Illusions of Bayh-Dole: “manufactured substantially” 2

The start of this article is here.

We are taking apart section 204 of the Bayh-Dole Act, the “Preference for United States manufacturing” that ends up being almost useless for its purpose. But to get there, we have to go step by step to demonstrate just how empty section 204’s gesture really is. I’m sorry that it takes time to work through professional-grade awfulness, but that’s in the nature of Bayh-Dole.

We are dealing with just the first sentence of the two that make up section 204. We have dealt with precedence, scope, and the fundamental restriction. Now let’s look at how section 204 puts its emphasis on licensing, and not just any licensing, but a rather useless sort of exclusive licensing.

Licensing, Not Ownership

Consider the licensing element in section 204. There is no restriction for US manufacture on the owner of the subject invention (or on patent rights to the invention). Section 204 does not require the owner of a subject invention to manufacture in the US any subject invention for its own use or to sell (using a patent to ensure exclusivity). Only the march-in provisions in Bayh-Dole (35 USC 203) might require an owner of a subject invention to grant licenses if the owner is not using or otherwise making available the benefits of the subject invention–and march-in has never been used in the history of Bayh-Dole.

Thus, section 204 notwithstanding, a business owner of a patent on a subject invention may, for instance, contract with a foreign supplier for product to use and to sell exclusively in the United States, and can even make that supplier exclusive, so long as the business is doing the using and the selling in the United States. As long as the small business does not grant an exclusive license to anyone else to use or to sell in the United States, section 204 doesn’t apply. Section 204’s focus on a preference for United States industry only in the narrow instances of exclusive licensing to use or exclusive licensing to sell, then, is very limited and very strange.

The general statement of prohibition in 204 makes Bayh-Dole appear to strongly support American industry and that any variation from this support is exceptional. But once the initial shape of 204 is made apparent, it is clear that 204 does very little and is easy to work around. It is a pebble thrown into a stream, but with a wonderful bureaucratic flourish. The stream does not change, and we are left with the mental image of the flourish (“Preference for United States Industry”!), and distracted from the inconsequential result.

What we describe in section 204 is not premised on a patent owner’s clever gaming of the system–that happens, anyway. Instead, section 204 itself games Bayh-Dole’s policy statement and expressly provides guidance to show patent owners how they may avoid section 204’s demand for American manufacturing. Here, Bayh-Dole’s most important section–the thing that will boost American manufacturing and restore American leadership in innovation–turns out to be the weakest part of the law–and the weaknesses are designed into section 204.

We are only half way through with the first part of section 204. There’s an immediate weakening of 204’s primary restriction on exclusive licenses to use or to sell:

unless such person agrees that any products embodying the subject invention or produced through the use of the subject invention will be manufactured substantially in the United States.

No one can grant exclusive licenses to use or to sell subject inventions in the US–unless. The clause that follows is presented as a broad demand for US manufacture. In part it is, but in large part it actually is not. We have already shown that the scope for the demand is almost uselessly narrow–section 204 applies only to exclusive licenses to use or to sell in the United States. The requirement doesn’t apply to owners or to non-exclusive licenses or to exclusive licenses for non-U.S. jurisdictions. What does it then matter if what follows on this narrow scope is stated broadly? Not much. Let me show you.

An institutional owner of a subject invention can license exclusively to sell or use in the United States only if the licensee agrees to use or sell certain product “manufactured substantially in the United States.” The licensee need not have an exclusive license to make the invention–the licensee might acquire product from others authorized to do the manufacturing, or the licensee may have only a non-exclusive license to do the manufacturing and thus both build its own product and acquire product from others. The exclusive licensee to use or sell complies with this requirement if the product that the licensee uses or sells has been “manufactured substantially” in the United States.

Let’s consider some oddities of this requirement:

  • depends on a private agreement
  • focused on products
  • does not define manufactured
  • invokes the qualifier “substantially”
  • makes no mention of import

Private Agreements

The walk back on the prohibition on granting exclusive licenses to use or sell in the United States takes a form similar to the “letter of assurance” in export control law. Here, the licensor is required to require the licensee to “agree” to substantially manufacture product in the United States (or to use or sell product manufactured in the United States, presumably from a company licensed to manufacture product in the United States).

Section 204 states an outright condition on owners of subject inventions–they cannot license exclusively the right to use or sell in the United States unless they get an agreement from the licensee regarding their use of manufactured product. We have proposed a more meaningful provision would place this limitation on exclusive licenses granted in foreign jurisdictions by American small businesses and nonprofits. Since section 204 is part of  federal patent law–and therefore applies to all those in the United States–why not just draft 204 to require anyone in the United States owning a subject invention or holding an exclusive license to use or sell product embodying a subject invention or made through the use of a subject invention to have that product manufactured substantially in the United States? That is, make it illegal–rather than just a matter of a private agreement–to import foreign manufactured products under an exclusive foreign license. You see the workaround? A patent owner licenses the right to manufacture exclusively to a foreign company. The patent owner then grants non-exclusive licenses to use and sell in the United States, or sells the product itself. But where does product come from? Ah, the foreign manufacturer. Section 204 couldn’t care less.

Why, if patent law could be used to assert what must happen, should a patent owner have to go on a scavenger hunt to obtain an agreement from each exclusive licensee, but only in the United States? And why does it matter when there are many workarounds?

A letter of assurance approach is much weaker than a statutory approach. A letter of assurance–in this case, that the licensee “agrees”–doesn’t provide much assurance that the licensee will in fact comply with the agreement. The agreement, after all, is a private agreement and does not appear even to rise to the level of being a contractual promise. The patent owner has no obligation to enforce the agreement, and there’s not really any dire consequence for the patent owner if the licensee fails to comply. It’s the licensee that has the problem. This is made clear in Bayh-Dole’s march-in provisions at 35 USC 203. There, the federal remedy (and public protection) for a failure to obtain the 204 required agreement or the failure to comply with it is that a federal agency can march-in and require (after a prolonged bureaucratic procedure) the patent owner to grant other licenses. The patent owner gets the benefit of income, then, from multiple licenses after having proven the market with an exclusive licensee, say, taking the risk of the development costs.

The agreement specified here in 35 USC 204 is made a condition of granting an exclusive license to use or sell in the United States. It does not require that the agreement become part of the grant of rights. That is, there is nothing in section 204 that stipulates that patent licensing agreements must contain this agreement language. The effect of making 204 compliance a matter of private agreement rather than statute is that enforcement then is routed to the march-in procedures–that is, the jurisdiction for administration is the federal agency providing the research or development funding, and not, say, the Attorney General tasked with enforcing federal law.

Products, Not Invention

The immediate walk-back in 204 allows exclusive licensing to use or sell provided that the licensor obtains the agreement of the licensee that certain products will be “manufactured substantially” in the United States. Thus, on its face, 204 does not apply to methods. An exclusive license to use a patented method in the United States lies outside 204’s requirements. Section 204’s requirement, however, is directed at products, not at the subject invention: products embodying the subject invention [compositions of matter, articles of manufacture] or produced through the use [methods, processes] of the subject invention. Thus, if a method produces a product, the product may come within 204’s requirement. But if a method provides information–such as quality testing or a disease diagnosis–then it’s difficult to see how 204’s requirement applies.

This is another instance of a broad scope on a narrow requirement, potentially broader than the reach of claims in an issued patent. A licensee of the exclusive right to use or sell has to agree to acquire product manufactured substantially in the United States. To comply, the exclusive licensee to use or to sell does not need an exclusive license to “make” inventions (or even a non-exclusive license to “make”)–the licensee just must have compliant product available for use or sale.

However, the “produced through the use of the subject invention” includes a much wider range of possible products, well beyond those that themselves embody an invention or even those that could only be produced through a patented process. Consider, for instance, an invention that improves the positioning controls on a precision milling machine. The milling machine may produce many products, but if that milling machine includes the inventive extension, then perhaps all such products come within the scope of 204’s requirement. This is the kind of stuff that might raise antitrust concerns, if a patent owner attempted to claim royalties on the sale of all such products, but Bayh-Dole’s disavowal of antitrust immunity in section 211 does not necessarily apply to section 204, since section 204 asserts its precedence over section 211. Think about that. Section 204 may create an immunity from or defense to actions under antitrust law.

To give a sense of the breadth of 204’s manufacturing requirement, consider an invention that adds a function to, say, a digital display. Let’s say the invention consists of components that might be packaged into an integrated circuit to improve display contrast. The overt requirement of 35 USC 204, then, is that any product that embodies this integrated circuit must be manufactured substantially in the United States. That would mean display panels that incorporate this invention, and any televisions, computer monitors, and other products that use these display panels–all these also would have to be manufactured substantially in the United States, if a patent owner granted an exclusive license to sell in the United States. For composite technological products having hundreds to thousands of components, this is an impossible demand, on the face of it.

Or, put another way–this demand is precisely where Bayh-Dole should have to be if it were really about American technological leadership. It would be exactly this kind of invention, basic to many applications, that ought to be manufactured in the United States, along with every product that includes such invention or is made through the use of the invention. Alternatives: license non-exclusively the right to use or to sell; don’t include or use the invention in making the product. But Bayh-Dole sets up section 204 compliance to be at federal agency discretion, not as a matter of federal law.

Consider, further, “product” since that’s the core of 204’s restriction on exclusive licenses to use and sell in the United States. In our example, is the “product” the integrated circuit–that would have to be fabbed in the United States? Or does “product” extend to any product that uses the integrated circuit? There is no guidance. No one, apparently, thought it through, or if they did, they deliberately chose not to explain what ought to happen. We might offer the conjecture: the purpose of the clause was to make legislators feel good about the law so they would vote for it or at least not oppose it, and not to mean anything in practice (other than to create more bureaucracy and bother).

A similar work up can be done for “produced through the use” prong of the scoping statement. The more general the method or process, the more products downstream will also have to be manufactured substantially in the United States. Clearly, the scope of the walk-back requirement is potentially much, much broader than simply whether a given subject invention is “made” in the United States.


We expect “to manufacture a product” to be a pretty straightforward thing. To manufacture means, in a typical dictionary definition tracking common usage:

Black’s Law Dictionary on-line has the following, for the noun form:

So far so good–to manufacture something is to make it, not gather it or birth it. But we might ask whether manufacturing involves just the inventive portion of the product, or the entire product. If an invention improves an automobile’s autopilot system, is the product a chip containing the new steering algorithm, the steering module of the autopilot system, the autopilot system itself, or the automobile? The section 204 scope is any “products embodying the subject invention”–and any of these elements might meet that definition, from the integrated circuit to the entire vehicle.

Bayh-Dole muddies the water with regard to manufacture by using the term in its definition of “practical application” (35 USC 201(f)):

The term “practical application” means to manufacture in the case of a composition or product, to practice in the case of a process or method, or to operate in the case of a machine or system; and, in each case, under such conditions as to establish that the invention is being utilized and that its benefits are to the extent permitted by law or Government regulations available to the public on reasonable terms.

The inventions that may be “manufactured” are compositions of matter and products. Processes, methods, machines, and systems are distinguished by “practice” or “operate.” Plants (some of which may be patented) are not mentioned at all. If the policy and objective of Bayh-Dole is to promote the practical application of subject inventions–and that’s what this definition implies, standing in for “utilization”–then it would appear that manufacture in Bayh-Dole is limited to compositions and products, and does not include processes or methods or the operation of machines, even if patentable.

In practice, it is not at all clear what “manufacture” means in the statute–we know the common meaning of the word, and we can categorize many activities as involving manufacture, but we can’t reason from the abstraction back to any particular instance of manufacturing and say with confidence this is the meaning intended–we cannot differentiate an integrated circuit embodying a subject invention from an automobile that happens to have that integrated circuit as one part among thousands. If one asserts that all of these variations are intended, then the scope of the requirement is truly broad.


The complication arises when we add the adverb “substantially.” This addition is ambiguous in multiple ways. Does “manufactured substantially” mean that most product is manufactured in the United States and some product could be manufactured elsewhere? Or does it mean that for each product, most of the parts? or value? or time spent manufacturing? must be from the United States. Who can say?

Furthermore how does one determine whether something is “substantially” manufactured in a given location? If all the parts are made in China, but the final assembly is done in the United States, is the product manufactured substantially in the United States? Even if the parts subject to the patent on the subject invention are made in China? What if the parts are made in the United States at great expense, but the time-consuming and specialized assembly is done in China? Where has the substantial part of the manufacturing been done?

There are plenty of variations on this theme. Other than grossly simple production work all done in one location, there is no good way at all to establish what “manufactured substantially” means. Does “substantially” track the number of parts, the time involved,  the importance of parts or function, assembly and packing, the relative size of parts, or merely location?

At the very least, we can be sure that “substantially” walks back the manufacturing requirement. Whatever the requirement is, it does not have to be fully met, just “substantially” met. For complex products where parts and assembly are sourced from multiple locations, it may be impossible to construct a reliable account of the time, cost, proportion, relative value of the pieces of manufacturing that are required to produce the final product in its packaging, ready for sale. While one may find it attractive to interpret an ambiguous statute against the drafting party, those willing to agree to manufacture substantially in the United States often don’t find much comfort in doing so. As we will see, however, there are process alternatives to circumvent the requirement, and in any event, the clause has never been enforced, since it can only be enforced by means of the march-in procedures in section 203–procedures that have never been used to march-in and which were designed not to operate.


More strangeness. If 204 were really pressing for a preference for United States industry, why not move the restriction so that no small business or nonprofit invention owner may grant a foreign right to use or sell without requiring substantial manufacture in the United States? That would be a clause favoring American exports based on American manufacturing. Wouldn’t that be the thing to address trade deficits and American loss of leadership in technological innovation? It might be ill-advised, but then most everything in Bayh-Dole is ill-advised, so why not here, also?

And if section 204 is intended to use the patent system as a trade barrier–keeping foreign manufacturers of American inventions out of the American market–then why not simply prohibit the importation of any foreign-manufactured product, whether licensed or not? “No small business firm or nonprofit organization shall import any products embodying the subject invention or produced through the use of the subject invention.” Make them here, or get federal government approval to import. That would be a meaningful preference for United States industry. But that’s not what 204–the most important provision of Bayh-Dole–requires.

Part 3 of this article is here.

This entry was posted in Bayh-Dole and tagged , , . Bookmark the permalink.