Too Many Patents?

First, I decided to take a bit of a break over the weekend.

What follows is a posting from a friend of mine, named PatentPanda, about some issues in patent law and policy.

I try to keep a good collection of friends with various expertise, and I am thrilled to have PatentPanda writing.

I'm thrilled to be the first guest post on Not for Polite Company. I'm a patent attorney in D.C., where I lead the emerging business practice of a small firm, and am excited to be writing about the mess with software patents (the opinions are my own). Basically, they are too many of them.

There are a lot of folks who really don't like software patents. Some, like google and microsoft, come to their opinions from a financial perspective. Quite simply, their defensive efforts (litigation, damages, due diligence) cost more than they earn by offensively, e.g. licensing out their patents or the competitive advantage that they receive from employing patented technology. This is as it is supposed to be. A bold statement, I know.

To appreciate the problem with too many patents, it's worth first taking a moment to look at what the patent system is supposed to accomplish. The constitution talks about securing rights to inventors to promote science. Securing the rights against who? Presumably competitors. Without patents, a technical person would need to invent something and then sell it. With patents, the technical person need only invent something, and can be paid once someone else has sold it. Just as the advent of the corporation lets people specialize in things like "human resources," patents remove the need to be good at both inventing and business.

Before we go further, I should note that the patent world exists in two halves: electrical and life sciences. "Electrical" encompasses not only electronics, but software, business methods, and even general mechanical stuff. Life sciences is the kind of thing where having a Ph.D. is really helpful: chemistry, biology, DNA, etc. The following stuff is applicable for electrical stuff, but is often completely wrong for life sciences.

Patents are an important barrier to entry. For a large software company, their value is relatively low. Say you want to write a competitor to Windows. You, not google. The sheer scale of such a project, much less launching it, is so huge that only very large players can hope to compete. More importantly, a small company can't take down Microsoft without doing something entirely new and better. So Microsoft gets relatively little protection from patents.
If you are a small company, it's not that hard for a big company (e.g. Microsoft or Google) to put you out of business by moving into your space. Whatever resources you command, they can muster. There are a few exceptions, and patent protection is often way up on that list. For a small company, patent protection can be vital.

Each of these approaches is driving the huge number of software patents. IBM began the picket fence approach, and Microsoft copied (shock, I know) by hiring away Marshall Phelps, who set up the program at IBM. When a small company would tell IBM that IBM was infringing a patent, IBM's response was often something to the effect of "here's 2000 claims, we're sure that you're infringing something."

This evolved, as capitalistic things tend to do. The patent portfolios of big companies continue to swell, and continue to cover lots of little things in an effort to make sure that they have some protection everywhere; they always have something to trade. The small companies have learned not to bother big companies until the small companies are not doing anything (and thus don't care how many patents the big company has). Not doing anything generally takes one of two forms: either the company tried doing something, failed and now has only the patents left, or never did anything but license the patents in the first place.

Patent portfolios have grown because patents are so powerful. There are significant damages, and even more fear inducing, injunctions. There are very few things that could shut down Windows, and a patent is on that list. This leverage is a key driver in the growth of patents, and the reason why there are more software patents than other types. Unlike mechanical inventions, digital ideas can be replicated nearly effortlessly, leading to the possibility of fabulous riches within several years of starting the company.

So now there are a lot of small companies generating software patents, and the big companies are generating lots of software patents, and both are working on scale far larger than either mechanical or pharmaceutical outfits. This raises everyone's transaction costs.
More patents makes it harder to tell if a new product that you are launching infringes a patent. If patents were rare, you could go poke around, and have your attorney review the half dozen important ones in your field. Due diligences for software often scope out thousands of patents to be considered before doing a deal. Now patent lawsuits against large software products are almost an inevitability.

The fear is that folks will have to spend more time dealing with patents than code, which would be bad.