Once you get bitten by the endogeneity bug…*

... you stay bitten.

I was reading Agatha Christie’s Appointment with Death the other day. The most unlikely place where you would imagine a Heckman selection issue to surface, right?

Wrong! The character of the horrible Mrs. Boynton is revealed, and it is also revealed that she had been a jail wardress. In the initial chapters of the  book, two psychologists are discussing her, when the younger one theorizes that perhaps her profession had made Mrs. Boynton the sadist that she was. The older one, however, opines this to be not quite true. Perhaps he did read Heckman’s seminal paper on the selection bias (though unlikely, since this book was published in 1938, whereas the paper was published in 1979, unless he had a time machine!). Nonetheless, he rationalizes that Mrs. Boynton became a jail wardress because she was a sadist.

Aha! Selection bias at play! Need a Heckman correction pronto!

I think I am too much enamored by Heckman.

For those who are interested, here is the citation: Heckman, J.J. 1979. Sample Selection Bias as a Specification Error, Econometrica 47(1), pp. 153-161.

* Cross posted from my personal blog here.