|Stop Bad Patents! |
A campaign by The Internet Association
The researchers first described 'a parsimonious model of an innovative economy in which NPEs endogenously arise as patent trolls', due to the combination of a number of factors (heterogeneity in innovation quality across the agents, commercialization costs, imperfections in the legal system of IP enforcement). In such model, an agent that seeks to maximize profits acts on the basis of a comparison between the expected profitability of litigation, on one side, and of commercialization, on the other. Low-type innovators, therefore, commonly choose to litigate; as their expected profits from litigation depend upon the likelihood of obtaining compensation from the alleged infringer, and the extent of such compensation, the model predicts that 'targeted firms should be those cash-rich enough to fund payoffs and those most likely to settle or lose the case for any reason (even if they have less cash)'. The model implicitly discloses the risk that some meaningful inventions may not be brought to the market, where the expected profitability of litigation is higher than that of commercialization, or, conversely, where the probability of being sued acts as a deterrent.
To validate its theoretical model, the study primarily examined data on NPE behavior from PatentFreedom (and firm-related data from several other databases). The researchers found that non-practicing entities, as expected, commonly target companies with high levels of cash balances, or which had a recent positive cash shock (steep increase in cash holdings, compared to the previous fiscal year). A one standard-deviation increase in these indicators correlates, respectively, with an 11% and a 2% increase in the likelihood of being sued by an NPE. Considering that the unconditional probability of being sued by NPEs is 2.18%, such findings indicate a five-fold increase of the risk of NPE litigation in the case of higher cash levels, and a two-fold increase in the case of positive cash shocks. Non-practicing entities also target conglomerate firms regardless of the source of the cash levels (e.g. even where the cash revenues are not derived from business segments related to the infringing patents): the study found that 'profitability in unrelated businesses is almost as predictive of NPE infringement lawsuits as is profitability in the segment related to the allegedly infringing patent'.
As predicted by the theoretical model, NPEs prefer targeting companies against which they have a higher ex ante likelihood of winning (or, conversely, that have a higher ex ante propensity to settling). Thus, litigation blossoms if the targeted company is already engaged in a number of other litigation events unrelated to intellectual property (a one standard-deviation increase corresponds to a 1.67% increase in the likelihood of NPE litigation), but is deterred by the presence of a large legal team (0.5% decrease). The validated model combines the potential proceeds with the probability of success to determine the expected profitability of NPE litigation. Thus, expected profitability positively correlates with high cash levels, small legal teams, and preexisting litigation events.
Evaluating the impact of NPE litigation, the researchers found that the companies that lost against non-practicing entities in court experienced a subsequent significant reduction in R&D expenditure and inventive activity (in terms of post-litigation patenting activity and citations to their marginal post-litigation patents), concluding that 'it really is the NPE litigation that causes this decrease in innovation' in the United States. The small amount of the damages awarded in NPE litigation that flows back to end-inventors (estimated at 5%) does not affect this conclusion, as it constitutes an insufficient incentive for innovators to carry out inventive activity under the threat of NPE litigation.
The scholars validated these findings by comparing them to results obtained from a poll of practicing entities, and concluded that they 'are not just reflections of general characteristics of IP litigation'. The study, however, could not take into consideration informal patent assertions by NPEs (usually in the form of demand letters), due to the lack of reliable data on the phenomenon. While the researchers note that 'it is widely believed that informal patent assertions has been in decline recently, and is projected to decline further', a comparison between informal and formal (in court) assertions (as well as between successful and unsuccessful formal assertions - although undeniably subject to the uncertain attribution of settlements to either category) would help further distinguish between the different patterns of behavior that emerge even within the supposedly homogeneous group of non-practicing entities.