Sample Selection Approaches to Estimating House Price Cash Differentials

Abstract

We extend the literature on house price cash differentials in important ways. First, our paper is the first to employ methods to correct for sample selection bias, using both switching regression and propensity score matching of cash vs. non-cash transactions. We use selection models to produce price counterfactuals for cash and noncash buyers. We also include both average treatment effect and a propensity score weighted selection models. From the selection models, we find that previous studies likely overstate the cash discount. Results from counterfactual tests examining cash discounts suggest amplified cash discounts in areas with close proximity to an environmental hazard; and also a pricing differential based on CBG level income, with purchasers in high income areas more likely to pay a cash premium compared to market participants in areas with comparably lower income, where a cash discount is detected. These results provide useful insights for market participants including real estate appraisers, brokers, and buyers and sellers of real estate.