After seeing his interesting new article, I asked Professor Andres Sawicki (University of Miami School of Law) to draft this short essay about his work. — Dennis.
by Andres Sawicki
In a recently published article, I report preliminary evidence supporting a novel view of what patents can do: keep inventive teams together. This evidence suggests that, in addition to their traditional role as incentives for innovation, patents may be doing important work in fostering collaboration in high tech industries.
To see how this works, suppose you’re the founder of a Silicon Valley start-up. After a few years, you’ve found modest success—a product launch, a small core of devoted customers. But it seems clear at this point that there’s no massive IPO exit on the horizon. Instead, your capital is running low and the venture capitalists who financed your firm are getting impatient.
Now Facebook shows up at your door: it wants to hire you and your team of engineers. The catch is that it wants the whole team. Facebook knows how hard it is to find talent that works well together. Plus, both you and Facebook know that if your team doesn’t go as a group, any team member that strikes off on her own is likely to soon become a competitor. What do you do?
One possibility is that you convince Facebook not only to hire your team, but also to buy the entire start-up. That way, Facebook can acquire the rights to any patents flowing from the work the team did at the start-up. These patents can then bind the team together by raising the costs to members of leaving—a departing team member won’t be able to continue working in the path set out by the team’s patents. Thus, while intellectual property is traditionally thought to prevent the entire public from freeriding on a creator’s investment in producing a public good, it can also regulate relationships among team members, as Robert Merges and Paul Heald have separately argued in the patent context, and as Tony Casey and I have jointly argued in the copyright context.
In my most recent article, I support this team-binding view of patents with data from Silicon Valley acqui-hires. Those transactions, illustrated by the founder-Facebook scenario posed above and explored in illuminating detail by John Coyle and Gregg Polsky, are understood to be driven by Silicon Valley norms of cooperation. Facebook and the engineers agree to send some money to the start-up’s equity holders to ensure that the VCs will be open to another pitch from the engineers a few years down the road. This usual understanding of the acqui-hire has little room for IP.
But sending money to the VCs may also be a way to ensure that the start-up’s IP follows the team from the start-up to the buyer. Before the transaction, the start-up will ordinarily have claims to the IP generated by its engineers. Diverting some of the purchase price to the VCs and other equity holders enables the buyer to obtain those claims.
To test the plausibility of this hypothesis, I examined a set of 63 acqui-hires over a two-year period, to see whether: (1) the start-up assigned some of its patents to the buyer; (2) the start-up assigned some of its pending patent applications to the buyer; or (3) the buyer filed, after the transaction, a patent application naming one of the start-up’s principals as an inventor. In 29 of the 63 acqui-hires, at least one of those three kinds of patent transfers occurred. In many of the remaining 34 acqui-hires, there were simply no pre- or post-transaction patents to speak of, indicating that patents cannot completely account for the acqui-hire trend. Still, there was only one transaction in which the start-up retained all of its existing patents and pending applications, and none of the start-up’s principals had been listed as an inventor on one of the buyer’s post-transaction patent applications. In short, when the start-up has rights to existing or future patents, the buyers consistently obtain those rights. This data thus indicates that patents are in fact an important part of the acqui-hire trend, without which it may be difficult for the founders and Facebook to consummate the deal. And, more generally, it suggests that patents are not only rewards for lone inventors, but also tools for keeping groups of inventors together.
Andres Sawicki: patents can then bind the team together by raising the costs to members of leaving—a departing team member won’t be able to continue working in the path set out by the team’s patents.
The major problem with this hypothesis is that other “costs” and “benefits” presented to joint inventors will almost always dwarf the “cost” associated with the remote possibility that one inventor will be “unable” to continue working “in the path set out” by a third-party owned patent on which that joint inventor is named.
People who are joint inventors on patents leave companies, all the time, every day, for any number of reasons. The companies typically own those patents and, if the joint inventor negotiated a decent deal with the company, the joint inventor has already benefitted. If those benefits haven’t flowed to the inventor for some reason, then the joint inventor can negotiate with the company that wants to hire him/her to account for any such “lost” benefits (assuming the companies aren’t engaged in an illegal wage fixing conspiracy).
As for the possibility that a joint inventor will be precluded from working on “a path set out by a patent” on which the joint inventor appears, it’s almost always the case that a departing joint inventor is aware of approaches that are not taught in the patent or the departing joint inventor is no longer interested in working on that project or the departing joint inventor has good reason to believe that, to the extent the patent precludes anyone from working in a broad “path”, the patent is invalid.
A more interesting question regarding the sociology of the silicon valley “innovation” communit would be to look at how often an “inventor” on two patents owns both (1) a controlling interest in a company A that is “working on” the “technology” of patent A; and (2) a controlling interest in a company B that is “working on” “technology” covered by both patent A and patent B. In other words, the “inventor” here is hedging his bets. The follow-up question is: how often do they employees of these two companies know about that arrangement when they are hired?
Your entire last paragraph is just propaganda nonsense devoid of any connection with reality MM.
link to youtube.com
There is an interesting video with Paul Krugman talking about the banks trying to turn back regulation to 1894. He says that it is one small battle after another. Just think Google and others are spending more money than Goldman Sachs to end patents or turn them into only a tool for them.
Your nonsense on this blog MM is just one part of the battle.
Interesting how often Malcolm rails against the 1%’s and then turns around and wants to deny patent protection to a form of innovation that opens itself to the average joe…
(and also repeatedly implies that patents should only be available for the Games of Kings “but for” expensive propositions)….
Things that make you go hmmm…..
So further to my comments below, I general agree with what the author is saying. I think too it creates a document that the team can agree on was a joint effort. I think a lot of the people have integrity and want to do the right thing. The document (patent application) then becomes part of their relationship. I think they understand that what is in the patent application is part of their joint effort and if they want to go it alone then they need to do something that is different.
That is one of the great benefits of patents. The formalization of disclosing what was done. Most people don’t want to go break their agreement with others. It also defines what they can do. They can leave and do something different than that document.
We want to avoid secrecy and non-competes that often extend so far to bind a person to company for work that is different, but that the worker would need an attorney to understand whether they can leave their current company.
We also want ways to fight Google, Apple, Microsoft who have an enormous advantage by market share and infrastructure. No one that I have ever met doesn’t think that patents are good who actually works with Silicon Valley start-ups (like me) and the large tech companies.
Good comment, Night Writer.
Andres,
How will the High-Tech Antitrust Litigation suit in US District Court of the Northern District of California regarding wage fixing against the major Silicon Valley firms impact your analysis? Will the $300 Million settlement, that was recently rejected and thrown out by the court for being too small, significantly influence your study?
For any seasoned, patented, technology professional, this may have a significant impact on them. They now may have a potential shift in the balance of power in negotiations going forward. This may also significantly increase the cost of acquisitions or may make many deals difficult to tie intellectual property developed as a team into a packaged deal.
Do you see this happening?
CB
The other big thing that is going on is the proposed trade secret laws in Congress. The law would create causes of action that would be effective in California.
I was a developer for a computer start-up in the early to mid-1980’s before the rise of patent protection and the way the tech companies tried to stop you from competing and switching to other jobs was through trade secrets. Onerous non-compete agreements came with stock and bonuses.
The danger is that we go dark again. Patents play a key role in allowing the inventors to switch companies because the patents disclose and define what the inventors cannot do. There aren’t giant murky areas like there used to be.
That particular legal issue is not the only way that the workers can be ‘subjugated.’
There is the co-opted philosophy path.
It is always amusing to read techdirt and slashdot and see the absolute failure to grasp just who benefits from the strong anti-protection stances that are exhibited there. Not only is “right to copy” NOT a form of innovation, it is also for the logical benefit of the owners of the companies, the Googles as it were, and NOT for the mere cogs, the mere bits, of the workers in the machine. And yet, the lemmings lap up the anti-protection mantra and march lockstep to the supposedly ‘@narchist’ tune, misguided to the end.
There are unreasonable restraints, however. The anti-patent reaction isn’t strange when the patents that have been allowed and litigated create restraints that seem, to developers, to prohibit any exploratory thought, reverse engineering, or creativity in a given field. If your career consists of writing code, seeking permission to write any and every line of code for fear of infringing a ill-granted patent would naturally spur the visceral rejection of patents as a whole.
I haven’t seen that in my practice. But, I agree that the PTO needs to be tightened. Unfortunately for us all, the Obama’s selection was based on Google pressure and not a rational assessment of the needs of the PTO. We needed someone that has vast experience in prosecution of patents. We got someone with no experience in prosecution of patents, but a strong belief that patents are bad. Toxic.
I would like some real-world examples of programs you think you couldn’t write. The reality is that there aren’t many patents that a good patent attorney can’t get you out from under.
No such thing as limiting thought.
None.
Now, if you want to do more than thought and actually copy – and you want to call that “thought,” well, we are just going to have to part ways.
The flip side of this coin is that you’re hosed if you hate your co-workers or if your boss(es) treat you like dirt.
Yeah, there isn’t necessarily an inherent “good” for the employee/inventor to be bound to a team.
The concern raised by Night Writer above is compelling though; onerous non-compete agreements and unclear boundaries of trade secret that restrict one’s employment may be worse than the “bindings” that are public.
Is Facebook a hypothetical example or has Facebook actually acquired the rights to a patent flowing from the work a team did at the start-up?
Facebook was used as the hypothetical because it has often engaged in acqui-hires, including those studied in the article. See, for example, link to businessinsider.com
[…] Binding Teams in Silicon Valley […]
What is meant by:
“a departing team member won’t be able to continue working in the path set out by the team’s patents”
The comparison here is between a setup where a company obtains patents versus one where it does not obtain patents: Without the employer obtaining patents, an employee can leave and start competing and working on the same type of technology right away. With patents, it is harder for the employee to leave and still work on the same of technology. The result is then that the patents tend to keep the team together because it limits options.
Overlapping all of this are other legal means for limiting employee options such as trade secrets, non-competes, and confidentiality agreements.
Worth noting that non-competes are rarely enforceable in California, which has obvious relevance to Silicon Valley.
Also worth noting that California is a community property state. I seem to remember an interesting post on here about how spouses may actually have interests in these patents and if not assigned, may prevent assertion of a patent.
Yes to Dennis and Jane. The fundamental point is that if the patents are in the buyer’s hands, the buyer can use those patents to make it more difficult for defecting team members to work on related technology.
Of course, there’s a lot of nuance I couldn’t capture in this post (or even the article!). For example, you could imagine that the defecting team member obtains a patent on an improvement to the technology, thereby setting up a blocking scenario with the buyer. Either way, the basic point stands–if the buyer obtains the patents, the buyer has greater influence on the original team members because the buyer can then make it more costly for those team members to go their own way (which, I should emphasize, may or may not be a good thing; from a team production perspective, it has the upside of keeping the team together, but that’s not the only thing that matters in the world).
And, yes, California’s refusal to enforce (most) non-competes is part of the story because it means one non-patent alternative for binding teams is taken out of play. Trade secrets are lurking too; I could not capture in this data whether trade secrets were also going to the buyers, although I suspect they were for the same reasons the patents typically did.
For now non-competes in CA aren’t enforceable. The new proposed (
Google) laws would create causes of action that are enforceable in CA. There have been some articles about that in the high-tech press.
As Prof Sawicki admits, the link that he draws here is impacted greatly by the cherry-picked location of Silicon Valley. Since that is the primary U.S. location where there are a sufficient density of start-ups and large companies to create a market for teams. This then offers the interesting suggestion that patents may be seen as more valuable based upon the location-of-invention.
Not sure that the suggestion necessarily follows.
“The fundamental point is that if the patents are in the buyer’s hands, the buyer can use those patents to make it more difficult for defecting team members to work on related technology.”
Does this not apply at a higher level – regardless of teams? Regardless of the model under study of buying the start-up in its entirety? I’ve worked on plenty of asset acquisitions to see the fundamental point in play well beyond the focus of the research here. The nature of the exclusionary power – and the concomitant issues of employee agreements and assignments of any inventions that a person is hired to work on are just not so limited.
I think you’re suggesting that patents have value apart from their potential value in binding team members. If so, then I agree. Patents can serve more than one function. The striking thing about the data here is that the patents are nearly always transferred to the buyer even though other things of value (e.g., office space, supply contracts, and other assets of the start-up) are not.
The puzzle, then, is to identify a reason why the buyer of the start-up is the highest value user of the patents. It’s not the most unusual pattern, and we can come up with other explanations. For example, maybe the buyer’s planned technology development would render it especially vulnerable if these patents ended up in others’ hands. That’s an issue for further investigation.
The data here primarily reduces the plausibility of one hypothesis: that the patents are no more valuable to the buyer than they are to the VCs or to the market as a whole. If that hypothesis were true, we’d expect to see patents going in different directions in acqui-hires, as is the case for the start-up’s non-IP assets. Because we don’t see that pattern, something else seems to be happening.
Thanks Andres,
My friend Occam would suggest that if one company wanted to purchase hard good assets, then a direct purchase from a traditional type of supplier would be where to look for for that type of confirmation.
I do agree that sometimes when one wants to be a type of intellectual asset, that such purchases will be impacted by factors such as already-staked out protections (patents and other IP), and sometimes the brains behind those items will also be desired.
But I think that “truth” applies regardless of the type of IP.
I have also seen what you did not focus on here: asset purchases that included IP but for which the IP was strictly an afterthought (it came with the package) and for which the client was unprepared – and often unwilling – to invest any additional capital in either further development or enforcement. I do not have anything more than anecdotal information on how pervasive (or not) this may be, but again, from what I have seen across several different art fields, this does not appear to be art field driven.
Thanks for sharing your research!
Sawicki: (1) the start-up assigned some of its patents to the buyer; (2) the start-up assigned some of its pending patent applications to the buyer; or (3) the buyer filed, after the transaction, a patent application naming one of the start-up’s principals as an inventor. In 29 of the 63 acqui-hires, at least one of those three kinds of patent transfers occurred.
What happens if you remove (3) from this analysis? What are the numbers? I ask because (3) doesn’t seem like a “patent transfer” at all. It’s just data showing that one guy that was hired by a company is doing the work that he was hired to do.
it suggests that patents are not only rewards for lone inventors, but also tools for keeping groups of inventors together.
I think you are confusing a correlation with causation. There is zero evidence presented here demonstrating that patents are “keeping groups of inventors together”. At best, the evidence described here shows that, for some short period of time, at least one “innovative” employee of a business that is purchased by a nearby business usually “innovates” for the purchasing company. Frankly, there’s nothing remotely new or interesting about that.
What’s next? A study showing that patents are “tools” for improving the quality of lunch consumed by silicon valley workers?
Hi MM,
Thanks for your comments. The full numbers are available in the article if you’re interested. Briefly, 13 of the 29 transactions involved the first two types of transfers. Another 16 involved only the third type. I realize (and note in the article) that it’s less clear whether the third type of transaction should count as an IP transfer. I think it’s plausible to categorize it that way because, unless the start-up is bought, its invention assignment agreements would ordinarily give it plausible claims to patents naming the start-up’s principals as inventors, even after those principals leave the start-up. That’s especially true where, as in this data, the patents are filed shortly after the principals leave. If the buyer wants to clear doubts about patent ownership, it has to acquire the start-up’s claims. None of this is to say that the third transaction necessarily drives the decision to pursue an acqui-hire; just that it is plausible to think of it that way.
I don’t think I’m making a correlation-causation mistake. I describe the evidence here and in the article as preliminary and suggestive. That’s what correlations often are. I’m not making strong claims about patents causing particular results, only showing that it’s possible.
Andres:
I don’t think I’m making a correlation-causation mistake. I describe the evidence here and in the article as preliminary and suggestive.
With all due respect, Andres, I don’t even see a meaningful correlation here. You looked at 69 so-called “acqui-hires” and you find that 13 of them (less than 20%) fall into one of two categories ((1) the start-up assigned some of its patents to the buyer; (2) the start-up assigned some of its pending patent applications to the buyer).
Somehow you get from that data to this: data thus indicates that patents are in fact an important part of the acqui-hire trend, without which it may be difficult for the founders and Facebook to consummate the deal.
But over 80% of the time patents aren’t involved in the acqui-hires.
more generally, it suggests that patents are not only rewards for lone inventors, but also tools for keeping groups of inventors together.
There is no evidence for the use of patents as such a “tool” in your data. Is the hypothesis that patents are “tools for keeping groups of inventors together” an interesting hypothesis worthy of testing? Apparently it is to you and a few other people. But you haven’t tested that hypothesis in any meaningful way.
Dennis, good insight as to the value of patents to a startup.
Now, we have to convince congress that what is good for startups is good for the USA. As you know, startups all but universally opposed first to file.
The overwhelming majority of new businesses have zero need for patents.
That’s a good thing.
Changing that would benefit patent attorneys at the expense of everybody else. That’s why the most loudest and most insistent attempts to effect such a change come from (surprise!) patent attorneys.
“The overwhelming majority of new businesses have zero need for patents.”
Taking for the moment the assumption that such is true (an assumption that is currently ungrounded), the result is a neutral thing rather than a good thing.
The spin here of course is that it must be the evi1 attorneys trying to “corrupt the system” by trying to “get rich” to the detriment of their clients that is to blame.
The nonstop mantra is deplorable. If you want to denigrate yourself personally have at it. But please stop denigrating the entire profession with such wildly over the top intimations. Please stop using the “lawyers are evi1” meme.
Are there bad ones out there? Sure? Is the industry so rife with these “bad ones” that your relentless edge of the field of rye tactics are called for? Absolutely not.
Just
Stop
The
Nonsense
… the neutral aspect being that joining the patent game is a completely voluntary thing.
And since the AIA, there is even the su generis PUR feature, so the aim of promotion does not even have to be joined.
So unless you think that the very idea of patents is bad (quite possible), the arguments you wield against one form of innovation are readily seen as arguments against the innovation system itself.
“must be the evi1 attorneys trying to “corrupt the system” by trying to “get rich” to the detriment of their clients that is to blame.”
I don’t ever think he said “to the detriment of their clients”. From what he’s said I would think he believes the opposite.
Taking for the moment the assumption that such is true
It is true.
The fact that you (or anoyone like you) would question this banal statement only demonstrates that you live in a tiny bubble with other confused, entitled people who believe that everyone should (or does) share their personal dreams about patent maximalization.
please stop denigrating the entire profession
I’m not denigrating the entire profession.
But I love it when you pretend that you care about how patent attorneys appear to other people! That is truly funny.
Are there bad ones out there? Sure
Thank you for acknowledging. Now if you’ll just point some of them out to everyone we’ll know that you can actually tell the difference. Good luck.
what is good for startups is good for the USA
A more accurate summation of the sentiment is “what is good for silicon valley venture capitalists is good for the USA.”
Because those people aren’t rich enough already. God forbid some guy with $10 billion dollars has to endure a little more risk in his gambling schemes. Why, he might actually lose some money and have to sell one of his islands. Oh, the humanity! The relentless persecution! We must hand out more patents, and quickly!
Paypal –> Tesla –> SpaceX.
What wonders startups bring.
What is good for startups is good for the USA!
I’m not sure Elon Musk and Tesla is the example that should be used to show the need for patents, given that Tesla has realized that patents were in fact hindering the wide adoption of the technology needed to grow their business.
Maybe there are other cases where this is not the case, but Tesla is a poor example to support the idea that patents are needed by startups.
Paypal –> Tesla –> SpaceX. What wonders startups bring.
Again: I’ve nothing against “start-ups”.
But there isn’t a compelling argument here to further corrupt the broken patent system because the 1 percenters want more profits from their gambling schemes and because some bigmouthed patent attorneys want to make more money while contributing as little to progress as is humanly possible.
Is a transparently broken and corrupted patent system “good for the USA”? Pretty sure the people on other side of the system don’t think so. And there’s far, far, far more of those people around than there are people who own patents or who can afford to gamble with them (and there always will be).
Egad, you’re one of the most negative people I’ve ever seen. Everything is phrased in this “evil versus good” language, with no one in the middle. Patent attorneys? They’re sc um and one percenters and liars and cheats, who are taking their client’s money to procure obviously invalid patents. Venture capitalists? The same, only worse because they make even more money.
When I was an engineer, I worked for a small start up that got VC money. Did the VCs make money? Probably. But they also grew a business, hiring way more engineers and staff than would have been possible without that seed money. Those employed people had families, car payments, etc. They benefited. And they were no where near Silicon Valley.
Also, this thread is an example of how MM can take over a conversation. The last time I checked this thread, there were about 7 posts, of which 4 were from MM. At least 90% of the words at that time were from MM.
Agreed that MM provides a lot of content, but disagree that this would be better without MM’s contributions.
Bear in mind, it is very easy to slip into a sort of “group think” when it comes to patents. “All patents good, more patents better, make patents easier to enforce, all the time.”There are many patent blogs where this thinking goes unchallenged. This isn’t such a blog, thankfully.
MM’s contribution is partly in that he forcefully challenges the group think so prevalent in the patent community.
Your cheerleading “spin” is obvious here, Go, as the “lots of content” is way way way too “nice” as it is by and large the type of “content” that could be eliminated without impacting the level of substantive content.
It is NOT groupthink – and it is most definitely true (if you take that we are talking valid patents) that indeed more patents IS better; that making patents stronger an easier to enforce WOULD be better.
If you understood the Qui Pro Quo and why we even have a patent system in the first place, them you would realize the error of your own post, and just why Malcolm’s venom is simply out of place.
And his manner of “forcibly challenging” is most definitely lacking in substantive content.
I would LOVE for him to “forcibly challenge” in some even remotely meaningful way. Alas, that is NOT what we get in such high volumes (and for such a long duration – eight years and running, well before I posted here – so this is NOT just one person’s opinion).
I would LOVE for him to “forcibly challenge” in some even remotely meaningful way.
Your defunct, relentlessly parroted dogma regarding Diehr was forcibly challenged by me and others for quite some time.
Your reaction to that is well-documented. Shall we remind everyone again how you behaved?
Ah, the internet tough guy speaks up.
You forgot: “in some even remotely meaningful way”
You have not don’t that part yet.
As to “and others,” and “we,” you don’t get to count the sock puppets like Vivika M and Friend(s) of the Court that you use over at PatentDocs. Repetition of the same b@nal C R P does not meet the “meaningful way” criteria.
Plus, you don’t believe that Diehr is good law anyway, as you have made clear, so your startin point is simply in error here.
Try again.
Or perhaps for the first time.
PatentBob: Everything is phrased in this “evil versus good” language, with no one in the middle.
False. Keep digging.
you’re one of the most negative people I’ve ever seen
Noting that some patent attorneys are bottom-feeders and recognizing that we shouldn’t let the richest people in the world decide what’s good for everybody isn’t “negative” unless, of course, you are one of those people.
You know what’s negative? Comparing people who discuss flaws in the patent system with “jihadists” and “rapers and killers.”
Wake me up next time you feel compelled to comment on that sort of negativity, PB. Then we’ll know that you aren’t a hypocrite.
Comments are closed.