Is it ok to have log transformation in difference-in-differences regression?

Is it ok to have log transformation in difference-in-differences model? It seems like we cannot as it violates the additivity assumption. Am I correct?

03 March 2015 2,212 3 View