[Carlo Lazzaro]

Friederike:
thanks for clarifying.
I thought that you had run a robust hausman test, such as the one provided by the user-written programme -xtoverid- (rnethelp "http://fmwww.bc.edu/RePEc/bocode/x/xtoverid.hlp")

[Friederike Osterndorff]

Hi Carlo,

thanks for the tip. I didn't know that command. However, I tried to install it, because I think it could be useful. But I get an error message:

package name: xtoverid.pkg
from: http://fmwww.bc.edu/RePEc/bocode/x/

checking xtoverid consistency and verifying not already installed...
installing into c:\ado\plus\...
could not rename c:\ado\plus\stata.trk to c:\ado\plus\backup.trk

r(699);

I tried the solution suggested here:

http://www.stata.com/support/faqs/da...ment-variable/

but its not working:
-set STATATMP- not allowed; 'STATATMP' not recognized.

Any other ideas?

Thanks

Friederike

[Nuzaba Tahreen]

Hello
I am using a dummy variable as my main explanatory variable but it time invariant. when I run xtreg with fe command, main explanatory variable omitted. Is it a problem to use such a variable for as main explanatory variable for panel data? I have 143 developing countries for 15 years. Again when I run it with re, it gives me estimates without omitting my main dummy variable.
I am really confused.
Best
Nuzaba Rahman

[Carlo Lazzaro]

Nuzaba:
welcome to the list.
Please, start a new thread. Thanks.

[Abu Tauheed]

Dear Maarten Buis,

I am doing an impact evaluation of a Maternal Health Policy for 18 Indian States. To find a treatment effect, i have created a treatment variable (1 for those who are treated and 0 otherwise). Similarly, i have created a time dummy (1 indicating the post policy period and 0 otherwise). My dependent variables are dichotomous, thus I am using a non-linear regression equation. Therefore, i am using panel logistic model. However, when I am, estimating the Average Treatment Effect (ATE) in STATA, it states that due to collinearity my treatment variable has been omitted.

Kindly, suggest how to overcome this issue.

[Arshad Khan]

Valuable and Informative discussion

[khadija mubarka]

Hi All,

Could you please help me in solving this issue. I am facing a problem in linear regression estimation. I have panel data, 471 firms with 9 years. When i run fixed effect model (because hausman suggest fixed effect instead of random) then dummy variable D3 drops along with all industry dummies included (stata says that (it omitted because of collinearity). Dropping down industry dummies is not a problem for me, but this D3 is my main variable. My model is:

xtreg Y X1 X2 X3 X4 D1 D2 D3 D4 industry_dummies, fe

D1 = firm is family firm or not
D2 = family member is present on board or not
D3 = firms is owned by foundation or not
D4 = foundation is for charity purpose or not

(all these 4 dummies are main variables)

industry dummies = here i included 11 industry dummies and also excluded one.

I shall be grateful for your help. please let me know if you need any other information.

[Carlo Lazzaro]

Khadija:
you should have started a neew thread..
That said:
-it's better to rely on -fvvarlist- for creating categorical variables and interactions rather than creating them by hand;
--fe- specification drops each time-invarying predictors;
-dropped variables are collinear with the fixed effects; hence there's nothing you can do but changing your model specification.

[Maria Kohnen]

Hi Carlo,
thats strange. Everything is the same. Exact that I made the interaction terms in the first regression by multiplyig the two regressions. and in the second one, I used

##

, like you proposed...
is it okay to also make interactions by just using the product of the two variables?
also, when I use outreg2, the regression including the

##

version looks very messy...any other/better way to export the output tables?
or is outreg2 the prefered way?
looks like this:

	(1)	(2)
VARIABLES	RDlog	RDlog
1.POST_FINE_DUMMY	0.105**	0.105**
	(0.0482)	(0.0482)
1o.LENIENCY_DUMMY	-	-
0b.POST_FINE_DUMMY#0b.LENIENCY_DUMMY	0	0
	(0)	(0)
0b.POST_FINE_DUMMY#1o.LENIENCY_DUMMY	0	0
	(0)	(0)
1o.POST_FINE_DUMMY#0b.LENIENCY_DUMMY	0	0
	(0)	(0)
1.POST_FINE_DUMMY#1.LENIENCY_DUMMY	-0.0354	-0.0354
	(0.0905)	(0.0905)
1o.index	-	-
2o.index	-	-
0b.POST_FINE_DUMMY#0b.index	0	0
	(0)	(0)
0b.POST_FINE_DUMMY#1o.index	0	0
	(0)	(0)
0b.POST_FINE_DUMMY#2o.index	0	0
	(0)	(0)
1o.POST_FINE_DUMMY#0b.index	0	0
	(0)	(0)
1.POST_FINE_DUMMY#1.index	0.0667	0.0667
	(0.0846)	(0.0846)
1.POST_FINE_DUMMY#2.index	0.0432	0.0432
	(0.113)	(0.113)
1997.year	0.142	0.142
	(0.0955)	(0.0955)
1998.year	0.242***	0.242***
	(0.0907)	(0.0907)
1999.year	0.225**	0.225**
	(0.0977)	(0.0977)
2000.year	0.399***	0.399***
	(0.0990)	(0.0990)
2001.year	0.343***	0.343***
	(0.107)	(0.107)
2002.year	0.320***	0.320***
	(0.111)	(0.111)
2003.year	0.261**	0.261**
	(0.111)	(0.111)
2004.year	0.256**	0.256**
	(0.113)	(0.113)
2005.year	0.252**	0.252**
	(0.117)	(0.117)
2006.year	0.271**	0.271**
	(0.122)	(0.122)
2007.year	0.240*	0.240*
	(0.129)	(0.129)
2008.year	0.178	0.178
	(0.134)	(0.134)
2009.year	0.166	0.166
	(0.137)	(0.137)
2010.year	0.233	0.233
	(0.142)	(0.142)
2011.year	0.272*	0.272*
	(0.149)	(0.149)
2012.year	0.394**	0.394**
	(0.159)	(0.159)
2013.year	0.136	0.136
	(0.173)	(0.173)
2014.year	0.163	0.163
	(0.182)	(0.182)
2015.year	0.517**	0.517**
	(0.211)	(0.211)
Constant	18.81***	18.81***
	(0.117)	(0.117)
Observations	1,446	1,446
R-squared	0.096	0.096
Number of ID	145
Time FE	YES
Year FE

Robust standard errors in parentheses

*** p<0.01, ** p<0.05, * p<0.1


i mean, i should first of all exclude the year dummies, right? thats way I used the

addtext

function.
but just looking at the interaction terms, it looks kind of confusion..?

thank you, best

[Salman Mallick]

This was a good discussion. I have started a new thread on a similar topic here: https://www.statalist.org/forums/for...nel-regression, and have thus removed my reply.

[Carlo Lazzaro]

Salman:
Thanks for pointing this out.