Re: st: xtabond

["Brian P. Poi" <bpoi@stata.com>]

Yesterday Jing wrote:

...

> I want to have arellano-bond estimator in my panel data analysis. I know
> the xtabond does the job, but, when I used it in my stata programming, it
> always gave me the wrong message. For example:
>
> . tsset prov time;
>        panel variable:  prov (unbalanced)
>         time variable:  time, 1978 to 1998, but with gaps
>
> . xtabond g_lypcc pop100cr a_pol inst, lags(2);
> timevar (time) may not contain missing values when option full is specified
> r(451);
>
> I have no idea what's wrong and what I shold try next. Could you please
> help me to figure it out? Your help would be highly appreciated!

Jing is apparently using Stata 9.  The error message is being produced because 
the time variable contains one or more missing values.  Typically,
when a panel dataset is unbalanced, the time variable records the times for 
which observations do exist, and there is no observation at all in the dataset 
for a time for which data is missing.

In Jing's case, the time variable actually contains missing values, and I would 
recommend first looking to see why that is the case.  Assuming those 
observations do just represent periods in which data is not available, 
Jing can

   . drop if time >= .

and then use -xtabond- as usual.

In Stata 10, -xtabond- excludes observations with missing values for the 
time variable.

  -- Brian Poi
  -- bpoi@stata.com


*
*   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/