The Idiot Index

The ratio and what a high one means

Elon Musk's pricing diagnostic decomposes a part into its constituent raw materials and asks what it would cost at the physics floor, then compares that number to the price actually being quoted. A commodity part should cost close to its raw materials; a large gap between the two means the price is almost entirely process, tooling, and margin rather than anything to do with the underlying physics, and in Musk's own phrasing, a high ratio means whoever set the price is either an idiot or working with idiots. The instinct behind it is a refusal to accept a market price as a given fact rather than as a decision made by a specific supply chain that can be examined and, if the ratio is bad enough, replaced.

The canonical proof

The example that made the diagnostic doctrine at SpaceX is a single actuator. A supplier quoted one hundred twenty thousand dollars for the part; Musk's budget for it was five thousand; Steve Davis built it in-house for three thousand nine hundred dollars, and Musk's entire response was "Ok."¹ A component with a raw-materials cost in the tens or low hundreds of dollars had carried a five-figure price because the aerospace supply chain priced small-volume custom parts on inherited process and margin rather than on physics. The diagnostic made that gap visible and turned it directly into a build decision. Applied across an entire vehicle, the same test is the mechanism behind SpaceX building far more of its own hardware than a typical aerospace company would, driving each high-ratio part toward its raw-materials floor rather than continuing to pay someone else's markup for it.

When the index loses to a different rule

The diagnostic has a documented limit, and Musk's own behavior supplies the counterexample. In Ashlee Vance's biography, he refuses to pay twenty-five thousand dollars for a part he believes can be built for two thousand, a clean application of the index, and in a separate instance approves sixty thousand dollars in jet fuel to fly a part to Hawaii solely to save the team a single day of work.² The two decisions look contradictory only if cost is assumed to be the only variable. Early SpaceX operated against a daily burn rate of roughly one hundred thousand dollars and an expected revenue, a decade out, of roughly ten million dollars a day, which means a single day of delay costs a future ten million dollars plus the unlocked value of an entire team sitting idle. Against that number, sixty thousand dollars is trivially the correct call. The two rules are not competing, they are ordered: the idiot index governs whenever cost is the binding constraint, and time overrides it the moment time becomes the binding constraint instead.

Where the diagnostic can mislead

The index assumes materials cost is the correct floor for a part's price, which understates the case for components where the real value is precision, reliability, or certification rather than bulk material; a cheap actuator that fails during flight is not a bargain regardless of what it cost to make. Insourcing every part with a high ratio also trades a supplier's margin for a company's own fixed cost and management complexity, so the index functions best as a heuristic for where to start looking rather than as proof that building something in-house is always the cheaper choice.