1. ECON734: Spatial Econometrics – Lab 2
Term I,
Yang Zhenlin

2. Boston House Price The Boston house-price data of
Harrison, D., Rubinfeld, D. L. (1978): Hedonic prices and the demand for clean air, J. Environ. Economics & Management, Vol. 5,
Used in Belsley, Kuh & Welsch: “Regression diagnostic ...”, Wiley, Various transformations are used in the table on pages
Augmented with latitude-longitude coordinates by:
Gilley, O.W., and R. Kelley Pace. (1996). On the Harrison and Rubinfeld Data. Journal of Environmental Economics and Management, Vol. 31,
It contains 506 census tracts.
See boston_readme.txt for more details on the data. The file boston.dat contains the augmented data, and the file boston.xy generates the spatial with matrix based on the latitude-longitude coordinates.

3. Boston House Price
Variables (in columns) order in the augmented data file boston.dat:
CRIM per capita crime rate by town
ZN proportion of residential land zoned for lots over 25,000 sq.ft.
INDUS proportion of non-retail business acres per town
CHAS Charles River dummy variable (= 1 if tract bounds river; 0 o.w.)
NOX nitric oxides concentration (parts per 10 million)
RM average number of rooms per dwelling
AGE proportion of owner-occupied units built prior to 1940
DIS weighted distances to five Boston employment centres
RAD index of accessibility to radial highways
TAX full-value property-tax rate per $10,000
PTRATIO pupil-teacher ratio by town
B (Bk )^2 where Bk is the proportion of blacks by town
LSTAT % lower status of the population
MEDV Median value of owner-occupied homes in $1000's
latt Lattitude coordinates
long Longitude coordinates

4. Boston House Price The model: MEDV = 0 + 1Crime + 2ZN + . . . +
The spatial weight matrix is constructed using the Euclidean distance in terms of longitude and latitude by boston.xy.
A threshold distance, e.g., 0.05, is chosen, which gives a Wn matrix with 19.08% non-zero elements.
The model: MEDV = 0 + 1Crime + 2ZN
13Lstat + error + spatial terms
where the spatial terms may contain SED, SLD, S-Durbin, SED+SLD, S-Durbin+SED, SLD+S-Durbin, etc.

5. Fitting house price data with SED model
The key Matlab commands in the main program Boston_Qmle_SED.m:
load boston.dat; % Harrison-Rubinfeld data
[n p] = size(boston); % no. of rows and columns in data
y = boston(:,p-2); % median house values
latt = boston(:,p-1); % lattitude coordinates
long = boston(:,p); % longitude coordinates
x = [ones(n,1) boston(:,1:p-3)];
vnames = char('hprice','constant','crime','zoning','industry', ...
'charlesr','noxsq','rooms2','houseage','distance','access','taxrate', ...
'pupil/teacher','blackpop','lowclass');
beth0 = pinv(x*’x)*x’*y; % OLS regression of y on x
rhoh = SED_ML(y,x,W); % returns the QMLE of rho

6. Fitting house price data with SED model
The generic m-function SED_ML.m, called by the main program Boston_Qmle_SED.m, produces the following results.
(Q)MLE se_MLE t_MLE se_QMLE t_QMLE
Const
CRIM ZN
Indus
Chas
NOX RM
Age
DIS
RAD
TAX
PTRATIO B
Lstat
SED 