Problem with the interpretation of PPML results of an augmented gravity equation

[Imke Taylor]

Hello,

I hope someone here can help me with the interpretation of my Stata PPML results of an augmented gravity equation. Here the dependent variable is the outward FDI position from investor country i to host country j in USD. My main variable of interest is a policy index that takes values form 1 to 7 (SEER_investor and SEER_host ). Even thought the results are obviously not significant, I would still like to interpret them economically. I would have said something like that: "For every unit increase in SEER_i, the corresponding difference in expected mean of outward FDI positions will amount to negative 0.033 units of FDI, holding all other factors constant." This doesn't seem really logical to me. Does any one have a tipp or recommendation for me?

Thank you very much in advance!

[Joao Santos Silva]

Dear Imke Taylor,

These coefficients are semi-elasticities, so, one unit increase on SEER_i lowers the expected value of the flow by approximately 3% = 100*(exp(-0.033) - 1)%.

Best wishes,

Joao

[Imke Taylor]

Dear Joao Santos Silva, thank you very much for your fast answer! Is this due to the PPML estimator? Sorry i am bit lost, i thought semi-elasticities are normally mentioned in the contexts of log variables.

[Joao Santos Silva]

Dear Imke Taylor,

It is due to the fact that the gravity equation is an exponential model and therefore the interpretation of the coefficients is as in a linear model where the dependent variable is in logs.

Best wishes,

Joao

[Imke Taylor]

Thank you very much Joao Santos Silva