Hello Stata community!
I'm currently writing my master's thesis on the evolution of the green energy mix following recessions/economic shocks and am very new to Stata. There are some papers that I'm attempting create my baseline specifications on, but quite a few questions arise when replicating their projections and the methods I should apply.
At first I used the https://www.stata.com/manuals/tslpirf.pdf instructions and created the following code:
tsset ifscode year, yearly
lpirf oil_ln electricity_ln primary_energy_ln coal_ln, lags(1/2) step(8) exog(L(0/2).recession)
quietly lpirf oil_ln electricity_ln primary_energy_ln coal_ln, lags(1/2) exog(L(0/2).recession)
irf set myirfs.irf, replace
irf create lpmodel
irf graph dm, impulse(recession) irf(lpmodel)
And it created reasonable projections of the 4 dependant energy consumption variables following a recession shock (dummy variable), somewhat in line with previous research.
However, when I run on just an individual dependant variable, oil, the projections are quite different (positive trend):
tsset ifscode year, yearly
lpirf oil_ln, lags(1/2) step(8) exog(L(0/2).recession)
quietly lpirf, lags(1/2) exog(L(0/2).recession)
irf set myirfs.irf, replace
irf create lpmodel
irf graph dm, impulse(recession) irf(lpmodel)
What impact would the other dependant 3 variables have on the 4th?
Another issue is that my data structure is panel, not time series and I would like to add country fixed effects. Does lpirf allow for these modifications? Would a looped xtreg be an alternative?
Thank you in advance!
Regards,
Rudolfs
I'm currently writing my master's thesis on the evolution of the green energy mix following recessions/economic shocks and am very new to Stata. There are some papers that I'm attempting create my baseline specifications on, but quite a few questions arise when replicating their projections and the methods I should apply.
At first I used the https://www.stata.com/manuals/tslpirf.pdf instructions and created the following code:
tsset ifscode year, yearly
lpirf oil_ln electricity_ln primary_energy_ln coal_ln, lags(1/2) step(8) exog(L(0/2).recession)
quietly lpirf oil_ln electricity_ln primary_energy_ln coal_ln, lags(1/2) exog(L(0/2).recession)
irf set myirfs.irf, replace
irf create lpmodel
irf graph dm, impulse(recession) irf(lpmodel)
And it created reasonable projections of the 4 dependant energy consumption variables following a recession shock (dummy variable), somewhat in line with previous research.
However, when I run on just an individual dependant variable, oil, the projections are quite different (positive trend):
tsset ifscode year, yearly
lpirf oil_ln, lags(1/2) step(8) exog(L(0/2).recession)
quietly lpirf, lags(1/2) exog(L(0/2).recession)
irf set myirfs.irf, replace
irf create lpmodel
irf graph dm, impulse(recession) irf(lpmodel)
What impact would the other dependant 3 variables have on the 4th?
Another issue is that my data structure is panel, not time series and I would like to add country fixed effects. Does lpirf allow for these modifications? Would a looped xtreg be an alternative?
Thank you in advance!
Regards,
Rudolfs
