Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: Predicted values from difference-in-difference regression


From   Jonas Eeg Madsen <jonaseegmadsen@hotmail.com>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Predicted values from difference-in-difference regression
Date   Sun, 9 Jun 2013 12:57:42 +0200

Dear All

I am trying to repost this question and hope that some of you can help me. I
have specified, a written what how I have tried to do in stata

I have made a difference-in-difference regression with data from 3 years:
2010 2011 2012. The reform I am testing was in 2010, and the next two years
is post reform. So 2010 is reference category. 
My data is 3 times the actual number of observations. So every local
government appears
3 times, one for 2010, 2011 and 2012. 

reg andel_pleje fisk i.aarstal c.fisk*i.aarstal enlige socio antal_børn
Region ireland

or

reg x1 i.x2 c.x1#i.x2 k1 k2 k3 k4 k5

x1                         are fiscal pressure or not, 0 or 1
x2_2011             are dummies for 2010, 2011, 2012 with 1,2 or 3
c.x1#i.x2 	are interaction terms
k1-k5                  are control variables


So my intent is to predict values on Y for the three years 2010, 2011 &
2012. The values has to be split on the two categories on x1 (fisk ) which
is fiscal pressure or not Basically I want the values to be shown in a graph
with y-values on Y and 2010, 2011 & 2012 on the X, and then two lines for
fiscal pressure or not

A graph as here: https://picasaweb.google.com/lh/phot...eat=directlink

Here is my regression output:

		(Std. Err.	adjusted for 98	clusters in Komnr)
				
		Robust
andel_pleje	Coef.	Std. Err.	t    P>t	[95% Conf. Interval]
				
1.fisk	8.589873	2.680336	3.20   0.002	3.270147     13.9096
	
aarstal	
2		4.237324	1.206105	3.51   0.001	1.84354
6.631108
3		3.675572	1.072663	3.43   0.001	1.546633
5.80451
	
aarstal#fisk	
2 1		-2.381917	1.376892	-1.73   0.087	-5.114666
.3508329
3 1		-2.087134	1.417584	-1.47   0.144	-4.900646
.7263768
	
Enlige		1.092447	.8246021	1.32   0.188	-.5441598
2.729054
Socio		.0460545	.0754113	0.61   0.543	-.1036159
.195725
Antal		-.0001882	.000101	-1.86   0.065	-.0003886
.0000123
Region		4.178592	1.330501	3.14   0.002	1.537915
6.819268
ireland		6.696347	11.2325	0.60   0.552	-15.59705
28.98975
_cons		17.09208	10.03276	1.70   0.092	-2.820169
37.00433
				

Here is what I have tried, but it doesn?t reflect the right colleration:

margins, at (fisk=(0 1) aarstal =(1 2 3) ireland=(0)) atmeans asbalanced
marginsplot


I hope I have made my question clear, otherwise please ask again

Thanks for any help

/Jonas

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


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