The Cartel That Killed the Lightbulb

In 1925, GE, Philips, and Osram convened in Geneva to solve a problem most manufacturers only dream of — their product worked too well — and emerged with a Swiss testing bureau, a fine schedule in Swiss francs, and a mandate to engineer the common bulb to die on the dot of 1,000 hours.

In Bay 6 of Fire Station 6 in Livermore, California, a hand-blown carbon-filament bulb manufactured by the Shelby Electric Company of Shelby, Ohio, has been burning continuously since 1901 [S5][S6]. Guinness World Records certifies it as the longest-burning light bulb on Earth [S5]. It was donated to the Livermore-Pleasanton Fire Department at the turn of the century and still hangs there now — dimmed to roughly the equivalent of 4 watts from its original 60, the filament so spent it barely glows [S5][S6]. It does not burn brightly. But it burns.

The Centennial Light exists as proof of what a bulb can do when no one is trying to limit it: carbon-filament technology, the Livermore bulb demonstrates, can outlast entire human lifetimes [S5][S6]. Tungsten-filament bulbs — the kind that replaced carbon designs in commercial production — are capable of similar longevity under the right conditions [S1]. Twenty-three years after this particular bulb was installed, in January 1925, executives and engineers from the world's largest lamp manufacturers gathered in Geneva to ensure that capability remained theoretical [S1][S4].

On 15 January 1925, representatives of Osram (Germany), Philips (Netherlands), Tungsram (Hungary), Associated Electrical Industries (Britain), ELIN (Austria), Compagnie des Lampes (France), Tokyo Electric (Japan), and General Electric's international operations incorporated Phœbus S.A. Compagnie Industrielle pour le Développement de l'Éclairage — a Swiss holding company designed to run until 1955 [S1][S4]. Members held shares proportional to their lamp sales [S1][S4]. The membership was genuinely global from day one, covering, as Krajewski documents from the cartel's own records, dozens of countries across multiple continents [S1]. The cartel's stated purpose, embedded in its French corporate name, was the development of the lighting industry [S4]. Its actual mandate was the opposite.

The problem these companies faced was one most manufacturers would envy: their product had gotten too good. Tungsten-filament bulbs had become so reliably long-lasting that customers were buying too few replacements [S1][S7]. The engineers assigned to fix this were not cutting costs or simplifying manufacturing. They were tasked with making a product that already worked perform worse [S1]. This was, as Krajewski documents from the cartel's own archived records, the founding charter of planned obsolescence as an industrial strategy [S1].

The instrument the cartel created was the "1,000 Hours Life Committee," which required member companies to submit sample bulbs to a central Swiss testing laboratory [S1]. The target was explicit: bulbs should last 1,000 hours, no more [S1]. Manufacturers whose bulbs lasted between 1,750 and 2,000 hours were fined 20 Swiss francs per 1,000 bulbs sold; fines escalated for bulbs that outlasted that threshold even further [S1]. A 1929 table in the cartel's internal documents published the exact penalty schedule — invoices, effectively, for the industrial sin of making something too durable [S1].

What the committee discovered was that engineering reliable failure is harder than engineering reliability. Making a tungsten-filament bulb last as long as possible follows well-understood principles [S1][S2]. Making one die predictably at exactly 1,000 hours — not 800, not 1,200 — required a more precise understanding of failure modes: specific filament thicknesses, controlled draw-ratios, calibrated crystallization rates [S1][S2]. The committee had to solve the engineering problem of death on a schedule. It turns out that's genuinely difficult.

According to the cartel's own internal records, as analyzed by Krajewski from archived documents, the average lifespan of bulbs shipped from member factories dropped from approximately 1,800 hours in 1926 to just 1,205 hours by fiscal year 1933–34 [S1][S2]. The committee achieved its target within two years and enforced it for fifteen [S1][S2]. The commonly cited figure of a 2,500-hour pre-cartel baseline represents what the technology could achieve — the Centennial Light still burning in Livermore is the evidence — but not what was being routinely shipped before 1925 [S2][S5].

The arrangement was designed to run for three decades [S1][S4]. World War II ended it early. The cartel dissolved in 1939–1940 as wartime coordination among companies headquartered in belligerent nations became impossible [S4][S7]. The U.S. Department of Justice filed an antitrust complaint on 27 January 1941, but the trial was suspended at the request of the Secretaries of War and Navy and did not resume until March 1946 [S3]. In 1949, the U.S. District Court for the District of New Jersey found General Electric had violated the Sherman Antitrust Act, in part because of its activities within the Phoebus cartel [S3].

Not every institution read the evidence the same way. In 1951, the UK Monopolies and Restrictive Practices Commission reported to Parliament that 1,000 hours represented a legitimate technical trade-off rather than deliberate planned obsolescence, arguing that "there can be no absolutely right life for the many varying circumstances to be found among the consumers in any given country" and dismissing the allegation as "misconceived" [S4]. The Commission's argument: brighter, more efficient bulbs inherently burn out faster, so 1,000 hours was a reasonable engineering compromise [S4]. Krajewski's counter, drawn from the cartel's own fine schedule, is difficult to dismiss: you don't invoice your own members for making bulbs that last too long unless longevity is the variable you are deliberately suppressing [S1].

When the cartel dissolved, something unexpected happened: nothing changed. The 1,000-hour expectation, embedded across decades in consumer markets and manufacturing processes, persisted as a voluntary industry norm [S7]. No ongoing coordination was required. Individual manufacturers had independently internalized the replacement-cycle revenue model — they had learned to profit from the rhythm of engineered failure, and the rhythm continued without a conductor [S7]. The cartel dissolved; its logic did not.

The second irony arrived decades later in the form of LED bulbs. Rated at 15,000 to 50,000 hours or more, LEDs represented a genuine engineering leap — the kind of longevity the Centennial Light's carbon filament had achieved the hard way, now available in efficient, solid-state form [S1][S5]. The problem was that a century of consumer experience had been spent learning that "long-lasting" was a marketing claim that couldn't be trusted [S1]. The Phoebus cartel's deepest legacy was not a fine schedule or a Geneva holding company. It was a trained suspicion in the global consumer market that durable products don't stay durable — a suspicion the cartel had carefully cultivated, and that LED manufacturers then had to spend years of effort working against [S1][S7].

The Shelby Electric Company bulb in Livermore still glows, having outlasted the cartel by more than eighty years [S5][S6]. It burns at roughly 4 watts now — too dim to illuminate much of anything [S5]. But it has made the argument the 1,000 Hours Life Committee spent fifteen years trying to suppress: that the light, if left alone, does not go out.