American Housing Survey II
Data Model drives variable selection Selection Criteria Certain MSAs Owners only Known metro status Known year built 103 variables on 21720 observations
Data Download Go to http://research.umbc.edu/~coates/ec611/owner_housing.zip Unzip the file using Winzip Read it into Excel Parse it into columns Save as tab delimited text
Codebook Http://research.umbc.edu/~coates/ec611/owner_housing_codebook.html
Stata Read into Stata insheet using “your file name” Be sure you have 103 variables, 21720 observations
Examine the Data Sort control famtyp famrel Summarize data note multiple observations per control what are famtyp and famrel? Summarize data several variables are strings note minimums and maximums
Housing Characteristics Goal: estimate a hedonic equation for house characteristics Consider Bedrms, Baths, halfb, kitch, dens, dining, famrm, laundry, living, rooms, unitsf, lot What do you observe? Look at maximums and minimums
Housing characteristics Functional form? Does each bedroom contribute equally to value? Does each bathroom? What about area? Price or log price? Construction cost or purchase price?
Regression OLS - lprice as function of house characteristics, unitsf and lot tabulate characteristics tabu varname, gen(varname) Sample weights [pw=pwt]