IV or GMM?

[Marcos Gonzalez]

I have a panel data set of 23 countries and 19 years , that provide an unbalanced panel data of 437 obs. My question is, which methodlogy is more appropiated for dealing with endogeneity? IV or GMM according to this sample?


Thanks