Hello guys,

I have an urgent question regarding marginal effect anaylsis:


about my model specification: (dependant variable is log of real GDP per capita (continous), corruption and democracy indp. variable are also continous

LogRGDPPCit = γ0 + γ1 CORRit + γ2 DEMOit + γ3 CORRit × DEMOit + γ4 (CAP/Pop) + γ5 EDUit + γ6 EconFree+γ7 MoneySupplyit+ γ8 RNDit + vi

I am interested in the marginal effect of corruption on growth and whether that depends on the level of democracy and how( so the statisticall significance ) and sign of γ3.
..
∂LogRGDPPCit / ∂CORRit = γ1 + γ3 DEMOit


I have estimated the parameters with Pooled OLS and Fixed effects and SYS GMM (There teh model specificatio is sltighly different) and found for example these results. So for the fixed effects restults i got:

Marginal effect of corruption on growth: ∂LogRGDPPCit / ∂CORRit = γ1 + γ3 DEMOit = -0.258+ 0.0367*3.41= -0.132853.

Is it correct to write that increasing the corruption score by 1, at the average level of democracy (which is 3.418), decreases real GDP per capita by 13.28%? And that therefore increasing the corruption score by one standard deviation (1.399) would decrease real GDP per capita by 18.57% at the average level of democracy?

thanks