Candidate Scoring System, Second Release

post by kbog · 2019-03-19T05:41:20.022Z · EA · GW · 27 comments

Contents

  Preface
  Summary for Voters and Activists
None
27 comments

This is the significantly improved second edition of CSS, again looking at the presidential primaries for the US 2020 election.

PDF report

Excel spreadsheet

Here I copy the introductory sections for convenience:

Preface

The Candidate Scoring System (CSS) is a method for selecting preferred candidates in elections. It is based on Effective Altruist (EA) ethics and methodology. Of course, opposing political positions are still valid in the EA movement and there is room for respectable disagreement. Other people in the EA movement may have different understandings of the factual impacts of various political actions, and they may have different values regarding the appropriate goals of government. But we approach the central, most important policy question – how to maximize global well-being – by gathering opinions and research from authorities in a wide range of domains, then modeling them together with our own careful judgment to fill in the gaps.

CSS1 was released on March 5, 2019, establishing basic policy preferences and providing tentative scoring of presidential candidates. For CSS2 we have deepened our analysis of policy questions, gathered more information about political candidates, expanded the number of candidates under consideration, added calculations of election probabilities and counterfactuals, and simplified the information into a single report with an accompanying Excel model.

This project is limited by the constraints of time and manpower against the vast breadth, depth and complexity of the problems that it tackles. Therefore, many arguments and evidence will be missing. This does not mean the project is necessarily wrong or biased, it just means we haven’t yet included as much content and research as we would like to. It is a work in progress and open to input from others. We are uncertain about much of this content, but we minimize hedging language for the sake of readability. If some relevant information is missing, please submit ideas and content to improve the next version – everything here is subject to revision and elaboration.

CSS is an independent volunteer project.

Summary for Voters and Activists

CSS2 makes the following recommendations:

· John Delaney should be supported if there are tractable opportunities to do so, particularly in Iowa.

· Cory Booker should be supported if Delaney’s candidacy is considered intractable.

· Potential Republican challengers to President Trump should be encouraged and supported if a real chance appears, especially John Kasich.

Our recommendations are based on estimates of the expected value of changing the outcomes of the primary races. We approach this question by first estimating the desirability of each candidate as a potential president, yielding presidency scores. We then factor in the nomination and election chances of all their competitors to produce nomination scores representing the difference in the expected election outcome when the candidate wins or loses in the primaries.

27 comments

Comments sorted by top scores.

comment by andrewpei · 2019-03-20T20:08:39.891Z · EA(p) · GW(p)

Fantastic report! I love this type of content and can't wait to sink my teeth into it

comment by MaxGhenis · 2019-03-24T05:59:19.591Z · EA(p) · GW(p)

Incredible report, bravo! Like probably anyone, I don't agree completely with the ratings, but the structure and research helped me think through my own priorities. I was already interested in supporting Delaney, so this motivates me to ask more people to give him a donation to get on the debate stage.

I have some minor suggestions, which I left in this copy.

Beyond that, my only non-minor suggestion is to consider mentioning domestic poverty as a (potential) priority area, even if it ends up not included due to the thresholds. Depending on the poverty line, US poverty contributes to some premature deaths, though I haven't researched what level would be associated with 100,000 per year. Better-designed antipoverty programs could also improve GWP through improved incentives (especially, I'm guessing, with respect to SSDI), though this could be slight.

CSS2 discusses antipoverty programs like UBI and EITC in the budgeting section, though with a different aim. Yang's UBI isn't fully funded (I've estimated it'd add $1.5T to the annual deficit in static simulation), and other antipoverty proposals like Harris's LIFT Act also don't include funding proposals, but I'd consider emphasis to predict antipoverty action.

Another antipoverty bill considered effective by economists is the American Family Act, which is essentially a child dividend. Cash transfers to families with children improve kids' long-term outcomes (Vox). All 2020 candidates in Congress are cosponsors, and all except Sanders cosponsored its predecessor in the last session. Columbia and Vox have summarized the core antipoverty bills.

Thank you for your stellar work.

comment by kbog · 2019-03-24T06:58:44.564Z · EA(p) · GW(p)

Thanks for giving such detailed feedback.

I am now leaning towards separating cash transfers/antipoverty programs away from taxation. When I next put major time into this (I'm not currently, actually) I plan to do that.

I'm always looking for other people's ratings, depending on the nature of the disagreement I can compromise between multiple ratings for better accuracy.

comment by Davis_Kingsley · 2019-03-24T01:19:54.027Z · EA(p) · GW(p)

I like that you've put the effort into creating this, but I'm not fond of the background assumptions here - there seem to be some elements that not all EAs might necessarily share. For instance, one section begins "Intrinsic moral rights do not exist" - that's certainly not what I believe and it seems inconsistent with other sections that talk about the "intrinsic moral weight" of animal populations, etc.

While the fact that you've "shown your work" with the Excel spreadsheet helps people evaluate the same issues with different weights, if someone is interested in areas that you've chosen to exclude it's less apparent how to proceed.

I do appreciate the work you've put into this, though!

comment by kbog · 2019-03-24T01:53:22.075Z · EA(p) · GW(p)
For instance, one section begins "Intrinsic moral rights do not exist" - that's certainly not what I believe and it seems inconsistent with other sections that talk about the "intrinsic moral weight" of animal populations, etc.

It's definitely consistent - animals can have interests without having rights, just like humans.

Rights can point in a bunch of different ways depending on the moral inclinations of the reader. And integrating and applying them to policy is a very murky issue. So even if I wanted to investigate that side of things, I would have little ability to provide useful judgments to EAs.

At some point, it would be nice to include full arguments about morality. But that's pretty low on my priorities, I don't expect to add it in the foreseeable future. Those arguments already exist elsewhere.

While the fact that you've "shown your work" with the Excel spreadsheet helps people evaluate the same issues with different weights, if someone is interested in areas that you've chosen to exclude it's less apparent how to proceed.

You can add a column besides the other topics, then insert a new row into the weight table (select three adjacent cells and press insert...). True it's a little complicated - but I have to make the spreadsheet this way in order to make the sensitivity analysis work well.

comment by Davis_Kingsley · 2019-03-24T02:19:39.824Z · EA(p) · GW(p)

I don't think there's much practical difference between "intrinsic moral interests" and "intrinsic moral rights", but that's not really the point - it's more that I think given such differences in perspective between EAs, I'm not sure that documents like this are great for EA as a movement. I would at least prefer to see them presented less... authoritatively?

comment by kbog · 2019-03-24T03:39:10.951Z · EA(p) · GW(p)

OK fine, in CSS3 it now simply says " Absolutist arguments for or against abortion disappear once we focus on well-being. "

comment by Davis_Kingsley · 2019-03-24T09:04:19.453Z · EA(p) · GW(p)

Like I said, that's not really the point - it also doesn't meaningfully resolve that particular issue, because of course the whole dispute is whose well-being counts, with anti-abortion advocates claiming that human fetuses count and pro-abortion people claiming that human fetuses don't.

I dunno, maybe I'm overly cautious, but I'm not fond of someone publishing a well-made and official-looking "based on EA principles, here's who to vote for" document, since "EA principles" quite vary - I think if EA becomes seen as politically aligned (with either major US party) that constitutes a huge constraint on our movement's potential.

comment by kbog · 2019-03-24T13:30:03.031Z · EA(p) · GW(p)

You said the problem was stating it authoritatively rather than the actual conclusions, I made it sound less authoritative but now you're saying that the actual conclusions matter. The document has sufficient disclaimers as it is, I mean the preface clearly says EAs could disagree. You don't see Givewell writing "assuming that poverty is the #1 cause area, which EAs may disagree on" multiple times and I don't treat politics with special reverence as if different rules should apply. I think there's something unhealthy and self-reinforcing about tiptoeing around like that. The point here is to advertise a better set of implicit norms, so that maybe people (inside and outside EA) can finally treat political policy as just another question to answer rather than playing meta-games.

the whole dispute is whose well-being counts, with anti-abortion advocates claiming that human fetuses count and pro-abortion people claiming that human fetuses don't.

If I care about total well-being, then of course people who say that some people's well being doesn't count are going to be wrong. This includes the pro lifers, who care about the future well being of a particular fetus but not the future well being of any potential child (or not as much, at least).

comment by Davis_Kingsley · 2019-03-24T14:24:34.458Z · EA(p) · GW(p)
You said the problem was stating it authoritatively rather than the actual conclusions, I made it sound less authoritative but now you're saying that the actual conclusions matter.

Sorry, I perhaps wasn't specific enough in my original reply. The "less authoritative" thing was meant to apply to the entire document, not just this one section - that's why I also said I wasn't sure documents like this are good for EA as a movement.

I think there's something unhealthy and self-reinforcing about tiptoeing around like that. The point here is to advertise a better set of implicit norms, so that maybe people (inside and outside EA) can finally treat political policy as just another question to answer rather than playing meta-games.

Strong disagree. Political policy in practice isn't "just another question to answer" - maybe it should be, but that's not the world we live in - and acting as if it is strikes me as risky.

comment by kbog · 2019-03-24T15:21:52.763Z · EA(p) · GW(p)
The "less authoritative" thing was meant to apply to the entire document, not just this one section.

In the preface I state that hedging language is minimized for the sake of readability.

Political policy in practice isn't "just another question to answer".

Neither is poverty alleviation or veganism or anything else in practice.

comment by Davis_Kingsley · 2019-03-25T01:38:46.363Z · EA(p) · GW(p)
Neither is poverty alleviation or veganism or anything else in practice.

Again, strong disagree - many things are not politicized and can be answered more directly. One of the main strengths of EA, in my view, is that it isn't just another culture war position (yet?) - consider Robin Hanson's points on "pulling the rope sideways".

comment by kbog · 2019-04-03T03:46:01.533Z · EA(p) · GW(p)
Again, strong disagree - many things are not politicized and can be answered more directly.

I think I'm losing track of the point. What does it mean to answer something "more directly"?

consider Robin Hanson's points on "pulling the rope sideways".

I'm not sure how that's relevant here since I'm clearly saying that we're not taking a position on abortion.

comment by Davis_Kingsley · 2019-04-03T11:45:18.860Z · EA(p) · GW(p)

Just posting to acknowledge that I've seen this - my full reply will be long enough that I'm probably going to make it a separate post.

comment by KeynesYouDigIt · 2019-03-22T19:40:04.146Z · EA(p) · GW(p)

Totally a choosing beggar here but I'd love a TLDR on each major candidate, where they were hightest and lowest. Why is ORourke wo low?

comment by kbog · 2019-03-22T21:58:41.344Z · EA(p) · GW(p)

I have written basic TLDRs for the presidency scores on page 74. Though, all I wrote for Beto was:

O’Rourke is inexperienced and has not supported animals as well as some other candidates.

He seemed pretty average in other ways.

I'll point to this section more clearly in the next version.

comment by Milan_Griffes · 2019-03-22T20:29:16.100Z · EA(p) · GW(p)

+1

comment by Bluefalcon · 2019-03-21T17:58:57.271Z · EA(p) · GW(p)

Delaney's hitting 2% on Predictit for the first time AFAIK. Did your quasi-endorsement move the markets?

comment by kbog · 2019-03-21T18:31:58.753Z · EA(p) · GW(p)

Heh, well at the time of this report he announced that he was giving $2 to charity for every donation he got, to try to qualify for the debates. So maybe it was that. (He still needs them - it would be useful to donate $1 to his campaign if you have a moment.)

comment by michaelchen · 2019-03-22T01:58:28.562Z · EA(p) · GW(p)

Just donated! For others' convenience, the link is https://go.johndelaney.com/page/content/this-is-about-america/.

comment by Milan_Griffes · 2019-03-22T20:19:47.414Z · EA(p) · GW(p)

Whoa really?

Could you link to the source of this so I could read about the specifics of his promise?

comment by Milan_Griffes · 2019-03-22T20:28:56.441Z · EA(p) · GW(p)

Oh, I see. It's not a 2:1 match, it's $2 flat for each new donor. So donating $1 is sorta gaming Delaney's intention (though adhering to the letter of the deal).

comment by kbog · 2019-03-22T22:01:06.075Z · EA(p) · GW(p)

Well, he really is trying to get people to make $1 donations. He's a pretty wealthy guy but he needs 65,000 individual donors in order to be allowed into the debates.

comment by Milan_Griffes · 2019-03-23T04:12:17.816Z · EA(p) · GW(p)

Donated.

comment by Milan_Griffes · 2019-03-22T22:26:24.377Z · EA(p) · GW(p)

Got it.

comment by Taymon · 2019-03-20T17:27:47.298Z · EA(p) · GW(p)

Links aren't working.

comment by kbog · 2019-03-21T07:04:24.963Z · EA(p) · GW(p)

Odd. Perhaps your browser does not accommodate OneDrive? I have Firefox; I have opened it in private mode (not logged in) and I can access them. Other people have also accessed them.

Issues may be caused by x32 Firefox or Kaspersky Password Manager: https://support.mozilla.org/en-US/questions/1115599

You can PM me your email address, and I will email the documents to you.