Bernard Salaine (Columbia University)
Kalai Family Workshop in Econometrics
Mar 22 2018
Back to All Events
Fast, Robust, and Approximately Correct: Estimating Mixed Demand Systems, with Frank Wolak
We propose an estimator for mixed demand systems that is exactly valid when there is no random variation in preferences, and becomes a coarser approximation as the amplitude of random variation grows. In the BLP framework it can be obtained as is a run-of-the-mill 2SLS regression, and can be computed in microseconds with off-the-shelf software. Moreover, it need assume very little on the distribution of the random variation in preferences, justifying the “robust” in our title. Since our estimating equation is linear, computing the optimal instruments is also straightforward. The method performs very well in Monte Carlo simulations.