The agreement
In December 1924 representatives of the major incandescent lamp manufacturers met in Geneva: General Electric from the United States, Osram from Germany, Philips from the Netherlands, Associated Electrical Industries from Britain, and a number of other national producers. What they created was registered in Switzerland as Phoebus S.A., and it was a cartel in the full sense — it divided the world into territories, allocated quotas, pooled patents, and set standards its members were bound to.
One of those standards was the service life of a lightbulb, fixed at 1,000 hours.
This was not a target or a guideline. Members submitted bulbs to a central laboratory in Switzerland for testing, and a member whose bulbs lasted materially longer than the standard was fined. The penalties were calibrated so that complying was more profitable than making a better product.
The effect is measurable. Average bulb life fell from around 1,800 hours in 1926 to about 1,200 by 1933, and by 1934 almost nothing on the market exceeded 1,500 hours. Over the same period member sales rose substantially.
Wikimedia Commons — KMJ, CC BY-SA 3.0
The part that is usually overstated
The Phoebus cartel has become the standard illustration of planned obsolescence, and in that role it is usually described as the suppression of an everlasting lightbulb. The evidence for the cartel is strong. The evidence for the everlasting bulb is not, and the reason is a genuine engineering constraint rather than a corporate excuse.
In an incandescent lamp, light comes from a filament heated until it glows. The hotter the filament, the more of its output falls in the visible range and the more efficient the bulb is in lumens per watt. The hotter the filament, the faster it evaporates and the sooner it fails. Efficiency and longevity are traded against each other directly, through the same variable.
So a bulb engineered to last 2,500 hours is a dimmer, less efficient bulb than one engineered for 1,000, and the owner pays the difference in electricity for the life of the lamp. For most of the twentieth century electricity cost far more than bulbs did, which means a shorter-lived, brighter, more efficient bulb was genuinely the better buy for the person using it. A thousand hours is a defensible answer to that optimisation, and engineers outside the cartel reached similar figures.
The Centennial Light in Livermore, California, is the exhibit most often produced against this. First illuminated in 1901, it is cited as proof that a bulb can last indefinitely, and it has indeed been burning for more than a century.
It is also no longer doing the job of a lightbulb. It was originally rated at 60 watts and now emits roughly the light of a four-watt bulb — a dim orange glow. Its survival is attributed to a thick hand-made carbon filament, a high-nitrogen atmosphere inside the glass, very low power, and the fact that it is essentially never switched off, which removes the thermal shock of switching that kills most lamps.
Every one of those is a reason it is not a counterexample. A bulb that is dim, hand-made, and never switched off describes a museum piece, not a product the cartel could have sold instead. It demonstrates the tradeoff rather than refuting it.
The claim that does hold: they were testing for it
There is one element of the planned-obsolescence reading that the record supports directly, and it is easy to lose while correcting the rest.
The cartel did not merely agree on a figure and leave members to interpret it. It built an enforcement apparatus: bulbs went to a central Swiss laboratory, results were recorded, and a member whose product ran long was fined. An organisation that fines its members for excessive durability has, as a matter of plain description, made durability something to be prevented rather than achieved.
That is a different claim from "they hid an eternal bulb", and it is the one with documents behind it. Whatever the right engineering answer was, the cartel's structure guaranteed that the answer would stop being revisited, because any member who improved on it was penalised for doing so. Standards are ordinary and useful. A standard with a fine attached for exceeding it is not a standard; it is a ceiling.
So what was the offence
Stating it precisely makes it worse rather than better.
The objection to Phoebus is not that 1,000 hours was the wrong number. It is that the number was set by agreement among competitors rather than by competition, and enforced by fines. That removed the thing that would have tested whether it was right. If a manufacturer had found a way to get longer life without sacrificing efficiency — through better filament metallurgy, gas filling, or manufacturing consistency — the cartel's structure gave it no way to profit from that and a penalty for trying.
The cartel also did the ordinary things cartels do, which is where the consumer harm is least arguable: it carved up markets, fixed prices, pooled patents to keep entrants out, and suppressed competition in a product every household bought. That is the documented injury. It does not require an everlasting bulb to exist.
Why the overstatement matters
This entry is unusual in this archive for being a case where the evidence supports a real conspiracy and the popular telling still gets it wrong — in the direction of making it larger.
That has a cost beyond accuracy. "They invented a bulb that lasts forever and hid it" is checkable, and it does not check out, so anyone who looks into it finds the famous claim unsupported and can reasonably conclude the whole story was overblown. The actual story — a documented international cartel that fixed prices, divided markets and fined its own members for building products that lasted too long — survives any amount of scrutiny. Trading the second for the first is a bad deal, and it is made constantly.
The cartel broke up around the outbreak of the Second World War. The agreement had been due to run considerably longer.