Table IV.

Results of first-generation panel unit root tests

TestTest-statistic
LevelsFirst differences
Null hypothesis: unit root
Levin–Lin–Chu (LLC) test−8.210* (0.000)−22.949* (0.000)
Breitung test (BT)−5.329* (0.000)−11.241* (0.000)
Im–Pesaran–Shin (IPS) test−5.453* (0.000)−21.713* (0.000)
ADF-Fisher x2 (MW) test84.901* (0.000)384.494* (0.000)
Phillips–Perron–Fisher x2 (PPF)74.314* (0.000)2,148.260* (0.000)
Null hypothesis: no unit root
Common unit process8.271*21.041*
Hadri test (HT)(0.000)(0.000)

Notes:

*

Indicates statistical significance at the 1% level. For levels, deterministic terms included are individual effects and linear trends. Lags are based on the Schwartz information criteria (SIC) criterion. In Hadri’s test, Heteroscedasticity-consistent Z-statistics are reported

or Create an Account

Close Modal
Close Modal