Estimating House Price Growth with Repeat Sales Data: What's the Aim of the Game?

Abstract Since the seminal work of M. Bailey, R. Muth, and H. Nourse (1963,J. Amer. Statist. Assoc.58,933–942), numerous articles have been written about repeat sales and other methods for constructing house price indices. Our justification for producing yet another paper on this subject is to reemphasize fundamentals. We focus on the basic building blocks—asking questions about what the underlying target is, how repeat sales goes about estimating the target, and when a particular index might be used in practice—rather than on more complex, higher level concerns such as statistical or modeling accuracy. We find that much of the debate over index methodology can be distilled to implicit and largely unrecognized disagreement over the desired target or the intended application. Consequently, we contend that paying greater heed to fundamental questions offers significant rewards to both researchers and practitioners.