Comparing growth rates within a panel

Question

I am conducting a POLS with a panel dataset of 27 countries across the time span 1950-2010 using Stata.

My main focus is on how age structure, in particular growth rates of respective population parts (older, younger, working-age, and so on) affect economic growth. In the regression output, I receive a negative sign for the older population growth rate and a positive one for the younger population.

Consequently, I want to compare which is likely to have a stronger effect on economic growth and therefore I essentially want to compare which grows faster.

How can I determine which growth rate is faster using Stata? Or do you have any other suggestions for comparing these two rates? Or is it even sufficient to use the p-value as indicator?

Answer 1

A $p$-value does not measure the size of an effect, so you should not compare those if you want to know which effect is stronger.

Since the two explanatory/independent/right-hand-side/x-variables you want to compare have the same unit (they are both growth rates) you can compare the coefficients and see which one is bigger.

You could also look for a single country at the actual change in the younger population and the resulting change in economic growth and compare that with the actual change in the older population and the resulting change in economic growth. Do that for a limited number of "example" countries: say a country with a large increase in the younger population but little change in the older population, a country with increasing older population and decreasing younger population, ... Discuss the differences as having two components: differences in effects and differences in changes in population structure.