A Simplified Klein–Spady Estimator for Binary Choice Models

One of the most cited studies within the field of binary choice models is that of Klein and Spady (1993), in which the authors propose an estimator that is not only non-parametric with respect to the choice density but also asymptotically efficient. However, while theoretically appealing, the estimator has been found to be very difficult to implement with poor small-sample properties. This paper proposes a simplified version of the Klein–Spady estimator, which is shown to be easy to implement, numerically relatively more stable, and with excellent small-sample and asymptotic properties.

Online_appendix_Hjertstrand_et_al.