Hi all, I would like to know if I can use ppmlhdfe and use my coefficients as given.
I am estimating the following model
As shown i include different levels of fixed effects.
My model returns the following output. I show only my variable of interest which is covid( where 1.covid is the onset of the pandemic).
In this case do I need margins or I can use the coefficients as it is. Secondly if I can use the coefficients as is, do I then calculate the semi elasticities by hand such that 100*(exp(-0.011) - 1)% = -1.1%.
I am estimating the following model
Code:
ppmlhdfe total_cpi $demog i.covid , absorb(month year household_id) vce(robust) v(-1)
My model returns the following output. I show only my variable of interest which is covid( where 1.covid is the onset of the pandemic).
Code:
HDFE PPML regression No. of obs = 1134816
Absorbing 3 HDFE groups Residual df = 1087519
Wald chi2(1) = 0.20
Deviance = 41630931.95 Prob > chi2 = 0.6516
Log pseudolikelihood = -21908385.05 Pseudo R2 = 0.4638
------------------------------------------------------------------------------
| Robust
total_cpi | Coefficient std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
1.covid | -.0108314 .0239846 -0.45 0.652 -.0578404 .0361775
_cons | 3.825952 .0117846 324.66 0.000 3.802855 3.84905
------------------------------------------------------------------------------
In this case do I need margins or I can use the coefficients as it is. Secondly if I can use the coefficients as is, do I then calculate the semi elasticities by hand such that 100*(exp(-0.011) - 1)% = -1.1%.
Comment