Hacker News new | past | comments | ask | show | jobs | submit login
Phoebus Cartel (wikipedia.org)
126 points by mbroncano 5 months ago | hide | past | favorite | 82 comments

The planned obsolescence of incandescent light bulbs is largely a myth. While there are plenty of real examples of planned obsolescence, with incandescent light bulbs there are engineering tradeoffs between longevity and efficiency. Basically, the longer lasting a bulb is, the dimmer and less efficient it is. If you've ever seen any of the supposed 100 year old bulbs, you'll notice that they're extremely dim. Before, during, and after the Phoebus cartel incandescent light bulbs were (and still are!) standardized to 1000 hours. Why? Because that happens to be a good trade off between longevity and efficiency. It just doesn't make sense to waste an extra $5 on electricity to make a $1 light bulb last longer.

You have always been able to buy long life incandescent light bulbs, and still can to this day. They're called rough service bulbs, made for ovens and closets and areas used infrequently. People just never used them for general purpose lighting because they're dim and power hungry.

If you read the report by the British monopolies commission quoted in this article, you'll notice that they came to the same conclusion [1]. Pop science articles and Reddit love to bring out the Phoebus cartel over and over as an example of planned obsolescence, but always completely ignore the factors that go into making a good incandescent light bulb.


> Before, during, and after the Phoebus cartel incandescent light bulbs were (and still are!) standardized to 1000 hours. Why? Because that happens to be a good trade off between longevity and efficiency.

I would believe this if the cartel fined members based on the inefficiency of their lightbulbs relative to light produced (lumens per watt), and fined according to this.

Or allowed longer life bulbs if they met a lumen per watt efficiency target.

Or even if they added a minimum life to ensure that consumers got quality, and fined producers if they produced bulbs that were too low-life.

They didn't, they fined members who produced lights that lasted longer regardless of bulb efficiency.

Because the decision to limit bulb life to 1,000 hours had obvious financial benefits for cartel members, I think it's pretty generous to think that this policy was solely based on wanting what was best for consumers.

Other factors affect light output like frosted glass, which they didn’t want to forbid. Setting a target lifetime seems like a simple way of only regulating the lifetime-efficiency tradeoff while giving freedom on the others.

I figure this is because at the time there wasn't an accurate way to measure light output en masse. This was before photoresistors and solar panels were invented. Also, why would manufacturers still target that figure even today in absence of the cartel?

They were still an anti-competitive cartel.

The Pheobus Cartel wasn't just about bulb lifespan, but also bulb price. They also used their power to increase the price of bulbs. The fact that they converged on a practical trade-off in terms of life span & energy costs doesn't change their anti-competitive nature.

I agree with you there, but most people bring up the Phoebus cartel as an example of planned obsolescence and rarely being up the price fixing aspect. Besides, the cartel didn't last for very long regardless.

> The fact that they converged on a practical trade-off

Is it even desirable to mass produce at one specific point in the tradeoff curve? Wouldn't society, or at least engineers, appreciate being able to choose where in the curve you land?

Only if the choice is clear and reliable. Right now there seems to be a lot of difference between LED bulb longevity, but it's basically a dice roll if you get a lasting one or not.

The IEEE has a good article on how quickly Phoebus failed: https://spectrum.ieee.org/tech-history/dawn-of-electronics/t... If not for Pynchon, no one would have ever heard of Phoebus - it's not even one of the successful cartels.

One amusing tidbit not mentioned in the IEEE article I noticed while looking at some congressional testimony: the Scandinavian competitors apparently did not just undercut Phoebus, but signed up first to get Phoebus to bankroll construction of their lightbulb plant and then began undercutting them with it!

> it's not even one of the successful cartels.

which are ?

De Beers or OPEC spring to mind.

The effectiveness of OPEC (at least since 1980) is a matter of some debate.

Somehow I thought De Beers was a monopoly, not a cartel.

OTOH, if I could discover a way to make a lightbulb that lasts 1100 hours, for the same electricity and brightness, then consumers would buy that, and I would make money!

But the cartel colluded to prevent such innovation. The British monopolies commission effectively colluded with them by first decided that there should be a single lifetime as a standard: with that decision already made, then of course they could easily claim to have reached a decision based on the facts provided by the companies they were investigating.

And I would happily pay an extra $5 in electricity for a bulb that reduced my need to change them. Should we ban the iPhone because you could easily use a cheap Android instead?

>OTOH, if I could discover a way to make a lightbulb that lasts 1100 hours, for the same electricity and brightness, then consumers would buy that, and I would make money!

This will never happen with incandescent bulbs because the lifespan vs efficiency trade off is fundamental to how they work. More power means more heat which causes it to fail faster. However, this did end up happening eventually with the invention of LED bulbs and now many big companies formerly in the cartel (including Phillips) are in the process of shutting down or selling off their lighting division because it's no longer profitable to sell light bulbs when LEDs last so long.

>The British monopolies commission effectively colluded with them by first decided that there should be a single lifetime as a standard

Just as companies today agree on standards like USB or WiFi, back then they also agreed to a standard bulb life and efficiency. This means that one company's bulbs are interoperable with another's and you don't end up with a house full of mismatched bulbs.

>And I would happily pay an extra $5 in electricity for a bulb that reduced my need to change them.

This doesn't account for the extra greenhouse emissions and environmental effects of running inefficient bulbs. Also, long life bulbs were never banned, even when the cartel was active. You've always been able to buy them if you wish, most people didn't because they had a poor lumens/watt

It could happen with a different filament.

As mentioned there's an inverse trade-off between life and brightness, so if you found a way to make an 1100 hours lightbulb, you could equally make a brighter lightbulb than the competition.

Except you couldn't, because the cartel has educated customers that 60W = medium brightness, 100W = very bright. Brightness has been linked to electricity use.

That also means that you couldn't make a bulb as bright as a 100W bulb that only uses 90 watts: if you advertised it as a 90W bulb, people would think it was less bright.

That's now solved by stating the actual light output in lumens as well as the power consumption in Watts (and it should have always been thus).

Essentially, every compact fluorescent and LED light bulb I've bought in recent years states both the power consumed and light output (which means that if one bothered one could calculate the overall efficiency of the bulb).

Also, for those who still have difficulty with the new specifications, manufacturers state the new bulb's light output to be equivalent to an incandescent of 'x' Watts (i.e.: the light output of say a 20 Watt compact fluorescent ≡ to a 100 Watt incandescent bulb). This is now rather irrelevant where I live as incandescent bulbs have been banned for so long that everyone has forgotten about them.

Your arguments notwithstanding, the fact that the Phoebus Cartel essentially operated in secrecy implies that there was more to this than just optimizing (standardizing) incandescent light bulbs to 1000 hours.

True, one has always been able to buy long life incandescent light bulbs and I've done so myself to use in certain industrial situations. However, I'd contend that that has little or nothing to do with planned obsolescence.

Moreover, there's more than a whiff of the Phoebus Cartel still about today with light bulb manufacture. Let me give you an example: when compact fluorescent bulbs first came out some four decades ago, I installed many of them. The principal brands I bought were Philips and Wotan and the majority of them lasted well in excess of 10,000 hours. I recall losing track of the life of one low power bulb of ≈8 watts after it had passed 50,000 hours operation (that is, after having being in service for over 25 years and used between 4 and 5 hours per day over this time, eventually it was replaced not because it had failed but rather to replace it with a bulb of higher wattage).

Take the current situation: almost every compact fluorescent that I've bought in the last 10 or so years has only lasted about 1000 hours despite the fact that the packets they came in usually advertised their life to be a nominal 8,000 hours or so. The fact is that none of them ever reached anywhere near these published figures. To make matters worse, often these bulbs would break (usually around the base) and sometimes the coiled glass would fall to the floor (no doubt dispensing its small amount of Hg everywhere in the process).

Moreover, a while back, so called 'green' environmentally friendly compact fluorescent bulbs were all the rage. They were advertised as having less mercury and thus more environmentally sound. What a joke that turned out to be. All the Greenies loved them because they contained less mercury. Unfortunately, most never knew how the bulbs worked anyway and thus they never realized that reducing their mercury levels below a certain concentration would dramatically reduce the bulb's life to such an extent that the overall result was a net loss for the environment. Right, environmentally friendly bulbs actually had an even shorter service life (hence, Watt for Watt, they produced much more environmental e-waste over the long run)!

Clearly, all manufacturers were disingenuous by not stating the consequences of lowering the mercury levels in compact fluorescent bulbs. They deliberately played on the fact that consumers weren't cognizant of the facts, thus the lowering of manufacturing standards was a win for them as there were few complaints. Manufacturers not only made savings by including less mercury but also other manufacturing standards were lowered such as using lower grade capacitors and the maximum working voltages on capacitors and semiconductors were reduced to much closer (lower) tolerances which made these bulbs much more prone to failure from over-voltage spikes, etc.

In the end, many more bulbs were consumed than there should have been. Concomitantly, we've ended up with even bigger piles of electronics waste—and the net amount of mercury used was actually higher than it would have been if the bulbs had been designed to have a service life of >10,000 hours (as had been the case some 20 years earlier). Incidentally, the same unacceptable logic is now being applied in the manufacture of normal fluorescent lights† in that current tubes have a much, much shorter lifespan than those made 50 or so years ago).

Yep, it wasn't kosher to go around advertising the fact that 'our bulbs have a little more mercury but they will last you many times longer'. Right, manufacturers must have been laughing themselves sick over the fact that they could double-dip their profits by playing on the fact of the public's current, almost irrational, phobia over Element 80 and that reducing the amount of it in bulbs automatically led to vastly increased bulb sales. Moreover, they gloated even more when some governments made compact florescent bulbs mandatory but did not also introduce mandatory manufacturing standards to ensure their long operational life.

The overall net effect being bulb manufacturers scoring two wins whilst consumers and the environment both scored nil. Right, the incarnation of the Phoebus Cartel is alive and well.


† Several years ago, I was involved in relocating a factory and the old site used T12-type fluorescent tubes the majority of which were made by Crompton Parkinson and were over 50 years old and still in good working condition!

We decided not to relocate those old lights and we replace them with new T8-type fluorescents of Eastern manufacture at the new site. These new tubes are now already starting to fail. The lifespan difference between the old and new types is so startling that I reckon it ought to be part of a study into changes into manufacturing techniques. These new tubes have to represent one of the most quintessential examples of planned obsolescence ever. Frankly, it's absolutely outrageous that this can happen.

In a bit of synchronicity, Dubai has sponsored development of more efficient and longer lived (via underdriving) LED lamps, and exclusivity enforcement means that you are prevented from being able to buy these longer-lived lamps: https://hackaday.com/2021/01/17/leds-from-dubai-the-royal-li...

This is a very different trade though, the Dubai lamp trades a higher upfront price (these lamps are more sophisticated and cost more to make, so unsurprisingly they also cost more to buy) for a longer lifespan and improved efficiency.

Whereas incandescents were trading lifespan versus efficiency. You can make 5000 hour incandescents, but your "60 watt" 5000 hour lamp will put out far less light, so it costs the same to run but it's not bright enough, then you buy the 100 watt version, now it's bright enough but you're paying two thirds more money to run it!

A hypothetical 5000 hour incandescent only makes economic sense if your electricity is basically free (greedily maybe it makes sense if you don't pay for it, e.g. rental inclusive of electricity bills) whereas a Dubai lamp makes sense regardless of electricity price, because over long enough periods the increased lifespan saves you money anyway.

these lamps are more sophisticated and cost more to make

Actually, they're simpler than a lot of other LED lamps, because they're not dimmable and use only a capacitor to limit current followed by a small post-regulator.

The driving electronics often fails before the LEDs, but they've minimised that with these.

Because they're under-driven, they actually need considerably more (typically four times as many) light emitters than a cheaper alternative, and those cost money. It's true that a capacitive dropper is a cheap way to make LED lamps work, but that's how the LED lamps you're offered in a typical store today already work, so doing that isn't saving them money over competitors, and since it's a long-life product they didn't skimp on the power supply design, check this Big Clive video for example:


> The driving electronics often fails before the LEDs, but they've minimised that with these.

I find a regular supply of cheap LED bulbs with that issue in the bin.. They really should put out a stupid modular model so you can swap the circuit for 1$

... or if I am personally more interested in the difficulties of changing the bulb than the electricity it costs to run it.

Veritasium made a video that reviews the history and presents the topic in an engaging way.

"This is why we can't have nice things" - https://www.youtube.com/watch?v=j5v8D-alAKE

The magical longer life and more efficient incandescent bulb was invented of course: The halogen bulb. It got around the previous physics barrier: That the hotter the filament (and therefore whiter and brighter the light), the more quickly it deteriorated to the point of failure.

To me, hot as they ran (fire hazard if used improperly) halogens were the pinnacle of artificial light quality. Of course I've changed all mine to LEDs now. The LEDs give adequate light and are better in every other way. But the light from halogen bulbs was beautiful.

Standardizing a 1000 hr bulb life vs standardizing a 30% commission to host an app, is it really a coincidence or are cartels just better at hiding it now?

Market forces often push participants towards an equilibrium value, so it’s hard to say. I don’t know enough about economic theory to know if there’s a way to tell.

The 30% commission has a long history. There are precedents going back to the early days of games consoles. Both the iTunes music store and Steam opened in 2003 charging 30% or very close to it (iTunes tracks were 99c so it doesn’t work out exactly).

It’s hard to argue the 30% in either of those cases were a matter of abuse of market power. Steam was struggling to establish itself against the incumbent distribution channels. The iTunes rates were negotiated with the major labels who were notoriously hard deal makers and very wary of online sales channels. They held all the cards is their negotiations with Apple and yet they seemed to think 30% was fair. So at that point I don’t think it’s possible to make a credible case that the 30% rate was extortionate.

If 30% was abusive, we’d expect to see it act as a brake on adoption of the App Store by developers, but is there really any evidence for that? On Android have any of the smaller stores tried to differentiate on price at lower than 30% to attract developers, or has Google tried to woo iOS developers with lower rates? If 30% was abusively high we’d expect to see something like that happening.

A percentage doesn't make sense, because cheap apps and expensive apps require exactly the same effort for hosting them, and there is no risk involved.

So the fact that these competitors use the same pricing model and the same price has a bad smell.

If the actual cost of service was pro rated across apps this would basically eliminate free and low cost apps. In fact how would you price this in the first few years of the App Store when it lost money? Furthermore it would make it incredibly expensive and therefore risky to develop launch an app. They would likely have had to charge hundreds of dollars, perhaps thousands per app in the early years, stifling innovation and strangling freemium or free offerings. The App Store would be a very, very different place today.

This would be like supermarkets charging product manufacturers a uniform fixed rental for shelf space. No supermarket would stock toilet paper, it’s too bulky and low revenue, or if they did they have to charge a lot more for it.

So Apple and Google decided that expensive apps should subsidize the cheap ones? That sounds almost like a governmental move. I'd rather have the real government regulate the market I'm operating in.

Have Microsoft and Google standardized on the 30% commission?

Or is it more that "Apple stuff" is not directly substitutable with "Google stuff", and that Apple has a monopoly on "Apple stuff"?

Yes, they have, please see this chart


Pretty much everyone has, with the occasional exception.

Interesting, thanks.

I wonder how much of this is deliberate collusion, as opposed to an equilibrium where companies all realize that it's more profitable to match each others' fees than to try to compete on price.

If other app stores charge 30%, you know that developers are willing to pay 30%, and there's not much to be gained by reducing your own rate below that.

I wonder how much of this is deliberate collusion, as opposed to an equilibrium

I saw a theory in the comments on HN once that companies can 'collude' without communicating with each other. If memory serves, it was different than just achieving an equilibrium. Maybe it's enough of a theory to have a name? Hopefully someone who knows what I'm thinking of can chime in.

Edit: I may be thinking of "tacit collusion". https://en.wikipedia.org/wiki/Tacit_collusion - "Tacit collusion is a collusion between competitors, which do not explicitly exchange information and achieving an agreement about coordination of conduct."

They read each-other's press releases, they reasearch competitors, employees and mamagers move between companies. There is no way for them to not communicate.

Just because bosses didnt sit down a and sign a deal in a smokey room doesnt mean no decision was made

Yes, although typically tacit collusion is introduced with the example of price matching at big box retailers. The setup in this case is somewhat different, although the outcome is still an equilibrium in which all sellers "agree" to keep prices high (and above marginal cost) without directly communicating.

That is remarkable, and I wonder if there are explanations other than a cartel.

Steam and GOG too

Wasn't Steam one of the first to implement such pricing? Or have they followed the lead of somebody else before them?

Steam has a much better case to argue, given that they're a fiscally independent company with no board of directors or shareholders to satisfy. Since they don't get monthly cash infusions, they need to do two things: shoot for an aggressive growth model, and shower the end user in features. In my opinion, they've succeeded in both regards: their 30% cut is pretty large, but it's justified given the scale of the operation they run. Plus, their generosity with things like cloud save, video streaming, and their assortment of community offerings, it's hard to really claim that Steam doesn't deserve their cut. If anything, I consider them the premier example of what a "premium" CDN pipeline looks like.

It’s unlikely there was collusion around app store commissions. It’s more likely a case of “how much can we get away with charging?” and existing competitors in market providing a proof point of what is palatable to developers. That is a somewhat stable state until someone like Epic comes in with a lot of money and a desire to quickly capture market share with a lower commission.

I'd be willing to defend Apple's 30% cut if any of it was justified. Instead, my experience working in the walled garden has been gruelling. Xcode is a pathetic disaster at this point, but Apple seems to be unwilling to extend any functionality to the editors I actually use. Getting an app through the guidelines is an exercise in arbitrary debasement. Their phone support is useless too, with most of them directing me to forum posts that basically say "don't do whatever you're doing". I think Epic has every right to offer their lower commission, because their service downright sucks. I still plan to buy games from Steam for the foreseeable future, because their work on Proton is worth it to me.

I am pretty sure it is established that there was, Amazon and Apple were the first 2 members

Apple and Amazon are far from the first storefronts.

Steam was charging 30% as far back as 2003.

Pynchon's "Byron the Bulb", the great example of American Sublime according to Harold Bloom, is a paranoid story about this very affair https://www.tildedave.com/byron.html

There's a fantastic Throughline podcast episode on the Phoebus Cartel: https://www.npr.org/2019/03/27/707188193/the-phoebus-cartel.

I often wonder is their is a Phoebus equivalent in the SmartPhone market.

I am unsure how we got to a point where the life span of a $900-$1,200 device is measured in months not years is acceptable but is seems to me it would have to involve some conspiracy especially when the technology they replaced or augmented (traditional phones and laptops) where all measured in years, multiple years.

> I am unsure how we got to a point where the life span of a $900-$1,200 device is measured in months not years

Because we didn't get to that point? My Apple phones last 3+ years unless I break them, and only the ones I hold on to for closer to 5 years aren't still working entirely fine when I upgrade; usually all they'd need to keep working for another couple years is a battery replacement.

It is entirely possible, but in the case of smartphones the hardware processes and systems are so new that we're also seeing the effects of "crapshoot integration..." The market is moving so fast that there isn't time to run enough tests to have strong expectations on the ten-year lifespan of the product. Batteries, in particular, are such bleeding-edge technology (and in a space of chemistry so poorly theoretically understood) that every new design is more or less a dice-roll on the long-term performance.

The hallmark of a Phoebus-cartel-style situation would be if smartphones from N years ago from multiple manufacturers were still going strong but newer phones were slower, or had crappier battery life, or broke down more often (consistently year-over-year). I'm not sure that's an easy comparison to make; go too far back in the space and you pass into the pre-iPhone, pre-Android era "smartphones" that were robust as hell but remarkably featureless relative to what came after.

That having been said, Apple has been sued over the planned obsolescence of its smartphones (https://www.macrumors.com/2021/04/08/apple-chile-planned-obs...)

> that every new design is more or less a dice-roll on the long-term performance.

Sometimes things happen to make companies uh take note:


You either had bad luck, are doing something to break them, or purchased particularly bad phones.

Phones last years now although first party software updates are frequently only available for 1-2 years and apps eventually are made that require more recent android versions effectively forcing you to update in 5 years to have access to all possible apps. Games may be more aggressive in their requirements as well.

>> first party software updates are frequently only available for 1-2 year

Bingo, that is the planned obsolesce part. too many people here are focusing on physical viability, not software viability.

From a security standpoint, if it is out of support it is a brick, and the number of vulnerabilities that are discovered on these devices NO ONE should be running a mobile computer with current updates

So sure the device may physically still turn on, and you can still "use" it, it is not something I would recommend.

This isn't how people actually use phones. People upgrade more frequently because they want new phones. The percentage that is considering the security of their device is a single digit percentage.

I have only just replaced my Moto 4g 2014 (£170) with a motog9 power - the battery just would not hold a useful charge.

I have the 7 which I got for almost nothing. FANTASTIC battery but its going to be interesting to replace someday.

My mother-in-law is still using my hand-me-down iPhone 6s, a 6 year old phone that still runs the latest iOS.

If there is, it's a pretty bad cartel seeing as the older devices do not actually break after a year or so :)

If new devices are more likely to break than old ones doesn't that suggest that planned obsolescence has been introduced? Probably not because of a cartel but almost certainly due to plain old greed and a lack of moral integrity.

With Android certainly. I wouldn't expect any Android phone today to still receive updates in 5 years.

Hm. Wondering how planned obsolescence (PO) can be fought.

The first step of solving a problem is defining it.

Maybe if there was a website where enough people self-reported device failures out of spite, wonder if that would help pinpoint PO cases. Incidentally it may solve the fake review issue.

An open source bug reporting tool but for IRL failures. BBB comes to mind.

Just a random and free idea.

> The first step of solving a problem is defining it.

Yep, and the issue is that most people will run from the opportunity to make their relationship with technology more complicated. Most people don't care if you can't replace the battery on the new iPhone because they intend to replace it in 2-3 years. As sorrowing of a statement as it is to say, it's hard to get first-class citizens to care about issues that don't affect them. In other news, the earth is round and gravity pulls downwards.

If your friendly neighborhood government has a fully empowered consumer protection agency, said agency can enforce reasonable warranties on products. If companies have to replace any appliance (or whatever) that fails within 10 (or whatever) years, you can bet they'll make them more reliable.

> you can bet they'll make them more reliable.

Or a lot more expensive. And if you don’t care about the 10 year reliability, you are being hurt by having to pay more.

how about forcing tech companies to state at the time of sale how long they plan on maintaining/ updating their product/program. Also, requiring hardware manufacturers to provide an estimate of the expected lifespan in cycles of their product, and/components. Figures that are too conservative must be punished with the same level of strictness as overestimations in order to maintain the integrity of the law.

Obviously, these laws will only apply to the tech sector, but it's a start.

France has outlawed the practice as part of an environmental law.

Easy to outlaw, difficult to prove unless the law also mandated minimum usable lifetimes & warranties for products.

A company doesn't have to build in some sort of off switch to force obsolescence, all they have to do is use cheap parts to cut costs and put the burden of paying for maintenance on the buyer. Basically, planned obsolescence can be indistinguishable from a product that is cheaply made.

They also outlawed managing cpu utilisation in order to prolong device lifetimes, so apparently it’s illegal to try to extend device lifetime and also illegal not to. Robespierre and Cardinal Richelieu would be so proud.

Yeah, and manufacturers are allowed to assign their own repairability indexes to their products. That's how Apple ended up scoring the Macbook an 8, and the iPhone a 7.

The Omnibus Podcast with Ken Jennings did an episode on this last year.


not sure how much things have changed. A couple years back i had a few brand new, expensive led bulbs. I expect my light bulbs to last at least a couple of years. There supposedly long life light bulbs began flickering after just a few months.

Nothing has changed. In the EU the expensive "eco friendly" LED bulbs had a stated lifetime of 2 years.

None has lasted even half a year here, which is shorter than incandescent bulbs. It is all a big fraud.

I have philips bulbs going strong for like 7 years

LG's rather misleading "10 year warranty (small print: on the inverter drive)" on some of their washing machines comes to mind. Your LEDs are fine, it's just that everything else fails heh.

Strangely Youtube suggested exactly this topic on my feed last week and weird enough it isn't the first time I've seen stuff on HN after it popping on there. Just an observation.

Possibly a case of https://en.wikipedia.org/wiki/Frequency_illusion

But then it's also not unlikely that other people got the same YouTube recommendations, after which they did some online research and posted interesting links to HN. Or maybe the causation was the other way around. Who knows?

Back in a day I kept receipts for all bulbs and pestered Home Depot with refunds for failed bulbs.

Can't affect fraudulent cartels but can apply pressure on merchants not to carry flawed merchandise

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact