Geoscience Reference
In-Depth Information
X
f
j
¼
1
YFWOR
j
ER
X
cm
k
KAPWOR
¼
1
PWM
k
QM
cm
þ
¼
X
ce
i
¼
1
PWE
i
ð
10
21
Þ
:
QE
i
hwor
govwor
X
f
m
1
factwor
m
¼
TOTSAV
¼
INVEST
þ
WALRAS
ð
10
:
22
Þ
where
FS
f
: Supply of factor f;
QQ
c
: Supply of composite commodity c;
YFWOR
f
: Foreign factor income;
factwor
m
: Factor payments from ROW (constant in foreign currency);
INVEST: Total investment expenditure;
WALRAS: Slack variable for Walras's Law.
10.5
Estimation Procedure
The estimation procedure of the study is carried out sequentially in the following
four steps.
10.5.1 Step 1 Spatial Autocorrelation Test
The spatial autocorrelation, which is measured by values of Moran's I, is tested for
in the capital-labor ratio variable and in wage-rental ratio variable using
GeoDa
,
developed by the Spatial Analysis Laboratory at the University of Illinois at
Urbana-Champaign (Anselin et al.
2006
). The universal global Moran's I is defined
as (Moran
1950
; Cliff and Ord
1981
):
X
n
i¼
1
w
ij
x
i
x
ð
Þ
x
j
x
n
X
n
i¼
1
X
n
X
n
i¼
1
w
ij
x
i
x
I ¼
ð
10
23
Þ
:
2
j¼
1
w
ij
ð
Þ
where n is the number of regions which includes 48 contiguous states and the
District of Columbia for most sectors except the sectors of pipeline and wate
r
transportation, which only contain 48 regions and 36 regions respectively.
x
and
x
denote the specific state and the mean of
x
respectively.
w
ij
is the spatial weight
matrix, representing the spatial relationship between region
i
and
j
. The spatial
relationship in this study is defined as being contiguous to each other. Thus the
spatial weight matrix is generated using the Queen Contiguity method.
Because Moran's I can only be tested on a yearly basis, Moran's I for each year
from 1997 to 2011 is calculated. The results are similar for each variable in each