Monday 19 October 2009

Can One Contract Around a Trade Secret?

I begin a new teaching term next week and that means two sessions with my MBA students, setting out the basics of IP law, before we delve into the challenges of the IP-managerial interface. I have been teaching IP for a long time and that should mean that there is very little angst connected with this part of the course. Sadly--au contraire! There is one segment of the presentation--trade secrets--that I always approach with the "fear and trembling" usually associated with more existential issues.

Put briefly, at some time during the treatment of trade secrets there is always at least one thoughtful, attentive student who asks the ultimate question--"So why do we need trade secret protection as a separate and distinct right? Surely it can be subsumed into other rights--such as contract and tort--that handle the subject matter." Over the years, I have worked up an answer that seeks to point the advantages of having a separately protectable right for valuable secrets. I was recently pleased to find that my notions about trade secrets were supported as part of a much wider-ranging article by the distinguished IP scholar Mark Lemley, in his most interesting article, "The Surprising Virtues of Treating Trade Secrets as IP Rights", Stanford Law Review, vol. 61, Nov. 2008.

From time to time on this blog I will consider various aspects of Lemley's analysis, because I am firm believer that trade secrets should be viewed more centrally as part of anyone's bundle of IP rights. In this blog post, I want to consider one aspect that has practical as well as doctrinal significance, namely, whether one can contract around trade secrets law (similar, e.g., to the question that arises under copyright law whether one cannot contract away the right to reverse engineer).

Thus Lemley writes:
"In trade secret law, [the question] comes up in three significant contexts: efforts to contract around the requirement of secrecy itself, whether in business disputes or in restrictive employment covenants, efforts to ban reverse engineering by contract, and the question of whether a confidential relationship can be implied absent a contract" (footnotes omitted.)
The rationale for not allowing one to contract away a trade secret right is stated in a footnote to Lemley's article, where James Pooley argues as follows:
"The law relating to trade secrets reflects a balance of public and private interests in the encouragement of innovation, the preservation of ethics and the maintenance of a free marketplace of ideas and movements of labor. The balance should not be upset in any given transaction by private understandings between the parties."
Lemley is a bit more circumspect, stating that "... my inclination is to prevent parties from opting out of particular rules of trade secret law, at least to the extent they rely on trade secret rather than contract remedies."

Sadly, there does not seem to a dispositive answer to this question. In my view, the issue of whether one can contract around trade secrets is a "big thing". I encounter one or another of the three scenarios described by Lemley on a frequent basis and I am uncertain about what to advise if the client asks about the ultimate enforceability of the "contract around" provision. The underlying difficulty is always the same--if there is a broad right to "contract around" the trade secret right, why exactly should trade secrtes be placed in same pantheon as patents, copyright and trade marks"? And while I am still stammering to give a cogent answer on that question, there then follows the ultimate query--"If so why should we be mentioning trade secrets in the same breath?"

I have less than a week to prepare myself for this year's inevitable bout with uncertainty about the answers to these questions. Sharon Sandeen has characterized trade secrets as "the Cinderella of intellectual property law". I am not sure, however, that discussing trade secrets with challenging MBA students is exactly "the ball" that Sandeen has in mind. Any advice will be welcome.

Char Girl or IP Right?


Anonymous said...

As an English lawyer, I am struggling to understand the question, let alone supply an answer.

We don't have trade secrets laws as distinct from general equitable laws of confidentiality, and there is no great clamour for an extra set of rules.

Confidential information is not a type of intellectual property but can be licensed in a similar way.

As for "contracting around", an English court is unlikely to protect by contract (ie in a confidentiality agreement) information that does not have any inherent confidentiality. When it comes to non-compete clauses, the courts have set limits (or at least principles governing the limits) on what is acceptable in an employment context and in a sale of business context, respectively.

Is this a US or Israeli law issue that doesn't translate into other legal systems?

Anonymous said...

I am also struggling with your question. It sounds like a fascinating topic but I'm not exactly sure what you're getting at? Can you provide a concrete example?

Neil WIlkof said...


Thanks for the instructive comments. I have a couple of thoughts:

1. I can't speak with any authority on the English law position. Lemley reports (fn. 167) that there is case law in the US that holds that a court will in some fashion hold a party to a contract to an undertaking of confidentiality even if the information is not secret.

2. As a matter of practice, we encounter from time to time "confidential" agreements under English law that include a clause that purports to obligate a party to maintain confidentiality of information and the like without a secrecy limitation to the obligation. We have often wondered how an English court would enforce such a provision where the "disclosure" was not of confidential information.

3. As for the differences between the US, Israel, and other countries,with respect to the treatment of trade secrets, my short response is that legal treatment of the subject differs materially from country to country, which makes it so difficult to explain to oneself, much less to students.

4. With respect to the EU specifically, I refer to another footnote in the Lemley article (n. 174)--"Kataryzna Czapracka suggests that the EU antiturst authorities give less deference to trade secret claims than their US counterparts because the European Union has no conception of trade secrets as property rights. Czapracka, "Antrust and Trade Secrets: The U.S. and the EU Approach, 24 Santa Clara Computer and High Tech L.J. 207 (2008)."

Michael F. Martin said...

Can it be true that there is no tort of trade secret misappropriation in England? I would be surprised.

When I read Lemley's paper, I was struck by the timing of when the doctrines of trade secret law seemed first to emerge -- in the mid-19th century. This was the same period in which the concept of patents as property rights seemed to really take hold. Prior to that period, patents of invention were treated much like other patents -- i.e, as privileges from the sovereign.

And perhaps that suggests the answer here. The patent system is the Coasean answer to the problem of contracting around trade secrets. Because torts, including torts of trade secret misapproporation, are in personam, it is too costsly to bargain with all that are either benefited or harmed.

Anonymous 1 said...

Neil, the more I look at your question, the more I think this is a US law issue that wouldn't be raised in an English law context.

I don't know much about US trade secrets laws (and I am not familiar with the case law that Lemley cites) but I would have the same question - why are they needed? I wonder whether the federal versus state jurisprudential tug of war has something to do with it.

As for the difficult questions that you raise, I am not sure that these question depend on there being trade secrets laws. They could be asked in relation to equitable duties of confidence. Or are you asking whether trade secrets laws really add anything to equity, if they can be contracted around?

I am surprised that MBA students would care that much about the detail of trade secrets laws; (to be a little provocative) isn't this a detail that technician lawyers would worry about and advise them on, when they are CEOs or commercial executives looking at the broader commercial picture? I am doubtful whether many law students would be that inquisitive.

Neil Wilkof said...


Thanks for the comments:

1. The point of Lemley's article is to address the question of whether trade secrets should be understood as an IP right. From your comments, I guess that the jury is still out whether the question arises under English jurisprudence.

2. As for MBA students, I tend to teach middle manager types within the greater start-up ethos of Israel; I suspect that the student body at places like Wharton, Harvard and Chicago are different. In any event, my students seem to find the subject to be of interest because they sense that they encounter a lot of inventions,developments and the like that are not or will not be the subject of patents, but still need to be attended do as part of their managerial function. I agree the technical question raised in the second half of the blog post will be of less interest to them.

By the way, in my view,trade secrets will become a much more central part of law school curricula, at least in the U.S., either under the rubric of IP or some other category.

Anonymous said...

In Australia, I think we're in the same position as England (as per the first comment) -- the information must be confidential, although this can include things like which of a number of publicly-known suppliers are reliable ones. Beyond that, remedies are largely in contract, usually under restraints of trade.

Courts here would not enforce an obligation not to use/disclose information that was once confidential but is now in the public domain. (The usual objection is that the court would be acting in vain in trying to do so.) The obvious exception is where the information is only in the public domain as a result of the breach complained of (something that is usually taken account of in a well-drafted confidentiality agreement, anyway).

I wonder whether the difference is explained by the fact that information here is not considered to be property. I think that is also the case in England. The position (I understand) is different in the US. Thus there can be no proprietary right in information unless one is created by statute (eg copyright) -- hence one is relegated to using confidentiality as a proxy. We run across this issues in clauses from US boilerplate contracts that talk of a party "owning" information...

See for an example of trade secret protection and confidentiality: Artedomus v Del Casale [2006] NSWSC 146; 68 IPR 577

And for the limits of the doctrine: Maggbury Pty Ltd v Hafele Aust Pty Ltd [2001] HCA 70; 210 CLR 181; 185 ALR 152; 76 ALJR 246 (13 December 2001)

Anonymous said...

Anon Australia mentions an interesting point. I am much troubled by references in US-drafted contracts to a party (usually the customer) "owning" the results of a project that is performed under the contract. Under English law, you can own resulting inventions, or copyright in the work setting out the results, or physical property in the piece of paper on which the results are printed, but you cannot own the results themselves - it is meaningless. Presumably it is meaningful if the results are protected by US trade secrets laws?

Renzo Marchini said...

I do not think that this arises under English law. As is the case for Australia, there is simply no “property” in information. Trade secrets are protected by the law of confidentiality.

And even if some information was disclosed in confidence, then quite clearly (under English law) you can contract out of the duty of confidentiality. I don’t see the point of the question. This would be the scenario: X is a secret; I disclose it to you; by the way – by contract – I agree that you don’t have to abide by the duty of confidentiality as regards X. You are in effect agreeing that there was no duty of confidentiality.

I also think that under English law you can protect by contract information that is not in actual fact confidential in nature. What you are doing is putting in place a contractual restriction upon the recipient to do something (keep it secret) or not do something (do not use for other than X). It is analysed as a restraint of trade and thus can be enforceable if “reasonable as between the parties” etc.

An example of how this issue arises under US law would be helpful.

Renzo Marchini said...

PS even if it is a property right in other jurisdictions. I simply don't see how the question of "contracting out" of it really arises. You can quite clearly contract out of other property rights (in effect, by granting licences); why not this one?

As mentioned, exmaples might be useful.

Unknown said...

It appears that folks are looking for an overview of how trade secrets are handled in the US, so here it is (I will post) separately on the contractural issue):

In the US, we protect trade secrets in one or both of two ways.

Under the first method, individual states define trade secrets by statute. Most states have adopted the Uniform Trade Secret Act (UTSA), to which a body of case law applies instructing courts and how this uniform act is applied in a particular state and as well as in general if there is no precise state law to apply in a specific case. The UTSA defines a trade secret as “Information” that derives independent economic value, actual or potential, from not being generally well known to, and not being readily ascertainable by proper means by, other persons who can obtain economic value from its disclosure or use and which is the subject of efforts that are reasonable under the circumstances to maintain its secrecy. Because trade secrets arise by statute, even if there is no contract between parties, a party can bring a suit under this UTSA (actually her respective state's specific statute conforming to the uniform act) contending that another party has misappropriated her trade secrets.

The second way of protecting trade secrets in the US is broader than trade secrets, and arises by contract. One can define confidential information as any information that is exchanged between parties, as long the "exceptions" are included in the contract (ie., independent development etc). To bring suit for disclosure of confidential information, there must be a breach of contract claim.

The main substantive difference between trade secret protection and confidential information that arises by contract is that trade secrets can go on forever, but confidential obligations that arise by contract must have a date certain by which the obligation ends, or else the underlying contract will be found indefinite and, therefore, invalid. In my state of Georgia, the general view is that any length of more than 5 years is too long, but I tend to say that 3 years is a solid time period for the information that my clients typically seek to protect.

For true trade secret information (which truly has value for an extended time--actually very little information falls under this definition), good US lawyers recommend the "belt and suspenders" approach by saying in the contract that obligations that arise with respect to information that falls under the definition of "confidential information" expires on a specified date, but any information that comprises "trade secret information" has an indefinite duration.

Unknown said...

With regard to the original issue of whether one can contract around trade secrets in the US, the answer is yes. The parties can say in a contract that no trade secrets or confidential information are being exchanged. Even if this is not true, the assertion in the contract would be dispositive. In other words, if I said there were no trade secrets involved, I could not later in good faith bring a case against the other party under the trade secret statute because I said in the contract this was not the case. As set out in the previous comment, the trade secret statute requires me to demonstrate that the information meets the legal definition of trade secrets and I can't say the information is if I said in the contract they were not trade secrets. (Of course, I could argue that the contract was based on fraud, etc, but that is far afield of the original question.)

Anonymous said...

Very interesting, Jackie. So:

1. general confidentiality laws, comparable to the English law of confidence, don't exist in the US - it is simply a matter of contract?

2. Are trade secrets considered a form of personal property in the US?

Unknown said...

1. General confidentiality laws, which appear to be a form of common law of confidentiality, do not exist in the US. My guess is that one could possibly figure out a legal theory that could support a claim of breach of confidence, but this would likely be in the vein of tortious interference. There is no legal claim for breach of confidence in the US.

2. Trade secrets are definitely a form of property in the US--one of the "sticks" in the bundle of rights that forms intellectual property in the US.

Neil Wilkof said...


The idea of contracting around trade secret law per the Lemley article includes a contractual obligation to keep something secret that is not secret under the applicable trade secret law, and to waive the right to engage in reverse engineering. As noted, the focus of the article is to argue in favor of a distinct property right for trade secrets, as opposed to combining trade secrets with contract law, tort law, fiduciary duty obligation of good faith and the like. IMHO, it is a fair question that Lemley asks. Whether you accept his reasoning and conclusions is of course up to each of us.

Mark Anderson said...

Having now (skim) read Lemley's article, I understand better the questions that Neil is raising. But I feel that several distinct issues are being unnecessarily and confusingly conjoined in favour of a "theory" of trade secrets as a form of IP.

In brief (this comments window is not well-designed for anything more):

1. Lemley seems to make much of the deficiencies of the law of confidence as a tort, citing cases where the courts seem to have ignored any requirement for the information to be inherently secret, rather than simply the subject of contractual obligations of secrecy. This issue may arise from how the US courts have defined the tort. Under English law, and in particular Coco v Clark, one of the three key requirements of a breach of confidence action is that the information must have the "necessary quality of confidence". To the extent that Lemley is arguing that we need to view trade secrets as a form of IP because this will introduce a requirement of inherent secrecy, I think that is a non-issue under English law in view of the Coco v Clark criteria. (Incidentally, Lemley's article starts by talking about the "Anglo-American" nature of trade secret doctrine, but he omits mention of some of the key 19th century English cases, eg Albert v Strange, so I am not convinced by this aspect of his argument.)

2. In relation to issues of reverse-engineering and contractual opt-outs, I didn't understand Neil's comments but I now see how they reflect Lemley's commentary. I am still not sure of their relevance to whether trade secrets are a form of IP. After discussing these issues, Lemley acknowledges (at page 351) that "understanding trade secrets as IP rights won't resolve this debate, any more than it has in the copyright context."

3. Lemley does, however, follow the above quote, a couple of sentences later, with the phrase "And it may strengthen the hand of those who argue that there is more at stake here than just the agreement of two private parties." In my mind, this raises the interesting question of whether, by recognising trade secrets as a form of IP, it may acquire a public policy element that is otherwise lacking. I am thinking, for example, of the theory that patents are a deal between State and inventor, to encourage disclosure of new ideas. Having said this, I don't think there is likely to be an appetite in the English courts for such theorising.

Michael F. Martin said...

I'm definitely missing some nuance here that everybody else seems to find interesting.

From a law & econ perspective, I see trade secrets as either tort or contract right in every sense that's relevant to the question about contracting around. Unlike copyright, trademark, and patent infringement, trade secret claims are not strict liability. Hence, while in principle these rights are in rem by statue, in practice they are asserted only in personam against bad actors (the misappropriators) or employees to enforce non-competes. The corner cases are angels dancing on the heads of a pin.

Thus, Neil's question as I understand is can you contract around being liable for torts or breach of contract?

But I do find interesting the question raised indirectly here about reverse-engineering. If courts were to rule consistently that reverse-engineering of trade secrets was trade secret infringement, then trade secrets would indeed be more like property rights. But in all of the cases of which I am aware, there were bad acts in addition to the reverse-engineering -- i.e., the reverse-engineers somehow misappropriated something before starting their reverse-engineering, often an employee.

The policy point here is that if you don't permit reverse-engineering of trade secrets than you substantially discourage patenting and freedom of inventors to move around from company to company.