How to check parallel trends assumption of the variables are all first difference of logarithms
Good afternoon,
I'm doing my thesis on the CAPs (agricultural eu policy) on TFP.
I'm doing a DID model with eu15 countries and USA, Australia and Canada as control group. I then have different dummies for the countries and years they enter the eu and therefore the policy.
I think this is considered a staggered DID. (these dummies basically overlap each other once the country starts being part of the treated group).
However my dependent variable, TFP, is the rate of change of productivity, all the variables in the model are also on ROC. this means that I can't graph the 2 groups to check what was going on before 1992, I basically get white noise since is the variation rate. Should I compare them before doing the variation? Can I compare the mean and variance of the ROC?
Sorry to bother and thanks!
I'm doing my thesis on the CAPs (agricultural eu policy) on TFP.
I'm doing a DID model with eu15 countries and USA, Australia and Canada as control group. I then have different dummies for the countries and years they enter the eu and therefore the policy.
I think this is considered a staggered DID. (these dummies basically overlap each other once the country starts being part of the treated group).
However my dependent variable, TFP, is the rate of change of productivity, all the variables in the model are also on ROC. this means that I can't graph the 2 groups to check what was going on before 1992, I basically get white noise since is the variation rate. Should I compare them before doing the variation? Can I compare the mean and variance of the ROC?
Sorry to bother and thanks!
