Rule-Based and Case-Based Reasoning in Housing Prices

People reason about real-estate prices both in terms of general rules and in terms of analogies to similar cases. We propose to examine empirically which mode of reasoning fits the data better. To this end, we develop the statistical techniques required for the estimation of the case-based model. It is hypothesized that case-based reasoning will have relatively more explanatory power in databases of rental apartments, whereas rule-based reasoning will have a relative advantage in sales data. We motivate this hypothesis on theoretical grounds, and find empirical support for it by comparing the two statistical models (rule-based and case-based) on two databases (rentals and sales). ∗We wish to thank Don Brown for the conversations that greatly influenced this work. We thank the Student Association of Tel-Aviv University and Professor Juval Portugali for the data. This work was supported by grants from the Sapir Center at Tel-Aviv University as well as ISF grants no. 975/03 and 355/06. 1 Introduction 1.1 The problem Mary wishes to sell her apartment in the city. How should she determine her asking price? She would probably base her decision on the data available to her regarding apartment prices as a function of the apartment’s characteristics, such as location, area, view, and so forth. Mary might wish to assess the value of her apartment based on more sophisticated theories. For instance, she might start with the price she paid for her apartment and add an annual appreciation that seems reasonable to her. Or she might try to predict market trends and …gure out how much the apartment should be worth. But such theories should also be consistent with the data regarding the sales of other apartments. Indeed, considering price as a function of characteristics directly might be viewed as a reduced form model that Mary can use as a proxy for more involved theories. How would Mary go about assessing the value of her apartment given a database of the prices of other apartments? Casual observation suggests that two modes of reasoning are very common in generating such assessments. The …rst relies on general rules, such as, “In this area, the price per squared meter is $3,000”. The second is case-based, as in the argument, “The apartment next door, practically identical to mine, was just sold for $300,000”. Indeed, in the US the standard assessment procedure involves two assessments, one that is rule-based and another that is case-based. It seems safe to assume that, for the most part, both types of reasoning are present when a person attempts to assess the market price of a realestate asset. The question we wish to address is whether one can make any qualitative predictions regarding the relative importance of rule-based versus case-based reasoning. In particular, what economic considerations might a¤ect the type of reasoning that agents tend to engage in?1 1The precise distinction between rule-based and case-based reasoning is an interesting 1 Gayer et al.: Rule-Based and Case-Based Reasoning in Housing Prices Published by The Berkeley Electronic Press, 2007