Last month, Biden’s proposed 2024 budget included a new DAME tax (Digital Asset Mining Energy), which would tax large-scale cryptocurrency mining operations by 30% of their electricity usage. This comes in response to the rapidly growing energy expenditure of cryptomining in the US, which became the world leader in cryptomining after China banned the practice in 2021. The 34 largest bitcoin miners in the US consume approximately 3.9 GW (per analysis by WattTime); global bitcoin mining consumes roughly 15 GW, of which about 38% is in the US, or 5.7 GW. This is about 1 / 75 of the US’s total electricity consumption, or 1 / 600 of the US’s total energy consumption. Note that estimates vary considerably, especially across different years.1Global mining rate of bitcoin can easily be found, currently hashes per second; the uncertainty lies in how much electricity is needed by various hardware to calculate each hash, as well as what fraction of mining is performed in the US.
Bitcoin mining involves a computationally-intensive search for valid “blocks” to add to the blockchain; these blocks contain a record of any new bitcoin transactions, and it is the computational intensity of finding valid blocks that inhibits people from inventing a fraudulent transaction history.2To prevent people from creating a single fraudulent block of transactions in isolation, each block refers to the previous block; hence the “chain” in blockchain. An attacker would have to recreate the entire transaction history, and therefore use as much computational power as all other bitcoin miners combined. Successful miners are rewarded with a small amount of bitcoin. Every two weeks the difficulty of finding a new valid block is dynamically adjusted so that a new block is found, on average, every 10 minutes.
This is a proof-of-work cryptographic system, where being able to produce a valid block proves you expended a certain amount of effort (say cpu time, or electricity) to find it.
So what happens if you tax the work in a proof of work system? Very little! There is nothing in the bitcoin protocol that requires the work to be in the form of cpu time or electricity; taxes serve equally well as work. With higher taxes driving out the least profitable miners, the mining rate will go down, and the mining difficulty will automatically decrease to compensate. As transaction fees are presumably extremely inelastic to the present mining rate, total mining revenue will be unchanged, but less electricity will be expended to find each block. The energy not being spent on mining instead gets captured as taxes, where it can be re-used by the federal government for useful purposes instead of being lost as waste heat in a datacenter.
If all mining is equally efficient, then a flat 30% tax applied to the US’s cryptomining sector would lower electricity usage by 23%, a savings of 1.3 GW.3Assuming that none of the sector is displaced internationally. Of course, one expects miners to move to seek lower taxes, but there is a significant global movement towards taxes / other mining restrictions. This also supposes non-electricity expenditures of mining are negligible. Assuming 10 cents per kWh (modestly below the US average, but this may be too high as presumably miners are located whereever the cheapest electricity is), the electricity savings are turned directly into tax revenues to make 1.15 GUSD per year.4“giga US dollars” feels natural after so much time reading about electricity consumption. A more sensible unit than “quad”s, anyhow.
However, raising taxes will first drive out the most unprofitable miners, who are presumably also the most inefficient. Thus electricity usage should drop by a bit more than 23%, and the tax revenue be somewhat less than predicted. The difference (i.e. the extra electricity savings plus the money not paid as taxes) goes to increased mining profits, except a tiny amount goes to lowered bitcoin transaction fees (depending on how elastic the fees are to mining rate). Further raising mining taxes simply further lowers electricity consumption and raises tax revenue, all with no harm to bitcoin nor to miners’ profits!5Again, assuming no tax evasion or international displacement; with no regulation legitimate miners would be driven out by tax evaders.
Another way to see this is that a 30% tax is equivalent to if all miners idled 23% of the time, and gave the money they saved to the government. Idling 23% of the time is equivalent to having 30% excess capacity. Of course, people will want to use this excess capacity, and not just leave it idle; the most profitable (and presumably efficient) miners will be able to exploit this extra capacity, driving out less efficient miners. Total mining profits go up and electricity usage goes down.
Bitcoin – defying reasonable market economics since 2009.
Post scriptum At insane tax rates the “mining” process becomes functionally equivalent to just handing the government a chunk of money to get the next block in the block chain, with no intensive computer search at all. While this may appear to make bitcoin vulnerable to cryptographic attack, this is only true to the extent to which attackers can evade the mining tax. To prevent tax evasion, the government can publish which blocks are authentic – and lo and behold, we’ve reinvented centralized fiat currency.
Follow RSS/Atom feed for updates.