (what impact to inheritance or other inter-generational taxes play?), or a host of other things?

It’s interesting data, and I’d love to explore the root causes more. But the conclusion seems really spurious. I think most people would agree that a lower correlation (or elasticity, which is *very* different) is more socially desirable. But maybe it’s worth following the old school of actually analyzing the data and testing hypotheses before moving on to prescription?

The article offers a rather cunning argument against the logic or meritocracy, and in the end, it does end up condemning the bonuses paid out to wall-street bankers. In short, they do not add the scale of value to their companies that they claim to (in fact it would be impossible for them to) and so do not deserve the exorbitant bonuses; the myth that they do add that value derives from the rather naive society wide myth of merit.

The author and webpage is published by people that I happen to rabidly disagree with on many issues (like capitalism and the free market), but at least as far as starting to get knowledge, skill, society and social interaction right, if the idea catches on, it could be a big step for the right :)

Commenting on the `why’ should probably not proceed before we figure out the `what’.

That said, the caption for the graph in the Times article explains their coefficients as follows: “The higher the value, the greater is the persistence of earnings across generations, thus the lower is the intergenerational earnings mobility.”

For example, if the statistics are estimated from after-tax estimates of income, the lower coefficients could be due to anything from intergenerational market mobility to government transfers and progressive taxation regimes.

Or, depending on what the authors controlled for, it could be measuring rural/urban inequality and migration patterns. This is exactly the kind of thing the second graph is supposed to be taking into account. See how Italy slid from one end to the other–that suggests that this is probably a major source of inequality in Italy. It’s probably different in other countries for different people.

Folks should consider going and reading the articles cited in the footnote to the Times article. I’m sure they probably have something to say about it.