We propose a two-stage instrumental variable estimator that is consistent if there is selective compliance in the treatment group of a randomized experiment and the outcome variable is a censored duration. The estimator assumes full compliance in the control group. We use the estimator to re-analyze data from the Illinois re-employment bonus experiment.

Instrumental variable (IV), Non-compliance, Randomized experiment, Treatment effect
dx.doi.org/10.1016/j.jeconom.2004.04.004, hdl.handle.net/1765/1998
Journal of Econometrics
Sociaal-Economisch Onderzoek Rotterdam BV (SEOR)

Bijwaard, G.E, & Ridder, G. (2005). Correcting for Selective Compliance in a Re-employment Bonus Experiment. Journal of Econometrics, 125(1-2 SPEC. ISS.), 77–111. doi:10.1016/j.jeconom.2004.04.004