What Blanket Cross-Licensing Says about Software Patents

by on November 15, 2006 · 8 comments

It occurs to me that the popularity of blanket patent cross-licensing agreements among software firms is pretty strong evidence that software patents don’t do much to incentivize creativity. This isn’t how healthy patent or copyright markets behave. For example, when’s the last time two big Hollywood firms announced blanket cross-licensing deals for their movie libraries, wherein each promises not to sue the other for copyright infringement if they pirate each others’ movies? Or how often do pharmaceutical companies sign blanket cross-licensing agreements permitting each to produce the other’s drugs without paying for them? Or, for that matter, how often do software companies sign blanket cross-licensing agreements for their software copyrights?

The reason these things almost never happen is that movie copyrights and drug patents represent genuine economic assets with clear boundaries. A movie copyright protects the right to distribute a particular movie. A drug patent represents the right to manufacture a particular drug. And a software copyright represents the right to distribute a particular bit of code.


This is economically efficient because, as Lewis explains, companies often face a question about whether to build technology themselves or buy them from somebody else. Movie and software copyrights and drug patents provide an elegant mechanism for one company to develop a valuable asset, claim ownership of it, and then license it to a third party.

In contrast, a software patent doesn’t cover any particular thing. They cover vague categories like wireless email and one-click shopping. These aren’t discrete economic assets that can be bought and sold. If a company licenses NTP’s wireless email patent, it doesn’t get any code or schematics for building a wireless email device. All they get is a promise from NTP not to sue. Likewise, if somebody licenses Amazon’s one-click patent, they don’t get to use Amazon’s one-click software (that’s already covered by copyright, which would require a separate license). All you’d get from licensing the patent is assurances that Amazon won’t sue you later.

In short, a software patent doesn’t represent anything in particular. Which means that in practice, a software patent amounts to nothing more than the right to hire a lawyer to harass your competitors. And that’s why successful software companies don’t buy and sell them like economic assets. Instead, they just sign agreements promising not to sue. Sometimes, the company with fewer patents agrees to pay some royalties to the company with more patents in recognition that the latter has more firepower on his side. But the primary goal of the agreement is to allow them to ignore the patent system and focus on actually building useful products. It’s hard to see what purpose that serves.

  • http://weblog.ipcentral.info/ Noel Le

    This returns to your theme of clear boundaries. I know it’s a libertarien concept, but many things exist in the world without bright lines. Clear boundaries are a political or ideological construct, not *necessarily* an economic one.

    the popularity of blanket patent cross-licensing agreements among software firms is pretty strong evidence that software patents don’t do much to incentivize creativity.

    Patents incent capital investment, not creativity unless you’re talking about prospecting patents for different commercialization opportunities. In any case, explain how this pertains to the rest of your post. How is your statement a reflection on the value of patents. Are blanket cross-licensing agreements the dominent form of licensing in the industry- that seems to be your presumption.

    a software patent doesn’t represent anything in particular. Which means that in practice, a software patent amounts to nothing more than the right to hire a lawyer to harass your competitors.

    Where do you gather this. Do you know how much licensing occurs every day in the industry. IBM rakes in a nice $2B plus.

    Licensing happens via several reasons. A company will find a patent or find out another firm is working on a patented technology. It contacts that firm, and lets say negotiations go well, and the company licenses the patent. They don’t just get the facial value of the patent though- the licensor will have incentive to see how their patent is commercialized or improved upon (they might even have a share in profits), and thus may share with the licensee some information not contained in the patent claim.

  • http://weblog.ipcentral.info/ Noel Le

    This returns to your theme of clear boundaries. I know it’s a libertarien concept, but many things exist in the world without bright lines. Clear boundaries are a political or ideological construct, not *necessarily* an economic one.

    the popularity of blanket patent cross-licensing agreements among software firms is pretty strong evidence that software patents don’t do much to incentivize creativity.

    Patents incent capital investment, not creativity unless you’re talking about prospecting patents for different commercialization opportunities. In any case, explain how this pertains to the rest of your post. How is your statement a reflection on the value of patents. Are blanket cross-licensing agreements the dominent form of licensing in the industry- that seems to be your presumption.

    a software patent doesn’t represent anything in particular. Which means that in practice, a software patent amounts to nothing more than the right to hire a lawyer to harass your competitors.

    Where do you gather this. Do you know how much licensing occurs every day in the industry. IBM rakes in a nice $2B plus.

    Licensing happens via several reasons. A company will find a patent or find out another firm is working on a patented technology. It contacts that firm, and lets say negotiations go well, and the company licenses the patent. They don’t just get the facial value of the patent though- the licensor will have incentive to see how their patent is commercialized or improved upon (they might even have a share in profits), and thus may share with the licensee some information not contained in the patent claim.

  • http://mcgath.blogspot.com Gary McGath

    “But the primary goal of the agreement is to allow them to ignore the patent system and focus on actually building useful products. It’s hard to see what purpose that serves.”

    Isn’t it obvious? You’ve just described a cartel setup — created, as always, with the aid of the government.

  • http://mcgath.blogspot.com Gary McGath

    “But the primary goal of the agreement is to allow them to ignore the patent system and focus on actually building useful products. It’s hard to see what purpose that serves.”

    Isn’t it obvious? You’ve just described a cartel setup — created, as always, with the aid of the government.

  • http://weblog.ipcentral.info/ Noel Le

    TIm has a funny way of putting it. Nobody is ignoring the patent system, they’re leveraging its flexibility through private contracting. Evidently, Tim thinks that if the patent system leverages other aspects of private ordering, its a broken system.

    Whats funniest about his comment is that apparently, like the Microsoft-Novell deal Braden Cox wrote about, Tim doesn’t like firms working with patent holding companies.

  • http://weblog.ipcentral.info/ Noel Le

    TIm has a funny way of putting it. Nobody is ignoring the patent system, they’re leveraging its flexibility through private contracting. Evidently, Tim thinks that if the patent system leverages other aspects of private ordering, its a broken system.

    Whats funniest about his comment is that apparently, like the Microsoft-Novell deal Braden Cox wrote about, Tim doesn’t like firms working with patent holding companies.

  • http://pharmacynewsblog.com/ Leo

    The pharmaceutical companies every year increase sales volumes who it supervises? WBR LeoP

  • http://pharmacynewsblog.com/ Leo

    The pharmaceutical companies every year increase sales volumes who it supervises? WBR LeoP

Previous post:

Next post: