Hi guys,
I was wondering if it is possible to test for unit root on a T=2 panel dataset. I only have 2 t's and so all errors say insufficient observations. If I can not test for a unit root, can I make any assumptions given my t=2?
The context of this is i am measuring effects of RD intensity on sales growth. One of my explanatory variables is lagged sales . For all the regressions I run lagged sales has a negative correlation with sales growth. Can I make any implications given this result?
Thank you

MY N=500 T=2