I am working on a panel data project for university and am running a growth regression on aid, population, government expenditure (% of GDP) and a few other control variables. I am investigating the effect of aid on growth. I am using a panel of all Sub Saharan african countries over a 30 year period.
Part of this is that i need to do panel unit root tests to test whether the variables are stationary or not. I don't think i need to do this on any variabels that are measured as a percentage of GDP do i?
Anyway, I have successfully imported all the data into eViews (good start, i know), and am stuck at the first hurdle. I double clicked my population series, went view - unit root tests in levels, and got the following output:
So i have p values of 1? Does that mean that the series is 100% non-stationary? (i expect it to be non stationary of course).
Also, I know that when I run my standard regression to see whether aid is statistically significant in determining growth, it will be impossible to conclude the direction of causality (whether aid causes growth or growth causes aid). So, I was thinking of including a lagged dependant variable (growth) in the regression, to control for unobservables that affect growth and may affect the about of aid received. Is that right? Thank you for your help.