Toni,
Date sent: Mon, 03 May 2004 14:12:20 +0200
From: "Antonio Rodrigues Andres" <[email protected]>
To: <[email protected]>
Subject: st: observations in xtabond
Send reply to: [email protected]
> I am using xtabond command in STATA. I specify a simple dynamic model
> xtabond y, maxldep(1). This means that I use as intrument yt-2. My panel
> is unbalanced and countries are not observed consecutively. The pattern
> is
> ab country
>
> country | Freq. Percent Cum.
> ------------+-----------------------------------
> 1 | 14 3.54 3.54
> 2 | 21 5.30 8.84
> 3 | 11 2.78 11.62
> 4 | 26 6.57 18.18
> 5 | 20 5.05 23.23
> 6 | 24 6.06 29.29
> 7 | 9 2.27 31.57
> 8 | 14 3.54 35.10
> 9 | 6 1.52 36.62
> 10 | 21 5.30 41.92
> 11 | 31 7.83 49.75
> 12 | 4 1.01 50.76
> 13 | 20 5.05 55.81
> 14 | 31 7.83 63.64
> 15 | 19 4.80 68.43
> 16 | 5 1.26 69.70
> 17 | 14 3.54 73.23
> 18 | 27 6.82 80.05
> 19 | 4 1.01 81.06
> 20 | 27 6.82 87.88
> 21 | 48 12.12 100.00
> ------------+-----------------------------------
> Total | 396 100.00
>
> when running xtabond command appears the following output
> xtabond y, maxldep(1)
>
> llano-Bond dynamic panel-data estimation Number of obs =
> 244
> Group variable (i): country Number of groups =
> 16
>
> Wald chi2(1) =
> 60.15
>
> Time variable (t): year Obs per group: min =
> 1
> avg =
> 15.25
> max =
> 46
>
> One-step results
> ------------------------------------------------------------------------------
> lsrt | Coef. Std. Err. z P>|z| [95% Conf.
> Interval]
> -------------+----------------------------------------------------------------
> lsrt |
> LD | .6781621 .0874415 7.76 0.000 .50678
> .8495443
> _cons | -.000442 .0010119 -0.44 0.662 -.0024252
> .0015413
> ------------------------------------------------------------------------------
> Sargan test of over-identifying restrictions:
> chi2(46) = 75.90 Prob > chi2 = 0.0036
>
> I loose two observations after taking lags and first differences. I dont
> see why the number of groups is 16.
>
> Any thoughts
The culprits are probably
country | Freq. Percent Cum.
------------+-----------------------------------
9 | 6 1.52 36.62
12 | 4 1.01 50.76
16 | 5 1.26 69.70
19 | 4 1.01 81.06
After the loss of observations because of lags and missing values,
these groups probably drop out completely. It looks like one other
does as well.
After the estimation, do
ab country if e(sample)
and you'll see what was used.
--Mark
> toni
> *
> * 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/
Prof. Mark E. Schaffer
Director
Centre for Economic Reform and Transformation
Department of Economics
School of Management & Languages
Heriot-Watt University, Edinburgh EH14 4AS UK
44-131-451-3494 direct
44-131-451-3008 fax
44-131-451-3485 CERT administrator
http://www.som.hw.ac.uk/cert
*
* 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/
