Sample Selection Approaches to Estimating and Allocating House Transaction Funding Price Differentials

Abstract

This study examines house transaction price differentials observed among funding type combinations; accounting for potential sample selection and spatial biases yields a better approximation of price differentials between group combinations. Consistent with expectations we detect, and correct for, selectivity and spatial biases. Transactions with conventional financing have superior characteristics compares to all-cash funded transactions, and Federal Housing Administration (FHA) and Veterans Affairs (VA) funded houses have inferior characteristics relative to all-cash characteristics. price counterfactuals for (1) all-cash financed property, (2) conventional, (3) FHA, and (4) VA property transactions reveal, consistent with expectations, unexplained housing/neighborhood characteristics, are superior relative to FHA and VA financed properties. Results reinforce the notion that credit matters in the provision of financial services with regards to housing prices, while Blinder-Oaxaca price differential decompositions provide additional insights.