Re: st: how to create panel data out of different sample each year

[Manasi Vydyanath <manasi.vydyanath@gmail.com>]

Therefore, you will not be able to create identifiers (or, for that  
matter, any of the conventional panel data techniques) to your  
dataset. However, there is a way around this: you might be able to  
look into are synthetic cohort methods, i.e., creating groups of  
individuals whose observations you aggregate to create a larger  
"individual". Estimators applied to this larger "individual" will be  
asymptotically consistent. Some papers you can look at for this are:  
Deaton and Irish (I think it's 1982, or 1985), Moffit (1993),  
McKenzie (2000) and Inoue (2005).
As far as I know, there are no existing estimators in Stata for  
dealing with this sort of thing, but it shouldn't be too difficult to  
program some of them. Let me know if this helps at all,
Manasi Vydyanath


On Dec 26, 2006, at 5:52 AM, Murat Karaoz wrote:


> Dear All,
>
> I have 5 years of data (1996-2000). I work with samples differ each  
> year.
> I try to create panel data out of this using all the samples in  
> five year
> period. So in a sense I have a panel with a different sample each  
> year.
> Since samples differ each year, the identifier become different  
> each year
> for all years. Thus you are not able to repeat an identifier for each
> year.
>
> My question is
>
> how to form the "id" (identifier) variable in this situation.  
> Because all
> observations have different id every year which create problem.  
> then xt
> functions dont work.
>
> HAPPY NEW YEAR!!!
>
> Murat Karaoz
>
> *
> *   For searches and help try:
> *   http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/