It’s hard to say what’s crucial from a longtermist point of view, or even from a cluelessness point of view.
A few considerations that might be crucial here:
High inflation might cause populism and political instability globally and this might be bad.
Frontier growth in high-income countries is very important for global development. Countries with cutthroat capitalism like the US have more inequality but also provide more of the global public good of (technological) innovation which grows the world economy and countries (e.g. Scandinavian countries) with cuddly capitalism and less inequality freeride on. See: Can’t We All Be More Like Scandinavians? Asymmetric Growth and Institutions in an Interdependent World. And so from a longtermist perspective, if growth decreases existential risk, then the lost growth from suboptimal macroeconomic policy is bad.
But growth and political stability are also of neartermist concern. Generally, I wouldn’t worry too much about longtermist. In macroeconomics there’s the ‘Tinbergen Rule—a basic principle of effective policy, which states that to achieve n independent policy targets you need at at least n independent policy instruments).’
This is because multi-objective optimization is generally harder than single-objective optimization.[286]
For a similar reason, we might not wish to analyze every grant from a neartermist and longtermist perspective but rather decide on the overall allocation on the portfolio level. In other words, we should focus on the highest possible neartermist impact with our neartermist grant portfolio, which will have some effect on the longterm, but the effect of this will probably be overshadowed by how much we invest in explicitly longtermist cause areas, like x-risk reduction.
Generally, I wouldn’t worry too much about longtermist. In macroeconomics there’s the ‘Tinbergen Rule—a basic principle of effective policy, which states that to achieve n independent policy targets you need at at least n independent policy instruments).’
Why have you categorised monetary policy as a neartermist intervention? Why not a longtermist? There’s a very clear and direct impact on economic growth which will naturally see the vast majority of value/disvalue in the future. For me this puts monetary policy in the longtermist intervention camp.
Other cause areas that focus on saving lives or reducing suffering in the short-term might naturally be seen to be neartermist cause areas because the link to economic growth / x-risk is very unclear. In this case a somewhat credible argument can be made that you can ignore future effects because you have no clear what they are (although some have pushed back against this view).
It seems to me however that judging an intervention that has a clear impact on economic growth based solely on very short-term effects is wholly inadequate. You’re ignoring too much in doing so!
I take your point. I agree that this is important, that’s why I engaged with this topic because it’s quite a large lever. However, ultimately, I see the first order effects of the advocacy on the macroeconomy are perhaps not as important as the second order effects of making suboptimal grants on OpenPhil and the wider EA community (including reputational effects). This because the overall amount of grants in 2021 were perhaps relatively low and likely didn’t cause contribute in a massive way to current inflation. Put simply, I’m not saying that OpenPhil is a major cause of current inflation. Maybe it’s important not to conflate magnitude of effect with whether something is longtermist (if you speed up growth by a tiny amount through macroeconomic policy, then surely it’s less important on the longterm future, than say, curing all disease).
Generally, maybe the difference between longtermist and neartermist interventions is a bit fuzzy and ill-defined at this stage.
It’s hard to say what’s crucial from a longtermist point of view, or even from a cluelessness point of view.
A few considerations that might be crucial here:
High inflation might cause populism and political instability globally and this might be bad.
Frontier growth in high-income countries is very important for global development. Countries with cutthroat capitalism like the US have more inequality but also provide more of the global public good of (technological) innovation which grows the world economy and countries (e.g. Scandinavian countries) with cuddly capitalism and less inequality freeride on. See: Can’t We All Be More Like Scandinavians? Asymmetric Growth and Institutions in an Interdependent World. And so from a longtermist perspective, if growth decreases existential risk, then the lost growth from suboptimal macroeconomic policy is bad.
But growth and political stability are also of neartermist concern. Generally, I wouldn’t worry too much about longtermist. In macroeconomics there’s the ‘Tinbergen Rule—a basic principle of effective policy, which states that to achieve n independent policy targets you need at at least n independent policy instruments).’
In macroeconomics you see that for instance where many economists disagree that central banks should take account of the environmental implications of its policy decisions.
This is because multi-objective optimization is generally harder than single-objective optimization.[286]
For a similar reason, we might not wish to analyze every grant from a neartermist and longtermist perspective but rather decide on the overall allocation on the portfolio level. In other words, we should focus on the highest possible neartermist impact with our neartermist grant portfolio, which will have some effect on the longterm, but the effect of this will probably be overshadowed by how much we invest in explicitly longtermist cause areas, like x-risk reduction.
Why have you categorised monetary policy as a neartermist intervention? Why not a longtermist? There’s a very clear and direct impact on economic growth which will naturally see the vast majority of value/disvalue in the future. For me this puts monetary policy in the longtermist intervention camp.
Other cause areas that focus on saving lives or reducing suffering in the short-term might naturally be seen to be neartermist cause areas because the link to economic growth / x-risk is very unclear. In this case a somewhat credible argument can be made that you can ignore future effects because you have no clear what they are (although some have pushed back against this view).
It seems to me however that judging an intervention that has a clear impact on economic growth based solely on very short-term effects is wholly inadequate. You’re ignoring too much in doing so!
I take your point. I agree that this is important, that’s why I engaged with this topic because it’s quite a large lever. However, ultimately, I see the first order effects of the advocacy on the macroeconomy are perhaps not as important as the second order effects of making suboptimal grants on OpenPhil and the wider EA community (including reputational effects). This because the overall amount of grants in 2021 were perhaps relatively low and likely didn’t cause contribute in a massive way to current inflation. Put simply, I’m not saying that OpenPhil is a major cause of current inflation. Maybe it’s important not to conflate magnitude of effect with whether something is longtermist (if you speed up growth by a tiny amount through macroeconomic policy, then surely it’s less important on the longterm future, than say, curing all disease).
Generally, maybe the difference between longtermist and neartermist interventions is a bit fuzzy and ill-defined at this stage.
A few random thoughts on this:
Maybe the difference is more that of broad vs. narrow interventions
Relatedly, Why Charities Usually Don’t Differ Astronomically in Expected Cost-Effectiveness
At OpenPhil, the macroeconomics program is not in the Longtermist department.