In Defense of Power-Hungry Services (That Aren’t Bitcoin)

Why Energy Consumption Is Usually Worthwhile


January 23, 2018 | Alex Trembath

Of the various backlashes to the Bitcoin craze, one keeps popping up: Bitcoin uses a lot of electricity—therefore, Bitcoin is a bad idea.

I would reverse this: Bitcoin and other cryptocurrencies are mostly a bad idea—therefore, whatever energy they use is mostly wasted.

This point may sound trivial, but it’s not. Many criticisms of Bitcoin launch exclusively from its energy footprint, as though energy consumption is always wasteful. In a prototypical example at the Guardian last week, Alex Hern calls Bitcoin a “climate threat,” writing that “bitcoin mining is a competition to waste the most electricity possible by doing pointless arithmetic quintillions of times a second.”

Hern’s critique would be far more compelling if he made a case that Bitcoin is “pointless” to society. But he doesn’t. His sole criticism of Bitcoin is its energy use, concluding that—were Bitcoin to disappear—“we might be able to return to worrying about more conventional sources of climate change, like the automotive industry, plane travel, and Donald Trump.”

This is silly. Goods and services that use a lot of energy (like the automotive industry or plane travel) can be worthwhile things to society. Bitcoin simply … isn’t that useful. (For more on why I think this, read Bloomberg’s editorial and this thorough takedown by Kai Stinchcombe.)

That is, I think Bitcoin is a bad, pointless idea, without even considering its energy usage. And, crucially, no one can seem to agree on how much electricity the Bitcoin network actually does use. Hern and others cite a Digiconomist estimate of 42 terawatt-hours per year, but crypto advocate Marc Bevand pegs it closer to 7 terawatt-hours per year. Whether it’s a lot or a little, I think it’s too much.

The problem with overemphasizing the energy footprint of cryptocurrencies is that it treats energy consumption a priori as a negative. But energy consumption is as good as the products and services it provides humanity—so, mostly pretty good. Energy enables economic activity, communications, security, travel, entertainment. As energy consumption has risen over time, so have education, democracy, human health, and global economic inclusion, while poverty, violence, and vulnerability to disaster and disease have all declined.

Yes, energy waste and inefficiency are real and should be addressed. But energy efficiency and energy consumption have a tricky relationship. It’s worth noting that as large-scale energy services have gotten more efficient over time, energy consumption in those sectors has gone up, not down. Lighting, power generation, steel, and other sectors have all gotten markedly more energy efficient over the decades, but these efficiency gains in these sectors have backfired to redound to higher energy consumption overall, since higher efficiency means higher productivity. This phenomenon, broadly known as the efficiency “rebound” effect, usually doesn’t entail an absolute increase in energy consumption in any given sector, and it does saturate in most cases—but it is persistent, and significant. We can eliminate some energy waste, and certainly become much more efficient—but don’t expect global energy consumption to fall anytime soon.

We should also refrain from treating large new sources of energy demand as automatically threatening or wasteful. Increased energy consumption in one sector can actually yield real benefits—synthetic fertilizer and mechanization of farming in the 20th century, for instance, increased agricultural energy use but enabled more productive farming and more efficient use of land. Bitcoin may be a bad idea, but desalination and wastewater treatment, large-scale materials recycling, direct-air carbon capture, vertical farming and aquaculture, and spaceflight have a lot to say for them—and all will come with significant energy demands. This is to say nothing of the latent demand for mundane technologies like refrigerators, cars, and air conditioners in places like India or sub-Saharan Africa—latent demand that could cause global energy consumption to triple or quadruple by later this century.

Of course, Hern and others do point to a legitimate concern: climate change. Cryptocurrencies use electricity, most of which still comes from coal and natural gas, and the associated carbon emissions hasten global warming. This is obviously worrying, but it is much more a reason to accelerate to low-carbon alternatives like advanced nuclear and advanced solar than it is to forego energy-hungry services like Bitcoin. The main reason to forego Bitcoin is because it’s dumb.

I have no idea how Bitcoin and other crypto- and blockchain-enabled services will scale, nor would I rule out the possibility of them proving useful in the future. But the panicked reaction to cryptocurrencies’ energy consumption sets a worrying precedent. We should not shy away from new technologies and new forms of consumption simply because they impose large new energy demands. Many of those new demands will be much more worthwhile than Bitcoin, and many (think urbanization, agricultural modernization, and even more far-afield things like carbon capture technologies) will consume lots of energy in service of making civilization more resilient to climate change.