Hi all,
I'm currently doing some research using a Panel VAR model and am following the Love & Abrigo (2016) using the PVAR command. My first question surrounds clustering, I'm using panels of local authority data (my country split geographically into groups of around 100k people) for three endogenous variables. I feel like I should be using clustering rather than the typical robust option due to panels being regional, but I'm not sure what type of clustering to use, the command for it and the intuition behind that specific clustering of the errors.
Secondly, I've used the normal tests for serial correlation, heteroskedasticity, omitted variable bias and multicollinearity in fixed and random effects models before, however I'm unsure of whether to use the same commands after PVAR. If not, what commands can be used to test these things (especially the first two) after PVAR as I couldn't find anything in the Love & Abrigo (2016) paper?
Thirdly, if anyone has any papers that contain the Stata code for PVAR models and its robustness checks, this would be of great help too.
Any help would be greatly appreciated,
Thank you
I'm currently doing some research using a Panel VAR model and am following the Love & Abrigo (2016) using the PVAR command. My first question surrounds clustering, I'm using panels of local authority data (my country split geographically into groups of around 100k people) for three endogenous variables. I feel like I should be using clustering rather than the typical robust option due to panels being regional, but I'm not sure what type of clustering to use, the command for it and the intuition behind that specific clustering of the errors.
Secondly, I've used the normal tests for serial correlation, heteroskedasticity, omitted variable bias and multicollinearity in fixed and random effects models before, however I'm unsure of whether to use the same commands after PVAR. If not, what commands can be used to test these things (especially the first two) after PVAR as I couldn't find anything in the Love & Abrigo (2016) paper?
Thirdly, if anyone has any papers that contain the Stata code for PVAR models and its robustness checks, this would be of great help too.
Any help would be greatly appreciated,
Thank you
Comment