Re: st: DiD

[inggrid <inggrid_ep@yahoo.com>]

Hi Federico,

Thank you for your response.

You are right, I have a panel data set between 2000 and 2007. I would like to assess the impact of a program introduced in 2005. I have information on treated households. It is a dummy variable (variable name: program), taking a value of 1 if a household covered by the program and zero otherwise (from data in 2007). Hence, I use a difference-in-differences estimator. 
My commands to run the DiD as follows:

gen treat=(program!=0)
label variable  treat "hh covered by the program"
gen post05 = year>2000
gen treat_post05 = treat*post05
regress  y treat post05  treat_post05 [aweight=hhweight],r cluster(ea)
Subsequently, my STATA told that:
(sum of wgt is   5.5336e+03)
note: post05 omitted because of collinearity
note: treat_post05 omitted because of collinearity

Any idea what is wrong?

Best regards,
Inggrid
--- On Tue, 3/29/11, federico.tedeschi@univr.it <federico.tedeschi@univr.it> wrote:

> From: federico.tedeschi@univr.it <federico.tedeschi@univr.it>
> Subject: Re: st: DiD
> To: statalist@hsphsun2.harvard.edu
> Date: Tuesday, March 29, 2011, 3:55 PM
> Dear Inggrid,
> I think I'd need to know what you are exactly doing in
> order to answer properly (the command, with the explanations
> of what the variables stand for), so I can give you just an
> idea. 
> My guess is that, if you're using DID, you're interested in
> the difference before/later a given threeshold for 2
> different groups. If you use year dummies, such difference
> is "absorbed" by them. I mean: you're interested in the
> difference across years, so you shouldn't treat it as a
> confounding factor.
> Federico
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
> 


      

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/