How to plot a figure with treatment and nontreatment after DID regression.

[Chenli You]

Dear Statalists,

I am trying to plot a figure after a difference-in-difference regression. The DID regression is as below.

reg dependent i.year treatment treatment*post var1 var2

After the regression, I want to plot a figure with two lines.
Line 1 = mean(dependent) - _b[var1]*mean(var1) - _b[var2]*mean(var2), for treatment group through years.
Line 2 = mean(dependent) - _b[var1]*mean(var1) - _b[var2]*mean(var2), for nontreatment group through years.
These are actually "the mean of dependent variables" minus "the factors impacted by var1 and var2" for both treated and untreated groups through years. So, I tried the code as follows.

egen depedent1=mean(depedent1) - _b[var1]*mean(var1) - _b[var2]*mean(var2), by(year), if treatment==1
egen depedent2=mean(depedent1) - _b[var1]*mean(var1) - _b[var2]*mean(var2), by(year), if treatment==0

Unfortunately, Stata reports "varlist not allowed". May I have your advice on how can I work on this?

Thanks a lot,
Chenli

[Rebecca Water]

Hello Chenli,
when clicking on the #-symbol you can wrap your code in [/CODE] [CODE] and it will be easier to read. Also, example data with -dataex- supports others helping you.

Concerning your regression, I think you should cluster the standard errors on the panel unit id, for instance:

regress variables, vce(cluster unitid)

See more here: https://www.stata.com/manuals13/xtvce_options.pdf

Besides this, you don't need to type "_b". Here is an example that I have from a tutorial that may be useful as a generel hint on how you can directly use regression results.

help predict
regress dep indep
predict dep_hat
//COMPARISON
graph twoway scatter dep indep || line dep_hat indep

[Rebecca Water]

Edit: In general for the egen command, your if statement is in the wrong place, it should be before the comma:
https://www.stata.com/manuals13/degen.pdf

[Chenli You]

Hello Rebecca,

Thank you very much for the help. Your suggestions work very well.

Best wishes,
Chenli