Comments | economics: oh god halp

tisiphone in economics

oh god halp

Nov 16, 2010 16:58

The proposal for my nice historical overview of the Cote d'Ivoire cocoa industry came back from my prof today, with comments. Why not focus on the Cocoa Marketing Board? he sez. How about an econometric analysis? he sez. So, I turn to you, because I suck at econometrics. I know what to do, just not how to pick my models. How would you analyze a ( Read more... )

Comments 4

everyoneiknow November 17 2010, 01:06:00 UTC

Honestly, what you use for variables will probably be determined largely by what you can find reliable data on. Have you started looking for data over that time period already? Are you looking at international demand or just demand within the country? (I don't know anything about cocoa trade.) Could you find another cocoa producing country and do a comparison of some sort? You might be able to look through WTO data/statistics and see if anything jumps out at you as being relevant (and having data available) that you wouldn't have thought of otherwise. Beyond that I'm not sure where to find reliable statistics about Cote d'Ivoire.

tisiphone November 17 2010, 01:10:18 UTC

The FAO has trade data (volume/price) for country-level trade and international demand. Domestic demand is negligible, cocoa is Cote d'Ivoire's major export commodity and it's all produced for exchange rather than internal use. I'm not real hung up on variables right now, I'm just struggling with what basic model might be best. I may just use OLS regression, but I don't know how to account for the pre and post conditions.

everyoneiknow November 17 2010, 01:17:57 UTC

Is there a reason why it needs to be more complicated than OLS with a DV for pre/post? I've only had two Econometrics courses (one undergrad and one graduate), but no other model is immediately coming to mind as something that would be more appropriate.

tisiphone November 17 2010, 01:37:28 UTC

I dunno, that's my question - whether something else would get at what I'm trying to learn better? I'm not sure. I'd like to do time series, because the 70s is going to throw everything off - there was a big spike in demand and price then. But that might just be too complicated, given how terrible I am at this. Thanks for helping me think it through.