[Retracted] Foreign Direct Investment, Regional Innovation, and Green Economic Efficiency: An Empirical Test Based on the Investigation of Intermediary Effect and Threshold Effect
Table 7
Empirical test of the threshold effect.
Variables
1
2
3
Threshold variable
LnGDPit < 8.748
8.748 ≤ LnGDPit ≤ 9.373
LnGDPit > 9.373
LnFDIit
LnGDPit < 8.748
8.748 ≤ LnGDPit ≤ 9.373
LnGDPit > 9.373
−0.0554
0.0347
0.0628
LnGECMit−1
[−2.97]
[0.60]
[5.67]
0.3135
0.1832
0.3468
Control variable
[2.80]
[0.27]
[7.54]
Individual effects
Control
Control
Control
Sargan
Yes
Yes
Yes
19.7496
1.1150
16.1430
AR(2)
(0.474)
(1.000)
(0.933)
—
1.2718
0.0453
Wald
—
(0.203)
(0.964)
201.6500
59.0700
2058.8500
When the LnGDPit < 8.748, since the two-step estimator variance-covariance matrix is not full rank, the one-step estimation is taken, and the results are still robust and without the AR(2) test.
