The only people who would be mining are the ones who are spending 0$ on electricity, because the electricity would otherwise be wasted.
By heater, I meant literally a heater in someone's house.
The idea is that the only electricity that would be spent on the blockchain is electricity that was going to be sent down the drain anyway.
Intuitively, you're right. However, I read an article (can't find it now, still looking) that presented a convincing case that if there is $X up for grabs for doing a certain amount of work (mining), then the amount of work that is done increases until $X is being spent on the work. So if you're getting heating out of it as well, X increases, but that just means more mining is done until it's no longer marginally profitable.
But intuitively you're right. So I don't quite know where the mistake lies.
I think it's a stretch to say the two are perfectly linked.
If you dropped someone's electricity price in half, would they buy twice the equipment to mine? If you dropped it to 1%, would they buy 100x as much equipment?
Right, but once people do that, mining becomes more efficient. If mining becomes more efficient, more people start mining, then the difficulty adjusts, and then you're back at the beginning.
It's not a bad thing: efficiency gains are good. But they don't reduce the total energy expenditure, because they incentivise increased mining.