I find this a rather challenging post, even if I like the high-level topic a lot! I didn’t read the entire linked paper, but I’d be keen to understand whether you think you can make a concise, simple argument as to why my following view may be missing sth crucial that immediately follows from the Harsanyi vs. Rawls debate (if you think it does; feel free to ignore):
The Harsanyi 1975 paper which your linked post also cites (and which I recommend to any EA), is a great and rather complete rebuttal of Rawls core maximin-claim. The maximin principle, if taken seriously, can trivially be seen to lead to all sorts of preposterous choices that are quite miraculously improved by adding a smaller or larger portion of utilitarianism (one does by no means need to be a full utilitarian to agree with this), end of story.
Thank you so much for your comment. Yeah, I think Harsanyi’s review is awesome, too (particularly his criticism of Ralws’s position on future generations); I don’t know why it seems to be quite often neglected by philosophers.
What you might be missing: notice everyone agrees that maximin sort of sucks as a decision principle—Rawls never endorsed it this way. However, I’d add: a) It’s only discussed as a decision criteria in cases where you don’t have probabilities (it’s Wald’s criterion). Now the standard response to that would be something like “but you can build credences by applying Laplace’s Principle,” and I tend to agree. But I’m not sure this is always the best thing to do; not even Savage thought so. Rawls thinks that, particularly in the original position, this would not be done… and really, it’s not clear why.
b) Notice that people often display higher risk-aversion when making decisions for the sake of others—they’re supposed to be “prudent.” Of course, I don’t think this should be represented as following some sort of maximin principle—but maybe as some kind of “pessimistic” Hurwicz criterion; yet, and I can’t stress this enough, my point is that people are not actually implying that negative outcomes are more likely. c) Actually, Harsanyi himself remarks (in the addendum of the review) that the maximin (or the difference principle) is a useful proxy to use, e.g., in the theory of the optimal taxation—even though it’s not a “fundamental principle of morality.” I think this point is more relevant than it might seem at first sight; actually, my interpretation / defense of the difference principle basically depends on that.
I think (almost) all I’d have to say on this matter is in the Section 3 of the paper (especially the first part). But TL;DR: the difference principle is not a basic moral principle (and maximin is not an alternative to expected utility theorey). And the problem of the original position should be seen as a complex bargaining problem—no originality here, as K. Binmore and D. Gautier are quite explicit about it (but they’re not so popular). I wouldn’t say this is the best solution for “the problem of justice” (I should deeply study the Kalai-Smorodinsky solution to bargaining games first), but I do think it’s a salient one—and this would explain its appeal.
Thank you! I was actually always surprised by H’s mention of the taxation case as an example where maximin would be (readily) applicable.
IMHO, exactly what he explains in the rest of the article, can also be used to see how optimal taxation/public finance should rather only in exceptional cases be using a maximin principle as the proxy rule for a good redistributive process.
On the other hand, if you asked me whether I’d be happy if our actual very flawed tax/redistribution systems would be reformed such as to conform to the maximin—es, I’d possibly very happily agree on the latter, simply as a lesser of two evils. And maybe that’s part of the point; in this case, fair enough!
I find this a rather challenging post, even if I like the high-level topic a lot! I didn’t read the entire linked paper, but I’d be keen to understand whether you think you can make a concise, simple argument as to why my following view may be missing sth crucial that immediately follows from the Harsanyi vs. Rawls debate (if you think it does; feel free to ignore):
The Harsanyi 1975 paper which your linked post also cites (and which I recommend to any EA), is a great and rather complete rebuttal of Rawls core maximin-claim. The maximin principle, if taken seriously, can trivially be seen to lead to all sorts of preposterous choices that are quite miraculously improved by adding a smaller or larger portion of utilitarianism (one does by no means need to be a full utilitarian to agree with this), end of story.
Thank you so much for your comment. Yeah, I think Harsanyi’s review is awesome, too (particularly his criticism of Ralws’s position on future generations); I don’t know why it seems to be quite often neglected by philosophers.
What you might be missing: notice everyone agrees that maximin sort of sucks as a decision principle—Rawls never endorsed it this way. However, I’d add:
a) It’s only discussed as a decision criteria in cases where you don’t have probabilities (it’s Wald’s criterion). Now the standard response to that would be something like “but you can build credences by applying Laplace’s Principle,” and I tend to agree. But I’m not sure this is always the best thing to do; not even Savage thought so. Rawls thinks that, particularly in the original position, this would not be done… and really, it’s not clear why.
b) Notice that people often display higher risk-aversion when making decisions for the sake of others—they’re supposed to be “prudent.” Of course, I don’t think this should be represented as following some sort of maximin principle—but maybe as some kind of “pessimistic” Hurwicz criterion; yet, and I can’t stress this enough, my point is that people are not actually implying that negative outcomes are more likely.
c) Actually, Harsanyi himself remarks (in the addendum of the review) that the maximin (or the difference principle) is a useful proxy to use, e.g., in the theory of the optimal taxation—even though it’s not a “fundamental principle of morality.” I think this point is more relevant than it might seem at first sight; actually, my interpretation / defense of the difference principle basically depends on that.
I think (almost) all I’d have to say on this matter is in the Section 3 of the paper (especially the first part). But TL;DR: the difference principle is not a basic moral principle (and maximin is not an alternative to expected utility theorey). And the problem of the original position should be seen as a complex bargaining problem—no originality here, as K. Binmore and D. Gautier are quite explicit about it (but they’re not so popular). I wouldn’t say this is the best solution for “the problem of justice” (I should deeply study the Kalai-Smorodinsky solution to bargaining games first), but I do think it’s a salient one—and this would explain its appeal.
Thank you! I was actually always surprised by H’s mention of the taxation case as an example where maximin would be (readily) applicable.
IMHO, exactly what he explains in the rest of the article, can also be used to see how optimal taxation/public finance should rather only in exceptional cases be using a maximin principle as the proxy rule for a good redistributive process.
On the other hand, if you asked me whether I’d be happy if our actual very flawed tax/redistribution systems would be reformed such as to conform to the maximin—es, I’d possibly very happily agree on the latter, simply as a lesser of two evils. And maybe that’s part of the point; in this case, fair enough!