Economics as usually taught is not a positive science at all. To be fair, no science is. If we claim that scientists should stick to the facts, this implies they should give up the goal of explaining or predicting real world events. This is true even in physics. Why has the pen I just dropped fallen to the ground? Because it is subject to the law of gravity (duh!). But why do we accept this explanation? It is because most of us have never left earth and therefore have never experienced an exception to the law of gravity and thus find it safe to assume that the law will hold for all objects dropped. Thus, our acceptance of scientific laws is ultimately an act of faith and this faith can be corroborated more or less by experience, but the law can never be accepted as such on the basis of experience alone.
Economists’ assumptions, however, are very different. The workhorse model of modern macro-economics for instance assumes ’a large number of [profit maximizing] identical firms, […that] hire workers and rent capital in competitive factor markets, […] sell their output in a competitive output market[, and] are owned by the [rational] households, so any profits they earn accrue to the [large number of identical] households’ (Romer 2012: 49-50). In sharp contrast to assumptions in physics, none of these assumptions are corroborated by experience. Firms are not identical, but instead try to stand out from the competition. Neither are households and in the light of much behavioral economics and psychology alike, the assumption of rationality does not appear to be very tenable either. If you have ever spent time in the real world, you also know that competitive markets are few and far between and if we all shared in companies’ profits, the developments Pikkety describes could never occur. Romer might as well write: ‘Ignoring reality completely, we may assume…’
Now, if these were just simplifying assumptions to be relaxed later (comparable to assuming friction away when discussing gravity), this lack of realism would just be a temporary matter and more advanced models would regain their positive footing. But as I stated before, we have known ever since Sonnenschein (1972), Mantel (1974) and Debreu (1974), that some of these assumptions cannot be relaxed at all without sacrificing the solvability and/or stability of the model. So we are indefinitely stuck with absurd assumptions that have no basis in reality and can never be relaxed.
Reasoning from those absurd assumptions however, we can ‘prove’ that markets tend to optimal allocative efficiency, a unique equilibrium that optimizes societal welfare. If you accept this conclusion, it seems to follow that society is best of if the assumed conditions can be realized in reality. But once you have reached that conclusion, you are engaging in wishful thinking: a normative agenda that outlines not what reality really is like, but what it should be like to reach optimal outcomes. This in itself already has a normative ring to it, but once you realize that reality cannot even in principle conform to all of the assumptions, it follows that belief in the policies such models prescribe, requires that the absurdities of the model are somehow immaterial to the applicability of their conclusions. So, even though the existence of differences in preferences, for instance, theoretically invalidates the existence of stable equilibria, we are somehow required to convince ourselves that this theoretical impossibility has no real bearing on reality. Now, that is a big leap of faith. All in all, much of neoclassical economic modelling hides a normative agenda, whose acceptability and applicability hinges on our acceptance of the impossible as either possible or negligible.
None of the above is exactly rocket science, I think. Why then do so many economists cling to the belief that economics is a positive science that studies ‘man as he really is’ (Hirschman 1977; cf. Mandeville 1724 or, more recently, Levitt and Dubner 2009)? Their conviction probably stems from historical inertia. When Adam Smith wrote The Wealth of Nations, he was up against the clergy and the widely held belief that avarice (which we now call profit maximization) was a mortal sin that should be held in check and fought tooth and nail, if we were to prevent chaos. This moral battle was notoriously difficult to win, especially for those living in poverty (which in Smith’s day was true for the vast majority). Pushing back against the clergy, Smith conjectured that if we were all guided by self-interest and avarice and interacted through markets, the market would not only constrain our vices sufficiently to prevent society from spiraling into chaos, but would actually ensure superior outcomes. This was a very liberating argument. No longer did people need to constrain their impulses, live in poverty and hope for salvation in the after-life. Instead, they could do what was best for them without fearing God’s wrath, as they could redress their attention to their self-interest as an act of biblical love for one’s neighbor. From this perspective it can be understood why economics at the time could be hailed as the study of ‘man as he really is’. Economics had after all replaced lofty moral ideals with the idea that giving in to one’s impulses is actually beneficial to all and doing so surely comes more naturally to most people than being a saint.
The early successes of the new world order Smith envisioned seemed to prove him right: wherever capitalism took hold, economic growth jumped up (Heilbroner 2012; cf. Fouquet and Broadberry 2015). This strengthened the belief in Smith’s conjecture and later on prompted attempts to prove it mathematically. From the outset, then, neoclassical economic modelling was a teleological endeavor in which the assumptions were at least as much premised on the conclusion as the other way round. The conclusion just had to be that communities of self-interested individuals interacting in markets were bound to achieve superior outcomes and the investigations and modelling took the form of establishing the assumptions that would generate this conclusion. Thus, realism of assumptions or establishing an objectively true theory was never the primary goal. But if you mutilate your assumptions to the point of absurdity in order to prove the (potential) superiority of a particular society, you are engaging in ideology rather than science.