(Jae C. Hong/Associated Press)

Innovation seems to be on the wane lately: “The quality of new patents around the world has fallen dramatically over the past two decades,” concludes a new OECD report on science and technology trends.

In nearly every country, patent offices are getting flooded with more and more patents for dubious or low-quality inventions. And that, the OECD report notes, leaves everyone worse off: “The rush to protect even minor improvements in products or services is overburdening patent offices. This slows the time to market for true innovations and reduces the potential for breakthrough inventions.”

Not everyone is suffering from these developments, though. Consider “patent trolls,” the derisive term for small companies that exist solely to stockpile patents — often vague patents on basic business methods — and then hurl infringement suits at firms trying to market new products. A new study by three researchers from the University of Boston finds that lawsuits from patent trolls (or, more politely, “non-practicing entities”) have surged in the United States since 1990, costing the economy a whopping half a trillion dollars all told. No, that’s not a typo.

Now, it’s possible that this is a good thing. Maybe there really are a bunch of small inventors out there getting wrongfully infringed by larger companies. In that case, these lawsuits would be a reasonable way to transfer wealth to its rightful owners. But the researchers found that this didn’t appear to be the case. They looked at 14 litigants who had caused companies to lose about $87.6 billion in value. Very little of that lost wealth — only about $7.6 billion — actually accrued to the plaintiffs. This suggests that these lawsuits aren’t actually improving the market for technology or boosting incentives to invent. It’s consistent with the view that patent trolls are just exploiting weaknesses in the legal system.

So why are patent trolls on the rise? The Boston University researchers found that about 62 percent of the lawsuits involved software patents. Patents on software and business methods tend to get litigated much more frequently because they often have, as the authors put it, “fuzzy boundaries.” A patent on a brand-new drug compound is relatively simple to delineate. But software patents aren’t always clear in scope or easy to understand. Sometimes they are patents for relatively common techniques that multiple companies might reasonable think up: Paul Marks, for instance, catches IBM moving to patent the “well-known” idea of partitioning software to protect against malware.

So these two trends appear to be connected. Patent offices are getting inundated with low-quality or overbroad patents. And those patents, in turn, are frequently being deployed in unproductive lawsuits that cost companies billions and hinder innovation. It may be true that genuine progress in science and technology is slowing down, as Tyler Cowen suggests in his book “The Great Stagnation.” But our current legal regime for patents doesn’t seem to be helping, either.