Eli Dourado is a research fellow at the Mercatus Center at George Mason University with the Technology Policy Program.
Silicon Valley firms have complained about "patent trolls" for years. As nonpracticing entities, trolls acquire patents but don't produce anything with them except lawsuits against companies that may (or may not) infringe on their patent portfolios. In recent years, trolls have even become a tool that practicing companies use to stifle competition. By transferring a patent to a troll and retaining a license, firms can inflict legal pain and distraction on their competitors.
To remedy this problem, Reps. Jason Chaffetz, a Utah Republican, and Peter DeFazio, an Oregon Democrat, have reintroduced the SHIELD Act, which would require nonpracticing entities who lose such lawsuits to pay for their adversaries' legal fees. While the SHIELD Act may provide some modest relief for Silicon Valley firms against spurious lawsuits, it doesn't solve the real problem that the Valley and other industries have—that patents do more harm than good in their particular area of innovation.
There is good evidence that patents spur innovation in some industries, like chemicals and pharmaceuticals. It is easy to find out if a particular molecule has already been patented, and therefore easy to avoid infringement. And to create a new, not-yet-patented molecule, it is not usually necessary to license very many other patents.
In contrast, to undertake other kinds of innovation, like in software or business methods, it is much more difficult to know when you are infringing;on someone else's patent. Patents in these areas are vaguely worded, hard to search for, and subject to honest, mistaken infringement. Furthermore, some products, such as smartphones, require the assembling of thousands of patent licenses, sometimes from competitors. The cost of negotiation and strategic maneuvering is significant.
While the SHIELD Act is probably better than nothing, it is time for Congress to consider more thoroughgoing reform of the patent system. In doing so, they should consider the two factors discussed above: 1) how easy is it to avoid unintentional infringement; and 2) could the patent slow down sequential innovation?
One good reform would be to simply eliminate patents in areas that fail these two tests. This would mean that software algorithms, business methods and even electronics would be patent-free zones. Innovation would be driven by the desire to get the best products on the market first, not by grants of monopoly that slow down other potential innovators. Specific software products would still be protected by copyright law. Meanwhile, patents could be retained in pharmaceuticals, materials science and other chemicals, since these areas pass the two tests above.
Another important long term reform would be to eliminate the Federal Circuit Court of Appeals as the only appeals court that is allowed to hear patent cases, which it has been since 1982. Because the court is partially specialized in patent law, it has been largely "captured" by patent interests. Its rulings have consistently expanded the scope of patent protection, including to software and business methods. By returning patent appeals to all of the Circuit Courts, patent jurisprudence might naturally become less lopsided and more sensible over time.
The sponsors of the SHIELD Act deserve credit for recognizing that there are severe problems with our patent system. Patent trolls do inflict unnecessary costs upon honest innovators who just want to provide a good product to the market. But it is important to recognize that trolls are merely a symptom of a greater patent disease. If we fix the scope of patent law, the troll problem will go away.