Difference in differences Method

[Neeraj Kumar]

Hello
I am working on unbalanced panel data for 2 year , in 1st year observation are 26734 and second year it's 27579, after doing an one to one merging total number of observation are 53549. In which I am running diff command for difference in differences. Dependent variable is shareCO1X and treatment variable is mgnrega1 and time variable is time ( i named my variable as time).
I am facing problem of treatment observations under Before column is coming 0 as it can be observed from table. I want to know why it is coming like that or where am i doing a mistake please guide me.

[Carlo Lazzaro]

Neeraj:
the usual advice is to -codebook- your data to see what went wrong.

[Neeraj Kumar]

Thanks Carlo for your prompt reply, here I am sending you the codebook results for those variables that I used in the command of diff. I don't think so anything is wrong with data.

[Carlo Lazzaro]

Neeraj:
my guess is that before treatment sets in, all patients are controls (and so you have 0 treated patients).

[Neeraj Kumar]

Yes I don't have treated patients in before treatment then how to generate treated patients in before treatment.

[Carlo Lazzaro]

Neeraj:
why don't you try:

xtset id time
xtreg shareCO1X i.treated##i.time, fe

[Neeraj Kumar]

Thanks for your suggestions. I got the mistake and trying to solve it because the problem is in my data.

[Abou Sarambe]

Dear all,
I would like to know how to get confidence intervals in the output of diff command.
Kindly,

Abou Sarambé