Does panel unit-root tests for heteroskedastic panels, like
perform any better than the official panel unit-root tests provided by started? I have tried using the
-- Panel unit-root tests for heteroskedastic panels command, but it is giving me very different results compared to the official panel unit-root tests provided by Stata.
Even when the variables are first differenced it fail to reject the null for panels contain unit roots. This is in sharp contrast to other common panel unit-root tests.
Code:
xtpurt varname
Code:
xtpurt
Even when the variables are first differenced it fail to reject the null for panels contain unit roots. This is in sharp contrast to other common panel unit-root tests.
