Voting is today (Tuesday March 3) in California and other states - here are recommendations
By kbog @ 2020-03-03T10:40 (+23)
Today is the last day to vote in a number of states including California, Massachusetts, Texas, and elsewhere. Voting is a pretty impactful activity in expectation so please find time to do it, and help lots of friends/family get to the polls (unless they have bad opinions, I guess). Here is the Ballotpedia page with details on how/where to vote.
Note: in other states, there will be similar elections coming up soon as well! This guide will still apply.
Democratic presidential primaries
For the presidency, I recommend voting for Joe Biden. This is supported by the comprehensive Candidate Scoring System for welfare maximization. The very short story for Biden is that he is pretty similar to Sanders in terms of overall merit for the job, but is a bit more likely to defeat Donald Trump. The conclusion is robust to most differences in opinion on EA priorities for things like long run vs short run benefits, humans vs animals and so on.
(Yes, last time I posted it, I recommended Bloomberg over Biden. Several things have changed. Biden became a bigger frontrunner as Bloomberg fell, I took into account the dynamics of a contested convention and corresponding legitimacy issues, I lost confidence in the expectation that Bloomberg would support animal welfare due to the greater evidence that he's a stubborn political centrist, Biden released a decent plan for reducing housing restrictions, and he posted this article on foreign policy.)
Max Ghenis has made a program for determining who is or isn't viable in your particular state/district, but Biden's probably viable pretty much everywhere, and giving him a higher popular vote is still helpful for legitimacy, so I don't think you need to worry about this.
Congressional primaries
There are many congressional primary elections. The Candidate Scoring System drafts for Senate and House of Representatives include a little guidance on the primaries. A good rule of thumb is that more moderate candidates tend be more electable, and we should nominate more electable Democrats who will beat Republicans. Incumbents are probably more electable as well, so it's good to keep them safe in their seats from primary challengers.
It seems to me that aside from the question of electability, very left wing Democrats are not systematically better or worse than moderate Democrats, but moderate Republicans are better than very right wing Republicans. However, there are exceptions for the Democrats. Some moderate Democratic politicians - especially men and especially in rural districts - can be particularly worse for animal welfare compared to leftists. And some leftist Democrats (such as Jamaal Bowman) can be particularly worse for foreign policy due to being very non-interventionist, anti-military and less serious about foreign aid. Female Republican politicians also seem to be better than male Republicans for animal welfare.
Local politics
In San Francisco, I'm not personally familiar with the specific elections, but you could probably do much worse than listening to Max Ghenis. Maybe other EAs have other recommendations for San Francisco, feel free to share in the comments.
Here are my recommendations for Los Angeles and Glendale races.
For the rest of the country, here is some general guidance.
For local city politics, the most important issue is relaxing zoning restrictions against dense housing. This is an issue which EAs (and policy wonks more broadly) seem to broadly agree on. If you want to see the evidence and arguments which justify it, see the "Housing" section of Candidate Scoring System.
Look at local candidates for how they express their commitment to allowing more housing. If they criticize "luxury housing" or "overdevelopment" then that's a red flag; in practice it often means that they are opposed to any new dense housing at all. Note that this is not a Democratic or Republican issue; candidates from either party can be good or bad here. See if you can find a local "YIMBY" group which recommends candidates and measures for better housing. And in every single ballot measure about urban development, it's probably best to pick whichever option allows more construction.
More/better mass transit seems to be a good idea. I haven't personally researched this, but it's widely supported among urbanist, YIMBY, and Neoliberal types who are pretty trustworthy on this kind of thing, and it makes plenty of common sense.
Local politics can also include issues of criminal justice. Generally speaking, US government tends to err on the side of being too "tough-on-crime" and putting too many people in prison; this is a consensus among social scientists and the Effective Altruist Open Philanthropy project, and this consensus appears to be reasonably robust as opposed to being the product of some kind of bias. But note that while harsh sentencing rules are bad, it can actually be good to hire more police officers.
One could argue that a few soft-on-crime politicians in very liberal cities, like Chesa Boudin, are going too far. I don't know enough about the matter to answer this. In more moderate and conservative areas at least, it is right to consistently press for softer policies. For this purpose, it is typically better to vote for Democrats rather than Republicans.
Caution regarding coronavirus
Exercise caution in the mass gatherings around voting stations; wash your hands properly and don't touch your face.
Aaron Gertler @ 2020-03-09T21:08 (+3)
Some moderate Democratic politicians - especially men and especially in rural districts - can be particularly worse for animal welfare compared to leftists [...]Female Republican politicians also seem to be better than male Republicans for animal welfare.
Why is it worth adding analysis of whether male or female politicians tend to be better on certain issues? If you're only voting on a few races at a time, shouldn't you just look directly at candidates' policies, rather than trying to guess at their views based on other characteristics?
The explanation that comes to mind for me is that a voter who doesn't have time to look at policy can make a better guess about which candidate to vote for based on gender. But it still seems like a weak signal at best, and a poor way to go about voting if you do expect your vote to be "impactful in expectation". Was that what you had in mind?
kbog @ 2020-03-23T20:46 (+2)
New candidates have never served in Congress and therefore do not have legislative track records on animal welfare, and it's such a minor issue to most voters that candidates almost never express their views on it while running for office.