Use country-by-country matrices from CPA to COICOP categories instead of global

[martincollignon]

This article gives country-by-country matrices: https://www.sciencedirect.com/science/article/pii/S2352340920302894?via%3Dihub

We may be able to use this instead of our European-wide one for the relevant COICOP categories.

N.B. Exiobase categories would need to be translated to CPA to use this. There is a concordance table available here, but there may be better available: http://carbonconsumptionsurvey.eu/wp-content/uploads/2018/09/Tables_de_correspondances.zip

[floriandierickx]

Hey @martincollignon, just as a sidenote/reference: a similar contribution/discussion was ongoing beginning this year on the correspondence tables you refer to: https://github.com/BONSAMURAIS/correspondence_tables/issues/28, in the context of the work of BONSAI to inventory all existing correspondence tables (https://github.com/BONSAMURAIS/correspondence_tables/tree/master/correspondence_tables/data/raw). Maybe @mfastudillo knows more interesting files/links ?

[martincollignon]

Hi @floriandierickx ! Good to see you here. Right now we're looking at the work of @ghardadi and want to expand that to all countries covered by Eurostat. Paper: https://onlinelibrary.wiley.com/doi/pdf/10.1111/jiec.13045 Data and code: https://github.com/ghardadi/correspondencematrix

What do you think about that?

[floriandierickx]

Hey @martincollignon, that sounds like a perfect way. Did you get in touch with @ghardadi already? He is the man you need for all things related to household expenditure and carbon footprint! When I have some more time I would be interested to follow-up on this issue and look a bit more in detail, but certainly Gilang knows the hidden 'catches in the deal' of it all.

[martincollignon]

We've been in touch. I think the next step would be to map all the data that is needed (happy to do that work) and then re-use @ghardadi 's work to scale from Germany to EUROSTAT covered countries.