The goal of patents is to encourage innovation. It does this by granting an inventor a temporary monopoly over his invention. This gives the inventor the security of knowing that he will have time to recoup his research and development costs without someone taking his invention and marketing his work against him. Conversely though, it prevents competition and promotes a monopoly. Patents attempt to balance the evils of a monopoly and the need to encourage investment in innovation. So how does this all play out in reality? More specifically, how does it play out in reality with regards to software patents?
In the last post, I pointed you to an article by former Sun CEO Jonathan Schwartz. In the article, he discusses how Apple and Microsoft came to him threatening to sue Sun for violating their respective patents. Scwartz didn’t deny that Sun was most likely infringing, but instead he responded with the claim that Apple and Microsoft were both infringing on Sun’s patents as well. This is the mutual assured destruction (or M.A.D.) approach to intellectual property and we might get to see the effects of this type of an all-out patent war with the Apple and Nokia lawsuit. Nokia is suing Apple for patent infringement claiming that the iPhone infringes on several of its patents. In response, Apple has filed suit against Nokia for patent infringement claiming that its phones violate patents Apple obtained for the iPhone. From this and other cases, it wouldn’t be that far of a stretch to assume that big software companies infringe on the intellectual property of other companies all the time. With this in mind, let’s discuss the effects that patents have on software innovation.
My general view is that state-enforced monopolies are a generally bad idea and can only be justified with good reason. While the government most certainly has the authority given to it by the U.S. Constitution to grant these monopolies for a limited time, the question should be whether these rules promote innovation or stifle it? In some industries that have massive research and development costs to produce a product that could easily be copied once it is on the market, I believe the answer is that patents help promote innovation. But what about in the case of software? First of all, software already has copyright protection, preventing competitors from being able to just duplicate existing products. This protection goes beyond just verbatim copying, giving software authors protection of the binary, structure, and design. So software developers don’t have to worry about the competition stealing their work. However, copyrights don’t protect the key process or algorithm of the software. Competitors can just recreate the algorithm themselves without having to do the hard work of determining how to create it. To put it a different way, all solutions seem easy… once somebody else figures them out. The protection of the algorithm is where the patents come in. Would software companies continue to innovate if anyone could take its algorithm? It depends on what percentage of the software development process the particular algorithm creation constitutes. To clarify, I would like to limit this to those algorithms that are patent worthy (non-obvious, novel, etc.). In my personal experience, I would have to say – it depends. If we are discussing what I consider to be patentable algorithms, then algorithm creation constitutes a good portion of the development time. If we are discussing all algorithms that the patent office has accepted, then algorithm creation time for most of the accepted patents probably doesn’t constitute a significant portion of the development time for any marketable application. I’ll discuss bad software patents and eventually what I think the bar for patentability should be later, but for now let me just say that I believe that most development companies would continue to create new software regardless of software patentability.
We know that large software companies will continue to create new software knowing that other big software companies will infringe on its patents. Companies will do this because it knows it is or will be infringing on the patents held by the original infringing companies. Given the number of software patents and the wide swath of technological areas they cover, it would be practically impossible (as well as time and cost prohibitive) to ensure that a large software product does not infringe on any existing patents. Large companies don’t seem to care because it has its defensive patent “war-chest” to protect it against claims (this does not apply when they are sued by “patent trolls” which I will discuss later). However, small developers do not have this protection. To make it worse, there is no exception for innocent infringement. It might seem fair to give immunity to a developer who discovers an algorithm all on his own, ignorant of the existence of a patent. In fact, one can argue that the fact that another developer independently discovered the algorithm makes the it non-novel and therefore unpatentable. The problem with this line of thinking is that novelness is determined at the time at which the discovery occurred. I reiterate, all solutions seem easy… once somebody else figures them out. So the current patent system seems to give preference to large, established companies while making it extremely difficult for small companies to create software.
Much like an abusive boyfriend who beats his girlfriend with one hand while gently wiping away her tears with the other, software patents hinder innovation while at the same time provide much needed protection to the innovation that occurs. The first inquiry asks whether or not developers would continue to create software if they were not afforded patent protection. I said most developers would continue. I stand by that. The problem is the few that wouldn’t continue. The developers who could continue are the ones who improve on the existing paradigm. The ones who wouldn’t are the ones who change the paradigm. They are the ones who dedicate teams of scientists and engineers to figure out problems. They are the ones who dedicate months and years to perfecting a process. As much as software patents hinder those who wish to improve upon the current state of software, they encourage and protect those who take software in a new direction. The second example demonstrates how small businesses are at a disadvantage when it comes to creating software because of software patents. However, those same patents provide small businesses with the only protection it has against its larger competitors. Imagine a small company that creates a word processing application. Imagine further that they add some productivity feature that no other word processor has implemented before. If the company is able to patent this feature, it now has a marketable product. Without the patent, Microsoft could take the feature and incorporate it into Word. Some would say that this is good, let the market decide the better product, give the people choices. But they fail to consider that Word is entrenched in corporations and Microsoft likely has a much larger advertising budget than its small competitor. Without patents, the small company would spend all of its time and money on innovation only to have the entrenched competitor steal its innovations. This leads to permanent monopolies, not the temporary ones that patents provide for.
One last note before I finish this post: it is the process that is patented, not the result. If a software mapping company patents a process by which geographical starting and ending points are converted into directions, it is the process by which it does this that would be patented. This means that another software mapping company can still create a process that converts the two points into directions, it just can’t use the same process to do so as the first company. This leads many to believe (e.g. Gene Quinn – but I can’t find the post where he said so) that patents encourage innovation by forcing competitors to find different ways to solve the same problem, hopefully coming up with better and more efficient methods along the way. If the bar to getting a patent on software was where I believe it should be, I would completely agree. Unfortunately, it is not there yet, more on this in a later post.
Monday I will discuss the effects that bad patents and patent trolls have on innovation. Then I will wrap this whole thing up with what I believe should happen to the software patenting process and what I believe the bar to getting a software patent should be. I will leave you with this tangent to think about, and if you know anything about corporate tax law, please comment. If a company refuses to enforce its patent rights against a competitor because the company is also infringing on that competitor’s patent rights, and this is construed as a barter, how would that affect its taxes? I know that when people barter goods or services, they are supposed to report the fair market value of the good or service to the Internal Revenue Service. I wonder if the IRS were to look into this more closely, would it effect the way companies report these patent standoffs.