Resource neutrality and levels vs kinds of demandingness
Caroline Ellison recently argued that effective altruist debates on demandingness often focus exclusively on frugality: how much to sacrifice in terms of money or material resources. She argues that there are many ways in which effective altruism could be demanding: e.g. in terms of time (how hard to work), epistemics (how hard to fight motivated reasoning), or status (willingness to take a less prestigious but higher-impact job). The notion that demandingness is solely about money conflates two issues, in her view: how demanding EA should be, and how EA should value different resources (e.g. time and money) relative to each other.
I agree with this analysis. Effective altruists have long argued that we should be cause neutral: that we shouldn’t prejudge the issue of what cause (e.g. local homelessness or global poverty or AI safety) to invest in, but should instead choose based on impartial assessments of impact. Similarly, effective altruists should be resource neutral. When we decide to allocate resources to others, we shouldn’t prejudge the issue of whether to give away money, time, or other resources. Instead, our choice should be based on impartial assessments of impact.
When effective altruism was started, there wasn’t a lot of money dedicated to effective altruist causes. This has changed drastically in recent years, with the growth of funding from, e.g. Open Philanthropy and FTX Future Fund. As a result, effective altruism now has a relatively lower need for more funding compared with dedicated direct work. That plausibly means that effective altruist demandingness should shift away from a focus on money towards a focus on time. If we can buy more time with money, we should to an extent do that.*
Should these developments also change how demanding effective altruism is in general — regarding all resources (time, money, etc) put together? The title of Ellison’s article is “Demandingness and Time/Money Tradeoffs are Orthogonal”. That might suggest they should not. According to this analysis, what level of demandingness to have is a completely separate issue from what kinds of demandingness to emphasise:
Level of demandingness: how many resources (e.g. time or money) to give to others
Kinds of demandingness: what kinds of resources (e.g. time or money) to give to others
I agree that these two questions are conceptually separate. A particular answer to the first question doesn’t in itself logically entail any particular answer to the second question, nor vice versa. But that doesn’t mean that they’re empirically unrelated. It’s entirely possible that a greater need for time than for money should cause the overall level of demandingness to go down.
Effective altruist organisations are now paying substantially higher salaries than they did initially. My understanding is that that increases their impact, since it enables them to recruit more qualified staff. Simultaneously, that makes effective altruism less demanding in terms of money.
And I don’t think that’s offset (at least not fully offset) by an increased demandingness in terms of time. One reason is that it’s hard to increase the number of hours much beyond a certain point without getting burnt out. Another reason is that such expectations would partly undo the recruitment boost that you get from higher salaries.
For these and related reasons I would guess that effective altruism has actually become less demanding as a result of the influx of money and the consequent switch to dedicated direct work becoming the key bottleneck in effective altruism. But that is a broad generalisation that depends both on the kind of job you have and your individual psychology (how much you value, e.g. time vs money). Some people may well view today’s effective altruism more demanding than that of ten years ago. In any event, that, too, is compatible with my main claim that the overall level of demandingness is influenced by what kinds of demandingness effective altruism emphasises. Resource neutrality does not entail that the optimal overall level of demandingness is fixed.
* Obviously it’s a tricky empirical issue to determine the precise extent to which we should do that, but that’s beyond the scope of this post, which is more conceptual.