By Professor Camilla A. Hrdy
Cross-posted on Written Description
One of the key purposes of trade secret law is to address the “Arrow information paradox.” The information paradox posits that there is a fundamental challenge in information exchange: It is difficult to assess the value of information without first sharing it, but once the information is shared, it becomes vulnerable to being copied, leaving the originator without compensation and without a competitive advantage.
Trade secret law, many argue, partially solves the “paradox” by making it easier to exchange information despite the risk of disclosure. By legally protecting secrets that are shared under the auspices of confidentiality, trade secret law creates a more secure framework for limited, protected exchanges of information.
Indeed, the Supreme Court itself implied in Kewanee v. Bicron (1974) that, without the legal protection of trade secrecy, “[t]he holder of a trade secret would not likely share his secret with [another] who cannot be placed under binding legal obligation to pay a license fee or to protect the secret. The result would be to hoard rather than disseminate knowledge.”
Scholars like Mark Lemley and Michael Risch have made similar arguments, suggesting that without the protections of trade secret law, companies might share less information and might make wasteful expenditures on keeping their secrets. Although other scholars like Michael Burstein have suggested some of this sharing would happen anyway through contractual arrangements like non-disclosure agreements, this is the conventional view of what trade secret law does: It makes protected sharing easier.
In their engaging new article, Trade Secrecy’s Information Paradox, forthcoming in Notre Dame Law Review, Chris Buccafusco, Jonathan Masur, and Deepa Varadarajan, argue that (irrespective of whether trade secret law actually solves the Arrow information paradox), trade secret law has generated its own information paradox.
“[T]rade secret law,” they write, “contains its own, nested information paradox. In many cases, determining whether a piece of information should be protected by trade secret law—that is, whether protecting it will be socially valuable as opposed to socially harmful—first requires exposing that piece of information to public scrutiny. But such exposure eliminates any chance of protecting it as a trade secret. This is a fundamental problem with trade secret law, and it is one that cannot be easily overcome.” (7)
Let’s unpack this.
So the authors begin by arguing that the main goal of trade secret law should be to “provid[e] incentives for the development of information that might not otherwise exist.” Trade secret law is supposed to “fill a hole left by patent and copyright law that, if left unfilled, would lead to underinvestment in socially productive knowledge…” (Not everyone agrees that creating incentives-to-innovate is the main goal of trade secretly, but it’s commonly identified as at least one of them). And yet, the authors observe, trade secret law does not distinguish between socially valuable and socially harmful information; it protects information as a trade secret irrespective of whether it is socially valuable. Therefore, trade secret law ends up shielding significant amounts of information that might, in fact, harm the public interest. (Compare patent law, which also does not distinguishes socially beneficial from socially harmful inventions…but requires disclosure of the information, thus avoiding the problem of hiding “bad facts” from the public).
Herein lies the authors’ “paradox.” It is impossible for courts, regulators, or others acting in the public interest to know whether information that is asserted as a trade secret is socially harmful or socially valuable, because figuring this out would require revealing the very information in question, which would destroy its secrecy.
In other words, the same “information paradox” that makes it difficult to discern information’s private value also makes it difficult to discern information’s social value. The result of this information paradox, the authors suggest, is that it is very hard if not impossible for government regulators to learn about and address socially harmful secrets; and it is very hard if not impossible for courts deciding cases that implicate a potential disclosure to decide whether the disclosure should be permitted or not.
Public-spirited actors are, in a sense, paralyzed. If they prohibit the disclosure, they may harm the public by allowing a firm to hide dangerous information that the public needs to know; but if they allow the disclosure, and they get it wrong (because the information is socially good and privately valuable), then they harm the public too, because “this firm and others like it” might consequently stop “investing in the creation of such innovations” (5).
I absolutely loved reading this article. It was beautifully written and extremely insightful, bringing together scholarship on trade secrecy, the expansion of the secrecy exemptions from public disclosure under FOIA, as well as scholarship on “costly screens.”
But I do have a serious hang up about the “information paradox” framing. Let’s assume there is an “information paradox” here, and that discerning information’s social value requires sharing it and thus destroying the information’s secrecy. To me it seems like trade secrecy is the solution, not the problem. The same arguments the Supreme Court made in Kewanee in favor of recognizing trade secret rights — to encourage safe sharing and prevent the “hoarding” of knowledge — apply here too.
Thanks to trade secrecy, disclosures are safer, allowing then to happen more frequently. If a court or a regulator is troubled by the “paradox,” worried they cannot figure out information’s social value without disclosing it, then trade secrecy provides a partial solution to that problem by ensuring that the owner of the information can make a legally-protected disclosure to the court or the regulator, while preserving its trade secret rights.
In fact, case law such as Ruckelshaus v. Monsanto Co. (1984) and Food Marketing Institute v. Argus Leader Media (2019) are based on the premise that the government can obtain information from a private actor in the course of regulating that private actor, while making a promise to retain the information’s confidentiality. When government makes such assurances of confidentiality, then the information can safely be shared with government, so that government can effectively regulate around it.
I don’t see why this solution does not work here too. Protected, limited disclosures of trade secrets should help solve the authors’ “social-value information paradox,” just as it helps solve the private-value one. Government regulators can review trade secrets in a protected environment to identify negative social value, and make any partial disclosures necessary. If government regulators breach their duties during this exchange, they can be sued for trade secret misappropriation as well as breach of contract, among other things.
I don’t see why full public disclosure is required. (The word “disclosure” probably should not even be used when discussing a protected disclosure; we should call it “sharing” to show that trade secrets are not necessarily lost, as Sharon Sandeen has observed). Scholars like Chris Morten and Elizabeth Rowe take the view that it’s perfectly possible to review trade secrets in private, and even partially disclose facts about them, without “breaking” the secrets. For example, to determine whether an algorithm used by a tech company is biased and illegally discriminatory would not require public disclosure of the algorithms. At worst, this might require public disclosure of the fact of a negative bias. This is very different from full disclosure of the algorithm.
To be fair, the paper does address the possibility of what it calls “in camera” review, but only very briefly on page 30, and I think the very phrase “in camera” makes this seem like a specific situation where a judge is viewing information “in camera” during litigation. But it’s much broader than that. Judges, regulators, or anyone needing to view the information can do so within the safety of the law; that is a big reason we have trade secrecy.
The authors also point out that regulators, such as federal agencies, would need more resources to complete this type of “in camera” review. But why is that trade secret law’s problem? Isn’t that a political problem of insufficient regulation? If a government wanted to do more regulation, trade secrecy rights might only make that easier by giving companies incentives to share their secrets with government.
One final point on “economic value.” The authors operate from the assumption that information with purely negative value is a trade secret. For example, they imply that the fact that an algorithm is biased or the fact that an airplane is unsafe is a trade secret. I think the paper should more clearly acknowledge the argument that negative facts, which play no role in a company’s business operations, are arguably not trade secrets at all, because they do not derive “independent economic value” from secrecy.
The mere fact that an algorithm is biased, or the mere fact that a plane is unsafe is not a trade secret. The algorithm might be, and the plane’s design might be, even if it is faulty. But the mere fact that information might be bad for the public is not a trade secret. And that is really all the public would need to know. Thus, regulators can assess details in private with assurances of confidentiality, while publicizing their ultimate finding that certain information is “bad” from a public health perspective, or that it violates some law, because those findings are not trade secrets. (It’s debatable, I concede, but I make this argument in my article on “independent economic value,” and not surprisingly I think it’s compelling).
All that being said, I highly recommend this article. I found it really thought provoking and an excellent addition to recent scholarship asking whether trade secrecy inappropriately interferes with disclosures that could be in the public interest by scholars like Dave Levin, Rebecca Wexler, Tait Graves and Sonia Katyal.