Hi all,

I use balanced panel data N=43, T=24 and I want to use Hadri LM unit root test for initial real GDP/capita which represents the initial year of the real GDP/capita. for instance, data from Brazil (real GDP/capita 50 in 1950) and Holland (real GDP/capita 100 in 1950) from 1950 to 2000, I created initial real GDP/capita, equal to 50 for all observations from Brazil, and 100 for all observations from Holland.

When I used Hadri LM unit root test P-valu = 1.000


. xtunitroot hadri initrgdp, robust

Hadri LM test for initrgdp
----------------------------
Ho: All panels are stationary Number of panels = 43
Ha: Some panels contain unit roots Number of periods = 24

Time trend: Not included Asymptotics: T, N -> Infinity
Heteroskedasticity: Robust sequentially
LR variance: (not used)
------------------------------------------------------------------------------
Statistic p-value
------------------------------------------------------------------------------
z -7.3314 1.0000

Can I say that real GDP/capita is stationary at p- value =1.000?

Help, please

Thank you