Armen,
Country pairs must be repeated; one for Vij and another one for Vji. The
bilateral trade value in Vij and Vji will be the same however when you run
the equation the explanatory variables will change (GDP, POP, Tariff..).
Start with the bilateral trade data. Expand this dataset with values Vij to
account for values Vji. Once you've done that merge the macro variables
(GDP, POP, Tariff, etc.) taking country i as the reference one. You will
have repeated macro variables for each observation of bilateral trade per
country i per year.
If you want to take a look to a ready to estimate dataset of this kind, let
me know and I will send you one.
I hope this helps,
Rafa
----- Original Message -----
From: "Nick Cox" <[email protected]>
To: <[email protected]>
Sent: Friday, August 13, 2004 12:36 PM
Subject: st: RE: working out data structure
> In Stata terms each observation should be a pair of countries,
> e.g. Germany and France. Some values would necessarily be
> repeated e.g. the GNP value for Germany would be repeated
> for each pair that includes Germany. Merging transactions
> data files with country property files may be a good
> way to go.
>
> In some fields similar models are called gravity models. If
> Newton had had Stata, or indeed any other decent software,
> he would have been to regress
>
> ln Force_ij = b0 + b1 ln Mass_i + b2 ln Mass_j + b3 Dist_ij
>
> and discover that b3 ~ -2.
>
> Nick
> [email protected]
>
> > -----Original Message-----
> > From: [email protected]
> > [mailto:[email protected]]On Behalf Of Armen
> > Khachatryan
> > Sent: 13 August 2004 11:33
> > To: [email protected]
> > Subject: st: working out data structure
> >
> >
> > T. Otsuki report in Food Policy 29 (2004). The model they
> > suggest looks
> > interesting also to my area. It looks like
> >
> > ln(Vij)=b0+b1ln(GNPi)+b2ln(GNPj)+b3ln(POPi)+b4ln(POPj)+b5ln(DISTij)+
> > b6*ln(MRL)+b7ln(100+TARIFF)+b8*Dbfa+b9*Dtacp+b10*Dntacp+b11*Dr
> > est+b12*Drta+b
> > 13*Dcolij+b14*D1997+b15*D1998+uij.
> >
> > Where i and j mean importing and exporting
> > countries, V denotes the
> > value of trade from j to i. GNPs are clear, the same POPs, and DIST is
> > distance between i and j, MRL the max residue limit imposed
> > by i. TARIFF is
> > applied tariff rate between i and j. The rest are dummies.
> > The data set
> > covers 1997-1999.
> >
> > I have just started to consider doing something like
> > that. Surely
> > one needs data. The data set structure is what I am concerned
> > with now.
> > The problem I have is how to build the data set.
> > Unless I have the
> > idea, I can't proceed with collecting the data.
> > Sorry for the lengthy writing, but I got a kind of
> > mental brake, I
> > need help to move forwards.
> > Question(s):
> > In the data matrix should importing countries appear
> > horizontally
> > and exporters vertically? How I fix the V, POP and others stuff for
> > importers and exporters, how with distance.
> > Do you think it is ok to contact Mr. Wilson and Mr.
> > Otsuki (the
> > author of the article mentioned above) asking them to send a
> > small piece of
> > their dataset?
> > Any suggestion/help/hint will be very much appreciated.
> >
> > Thanks in advance.
> >
> > Armen
> >
> >
> > University of Hohenheim
> > Stuttgart, Germany
> > Tel.: +49 711 459 2600 or 3303
> > Fax: +49 711 459 3709 or 3762
> >
> > ---
> > Outgoing mail is certified Virus Free.
> > Checked by AVG anti-virus system (http://www.grisoft.com).
> > Version: 6.0.737 / Virus Database: 491 - Release Date: 11.08.2004
> >
> >
> > *
> > * 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/
> >
>
> *
> * 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/
>
*
* 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/
