After publishing the previous post, I received the following email from Christoffer Stjernlöf, which I thought was worth sharing:
I enjoyed your latest article on the September Jobs Report – these statistics come with interesting connections and assumptions which I lack the training to realise, so it’s always nice to have someone knowledgeable comment on them at a deeper level than what’s typically found in mainstream media.
I especially found the stalling analogy apt, and it’s probably more appropriate than you think! You wrote:
An airplane has a stall speed: if it slows down a bit, it flies a bit slower, but if it slows down too much then it stops flying entirely and falls to the ground.
The stall speed needn’t be the speed at which the plane drops out of the sky – it could simply mean the beginning of an entire flight regime called the stalled regime.
In the normal, cruise regime, the wing produces lift, and if the wing is tilted upwards it produces more lift by virtue of meeting the air more aggressively. This means if the plane slows down (and the wings cut through less air per unit of time), the wings need to be tilted up to compensate and maintain level flight.
What happens in the stalled regime is not that the wing stops generating lift altogether, but that further increases in wing angle start decreasing (!) the amount of lift produced! So if a pilot finds themselves in this regime, and want to increase lift, they need to angle the wing downwards rather than upwards – contrary to normal expectations.
The reason I bring this up is that it mirrors the feedback loops of the economy – on the good side of a threshold, the natural feedback loops keep things somewhat stable, but on the bad side of the threshold, the natural feedback loops instead conspire to make things worse.
The reason we think of the plane falling out of the sky in the stalled regime is that it no longer has positive stability in that regime, so perturbations tend to worsen the condition until the wing indeed no longer produces enough lift to keep the plane in the sky.
For more details, the words to search for are “back side of the power curve”, “region of reversed command” or this excellent text: https://www.av8n.com/how/htm/vdamp.html#sec-stall-intro
The linked text is indeed fascinating.
In his “Politics and the English Language,” George Orwell warns against drawing from “the huge dump of worn-out metaphors,” which are used without regard for the object they notionally describe. Language, to be sure, is full of fossilized metaphors, which are now just words. It’s the dying metaphors that Orwell wants us to avoid. If a term still conveys a comparison to a physical object or concrete situation, one should have the real object or situation in mind, and be sure that the thought being expressed really applies to it. It’s good advice, which I strive to follow.
Here we have the opposite, or the contrapositive, of Orwell’s dying metaphor. What makes for a living and vigorous metaphor is that more careful attention to the physical thing, will give us a deeper understanding of the idea to which it is being compared.
The transition from a domain of stabilizing negative feedback to destabilizing positive feedback (or from dampened to amplified disturbances) is exactly what Leijonhufvud had in mind with the idea of a corridor of stability. Thinking more precisely about what an airplane stall entails, as Stjernlöf suggests, captures this better than a vague idea that the plane just stops flying.
Or in a slightly different vein:
Paul Krugman used to write a lot about how when interest rates are at the lower bound (synecdoche for an economy facing binding demand constraints), normal economic theory gets turned upside down: prudence is folly and virtue is vice. It’s interesting to learn that the idea of an economy having a “stall speed” is, if the metaphor is taken seriously, a description of exactly a situation where moving the levers has the opposite of the usual effect.
