GDP on Gravity Model

[Stephen Apostol]

I really need help on my panel regression. My dependent variable is bilateral trade flow. My Independent variables are GDP, Exchange Rate, Inflation, and Transportation cost (as a proxy variable to distance, since distance is constant so numbers are fixed and can make my results worst). Gravity model states that GDP and Distance must be negative. I log my variables, I tried experimenting and combing logs to constant variables. I did make Transportation cost turn to negative, my only problem is that the GDP holds its positive sign in anything that I do. Please help, I really need to finish this because I want to graduate. Thanks to anyone who is/are willing to help.

[Joao Santos Silva]

GDP should be positive!

Joao

[Stephen Apostol]

GDP should be positive!

Joao

The variable GDP on my regression is always positive. I log some variables and experiment, trial and error basis, to make my transportation cost negative. If transportation cost is negative then GDP is positive, Am I done with it? Thanks Joao

[Joao Santos Silva]

Dear Stephen Apostol,

In the gravity equation for trade, the trade goes up with the size of the partners (measured by GDP and other variables) and goes down with the trade costs (measured by distance and other variables); have a look at the Wikipedia page on this. So, it is natural that your transportation costs have a negative coefficient and GDP a positive on.

Best wishes,

Joao

[Francis Asare]

Hello,
I am student and I am hoping to defend my thesis soon because I have delayed. I am using gravity model
to estimate effects of trade dispute.
The signs for the co-efficient does not conform to the gravity model. eg. GDP becomes negative and Distance becomes positive.
I would be much appreciated if anyone can help.