I had a fun conversation with another economist the other day, which helped me suspect there may be yet another pathway of fake distributional indifference in bad economics I haven’t really discussed before, in this case centered on the theoretical concept of “initial endowments.”
Recall “fake distributional indifference” is just my term for the great assortment of errors and rhetorical tricks by which purveyors of bad economics seek to portray neoclassical welfare economics as saying more about distributional issues and ethics in the real world than it really does. The initial endowment variant of fake distributional indifference runs something like the following: “In economic theory, we can theoretically get to any Pareto optimal, economically efficient outcome corresponding to any distributional ethics by changing the initial endowment. However, in real life we can’t do that. In real life, we must consider the cost of changing the distribution of economic power in terms of lost “efficiency,” the deadweight loss of “utility” that accompanies any tax, for example, which must be balanced against any purported benefit based on distributional ethics, in other words, the famous “efficiency versus equity” tradeoff I’ve discussed in past posts. To avoid this loss of utility, we must accept the status quo Pareto optimal, economically efficient outcome, or if we’re not already at one, whatever particular instance results from policies designed to foster one from any given starting point. Whether or not to accept the loss of utility associated with “re-distribution” is exogenous to economic theory. Other people may decide the utility loss is worthwhile, including economists talking as regular people, but the job of economists qua economists is simply to point out the utility loss.” That’s basically just repeating back what this fellow was telling me, except he wasn’t very clear on the last bit, so when he informed me he would be willing to “tolerate” the utility losses if the taxes generating the “deadweight loss” of utility were used for certain purposes, I couldn’t really tell if he was still talking as an economist or just a random person with ethical beliefs, but I’m giving him the benefit of the doubt on that one.
Oh boy, here we go again, right? Bad economics rears its ugly head. Distributional indifference in neoclassical welfare economics holds even in the real world, when initial endowments are off the table. It holds even between Pareto optimal, economically efficient outcomes and any non-Pareto optimal, non-economically efficient outcomes (economically inefficient outcomes?) not Pareto dominated by them. That is to say, it holds between any two outcomes related in such a way that one cannot get from the one to the other via Pareto improvements. (That’s sufficient for my purposes, although honestly there are issues even if one can, relating to the fact particular real world instances of apparent Pareto improvements inevitably correspond to some subject’s distributional ethics but not others, which means they may themselves generate utility losses for some people, and hence are not real Pareto improvements. In other words, Pareto improvements are denizens of the Fairy Land, not the real world, a point I’ve discussed previously and will surely discuss again.) One doesn’t “lose” utility when moving from a particular real instance of a Pareto optimal, economically efficient outcome to a non-Pareto optimal, non-economically efficient outcome not Pareto dominated by it. One simply cannot compare the utility of the two outcomes. It’s that discontinuity, that incommensurable quality of outcomes differing along the distributional dimension, masked by things like “welfare losses” and “deadweight losses” defined relative to one distribution, that leads so many down the merry path to bad economics.
Put another way, there is no normative or ethical proposition in neoclassical welfare economics restricting distributional indifference to changing initial endowments, thus rendering it irrelevant in the real world where changing initial endowments is not as option. The idea true distributional indifference is restricted to changes in initial endowments because other ways of addressing distributional issues involve an “efficiency equity” tradeoff involves a form of fake distributional indifference I’ve discussed many times. “Economic efficiency,” defined using “utility,” cannot correctly be used as an independent criterion from utility. One can’t use economic efficiency to break distributional indifference by saying the latter only concerns utility, but we must also consider economic efficiency. I wrote a blog post or two on that, which I may have mentioned once or a million times. (See The Equity Efficiency Tradeoff and Fake Distributional Indifferences, July 22, 2020, or The Equity Equity Tradeoff, September 9, 2020.)
This issue is linked directly to the attenuated nature of “utility” in neoclassical welfare economics, the ethical half-theory structure of neoclassical welfare economics with some relevant normative issues in and some out, and the resulting split between the ethics of the Fairy Land of the Economic Theory and the ethics of reality. In the ethically attenuated, partially unspecified, unreal Fairy Land, one can make normative propositions about Pareto optimal, economically efficient outcomes as a general class, about collections, frontiers, boxes of equivalent outcomes, and one can flit easily between them at the stroke of the pan, by changing initial endowments or making market structure preserving transfers. In the real world, one is inevitably dealing with one particular instance of a Pareto optimal, economically efficient outcome at a time, in a fully specified world in which variable, subjective, full ethics apply. One. Not all of them. One. That’s it. And that one will appeal to some people, some subjects, but not others. The relevant real world comparisons, discussions, propositions must relate to particular, fully specified instances of economically efficient outcomes and other potential outcomes under realistic conditions. And forget flitting. If society prefers something different in the real world, it must do something, change something. Initial endowments are obviously out, and I’ve discussed the related problematic idea of market structure preserving transfers in previous posts. (See Appeals to Nonexistent Mechanisms for Addressing Distributional Concerns, June 10, 2020.) Labor and capital markets, even perfectly competitive ones, are part of a system for distributing economic power. People may have ethical opinions about them that are exogenous to neoclassical welfare economics. Distributional indifference applies to those distributional mechanisms just as much as to initial endowments modified by them.
Normative or ethical thinking about economic issues in the Fairy Land is different from normative or ethical thinking about economic issues in real life in that respect, and also with respect to the false factual premises that arguably apply in the Fairy Land. Normative or ethical arguments developed in the Fairy Land may be perfectly sensible when applied to the Fairy Land, but they can create all manner of confusion and conflict when inappropriately applied to the real world. It’s a misuse of economics. It’s bad economics.
Neoclassical welfare economics cannot properly be used to thwart other people’s efforts to revise economic systems, including perfectly competitive, Pareto optimal, economically efficient systems, to accommodate their distributional ethics or really other ethics. Ethical decisions about the distribution of economic power; whether to resolve particular or all interpersonal conflicts of preferences, needs, desires by using economic power in markets (extent of the market); even how to treat others expressing preferences that don’t conflict with the preferences of anyone else, under realistic conditions, are exogenous. Using neoclassical welfare economics to argue explicitly or implicitly for particular methods of resolving interpersonal conflicts of preferences or particular distribution or allocation systems or ethics, such as those expressed in perfectly competitive labor and capital markets and product markets, is bad economics.
Bad economics creates confusion and conflict. We can fix bad economics in the short run by accurately describing what neoclassical welfare economics really says, the normative inputs, the limits of the ethical half-theory structure. We can fix bad economics by accurately portraying the important role of activist democratic government in economic systems, including market systems, in terms of addressing the subjective, normative issues exogenous to neoclassical welfare economics, based on the subjective ethical views of the voters. We can fix bad economics by exerting some pressure on the typically complacent, self-satisfied, self-congratulatory residents of the Villa of Academic Economics to finally review their pedagogy and enforce some standards of professional behavior. In the long run, we can fix bad economics by revising the confusing bits of neoclassical welfare economics that generate bad economics, formalizing the role of democratic government in market systems, and hopefully removing the normative content entirely. Resolving the normative or ethical issues associated with evaluating economic systems and outcomes shouldn’t be an intellectual football contested by teams of clueless, insular, defensive, clannish, technocrats speaking in arcane, inaccessible terms as though they belonged to some all powerful ancient priesthood. It should be open, honest, public, forthright, in plain language, a matter for the people and democratic government. Resolving those issues will necessarily be an ongoing process, as some people die and others are born, and the living revise their ethical positions in response to the arguments of philosophers and others, events, experience, knowledge, science, technology, etc. There is no final resolution, only temporary agreement.
