“The rise of Bitcoin mining has inadvertently sparked a surprising ally for coal conservation, as renewable energy sources join forces to save the environment – one digital coin at a time.”
Bitcoin Miners Are Making The World Even Worse
New Book Tuesday is the best day of the week Chez Auffhammer. Especially when there is a new Michael Lewis release in the pile. Going Infinite describes the rise and fall of Sam Bankman-Fried, the founder of Alameda Research and the Crypto exchange FTX. Lewis got unprecedented access to Bankman-Fried and in nail-biting fashion describes one of the bigger implosions in finance history. I complained about crypto in 2018, but had sort of lost interest. Severin had pondered grid stability and crypto mining in another post. But this book made me wonder how the ole mining business is doing.
GPUs are amazing things. They are gadgets that allow you to do massive amounts of computations in parallel, but they consume an enormous amount of power and generate significant heat! I moved mine into a refrigerated basement across campus, as my office resembled a sauna. And no, I am not mining crypto. I am using these gadgets to figure out climate impacts.
So in my recent reading journey across the nerdverse, I came across a piece by Anna Papp, Doug Almond and Shuang Zhang called “Bitcoin and Carbon Dioxide Emissions: Evidence from Daily Production Decisions”. The first thing I learned is that the US overtook China in 2021 as the number one location for cryptomining, since China banned mining for crypto. Some US states even went as far as providing significant incentives for miners to locate within their borders, which reminds of an older literature on jurisdictional competition. You might think that this is a good thing, as on average, the grid here is cleaner than in China..
Turns out things are trickier than they appear. What really matters is how much actual mining is done and how the electricity used to mine is generated. You might think that these miners just buy from the grid, so we should think about this issue the same way we think about plugging in an EV and contemplating which type of power is at the margin.
The paper shows a worrying case study of the Scrubgrass Power plant in Kennerdell, Pennsylvania. This is the type of plant you want to phase out if you care about mother earth. It burns waste coal and is relatively inefficient. It’s a literal dumpster fire. It was on its way to be phased out (yay) until a cryptominer bought it and moved in their GPUs. Yes, you read that right. This is a power plant that generates electricity and mines crypto on site. In the alternate crypto free universe, this power plant would have been mothballed.
So is this bad? This is where the paper gets interesting. The cryptobros have all kinds of delusional stories about why this currency is making the world a better place. This paper is a pretty powerful counterexample. What they do is look at the price of crypto, which we observe and look at what happens to the emissions at Scrubgrass. They have thousands of price movements and observe emissions (since the plant has Continuous Emissions Monitoring System, CEMS). These serious scholars think carefully through issues that might confound the effect of price on emissions and I am convinced by what they find.
Are you ready? They find a long-run price elasticity of 0.33-0.40 and a bigger short-run elasticity of 0.69-0.71. What does that mean Max? If you take the most recent estimate of the social cost of carbon of $190 a ton, this implies that a $1 increase in daily Bitcoin price leads to an additional $3.11-$6.79 worth of daily damages from carbon emissions at Scrubgrass. Year to date the price of Bitcoin has gone up almost $20,000, so their results imply large emissions increases
Do I really need to point out that this is really dumb and something that should not be allowed to happen? Miners are creating value which they get to deposit in their bank accounts, while the rest of us pay $3-6 for each dollar they make. This is “I literally spit out my coffee when I read that abstract” stupid. So what do we do?
Two things. Price carbon. Globally. Not gonna happen just yet (although I am curious what will happen when the Trump tax cuts expire), or follow China in banning the mining of crypto domestically. If we don’t like bans, at least prevent these firms from buying old dirty power plants. If the estimates from this paper are right, then it might make sense to buy these out using public funds and mothball them. That would sure beat turning them into carbon mines. And along the way, it might help discourage scaling up cryptononsense.