3 ms·
This is true, but the problem is that you rarely know if political advocacy will work until after it succeeds, and even afterwards, it may take a while to study
by skybrian 6y ago
This is true, but the problem is that you rarely know if political advocacy will work until after it succeeds, and even afterwards, it may take a while to study the impact of the change.
GiveWell did study some political advocacy groups, but apparently it’s just hard to know what to recommend with confidence. There is a spin-off organization (Open Philanthropy) that uses funds from rich philanthropists to do more experimental work.
- seoaeu 6y agoYou don't need to know whether a specific advocacy effort will work, only to estimate whether the overall expected value is high enough. And we know that when advocacy succeeds it can have upsides many many times the cost put in (and there aren't that many advocacy movements) so there is at least some hint that the expected value is large. It feels kind of like a cop-out to say that you shouldn't pursue political change because you can't quantify how much good you'll do.
- skybrian 6y agoSure, you can guess at expected values, but what makes that guess interesting enough to share with others as a recommendation? Why would GiveWell's guesses about political causes be more interesting to read than anyone else's? GiveWell's job is attempting to make predictions about fairly definite questions that can be answered by doing a lot of homework. They can learn a lot about charities and read scientific studies and make predictions that they can stand behind. It's figuring out what the safe choices are, not guessing at speculative winners. But maybe a prediction market could do what you want?
- lmm 6y agoWe also know that advocacy movements can have downsides many times the costs put in. Estimating the expected value of political advocacy is a good idea, but I would expect the result to be negative.