Unfortunately, I am not very familiar with panel data methods (yet) but it seems important to check if my data display stationarity (aka time trends). Assuming that my regression function looks like this:

$π‘Œ_{𝑖𝑑} = \alpha + X_{𝑖𝑑} + W_{𝑖𝑑} + \varepsilon_{𝑖𝑑}$

$i$ = country,
$t =$ year,
$Y =$ dependent variable,
$X =$ independent variable of interest,
$W =$ control variables,
$\varepsilon =$ error term

How should I check stationary? Only for $Y$? Or also for each individual regressor? By using the global mean of $Y$ for each year (i.e. the annual average $Y$ over all countries in the data set to see if there is a global trend) or should I been looking for a countrywide time trend by plotting the time evolution of $Y$ for each country individually?

I really have no idea what the correct approach is.



Your Answer

By clicking β€œPost Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.