Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: how to balance a panel?


From   "Jeremy Page" <miahpage@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: how to balance a panel?
Date   Tue, 20 May 2008 15:59:41 -0400

Dear Claudio,

I ran into a similar question earlier this year.  I was helped by
people on this list.  Here is the link to the post.

http://www.stata.com/statalist/archive/2008-03/msg01106.html

Best,
Jeremy

On Tue, May 20, 2008 at 12:13 PM, Davide Cantoni
<davide.cantoni@gmail.com> wrote:
> Hello Claudio,
>
> this is the trick I use when I want to balance a panel (y is the
> outcome variable):
>
> *** begin code
>
> sort id year
> by id: egen nonmissing1to7 = count(y) if (year==1 | year==2 | year==3
> | year==4 | year==5 | year==6 | year==7)
> keep if nonmissing1to7==7
>
> *** end code
>
> Maybe there are more elegant solution around, but this one always
> works fine for me.
>
> Best
>
> Davide
>
>
>
> 2008/5/20 Claudio Cruz Cazares <Claudio.Cruz@uab.cat>:
>> Dear all!,
>>
>> I`m working in my Phd. dissertation but a have not find a solution for a crucial problem, hope you could help me.
>>
>> I have an unbalanced panel data of 4050 firms in 16 years but I have monotone and intermittent (since I clean the data from missing values of my dependent variables) missingness and as far as I know, the intermittent missingnes could cause problems when estimating fist differences.
>>
>> I wonder if someone could tell me how could I clean the sample. The easiest thing would be to balance it, but even I can see, after xtdes command, how many firms are in all periods, I can not find de way to keep just them. Is there any way to select the firms based upon their pattern?
>> Other solution is to eliminate the intermittent observations and keep those who died during the panel and those who growth within the panel, but how can I do that?
>>
>> I have generated a new variable in order to identify the missingness of a firm in a year.
>>
>> quietly by Identi: gen dy=year-year[_n-1]
>>
>> dy would be = 1 if there are non missing observation of the firm. Then, I drop the firms with a dy > 1. But I still have intermittent missigness!
>>
>> xtdes
>>
>>  Freq.  Percent    Cum. |  Pattern
>>  ---------------------------+-----------------
>>      577     18.89  18.89 |  111111111111111
>>      158      5.17   24.06 |  11 . . . . . . . . . . . . .
>>      150      4.91   28.97 |  1 . . . . . . 11111111
>>      136      4.45   33.42 |   . . . . . . . . . . 11111
>>      131      4.29   37.71 |  111111111111  . . .
>>      126      4.12   41.83 |  1 . . . . . . . .  . . . . . .
>>      126      4.12   45.96 |  1111 . . . . 111 . . . .
>>      116      3.80   49.75 |  111 . . . . . . . . . . .  .
>>       88      2.88   52.64  |  111111 . . . . . . . 11
>>     1447     47.36  100.00 | (other patterns)
>>  ---------------------------+-----------------
>>     3055    100.00         |  XXXXXXXXXXXXXXX
>>
>>
>>
>>
>> Thanks in advance
>>
>> Claudio
>>
>>
>> *
>> *   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/
>
*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index