To determine if “micro
markets” exist in the real estate
market, I must design and implement an experiment that compares price
performance over time for several different property types. The first step of
this is to determine how I will track prices for each property type I am
comparing. This will be done through a price index. For years, price indices
have been used to track the market performance and real estate prices. There
are two main methods of creating a price index.
The Hedonic Model
The hedonic model is a method of real estate valuation that
attempts to account for heterogeneity by appraising a property per the value of
the various attributes/characteristics. To do this, it considers a house simply
as a collection of multiple price-influencing characteristics. The house value
can then be determined by the summed value of all characteristics.
Despite being commonly used, the hedonic model has many
criticisms. To start, a house can be defined by infinite characteristics. Thus,
it can be difficult to determine which characteristics to consider and to
accurately value them. Inaccurate regression techniques can also result in
overestimation or underestimation of a characteristic’s value, making
insignificant variables become significant and vice versa.
The Repeat Sales Method
Unlike the hedonic model, the repeat sales method only uses
properties with multiple sales to track the market. Heterogeneity is controlled
for by looking at changes in prices for the same houses over time. This
eliminates the need for data on multiple variables and complicated regression
techniques, making it easier than the hedonic model.
A criticism of the repeat sales method is the limited
sample size. Since only properties with multiple sales are being considered, it
is important to consider the accuracy of this process. Multiple researchers
have criticized it for ignoring too much data and creating a large amount of
sample selection bias.
So, what process will I be using for my study? I have
decided to use the repeat sales method for two reasons. First, this method is
much easier than the hedonic model. The hedonic model requires extensive
knowledge of regression techniques as well as detailed data on a multitude of
house characteristics. With the time and knowledge I currently have, using this
model would be impractical. Rather, the repeat sales method requires less data and
less regression techniques.
Secondly, the repeat sales method is comparable to the
hedonic model in regards to its accuracy in portraying price dynamics. While
both models have their difficulties and criticisms, they are still used today
and supported by researchers around the world. Moreover, the differences in
accuracy between the two is more in the type of bias they cause, not the
amount. Recognizing this, I have decided to use the repeat sales method for my
study.
Over the next few days, I will be collecting sales data and begin organizing it
to analyze. Sometime this week, I will have another post with specific details
of this data and my completed methods.