Limiting term length solves over-patenting and reduces prices.

The impact nonpracticing entities (NPEs), or “patent trolls,” have on innovation may be the most important empirical question in patent law today. So far, however, scholars have analyzed litigation brought by various types of patent owners in a fragmented and indirect fashion. Some scholars have studied only the most litigious or easily identifiable trolls.1 Such studies miss as much as 85% of NPE-asserted patents.2 Others have focused exclusively on litigation filed during a handful of years in the last decade.3 These studies fail to account for the vast differences among patents that happen to be litigated at the same time. It makes little sense, for example, to compare the first and only litigation of a one-year-old medical-device patent to the twentieth litigation of a nineteen-year-old software patent, even if both proceed contemporaneously.

Perhaps not surprisingly, these narrow studies have produced results at odds with one another. The empirical literature examining NPEs is, to put it mildly, internally inconsistent. Some studies strongly suggest that NPEs are every bit the tail that wags the dog. NPEs, for example, assert the lion’s share of “most-litigated” patents4 and are especially dominant in high-tech fields, where patents tend to be plentiful, cheap, and broad.5 Others report the exact opposite. Trolls really don’t exist at all,6 exist but are exceedingly rare,7 or exist in modest numbers but hold few of the traits attributed to them by their detractors.8

This Article fills these gaps in the existing literature by studying a broad cross-section of patents over the entire patent term. Rather than studying a subset of patents linked by litigiousness or contemporaneous court filings, I study all patent enforcement for a random sample of recently expired patents.

With this data, I can for the first time account for the relative timing of lawsuits filed by practicing and nonpracticing entities. My findings are dramatic: opposing views of NPEs in the literature ring true but at opposite ends of the patent term. Product-producing companies predominantly enforce their patents soon after they issue and complete their enforcement activities well before their patents expire. NPEs, on the other hand, begin asserting their patents relatively late in the patent term and frequently continue to litigate their patents to expiration. Indeed, I find that the average product-company patent has been shelved by its owner before the average NPE patent has even been asserted.

The degree to which NPEs dominate the final few years of the patent term is especially surprising. Though asserting just over 20% of all studied patents, NPEs account for more than two-thirds of suits and over 80% of infringement claims litigated in the final three years of the patent term. Notably, NPEs’ domination of late-term litigation is almost completely attributable to firms that do nothing more than hold patents. NPEs that many do not consider trolls—universities9 and individual inventors10 in particular—do not drive the results reported below.

I also compare the relative litigiousness of product-producing companies and NPEs, as well as differences in the subject matter and strength of their infringement claims. In addition to my overall findings, I report how these statistics change among patents litigated in the final years of the patent term. I find that NPEs are especially litigious: they overwhelmingly assert high-tech patents and lose at a relatively high rate when their infringement claims are adjudicated on the merits. The same is true for patents litigated late in the patent term: NPEs are more litigious, more high-tech focused, and more likely to lose on the merits of their infringement claims. Interestingly, an outsized percentage of product producing–company patents litigated late in the term are high-tech related. These patents are asserted by a unique group of companies, which sell a product and yet blur the line between practicing entities and trolls.

My findings add to mounting evidence that the costs of NPE litigation outweigh their benefits. In fact, they cast serious doubt on NPEs’ chief alleged benefits: that paper patentees help create a market for innovation and contribute to the dissemination of useful technology. Instead, NPEs overwhelmingly wait to assert their rights until the underlying technology is stale and unlikely to be of much use to accused infringers that independently developed the technology themselves years earlier.11 Overall, my findings suggest that Congress should shorten the patent term by three years or even longer.12 In these final years of patent protection, more than 80% of patent assertions are brought by patent-holding firms that have no intention of commercializing a product. Much of the remaining litigation is brought by product-producing companies asserting high-tech patents, often with far less than ideal motivation. At the very least, Congress or the Patent and Trademark Office (PTO) should consider increasing the frequency and magnitude of maintenance fee payments required in the latter half of the patent term.