Comments Elsewhere

Brandon Berg: In fact, I think Scott did just this, as he said that he was not endorsing laws against suicide.

Again, this by itself does not prove that Scott regarded his claim as merely positive and not normative. All that it proves is that he’s open to regarding whatever value (whether some or none) efficiency might have as trumped by the value that something else has. (Normatively, being wealthy is better than being poor. That’s part of what “wealth” and “poverty” mean. But there are plenty of cases in which it may better to realize some other value, at the cost of being less wealthy, than it would be to maximize wealth. That doesn’t mean that “being wealthy” is a strictly non-normative category. It just means that it’s an overridable norm.)

David: Efficiency is not a normative term. In most usage it is actually a quantitative measurement, a car with greater efficiency gets better miles per gallon than a less efficient car.

This is obviously not true. Your ability to quantitatively calculate efficiency, as the term is ordinarily used, depends on you first identifying what counts as a benefit and what counts as a cost. An increase in miles per gallon, for example, only counts as an increase in efficiency because, in this case (but not in all) getting more of the dependent variable (miles traveled) for the same or less of the independent variable (gasoline burned) counts (prima facie) as a good feature for the car to have. Which is a normative judgment. (Which in turn depends on your identifying the dependent variable as a benefit, i.e., a good thing for the car to produce, and the independent variable as a cost, i.e., a bad thing for the car to demand, or something which is bad in itself and valuable only for its consequences. These are, again, normative judgments.)

There are in fact plenty of cases where efficiency involves getting less for the same or less for more (e.g. less exhaust for the same amount of gasoline, less waste heat for more revolutions of the turbine). The only thing that all of the cases of efficiency have in common with each other and not with cases of inefficiency is not any kind of quantitative positive relation, but rather the normative relation of increasing things good to have and decreasing things bad to have. The ordinary use of the term “efficiency” simply has no cash value denominated in purely quantitative terms.

Brandon Berg: If we suppose that we have a general idea of what people like and how much they like it, we can say that a particular arrangement will or will not be Kaldor-Hicks efficient. That’s a positive statement, not a normative one. Yes, we take preferences into account (possibly inaccurately). But we simply consider these as positive facts.

David: Better off and worse off are to the economist just a tally of what each individual under study thinks (or more appropriately reveals) of his situation. Yes, each individual must make a normative claim, but the economist makes no such claim, he just tallies those claims.

Fine. So here’s a conventional textbook definition of Pareto efficiency. (Since Kaldor-Hicks efficiency is defined in terms of Pareto efficiency, we’ll leave that as a further exercise.)

(PE) A situation is Pareto efficient if and only if there are no available changes that would make at least one person better off and make nobody worse off.

Your suggestion is that we make Pareto efficiency non-normative by making “better off” and “worse off” refer to positive facts to the effect that the preferences that the people in question happen to have are satisfied or frustrated. So the more explicit definition is something more like:

(PE′) A situation is Pareto efficient if and only if there are no available changes such that (1) there is at least one person for whom the change would satisfy at least one currently unsatisfied preference, and (2) there is no one for whom the change would frustrate at least one currently satisfied preference.

Is this an accurate way of spelling out what you mean when you claim that economic efficiency, as Scott was using the term, is non-normative?