The United States patent system was designed to encourage innovation by protecting inventors. Instead, it has become a vehicle for companies that produce nothing to extract billions from companies that produce everything. By the late 2000s, patent troll lawsuits were associated with over $80 billion in lost shareholder wealth annually for defendant firms[s], a staggering levy on American innovation that flows not to inventors but to legal operations designed to exploit weaknesses in intellectual property law.
This is not a side effect of a well-functioning system. It is the system working exactly as patent assertion entities have engineered it to work. The result is a tax on innovation, paid by companies that build products and collected by companies that build lawsuits.
The Scale of Patent Troll Costs
The numbers are difficult to comprehend. From 1990 through 2010, non-practicing entity lawsuits destroyed over half a trillion dollars in shareholder value[s]. That figure captures only the direct market impact measured through stock price declines; it excludes the indirect costs: diverted engineering resources, delayed product launches, abandoned research programs, and the chilling effect on ventures that never get funded because investors fear patent exposure.
The trajectory has been relentless. Patent litigation costs in the United States grew from $3.6 billion in 1984 to $61 billion in 2009[s]. By 2011, patent trolls alone cost defendants approximately $29 billion in direct legal expenses[s]. Set against the $247 billion that US businesses spent on research and development in 2009[s], patent assertion entities were extracting nearly 12% of the nation’s R&D budget without contributing a single product, service, or technological advance.
The problem has not improved. A 2023 industry analysis reported that patent troll cases had risen roughly 500 percent over the prior decade[s]. In the first half of 2025, non-practicing entities filed 91% of all patent cases in the high-tech sector[s]. Nine of the top ten patent plaintiffs in district court were NPEs[s]. Non-practicing entities filed 56.1% of U.S. district-court patent cases in the first half of 2025, and 91% in high-tech district-court cases[s], with nearly 64% of their cases concentrated in the Eastern and Western Districts of Texas[s], jurisdictions they favor for plaintiff-friendly procedures and juries.
How the Extraction Machine Works
The patent troll business model exploits a fundamental asymmetry. Defending against a patent lawsuit costs an average of $2.8 million[s]. For small and medium enterprises, the average defense costs $1.5 million[s]. A patent troll can acquire vague patents cheaply, send demand letters to hundreds of companies, and settle for amounts below the cost of defense. Most targets pay. The rational calculation is simple: settlement is cheaper than justice.
This targeting is strategic, not random. Research shows that a large increase in cash on a company’s balance sheet increases the probability of being sued by an NPE by 11 percent[s]. Understanding patent troll costs requires recognizing that these entities hunt balance sheets, not infringers. They also avoid companies with substantial legal departments: firms with large legal teams are less likely to be targeted[s]. The business model depends on asymmetric vulnerability. Deep pockets with thin legal resources are ideal prey.
The patents themselves are often problematic. A 2020 study found that patent trolls strategically file with patent examiners who have track records of issuing poorly crafted, vague patents[s]. The trolls are not exploiting legitimate intellectual property; they are exploiting the ambiguity of intellectual property that should never have been granted. When these patents reach court, they frequently fail. But most cases never reach court. Early settlement reduces litigation costs by 50%[s], which means both sides have strong incentives to settle before the patent’s validity is ever tested.
The Innovation Drain
Patent troll costs do not merely transfer wealth from productive companies to unproductive ones. They actively suppress innovation. A National Bureau of Economic Research study found that downstream innovation drops in fields where non-practicing entities acquire patents[s]. The 2013 White House report on patent assertion documented cases where health information technology companies ceased all innovation in technologies subject to PAE lawsuits, causing their sales to fall by one-third[s].
The mechanism is straightforward. Companies that successfully defend against patent troll lawsuits spend an average of $211 million more on research and development than companies that settle or lose[s]. The money paid to trolls comes directly from R&D budgets. Every dollar extracted is a dollar that could have funded an engineer, a prototype, or a clinical trial. The aggregate effect is a measurable reduction in American technological output.
The burden falls disproportionately on smaller firms. Large technology companies now treat patent troll costs as a predictable expense, licensing fees as routine as insurance premiums. They can afford the legal departments to fight back. Startups cannot. Small and medium businesses account for 37% of direct costs from NPE litigation[s], and for them, these costs represent existential threats rather than line items. The White House estimated that PAEs threatened over 100,000 companies with patent infringement in a single year[s]. Most of those threats targeted small businesses without the resources to fight.
The Defense: Do Trolls Help Inventors?
The strongest argument for non-practicing entities is that they provide liquidity to inventors. An individual inventor or small company that patents a breakthrough technology may lack the resources to commercialize it or enforce it against infringers. By purchasing patents and pursuing enforcement, NPEs could theoretically transfer value to these inventors, increasing incentives to innovate.
This argument deserves to be taken seriously. Markets for intellectual property are genuinely illiquid. Patent brokers and aggregators could in principle perform a valuable matching function, connecting inventors to the companies best positioned to use their technologies. Historical evidence suggests such intermediaries did provide these services in earlier eras of American industrialization.
The question is whether today’s patent assertion entities actually perform this function, or whether patent troll costs simply represent extraction without corresponding benefit.
The Evidence Says No
The empirical record is damning. Research examining the financial flows from NPE litigation found that less than 10% of the value lost by defendant companies flows to patent holders[s]. The rest is pure deadweight loss: legal fees, court costs, settlement payments to intermediaries, and destroyed business value that benefits no one. A Boston University study found that very little of the wealth destroyed by NPE lawsuits represents transfers to inventors[s]. The loss of incentives to defendant firms is not matched by any increase in incentives to other inventors.
Moreover, NPEs primarily acquire patents from small entities[s], but they target them specifically for litigation potential rather than technological value. They purchase patents with vague claims that can be construed to cover established technologies, not patents that represent genuine innovation in need of commercialization. The business model depends on ambiguity, not on facilitating technology transfer.
The theoretical case for NPEs requires them to increase returns to inventors. The evidence shows they do not. Patent troll costs are a deadweight loss on the American economy, not a transfer payment that could be justified on distributional grounds.
What Must Change
The patent system needs structural reform to address these costs. The America Invents Act of 2011 created inter partes review, allowing patent validity to be challenged at the Patent Trial and Appeal Board rather than in expensive district court litigation. This process has been critical for small businesses[s]: it provides a path to fight back against invalid patents without spending millions on courtroom defense.
Yet this avenue is now under threat. Recent proposed rule changes at the USPTO would dramatically restrict access to inter partes review. When the public comment period closed, 97% of all comments opposed the changes[s]. Major manufacturers, including Ford, Tesla, Intel, and Apple, submitted comments warning that the changes would benefit patent trolls at the expense of legitimate innovators[s]. The public understands what is at stake, even if policymakers do not always act accordingly.
Fee-shifting would also help reduce patent troll costs. Under current rules, each side typically pays its own legal costs regardless of outcome, and courts award fees only in narrowly defined exceptional cases. This allows trolls to pursue weak claims with little downside risk. Requiring losing parties to pay the winner’s legal fees would make frivolous litigation more costly and settlement less automatic. The evidence suggests that troll patents fare poorly when actually tested in court; fee-shifting would force more of them to be tested.
Most fundamentally, patent quality must improve. The Patent and Trademark Office should allocate more resources to examiners who scrutinize claims carefully, and fewer patents should issue with vague boundaries that invite opportunistic enforcement. Every ambiguous patent is a lottery ticket for future trolls.
The patent system was created to promote innovation. When patent troll costs consume a significant fraction of R&D budgets without producing any corresponding benefit to inventors or the public, that system has failed its constitutional purpose. The fix requires acknowledging what the evidence has made clear: non-practicing entities, as they currently operate, are not intermediaries facilitating technology markets. They are rent extractors imposing a tax on innovation. The question is whether American policymakers will muster the will to stop them.
