My father owned one washing machine for twenty-two years. It broke four times that I remember, and four times a repairman came, opened a panel with a screwdriver, replaced a belt or a pump or a switch, and left it running for another five or six years. The machine was heavy, ugly, and by the standards of anything sold today, embarrassingly inefficient with water. It was also, in a way that has become almost impossible to describe to anyone under thirty without sounding sentimental, a machine that assumed it would be fixed rather than replaced.
I don’t think this was a virtue of the era so much as an accident of the technology available at the time. Mechanical systems are, by their nature, legible: a belt is a belt, a bearing is a bearing, and a competent person with the right tools can generally work out what has failed by looking at it. The shift toward integrated electronics, proprietary firmware, and components bonded rather than fastened together didn’t happen because manufacturers set out maliciously to make things unfixable. It happened because those methods are cheaper to manufacture at scale, and because for a long time, nobody built the friction of unrepairability into the price a customer actually saw.
Repair as a designed-out feature, not an oversight
What has changed in the last decade is that the unrepairability stopped being incidental and became, in a meaningful number of cases, a design choice defended on its own terms. Batteries glued rather than screwed into place. Diagnostic software locked behind manufacturer-only tools, so that even a competent independent repair shop can’t tell you what’s actually wrong with a device, only an authorized technician can, at an authorized price. Parts pairing, where a replacement component, even a genuine one, is rejected by the device’s own software unless it’s been registered through the manufacturer’s system first.
The right-to-repair movement, which has produced legislation in a growing number of jurisdictions requiring manufacturers to make parts, tools, and documentation available to independent repair shops and consumers, is often framed as a consumer protection issue, and it is one. But I think the more important argument is a quieter one, about what kind of relationship a person is allowed to have with the objects they depend on. A washing machine you can open, understand, and fix is a machine you have some authority over. A washing machine that must be returned to the manufacturer, or replaced entirely, the moment a single sensor fails, is a machine that owns a piece of your authority instead.
What repairability actually costs, and who pays it
None of this is free. Manufacturers correctly point out that designing for repairability, accessible fasteners instead of adhesive, modular components instead of integrated boards, standardized parts instead of proprietary ones, adds cost and can work against efficiency, weight, and water resistance targets that consumers also say they want. A phone built to be opened with a screwdriver is, all else equal, a slightly thicker, slightly less water-resistant phone than one sealed at the factory. That tradeoff is real, and it’s reasonable for a company to make it explicitly, provided the customer is actually told the tradeoff is being made, and provided repair, when it’s needed, remains a genuine option rather than a theoretical one gated behind cost and inconvenience calibrated to make replacement the obviously easier choice.
That’s the test I’ve come to use, in the absence of a better one: not whether a manufacturer has made repair difficult, some difficulty is inherent to sophisticated machines, but whether they’ve made it difficult on purpose, as a way of directing a customer back toward a purchase rather than a fix. My father’s washing machine repairman is retired now. The trade he worked in still exists, but it exists at the mercy of decisions made in product design departments, years before any specific machine breaks, about whether a repairman is meant to be able to help at all.