Oh my god I am so excited for this, I've been trying to put together a thesis paper on this exact subject! I have had such a hard time finding prior relevant work.

Fair, fair, and fair. I do think there are mitigating responses to all of these points as well, but I’ll concede the point that these are cases on the fringes of convenience for him. I was personally more thinking about IQ if I had to think of an example - he seems to place more importance on it than most people, but as I think he pointed out in a blog post I can’t find now, this leads just an awful lot of people to really statist and quasi or outright fascist views, so even if it doesn’t actually imply fascism, it’s an area where adopting a view closer to the average would be more convenient, provide an additional reason he could give against such people.

Thanks, these are interesting examples (and if I’m commenting too much someone please tell me, I can do that sometimes I think), but I range from somewhat to very skeptical on them as counterexamples:

1. This is the most plausible one I think, it really does seem like it lends support for greater intervention on certain views. However, it’s hard to find a view of population ethics/population sciences that does not have some population it prefers, or that gives a good account of why incentives will produce it naturally. My impression is that most people either have quite implausible views that are completely neutral, or just, as with Caplan, think this isn’t a road we want to go down.

2. I think Caplan thinks education would be pretty fine if you took away the public funding/subsidies, it would just naturally become much less common (though he does make note of the issues with a market quickly optimizing “conformity” signals specifically, which might be the greatest source for market inefficiency here for him)

3. He seems to think humans are primarily irrational in the areas where anarcho-capitalism takes away our power and primarily rational where it would leave us power, see his arguments about for instance how much people are willing to spend in rent to live in immigrant free neighborhoods versus what they actually vote for in immigration policy, or more broadly his work on the irrational voter. His views aren’t always convenient in this area, but some amount of human irrationality is very hard to plausibly deny, and the version he believes in is pretty convenient for him imo.

I think any question that attempts to get at the heart of the strongest objection to a public figure's worldview is going to sound like an accusation, because in a way it is, mostly I hope it's taken as an ultimately good natured, curious, and productive accusation. On the point of libertarianism being a "good lens", I mean libertarianism as a policy suggestion. I am voicing suspicion that there isn't a plausible lens behind this policy view that generalizes so well in both philosophy and the real world that it doesn't leave Caplan's slate of opinions looking suspicious, but for what it's worth my second question was basically asking him to propose one.

Part of my second question is that I think in order to beat these two challenges, the best he can do is say that there is one fairly simple principle that is behind anarcho-capitalism, and that it generalizes so robustly, both when thrown into the real world, and when thrown into philosophical controversies, that it causes all of them to conveniently point in a similar direction. It would have to be one he believed in from a young age and saw vindicated more and more over time in practice, and it needs to be remarkably unpopular to, despite having unusually powerful application in so many controversies, escape the sympathies of so many other experts. I suspect he will suggest something like this, but I am suspicious a principle that actually meets these criteria doesn't exist, and that much of his worldview is best explained by bias. This is why I think a question on this level is one of the best challenges to pose him.

Two reasons I disagree:

1. It is suspicious to just happen to have a whole bunch of views that support one's pre-existing politics, but it is only a little less suspicious to have a whole bunch of not that related views that all conveniently support one coherent political view
2. I'm taken to understand that Caplan has been a libertarian since he was a kid, and an ancap almost as long. Insofar as he considers most of or many of the listed positions to come from careful academic reflection, most of the arguments he makes about them are probably ones he didn't have when he first became sympathetic to his current politics

Forgot mental illness, which again is suspiciously convenient, and maybe on the lower end of the plausibility spectrum among his views.

Because my draft response was getting too long, I’m going to put it as a list of relevant arguments/points, rather than the conventional format, hopefully not much is lost in the process:

-Ethics does take things out there in the world as its subjects, but I don’t take the comparison to empirical science in this case to work, because the methods of inquiry are more about discourse than empirical study. Empirical study comes at the point of implementation, not philosophy. The strong version of this point is rather controversial but I do endorse it, I will return to it in a couple bullets to expand it out

-Even in empirical sciences, the idea of theories just being rough models is not always relevant. it comes from both uncertainty and the positive view that the actual real answer is far too complicated to exactly model. This is the difference between say economics and physics – theories in both will be tentative, and accept that they are probably just approximations right now because of uncertainty, but in economics this is not just a matter of historical humility, but also a positive belief about complexity in the world. Physics theories are both ways of getting good-enough-for-now answers, and positive proposals for ways some aspect of reality might actually be. Typically with plurality but not majority credence.

-Fully defining what I mean by ethics is difficult, and of less interest to me than doing the ethics. Maybe this seems a bit strange if you think defining ethics is of supreme importance to doing it, but my feeling of disconnect between the two is probably part of why I’m an anti-realist. I’m not sure there’s any definition I could plug into a machine to make an ethics-o-meter I would simply be satisfied taking its word for it on an answer (this is where the stronger version of bullet one comes in). This is sort of related to Brian Tomasik’s point that if moral realism were true, and it turned out that the true ethics was just torturing as many squirrels as you can, he would have simply learned he didn’t care about ethics and it wasn’t what he was doing all along. I feel part of my caring about ethics is constituted by my understanding of how I got there more than it is about extrapolating from exact definitions. I know it when I do it, and it is a project that, to my understanding of it, I care about deeply right now.

-I don’t think this answer quite fits any of Greenberg’s proposals exactly, but he is definitely confused, and fair enough, as he is confused about a confusing topic. I just want to note that it is meta-ethics that is confusing, not anti-realism. I think he blows past moral realism sort of quickly, expecting that what realists who subscribe to theories like these are doing is perfectly understandable, but I think it is still extremely weird. Most initial approaches one can take to moral realism either start out apparently collapsing into normative ethical theories instead, or else require some extremely unlikely empirical assumption. In order to rescue realist theories, you need to start getting ideas that are more complicated and recognize the dilemmas. I originally wrote two example dialogues to get at this point, but they wound up going on too long for a comment, so I just want to start by positing that, in my experience, this is the case. The obvious first approaches either in some way posit one’s normative theory to be what “value” is despite disagreement from other people who are using the same words, or else there is some sense in which the disagreement is explained away as coming from some source of irrationality that, if spelled out with an empirical prediction, requires a probably bad prediction. Meta-ethics always faces a foundational dilemma in spelling out what exactly moral disagreement is.

-Since this is getting long winded and it seems like it’s pretty much only us here at this point, I was wondering if you wanted to migrate this conversation in some way, for instance we could chat more via video call or something at some point. If not I’m also fine with that, we could call it here or keep going in the comments. I just thought I would mention that I’m open to it.

I endorse moral uncertainty, but I think one should be careful in treating moral theories like vague, useful models of some feature of the world. I am not a utilitarian because I think there is some "ethics" out there in the world, and being utilitarian approximates it in many situations, I think the theory is the ethics, and if it isn't, the theory is wrong. What I take myself to be debating when I debate ethics isn't which model "works" the best, but rather which one is actually what I mean by "ethics".