This One Is True. They Met In Geneva And Shortened The Bulb.

GENEVA — On 23 December 1924, representatives of four lamp manufacturers met in Geneva and agreed to limit the working life of a household bulb to 1,000 hours.

Before that, 1,500 to 2,000 hours was normal and long life was advertised. After it, members who made lamps that lasted too long were fined — on a sliding scale, per lamp, per hour over the standard.

The schedules survive. They are ordinary office paperwork, ruled and totted up, and they are the most chilling documents this newspaper has photographed all year.

Why we are leading with this

Because it is true, and because being true is not what gets a conspiracy attention.

This newspaper receives about eleven hundred letters a year about a tablet, and roughly none about the Phoebus cartel. The tablet is a fragment with an eleven-month hole in its provenance. The cartel is a signed agreement, an enforcement regime with a fine schedule, and a finding by a United States federal court.

United States v. General Electric Co., 82 F. Supp. 753. New Jersey, 1949. Seventy-four pages, free online, and the court states plainly that the company had the power to determine lamp life across the industry and that profit was its main consideration in setting it.

You can read it this afternoon. Almost nobody has.

The part that is genuinely arguable

Here is the counter-argument, at the length it deserves, because a story that leaves it out is doing the cartel's work.

Filament life and efficacy trade against each other. Run a tungsten filament hotter and you get more light per watt and a shorter life. Run it cooler and it lasts longer and gives you less light for the same electricity.

Professor Emeka Sørensen-Bhatt walked our correspondent through the arithmetic. A lamp engineered for 2,500 hours at the same wattage is a dimmer lamp, and over the life of the bulb the extra electricity costs the owner more than the bulbs saved.

So 1,000 hours is not an arbitrary number. There is a real optimum somewhere near there, and it depends on the price of electricity.

He then said the thing that matters.

"All of that would be a perfectly good defence if they had published the calculation and competed. They did not publish it. They fined each other. You do not need a fine schedule to enforce physics."

The bulb in Livermore

There is a lamp in a fire station in California that has been burning almost continuously since 1901.

It is real, it is remarkable, and it is about four watts — dim enough that you would not read by it. It is a hand-blown carbon-filament lamp run far below its rating, which is exactly the tradeoff above, taken to its extreme.

It is evidence that a bulb can last forever. It is not evidence that a bulb can last forever and light a room, and the accounts that use it as the second thing are being careless with something they could check with one telephone call to the fire department, which is what we did.

What this desk takes from it

The best-documented conspiracy on our beat is a hundred years old, was settled in court seventy-seven years ago, is taught in competition law, and generates almost no correspondence.

Our correspondent has a theory about why, and it is only a theory, so it is in the last paragraph where a theory belongs: a conspiracy that has been proved is over. It has a citation and a verdict and nothing left to argue about, and the argument was always the point.