If voting is serious business, we need to treat it as such.
Right before the US 2020 election, Gelman argues that PA voters have a 1 in 8.8 million chance of breaking a tie. TX was 1 in 100 million. DC 1 in 240 trillion.
Showing some votes have high expected utility means showing those same votes can have high expected disutility.
It’s weird that Wilbin and MacAskill will be like, “Hey, careful! Before you donate $50, make sure you are doing good rather than wasting the money or worse, harming people. We are beset by biases that make us donate badly and we need to be careful.” But then when it comes to voting, they often advise people to just vote, or to guesstimate effects, when in fact the empirical work shows that are much more biased and terrible at judging politics than almost anything else.
Most people do not know enough to vote well, and voting well is hard. Believing it is easy is itself evidence of bias—that’s what the political psych shows. (Partisans downplay difficulty and think they are obviously right.) So if some people’s votes matter, rather than advising them to vote, period, we should advise them to be good EAs and be very careful about their votes.
I answered this before and it didn’t post. I’ll try again.
If voting matters, we have to treat it like matters.
EAs warn people, “Don’t just donate $500! Be careful. Learn what works and what doesn’t. Make sure you give to an effective charity rather than an ineffective or harmful one. Be aware that you are biased to make bad choices!”
But all that applies to voting. If voting can be like donating $50,000, it can also be like robbing a charity of $50,000. But oddly I see EAs telling everyone to vote and telling them to guesstimate, even though our evidence is that people are much worse at judging politics than charities, and even though guestimating a presidential candidate is orders of magnitude more difficult than judging a charity.
What do you make of Rob Wiblin’s post on the value of voting—https://80000hours.org/articles/is-voting-important/
If voting is serious business, we need to treat it as such.
Right before the US 2020 election, Gelman argues that PA voters have a 1 in 8.8 million chance of breaking a tie. TX was 1 in 100 million. DC 1 in 240 trillion.
Showing some votes have high expected utility means showing those same votes can have high expected disutility.
It’s weird that Wilbin and MacAskill will be like, “Hey, careful! Before you donate $50, make sure you are doing good rather than wasting the money or worse, harming people. We are beset by biases that make us donate badly and we need to be careful.” But then when it comes to voting, they often advise people to just vote, or to guesstimate effects, when in fact the empirical work shows that are much more biased and terrible at judging politics than almost anything else.
Most people do not know enough to vote well, and voting well is hard. Believing it is easy is itself evidence of bias—that’s what the political psych shows. (Partisans downplay difficulty and think they are obviously right.) So if some people’s votes matter, rather than advising them to vote, period, we should advise them to be good EAs and be very careful about their votes.
I answered this before and it didn’t post. I’ll try again.
If voting matters, we have to treat it like matters.
EAs warn people, “Don’t just donate $500! Be careful. Learn what works and what doesn’t. Make sure you give to an effective charity rather than an ineffective or harmful one. Be aware that you are biased to make bad choices!”
But all that applies to voting. If voting can be like donating $50,000, it can also be like robbing a charity of $50,000. But oddly I see EAs telling everyone to vote and telling them to guesstimate, even though our evidence is that people are much worse at judging politics than charities, and even though guestimating a presidential candidate is orders of magnitude more difficult than judging a charity.
Is it sufficient for it to be good to vote for EAs to be better than the median voter? (which I think is probably true.)