I thought it might be worthwhile to go over the issue of false factual premises in normative neoclassical welfare economics versus false simplifying assumptions in positive neoclassical economics again. Pretty sure I’ve done it before, a number of times, and I consider it a bit of a sideshow compared to other issues in normative economics, but I keep reading funny commentary relating to the issue, such as a recent exchange about how odd it is we’ve been hearing the same criticisms of neoclassical economics based on these false simplifying assumptions (positive economics) / false factual premises (normative economics) for at least the past fifty years or so with no evidence of any advancement of the conversation. That commentary was from an economist suggesting critics of neoclassical economics consider giving it a rest, but I think it’s rather more appropriate as a criticism of academic economics, which has had plenty of time to address the criticisms and clear up the confusion but, alas, appears to have done very little in that area.
In positive economics, the presence of simplifying but false factual assumptions is generally accepted because of the understanding the model will be empirically evaluated as a whole, and some concessions may be made to practicality, tractability, etc. This, of course, gets to the whole distinction between a falsifiable, incremental scientific theory designed to reflect reality and a “model” that can be studiously, purposefully false in every particular. I’ve mentioned that issue before and surely will again, but another day.
The great danger, of course, is a sort of backward reasoning in which, if such a model proves empirically useful, any false simplifying assumptions are then considered true. In technical philosophical terms, that’s called having your cake and eating it too, pseudo-scientific style. The assumptions in question are either true or false. If we’re going to assess that issue, we must use whatever methods we have to determine the facts of the matter. A model with those assumption beat competitors? Nice. But that’s the only information we have? Because we just said five minutes ago a model can be highly evaluated in terms of empirical prediction even though it contains false, simplifying assumptions.
One might argue the use of simplifying assumptions in economics is simply an excuse to do bad science, and real science isn’t about making things easy for the investigator or muddying the water of empirical fact. Be that as it may, other issues may also be involved. Like what? The particular needs of normative economics, of course. The eternal bane of economics as a science in the persistent conflation of positive and normative issues one finds in economics. It creates problems here, as in so many other areas.
False factual premises like “perfect information” and “perfect rationality” simplify normative arguments no less than positive models, they eliminate potentially problematic situations, but their status is entirely different because ethics, of course, is not science. If one presents a normative or ethical argument with false factual premises, it’s simply inapplicable to the real world. One may evaluate and approve the conclusions for the real world, but one’s approval evidently involves some basis other than the inapplicable argument. For example, if someone is about to eat a poisoned apple he or she doesn’t realize has been poisoned, but you do, an ethical proposition based on “perfect information” isn’t approximately correct, it doesn’t simplify anything, it’s just inapplicable, inappropriate, incorrect.
Applying normative propositions to the real world that are based on false factual premises from the Fairy Land will never be correct, appropriate, honest, simple. It’s bad philosophy, bad ethics. Oh, as an economist, you don’t do philosophy? You’re not interested in ethics? You’re a scientist, like a physicist? Glad to hear it. In that case, stop telling people what’s socially “optimal” and dispensing policy advice. Explain what you’re doing when you do positive analysis, internal logic checks, of normative or ethical arguments based on random normative inputs and random, false factual premises. Because there may be some confusion about that, and that confusion may be generating a certain amount of conflict. In other words, you may be doing bad economics. Just saying.
