Donations are kind of like taxes. They fund public goods, and generally individuals could get more personal value from keeping their money than by that money going to taxes/donations.

EAs often think about how much people should give (both themselves & others). Government wants to know how much to tax, and what to tax.

Economics has put a lot of thought into what is optimal tax policy. Applying this thought to optimal donation policy is useful. Claude/Gemini/your LLM of choice can probably do this better than me, but let's try anyways.

The rest of the post analyzes particular features of optimal tax policy, and how that applies to donation policy.

I'm going to pretend that we are fairly soul-less money-to-optimal-causes maximizers, because it fits better with simple econ models. Extensions to more complicated economic models are left as an exercise to the reader :)

Efficiency

Governments generally want to maximize tax revenue while minimally distorting people's behavior. If we are taxing for revenue maximization, we especially want to avoid distortions which cause people to produce less revenue. For example, a tax rate of 100% would cause fewer people to work, which would reduce net revenue.

There are important exceptions for negative externalities. For example, the government believes professional gambling is bad for society, it might want to set an unusually high tax rate on gambling winnings, as a way to discourage professional gambling.

How does this apply to how much one should donate?

Well, first we should note that we are in the position of an incredibly weak government. We can propose a tax (donation) policy, but our enforcement tools are very limited. Within the EA community, we can applaud those who follow our policy & shame or ostracize those who don't. The community is fairly porous, so folks can just choose to leave if they don't agree with the policy. So positive reinforcement of rule followers is probably more viable than negative enforcement for shirkers.

OK, so let's say our only tool is praise. Who should we praise?

We want the most bang for our praise buck. If we praise everyone who donates anything, big donors are getting less praise per dollar. So we should praise the largest donors. To reflect on the tax analogy, this is consistent with IRS getting the most value from auditing the richest folks.

Equity

Governments generally want tax policy to be (or seem) fair. Fairness usually amounts proportional taxes (everyone pays the same percent) or progressive taxes (richer folks pay an increasing percent).

For donations, this means we should encourage the richer to donate more. Uncontroversial point, and aligns with using our limited praise budget on them anyways.

Compliance & administrative simplicity

Governments want people to pay their taxes, and want to not have to spend a lot on tax collection. It is better to tax simple, easily visible things (like income) rather than complex or ill defined things (like malcontent).

For donation policy, we want a rule that maximizes donations for minimum effort. Encouraging 10% donations is a common rule of thumb here. This is simple to understand (just take your income, and slide it over one decimal), and a common Schelling point (Christianity, Giving What We Can, etc).

Compliance is interesting. We generally praise folks for saying they donate, and we require no further verification. For small dollar or praise amounts, this seems reasonable (like how the IRS shouldn't audit someone making 50k USD). For large amounts, we should raise our bar. If someone claims to have donated millions of dollars, and we are considering adding them to a celebratory mural of donors, we should probably followup with the charities nominally receiving those funds. For even larger figures (cough FTX cough cough), we should consider doing work to verify the money is legitimate & won't be clawed back.

OK, but who is this 'we' who will do this investigation. Well, it is probably those giving the praise. So you & me. Ah, but who praises the praise givers? Conveniently, we already have the EA Forum karma system for that.

Revenue adequacy & stability

Governments want enough money to function, and they don't want to squeeze companies dry in one year. So the best tax policy is one with high, steady revenues.

For donation policy, the analogous logic would be that we don't want to donate too little (inadequate) or all at once (unstable). For personal donation, this is the impact vs burnout tradeoff. For other donors, this is the problem of free-riders vs the problem of folks leaving due to pressure.

How is donation policy different from tax policy?

Donors (or even EA donors) are a much less well defined group than a country's citizens. This makes enforcing compliance harder (since you don't know who to target).

There is not a centralized authority for enforcing donation policy, in contrast to governments. This makes enforcing compliance harder (since who are 'you' to enforce it?).

The public goods from donations are generally experienced by non-donors, whereas a higher proportion of the benefits of government supplied public goods are experienced by taxpayers. IE, I am unlikely to ever sleep under a bednet but very likely to visit my local library. This makes compliance harder (since you don't feel the benefits as much).

Final take-aways

In encouraging donation norms, imagine what you would tell an extremely weak government.

Probably you focus your limited efforts on the wealthiest.

Probably you encourage a fairly low tax rate, which is set around some Schelling point already agreed upon.

Probably you would spend more time praising those who did pay their taxes, since you don't have the resources to punish shirkers.