it's my first time using xtabond, so I would like to run this by folks who are familiar with panel data.
Per advice from Phil in my previous inquiry, I am using xtabond for panel data of U.S. states (51) (N) and eight years from 2010 to 2017 (T). So I don't have a big sample size.
For xtabond, my question is whether to and how to incorporate time dummy or time progress variable. It reads to me like I really don't need to include time variable in the model, like I did below.
Code:
depvar IV1 IV2 IV3 IV4 IV5, vce(robust)
Code:
depvar IV1 IV2 IV3 IV4 IV5 YR11 - YR17, vce(robust)
Code:
depvar IV1 IV2 IV3 IV4 IV5 wave, vce(robust)
If anyone, please help me with 1) whether I should include time variable, 2) if I should, whether it should be year dummies or wave, and 3) if it's not necessary, what should be my consideration in making the decision?
Thanks
0 Response to xtabond time dummies
Post a Comment