(Cross posted on Project DisCo)
First of all, thank you Newegg!
Everybody go out and buy an extra computer monitor, TV or a bulk pack of styluses from them right now (FYI, styluses are retro cool!). I’ll wait…
Done? Good. Now I will explain why.
As it was widely reported earlier this week, perhaps best by Joe Mullin at Ars Technica, Newegg — the online technology retailer — fought Soverain, a notorious patent trollAn entity in the business of being infringed — by analogy to the mythological troll that exacted payments from the unwary. Cf. NPE, PAE, PME. See Reitzig and Henkel, Patent Trolls, the Sustainability of ‘Locking-in-to-Extort’ Strategies, and Implications for Innovating Firms.. Soverain was holding the industry hostage with some pretty shoddy patents with which they claimed the entire idea of a “shopping cart” in e-commerce. Newegg’s fighting spirit is commendable and part of its broader “We never negotiate with trolls. Ever.” policy.
[Newegg’s strategy may also be a risky calculated attempt to deter trollAn entity in the business of being infringed — by analogy to the mythological troll that exacted payments from the unwary. Cf. NPE, PAE, PME. See Reitzig and Henkel, Patent Trolls, the Sustainability of ‘Locking-in-to-Extort’ Strategies, and Implications for Innovating Firms. lawsuits in the future — the patent world’s equivalent of “we do not negotiate with terrorists.” However, the first two companies who fought Soverain — Victoria’s Secret and Avon — actually lost. Soverain lost its initial case in the same patent friendly court that rendered the Victoria’s Secret and Avon verdicts, the Eastern District of Texas, before winning on appeal.]
Unfortunately, under our current patent system Newegg’s act was exceptional. The retailer’s stand was the product of a fed up company willing to make a banzai charge on behalf of a besieged industry (or the Howard Beale of the tech world). If the Department of Commerce gave out medals of valor: Newegg should get one. It would have been easier (and likely cheaper) for them to settle early, but they stood on principle.
Unfortunately, economics, and actual outcomes, does not give much weight to principled stands. Incentives dictate market outcomes. Given the huge expense and uncertainty of patent litigation, patent friendly courts and the legal uncertainty around concerted boycotts of trolls, standing up to patent trolls is difficult. Usually this leads to companies, particularly small and medium size companies, settling with trolls instead of undertaking a costly legal fight that could put them on the hook for a huge sum of money if they lose. (Patent attackers often “punish” firms that fight the validity of a patent, offering significantly better deals to the firms that choose to settle early.)
In economic parlance, busting trolls with bad patents is a “public good.” In other words, the entire industry benefits when an individual company invalidates a bad patent. Or, as Dan Ravicher of PUBPAT pointed out in an amicus brief in the pay to delay pharmaceutical antitrust case:
Accused infringers who prove a patent invalid perform an important public service by correcting the PTO’s errors on their own nickel.
Economic theory also tells us that public goods are generally underprovided (by the private market at least) because individual market actors do not reap the full benefit of their investments. In this instance, Soverain had cases pending against over 30 companies, including such household names as eBay, RadioShack, Nordstrom, Bloomingdales and Walgreens. Newegg’s stubborn refusal to rollover means that those other companies are also saved the significant time and expense of fighting to knock out Soverain’s patents.
This “underprovision” is well-documented in economics literature:
In many cases where each party’s respective incentive is expressed by the difference in its profits between winning and losing, a patentee’s incentive to defend its patent grossly exceeds an alleged infringer’s incentive to challenge it….
Thus, for instance, if there are five infringers of equal size, each gets only a fifth of the gains from a successful challenge because each is paying only a fifth of the patentee’s total royalties. Therefore, the patentee has five times more incentives to prevail in litigation than any one challenger has.
Going forward, we can’t rely on altruistic companies like Newegg to reliably provide the “public good” for the industry at large. Academic literature suggests several ways of better aligning private incentives with the public good:
* * *
1) Fee shifting
One potential fix that would help on the margins is making it easier to recoup litigation fees when a patent is invalidated. This gives individual companies more incentive to challenge bad patents, as the expected value of the loss decreases significantly. It also makes it much riskier and more costly for firms to assert low quality patents. Under the current system, companies — particularly patent trolls and NPEsNon-Practicing Entity. A broad term associated with trolls but now disfavored because it includes universities and legitimate technology developers that seek to license technology in advance rather than after a producing company has independently developed it. More — can assert low quality patents against vulnerable companies (oftentimes startups) and command settlements because even winning a patent case is very costly. This is the premise of the SHIELD Act, which would target patent trolls in high-tech industries where the problem is currently acute.
2) Coordinated Defense
As I discussed above, the core of the public goods problem here is that the costs of challenging a patent are borne by an individual company, but the benefits are conferred to the industry at large. This incentivizes “free riding” where individual companies count on other people to do the dirty work of challenging a patent’s validity. One way around the free riding problem is for firms to make credible upfront commitments to fight patent trolls and share the burden of defense. This is especially important for vulnerable firms, like startups, who don’t have the time or money to fight individual patent suits themselves. As a result, they usually settle — oftentimes for a fee that is about the same as it would have cost to successfully defend the patent — so there is virtually no incentive for them to fight a shoddy patent on their own. If an entire industry, or a similarly situated group of competitors, could band together to share the costs and present a unified front this would help align the costs of challenging suspect patent or patents (which are borne by the industry) with the benefits of defending the patent (entirely falling upon the “company” asserting the patent). The result, from a social welfare standpoint, would be closer to the ideal outcome.
However, an older court decision — Jones Knitting — casts antitrust uncertainty around these coordinated actions because it wrongly viewed such activity as collusive arrangement to fix the price of a commodity input. Although firms are entitled to mount a joint defense after they are sued, the act of banding together pre-lawsuit is legally questionable. It shouldn’t be. Because trolls can pick off individual companies in successive waves of suits (as one patent scholar calls it, the “divide-and-conquer” strategy) — as is typical practice — they can take advantage of the misaligned incentives by isolating companies. Also, banding together to use declaratory judgements to preemptively knock out weak, but threatening patents would be socially beneficial.
The good news is that courts are beginning to recognize that such practices can be procompetitive — and they don’t cause consumer harm. In fact, quite the opposite. Two recent court decisions have cast doubt that courts will view such coordinated defensive actions (antitrust lawsuits filed by trolls against the industry self-help groups RPX and AST were recently dismissed) as actionable under antitrust law.
Going forward, courts should affirm (1) that group boycotts of trolls and coordinated patent busting are not per se illegal (automatically illegal without needed to show anticompetitive harm) under antitrust law and (2) that such activities are more like joint research ventures rather than “buyers cartels” because they involve companies banding together to produce a public good (in this case, freeing knowledge that should be in the public domain already) rather than companies conspiring together to restrain trade.
[Another similar coordinated mechanism that has been discussed as a partial solution to this problem is insurance reciprocals where companies agree not to individual settle with trolls, pay premiums and, if sued, the insurance company fights the patent in court. This is useful because the vast majority of trolls forced to defend themselves in court lose. Therefore, pooling risk would disincentivize trolls from asserting bad patents and, in the long run, it would also drastically decrease the market value for bad patents, which is largely based on their capacity to generate nuisance lawsuits.]
Another way that legal systems align the incentives of private actors with the public good is to give rewards–usually in cold hard cash. Whistleblowers are often given part of the penalty assessed if the information they provide leads to a conviction and fines (under the Federal Claims Act, the U.S. government has used such provisions to claw back $27 billion in fraudulently dispensed taxpayer money) and the IRS has used bounties to successfully catch major tax evaders. Rewarding companies for knocking out a bad patent isn’t even unprecedented. Under the Hatch-Waxman Act (aka Drug Price Competition and Patent Term Restoration Act) which became law in 1984, generic drug manufacturers who win a validity challenge of a patented drug are given an 180 day “exclusivity period,” which gives them a commercial head start as a reward for their effort.
Two types of bounties have been proposed. The first is an examination-stage bounty that rewards people for submitting prior artPrior art is the knowledge in the field of a patent that was publicly available before the patent was filed. that is used to successfully knocking out a patent application. The second type of proposed bounty (a litigation-stage bounty) would be awarded to a company or companies that void a patent in court, either through the successful defense of a patent infringement suit or through a declaratory judgement action (where a company or companies petition a court to invalidate a patent when they suspect that they are going to be sued in the near future).
* * *
It is not just companies that should be thankful to Newegg, but consumers as well. In another case of poorly aligned economic incentives that afflict the patent system — in this case the “pass through problem” — much of the costs that trolls, particularly trolls like Soverain, force on an industry at large are borne by consumers. The quick and dirty economic explanation: if an entire industry pays a near uniform extra cost then the costs don’t come out of the companies’ margins (as they would if only one company paid a higher cost for an input), but instead are largely passed through to consumers. So, as I was saying, go buy something from Newegg. They just saved you money!