I had an interesting thought the other day relating to the concept of Pareto improvements, that is to say, changes that make one person better off and no one worse off. Well, interesting to me, anyway. In a previous post, I explained that in the Fairy Land of Economic Theory, that is from the point of view of economic theory or economists qua economists, any Pareto improvement is equivalent to any other and they’re all equally ethically uncontroversial because issues relating to distributional ethics have been explicitly set aside, but in real life everyone evaluating economic systems or outcomes holding any form of distributional ethics, so basically everyone including regular folk and economists in their more general role as regular folk, will not only prefer some Pareto improvements to others, but may feel any particular Pareto improvement makes their society less ethical, less consistent with their distribution beliefs than before the improvement. In that sense, Pareto improvements are like Pareto optimums. They’re all the same in the Fairy Land, but in real life some people will prefer one and some another, and the decision between them, or even the decision to move toward any particular one in the presence of distributional consequences, will be external to the normative scheme presented in neoclassical welfare economics. However, it occurs to me now there may be even more serious problems with the concept that apply equally to real life and the Fairy Land.
Let’s say we’re happily ensconced in the Fairy Land, so we’re not going to take up controversial distributional ethics ourselves. Let’s talk about how Pareto improvements in that land work. The first notable thing about these Pareto improvements is that one must take some care about the basis on which one is evaluating being “better off.” In particular, the ostensibly non-controversial quality of accepting Pareto improvements in the Fairy Land requires that we define “better off” in terms of “utility,” not as more conventionally in real life in terms of money. This is an issue that slipped me up on a regular basis to comical effect before I sat down and thought about it for five minutes. The problem with trying to define “better off” in terms of money is that money in the form of relative economic power can have real consequences in terms of “utility,” that is, the expression of individual preferences that forms the basis of neoclassical welfare economics. For example, say we have two people, A and B, who let’s say both like some precious one off gem. Currently, A has more money, economic power, and hence can gain possession of the gem by wielding that superior economic power and buying it on the market. However, suppose we then drop a shipload of money into B’s lap so B now has the upper hand in that department while A has just as much money or economic power as he or she had before. One might be tempted to think of this as a Pareto improvement, but that’s not really correct, is it? The change in relative economic power means B now gets the one off gem that A had gotten formerly. A must have less “utility” than previously in the sense A cannot fulfill preferences of the same rank he or she fulfilled before the change. A has been made worse off in terms of “utility.” No, trying to define Pareto improvements in terms of money or economic power does’t really work in a logical sense if one intends to return at some point to the “utility” used in neoclassical welfare economics.
However, we’re not out of the woods even if we define the concept of Pareto improvements in terms of changes in “utility.” One issue is what exactly is the real world equivalent of increasing one person’s “utility” while keeping everyone else’s “utility” the same in the context of an economic model? We may have a bag of money we can imagine dropping on someone’s lap, but we don’t have a bag of “utility.” But we’re talking about the Fairy Land right now so let’s not fret about it. I suppose it must just mean some policy by which one person does better in terms of moving up his or her ranking of preferences fulfilled than formerly, and everyone else fulfills preferences of the same rank as formerly. However, if that’s what we have in mind, then it seems logically impossible to pull off a Pareto improvement if the people involved, the subjects economists are studying, not the economists themselves looking in from on high, have preferences in the form of ethical beliefs about relative “utility,” or to address the particular expression of a Pareto improvement in which everyone is made better off together and no one’s “utility” is held constant, there is any variation in those preferences. To make the issue here a bit more intuitive or obvious, let’s say A and B are quite different sorts. A is relatively poor but hard working and conscientious and just in general full of everything anyone might associate with merit, while B is your classic silver spoon playboy, shiftless, lazy, totally lacking in what anyone might consider merit. Say we drop a shipload of “utility” into B’s lap so now B is fulfilling preferences way up his list or ranking of preferences while A remains stuck back at level one. Is there anything about this result that may appear ethically controversial not to us but to A? Well, I think if A has any ideas linking relative “utility” to merit or fairness or justice or has any utilitarian concerns involving human welfare or let’s just say any preferences relating to distributional ethics at all, A might actually be a little annoyed or put off at the change. A might suppose that as a matter of ethics we should have dispensed the shipload of “utility” some other way. More troubling, if we accept that subject A may have preferences on such matters, then it becomes clear A may no longer be attaining or expressing preferences of the same rank as formerly with respect to that particular aspect of the society in which he or she lives. In other words, I think this concept of a Pareto improvement really only makes sense if no one involved as subjects of economic theory holds any form of distributional ethics or, I suppose, if everyone involved holds the same form of distributional ethics so everyone is on board with the change in question. That would clearly not be the case in real life, in which we can safely say everyone holds beliefs relating to distributional ethics and there is variation in those beliefs, which is what generates the controversy “utility” in economics was designed to avoid, so Pareto improvements based on “utility” would be logically impossible in that context, but I’m suggesting the same result seems to apply equally to the ciphers standing for human subjects in the Fairy Land of Economic Theory itself. At least, I’ve never run across an explicit assumption or proposition that the theoretical constructs that stand in for people in the economic models of the Fairy Land are prohibited from holding preferences relating to distributional ethics or must all hold the same distributional ethics. The preferences that underlie “utility” in modern neoclassical welfare economics are famously meant to be neutral with respect to motivation, thus allowing preferences motivated by ethical beliefs to be associated with “utility” just as much as preferences based on other motivations. “Utility” generating preferences are no longer limited to amoral expressions of narrow self interest as in the old moral philosophizing of the Deists struggling to explain certain seemingly unfortunate characteristics of their fellows in light of the good lord’s benevolence, although of course the factual premise of profit maximization on the part of business people qua business people remains parts of the theory. It would be a little strange if ethics were allowed with respect to some issues but not others.
It’s an interesting area because it seems to me to get back to one of the distinctive claims about neoclassical welfare economics one finds in bad economics, which is that rather than economists qua economists setting aside controversial distributional ethics and by extension other less commonly recognized controversial ethical issues, neoclassical welfare economics proposes the very controversial ethical proposition regular folk evaluating economic systems or outcomes, including economists in their more general role as regular folk, should not have any beliefs relating to distributional ethics or any beliefs relating to any of the other controversial ethical issues. I’ve discussed before how that error can result from a failure to acknowledge or come to terms with the distinctive ethical half-theory structure of neoclassical welfare economics in which some ethical inputs relevant to evaluating real instances of market systems and outcomes are endogenous to the theory and some exogenous, or to put it another way, a failure to distinguish reality, in the form of defined situations in which all relevant ethical considerations must be addressed, from the Fairy Land of Economic Theory, partially undefined situations hedged about with false factual premises in which one can ignore certain relevant ethical considerations one finds in real life. Is the entire concept of ostensibly ethically uncontroversial Pareto improvements built on this error? Is it based on some strange journey from the Fairy Land to reality and then back to the Fairy Land? Or perhaps on a journey from modern economics to archaic Deism and then back to modern economics? In other words, does it depend upon an assumption both the real and theoretical subjects of economic theory follow the prescription of what is now bad economics and hold no distributional ethics or, if they do, they must all hold the same distributional ethics?
Actually, now I’m thinking about it, I may have over complicated the situation a bit just then. What if the postulated decline in A’s “utility” following B’s windfall were not generated by preferences relating to distributional ethics at all but simple greed or envy as purveyors of bad economics typically propose? What if A follows the dictates of bad economics and holds no ethical beliefs including those suggesting the emotion of envy is not a proper basis for ethical thought? What if A simply prefers B not get one over on him or her? A greedily prefers the windfall himself or herself? Would the change in rank of A’s preferences being fulfilled relating to the characteristics of the world in which he or she lives, his or her relative situation in that world, and the “utility” associated with the fulfillment of those preferences no longer count? Are we in the business now of assessing the motivations that underlie preferences expressed by the subjects of economic theory and determining in our role as economists which preferences are acceptable and generate “utility” and which are not and do not? The de gustibus principle is out the window? We economists declare being greedy is fine, but not that greedy? Or must we also banish envy from the Fairy Land and institute a corresponding false factual premise in terms of our description of the real world in an applied context, not just deliver our own normative assessment of the sentiment, to make sure everything turns out right?
It’s funny when economists try to do ethical philosophy, isn’t it? Has there ever been a more comically confused, non-rigorous subject than normative neoclassical welfare economics? We should fix bad economics by taking out the casual and typically just bad normative or ethical philosophizing, putting value judgments and ethics back in the hands of the people and democratic government, and making economics a true science.
