Distribution, Competition, and Antitrust / IP Law

Potential Patent Licensing “No-No” #7: Royalty Provisions Not Reasonably Related to the Licensee’s Sales

Yes or No?

This is the seventh in a series on potential licensing “no-nos.” You can find the previous installment here.

Even monopolists are entitled to engage in business and earn a profit. Hence, as a general rule, patentees – even assuming that they have a monopoly (and as I’ve noted repeatedly, just having a patent does not mean that a patentee is a monopolist) – can charge what this wish for their patents.

Often, a patentee will charge royalties based on the number of units of product sold, or that equal a percentage of licensee revenues. If the license is conditioned on royalties on products “which do not use the teaching of the patent,” then it may be unlawful as a matter of patent law. See Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U.S. 100, 135 (1969). “[J]ust as the patent’s leverage may not be used to extract from the licensee a commitment to purchase, use, or sell other products according to the desires of the patentee, neither can that leverage be used to garner as royalties a percentage share of the licensee’s receipts from sales of other products; in either case, the patentee seeks to extend the monopoly of his patent to derive a benefit not attributable to use of the patent’s teachings.” Id. at 136.

But this rule is probably no longer per se.  Under the Patent Misuse Reform Act of 1988, 35 U.S.C. § 271(d)(5), a patentee can condition a patent license on the purchase of a separate product, unless, in view of the circumstances, the patent owner has market power in the relevant market for the patent or patented product.  If a patentee has the greater power to condition purchase on a non-patented product, it would seem to have the lesser power to charge royalties based on purchases of non-patented products.

Despite the general Zenith rule, royalties can be calculated according to sales of non-patented goods when the “convenience of the parties rather than patent power dictates the total-sales royalty provision.” Id. at 138. If the licensee is given the option of paying only for patented goods, but for convenience’s sake chooses to pay a fee calculated according to all goods sold, the arrangement is permissible.  However, determining when a royalty is “conditioned” as opposed to when it is “convenient” can be difficult to do in practice.

The Supreme Court in Zenith was addressing patent misuse doctrine. There is a separate inquiry as to whether “metered tying” – where a patentee charges a relatively low price for a patented product, and then ties the product to non-patented products (usually parts or supplies) – should be viewed as an antitrust violation. I covered that issue in a series of three posts here, here, and here.  Under current tying law, such a tie could indeed violate the antitrust laws. I won’t repeat the full tying analysis here.  Feel free to check out my recent presentation on tying law and avoiding liability, which is available in the downloads section.  However, a broad royalty base — one which includes non-patented products — doesn’t really amount to a tie; the element of forced purchases from the patentee seems to be lacking.

Instead, the potential antitrust danger with a broad royalty base is that it could, at least in theory, curtail competition in the non-patented product market, because if a licensee is already paying a royalty on non-patented products, it may have a disincentive to purchase or use competing products.  In that connection, I should note the Microsoft case (United States v. Microsoft Corp., 56 F.3d 1448 (D.C. Cir. 1995)). There, the government challenged operating system license fees paid by OEMs calculated according to the number of computers shipped, regardless of whether the computers were loaded with Microsoft’s OS. In other words, Microsoft licensed the OEMs on a “per processor” basis.  Since OEMs had to pay Microsoft for each computer shipped, they were arguably less likely to pay to install a competing OS on their computers. Microsoft agreed in a consent decree to charge the OEMs license fees only for computers actually loaded with the Microsoft OS. The precedential value of a consent decree is of course quite limited.

From an economics point of view, a broad royalty base may have no significance if the non-patented product and the patented product are used in fixed proportions.  For example, if for some strange reason a patentee had a patent on a type of left shoe, and charged royalties on left shoes and right shoes, the royalties would not affect competition for right shoes (assuming left and right shoes are always sold together).

Related posts:

About Howard Ullman

Antitrust, competition, and IP law enthusiast.


  1. […] This post is the eighth in a series concerning potential patent licensing “no-nos.” The previous post can be found here. […]

Leave a Reply

Optimization WordPress Plugins & Solutions by W3 EDGE
%d bloggers like this: