Gravity model - tariffs not significant - ppmlhdfe with exp-time, imp-time, and country pair FEs

[Gerome Retamal]

I want to get elasticity estimates, through the coefficient of the tariffs. I have an intra-national and international trade data, every 3 years from 2001-2019, with corresponding tariffs from the MAcMap-HS6 database of the CEPII (Guimbard et al., 2012).

Can anyone help explain why, my variable of interest tariffs, is only significant if and only if I include country pair fixed effects (FEs) only, and not in combination with exporter-time and importer-time FEs?

• Significant - only country pair fixed effects
  ppmlhdfe exports lntariff, absorb(Exporter#Importer) cluster(Exporter#Importer)
• Significant, but wrong sign (coefficient of tariffs is positive; with exporter-time and importer-time FEs)
  ppmlhdfe exports lntariff, absorb(Exporter#year Importer#year Exporter#Importer) cluster(Exporter#Importer)
• Insignificant (only exporter-time and importer time FEs)
  ppmlhdfe exports lntariff, absorb(Exporter#year Importer#year) cluster(Exporter#Importer)

If it matters, I run an OLS with exp-time and imp-time FEs, and the signs of my coefficients are correct.

[Joao Santos Silva]

Dear Gerome Retamal,

First of all, ignore OLS results :-)

If you have intra-national trade, you need to account for that. What is becoming standard is to create a dummy for international trade, say, inter, and absorb inter#year. You should also make sure you have a large number of countries.

Best wishes,

Joao

[Gerome Retamal]

Dear Joao Santos Silva,

That is helpful. I did the dummy for international trade. Although, the coefficient of tariffs is still non-significant.

May I ask, how to interpret the inter#year? Is it a fixed effect of intra-national (domestic) flow? What does it represent?

And do you have recommended study to cite that used dummy for international trade, and does becoming standard?

Thank you for the help.

[Joao Santos Silva]

Dear Gerome Retamal,

It is simply a way to take into account that intra- and inter-national trade are very different and that the difference can change over time. See

Baier, S., Y. Yotov, and T. Zylkin (2019): “On the Widely Differing Effects of Free Trade Agreements: Lessons from Twenty Years of Trade Integration,” Journal of International Economics, 116, 206–226.

Bergstrand, J., M. Larch, and Y. Yotov (2015): “Economic Integration Agreements, Border Effects, and Distance Elasticities in the Gravity Equation,” European Economic Review, 78, 307–327

Best wishes,

Joao