18
Table 6.2: Okun’s law: estimation and testing results
|
Non-spatial GLS estimation |
Spatial SUR estimation | |||
|
Coefficient |
t-statistic |
Coefficient |
t-statistic | |
|
Output growth |
-0.0510 |
-11.336 |
-0.0453 |
-9.032 |
|
Constant |
0.0020 |
11.036 |
0.0010 |
7.781 |
|
D93 |
0.0996 |
16.862 |
0.0157 |
29.453 |
|
D00 |
-0.0090 |
-26.983 |
-0.0055 |
-18.740 |
|
Unempl. threshold |
0.0401 |
10.946 |
0.0219 |
6.125 |
|
R2 SER |
SSR |
R2 SER |
SSR | |
|
0.305 0.010 |
0.133 |
0.524 0.008 |
0.091 | |
|
Non-spatial GLS |
Spatial SUR estimation | |||
|
Moran's I residuals |
Moran's I residuals |
Significant eigen- |
Moran's I spatial | |
|
1993 |
0.180 (4,043) |
0.002 (1,278) |
2, 4, 7, 12, 14, 16, 21, 22, 27, 33, 36, 37 |
0.702 |
|
1994 |
0.518 (11,199) |
0.043 (2,201) |
1, 2, 3, 5, 6, 7, 10, 12, 14 |
0.965 |
|
1995 |
0.375 (8,081) |
0.081 (2,851) |
1, 2, 3, 4, 7, 10, 13, 31 |
0.938 |
|
1996 |
0.321 (6,931) |
0.020 (1,374) |
1, 2, 3, 9, 10, 16, 17 |
0.949 |
|
1997 |
0.693 (14,756) |
0.061 (3,230) |
1, 2, 3, 4, 6, 7, 14, 21, 25, 29, 31, 32, 33, 34, |
0.891 |
|
1998 |
0.402 (8,645) |
0.025 (1,237) |
1, 2, 6, 9, 10, 23, 25, |
0.802 |
|
1999 |
0.236 (5,114) |
0.026 (1,878) |
6, 8, 10, 13, 14, 16, 17, 26, 27, 28, 33 |
0.721 |
|
2000 |
0.559 (12,058) |
0.024 (2,136) |
1, 2, 3, 4, 5, 7, 17, 19, |
0.939 |
Notes: Non-spatial GSL estimation is based on the SUR model (3.4) without applying the spatial filtering
approach. .D93, D00: Dummies for 1993 and 2000, R2: coefficient of determination, SER: stan-
dard error of regression, SSR: sum of squared residuals
Table 6.2 reports the results of the FD specification of Okun’s law. The law is supposed
to hold over the entire sample period. However, although all coefficients are significant
with the expected signs, their magnitude appear to be very low. Due to persistence ef-
fects, unemployment does not react very much in response to a change in the growth
rate.