Jack Chou commented yesterday on this blog about the patent arms race and the need for patent reform. On Monday, Google announced its $13B bid to acquire Motorola’s patent portfolio to defend Android, the latest example of companies spending billions on IP defense. I agree with him.

The Patent Office is a mess

Patents take years to be evaluated because the Patent Office is overwhelmed with increasingly technical patents across many fields. There are 1.2M patent applications awaiting evaluation. To put that in perspective, 7 million patents have been granted since the Patent Office’s inception in 1790. I make no mention of the billions of dollars spent annually on legal fees for patent disputes detailed beautifully in this infographic from Focus.com.

As a result, there have been claims the current patent system is hindering innovation and there have been many suggestions for patent reform. Most focus either on (1) streamlining the patent filing process or (2) removing patent protection from certain kinds of inventions.

A typical process improvement might shift the burden of prior art discovery to inventors. Today, the beleaguered Patent Office spends days researching previously filed relevant patents, called prior art, for each application. Requiring inventors or private companies to perform this research might speed the approval process along.

Fred Wilson, among others, has called for the elimination of the business process patents. These patents protect novel ways of doing things, not new technologies and are often quite broad, practically ensuring they will be used in patent litigation in the future.

That’s all pretty complicated. But I think patents have a more fundamental problem: they last for 20 years.

Patent History

The first recorded patent granted in 1421 for an assembly to hoist materials across the Arno River, was granted for 3 years. 28 years later Henry VI granted the first 20 year patent for the production of colored glass.

Patents are designed to ensure that the inventor, who toiled away to develop an invention, has an appropriate time to profit from the innovation. In the days of colored glass and basic cranes, companies moved slowly and it took years to commercialize ideas. Hence the 20 year patent period.

It’s time for a change

Unsurprisingly, this 600 year old policy of 20 year patent terms is outdated. Startups bring technology to market in 3 to 12 months. Technology adoption life cycles are shrinking dramatically meaning inventors profit from their innovations faster than ever. The New York Times published a beautiful chart demonstrating this compression. Click on the image for a bigger copy.

Faster technology adoption means an inventor should be able to capture the market value of a patent in a shorter period than 600 years ago. Unfortunately, the World Trade Organization still requires patents to be respected for at least 20 years.

The benefits of shorter patent durations

If we reduce the lifespan of a patent, technologies (and business processes) could enter the public domain more rapidly than before spurring more innovation. Most of the patent battles I’ve watched have been over older patents fought by large incumbents or patent trolls seeking revenues from tech monoliths. IT’s speed of innovation is tremendous and our patents should respect its cadence.

Once patents expire, innovation blooms. The 3D printing revolution, exemplified by MakerBot, exploded because of key patent expirations. MakerBot makes 3D printers, also called Rapid Prototyping machines, that cost 95% less than their competition. MakerBot sprang onto the market in 2008 because the 3D printing patents filed in 1988 had expired. Think of how much change we’ve seen since 1988. No one had mobile phones back then!

A corollary to this previous point is shorter patent periods require incumbents innovate quicker. If the 3D printing patents had lasted 10 years, we might have seen MakerBot in 1998, or seen lower cost 3D printers ten years ago!

Respecting industry differences

Reducing patent lifetimes makes sense for IT but because Innovation cycles vary by industry, it may not be a valid policy across all sectors. Biotech companies struggling to pass through FDA trials require longer protection to recoup R&D costs. And the patent system should respect those particularities of industry. But I believe we should reduce the duration of software patents.

Determining the proper duration of a patent is a tricky issue but reducing the duration of a patent is one way of pursuing broad patent reform and spurring innovation while respecting the work of inventors.


5 thoughts on “20 years is too long for patent protection

  1. “Biotech companies struggling to pass through FDA trials require longer protection to recoup R&D costs viable. ”

    Ha! A myth that big pharma loves to hear whispered into its ear! The fact is that much of the costs of developing new drugs and treatments arise FROM patents (limiting what can be researched without licensing costs) and is mostly paid for by government grants anyway.

    We tax payers end up paying for research, then paying monopoly rents for the drugs and treatments that result.

    Limiting/eliminating patents for drugs and medical treatments would go a long way towards addressing the rising costs of healthcare in the U.S., and would spur innovation and better treatment as well.

    • Thanks, Paul. You know more about this area than I do! I referenced it because it seems to be a hot button topic for these companies in particular.

  2. One problem with shorter patent durations is that people who want to use those inventions without paying for them have less time to wait. This is especially true when strong network effects make the invention impossible to commercialize individually, or when the incumbents have a de facto monopoly in their space. Similarly, this disadvantages innovators who cannot commercialize their inventions, such as universities, private research labs and individual inventors.

    Basically, if nobody else can or will bring that invention to market and compete with them, the incumbents can afford to wait as long as they want, and they’d only be happy to wait less.

    Imagine you invented a novel wireless technology. Or an useful feature for cable TV. Or an awesome UX for a smartphone. Or even something as small as a valuable extension to MS Word. Are you really going to establish your own wireless standard and get manufacturers to adopt it, or start your own cable network, or build your own smartphone, or write your own word processor to compete with Word? Or would you rather have the incumbents bring those to market and pay you for them? (Each of these examples is based on actual cases. Cable companies, after being shown the invention, waited *17 years* to introduce the feature everywhere! Eventually the inventor had to sell the patent to a “patent aggregator.”)

    Also, I’d like a citation on the MakerBot example. As far as I can tell, MakerBot’s 3D printer never infringed any patents because their approach (fused deposition modeling) is very different from those invented and patented in the 80s (such as stereolithography and laser sintering). The earliest, “broad” 3D printing patent expired in 1999. I highly doubt any of these patents were what held MakerBot or RepRap up, as you seem to believe.

    • I agree that shorter patents may not serve inventors who cannot commercialize those technologies. But patents are meant to spur innovation. Holding onto ideas that are never brought to market leads to patent trolling. The alternative for these inventors is patent licensing or sale.

      You have a great point that antitrust law and patent law are intertwined and there are considerations for market dynamics particularly in the case of network effect dominated industries. Longer patent periods don’t solve this, though. In both cases, patents should be sold or licensed. If they cannot be brought to market, then the case that the patent is innovative and useful is weakened.

      Re: MakerBot, I chatted with the company. That’s what the CEO told me.

      • If the invention cannot be bought to market because of physical limitations or feasibility considerations, then, yeah, the patent is not very useful. But if it is network effects or monopolies or other economic effects preventing bringing the invention to market, it does not mean the patent is less useful.

        Let me assure you that no inventor ever want to hold on to their ideas, but really want to bring it to market. Even the wackiest guy knows the idea itself is worth nothing unless brought to the market somehow.

        One problem is, they may not be the right entity to commercialize it. Techies and researchers are rarely also good businessmen. Or maybe the relevant market is dominated by incumbents. Patents are, then, in theory, a good way for the little guys to monetize their technical talents.

        But this is not what leads to patent trolling. There are many, many reasons, none of them simple. Here’s one: individual inventors vastly overestimate the value of their ideas 🙂 Everyone thinks they have the next million dollar idea. As you can imagine, this leads to friction during negotiations, leading to complications… like companies simply going ahead without licenses. If that happens, individual inventors are ill-equipped to enforce their patents, which end up being sold to “patent trolls,” who make a business out of litigating them.

        While I agree that mandatory licensing may alleviate many problems, I don’t think that, or reducing the term of patents, will help much overall. I think raising the bar for patents is the right answer, and we’re already on the right track these days with the KSR decision in 2007.

        BTW, was it Bre Pettis you talked to? I met him in person at an event, but we never talked. I saw the MakerBot 3D printer in action that time, and it was fascinating.

Comments are closed.