The persistence of migration outcomes

How persistent are economic gaps across ethnicities? The convergence of ethnic gaps through the third generation of immigrants is difficult to measure because few datasets include grandparental birthplace. I overcome this limitation with a new three-generational dataset that links immigrant grandfathers in 1880 to their grandsons in 1940. I find that the persistence of ethnic gaps in occupational income is 2.5 times stronger than predicted by a standard grandfather-grandson elasticity. While part of the discrepancy is due to measurement error attenuating the grandfather-grandson elasticity, mechanisms related to geography also partially explain the stronger persistence of ethnic occupational differentials.

That is the abstract of a piece by Zachary Ward, from American Economic Journal: Applied Economics.  In a number of regards this paper goes well beyond the previous literature.  Here is another interesting sentence:

…I find that 51 percent of initial ethnic gaps in occupational income remained after three generations.

The author also notes:

Rather than argue for an ethnic-specific causal mechanism, I instead point to measurement error and geography as key reasons for the stronger persistence of ethnic differentials across three generations.

I am not so convinced, as where you choose to live is endogenous to your expected labor market quality.  I am somewhat more persuaded by this point:

…the ethnic mean provides more information about the father’s true occupational status.

Iin other words, what appears to be an influence of ethnicity might instead be a transmission channel through the background of one’s own father.

At times the author seems naive, at other times Straussian, or maybe just afraid?  To be clear, I am myself an extreme culturalist, and that is not a Straussian remark.  This is in any case a major contribution to a contentious debate.