An EA-specific agency would have to be low-bono, offering major discounts to EA orgs—otherwise it would be indistinguishable from the countless existing for-profit agencies.
This is building a bit on what Sanjay said, but I think this sentence deserves more highlighting.
If the only advantage of an EA agency is that it has lower prices, then it’s equivalent to donating more money to the organization which hires the developers.
Donating more money has a lot of advantages over building an entire new agency: we don’t have to go through all the hassle of identifying talent, structuring contracts, etc. (Indeed, the benefit of avoiding all that hassle is why contract agencies exist.)
An adjacent project I would be excited about is someone creating a vetted list of non-EA agencies. This seems to provide many of the benefits of an EA agency, without some of the costs.
An EA-specific agency would have to be low-bono, offering major discounts to EA orgs—otherwise it would be indistinguishable from the countless existing for-profit agencies.
I guess I overstated this. There would be some advantages to a bunch of EA developers working together.
If the only advantage of an EA agency is that it has lower prices, then it’s equivalent to donating more money to the organization which hires the developers.
This isn’t true if you’re getting taxed VAT on the price difference.
Let’s suppose there’s a way around that, though. As you say in the sentence after this, changing the dynamics of money isn’t necessarily neutral, even if it comes from and ends up with the same groups of people. There are the costs you mentioned of an agency, and there are costs of centralisation: eg concentrating risk, diluting focus, and encouraging intra-org groupthink and inter-org siloisation.
What you mention as costs of an agency are mostly (perhaps all) fixed costs, so you would have to believe that it would offer close to 0 or negative counterfactual benefit for them not to be worth paying, especially given the extra value of information from trying. Either a low-bono or full-profit agency would have those discussed in part 2, so which of those would be better depends on what you think of centralisation (and whether you can avoid the extra VAT).
Meanwhile, I think a donor-funded agency has very substantial counterfactual benefits, as listed here.
Thanks for writing this up!
This is building a bit on what Sanjay said, but I think this sentence deserves more highlighting.
If the only advantage of an EA agency is that it has lower prices, then it’s equivalent to donating more money to the organization which hires the developers.
Donating more money has a lot of advantages over building an entire new agency: we don’t have to go through all the hassle of identifying talent, structuring contracts, etc. (Indeed, the benefit of avoiding all that hassle is why contract agencies exist.)
An adjacent project I would be excited about is someone creating a vetted list of non-EA agencies. This seems to provide many of the benefits of an EA agency, without some of the costs.
I guess I overstated this. There would be some advantages to a bunch of EA developers working together.
This isn’t true if you’re getting taxed VAT on the price difference.
Let’s suppose there’s a way around that, though. As you say in the sentence after this, changing the dynamics of money isn’t necessarily neutral, even if it comes from and ends up with the same groups of people. There are the costs you mentioned of an agency, and there are costs of centralisation: eg concentrating risk, diluting focus, and encouraging intra-org groupthink and inter-org siloisation.
What you mention as costs of an agency are mostly (perhaps all) fixed costs, so you would have to believe that it would offer close to 0 or negative counterfactual benefit for them not to be worth paying, especially given the extra value of information from trying. Either a low-bono or full-profit agency would have those discussed in part 2, so which of those would be better depends on what you think of centralisation (and whether you can avoid the extra VAT).
Meanwhile, I think a donor-funded agency has very substantial counterfactual benefits, as listed here.