Almost nothing in the world is in every way, shape, and form, bad. (Even The Phantom Menace gave us this glorious review from “Mr. Plinkett.”)
A new NBER paper identifies the role of non-practicing entities (NPEs) or “patent trolls” in innovation and finds that, while there can be serious negative effects on downstream innovation, these effects vary based on firm and patent characteristics and the role of NPEs as middlemen can yield positive effects for some innovators:
How do non-practicing entities (“Patent Trolls”) impact innovation and technological progress? Although this question has important implications for industrial policy, little direct evidence about it exists. This paper provides new theoretical and empirical evidence to fill that gap. In the process, we inform a debate that has historically portrayed non-practicing entities (NPEs) as either “benign middlemen”, who help to reallocate IP to where it is most productive, or “stick-up artists”, who exploit the patent system to extract rents and thereby hurt innovation. We employ unprecedented access to NPE-derived patent and financial data, as well as a novel model that guides our data analysis. We find that NPEs acquire patents from small firms and those that are more litigation-prone, as well as ones that are not core to the seller’s business. When NPEs license patents, those that generate higher fees are closer to the licensee’s business and more likely to be litigated. We also find that downstream innovation drops in fields where patents have been acquired by NPEs. Finally, our numerical analysis shows that the existence of NPEs encourages upstream innovation and discourages downstream innovation. The overall impact of NPEs depends on the share of patent infringements that come from non-innovating producers. Our results provide some support for both views of NPEs and suggests that a more nuanced perspective on NPEs and additional empirical work are needed to make informed policy decisions.
The paper uses a game-theoretic model that is then tested using an extensive (though by no means complete due to the nature of patent litigation) data set to find the effects, both positive and negative of NPEs.
First, let’s examine the positives, where NPEs function as “benign middlemen.” In their capacity as middlemen, NPEs are able to purchase patents from firms with business models that do not align well with a given patent, allowing the NPE to sell or license the patent to a firm that is more likely to make use of the patent. Additionally, the sale of a patent may provide smaller firms with capital to produce further innovation. As the authors write:
Serrano (2010, 2018) finds that patent renewal and transfer is an important component of innovation. As such, there seems to be a role for the NPE in serving as a important player in this market. Indeed, Akcigit et al. (2016) find that the secondary market in patents is mostly governed by middlemen. Griffith et al. (2003, 2004) find that R&D investment also allows firms to generate “absorptive capacity,” as firms may spend more on R&D with the expectation that they can engage in technology transfer through middlemen. This absorptive capacity may also make firms more likely to be linked to networks of patent middlemen–with capabilities to move on ideas the firm itself does not create.
On the other hand, because NPEs’ business model is built on preserving the value of their assets (patents), they have strong incentives to bring infringement cases against other firms, with negative effects on downstream innovation. The results of their empirical tests:
suggest that downstream innovative activity slows down with the acquisition of a patent by NPEs. Given our study design, it seems likely that the effect of NPE acquisition on downstream citations is causal. A pure selection effect would predict an increase in forward cites among the most valuable patents after NPE acquisition…If most of the infringement is due to non-innovators, the NPEs’ role is that of protecting innovators and will tend to increase innovation. If most of the infringement comes from innovators, the NPEs could slow overall innovation, although not by a great deal according to our calibration. Our results provide preliminary evidence on these distinguishing factors in a general equilibrium setup.
Of course, all of this analysis is based in the status quo, where patent use and issuance are widespread to the point far beyond the social optimum. If the scope of our current patent system were pared back significantly, based on these findings I suspect that the role of NPEs would be more beneficial for society, as it would manage a secondary market with patents of higher quality. This would be particularly beneficial to small firms, as the paper finds that NPEs are better able to help these firms defend their innovations.
Of course, this is just speculation, and it would be a much heavier political lift than reforms that target NPEs directly, but it is worth considering how we can make a system where such institutions contribute more to society rather than going after the institutions themselves.