Great post! Here’s another possible counter-point: The traditional EA interventions have been easy to quantify: bed nets, cash transfers, deworming, online-ads, leaflets, etc. As we get better at evaluating interventions we tend more towards harder-to-quantify stuff such as influencing politics. What makes the former interventions easy to quantify? One attribute is the fact that they consist of small things bought in large quantities. These are easy to study with RCTs. Running RCTs on areas where salaries are the thing to be funded is impractical.
So if the trend away from easy-to-quantify areas continues we can expect to put more of our money into salaries. This yields two reasons we may need more direct work and fewer EtG: 1) Hiring people is a lot less scalable which means less money is needed per intervention, 2) We may have to create new positions and fill them with EAs (e.g. what xrisk orgs do) or we may have to fill areas with EAs (e.g. politics).
Thanks! To be fair, I do feel like my argument takes into account this counter-point pretty fully, especially the section “The Problem With Funding Salaries”. But you’re right that the more we fund salaries, the weaker this argument becomes.
we may have to fill areas with EAs (e.g. politics).
Based on how each subsequent election cycle seems to be more expensive than the last in, e.g., the United States and the United Kingdom, I’m terrified by the possibility by how much it would cost those earning to give to fund a campaign by themselves. Like, thinking about how many lives that money could counterfactually save, and there isn’t even a guarantee an EA-funded candidate would get elected. Depending on how serious EAs interested in politics are, we better figure out how raise funds from outside effective altruism. and run successful campaigns before one of us starts running. With its connections to other researchers who could help on such a project, and their current research experience with normative rationality, evidence-based decision-making, and counterfactual reasoning, 80,000 Hours seems to be best poised to carry out this research among EA orgs.
Interesting, I hadn’t thought of the possibility to use EtG money to fund campaigns rather than just having EAs raise the money as other politicians do.
Also, as we get better at measuring things we might open up new giving opportunities, such as those currently being looked at by The Open Philanthropy Project.
Great post! Here’s another possible counter-point: The traditional EA interventions have been easy to quantify: bed nets, cash transfers, deworming, online-ads, leaflets, etc. As we get better at evaluating interventions we tend more towards harder-to-quantify stuff such as influencing politics. What makes the former interventions easy to quantify? One attribute is the fact that they consist of small things bought in large quantities. These are easy to study with RCTs. Running RCTs on areas where salaries are the thing to be funded is impractical.
So if the trend away from easy-to-quantify areas continues we can expect to put more of our money into salaries. This yields two reasons we may need more direct work and fewer EtG: 1) Hiring people is a lot less scalable which means less money is needed per intervention, 2) We may have to create new positions and fill them with EAs (e.g. what xrisk orgs do) or we may have to fill areas with EAs (e.g. politics).
Thanks! To be fair, I do feel like my argument takes into account this counter-point pretty fully, especially the section “The Problem With Funding Salaries”. But you’re right that the more we fund salaries, the weaker this argument becomes.
True, I totally overlooked that - I shouldn’t write comments when I’m sleepy ;)
Based on how each subsequent election cycle seems to be more expensive than the last in, e.g., the United States and the United Kingdom, I’m terrified by the possibility by how much it would cost those earning to give to fund a campaign by themselves. Like, thinking about how many lives that money could counterfactually save, and there isn’t even a guarantee an EA-funded candidate would get elected. Depending on how serious EAs interested in politics are, we better figure out how raise funds from outside effective altruism. and run successful campaigns before one of us starts running. With its connections to other researchers who could help on such a project, and their current research experience with normative rationality, evidence-based decision-making, and counterfactual reasoning, 80,000 Hours seems to be best poised to carry out this research among EA orgs.
Interesting, I hadn’t thought of the possibility to use EtG money to fund campaigns rather than just having EAs raise the money as other politicians do.
Also, as we get better at measuring things we might open up new giving opportunities, such as those currently being looked at by The Open Philanthropy Project.