Repair economics
A Socratic walk-through of repair economics — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why is it so often cheaper to replace a broken device than to mend it?
A toaster fails. The repair shop quotes more than a new toaster costs, and the quote is not a swindle — it is an honest hour of a skilled person's time. Your grandparents would have had it mended without a second thought.
The reflex is to blame the manufacturers: things are built to fail, built not to open. Some of that is true and we will get to it. But notice that the puzzle exists even for an old, well-made, entirely repairable object. Something moved underneath the whole comparison, and design is not the first thing that moved.
Reasoning it through
REASONING #Set the two options side by side as sums. Replacement costs one price: the new item. Repair costs three things — diagnosis, labour, and a part. Ask which of those has changed over seventy years, and in which direction.
The new item has become dramatically cheaper in real terms. Automation, enormous production scale, global supply chains, manufacture shifted toward lower-wage regions, and the ordinary learning effects of making the same thing ten million times. A mass-produced good is one of the few things whose cost falls year after year.
Now the repair. Diagnosis is a person thinking. Labour is a person working. Both are priced at the wage of a skilled worker in your country, and those wages have risen, not fallen. Worse, repair resists every trick that made manufacturing cheap. A factory makes identical units in sequence; a repairer receives one unit at a time, each broken in a different way, each requiring investigation before any work begins. There is no scale to exploit. Economists call this pattern cost disease — a sector with little productivity growth becomes relatively more expensive as the sectors around it get cheaper, even though nothing about it got worse.
So the ratio moved, and it moved without anyone deciding it should. That alone accounts for a great deal of what we observe.
Then add design, which sharpens it. Parts glued rather than screwed, so opening the case destroys it. Components soldered to a board and no longer socketed. Failures sold only as modules — you cannot buy the failed capacitor, only the whole board that carries it, at a price approaching the device. Proprietary fasteners, absent service manuals, no published schematics. Software that pairs a part to a specific unit, so a genuine replacement is rejected. Diagnostic complexity that has moved from a visible mechanism to firmware.
Each of these adds labour hours or removes a cheap part — pushing on exactly the term that was already the problem.
The analogy
THE ANALOGY #Think of tailoring against off-the-peg clothing. A garment is now made in a fraction of the time it once took, but taking one in still needs a person at a machine for an hour. Nothing about alteration became worse — the two activities simply stopped being priced on the same scale, and the seam that was once obviously worth mending stopped being so.
A tailor can always open a garment, whereas much of the modern repair problem is that the object cannot be opened at all — so design restrictions are a second, independent force on top of the wage arithmetic, not just more of it.
Clarifying the model
THE MODEL #The refinement that connects the two halves: labour is the binding cost, and design choices matter chiefly through how many labour hours they add. That is why a five-minute fix on a costly appliance survives while a two-hour fix on a cheap one cannot, whatever the parts cost.
Now the honest part, because this is where the argument usually goes wrong in both directions.
Some of the design is deliberate. Restricting spare parts to authorised channels, pairing components by serial number, and withholding manuals are choices that protect service revenue and new sales, and manufacturers have defended them under pressure rather than abandoned them. Calling that an accident is not credible.
But some of it genuinely is a side effect. A phone that is thin, sealed against water, and shock-resistant is easier to build with adhesive than with screws and gaskets. Integrating twenty components into one chip makes the device cheaper, smaller and more reliable — and simultaneously makes component-level repair impossible. The very miniaturisation that lowered the replacement price is part of what raised the repair cost. Both stories are true at once, and separating intent from consequence in any particular product is genuinely contested rather than obvious.
There is also a term neither price contains. The material and energy cost of manufacture, and the disposal cost afterwards, sit outside the comparison a household makes — which is the standard argument for policy rather than persuasion. Right-to-repair rules take that route: requiring spare parts to be available for a period after sale, requiring manuals and tools to be sold to independent repairers, and in France requiring a repairability score on the product itself. They work by changing the arithmetic, not by asking people to ignore it.
A picture of it
THE PICTURE #How to readRead each point by dropping to the horizontal axis for what a replacement would cost you, and across to the vertical for how many skilled hours the repair needs. Repair survives in the lower right, where the item is dear and the job is quick, and dies in the upper left, where the item is cheap and the job is slow. The two forces in the argument move a point in different directions: falling manufacturing cost slides everything leftward over time, while sealed and integrated design pushes it upward — and the toaster is the ordinary case of an object that used to sit lower and further right than it now does.
What became clearer
WHAT CLEARED #Repair did not become uneconomic because things got worse. It became uneconomic because manufacturing kept getting cheaper while an hour of skilled human attention did not, so the two sides of a comparison that once balanced quietly came apart. Design decisions then leaned hard on the same term — some of them chosen to, some of them an unavoidable consequence of making the object cheap and small in the first place.
Where to go next
ONWARD #- Why cost disease hits repair, care, teaching and live performance alike, and what that implies for their prices.
- Whether modular design can survive commercially when integration is what makes a product cheap.
Key terms
TERMS #| Term | What it means |
|---|---|
| Cost disease | the tendency of sectors with little productivity growth to become relatively more expensive as productivity rises elsewhere. |
| Module-level replacement | servicing by swapping a whole assembly rather than the individual failed component within it. |
| Parts pairing | software that binds a component to a specific unit, so an otherwise genuine replacement part is refused. |
| Right to repair | policy requiring makers to supply spare parts, tools, and documentation to owners and independent repairers. |
Every term the collection defines is gathered in the glossary.