Why Decision Science Proves Rory Stewart's Economic Consensus Is Flawed

This podcast episode critically analyzes Rory Stewart's approach to economic debate, arguing that his views are flawed not due to a lack of intelligence or integrity, but rather a significant blind spot rooted in the 'ideology of expertise' and 'homophily.' The host, Barry, contends that Stewart, like many in positions of power, operates within a closed intellectual ecosystem where similar backgrounds and education lead to shared assumptions and 'correlated errors.' This phenomenon, where clever people reinforce each other's mistakes, is contrasted with the scientific principle that accuracy in complex problem-solving arises from 'uncorrelated error,' meaning diverse perspectives making different mistakes that cancel each other out.\n\nThe episode draws on decision science and historical examples to illustrate its points. It references Francis Golton's century-old experiment where the average guess of a diverse crowd accurately predicted a cow's weight, and Lu Hong and Scott Page's research showing that diverse, less individually brilliant teams consistently outperform homogeneous, top-tier teams in complex problem-solving. A key distinction is made between individual intelligence and collective wisdom, emphasizing that groups fail not because individuals lack intelligence, but because they make the same mistakes. Rory Stewart's dismissal of economists outside his 'canonical' definition, such as Gary Stevenson, Grace Blakeley, and Richard Murphy, is presented as a prime example of this homophily, where conformity is prioritized over accuracy.\n\nThe practical implications are profound for economic policy and public discourse. The podcast argues that when the economic debate is dominated by a narrow group, crucial perspectives and 'invisible' data—like the experiences of renters, nurses, or those locked out of cities—are never brought into the room. This leads to a system that is 'calm, confident, and catastrophically blind' to urgent issues like rampant inequality and the need for wealth taxes. The Abraham Wald example from WWII, illustrating survivorship bias, highlights how focusing only on 'survivors' (visible data) can lead to counterintuitive and dangerous conclusions, missing the most critical information.\n\nUltimately, the episode concludes that Rory Stewart's error is not a moral failing but a scientific one: he is 'reading consensus as evidence.' His confidence is produced by the absence of dissenting voices, not a comprehensive understanding of the entire system. The podcast advocates for recognizing that intelligence within a closed system does not protect against bias; instead, it makes bias confident. It underscores the importance of valuing diverse lived experiences as 'hard data' and challenging established thinking to prevent systemic failures in economic policy, which currently ignores millions of valid perspectives due to an entrenched ideology of expertise." "concepts": [ "Homophily

Idiots crash the car, but clever people design a very robust and efficient road that leads straight off a cliff.

Decision science says that actually when clever people come together, they've often had a similar education or similar backgrounds and therefore they share the same blind spots.

Homophily... doesn't make people stupid. But it makes them miss the target. And in fact, they don't just miss the target. They reinforce each other's mistakes or in the technical terms their errors are correlated.

Accuracy in complex problems comes from uncorrelated error.

Groups, they don't fail because individuals lack intelligence. They fail because everyone makes the same mistake. Diversity helps by preventing correlated error.

This is not an attack on experts. It's an attack on the ideology of expertise about who is allowed to talk on the economy, who is allowed to debate.

When the response shifts from this idea is flawed to this person is flawed, you are no longer measuring accuracy. You are measuring conformity which is a textbook signature of homopholy.

The most important information, it wasn't loud, it wasn't dramatic, and it wasn't obvious. It was deeply counterintuitive, invisible.

When you only average the experiences of survivors, the system becomes calm, confident, and catastrophically blind.

Rory Stewart isn't wrong because he doesn't care. He's wrong because he's reading consensus as evidence.

Intelligence inside a closed system isn't protection against bias. It's how bias becomes confident.