The issue: The rule of law is failing when it comes to patents.
What’s wrong? Patent laws promote lengthy, expensive litigation.
The law decides. It doesn’t encourage settlement. It should be so oriented for patents. It’s efficiency. It’s common sense.
Last month, the author explained why settlement and
license agreements are more important to a well-functioning patent
system than decisions of the patent office and the courts, and
summarized five reasons why agreements are more efficient. (John
Schlicher, “Improving the Patent System by Removing Legal Obstacles to
Agreements,” 31 Alternatives 1 (January 2013).) In this
concluding Part 2, the author discusses some of the specific obstacles
the law places in the path of patent agreements.
* * *
If all patent disputes had to be resolved by litigation
to the bitter end, the patent system could not function and would be
abolished. The patent system makes a positive contribution, and
survives, only because business people avoid or resolve privately an
extraordinarily high percentage of potential and actual disputes.
For that reason, we would expect the law to develop at
every opportunity to facilitate and encourage patent agreements. At the
most abstract level, the law says the policy in favor of settlement of
litigation applies to patents. The law, however, developed in ways that
discourage private resolution of patent disputes, particularly by the
most efficient form of settlement, a license prior to a lawsuit.
To secure the benefits of settlement and license
agreements, the law should facilitate the discussions needed to
negotiate agreements, provide sensible incentives for patent owners and
product suppliers to reach agreements, enforce these agreements, and
limit enforcement costs and risks. These are some of the ways the law
LICENSE ROYALTIES AND THE VALIDITY OF THE PATENTS.
The law should encourage patent owners and producers to enter
settlement and license agreements, and permit all such agreements to be
true alternatives to litigation. Instead, the law sometimes forces
people to litigate patent issues rather than resolve them by agreements.
In almost every patent dispute, the potential infringer
contends the patent is invalid. Before the late 1960s, patent owners and
their licensees could resolve a patent validity dispute with a license
requiring payment of royalties without regard to validity, and requiring
the licensee not to commence an action to have the patent declared
Under those terms, the parties could agree to payments
discounted to reflect their views of the likelihood that the patent
would be found valid if litigated. They could also take into account the
lower costs each would bear operating under the agreement rather than
litigating, and the lesser risk each would bear because future payments
would not depend on validity issues.
This changed shortly after the U.S. Supreme Court’s decision in Lear v. Adkins,
395 U.S. 653 (1969). See “Judicial Regulation of Patent Licensing,
Litigation and Settlement under Judicial Policies Created in Lear v. Adkins” (1985)(available at http://bit.ly/VXRgG4). After Lear and lower court decisions extending Lear,
if a patent owner grants and a potential infringer accepts a license to
avoid litigation, the law insists that the license not prevent
litigation of validity.
Properly understood, the Lear decision merely
eliminated a contract rule on royalty obligations called licensee
estoppel. Where a license did not provide for the effect of validity on
payments (as in the Lear license), the estoppel rule was that
payments must be made without regard to validity. The law sensibly
assumed that is what the parties intended.
Lear eliminated that default rule. The effect was
that, where the agreement was silent, payments were dependent on
validity and the licensee could defend a contract action for royalties
based on invalidity. Lear did not say or necessarily imply that
patent owners and licensees could not lawfully provide for a different
result. The lower courts, however, read Lear to mean that license
agreements to make royalties independent of validity or to require that
the licensee not challenge validity were unenforceable.
This interpretation prevents patent owners and potential
infringers from eliminating the risk and cost of validity litigation
and agreeing to a license with payments at discounted rates based on
their views of the likely outcome of validity litigation. The law
prevents them from achieving several of the benefits of agreements
discussed in Part 1 last month.
Curiously, the law insists on that inefficiency only if
they enter a license before litigation begins. The day after an action
is filed, the law permits these agreements. In short, the law disfavors
licensing when the benefits are greatest—that is before litigation is
possible or before litigation begins—and favors licensing only when the
benefits are smaller, after litigation begins.
The result is more litigation.
DECLARATORY JUDGMENT ACTIONS BY POTENTIAL INFRINGERS OR LICENSEES.
The ability of a patent owner and potential infringer or licensee to
discuss a patent dispute and resolve it without litigation also depends
when the law permits the potential infringer or licensee to commence a
declaratory judgment action to have the patent declared invalid or not
infringed. Until the middle 1960s, patent owners and suppliers could
discuss possible agreements without creating the conditions for the
suppliers to commence declaratory judgment actions.
Today, declaratory judgment jurisdiction rules permit
settlement and licensing discussions to be used as the basis for a
supplier to commence a declaratory judgment action. If the patent owner
says during those discussions that the supplier is infringing, the
supplier may start litigation, even if the patent owner has not
threatened to sue.
Because it nearly is impossible for a patent owner to
negotiate without in some way indicating that it believes an agreement
is necessary, this rule poses a major obstacle to holding these
discussions. This leads to fewer discussions and, with fewer such
discussions, fewer agreements and more unnecessary litigation.
DECLARATORY JUDGMENT ACTIONS BY EXISTING LICENSEES.
The law also attempts to define the circumstances in which an existing
licensee may commence an action to have the patent declared invalid.
As discussed, Lear has been interpreted to mean that license royalties must depend on validity. Lear,
however, does not necessarily mean that patent owners lose all control
over how much validity litigation will occur. The patent owner and
licensee are entirely free to agree that the owner may terminate the
license if the licensee stops paying. With termination rights, a
licensee that stops paying and continues selling products may face an
infringement action and post-termination damages not necessarily limited
to the amount of pre-termination royalty rates.
Under those conditions, many licensees will continue to
pay. Validity litigation, with its risks and costs, may be rare. This
changes if the jurisdiction rules permit a licensee to start an action
to have the patent declared invalid without non-payment that jeopardizes
the continuation of the license. Before the early 1970s, the law seemed
clear that the federal courts had no jurisdiction to decide an action
begun by a company that was operating under a license and paying the
Unfortunately, the Supreme Court’s MedImmune Inc. v. Genentech Inc.,
127 U.S. 764 (2007) decision may be read to say that any patent
licensee may commence an action to have a licensed patent declared
invalid after simply objecting to further payments on the basis of
invalidity, and without stopping payment. See “Patent Licensing, What to
Do After MedImmune v. Genentech” (2007) (available at http://bit.ly/SoPC4v).
While this author believes MedImmune is more
limited and applies only where the license expressly makes the royalty
obligation conditional on validity, others may disagree. Under the
broader view, any licensee may force litigation to eliminate its payment
obligation without jeopardizing the continuation of the license. This
jurisdiction rule removes the disincentive to litigation provided by
Combined with Lear, this rule poses a great
obstacle to providing royalties at discounted rates and rates based on
the assumption there will be no litigation costs or risks. A patent
owner would be foolish to agree to the lower rates when even the lower
rates may not be paid and it may be forced to litigate. Again, this
poses an unnecessary barrier to licensing.
LICENSING MAY LIMIT DAMAGES. The law measures
patent damages in several ways that discourage agreements. See “Patent
Damages, the Patent Reform Act and Better Alternatives for the Courts
and Congress,” (2009) (available at http://bit.ly/UViCBQ). Only one is discussed here.
The law permits royalty terms of existing licenses to be
an important and occasionally decisive factor in determining damages
against other infringers. If a patent owner grants licenses to avoid
litigation, the law insists that the licenses may be used by infringers
to limit their damages. Since the 1850s, a so-called established royalty
has been an available measure of damages. In general, an established
royalty exists when a patent owner has granted a number of licenses at a
uniform rate for the same activities that constituted the infringement.
In addition, since the early 1900s, damages also may be measured by a
so-called reasonable royalty, and other licenses granted by the patent
owner may be considered in arriving at a reasonable amount.
Patent damages should be measured by the difference
between (1) the value an invention would have had to the owner, if there
had been no infringement by the defendant in a particular case or
anyone else, there were no uncertain patent issues and no litigation,
and the invention was used in the most productive and profitable manner
(the “full economic value of the invention”), and (2) the value the
invention had to the owner with infringement. Limiting damages based on
royalties in past licenses is a problem because those royalties are
highly likely to be less than the full economic value of some invention.
When the courts began measuring damages by an
established royalty, they did so by observing that market transactions
are the best measure of the value of things. Market transactions in
patents, however, will almost invariably undervalue inventions for
purposes of damages.
For a moment, put aside the possibility of infringement
and litigation. As with any agreement, a patent owner and potential
licensee will enter a license if each perceives that the profits it will
earn under a license exceed the profits that it would earn without the
license. Those profit levels depend on the economic value of the
invention, when used by the licensee, the patent owner, or some other
potential licensee. Agreed royalties, however, are unlikely to be the
same as the full economic value of an invention.
One disconnect between license royalties and damages is
that damages should be based on an invention’s actual value during the
period of past infringement and license royalties are set not based on
past value. They are based on expected future value. License royalties
depend on the future economic value of an invention the patent owner
expects to capture without licensing, and the future value the potential
licensee expects to capture with licensing.
It is difficult for a patent owner and potential
licensee to predict future economic value due to uncertainty about
future economic conditions and technological changes. Their estimates of
an invention’s future value are highly likely to be different from the
value the invention proves to have.
In addition, where a patent owner or licensee or both
are averse to risk, they will perceive the risk-discounted value of the
invention to be less than its expected value and will license for
royalties measured by the lower risk-discounted value. For that reason,
the agreed payment is likely to be lower than the value an invention
proves to have.
More important, royalties do not merely measure the
economic value of the invention to a patent owner or a licensee.
Litigation provides an option unavailable in almost all other commercial
A potential licensee often has the option to obtain a
license or risk infringement litigation, and a patent owner has the
option to license or engage in infringement litigation. If the owner
believes a potential licensee will use the invention without a license,
the patent owner will license only if the expected royalties are greater
than (1) the economic value of the invention to it without licensing
and (2) the risk-discounted expected value of an infringement action.
Likewise, a potential licensee will accept a license only if expected
royalties are less than (1) the economic value of the invention to it
with licensing, and (2) the risk-discounted expected cost of an
The patent owner’s expected value and in-fringer’s
expected cost of a patent infringement action depend on their
perceptions of the likely outcome and remedies of litigation. Just as
future market conditions are uncertain, the outcome of future patent
litigation is uncertain.
The expected value and cost of infringement litigation
are based on the parties’ perceptions of probabilities of winning and
losing. When royalties are agreed to in some amount greater than the
patent owner’s expected value of litigation and less than the
infringer’s expected cost of litigation, those royalties are likely to
be less than the actual value of an invention, because the probabilities
are usually significantly less than one.
For example, assume there is an invention that would
permit a potential licensee to earn an additional $30 per unit profits.
If the potential licensee proceeds without a license and is sued for
infringement, assume there is a 50% chance that the patent owner will
prevail. At first approximation, the potential licensee will pay no more
than $15 per unit for a license (that is, $30 per unit times its
probability of losing, 0.5).
If a risk-neutral patent owner and potential licensees
generally believed that a particular patent owner had about a 50% chance
of winning the infringement action, and we observe established
royalties for licenses of $15 per unit, the value of the invention is
more likely to have been around $30 per unit (that is, $15 per unit
divided by 0.5).
Third, a royalty is deemed “established” only where
licenses have been granted to a number of companies, and the royalty
provisions of those licenses are the same.
If a patent owner licenses a number of companies with
the same royalty terms, those terms are likely to underestimate the
invention’s value to a particular licensee where the invention
contributes different amounts to the expected profits of different
If different licensees value the use of the invention in
different amounts, a uniform royalty amount for each of them will
understate the invention’s value to some of them. This is because the
royalty terms must be acceptable to the marginal licensee having the
lesser value for the invention. Since the profits different companies
make as the result of using an invention are likely to differ, an
established royalty satisfying those two legal requirements is likely to
understate the invention’s value to some companies.
For these and other reasons, actual royalties are likely
to be lower than the values of inventions. Hence, damages derived from
rates in actual licenses create another unnecessary barrier to
licensing. A patent owner considering granting licenses to one or a few
licensees limits damages against infringers. For that reason, patent
owners will be less willing to grant licenses. If the patent owner
litigates against later infringers and wins, the damages may be limited
by the lower amounts given to licensees without litigation.
A patent owner may have the same concerns if it grants a
license to even one licensee. Suppose a patent owner believes that the
amount specified in that license will limit the amount of damages
available against that licensee, if the license terminates and the
former licensee infringes. If the owner agrees to a lower rate dictated
by expected value and cost of litigation, this rate may limit
infringement damages even after a court has found the patent valid and
infringed. The lower rate may apply to a time period for which that rate
Again, the result is less licensing and more litigation than there should be.
UNCERTAINTY ABOUT INJUNCTIVE RELIEF AND THE EFFECT OF LICENSING ON THE AVAILABILITY OF AN INJUNCTION. Injunctive relief is vital to provide proper incentives for patent settlement and license agreements. The Supreme Court’s 2006 eBay decision has caused unnecessary confusion about how judges decide whether to issue permanent injunctions. eBay Inc. v. MercExchange LLC, 547 U.S. 388 (2006)(available at http://bit.ly/UdQnuV). See Comments for FTC Hearings on the Intellectual Property Marketplace (2009)(available at http://bit.ly/UViRg5).
As noted in Part 1 last month, one reason agreements
produce better results than litigation is that settlement with a license
for the future negotiated by the parties free of judicial compulsion
produces greater benefits than any of court orders. Consider the
If the court grants an injunction against future
infringement, either an infringer will not use the invention in the
future or there will be an agreement in which the patent owner grants
the infringer a license. After an injunction, a patent owner will grant a
license only where future use by an infringer will generate larger
profits than available under any alternative state of affairs, such as
use only by the patent owner or some other licensee. The parties will
license when a license increases the value of a patented invention. An
injunction alone does not produce this beneficial result.
If the courts deny an injunction, the incentives to
license are lower, profits generated from future use of an invention are
likely to be lower, and the invention’s value is likely to be lower.
With no injunction, there are two basic outcomes.
First, the infringer may continue to sell the infringing
product, and a series of lawsuits follows in which the patent owner
captures some of the value of its invention as damage awards. The
infringer’s continued sales, unrestrained by any payment obligation,
continue to depress the value of that invention the patent owner may
capture by licensing others or using the invention itself.
The portion of the remnant of the invention’s value
ultimately captured by the patent owner will depend on patent damages
standards and in all events will be reduced by the added litigation
costs. Even in this situation, the parties may agree to license. The
license payments, however, will no longer be based on the full economic
value of the invention when used by the licensee. They will depend on
expected damages awards and litigation costs.
Second, the court orders the parties to attempt to enter
a license on agreed terms or on terms the court orders if they fail to
agree. In these actions, federal district judges become the government
regulators of patent licenses. Judges decide which companies obtain
licenses, the prices for licenses, and perhaps even other non-price
For many reasons, private negotiations between patent
owners and potential licensees will produce much better outcomes than
orders by judicial regulators. Again, the parties may agree to a license
to avoid having to operate under whatever license the court might order
them to enter. But again, the payments will no longer be based on the
full economic value of the invention. They will depend on whatever the
parties think the court might order.
Uncertainty about when an injunction will be granted and
the role of the court if an injunction is denied has a profound affect
on agreements because the parties do not know what will happen if they
litigate. The patent system is designed so that decisions by a patent
owner and users of inventions, and not judges, govern by whom and how
inventions are used and at what prices. Subject to one special
situation, only injunctions provide those people with the right
incentives to make those decisions.
The eBay decision has a second effect on
agreements. While the Court expressly declined to decide the effect of a
patent owner’s willingness to license on the availability of an
injunction, the decision permits a patent owner’s expressed willingness
and incentives to grant an infringer a license for the future to be
considered as a factor pointing to denial of an injunction.
Suppose a patent owner would be willing to license an
infringing company without litigation, if an injunction was certain to
issue in litigation against that infringer absent a license. If an
injunction is certain to issue and the patent owner and infringer
believe that patent owner is virtually certain to win, the value of
litigation to the owner and its cost to an infringer for the period
after judgment are equal to the full value of the invention.
If successful litigation will give the owner an
injunction, the owner will license only if the value the owner captures
from the license is greater than the value the owner may capture by
exploiting the invention in any alternative way, such as using it or
licensing someone else. In that situation, the infringer will be willing
to license for payments slightly less than the full value of the
If successful litigation may not result in an
injunction, the value of litigation to the owner and its cost to the
infringer are lower, and the payments from licensing and therefore the
incentives to license are lower. Again, the consequence is more
litigation and fewer agreements.
Consider the simple situation of a patent owner with one
potential licensee. If the patent owner expresses a willingness to
license that company, this reduces the owner’s ability to obtain an
injunction if the settlement or licensing discussions fail. Without an
injunction, the value of the invention the owner may obtain by licensing
is lower. The natural response for the owner is not to express a
willingness to license, because this increases the likelihood that no
injunction will issue after litigation. The unfortunate consequence is
again fewer licenses and more litigation.
Subject to a special class of situations mentioned
below, the law should be that a patent owner’s willingness to license
some infringer is not a reason to deny an injunction. It is the reason
to grant an injunction. Judges should not be regulators of patent
licensing. The choice between private control of licensing and judicial
control should be easy. As U.S. Circuit Judge Frank Easterbrook observed
in Matter of Mahurkar Double Lumen Litigation, 831 F. Supp. 1354, 1396-97 (N.D.Ill. 1993),
A patent conveys the right to exclude others from
making, using, or selling the invention, and this right implies the
propriety of an injunction enforcing exclusivity. The injunction creates
a property right and leads to negotiations between the parties. A
private outcome of these negotiations—whether they end in a license at a
particular royalty or in the exclusion of an infringer from the
market—is much preferable to a judicial guesstimate about what a royalty
should be. The actual market beats judicial attempts to mimic the
market every time, making injunctions the normal and preferred remedy.
See Schlicher, Patent Law: Legal and Economic Principles §§ 1.14,
The only difficult issue is whether an injunction should
be denied where it is clear that it is in the patent owners interest to
grant the infringer a license and in the infringer’s interest to
operate under a license, and there is good reason to believe an
injunction would distort the amount of the royalty upward in an
undesirable way. The potential distortion is that an injunction would
permit a patent owner to negotiate a higher royalty simply due to an
infringer’s large past invention-specific investments, and this prospect
would adversely affect the incentives of patent owners and producers in
The laws that pose unnecessary barriers to resolving
patent disputes should be changed. Changing these laws would encourage
agreements and avoid litigation and administrative procedures, costs and
risks. The courts devised every law mentioned in this article. None
were required by the Patent Act or any other act of Congress. Therefore,
the courts have the power to correct these and, indeed, most other
Judicial change, however, comes slowly and with
considerable institutional resistance. Unfortunately, legislation is
probably needed to achieve many of them. Also unfortunately, the America
Invents Act, P. L. 112–29, 125 Stat. 284-341 (2011), as noted at the
outset of Part 1, tries to improve the patent system only by more
government rather than a better way, more agreements.
The author is a Lafayette, Calif., attorney who works
with companies involved in patent disputes. This article is based on
Chapter 10 of his book, “Settlement of Patent Litigation and Disputes:
Improving Decisions and Agreements to Settle and License” (American Bar
Association 2011). The book describes tools and data to help people make
better, quicker, and cheaper settlement and licensing decisions.
©John Wiley & Sons, Professional and Trade Subscription Content
989 Market Street, 5th Floor, San Francisco, CA 94103 | Phone: 888.378.2537 | Fax: 888.481.2665