Qualcomm, National Security, and Patents
stratechery.com
stratechery.com
"There is a certain amount of irony here: the government is intervening in the private market to stop the sale of a company that is being bought because of government-granted monopolies. Sadly, I doubt it will occur to anyone in government to fix the problem at its root, and Qualcomm would be the first to fight against the precise measures — patent overhaul — that would do more than anything to ensure the company remains independent and incentivized to spend even more on innovation, because its future would depend on innovation to a much greater degree than it does now.
The reality is that technology has flipped the entire argument for patents — that they spur innovation — completely on its head. The very nature of technology — that costs are fixed and best maximized over huge user-bases, along with the presence of network effects — mean there are greater returns to innovation than ever before. The removal of most technology patents would not reduce the incentive to innovate; indeed, given that a huge number of software patents in particular are violated on accident (unsurprising, given that software is ultimately math), their removal would spur more. And, as Qualcomm demonstrates, one could even argue such a shift would be good for national security."
IMO this is the more controversial claim he makes and nobody discusses it. I think I am in favor but I'm no legal expert. We just take it on faith that the patent system is the way the world works, but what if it's only the way the world -used to- work?
This is a silly argument that only really favors large incumbent tech companies of a certain kind, internet businesses. It basically eliminates anyone that doesn't derive their moat from network effects or high capital intensity or any business that focuses on a specific part of the value chain away from the final good or end user, such as ARM or Qualcomm. The world is better off because ARM's business model exists. The world is even better off with Qualcomm. There's large antitrust issues with Qualcomm that's being litigated by regulators and Apple, but that doesn't mean they shouldn't be able to patent their technology.
It would certainly be better for technology in the short run, but the outcomes for human civilization could be catastrophic since competition is so damaged any time big players have so much control that they can undermine the free market.
No, the solution isn't to eliminate patents. It's to reduce the term of patents and copyright in general to more reasonable lengths of time. 20 years is at least 5 generations in most technology fields. If we want to stimulate tech, then we should reduce a hardware patent to 7-10 years, and a software copyright to 14-25 years. We should not reward rent-seeking behavior that does not provide any value to the economy.
Keep in mind, too, that copyright is what backs the GPL and other free software licences. You thought Tivoization was bad? If we set the expiration date to 5 years, then the 4.0 Linux kernel would be eligible for proprietary use by 2020.
Then, after that, an entity must actively pay an exponentially increasing rate of fees to retain the copyright.
Trademarks, as a consumer protection, should just have a low, flat fee, on a fixed basis (with early renewal allowed).
Software patents are different beast than other patents. In the realm of hardware patents and IP are enable profits for fabless R&D.
It's true that patent system requires overhaul:
* A some form of compulsory licensing for all patents after 5 years. Pricing trough auctions or arbitration. IP owners can't decline to license but they can still enjoy license fees.
* Software patents with much stricter requirements and very low priced compulsory licensing (discourages patenting when profits are low). No license fees for open source software.
Patents are a mechanism for rewarding useful inventions.
The definition of "useful invention" is difficult to get right, and the current US patent system is quite a ways off from a reasonable definition.
However, I'm pretty sure the most societally beneficial definition of "useful invention" is not "perfect fodder for the ideal VC-funded startup" (aka fixed costs with exponential user base growth).
The author is correct that if an invention is perfect VC fodder, then patents aren't necessary. But the author is incorrect in the implicit assumption that any "useful invention" is necessarily perfect VC fodder.
>> the current US patent system is quite a ways off from a reasonable definition.
But the road to effective IP law in the United States starts with understanding how invention happens. And "all useful inventions are VC-backed product-oriented B2C internet companies" is not a good model for how invention happens for the vast majority of useful inventions. Even in the past decade.
Without IP and the artificial scarcity enforced by it, capitalism simply cannot function for digital goods. In particular, "digital capital" cannot be accumulated, because all digital goods can be replicated infinitely.
I understand property law as a means of dealing with real scarcity but it seems so antithetical to capitalism to conjure up some laws to create scarcity.
That is precisely what Enclosure(pretty much the birth of capitalism) was. Land was made artificially scarce by concentrating it into the hands of the few, creating a landless working class that had to move to the cities and work in the factories to earn their living. At the time Britain also had something akin to an internal passport system, so that movement of labor between parishes was also restricted.
For a bigger, well researched exposition of this, see Kevin Carson's essay "The Iron Fist behind the Invisible Hand".
The R&D vs dividend decision is just about the expected ROI and risk of R&D vs Qualcomm's cost of capital. (Like any other investment decision). Note that the decision is a spectrum from "raise money and do R&D" to "pay a huge dividend and do no R&D".
The disagreement between Qualcomm execs and Jana (or Broadcom) is that the former say they can deliver returns over their cost of capital by investing in R&D, while the latter don't think so.
Patent reform (assuming it makes patents less attractive), would only bolster the investor case, as it would make new investment in R&D less attractive. (Ignoring 2nd and 3rd order effects)
[1]: Once you figure out how much R&D to do, the licensing vs selling vs spinning off is largely a financing question. If you need more cash to invest today, selling some patents replaces issuing stock. If you need cash to pay a dividend, you can give shareholders your patents directly (i.e., do a spin off) or sell the patents and give them cash.
I agree he clearly points towards patents as being an issue, but it's not clear to me he doesn't care about Qualcomm --- he clearly knows a lot about them.
Our trade policies with China are grossly unfair and I agree that they are effectively self-sacrificial. China is extremely protectionist and economically nationalistic, and is benefiting greatly by playing this off against America's openness and globalism. We keep expecting them to come around and play our game, but as has become apparently with recent political shifts in China (no more term limits, etc.) they aren't and they won't. They're not playing the globalist game. They're playing for their own national interests.
It's interesting to imagine their reactions when foreign governments intervene to block American companies acquiring their native companies.
Eg. the Dutch government blocking Qualcomm's takeover of NXP. (EU stalling the acquisition to investigate antitrust issues is different.)
Hate these type of moves.
How have you established that this is the most likely outcome?
1. There's too much damn consolidation going on, and any barrier to more is good, IMHO. Preventing consolidation is pro-competitive, and I support that.
2. While this decision isn't directly about China, that country has long been pretty egregious in blocking acquisition of its domestic companies (IIRC, a 50-50 partnership is the best you're permitted as a foreigner). It makes no sense to be open to acquisitions by countries who aren't willing to extend the same rights to you.
3. National security is a thing that shouldn't be sacrificed for high-roller business deals. I'm all for preserving American capabilities in these areas for that reason.
4. I don't really give AF about any American corporate whining about acquisitions getting blocked and other stuff like that.
As for your question, my understanding is a Broadcom acquisition of Qualcomm would change the market such that Chinese competitors would be able to force themselves deeper into the American market. So it's more about the second-order effects of the deal rather than the first-order ones.
Even a broken clock (current WH admin) is still right twice a day.
How would they force Qualcomm to continue its R&D for 6G even if they stay on their own?
How would they know, even if Qualcomm continue its R&D, those tech will be made into 6G standard, and not any Chinese competitors? Or any European Allies, like Ericsson or Nokia.
Broadcom is doing a leveraged buy-out via debt. So they need some sort of quick cash to avoid being swamped by debt. With that sort of debt load, the short term trumps the long term because if you don't deal with the debt in the short term, there is no long term to worry about.
"How would they force Qualcomm to continue its R&D for 6G even if they stay on their own?"
They don't, but Qualcomm has a good history of ignoring activist investors who call for them to split their chip-making and patent trolling arms.
"How would they know, even if Qualcomm continue its R&D, those tech will be made into 6G standard, and not any Chinese competitors?"
They don't, but they do know that Qualcomm has successfully done so in the past, and that it won't happen if they don't even try.
Qualcomm generates a lot of IP, but don't they also implement that IP in their chipsets? Does "patent trolling" now just mean enforcing their IP rights?
And that's really only the tip of the iceberg for the shenanigans that Qualcomm has engaged in. By spreading FUD, they were instrumental in ensuring that the US used their CDMA standard rather than the (arguably superior) European GSM standards which they didn't have any patents on.
IOW, patent trolling is legal. Qualcomm's patent trolling was so egregious that they've paid billions in antitrust fines and lost lawsuits.
Qualcomm doesn't licensee ONLY the wireless FRAND patents, they licenses everything from CPU to software design , battery, antenna etc patents.
Now the problem is they dont allow you to choose which one parts to license. They lump it together, and they say you are guarantee to be using one of our CPU / Antenna / Battery patents we have "invented". So just paid for the whole thing.
While you can put a price on FRAND patents, you cant fairly price the other patents which is not part of it. And, just on the defence of Qualcomm, they do actually do R&D and invent these stuff, so patents trolling isn't exactly a fair word to them.
According to Apple CEO Tim Cook spoken on record, for every 6 dollars of wireless patents, 1 dollars goes to 5 companies that is Huawei, Ericsson, Nokia, etc, and 5 dollars goes to Qualcomm.
Personally I think they are charging too much. But we dont really have a way to solve this pricing problem. So legal becomes the only choice.
No insight into standards adoption so I can't help you on point 2.
I have a question about this. If the US government believes Chinese dominance in standard would be a threat to US national security would it not be fair to assume the opposite is true? Would US dominance in standard setting would pose a threat to other countries' national security?
The claim is that this would effectively reduce investment in wireless tech innovation by US companies. This would make foreign investment more effective/more likely to dominate.
I don't know enough about this domain to really judge the article's reasoning (though it was certainly an interesting read), but as a US citizen, if our government held the interests of foreign powers equal to our own in its decision-making process, I'd be very concerned. Actually, something along these lines has been in the news a lot since Trump took office.
Imagine a card game with the object of getting the best hand. Three communal cards are placed on a table face up. China and U.S. each hold their own cards which they don't show each other.
Now imagine that China, the U.S., and a set of international researchers acting in good faith all interact to decide which communal cards are placed on the table face up.
The researchers can give whatever constraints they want-- maybe they want all face cards, or all the same number or suit. The U.S. and China can "play along" while at the same time making choices within that set of the researcher's constraints to maximize the chance that they have the winning hand.
U.S. and China can even guide the researchers away from one set of constraints to a different set if it improves their odds of winning.
So if U.S. gets to guide the constraints exclusively they greatly increase their chances of winning.
If you believe this (or a variation of this idea) then it's reasonable to consider the possibility that China is using, or could possibly use, economics as a weapon.
Today:
US GDP per capita: $60,000
Japan: $38,000
If China does everything right, somehow avoids any recessions or depressions, sees no major consequences from dictator Xi, magically stops accumulating debt at a torrid pace - in about 20 years they'll finally catch up to the US economy in size.
Given their population base, maybe it is inevitable; it won't be easy however. The US has been a comparatively very well oiled economic juggernaut for 150 years, and it's still rolling, commonly outpacing far smaller economies in growth. No country in world history has put together an economic stretch like the US has since the Civil War. China has a lot to prove about being able to manage something like that, especially under a power hungry dictatorship and utilizing a command & control tilted model. Filling in economic slack is the easiest part of growing an economy, that's where all the talking heads got Japan so wrong. Today all the talking heads are extrapolating China's easy fill-in-the-slack growth forward perpetually. They're going to be wrong again, the question is to what degree.
I think this should absolutely be a concern for US national security. But I also think there's something they're not telling us here. A lot of this sounds like the US government may already have its own backdoors in Qualcomm modems, and they don't want to give that up.
Whether that's real or not, the #1 concern is still that China would be able to backdoor any wireless equipment in the US from factory.
I don't see any reason why that shouldn't be taken seriously, especially when China requires by law backdoors in Chinese products and services.
> The US government worries that 90% of the modems will be produced by China (Broadcom/Qualcomm + Huawei + ZTE).
> There is a certain amount of irony here: the government is intervening in the private market to stop the sale of a company that is being bought because of government-granted monopolies.
wait, what the author expect the US government to do? To give the deal green light so bigger monopolies can be created by such merges? Maybe Qualcomm should learn something and start talking to banks to acquire Intel. Just borrow a couple hundred billion from banks, like what Broadcom did, what can possibly go wrong?
"It is ironic that the US is concerned about the patents. Patents only have weight because governments give them weight - They could declare the patents public domain at any time."