Page 1 of 1

Hedonic model - panel data

Posted: Thu Feb 26, 2009 5:53 am
by siviq
Hi,

I am doing a research on real option components' influence on agricultural land prices in Bulgaria. I have gathered data for 238 municipalities and 28 regions over 8 years. I am planning to run several regressions.
1. lnPrice = b + YD + lnPop + Migr + u
2. lnPrice = b + YD + lnPop + Migr + RD + u
3. lnPrice = b + YD + lnPop + Migr + RD + DSof + DSea + u
4. lnPrice = b + YD + lnPop + Migr + RD*DSof + RD*DSea + u
5. lnPrice(reg) = b + YD + lnPop(reg) + Migr(reg) + RD*DSof + RD*DSea + Urban(reg) + GDP(reg) + u

YD - year dummies
Pop - population
Migr - Migration
RD - regional dummies
DSof - distance to Sofia
DSea - distance to sea
Urban - urbanisation

My question is: How do I structure my Eviews workfile? (I use Eviews 5)
And more importantly: How do I import the spatial variables - distance to Sofia and distance to sea?

Here are my Excel worksheets:
Eviews data.xls
(401 KiB) Downloaded 546 times
Eviews data distances.xls
(169 KiB) Downloaded 485 times

Re: Hedonic model - panel data

Posted: Thu Feb 26, 2009 9:09 am
by EViews Gareth
Have you followed the following guide:
http://forums.eviews.com/viewtopic.php?f=7&t=74
?

Re: Hedonic model - panel data

Posted: Thu Feb 26, 2009 10:15 am
by siviq
Thank you for the quick answer.
The second question is the one that really bothers my. How do I import non-time varying variables? As distances from municipalities don't change overtime I can't import them like the other variables.

Re: Hedonic model - panel data

Posted: Thu Feb 26, 2009 10:21 am
by EViews Gareth
Two options.

1) In Excel just copy the data so that every year has the same value (this is probably the easiest)

2) Open the distance data into a new cross-section EViews page (i.e. non-panel with no dates), then copy it over to the panel page. As long as the same cross-section ids are used on the cross-section page and on the panel page, then when you copy from the cross-section page to the panel page, EViews will do the repeating for you.

Re: Hedonic model - panel data

Posted: Thu Feb 26, 2009 10:34 am
by siviq
So both ways the values of distances should be repeated for all years. I guess the first way will really be a lot easier then.

Thank you very much for your help.