Neighborly Competition in Real Estate Transactions

This paper studies price competition in residential real estate with a spatial framework using micro-level data. Using a sample of real estate transactions from 1993-2014 in Philadelphia, PA from the Multiple Listing Service (MLS) sales reports, a multinomial logit model is used to measure effects of competition on probability of sale and a repeat sales approach is used to measure effects of competition on sale price. The results indicate a that a higher measure of competition decreases the probability of selling and increases the probability of dropping out of the market without a sale relative to remaining an active listing. Additionally, competition has a significant negative effect on sale price prior to 2004 and a positive effect after 2006, suggesting that the nature of competition in local real estate markets has changed over time