Do Patent Trolls Exist? Two Studies Reach Different Conclusions (Part 1)

They are called many things. Patent Assertion Entities (PAEs), Non-Practicing Entities (NPEs), Patent Trolls, and on occasion, names not suitable in polite company. They often are accused of harming innovation and the economy, while providing nothing useful in return. They, less often, are said to promote innovation, in part by helping small inventors monetize their discoveries. Two recent academic studies attempt to shed light on this dispute, but their findings seem contradictory – at least at first.

Hoover Tower at Stanford University

Supporters of PAEs will be cheered by a revised working paper [pdf] released on 29 October by Professors Noel Maurer of George Washington University and Stephen Haber of Stanford University. This paper, published by Stanford’s Hoover Institution, found that large, publicly-traded PAEs invest a hefty percentage of their revenues in R&D. These PAEs, moreover, pose no threat to America’s high tech sector, costing that sector on average just 0.28 percent of its revenues.

Critics of PAEs can point to a revised working paper released on 21 June by Professors Lauren Cohen of Harvard Business School, Umit G. Gurun of University of Texas at Dallas, and Scott D. Kominers of Harvard University. This paper, published by Harvard Business School (HBS), found that PAEs, on average, behave like patent trolls. They target cash-rich firms that are often short of legal resources to defend themselves. They assert patents of lower quality than average patent litigants. They harm innovation, by significantly reducing R&D spending in many targeted firms. And they provide little money to the inventors whom they supposedly exist to protect.

While these two studies appear to contradict one another, that is largely because they examine different sets of data. Upon closer scrutiny, it is possible to not only reconcile these studies, but to see that each reveals some worthwhile information about PAEs.

Researching R&D

The Hoover Institution study examined 26 large, publicly-traded PAEs and found that these PAEs “spent nearly twice as much on R&D (as a percentage of revenues) than the weighted average of the 153 largest American high technology companies over the period 2011-16.” That sounds impressive. In reality, it isn’t.

Start with the numerator: R&D spending. These PAEs don’t make anything, so they “don’t need to spend money on other capital costs, such as marketing or manufacturing,” said Cohen, one of the authors of the HBS paper. Thus it is easy for them to spend a goodly amount of their available funds on R&D. Companies in the tech sector need to spend much of their money on making and selling their products, which squeezes their budgets for R&D.

Next, consider the denominator: revenues. The 26 publicly-traded PAEs tend to have low revenues. Most have “revenues lower than a typical Safeway supermarket,” according to the Hoover study. This contrasts with high tech firms, which often have huge revenues.

Calculating R&D as a percentage of revenues would therefore flatter publicly-listed PAEs. Although these firms spend a significant fraction of their revenues on R&D, that amounts to be a tiny amount of money – certainly as compared to the R&D spending of high tech firms. Thus, the Hoover study provides little evidence that these PAEs are a significant driver of R&D in America.

Mostly Harmless?

The other major finding of the Hoover study is that, while PAEs often are claimed to hurt the high tech sector, publicly-traded PAEs pose no threat to that sector. More specifically, the study found that these

…PAEs as a group are too small to have much effect—either positive or negative—on the US high technology sector. The total size of the transfer from the US high technology sector to the … PAEs (their revenues, plus our estimate of the litigation costs they might impose on other companies) averaged only 0.28 percent of the high technology sector’s revenues over the period 2011-16.

While interesting, these findings should be put in context. They concern only 26 publicly-traded PAEs, and these “aren’t representative of PAEs as a whole,” said Prof. Brian J. Love of Santa Clara University School of Law. These 26 PAEs account for only 10 percent of PAE litigation, stated Prof. James E. Bessen, executive director of the Technology & Policy Research Initiative at Boston University School of Law. Thus the study does not examine how the overwhelming majority of PAE suits might affect high tech firms.

The study, moreover, understates the harm caused by publicly-traded PAE suits, according to Bessen. “Litigation costs … are only a small part of the overall costs imposed on allegedly infringing firms,” he said. Most of the costs of PAE litigation include such things as: time lost by engineers and administrators, who have to deal with the litigation instead of working productively for the company; loss of business from customers leery of buying products from a firm being sued for infringement; and lower stock prices for firms being sued. The majority of litigation harms are thus not included in the Hoover Institution study, stated Bessen.

Very Interesting, but….

Finally,  by examining merely the harm that these 26 PAEs do to the high tech sector, the Hoover study fails to consider the harm done to firms in other sectors of the economy. That is a significant omission, because only a minority of those sued by PAEs are high-tech firms.

“As a whole, NPEs sue more non-tech companies, like retailers and service providers, than tech companies,” Prof. Jay P. Kesan of University of Illinois College of Law wrote in a chapter he contributed to an upcoming Intellectual Property Handbook.

The Hoover study thus provides some interesting data about a subset of PAEs, but sheds little light on the overall value or danger of PAEs. The authors of the paper “are reporting some useful facts, but it is nothing that is either comprehensive or that you could draw strong policy conclusions from,” said Bessen.

Does the HBS paper do any better? That will be examined in the next article.

Image Credits: Wikimedia