Thursday, November 13, 2008

NBER paper on IV

For those of you with dodgy or no instruments (we know who you are) see this recent NBER paper
Identification with Imperfect Instruments
A Nevo, A Rosen
Dealing with endogenous regressors is a central challenge of applied research. The standard solution is to use instrumental variables that are assumed to be uncorrelated with unobservables. We instead assume (i) the correlation between the instrument and the error term has the same sign as the correlation between the endogenous regressor and the error term, and (ii) that the instrument is less correlated with the error term than is the endogenous regressor. Using these assumptions, we derive analytic bounds for the parameters. We demonstrate the method in two applications.


Liam Delaney said...

if not already written, someone should write a paper on the assumption that the author has just searched for an instrument that works on something rather than actually tried to answer a question.

Kevin Denny said...