# AW: st: AW: RE: estimating an unbalanced panel

 From Jessica �lschl�ger <[email protected]> To <[email protected]> Subject AW: st: AW: RE: estimating an unbalanced panel Date Thu, 24 Jan 2008 16:43:12 +0100

```
Thank you very much for your suggestions.
I'll try with the imputation.

Regards

Jessica

-----Urspr�ngliche Nachricht-----
Von: [email protected]
[mailto:[email protected]] Im Auftrag von Maarten buis
Gesendet: Donnerstag, 24. Januar 2008 14:41
An: [email protected]
Betreff: Re: st: AW: RE: estimating an unbalanced panel

It looks like your data is balanced but that on some variables you have
lots of missing data. There are two solutions: 1) don't use the
variables with missing data 2) do some form of imputation as I
suggested earlier.

-- Maarten

--- "Jessica [iso-8859-1] �lschl�ger" <[email protected]>
wrote:

> Dear Maarten,
>
>
> when I'm typing xtdescribe for my dataset I get the following output:
>
>
> . xtdescribe
>
>     coid:  1, 2, ..., 30                                     n =
>    30
>     year:  1950, 1951, ..., 2005                             T =
>    56
>            Delta(year) = 1 unit
>            Span(year)  = 56 periods
>            (coid*year uniquely identifies each observation)
>
> Distribution of T_i:   min      5%     25%       50%       75%
> 95%     max
>                         56      56      56        56        56
> 56      56
>
>      Freq.  Percent    Cum. |  Pattern
>
>
------------------------+---------------------------------------------------
-------
>     30    100.00  100.00 |
> 11111111111111111111111111111111111111111111111111111
>
>
---------------------------+------------------------------------------------
----------
>     30    100.00
> |XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
>
> Does this tell you something I haven't recognized yet?
>
> I also typed "sum if year >=1990" because this is the time span
> (1990-2005)for
> which I'm doing my regressions at the moment. The least numbers of
> observations
> for a single variable that stata returns is 85 (see below).
>
> Do you have any suggestions?
>
> Jessica
>
>
> . sum if year >=1990
>
>     Variable |       Obs        Mean    Std. Dev.       Min
> Max
>
-------------+--------------------------------------------------------
>         coid |       480        15.5    8.664472          1
> 30
>        eu_15 |       480          .5    .5005216          0
> 1
>         year |       480      1997.5    4.614582       1990
> 2005
>     corprate |       366    33.99372    5.663233         16
> 50
>     persrate |       268    44.28313    9.637002         13
> 68
>
-------------+--------------------------------------------------------
>      vatrate |       436    17.50532    5.693931          3
> 25
>         pitr |        85    17.00588    9.002372        4.6
> 41
>         cetr |       204    26.71569    7.652872         10
> 43.7
>         ketg |       204    22.23529    5.554781       11.1
> 37.4
>         nwlc |       204    23.36667    8.837636        4.1
> 41
>
-------------+--------------------------------------------------------
>          hbs |       474   -.4729114    6.205395     -63.96
> 16.16
>       fdiout |       450    15767.68    33245.51     -11067
> 252012
>        fdiin |       450     12668.6    30055.53     -38581
> 321274
>      fdioutp |       450    124818.3    261967.7          0
> 2069013
>       fdiinp |       450    96719.37    196991.8          0
> 1553955
>
-------------+--------------------------------------------------------
> gdp1955_2005 |       480    34067.26    127258.6     .39306
> 806621.9
>        toexp |       245    2.93e+11    2.24e+12   1.08e+07
> 2.95e+13
>     toexpppp |       245    1.58e+12    7.82e+12   1.10e+07
> 6.19e+13
>          lbs |       433   -6.227229    75.19171     -791.5
> 172.16
>        cpiti |       320    7.017344    2.086486       2.66
> 10
>
-------------+--------------------------------------------------------
>       totrev |       463    481001.3    940469.4    78.7146
> 8091442
>  totaltaxgdp |       463    35.67134    7.816183   16.57135
> 53.38145
>     t1000gdp |       462    12.81689    5.196823   4.047751
> 30.31366
>     t1100gdp |       462    9.469915    4.958679          0
> 26.37211
>     t1200gdp |       462    2.974308    1.556358          0
> 12.81173
>
-------------+--------------------------------------------------------
>     t2000gdp |       462     9.03067    5.152212          0
> 18.86867
>     t3000gdp |       446    .3676013    .7287513          0
> 5.821948
>     t4000gdp |       462     1.88998    1.048129          0
> 8.41538
>     t5000gdp |       462    11.20277    2.966812   3.970961
> 17.90243
>     t5111gdp |       462    6.187076    3.719843          0
> 62.55744
>
-------------+--------------------------------------------------------
>     t5112gdp |       462    .2805195    .8740527          0
> 8.396804
>     t5120gdp |       462    3.954499    1.534671          0
> 9.740898
>     t6000gdp |       462    .1375409    .3561737          0
> 2.448328
>      dirtgdp |       462    24.09241    6.522121   7.083571
> 40.22071
>    indirtgdp |       462    11.20277    2.966812   3.970961
> 17.90243
>
-------------+--------------------------------------------------------
> (shortened)
>
> Zitat von Maarten buis <[email protected]>:
>
> > --- Maarten Buis asked:
> > > Are you using -xtreg- with the -fe- option to estimate your fixed
> > > effects model?
> >
> > --- Jessica Oelschlaeger wrote:
> > > Yes, I do.
> >
> > Than I am confused, -xtreg- with -fe- does not require balanced
> data,
> > so you should not loose that many units. See the example below:
> >
> > *--------------- begin example -------------
> > webuse nlswork, clear
> > xtdescribe
> > xtreg ln_w union south ttl_exp age , fe
> > *----------------- end example -------------
> >
> > -- Maarten
> >
> > -----------------------------------------
> > Maarten L. Buis
> > Department of Social Research Methodology
> > Vrije Universiteit Amsterdam
> > Boelelaan 1081
> > 1081 HV Amsterdam
> > The Netherlands
> >
> > Buitenveldertselaan 3 (Metropolitan), room Z434
> >
> > +31 20 5986715
> >
> > http://home.fsw.vu.nl/m.buis/
> > -----------------------------------------
> >
> >
> >       __________________________________________________________
> > Sent from Yahoo! Mail - a smarter inbox http://uk.mail.yahoo.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/
> >
>
>
> --
> Dipl. oec. Jessica �lschl�ger
>
> Forschungsstelle Europ�ische Integration
> Universtit�t Hohenheim
>
> mailto:[email protected]
>
>
> *
> *   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/
>

-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

http://home.fsw.vu.nl/m.buis/
-----------------------------------------

__________________________________________________________
Sent from Yahoo! Mail - a smarter inbox http://uk.mail.yahoo.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/

*
*   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/
```