Handling Time Invariant Covariates

[dhammo2]

When using the gsynthpackage it can handle time-invariant covariates but it does not allow treatment reversal.

Is there a way to run analysis in fect (which allows treatment reversal) without having issues with time invariant errors on socio-economic graphic covariates that are not changing?

[xuyiqing]

Thank you for your suggestion. We'll address this in the next update.