Power law distributions have previously been observed in data like income distribution, financial asset prices and city size distribution. In this paper we build on the limited research on power laws and real estate prices by exploring the distribution of real estate prices in Charleston County, South Carolina.
Data/Methods:
Using a data set with over 350,000 property value observations, we fit a power law distribution to the data and observe how the estimated power law exponent changes over time. We also compare the goodness of fit of the power law distribution to a lognormal distribution.
Expected Results:
We hope to show that a relationship exists between the estimated exponent on the power law distribution and the presence of price bubbles in the real estate market. Previous work has shown that property values follow a power law distribution at high prices (in the tail of the distribution). We expect to find that a lognormal distribution fits the entire data set well, but that a power law distribution fits the data at high prices better than a lognormal distribution.