Hi,
I am replicating one paper. In the paper, a system of the panel version of
the following regression* is estimated in differences and levels. This
regression is differenced and the lagged values in levels of all
explanatory variables are used as instruments. Similarly the lagged
differences of all explanatory variables as instruments for the level
version of the following regression is used in the estimation. Difference
and level regressions are combined in aystem. Thus the panel estimator
uses instrumental variables based on previous realizations of the
explanatory variables(internal instruments)
*yi,t − yi,t−1 = (�� − 1)yi,t−1 + ��FDi,t + ��Xi,t
+ ��i,t.
��n order to apply DPD using Stata,firstly I run the following commands
.encode countrycode,gen (countrys)
.tsset countrys year
panel variable:countrys, 1 to 88
time variable:year 1958 to 1998 with gaps
xtabond growthingini logprivatecredit inflation logtrade logschooling
loginitialgini , lags(1) artests(2)
no observations
r(2000)
since there are gaps within years,lag of dependent variable,say
growthingini can not be generated. So error of no observation takes place
and the command xtabond doesn't work with this unbalanced panel.
country year var 1 var 2 var 3 var 4
. .
. .
AUS 1986 4.621401 3.546704 2.339785 -.0403954
AUS 1993 1.197454 3.683009 2.367061 .0228291
AUT 1987 2.860833 4.249761 2.049111 .0111748
AUT 1994 1.087343 4.330157 2.086045 -.04415
BEL 1979 5.283737 4.885247 2.109486 -.0124152
BEL 1985 3.148579 4.909156 2.149084 .0037365
BEL 1992 2.223942 4.897898 2.208274 -.0008712
BGD 1963 11.59548 3.013667 -.1875351 -.0035743
BGD 1973 31.88276 2.874345 .247641 -.00466
this is just part of my dataset. How can I solve and reshape data in order
to apply DPD?
Thanks in advance
--
Sy
*
* 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/