Some Fundamental Differences between Ludwig von Mises and Nassim Taleb
In my previous post we saw some of the areas where Taleb’s writing is strikingly Misesian: the hidden value and emergent properties of spontaneous orders, the insightfully realistic view of uncertainty and risk-taking, and the aggressive despising of econometrics and other unproper uses of math in the domain of economics.
So far the similarities. As for the differences, most of them are quite profound.
To begin with, Mises's work, as was more customary during his time, spanned many more domains than Taleb’s. Not only an economist or a political scientist, Mises wrote intelligently on philosophy, epistemology, and history. Taleb, primarily a mathematician with a huge appetite for classics and poetry, restricts his writing to topics surrounding his One Big Idea: the poorly-understood subject of uncertainty and the mistaken models, practices and opinions that flow from it. The NYT best-selling book that made Taleb a household name, The Black Swan, explores the impact of under-appreciated tail risks.
In Fooled by Randomness he describes how during his years as a trader his risk-approach made him an object of derision among his more high-returning colleagues — until the various financial crises in the 1990s (Peso crisis of 1994, Asian financial crisis in 1997, the Russian Ruble conundrum in 1998) wiped them out (“they blew up”, in Taleb’s words) and laid the foundation for his financial and intellectual wealth .
While neither is a stranger to lots of writing (Mises's most famous books — Theory of Money and Credit, Human Action, Theory and History, and Socialism — amount to around 2,500 pages, comparable to Taleb’s Incerto series), Mises’s writing is wide and comprehensive, Taleb’s hyper-focused with lots of illustrations derived from ancient texts and trading experience alike. Using the Oxford philosopher Isaiah Berlin’s playful analogy, we might say that Mises is a Fox and Taleb a hedgehog.
Very similar to his objection to econometrics, Mises objects to probability theory in social sciences because economic or political actions are not homogenous events. Class probability, as elaborated by Mises’s brother Richard von Mises, is essentially meaningless; while knowing something about the class of events (say, rolling a die) is interesting, it tells us nothing about the individual event. In discussing this, Rothbard even maintained that “the Mises view demonstrates that all use of probability theory in social science is illegitimate.”
Applying probabilities to the science of human action is to Mises a category error. To Taleb, it’s a rudimentary and epistemic sample-selection error; on the basis on past observations, we simply cannot — unlike Wall Street commentators or your average econometrician — reliably draw inference on the set from which the observations are drawn. Mises rejects probability theory in the social sciences; Taleb rejects the use of the central limit theorem that underlies standard Gaussian distributions.
There’s a further disagreement between Taleb and Mises when it comes to epistemology, particularly Popperian falsificationism and the use of psychology in economics. Mises doesn’t believe that psychology can provide insights about economics, whereas Taleb is squarely on the behavioral train. In Theory and History Mises writes: “Most of [the schools of experimental psychology] are even of no use to praxeology, economics, and all the branches of history.” (p. 264)
And a few pages later he writes: “All that thymology can tell us is that in the past definite men or groups of men were valuing and acting in a definite way. Whether they will in the future value and act in the same way remains uncertain.” (p. 273)
In contrast, Taleb’s trading strategies relied on people underestimating and therefore irrationally underpricing tail-risk events, events that — when they ultimately occurred — allowed him to profit immensely. The very same biases that behavioral economists love to advance underlies Taleb’s explanations for our improper use of statistics in life and markets alike. In chapter 3 of Fooled by Randomness Taleb tells us that his two most influential role models were Einstein and Keynes — but Keynes “the probabilist,” not the amateur macro economist. Taleb was attracted and informed by the "irrational" behavior and folly of man. His idolizing Charles Mackay’s poorly-researched 1841 book Extraordinary Popular Delusions and the Madness of Crowds gets at the very difference between Taleb and Mises: Mises frequently thinks very highly of the layman in contrast to the Ivory Tower (read: socialist) intellectual; Taleb comes across as denouncing intellectual and layman alike — the educated man simply has more ways of fooling himself than the layperson, but behavioral fools they all are!
As for methodology, several things connect Karl Popper to the Austrian school. First, he was a student of Mises and well-versed in the economic ideas thriving in early 20th century Vienna. Second, Hayek’s abandoning praxeology for Popperian falsificationism tied the two doctrines closer to each other. On the other hand, Mises strongly objected to Popper, maintaining that there were meaningful synthetic a priori truths about the world that we could derive from first principles. Taleb, on the contrary, is as pure a Popperian as they come. In chapter 7 of Fooled by Randomness he admits to an “extreme and obsessive Popperianism” that he only acquired upon religiously reading Popper later in his career.
While Taleb refuses to be high-jacked by this or that heterodox economic school of thought, maintaining that he’s a statistically-oriented orthodox economist, I don’t think he understands what people mean by “orthodoxy”; Taleb literally specializes in blowing up most orthodox economic thinking. He writes:
I want to take the charlatanism out of economics and there is a way to do it: examine layers of stochasticity. Detect fragility, and remove offending models. […] I just do not like unreliable models that use "some" math like regression and miss a layer of stochasticity, and get wrong results, and I hate sloppy mechanistic reliance on bad statistical methods. I do not like models that fragilize. I do not like models that work on someone's computer but not in reality. This is standard economics. I showed in 1.7 that we cannot use standard deviations and it is not a matter of taste. Being an economist does not mean being a turkey. Yet all economists persist in these bogus methods.
He does have a point that criticizing mainstream does not automatically make him a detractor or follower of some fringe theory. Indeed, he doesn't subscribe to any idea or theory, setting him even further apart from Mises. At one point in Antifragile, Taleb takes economists to task for being too tied to their theories. Discarding all of them, he writes "a theory is a very dangerous thing to have. And of course one can rigorously do science without it." (2014:116)
The Misesian would not approve. Theory, Mises teaches us, is how we make sense of reality — how we know what to look for, how we separate signals from noise and how we understand the causal mechanisms producing the outcomes we observe. Theory is what makes economics a science — all else is sociology or political science or economic history.
In sum, many things unite Taleb and a Misesian-grounded economic and epistemological worldview: knowledge in emerging spontaneous orders, Antifragile systems, objection to misused statistics in the domain of economics. The student of Austrian economics can easily find topics of agreement with this strange public intellectual.
There are also very important differences: The Talebian rejection of statistics is based on sample selections, not category errors as with Mises; Taleb happily rejects theory, which is clearly anathema to Mises; and Taleb is a Popperian, whereas Mises's moral and scientific foundations more closely align with Kant.
Read Taleb with caution, but read him you should.