Buying a Bitcoin emits 195x as much CO₂ as buying an iPhone

Today a discussion on Discord about NFTs and environmental impact respectfully erupted into a situation XKCD portrays well:

https://xkcd.com/386/

Well, it turns out I was the wrong someone on the internet.

My hypothesis was that CO₂ emissions of cryptoassets are much lower than a large US tech company like Apple when adjusted by market capitalisation¹.

At the end of 2020, Bitcoin (BTC) had a market cap of $539 billion², Apple (AAPL) $2.3 trillion³, Microsoft (MSFT) $1.9 trillion⁴.

And the estimated total CO₂ emissions (CO₂e) for 2020 for Bitcoin 35.89 Mt CO₂e⁵, Apple 22.6 Mt CO₂e⁶, Microsoft approx. 12 Mt CO₂e⁷.

While the carbon emissions for Bitcoin came up higher than I’d expected to find, my confirmation bias had convinced my limited ability in mental arithmetic that low emissions and high market cap meant I was about to nail this stat. I wanted to show crypto has a fraction of the emissions relative to tech giants. But the opposite is true.

Rounding to one decimal place, Bitcoin emits 6.7 times as much CO₂ as Apple, and 10.6 times as much as Microsoft. That is staggering.

And thus circling back to the headline you clicked on. An Apple iPhone 12 Pro late-2020 would set you back $999⁸ and one Bitcoin would cost $29,001.72⁹. On the basis that either choice supports the underlying institution and their overall carbon emissions, the decision to buy one Bitcoin versus an iPhone 12 Pro emits 195 times as much CO₂.

Does that mean cryptoassets are doomed technology? No, this spells the need for established institutions to adopt and support crypto. The demand is there. Meanwhile, we need to be more cautious of the actors we currently enable in crypto networks and the corners they’re willing to cut in the pursuit of profit.

I’d also like to highlight fair objections to the above.

Publicly traded corporations under-report their carbon emissions and crypto mining emissions are over-reported.

While I can’t find concrete evidence to support either scenario, there’s no mistaking the incentive. Apple or Microsoft will favour the leanest calculation of their emissions to report progress towards being carbon neutral (or carbon negative) by 2030. Whereas there’s a current trend on the web to highlight research that overstates crypto mining emissions and bury reports to the contrary. Sensational headlines win the attention economy.

Emissions for proof of stake are substantially less than a proof of work model.

Absolutely, though I’m inclined to say with regards to NFTs we will see PoW take precedence since it reduces fraud. Emissions projected with proof of stake would be much lower. In this case, I’m not using projections, I’m using the estimated historical emissions.

This compares apples and oranges! What about gold mining, paper currency and minting, and the entire banking system?

I’d love to see the same comparison drawn against gold and fiat currencies, but I’m less interested in carbon footprint for the sake of accumulation of wealth. It makes more sense to count our carbon footprint beans in terms of productivity. Outputs of tech companies like Apple and Microsoft also probably outweigh cryptoassets in terms of productivity value for now.