I'm currently planning to resolve this NO. Final opportunity for counterarguments.

I think the two best arguments provided so far in favor of YES are:

1. Scott believes that white people are on average more intelligent than black people. Scott believes that important decisions should be made by more intelligent people. Scott believes that people being asked to make important decisions should be compensated highly for it. Therefore, Scott would prefer for disproportionate power and money to be given to white people over black people.

   I don't think this counts because Scott's belief is that intelligence is only moderately correlated with race, so it's not an effective proxy measure for race. Also, by this metric, anyone who believes something like "systemic racism has led to black people scoring worse on intelligence tests for environmental reasons, and the way to fix this is to hire a lot of dedicated college-educated antiracists to positions of power" would also be a white supremacist.

2. Scott would like people to be more willing to engage in object-level discussion with white supremacists about their beliefs. This is normalizing white supremacy, by removing the stigma around it.

   I think this argument is correct. However, it's not what I had in mind when I wrote the description, and I think the market title makes that clear. I meant "normalize X" as in "trying to get more people to adopt or consider X over other alternatives", not the much more general "people should be more open to considering a huge class of ideas, which happens to include X". I think the market title makes it clear that this isn't what I meant. Since the title contradicts the description, I fall back on my policy of resolving based on the "spirit of the market", and I think the spirit of this market pretty obviously doesn't include that.

Also, by this metric, anyone who believes something like "systemic racism has led to black people scoring worse on intelligence tests for environmental reasons, and the way to fix this is to hire a lot of dedicated college-educated antiracists to positions of power" would also be a white supremacist.

Sometimes people on the far left say stuff like "liberals ruin everything", and I think one instance of what this might be meant to communicate is that hiring lots of college-educated antiracists to positions of power in order to solve racism is bad. My model would not be surprised if they had a model wherein this is white supremacy.

I should probably give my model of their model:

Liberal elites are mostly white, and they gain their authority from education and politics. Within politics, rather than solving real problems that disadvantaged people face, they mainly attack conservatives (so that e.g. they can get reelected or gain political power). If e.g. the police do something illegal and immoral, and then protect each other from accountability, liberal elites will at best diagnose this as being due to a lack of education and progressivism, and will be averse to admitting that the police are criminal and fradulent, and to create structural changes (e.g. separate institutions to hold the police accountable and respecting the rights of the people the police are offending against) in the government that prevent these sorts of things in the future.

While liberal elites would mostly allocate status based on education and support of progressivism, ordinary black people would mostly allocate status based on addressing real problems they face. So the choice of having educated whites address it disregards black people's priorities (solving real problems rather than making the Democrats win), mainly for the purpose of making white liberals more powerful. White supremacy wrapped in an aesthetic of antiracism.

My model may be completely wrong btw. Maybe someone on the far left can clarify.

@tailcalled In my experience the "liberals ruin everything" critique is usually aimed towards centrist "classical liberals", such as those who control the DNC and consistently get elected to political office. The frustration is over the lack of support for more radical leftism in politics, such as of Bernie and AOC.

@IsaacKing belated: seems reasonable to me to use the title to disambiguate the meaning of "normalize" in the description, but I think it would have been good to clarify earlier.

Which probably means I should go read commentary on

/MartinRandall/did-donald-trump-argue-for-white-su

and see if I need to clarify anything.

The market on David Duke has resolved to 7 years.

https://manifold.markets/tailcalled/for-how-many-years-has-david-duke-b

Critical examples of white supremacy that David Duke has engaged in include:

• Founding a branch of the KKK

• Leading a branch of the KKK

• Saying that white people don't rape but black people do because they are wild animals

• Endorsing a goal of advancing white people and separating white people from black people

As far as I know, Scott Alexander has not done these sorts of things, so Scott Alexander does not appear to be a white supremacist in the sense that David Duke is.

I created the David Duke market because I wasn't convinced about @IsaacKing 's resolution criteria. I am still not convinced about his resolution criteria, but it seems to me that he doesn't follow them slavishly, but instead applies some nuance in guessing whether someone is truly a white supremacist.

I think that's fair enough, as exact resolution criteria are hard, but I think it would be better if it was reflected in the resolution criteria. E.g. instead of saying that it resolves based on

Scott, at any point of their life, clearly attempting to normalize or encourage white supremacy (defined as the belief that white people should be treated better than people of other races simply due to their race)

and then not resolving it when clear examples of him normalizing this are provided, it would have been better to say "Scott Alexander is an actual white supremacist as I imagine them rather than a consequentialist who believe in intelligence differences by race". Or perhaps "Scott Alexander is an actual white supremacist similar to people like David Duke". (I don't know if you have any non-KKK type white supremacists in mind, but if so you could also reference those.)

I think there is value in having objective criteria as they make the resolution fairer, and I think there is value in having subjective criteria as they make the resolution more accurate, but I don't understand the value in writing down objective criteria and then resolving according to subjective criteria, other than to provide a faux-objectivity to improve the perceived legitimacy of the market (which has the potential to backfire).

Hmm the end of this comment might have come off as a bit aggressive. I meant stuff like "faux-objectivity" in a descriptive way: https://www.lesswrong.com/posts/N9oKuQKuf7yvCCtfq/can-crimes-be-discussed-literally

@tailcalled a market resolving that david duke has only been provably a white supremacist for 7 years surely only demonstrates the ridiculousness of the resolution criteria. maybe that was your point.

but that said, i basically agree with you that it's better for a market like this to resolve based on clear resolution criteria even if they don't completely accurately represent the underlying concept than to continually move the goalposts to match what the resolver intuitively believes the resolution ought to be.

@EllieHigh The 7 year result was a lower bound based on the quotes people cited. It's quite possible it could have been higher if people had commented more quotes or similar in the market; it's just that the resolution criteria specified that the resolution was based on the evidence commenters provided.

@EllieHigh I'm not sure resolving positively due to clear resolution criteria are the best solution.

The market partly functions to generate a common record about Scott Alexander's view on race. People may take this record to form their views on Scott Alexander and his associates, which might have quite severe consequences for Scott Alexander. Of course if the resolution is valid, then probably those severe consequences should not stop us from doing the resolution, but if the resolution is invalid, then presumably it is best to put in extra effort to avoid misleading people in harmful ways.

As an analogy, imagine if the prediction market was not about Scott Alexander, but instead some other charged topic, such as e.g. whether black people are less intelligent than white people. In that case it seems like it would be justified to be careful about the resolution; people might have react in all sorts of racist ways if such a market resolved positively, and so if someone carelessly resolved it positively, we might suspect them of being hostile to black people.

I don't know what the best solution to this issue is, but it seems like prediction markets ought to be accountable so they don't promote inaccurate hostility towards Scott Alexander or black people. Maybe a careful subjective evaluation would be the best, though it is a bit late now as people have already made bets based on objective criteria.

@tailcalled well maybe if the market was resolved by morgan stanley. but as it is the resolver is just Some Guy who's already Scott's fan operating on a site that Scott has had a big part in promoting. so in this case the subjective resolution just makes a mockery of the market

@EllieHigh I agree with you that the factors you mention make it less of a deal if the market falsely resolves NO, because the one can just dismiss the NO resolution as pro-Scott bias induced by the general environment.

However, I think that conversely, this would make it more of a deal if the market falsely resolves YES, because then the story is that a Scott Alexander fan on a Scott Alexander endorsed site declared Scott to be officially a white supremacist.

Why do you believe a NO resolution would be incorrect according to my stated criteria?

@IsaacKing If someone decides that they don't want to hire black people because black people are criminal, then that seems like treating white people better than black people on the basis of their race, making it white supremacy by your definition. Scott Alexander was definitely normalizing that position in his murderism post. So he has at some point been normalizing white supremacy by your stated resolution criteria.

@tailcalled Scott's daycare scenario seems hypothetical or based on false information. Daycare providers are required by federal law in the US to run background checks on employees, and in the UK they must ask the police to carry out the highest level of background check possible, which can even consider suspicions and unfounded allegations if the police believe they are credible, and can even consider known associates. That said, there aren't such requirements for all jobs, so we can consider jobs where background checks are not done instead. The way you described the reasoning of the people in the post is inaccurate and is itself racist. It's not that "black people are criminal", but that, based on 1991 figures - so at a time when lead poisoning from car exhausts and other sources was increasing criminality among all races, but disproportionately among black people, black people in the US were much more likely to go to prison at some point in their lives. These people and their reasoning was not hypothetical, studies have shown this to be a real unanticipated effect of reducing the use of criminal background checks in the US. Scott says denying people jobs based on this an actual racist action, and I agree. Scott also says that the reasoning for the racist action is not racist. So, if you only read the latter part, you might think that Scott is trying to normalise this behaviour, but he is not. He wants to stop this behaviour by reintroducing criminal background checks to jobs that do not use them to assess candidates. He is literally performing political activism to attempt to stop racist actions in that post. His claim is that his preferred policy is the only effective way to stamp out this kind of racism. This seems like the opposite of white supremacist to me. It seems, in fact, like genuine antiracism, not the faux antiracism of CRT-inspired people who make racism-as-consequences arguments that he critiques in the post.

@RobinGreen I feel like your dispute is mainly that my description of Scott Alexander's behavior is misleading with respect to what his behavior indicates, rather than what his behavior was.

And I agree! Summarizing Scott Alexander's view on race as "He thinks black people are less intelligent than white people, he thinks more intelligent people might have greater moral weight and ought to be in control of society, he thinks white supremacists should get to make their own all-white society, and he thinks it's bad when people freak out about racism just because some company won't hire black people due to thinking they are criminal since black people actually are criminal" makes him come off as some crazy racist in a way that I don't think reflects his actual feelings about the topic. But just in terms of the plain meaning of the words, it appears to me that the description is accurate (or was accurate? potentially different parts of the description were accurate at different times).

Specifically, it makes him come off as some sort of KKK-like person, or at least, a Steve Sailer-like person. Like it makes him come off as someone who doesn't believe that there is much value to egalitarian cooperation with black people, or who thinks of black people as potential political enemies, or who is trying to build a society which legally oppresses black people.

But notably, the resolution criteria says nothing about Scott Alexander's beliefs about the value in egalitarian cooperation with black people, or about treating black people as potential political enemies, or about trying to build a society which legally oppresses black people. Presumably because these are latent motivational factors that cannot be directly observed and therefore are hard to do resolutions based on.

Instead, the resolution criteria talks about objective behaviors, which might be motivated by the KKK-like motives above, but which might also be motivated by other motives, such as sympathy towards black people combined with concern about progressive overreach. Your argument seems to emphasize Scott Alexander's desire to help black people with the post, and I agree that may be a contributing factor... but it seems to have nothing to do with the actual resolution criteria of the market?

Or I should say, it seems to have nothing to do with the formal resolution criteria of the market, but it seems to have everything to do with the actual, informal resolution criteria. I think most people here expect Isaac King to resolve the market NO, presumably because Scott Alexander's objective behaviors don't indicate the sort of white supremacy that he had in mind.

@tailcalled Please stop saying "black people are criminal", this is highly offensive and inaccurate.

@tailcalled not sure it's worth it to respond to people who may not be writing in good faith

@RobinGreen Hm, would it be better to say something like

"He thinks black people are less intelligent than white people, he thinks more intelligent people might have greater moral weight and ought to be in control of society, he thinks white supremacists should get to make their own all-white society, and he thinks it's bad when people freak out about racism just because some company won't hire black people due to thinking they could be criminal since 28% of black people have been imprisoned at some point in their life and so there's a good chance that a random black person is actually criminal"

?

@tailcalled To elaborate, if someone stated that in the US Asians are less likely to be imprisoned than white people, would you think that a fair way to characterise what they had just said was "white people are criminal?" As a white person I would find that a highly racist and offensive summary of the evidence they had provided - even if the intent of the first person pointing it out had been to justify only hiring Asian people. It is foolish to amp up the level of racism in the discussion to an absurd level, in the service of criticising racism.

And no, I wouldn't think it would be much better to say something like your alternative proposal, because that still mischarecterises what Scott is saying. Although, it does have the merit of no longer including the really offensive phrase that I objected to. Scott is saying racist behaviour is bad and we should have policies that lead to such behaviour not happening, such as permitting or encouraging criminal background checks for jobs.

@ForrestTaylor I disagree. If it is ambiguous whether someone is writing in bad faith, and they also seem to be unsure whether you are writing in good faith, if people just dismiss their counterparties as being in bad faith then that's going to start a runaway fracture of the network of trust, where people who were on the fence end up rejected by partisans on one side or another, and reject them in turn.

If you want to establish good faith, you have to acknowledge distortions made by your own side and take the concerns of the people on your opponent's side into account. Otherwise you are just trying to destroy your opponents using rhetoric.

While we can argue about whether trying to explain the thought process behind some people not hiring black people in a particular set of circumstances is "normalising racism" or "normalising white supremacy", Scott's argument is we should just eliminate the actual racist behaviour, which he claims can be done by introducing his preferred policy (I make no claim about whether it would in fact eliminate it, but I do think it would reduce it). Thoughts inside someone's head don't directly hurt anyone, unless they also lead to racist actions or speech. It seems to me from a rationalist perspective, from a pragmatic perspective, from a compassionate perspective, from an effective altruist perspective, from a common sense perspective, from any political perspective except a racist one, we ought to prioritise reducing racist ACTIONS over reducing racist thoughts. So I think interpreting the resolution criteria in a holistic way, and taking into account the relative harms of both prongs of Scott's argument, leads to the conclusion that he is not being white supremacist in this post. I submit that a reasonable person on the street who was capable of understanding the resolution criteria and was not highly caught up in this discourse already, would agree with me.

Here is an analogy. Some sexist men in the 19th century really did argue that women were inferior to men and could not be trusted with serious and important jobs, which had to be given to men. Suppose a Victorian version of Scott Alexander argued in 1830 that this was a rational thought process for those men to follow, because women had not been given a proper full education and thus could not in fact be admitted to jobs which required a university education. Suppose he then proposed allowing all women who wanted to, to complete a full schooling and enter and graduate from university, so that they might prove their merit to enter such jobs and so that people hiring for such jobs might stop disqualifying women from consideration, because they would see over time that women were just as intelligent and capable as men on average for those types of jobs. This is a two-pronged argument of an analogous form. Clearly, the second part is not misogynist - in fact, on the contrary, it is a textbook liberal feminist argument (anyone who does not know what "liberal feminism" means should familiarise themselves with the distinction between liberal feminism and radical feminism, by consulting the Wikipedia article on liberal feminism). Would the first part have been "normalising misogynist arguments"? I don't know - maybe - but even if it had been, the entire argument - the first part combined with the second part - could have been one that a liberal feminist could have easily made, and perhaps they did make such arguments.

@RobinGreen I'm... not sure we disagree? Like in my original comment, I acknowledged that the giving summary was misleading because it made Scott Alexander come off as a raging racist. This seems similar to the point you are making: Scott Alexander wasn't trying to screw over black people, he was trying to push for some improvement where the bad black people could be distinguished from the good black people, so that good black people could be treated as well as good white people.

My issue is that the resolution criteria did not specify something like "I will attempt to do an advanced inference where I distinguish KKK-style motivations from racially-cooperative motivations combined with beliefs that white people are on average superior to black people, and I only count the former and not the latter as white supremacy". Rather, the resolution criteria specified "Scott, at any point of their life, clearly attempting to normalize or encourage white supremacy (defined as the belief that white people should be treated better than people of other races simply due to their race)". So it seems incorrect to resolve the market based on the former rather than the latter.

@RobinGreen

It's not that "black people are criminal", but that [...] black people in the US were much more likely to go to prison at some point in their lives.

What exactly do you think the word "criminal" means?

@IsaacKing How on earth do you not see the problem here? It's a universal quantification.