Homeowners? Repeat-Sale Gains, Dual Agency and Repeated Use of the Same Agent

Previous studies of dual agency, where one agent serves both buyer and seller in a transaction, use hedonic models. Repeat-sale methods can test for the price effect of accepting dual agency. Dual agency does not show convincing effects on expected gain, which would occur if there was a systematic bias, or on heteroscedasticity, which would occur if there are large effects that are rare. Earlier researchers could not test for the effect of an owner picking a listing agent who was the earlier selling agent. Consistently positive mean abnormal price gains come from this choice, as well as signi?cant heteroscedasticity.