The first question is highly suspect in its value in testing one’s economics knowledge, considering that all basic economics texts (left, right, center, libertarian, or whatnot) teach that preferences are ordinal (ranked rather than real-valued) and interpersonally incomparable. So people who took economics would have been explicitly told to answer oppositely from what Klein considered correct!
This point was raised in the MR discussion and other places. (Bob Murphy wrote an introductory text that uses Klein’s exact sentence as an example of an economic error in reasoning.)
I agree that there’s a sense in which a dollar means more to a poor person than a rich, but that sense is not an economic one, but one of morality or cognitive science.
I agree that there’s a sense in which a dollar means more to a poor person than a rich, but that sense is not an economic one, but one of morality or cognitive science.
Consider the following thought experiment. Depending on a coin toss, you’ll become either rich or poor. Your decision is between granting $1 to the hypothetical version of yourself that’s poor, or to one that’s rich. Which do you choose? It seems like a perfectly clear-cut decision problem, with an equally clear-cut correct answer. There is also an obvious Counterfactual Mugging variant, where we won’t need to stipulate actual expected change in your rich/poor status and can consider granting money to your present (vs. different counterfactual) self, rather than different future self.
(This is meant as a focusing statement, not an objection to your comment.)
That’s just a rephrasing of the general intuition about the decision-theoretic optimality of giving a marginal unit of money to poor vs. rich people. It doesn’t change the fact that people are taught in economics that preferences are ordinal, making the claim in the survey economic nonsense and therefore uninformative about the answerer’s reasoning skills or economic knowledge.
A large number of causes could explain either answer: perhaps the respondent was parroting their economic “teacher’s password” when they said no, rather than answering the decision-theoretic analog.
I agree that there’s a sense in which a dollar means more to a poor person than a rich, but that sense is not an economic one, but one of morality or cognitive science.
I would have phrased it differently—I would have said that the standard economics is just wrong here, but it doesn’t matter for observable implications of preferences, so they can get away with it. I’m not sure if this means I disagree with you or not—it may just be semantics.
edit: I should have said observable behavioral implications of preferences. There are, in fact, neurologically observable implications of cardinal preferences, but the fact that preferences are cardinal doesn’t have an impact on the behavior of an agent. Well, at least if we assume that agent is rational. Now that I think about it, there’s the potentially a nonrational agent that’s a counterexample to this statement.
I think we agree. I refer to Murphy et al’s position as “hyper-anti-IUCism” (IUC = interpersonal utility comparison). If you restrict the topic to a specific domain, you have to throw out IUCs, but in the general case they are a meaningful, helpful concept. See my remarks in the second link.
(Btw, why do I get URL bloat when I copy a URL shortcut from this site?)
The first question is highly suspect in its value in testing one’s economics knowledge, considering that all basic economics texts (left, right, center, libertarian, or whatnot) teach that preferences are ordinal (ranked rather than real-valued) and interpersonally incomparable. So people who took economics would have been explicitly told to answer oppositely from what Klein considered correct!
This point was raised in the MR discussion and other places. (Bob Murphy wrote an introductory text that uses Klein’s exact sentence as an example of an economic error in reasoning.)
I agree that there’s a sense in which a dollar means more to a poor person than a rich, but that sense is not an economic one, but one of morality or cognitive science.
Consider the following thought experiment. Depending on a coin toss, you’ll become either rich or poor. Your decision is between granting $1 to the hypothetical version of yourself that’s poor, or to one that’s rich. Which do you choose? It seems like a perfectly clear-cut decision problem, with an equally clear-cut correct answer. There is also an obvious Counterfactual Mugging variant, where we won’t need to stipulate actual expected change in your rich/poor status and can consider granting money to your present (vs. different counterfactual) self, rather than different future self.
(This is meant as a focusing statement, not an objection to your comment.)
That’s just a rephrasing of the general intuition about the decision-theoretic optimality of giving a marginal unit of money to poor vs. rich people. It doesn’t change the fact that people are taught in economics that preferences are ordinal, making the claim in the survey economic nonsense and therefore uninformative about the answerer’s reasoning skills or economic knowledge.
A large number of causes could explain either answer: perhaps the respondent was parroting their economic “teacher’s password” when they said no, rather than answering the decision-theoretic analog.
I would have phrased it differently—I would have said that the standard economics is just wrong here, but it doesn’t matter for observable implications of preferences, so they can get away with it. I’m not sure if this means I disagree with you or not—it may just be semantics.
edit: I should have said observable behavioral implications of preferences. There are, in fact, neurologically observable implications of cardinal preferences, but the fact that preferences are cardinal doesn’t have an impact on the behavior of an agent. Well, at least if we assume that agent is rational. Now that I think about it, there’s the potentially a nonrational agent that’s a counterexample to this statement.
I think we agree. I refer to Murphy et al’s position as “hyper-anti-IUCism” (IUC = interpersonal utility comparison). If you restrict the topic to a specific domain, you have to throw out IUCs, but in the general case they are a meaningful, helpful concept. See my remarks in the second link.
(Btw, why do I get URL bloat when I copy a URL shortcut from this site?)