My greatly esteemed friend Vernon Smith turned 86 years of age yesterday. Vernon is, among other things, the leading figure in the development of experimental economics, for which he shared the Nobel Prize in 2002. For various methodological reasons, I have never been a fan of experimental economics. To me, it represents the sort of positivism that Austrian economists reject for good reason.
However, it has struck me recently that Austrian economics itself may be characterized as a kind of experimental economics in that its basic tool of analysis is the thought experiment—the careful thinking through of hypothetical alternatives. In ordinary experimental science, such as physics, chemistry, and Vernon’s experimental economics, one tests hypotheses (whatever their theoretical or other source and however plausible or fantastical they may bea priori) by observing the relation of changes in X to changes in Y, all other relevant things being held constant. The trouble is that, in all cases, one can never be sure that all other relevant things have been controlled. Hence, every experiment is inconclusive; its findings are only as good as the controls imposed by the latest experimental setup, and an experiment with better controls may overturn one’s current conclusion.
In Austrian economics, in contrast, one deduces the relation of X and Y from first principles so compelling that they are accepted as axioms, especially the Action Axiom (that people purposefully use means to obtain ends desired for the removal of their felt unease). In Austrian thought experiments, all other things whatsoever are held constant by a mental strait jacket that immobilizes them completely. Therefore the deduced relation of X and Y is always as strong as the axioms of analysis, provided that no logical mistakes have been made.
Of course, such thought experiments cannot serve as necessarily effective forecasting devices because the real world is in constant flux—that is, the “other things” are constantly changing. However, no matter how much they might change, the derived relation of X and Y remains an element of the real world’s operation. For example, if the money stock is enlarged, then all other things being equal, the general purchasing power of a unit of money declines. In an empirical situation, the money stock may be enlarged, yet a unit of money not lose general purchasing power because other things (e.g., the public’s demand for money to hold) have changed in an offsetting way. Nonetheless, we may still conclude that in the actual situation we observe, the general purchasing power of a unit of money—whatever it may be—would have been greater had the money stock not risen. Thus, we understand the workings of the world, at least in some part, even if we cannot predict empirically how the world or some specific element of it will change amid its constant flux.
So, all hail Austrian economics—the Queen of the Experimental Sciences. Long live the properly grounded thought experiment!
[Cross-posted from The Beacon.]