matchdata: Matched samples of house sales in Chicago for 1995 and 2005
In McSpatial: Nonparametric spatial data analysis

Description Usage Format Details Source References Examples

Sales prices, structural characteristics, and location variables for 1602 single-family homes in the City of Chicago in 1995 and a matched sample of 1602 homes in 2005.

1	data(matchdata)

A data frame with 3204 observations on the following 18 variables.

year: Year of sale, 1995 or 2005
lnland: Log of land area in square feet
lnbldg: Log of building area in square feet
rooms: Number of rooms
bedrooms: Number of bedrooms
bathrooms: Number of bathrooms
centair: Home has central air conditioning
fireplace: Home has one or more fireplaces
brick: Brick or brick/frame construction
garage1: Garage, 1 car
garage2: Garage, 2+ cars
dcbd: Distance from the central business district or "CBD" - the traditional center of Chicago at the intersection of State and Madison Streets, at approximately -87.627800 longitude and 41.881998 latitude
rr: Within .25 miles of a rail line
yrbuilt: Year the home was built
carea: a factor with levels. Community area, a traditional definition of neighborhood in Chicago.
latitude: Latitude in degrees
longitude: Longitude in degrees
lnprice: Log of sales price

Includes all sales of single-family homes on the Far North Side of Chicago listed in the cleaned Illinois Department of Revenue file for 1995. A matched sample is created from comparable 2005 sales using the MatchIt package. Matches are created based on propensity scores estimated using a logit model for the probability that a home sold in 2005 rather than 1995. The commands used to create the matched sample are the following:

hedonic$carea <- as.factor(hedonic$cname)
m.out <- matchit(y~lnland + lnbldg + rooms + bedrooms + bathrooms + centair + fireplace + brick + garage1 + garage2 + dcbd + elstop + lake + rr + yrbuilt + carea + latitude + longitude, data=hedonic,method="nearest",discard="both")
mdata <- match.data(m.out)
attach(mdata)
matchdata <- data.frame(year, lnland, lnbldg, rooms, bedrooms, bathrooms, centair, fireplace, brick, garage1, garage2, dcbd, rr, yrbuilt, carea, latitude, longitude, lnprice)

The elstop and lake variables, which are not included here, indicate whether a home is within .25 miles of and EL stop and within .5 miles of Lake Michigan.

Daniel McMillen. Sales data were provided originally by the Illinois Department of Revenue. Structural characteristics are drawn from the 1997 assessment file from the Cook County Assessor's Office.

Deng, Yongheng, Sing Tien Foo, and Daniel P. McMillen, "Private Residential Price Indices in Singapore," Regional Science and Urban Economics, 42 (2012), 485-494.

Ho, D., Imai, K., King, G, Stuart, E., "Matching as Nonparametric Preprocessing for Reducing Model Dependence in Parametric Causal Inference," Political Analysis 15 (2007), 199-236.

Ho, D., Imai, K., King, G, Stuart, E., "MatchIt: Nonparametric preprocessing for parametric causal inference," Journal of Statistical Software 42 (2011), 1-28..

McMillen, Daniel P., "Repeat Sales as a Matching Estimator," Real Estate Economics 40 (2012), 743-771.

data(matchdata)
matchdata$year05 <- matchdata$year==2005
matchdata$age <- matchdata$year - matchdata$yrbuilt 
fit <- lm(lnprice~lnland+lnbldg+rooms+bedrooms+bathrooms+centair+fireplace+brick+
  garage1+garage2+dcbd+rr+age+year05+factor(carea), data=matchdata)
summary(fit)