You Shouldn't Have To Hire A Patent Lawyer Before You Can Innovate
from the predictability dept
One of the persistent themes I noticed at Wednesday's patent conference at the Brookings Institution is that most of the lawyers seemed to assume that if the legal system ultimately reaches the right conclusion—invalidating a bad patent, say—that this means that the patent system is working well. Some panelists suggested that the Bilski decision, which struck down one particularly egregious "business method" patent, shows that there's not really a problem, because the courts are recognizing the problems with bad patents and correcting them. They seemed not to fully appreciate how slow and expensive the legal system is. One only has to think back to the great BlackBerry showdown to see that having the legal system eventually invalidate a bad patent may not be good enough. Even if the law is on the side of an accused patent infringer, the time, expense, and uncertainty of litigation can kill the firm before its rights can be vindicated in court.
I think the right way to think about patent reform is not whether the courts eventually reach the right result, but whether the system is predictable enough that you can tell in advance what the law requires, without hiring a patent lawyer. After all, this is how well-designed property rights systems work. I didn't need to hire a property lawyer to tell me who owns the apartment I'm living in—the rules of real property are predictable enough that I could figure it out on my own. The vast majority of property transactions are the same way—lawyers only get involved in exceptional cases that involve large sums of money or tricky legal issues. By the same token, if we're going to have patents on software (or in any other industry), they should be few enough and clear enough that a smart entrepreneur can figure out in advance, without the help of a lawyer, which patents he needs to license. If our current patent system isn't living up to that standard, the solution isn't to come up with ever-more-complex legal doctrines trying to separate the "good" vague patents from the bad ones. Rather, the solution is to restrict patenting to those fields where it's possible to make things clear and predictable. If that's not possible in some industry (and I suspect it's not in software), then that's a sign that patents aren't an effective way to promote innovation in that industry.