Re: st: RE: Testing for instrument relevance and overidentification when the endogeneous variable is used in interaction terms

[Jason Wichert <jasonw8907@gmail.com>]

Just as FYI for the silent followers of the discussion and future
researchers browsing the stata archives for instrumentation of
non-linear functions (and, admittedly, an attempt to revive this
discussion :-) ), since the approach doesn’t seem that common:

I’ve tried to read up more on the “control function approach” as
dubbed by Wooldridge (2002), and building on – or at the very least
closely related to - Hausman’s (1978) test for endogeneity.

Another name is “Two-stage Residual Inclusion” or 2SRI, and it’s
explained in more detail by Terza et al. (2008). While they show how
to apply this method for nonlinear models, they do not explain any
application for endogenous interaction term, ‘though.

Terza, Joseph V., Anirban Basu, and Paul J. Rathouz (2008): Two-Stage
Residual Inclusion Estimation: Addressing Endogeneity in Health
Econometric Modeling, Journal of Health Economics, Vol. 27, Issue 3
(May), pp. 531-543.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2494557
http://ideas.repec.org/a/eee/jhecon/v27y2008i3p531-543.html

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/