Economics As A Purportedly Neutral Discipline

I was part of a fun little exchange the other day about whether “economics” is a “neutral” discipline, as opposed to “political economy,” which was presented as not. Maybe I'll take a look at that this week.

Defining “economics” is famously tricky because the discipline involves disparate components. Some components, such as mathematical or engineering optimization techniques, may fairly be considered neutral tools applicable to a variety of different problems. Other components, such as the scientific or empirically predictive component, are often meant to be neutral, cast as neutral, but there is quite often some question about the role of neutral scientific evaluation based on empirical success versus other potential criteria. These other criteria, theoretical tractability, “understanding,” etc., can introduce non-neutral, normative evaluative elements to even ostensibly positive economics. For example, “understanding” often seems to equate roughly to intuitively appealing storytelling.

When one considers explicitly normative economics, such as neoclassical welfare economics, a traditional component of “economics,” the idea of “neutrality” becomes even more confusing and questionable. As with any ethical theory, the logic of the argument presented may be considered “neutral,” but the theory itself would not normally be considered “neutral.” It’s of interest to those who accept certain normative propositions and reject others; it expresses an ethical viewpoint. This is what I often refer to as “positive normative” economics, the conceit that economists aren’t presenting neoclassical welfare economics as a normative theory anyone should care about, they’re just analyzing the logic of the argument, in case anyone is interested. They then turn out around and give their policy evaluations, recommendations, based on that normative theory, but still, ostensibly just in case anyone is interested, which they’re not claiming anyone should be. Why do they talk in that funny, insincere way? Because while they appear to like and agree with the resulting policy recommendations, they don’t want to or can’t do proper ethical philosophy, evaluate their normative theory, or even analyze it or describe it well enough for others to do so. In their view, it’s not their job.

Neoclassical welfare economics is not “neutral” as far as normative content, it expresses a certain normative viewpoint and is of interest to those who agree that viewpoint. So is neoclassical welfare economics not really “economics” but “political economy?” Ironically, the ethical issues associated with political decisions about the definition, distribution, use of economic power are purposefully, explicitly excluded from neoclassical welfare economics, which is what makes it an ethical half-theory in relation to reality. Indeed, that’s typically the context for people who advocate the study of “political economy,” a desire to include those otherwise excluded issues, albeit usually introduced in the context of improving positive rather than normative economics. Of course, those are the issues lurking in the background of anti-democracy bad economics in the conservative style, where people try to apply neoclassical welfare economics in realistic setting without taking proper note of the exogenous positive and normative issues. Thus, anti-democracy bad economics in the conservative style may well be considered a sort of political economy, if an opaque, misleading sort, which some propose may be countered by a more real, conscientious, explicit sort of political economy. But as I often point out, anti-democracy bad economics in the conservative style, although perhaps a sort of opaque, bad “political economy,” is not really part of “economics” as much as a clumsy misinterpretation of one part of it, used for purposes of political rhetoric. 

So, is “economics,” as distinct from “political economy,” a “neutral” discipline? Seems rather misleading to say so, in part because neoclassical welfare economics is conventionally part of “economics,” while anti-democracy bad economics in the conservative style isn’t really.

Ethics For The Fairy Land

An interesting distinction between ethics for the Fairy Land of Economic Theory and conventional ethics for reality is that in the former some of the ethical content is expressed not in normative propositions but in the characteristics of the Fairy Land. I thought of the topic the other day when wondering, as I sometimes do, whether the core ethical proposition of neoclassical welfare economics is that we, the observers, should “maximize” total social “utility” under real conditions, or only under certain unreal conditions. 

Recall that “utility” in modern neoclassical welfare economics is all about the expression of preferences, whether defined as preference rank “utility” or inaccessible inner perceptions of satisfaction from preference fulfillment “utility.” If someone with mental illness prefers to throw himself off the astronomy tower under the mistaken belief he will fly, are we meant to agree to let him because it “maximizes” total social “utility,” or not really? If someone prefers to eat an apple we know has been poisoned but she does not, are we meant to agree to let her eat it because it “maximizes” total social “utility,” or not really? If someone has become addicted to some drug and prefers another fix, which we know will likely kill him, are we meant to agree to let him because it “maximizes” total social “utility,” or not really?

One might suppose it would be easy enough to figure out what normative neoclassical welfare economics says. I mean, how else can we evaluate the proposition we should maximize social “utility?” But surprisingly, no, not really. In neoclassical welfare economics, there’s no formal ethical proposition saying we should only “maximize” total social “utility” under a set of arbitrary conditions like people are perfectly rational, fully informed, not manipulated, not addicted, not desperate, etc. However, one might suspect those conditions are built into the structure of the Fairy Land. Can the ciphers, the subjects of Fairy Land, be ignorant, irrational, deluded, manipulated, addicted, desperate, etc.? I’m not entirely sure. Certainly as a condition of “perfectly competitive markets,” we know all about the assumptions of “perfect information” and “perfect rationality.” But what about at the level of just evaluating the normative goal of “maximizing” total social “utility?” 

If economists working with normative neoclassical welfare economics were careful ethical philosophers, they would be very clear and rigorous about the conditions under which we’re meant to evaluate the ethical proposition we should “maximize” total social “utility.” But they’re not, and they don’t want to be. They want to be scientists and mathematicians. They don’t care about themselves or anyone else evaluating normative propositions. And that’s really the whole problem with what I sometimes facetiously call positive normative economics. One can’t do proper ethics without clearly specifying the factual premises involved in one’s argument. Real ethics are understood to apply to reality, even if not all characteristics of reality are formally introduced as factual premises. One may fairly bring up considerations that apply in reality. Ethics for an unreal, artificial Fairy Land are different.

Economists want to play with arbitrary mathematical models of their own design, which purveyors of bad economics then use slyly, opaquely, inappropriately, to make dodgy ethical or normative statements about the real world, and many economists simply don’t care. There’s no one really watching what they’re doing, keeping tabs on them, keeping them honest, speaking up for real neoclassical welfare economics. Well, maybe not no one, exactly. A few. But there should be more. If you’re committed to doing normative economics, playing ethical arbiter for society, get serious about philosophy, ethics. If you want to be a scientist, engineer, mathematician, drop the normative content, the ethics, and learn to talk right.