Patent damages in US courts: overview of current state of play

This is an Insight article, written by a selected partner as part of IAM's co-published content. Read more on Insight

US patent litigation trends

More patents, lots of litigation

In June 2018 the US Patent and Trademark Office (USPTO) issued its 10 millionth patent. It took seven years to grow from 6 million patents (in 1999) to 7 million (in 2006), but it took only three years to go from 9 million (in 2015) to 10 million (in 2018). Although there are many explanations for this accelerated growth, some of it is undoubtedly due to a perceived increase in the value of patents, arising from and exemplified by a few eye-catching litigation damages awards.

For many years, growth in the number of patent cases filed closely tracked the number of patents granted. However, since the peak of over 6,000 litigation cases in 2013, the number of patent cases filed and the number of patents granted have moved in different directions – with the number of patent cases filed experiencing a noticeable decline. This is perhaps due in part to many litigations being diverted or deferred following the full implementation of the America Invents Act in 2013. The act established inter partes review before the USPTO Patent Trial and Appeal Board (PTAB) as an alternative to the courts for challenging patent grants on the basis of novelty or obviousness.

Nonetheless, patent litigation activity remains far above historical levels. Since 2013 more than 8,000 inter partes review petitions have been filed in total, compared to approximately 28,000 patent cases filed in district courts. About 60% more patent cases were filed in 2017 than in 2009.

Size of damages awards

Although the number of filings has fallen in recent years, on average the size of damages awards has not. According to a 2018 PwC study, the median damages award in 2017 increased by almost 70% to $10.2 million, from $6.1 million in 2016. Although large, these amounts are not eye-popping and have not grabbed nearly as much attention as the much rarer damages awards of several hundred millions of dollars.

In 2017 cases brought by non-practising entities (NPEs) still outnumbered cases brought by practising entities, and the median damage award for NPEs continued to outstrip award sizes for practising entities. According to the PwC study, the median damages award for NPEs was $14.8 million from 2013 to 2017, compared to just $4.2 million for practising entities. NPEs also have been shown to have more success with juries than in bench trials, though their success rates are lower than for practising entities in both settings.

Jury trials accounted for about three-quarters of all patent litigation decisions over the past decade, with the median size of such awards more than five times that of bench awards. In 2016 a jury delivered the largest damages award on record – more than $2.5 billion in royalties for infringement of a method for treating hepatitis C (Idenix v Gilead). Large awards like this have caught the public’s attention.

However, the jury award in Idenix v Gilead, like many large awards before it, was overturned. Many of these rulings have focused specifically on the methods used at the lower-court level to calculate lost profits or reasonable royalties (the two primary forms of relief available in US patent infringement matters). The fact that none of the patent damages awards granted in 2017 cracked the top 10 in terms of size may have something to do with the increasingly critical eye being used by courts to evaluate patent damages.

Rise and fall of Eastern District of Texas

One of the more consequential changes seen in 2018 has been the impact of the Supreme Court’s May 2017 ruling in TC Heartland v Kraft Food Group Brands LLC. In a rare unanimous decision, the court struck a blow against ‘forum shopping’ by re-establishing restrictions on the appropriate venue for patent infringement suits. Referring to the pre-1990 standard, the court ruled that a case must be brought in the jurisdiction where the company “has a regular and established place of business”, instead of simply any place in which it has sales.

Since this ruling, there has been a noticeable shift in patent litigation filings away from the Eastern District of Texas, which had gained a reputation for being a favourable venue for patent owners and their damages claims. A year after the decision, the number of cases filed in the Eastern District of Texas had fallen by two-thirds, from more than 1,600 to 525. Delaware, where many businesses are incorporated, appears to have picked up the largest percentage of those filings, increasing by two-thirds (550 to 907) during the same period. Northern California (the location for many patent-heavy Silicon Valley companies) has also seen a substantial increase, with filings almost doubling to just under 300. In recent years, the rate of success for plaintiffs in patent infringement cases in the Eastern District of Texas has been lower than that in both Delaware and California’s Northern District.

Growth of Daubert

Daubert challenges have become increasingly common in patent cases, allowing for possible exclusion prior to trial, in whole or in part, of an expert’s method of calculating patent damages. According to another 2018 PwC report, from 2000 to 2017, IP cases had the second largest number of Daubert challenges (behind breach of contract and fiduciary duty cases) among all types of litigation. This study shows that Daubert challenges in patent damages cases continue to rise. The success rates are staggering – roughly one out of every two Daubert challenges leads to a partial or entire exclusion of patent damages testimony.

The combination of the large number of patent cases being filed and the increasing complexity of the technologies and products underlying patent claims may be at least partially responsible for the growth of Daubert challenges to patent damages analyses in recent years. Moreover, it has been suggested that the signals from the various courts that hear patent damages cases are not always clear and consistent, leading to underlying uncertainty as to what is acceptable methodology. And uncertainty often leads to challenge.

US patent litigation hot topics

Calculating patent damages has always been difficult and uncertain. Four issues in particular have been at the heart of recent uncertainty.

Non-US damages

In a seven-to-two decision, the Supreme Court overturned the Federal Circuit (the single court that hears all appeals of district court patent cases) decision in WesternGeco LLC v Ion Geophysical Corp. The court ruled that in some instances patent owners may be allowed to claim lost profits for third-party sales taking place outside the United States.

WesternGeco involved US manufactured components that were exported and used in infringing products assembled outside the United States. Even though the lost profits that WesternGeco sought were generated overseas by use of the infringing products, the Supreme Court determined that the legal injury was the result of domestic activity, and so the “damages adequate to compensate for infringement” allowed in Section 284 extended, in this case, to lost profits from overseas sales.

Although technically the decision is limited to claims of lost profits, the court’s reasoning has led to speculation that the same arguments may apply to reasonable royalty damages. If so, the consequences may extend far beyond WesternGeco. Because the number of cases involving reasonable royalty damages far outnumbers lost profits cases, this could have a significant impact on the future of patent damages litigation.

Entire market value rule and smallest saleable patent-practising unit

Apportionment is at the heart of virtually every patent damages case. The issue arises in at least two forms. The first is apportioning the value of goods or services to the value attributable to patented technologies that are embodied in the goods or services. The second is apportioning value among patented technologies to the individual patents that are at issue in a particular case.

A decade ago, in the Cornell case, the court “reaffirmed that in any case involving multi-component products, patentees may not calculate damages based on sales of the entire product, as opposed to the smallest salable patent-practicing unit, without showing that the demand for the entire product is attributable to the patented feature”.

However, it appears that the parameters for deciding when to use the entire market value rule (EMVR) and when to use the smallest saleable patent-practising unit (SSPPU) are still being defined on a case-by-case basis.

Some case law has established that the threshold for applying the entire market value rule is reasonably high. According to some rulings, a royalty base for a multi-component product – that is, a product with many features and many patented components – can be based on the product’s sales price only if the specific patented feature in question can be proven to be the factor that drives demand. If instead the value of the product is driven by an aggregation of features and capabilities, the royalty base should reflect only the portion of the product’s total value traceable to the infringing component.

According to some observers, two cases decided early in 2018 – Finjan and Exmark – reflect the existing uncertainties on how and when to use the EMVR versus the SSPPU.

In Finjan the Federal Circuit rejected the jury award despite the fact that the reasonable royalty damages determination was based on the SSPPU. In overturning the award, the court indicated that, because the patent-practising component itself included patented and non-patented technologies, further apportionment was required. In Finjan some observers have suggested that the court seemed to move in the direction of a royalty base even narrower than that suggested by the SSPPU.

However, several days later the Federal Circuit upheld a plaintiff’s use of a broad royalty base. In Exmark the jury awarded reasonable royalty damages based on a percentage of sales revenues associated with the entire value of the infringing lawn mowers, in spite of the fact that the patented technology related to a single component. In rejecting the infringer’s argument on appeal that the royalty base should have been limited to the value of the flow control baffle component (ie, the SSPPU), the Federal Circuit found that “apportionment in this case can be done through a thorough and reliable analysis to apportion the rate”. The Federal Circuit further found that the use of the entire end-product price was particularly appropriate because the asserted claim was “directed to the lawn mower as a whole”. Moreover, according to the court, industry licensing practices supported using the value of the entire lawn mower as a royalty base. Ultimately, however, the court rejected the royalty rate (as opposed to the royalty base) calculated by the expert, saying that the expert failed to tie the royalty rate to the facts of the case.

While choosing the right royalty base can be difficult, it is extremely important. In the Federal Circuit’s review of Power Integrations, the original $140 million patent damages award was overturned in July 2018 because the award was based on the entire value of the infringing power supply controllers. The court found that the infringing products had other valuable features that affected consumer demand and that Power Integrations did not meet its burden to show that the patented feature was the “sole driver of consumer demand, i.e., that it alone motivated consumers to buy the accused products”. Accordingly, it ruled that the evidence was insufficient to invoke the EMVR.


Reasonable royalty damages are often based, at least in part, on reference to existing licences in the marketplace. Comparable licences can indeed be useful in analysing reasonable royalty damages, but assessing comparability can be a challenge. Comparability is often evaluated from both a technical and an economic standpoint. Is the underlying technology similar to that which is at issue? Are the economic circumstances surrounding existing licences similar to those at issue?

For years, many Daubert challenges have centred on issues of comparability. Some observers have argued that there is still a lack of clarity in the courts’ guidance regarding what constitutes a sufficiently comparable licence. For example, in ResQNet the Federal Circuit determined that “the most reliable license in this record arose out of litigation”. In LaserDynamics that same court rejected the use of settlement licences, finding that “this case is therefore well outside the limited scope of circumstances under which we deemed the settlement agreement in ResQNet admissible and probative”.

In the TCL/Ericsson case the litigants and the court took extraordinary care in determining whether industry licences were sufficiently comparable to the case at hand. Assessing whether the licensing parties were similarly situated to the litigating parties entailed an extensive fact-based examination, focusing on such things as the size and nature of the licensee’s business, as well as the timing of the licence. Even after such careful examination, the opposing experts reached significantly different conclusions as to interpretation of important licence terms and about which licences were comparable to the case at hand. In the end, the court considered the work of each expert, but modified each approach, adopting its own set of comparable licences. Even a detailed and careful analysis does not guarantee that a licence will be accepted as an appropriate comparable by a court.


Consumer surveys have become increasingly important in establishing the portion of value attributable to an infringing feature or technology. The surveys that are being used generally fall into one of two types:

  • usage surveys; and

  • demand or value-based surveys.

Usage surveys provide information about how often certain features are used relative to other features. Demand or value-based surveys (eg, conjoint surveys) provide information about how much consumers are willing to pay for different features and whether their purchase decision is based on the presence or absence of the patented feature. These surveys are often direct inputs into an apportionment analysis.

As with patent damages in general, surveys are difficult to do, and many are subject to Daubert challenges. According to the district court in Unwired Planet, an inadmissible survey is one that does not properly define the patented feature and does not target the patented invention. Surveys are an important tool, but they must be implemented with care.

Patent litigation has become big business in the United States. Fortunately, courts have constrained unreasonable growth. However, the bounds of those constraints remain uncertain as legal strategies and damages methodologies adapt to evolving product and marketplace realities, as well as to evolving interpretations and applications of patent law.

Analysis Group Logo

Analysis Group

800 17th Street NW, Suite 400

Washington DC 20006

United States

Tel+1 202 785 6300

Fax+1 202 530 0436


Unlock unlimited access to all IAM content