Normative and Positive

Can we take a step back this week and discuss the normative versus positive distinction? I hardly know what’s necessary to discuss these days, but I have noticed a certain amount of confusion relating to normative and positive propositions, so maybe worthwhile?

Positive propositions are about what is, in a factual sense, not what one supposes ought to be the case or what one ought to do. So a statement about observable reality, or at one step removed science depending on empirical verification, or what the heck, maybe also formal systems as math. The point of a proposition being called positive is it is capable of being refuted with reference to some objective standard. So statements of fact, science, are meant to be assessed relative to shared reality, while statements in formal systems relative to some agreed definitions, rules. Everyone around the world can discuss proposed facts, scientific theories, mathematical or logical problems, and agree on their proper mode of evaluation or assessment. 

Normative propositions are about ought rather than is. They’re humans expressing their own personal judgments about what they or others ought to do based on their subjective moral senses or sensibilities. They’re not really refutable per se, although one may evaluate, discuss, dispute them. Now in a normative context one may attempt to create a false criterion for objective evaluation by simply declaring some theory, book, creed to be dispositive in a normative sense, write it into law, enforce it, etc. However, that’s not agreement of criteria in the philosophical sense. No breaking of prior agreements relating to arbitrary terms or definitions, no bit of illogic, no affront to shared factual reality is involved if someone says he or she thinks it should really be some other theory, book, creed that should be normatively dispositive. It’s subjective opinion.

A while back, a great deal was made of interactions between positive and normative in generally positive or normative theories. For example, ethical theories are normative theories but if they’re addressed to the real world they will involve factual premises about the world, which are positive. Or again, ethical theories typically have some logic to their arguments. Do the positive elements mean ethical theories are not “normative theories?” I wouldn’t say so. The conclusions are normative propositions about one ought to do and must involve normative evaluation at some point. On the positive side, a great deal was made about potential normative influences on the conduct of positive science as far as motivations, interests, imagination, resistance to findings and theories, and so on. I recall a great deal was said about in the philosophy of science circa 1970s. Does it suggest we can’t really call scientific theories as positive? Again, I wouldn’t say so. At the end of the day, they’re about reality and intended to be evaluated on that basis. They’re positive theories, but they can interact with normative propositions via bias, motivations, etc.

The culmination of that bad philosophy was the contention there is no difference between positive and normative propositions, which the scientifically illiterate embraced as justifying their anti-science views and the philosophically illiterate embraced as suggesting ethics are objective facts. One still sees the influence of that bad philosophy especially in the philosophically clueless field of academic economics, in which some suppose neoclassical welfare economics is not a normative theory because it involves math or logic, others conflate it with predictive modeling. A typical example of such confusion in the context of normative and positive economics is conflating “false simplifying assumptions” in the predictive engineering models of positive economics with “false factual premises” in the ethical arguments of normative economics.

My primary concern is normative economics, the normative or ethical arguments made in neoclassical welfare economics about socially “optimal” economic arrangements, policies, what people ought to do, and how that is twisted in normative anti-democracy bad economics in the conservative style. I contend that is the bit of economics that accounts for most of the controversy, confusion and conflict relating to real economic issues and policies, the bit academic economics typically fail to adequately address, the bit fueling anti-democracy sentiment and dodgy ethical thinking. I’m not very interested in positive economics, predictive modeling mostly, although I have some residual interest in the difference between predictive engineering models (and related storytelling) and real cumulative science, from a youthful brush with the philosophy of science. In the old days, if a student expressed any interest in normative economics, they were shuttled into philosophy of science under the conceit economics is a science, so the only normative content must be potential interactions with normative propositions as discussed in the philosophy of science. I didn’t know enough at the time to object, so chalk that up as one more intellectual mugging at the hands of academic economists in the devious old men versus naive teenagers division. But I do now. That’s not the “normative economics” I mean to discuss. I mean the other one: the ethical theory.

Utility

Maybe we can start reviewing some topics in neoclassical welfare economics and associated misinterpretations in anti-democracy bad economics in the conservative style this week. How about “utility,” am I right? Utility is a fun one because it’s the focus of a great deal of idiosyncratic oddness in neoclassical welfare economics and unsurprisingly a great deal of misleading, manipulative rhetoric, word play, equivocation, conflation of issues, in anti-democracy bad economics in the conservative style. An academic economist once told me to not worry what “utility” means because it can mean anything one likes. I thought that funny. To assess the concept in a normative or ethical context one must know the definition and no, not every possible definition fits with neoclassical welfare economics.

In classical economics, “utility” was handled in a fairly conventional way as one might find in the ethical theory of utilitarianism in proper philosophy, so some notion of human welfare that policies might have some “utility” in terms of advancing, so “utility” morphed into “welfare.” Classical economics came under fire as critics wondered if human welfare was being advanced as advertised, and in response conservative economists did what they do best, played word games, in this case re-defining welfare, utility, to ignore commonly accepted observable indications of welfare.

An intermediate formulation translated “utility,” still as welfare, into a sort of internal sensation of satisfaction from preference fulfillment. In that context, “utility” was meant to exist but be inaccessible by definition, so external indications, technology, brain scans, etc., irrelevant. This was the initial context in which a very important principle in neoclassical welfare economics was formulated, which was that the relevant definition of “utility” cannot be applied in interpersonal contexts, that is, no “interpersonal comparisons of utility.” At that point, “utility” as a sort of internal “welfare” was still meant to be defined in interpersonal contexts, it was meant to exist and theoretically be subject to interpersonal comparison, it’s just that there was meant to be no practical, scientific, real way to do it. The point was to remove “utility” from the lion’s share of ethical thinking, which is about resolving interpersonal conflicts of preferences, allocating resources, etc. For that reason, neoclassical welfare economics became only trivially “utilitarian,” really more about opposing utilitarianism. This is the reason I tend to call references to “utility” in neoclassical welfare economics as fake utilitarianism, or another way I often say it is that unlike what is typically meant to be the full ethical theory of utilitarianism, it’s an ethical half-theory (really less than half surely). Comically, it’s not uncommon at all to hear confused critics inappropriately lambasting serious ethical utilitarianism on the basis of rejecting the conclusions of bad economics in the conservative style’s incorrect take on the fake utilitarianism in neoclassical welfare economics. Who said what?

Later and now conventionally “utility” in neoclassical welfare economics came to be seen as simply an indication of preference ranks of a given individual, so indexed by individual. Thus, it is not really defined in an interpersonal context, so a utility(A) and a utility(B), but no just “utility.” Comically, economists seem often confused about which definition of utility they’re using, awkward because the fundamental arguments in favor of socially “maximizing” it differ depending on which definition one is using, which tellingly seems a matter of utmost indifference for most of them. Specifically, one often hears economists who profess to be using preference rank utility discussing the impossibility (rather than the undefined nature) of interpersonal utility comparisons, or even more notably providing schemes to “weight” preference rank utility, add it up across people. Again, it’s indicative of an insincere, unserious use of “utility” as an ethical concept in a philosophical sense. Really, they’re just using the term to express a rather different ethical argument about respecting preferences of individuals, and both definitions get the job done well enough. One understands the waffling. Inaccessible internal “utility” cannot support a plausible ethical theory (imagine dreaming the preferences of a super conductor of “utility could be measured), but preference rank “utility” makes it painfully obvious they're dealing with an ethical half-theory.

Bad economics in the conservative style then plays off the effective removal of utility and welfare from interpersonal ethics by underhandedly promoting certain ethical views on economic power, resolving conflict, allocating resources in realistic conditions, in rhetorically clever ways. Needless to say, the peculiar quality of “utility” in neoclassical welfare economics and the ample opportunities to equivocate on the term as used in serious ethical philosophy and to conflate issues relating to preferences and welfare are of great use in bad economics in the conservative style. I’ll talk more about it in the future, don’t worry about that. 

Perhaps one last indication of the insincerity of academic economists? I suggested some time ago a Don’t Say Utility Challenge, in which I challenged economists to replace the fraught term “utility” with what they meant by it. I assumed for most that would mean replacing “utility” with “preference rank of individual X,” although in some cases I suppose with “inaccessible internal sensations of satisfaction from preference fulfillment.” I suggested it to avoid equivocation, conflation. Do you suppose many took me up on the offer? Have you heard many expressing themselves in those terms? Why do you think not? Perhaps because they often rely a lot on misleading word play to express their own normative or ethical views, to promote bad economics in the conservative style?