Real estate economics 180 Foldvary

Research project

100 points (10%).

March 24 first draft due.

May 5, final draft due.

Select three residential communities anywhere in the USA.

Find data for at least 20 houses in each community.

The home type (such as single-family residence) should all be the same.

The paper should cite the source of the data and list all the addresses.

You may contract real estate brokers for information, or search in the Internet, such as http://www.zillow.com

Use regression analysis to learn what determines the price of a typical house.

For example, P = L + aD + bT + cA + dG +eX +fL1 +gL2 + hL3

where P is price, L is the land value, D is bedrooms, T is bathrooms, A is age, G is growth rate of the house, X is the property tax, L1 is 1 for community 1, and L2 and L3 are zero for the other communities.

Variables used should include, if available:

1) estimated house price (the dependent variable)

2) number of bedrooms.

3) number of bathrooms

4) total number of rooms

5) square footage of the building

6) square footage of the lot

7) age of the house

8) average rate of growth of the house price during the past year or five years

9) average rate of growth of house prices in the county during the past one or five years.

10) most recent property tax

Use a dummy variable for the location in each of the communities,

thus two locational dummy variables.

You may use Excel for the regression, or an other software.

Determine the value added for each variable, e.g. how much value does a second bathroom add?

How much real estate value is based on the location relative to other communities?

Does the intercept provide a rough measure of the land value?

Note: some of the variables are correlated.

You should therefore avoid multicolinearity in any regression.

Regress the correlated variables together to determine the degree of correlation.

Try using different correlated variables, such as square footage of the lot and of the building to see the differences in significance.

Use both statistical and economic significance, the latter referring to the size of the coefficient.

Use any techniques you know that can enhance the results.