Simple regression

[Christine Everaerts]

Hi all,

I want to look at which sectors have a siginificant impact on total trade between two countries.
My dataset has the following format:

Origin	Destination	Year	TotalTrade	Sector	SectorTrade
EU	JPN	1998	1000	0	700
EU	JPN	1999	2000	0	300
EU	JPN	1998	1000	1	300
EU	JPN	2000	100	3	80
EU	JPN	2000	100	2	20
EU	JPN	1999	2000	4	600
EU	JPN	1999	2000	5	1100
EU	JPN	2002	100	9	20
EU	JPN	2002	100	7	80

TotalTrade is the trade flow in one particular year between one country pair (here EU - JPN).
Sector depicts the sector type and SectorTrade is the trade that is done in one particular year between one country pair, all SectorTrades together for a specific year make up the TotalTrade for that year.

I now want to do a difference in difference estimation where I want to look at how the enforcement of the FTA between EU and Japan has had an impact on the sector trade, but I only want to consider the sectors that make up a significant amount of the trade between the countries.

How can I check which sectors are significant?

Thank you!

[Jared Greathouse]

Difference-in-differences is impossible given this current data setup. DD (usually) demands a binary variable of treatment. I don't see this here. you don't have that. You also don't seem to have a comparison group, which you'd base your treatment effects on.

This sort of study might be possible if you were looking at imports for only Japan, where the imported good of interest would be the unit of analysis and the important value the outcome, where you can argue FTA impacted some products more than others. But as it is here, DD isn't on the menu.

[Christine Everaerts]

Yes, this is only a part of my dataset.
I first wanted help setting up a code such that I can look at the significance of each sector in the trade between two countries, here EU and Japan.