st: Re: how to deal with variables...

[Kit Baum <baum@bc.edu>]

I think what is being considered here is attrition: certain  
individuals drop out of the panel. That is not a problem. Imagine  
that your panel ID is the individual's SSN (or social insurance  
number). What you will find is that the panel is unbalanced, with  
more individuals in the first wave of the survey than in the second,  
more in the second than the third, etc. if there is no replacement.  
That is a very common setup -- it happens with firm-level data, for  
instance, due to mergers, liquidations, etc.

If you do not have preassigned ID codes for each individual (but  
somehow you know who is whom in each year) then just assign 1...N to  
the individuals in the first year, and assign the appropriate IDs in  
the second year, and so on. That is, if somehow you know that  
individual 5 has left the sample in year 2, just leave it out. Stata  
does not need a balanced panel.

Kit Baum, Boston College Economics
http://ideas.repec.org/e/pba1.html


On Dec 7, 2005, at 2:33 AM, statalist-digest wrote:


> But now that you are talking about obs, I have another
> question:
> when I create the panel, I assign an id to each
> individual, but as we move through years, the number
> of obs decreases so I cannot assign the same panelid
> so how do we treat the missing observation?
> Thanks for your help
> Feno
*
*   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/