The Market for Doom

Roger Partridge's essay on why every generation predicts technological mass unemployment — and why every generation is wrong. A lucid synthesis of economic history, cognitive bias, and the institutional preconditions that let labor markets absorb disruption. The punchline: economists who take scarcity and adaptation seriously have been the optimists all along; "dismal science" stuck to the wrong discipline.

Key Quotes

"Each prediction was a snapshot extended into a film."

Partridge's central diagnosis of the doom merchant's cognitive error. The prediction takes current conditions, removes one category of activity, and projects linearly forward. The actual economy doesn't work that way — it adds what the displacement releases. #concept

"The doom merchant works in two dimensions, the economy in three."

The two dimensions: the visible activity being displaced, and time. The third dimension is the new activity that emerges from the labour and capital released. This is the Bastiat point — the seen vs. the unseen — applied to technological unemployment. #concept

"Horses, who could not retrain, are the only species in the record where the doom prediction came true."

The sharpest line in the essay. Britain's horse population peaked over a million animals and collapsed to near zero. Humans retrain; horses don't. The distinction sounds glib but it's the whole argument in nine words. #pattern

"This is the worry the doom merchant is too distracted to notice: not that AI will be different, but that we will."

The real risk isn't that Schumpeterian adaptation stops working — it's that we dismantle the institutional machinery that makes it work. Property rights, open markets, rule of law, the freedom to start a firm. Venezuela and Cuba aren't counterexamples to the mechanism; they're demonstrations of what happens when you switch it off. #concept

"Economics is not the dismal science. Humanity is the dismal species."

The closing inversion. Carlyle called economics "the dismal science" for insisting that resources are scarce. Partridge argues the real pessimists are the doomsayers who keep betting against human adaptation, and the economists are the ones saying "actually, we'll figure this out — if we defend the institutions that let us."

Key Themes

#pattern — The doomsayer lineage repeats with clockwork regularity. Rifkin (1995) → Ehrlich (1968) → Club of Rome (1972) → the Triple Revolution memorandum (1964). Each "made with confidence, marketed with skill and falsified within a generation." The prediction is always the same shape: identify a displacing technology, project current employment structure forward, ignore the new things people will want and the new ways they'll work.

#concept — Static vs. dynamic reasoning as the root cognitive error. Partridge's most valuable analytical move: the doom prediction is "cognitive before anything else." It's not that doomsayers are bad at economics — it's that they're doing a different kind of thinking. They take a snapshot and play it as a film. The availability heuristic (Kahneman) amplifies the visible destruction and obscures the dispersed creation.

#concept — Human wants are infinite, and that's the engine. Partridge's first law: "Human wants exceed available resources, always, and at every level of material progress." The 1600 peasant wanted shoes; the 2026 citizen wants broadband, dental care, foreign holidays, and yoga. Every displacement releases resources toward newly-visible wants. The nail bar, the software engineer, the professional gamer — none were predicted by central planners.

#person — The optimist economists. Schumpeter (creative destruction), Bastiat (seen vs. unseen), Hayek (dispersed knowledge), Joel Mokyr, Philippe Aghion, Peter Howitt (2025 Nobel laureates). Partridge assembles them as a counter-tradition to the doomsayer lineage, arguing they've been right for two centuries while the catastrophists have been wrong with equal consistency.

#concept — Institutional preconditions are the real variable. The adjustment mechanism isn't magic. It requires "secure property rights, open markets, the rule of law, the freedom to start a firm and the freedom to leave one." Where those hold, adaptation works. Where they don't, it doesn't. The policy implication: protect the institutions, don't try to centrally manage the transition.

#concept — The ATM is the canonical counterexample. ATMs were predicted to eliminate bank tellers. Instead, banks opened more branches, and the U.S. employed more tellers in 2010 than in 1980. Automation reduced the cost per branch, which increased the number of branches, which increased demand for the human work the ATM couldn't do. This is the pattern doomsayers consistently miss.

Critical Analysis

Partridge is very good on the pattern — the historical recurrence of failed doomsaying, the cognitive explanation for why smart people keep making the same mistake. The essay is strongest as a taxonomy of error and a synthesis of economic history. The ATM example, the horse line, and the "two dimensions/three dimensions" framing are genuinely sharp.

Where the essay is weaker: it treats "we" as the variable while knowing almost nothing about what "we" might do. Partridge is vivid about the fifty-year-old compliance officer in 2028, but offers no mechanism for getting from that individual pain to the collective response that might dismantle the institutional preconditions he correctly identifies as necessary. The Venezuela/Cuba examples are instructive but they're extreme cases — the subtler erosion of open markets through regulatory capture, licensing cartels, and incumbent protection is more likely in the developed world and gets almost no attention.

The Acemoglu citation is worth scrutinizing. Partridge quotes him saying "no material impact on employment from AI" in May 2026, but doesn't engage with Acemoglu's actual thesis — that AI's employment effects depend on which direction we steer it. Acemoglu is worried about surveillance AI, automation AI, and AI that replaces workers rather than augmenting them. Partridge lumps him with the "conditional, prospective and policy-shaped" camp and moves on, but that camp is saying something more interesting than "don't worry."

The essay's framing of economics-as-optimism is rhetorically effective but historically selective. The same tradition that produced Bastiat and Schumpeter also produced Malthus (whom Partridge doesn't mention) and plenty of economists who were dismal about specific transitions. The "dismal science" label came from Carlyle's objection to economists opposing slavery — a fact that complicates Partridge's neat inversion.

That said, the essay lands its central point: doom sells. There has always been a market for predictions of catastrophe, and the predictors have always been credentialed, confident, and wrong. The essay is worth reading alongside 2026 Global Intelligence Crisis (Citadel's S-curve rebuttal to AI doomerism) and The Dead Economy Theory (which Partridge would likely diagnose as the latest entry in the doom lineage). For the institutional side, Why We Fear AI makes the complementary argument that AI anxiety is really capitalism anxiety.

Rodney Brooks' four-time-scale framework reinforces Partridge's historical argument from a different angle: if every transformative technology — from domesticated animals to containerization — took 50+ years of continuous at-scale deployment to reshape the economy, then the doomsayer's error isn't just cognitive (snapshot extended into a film) but temporal. They're predicting the economic dislocation of Time Scale 4 from the research demonstrations of Time Scale 1, skipping the two intermediate stages where technologies typically fail, stall, or evolve into something unrecognizable. The horse that couldn't retrain is the exception that proves Brooks' rule: technologies that survive all four time scales do reshape the economy, but on a schedule measured in human lifetimes, not quarterly earnings. Four Time Scales for Technology Development and Deployment

Sources: The Market for Doom: Why We Keep Predicting the End of Work Last updated: 2026-06-21