The Banana Index (from The Economist)

[aguformoso]

The Economist recently released an article titled “A different way to measure the climate impact of food”, which proposes a reference framework to compare the impact food has on the environment. The creators of the Big Mac index proposed a simple approach: let’s compare the footprint of food against… that of bananas. And that’s how the Banana Index came to life

I thought it would be useful to bring the data into Observable and start playing with it. I’m looking forward to new and creative ways of slicing and dicing the data! What types of conclusion would you be interested to reach based on this data? How would you visualise them? Let me know your thoughts through this post or as a comment in the notebook itself.

The Banana Index (from The Economist)

[Fil]

Beef is so bad, no wonder the Big Mac index is causing problems

How do you read the tooltips? At first glance it seems to be meaning the opposite (you’d need 1 banana or 551 tea serves to get as many proteins)?

[mootari]

Never thought I’d see a serious application of Banana for Scale.

[aguformoso]

Right, some more text will definitely help the reader The tooltip should be read as “Tea contributes 551.4 more times CO₂ than bananas in order to get the same amount of proteins”. When choosing to show absolute numbers we can see tea=4257.8 and bananas=7.7 kg CO₂ per 100 grams of protein provided.

Thanks for checking the viz out!

[yurivish]

Only tangentially related, but you might also enjoy this venn diagram from an article about the banana genome where a banana is one of the venn areas:

[CAYdenberg]

Am I interpreting correctly that the size (radius? area?) of each circle is proportional to the banana index by kg? In other words, in the middle two charts, size and position along the horizontal axis are redundant?

[aguformoso]

Hey, indeed the second row has redundant x and r (radius) attributes.

Vertical faceting (fy) is applied alphabetically, making “kg” the second row. I think users might find it easier to read redundant data first, and only then the other metrics… what do you think? Would that help to tell the same story in a better way?

Thanks for your interest!

[CAYdenberg]

I would make kg the first row if you can. It’s a small point, just took me an extra beat to figure out what I was looking at.

One other point I learned from Tufte: if radius is proportional to emissions per kg, then area is proportional to the square of emissions per kg, which isn’t precisely fair (the big circles are bigger than they should be).

OTOH if r is a log of emissions per kg (as x is) then … hm … I would have to think more about it …