A simple idea for making donation matching counterfactual
Donation matching is often dismissed in EA because the counterfactual impact is unclear. The matching funder would typically have donated anyway, so the match isn’t causing any counterfactual donations. But what if we could design matching programs where counterfactual impact is clearer?
The basic scheme: a funder locks crypto in a smart contract that releases the funds to a specific charity only if a match is found within some timeframe. If no match is found, the contract sends the funds to a null address, permanently destroying them.[1] This creates a credible commitment where the charity receives funds if and only if a matcher steps in. So the matcher can be confident that their donation actually caused the charity to receive the funder’s contribution.
When would this beat a direct donation? Assuming risk-neutrality, the expected value exceeds a direct donation if the probability of finding a match, multiplied by the total funds received (the original commitment plus counterfactually new money from the matcher), is greater than the committed amount alone. Thus, for a 1:1 match where the matcher is fully counterfactual, the probability of finding a match must exceed 50%.
There might be potential negative second-order effects though. In my proposed scheme, the funder is essentially holding charitable value hostage to extract more donations, which could be seen as adversarial behavior. Overall though, I’d be excited for some funder to try this approach.
A simple idea for making donation matching counterfactual
A simple idea for making donation matching counterfactual
Donation matching is often dismissed in EA because the counterfactual impact is unclear. The matching funder would typically have donated anyway, so the match isn’t causing any counterfactual donations. But what if we could design matching programs where counterfactual impact is clearer?
The basic scheme: a funder locks crypto in a smart contract that releases the funds to a specific charity only if a match is found within some timeframe. If no match is found, the contract sends the funds to a null address, permanently destroying them.[1] This creates a credible commitment where the charity receives funds if and only if a matcher steps in. So the matcher can be confident that their donation actually caused the charity to receive the funder’s contribution.
When would this beat a direct donation? Assuming risk-neutrality, the expected value exceeds a direct donation if the probability of finding a match, multiplied by the total funds received (the original commitment plus counterfactually new money from the matcher), is greater than the committed amount alone. Thus, for a 1:1 match where the matcher is fully counterfactual, the probability of finding a match must exceed 50%.
There might be potential negative second-order effects though. In my proposed scheme, the funder is essentially holding charitable value hostage to extract more donations, which could be seen as adversarial behavior. Overall though, I’d be excited for some funder to try this approach.
Other options are also possible, such as anti-charities.