Table 2.

Panel ARDL results (Misalignment_30)

Pooled mean groupMean groupDynamic fixed effects
Coef.Std. errorCoef.Std. errorCoef.Std. error
Long-run coefficients      
Mis_30−1.8741***0.4886−5.3945***2.6861−1.1599*0.6319
Pop0.4923***0.09190.32501.67060.4366***0.1202
HK0.4432***0.08050.79890.64270.08180.0635
INF−0.0897***0.0120−0.2395***0.0666−0.0244***0.0048
OPEN0.0124***0.00510.0898***0.0244−0.00270.0094
UN−0.0588**0.0256−1.4520***0.6168−0.01100.0403
Error-correction coefficient      
ECT (−1)−0.8756***0.0311−1.1504***0.0310−0.8461***0.0342
ΔMis300.02922.08033.30102.75022.6291**1.2614
ΔPop0.35301.13991.38641.9228−0.1352*0.0754
ΔHK0.52200.3854−0.18290.64170.4792***0.1042
ΔINF0.02990.02750.1340***0.04730.00150.0022
ΔOPEN0.1178***0.01870.0587**0.02580.0562***0.0120
ΔUN−1.7002***0.2837−0.31090.5998−1.3283***0.0972
Intercept−23.0660***1.0435−51.055843.8662−1.00283.6662
BIC9,783.899,856.749,976.52
AIC9,696.149,816.019,880.12
Observations25,662,5662,566
Hausman test 5.99(i)
(0.4238)
8.36(ii)
(0.5596)
Note:

*, ** and ***indicate significance at 10, 5 and 1%, respectively;

aUnder the null hypothesis, PMG is more efficient estimation than MG;

bPMG is more efficient estimation than DFE under the null hypothesis. In brackets are the associated p-values

Source: Authors’ own elaboration

or Create an Account

Close Modal
Close Modal