[Sergio Correia]

Note that you are not using absorb() to partial out the FEs, so in practice you are just using ivreg2.

If what you want is to partial out some variables such as "time" then instead of

ivreghdfe righttotax i.time (i.treatment i.time#i.treatment= i.instrument i.time#i.instrument) [pw = ate_wgt2], vce(cluster district) You would want to do:
ivreghdfe righttotax (i.treatment i.time#i.treatment= i.instrument i.time#i.instrument) [pw = ate_wgt2], vce(cluster district) absorb(time) However, not sure if this is what you want

[Wei LIIU]

Note that you are not using absorb() to partial out the FEs, so in practice you are just using ivreg2.

If what you want is to partial out some variables such as "time" then instead of

ivreghdfe righttotax i.time (i.treatment i.time#i.treatment= i.instrument i.time#i.instrument) [pw = ate_wgt2], vce(cluster district) You would want to do:
ivreghdfe righttotax (i.treatment i.time#i.treatment= i.instrument i.time#i.instrument) [pw = ate_wgt2], vce(cluster district) absorb(time) However, not sure if this is what you want

Hi dear, my estimation model looks like following:

Yijkt = β0 + β1 ∗ treat_j + β2 ∗ post_t + β3 ∗ treat_j ∗ post_t + β4X_it + ξ_i + δ_t + ψ_k + ϵ_ijkt
where treat equals 1 if in the treatment group, post equals 1 if the year is equal to or greater than 2012 (policy takes effect in 2012), and X_it are city level time varying variables. ξ_i is the city fixed effect, δ_t is the time fixed effect, and ψ_k is the product fixed effect.
The following is my regression command:
reghdfe sales_spec i.treat_1##i.post lnp_adj lngdp lnage nicotine tar CO , absorb(i.code i.year i.product) vce(cl province)
where code represents the city fixed effect, year represents the time fixed effect, and product represents the product fixed effect.the regression result looks like as following:

reghdfe sales_spec i.treat_1##i.post $X1list, absorb(i.code i.year i.product) vce(cl province)

(MWFE estimator converged in 10 iterations)

note: 1bn.treat_1 is probably collinear with the fixed effects (all partialled-out values are close to

> zero; tol = 1.0e-09)

note: 1bn.post is probably collinear with the fixed effects (all partialled-out values are close to zer

> o; tol = 1.0e-09)

note: post is probably collinear with the fixed effects (all partialled-out values are close to zero; t

> ol = 1.0e-09)

note: nicotine is probably collinear with the fixed effects (all partialled-out values are close to zer

> o; tol = 1.0e-09)

note: tar is probably collinear with the fixed effects (all partialled-out values are close to zero; to

> l = 1.0e-09)

note: CO is probably collinear with the fixed effects (all partialled-out values are close to zero; tol

> = 1.0e-09)

HDFE Linear regression Number of obs = 104,585

Absorbing 3 HDFE groups F( 4, 25) = 12.29

Statistics robust to heteroskedasticity Prob > F = 0.0000

R-squared = 0.3445

Adj R-squared = 0.3399

Within R-sq. = 0.0013

Number of clusters (province) = 26 Root MSE = 12.9509

(Std. err. adjusted for 26 clusters in province)

Robust

sales_spec Coefficient std. err. t P>t [95% conf. interval]

1.treat_1 0 (omitted)

1.post 0 (omitted)

treat_1#post

1 1 1.205053 .2614475 4.61 0.000 .6665916 1.743514

post 0 (omitted)

lnp_adj -14.86863 3.158481 -4.71 0.000 -21.37364 -8.363616

lngdp .5909204 .3428594 1.72 0.097 -.1152117 1.297053

lnage -.0493869 .9642361 -0.05 0.960 -2.035268 1.936495

nicotine 0 (omitted)

tar 0 (omitted)

CO 0 (omitted)

_cons 73.1776 12.36098 5.92 0.000 47.71968 98.63552

Absorbed degrees of freedom:

Absorbed FE Categories - Redundant = Num. Coefs

-

code 268 268 0 *

year 8 1 7

product 449 1 448

* = FE nested within cluster; treated as redundant for DoF computation

my confusion is how to get the coefficients of treat and post when using reghdfe as well as absorb these fixed effects. Theoretically, treat and post be absorded by fixed effect and thus cannot be estimated. Is there any empirical method to get the coefficients of the two variables?

[Clyde Schechter]

#17 is a duplicate post, asked and answered at https://www.statalist.org/forums/for...e-fixed-effect.