Broadtermism
Lots of the big ideas in EA came from noticing that certain numbers were large. In the early days, EAs noticed that number of people we can help is quite large, especially compared to the number most people actually help.

Lots of the big ideas in EA came from noticing that certain numbers were large. In the early days, EAs noticed that number of people we can help is quite large, especially compared to the number most people actually help.
Lots of the big ideas in EA came from noticing that certain numbers were large.
In the early days, EAs noticed that number of people we can help is quite large, especially compared to the number most people actually help. Then they noticed that there were a lot of farm animals too. Then they noticed there are even more wild animals than that—and even more expected future sentient beings than that.
Another number that is large: the number of stars. Rumor has it there are well over a hundred stars. Might this be another source of important moral conclusions?
These days, everyone and their mother is a Longtermist. The core argument for Longtermism is simple: the future is so big that nearly all those affected by our actions, in expectation, will reside in the far future. Thus, if we care about those whom our actions affect, we should primarily be concerned about the impacts our actions have on the far future—especially when harmful actions threaten to extinguish the future entirely. Or, put in a sentence: most people will be far from us in time, so the most important features of our actions are their effects on those in distant times.
But an overlooked fact: there’s an exactly parallel argument for space (thanks to Toby Ord for pointing this out and coming up with the name Broadtermism). Rumor has it that the universe is a pretty big place. Nearly all the stuff in the reachable universe is far from us.
In expectation, then, only a tiny slice of what matters is what will happen in the nearby bubble—even our galaxy only represents a tiny fraction of the affectable universe. What happens on Earth represents only a tiny fraction of the morally important things that will happen throughout the history of the universe. Far more important is ensuring things go well in the distant star systems where nearly everyone, in expectation, will live.
Call this view Broadtermism.
The case for Broadtermism is a lot like the case for Longtermism. Longtermists say: when in time you live doesn’t affect how much you matter. Thus, because nearly everyone will live in the far future, and those people matter as much as the rest of us, almost all of what matters is what will occur in the far future. Broadtermists say: where in space you live doesn’t affect how much you matter. A tiny fraction of future people will dwell on Earth.
I think that Broadtermism is correct for basically the same reason Longtermism is correct. People in distant galaxies matter. Almost everyone will be in a distant galaxy. So almost all of what matters is what happens in distant galaxies. Because things we can do now affect distant locations—extinction would prevent us from settling any of them—nearly all of the expected impacts of our actions lies in how they affect distant locations.
Like with Longtermism, there is both strong and weak Broadtermism. Weak Broadtermism merely says that making things go better in distant spatial regions is very important and the world should do a lot more of it. Strong Broadtermism says that it’s nearly all of what matters—vastly more important than what happens in nearby regions.
Similarly to Longtermism, there is both axiological and deontic Broadtermism. Axiological Broadtermism says that most of what matters morally is what will happen in the distant reaches of space. Deontic Broadtermism says that the most important features of our actions are how they affect distant regions of space—so the main determinant of whether an action is worth taking is how it will affect far out spatial regions.
What implications does Broadtermism have?
First, it means we should be a lot more willing to compromise with respect to Earth. Conventional morality primarily cares about what happens on Earth. So if we can get a compromise, where conventional morality reigns nearby and value maximization reigns in distant star systems, to the Broadtermist, that looks pretty good. After all, Earth is less than 1/10^20 of the affectable universe.
If we mostly cared about what happened on Earth, then we’d find a compromise solution where what happens on Earth is hugely suboptimal to be a horrible catastrophe. To the Broadtermist, it’s vastly less significant. The scale of the space we might develop is so vast and the stakes so giant, that what happens to Earth specifically is of little comparative importance.
It also means that we should be more focused on values spreading. Influencing the values of present humans can affect the trajectory of space settlement. Suppose you’re choosing between animal welfare interventions and global health ones. Global health charities have long-lasting environmental effects that benefit animals potentially for millions of years to come. But this is outweighed by the positive value effects of animal welfare charities, which could influence things across many planets.
Being a Broadtermist also supercharges the case for Longtermism. If almost all of what matters is what will happen on distant planets, and all settlement of distant planets will happen in the far future, then the far future’s importance relative to the present goes up. Probably this won’t be that action relevant—if you take future interests seriously, the numbers are extreme enough that you should already be a Longtermist—but it’s something.
I also find Broadtermism to be pretty intuitive. Odds are non-trivial that at some point, we will have a civilization that spans countless star systems. Once we have that, so that almost 0% of people live on Earth, it seems clear that what will happen outside of Earth will matter a lot more than what happens on Earth.
Send this story to anyone — or drop the embed into a blog post, Substack, Notion page. Every play sends rev-share back to Bentham's Newsletter.
We’ve simplified responses to 👍 / 👎. Past comments are archived but no longer visible.