Berkeley Law

Berkeley Law Scholarship Repository

Berkeley Law Books

2014

The Falcon's Gyre: Legal Foundations of Economic

Innovation and Growth
Robert Cooter
Berkeley Law

Follow this and additional works at: http://scholarship.law.berkeley.edu/books

Part of the Growth and Development Commons, and the Law Commons

Recommended Citation
Cooter, Robert, "The Falcon's Gyre: Legal Foundations of Economic Innovation and Growth" (2014). Berkeley Law Books. Book 1.
http://scholarship.law.berkeley.edu/books/1

This Book is brought to you for free and open access by Berkeley Law Scholarship Repository. It has been accepted for inclusion in Berkeley Law Books
by an authorized administrator of Berkeley Law Scholarship Repository. For more information, please contact [email protected].
 The Falcon’s Gyre
Legal Foundations of Economic Innovation and Growth
Version 1.4
 The Falcon’s Gyre
Legal Foundations of Economic Innovation and Growth
Robert Cooter
with Aaron Edlin

Version 1.4 posted August 2014

Version 1.4, posted: August 2014

Copyright © 2014 by Robert Cooter

Creative Commons Attribution-ShareAlike 4.0 International License:

http://creativecommons.org/licenses/by-sa/4.0/legalcode

Version 1.0, originally posted: June 2013

Robert Cooter ISNI 0000 0001 1867 4300

http://works.bepress.com/robert_cooter/

This book available electronically through the Berkeley Law Scholarship Repository
http://scholarship.law.berkeley.edu/books/1/
 Contents
viii Acknowledgements
ix Preface
Chapter 1. Overtaking
1.6 Innovation
1.9 Depletion
1.11 Freedom
1.13 Welfare
1.16 Conclusion
Chapter 2. Ventures
2.2 Life-Cycle of a Business Venture
2.4 Imitators or Enthusiasts?
2.8 Rate of Innovation
2.9 Conclusion
2.11 Appendix
Chapter 3. The Double Trust Dilemma
3.5 Relationships
3.7 Private Law
3.9 Public Law
3.11 Three Stages of Finance in Silicon Valley
3.13 Small and Large Firms
3.15 Ontogeny Recapitulates Philogeny
3.17 Conclusion
Chapter 4. Separation
4.2 Paradox of Growth
4.4 Producer’s Market Power
4.6 Innovator’s Market Power
4.8 Conclusion
Chapter 5. Fertility
5.5 Applying the Fertility Principle
5.7 Bargaining and Transaction Costs
5.10 Conclusion
5.11 Appendix: Welfare Triangles
Chapter 6. Separation and Fertility in Intellectual Property Law
6.2 Strength
6.5 How Strong?
6.7 Applications
6.8 Reach: Invention or Discovery
6.10 Breadth
6.13 Duration
6.16 Optimal Remedy
6.20 Thickets and Holdup
6.23 Middlemen and Trolls
6.24 Conclusion
 Acknowledgements
For the best feast, the host who supplies the main course should ask each
guest to contribute a dish. I have supplied the main course to the table of
my scholarship and many others have contributed their ideas. Most im-
portant, Aaron Edlin provided fundamental ideas and criticism, especially
to the initial chapters of this book where he is credited. I would also like
to thank Blair Dean Cooter, Robert Litan, Brian Broughman, Merritt Fox,
Robert Atiah, George Geis, Steve Maurer, Patrick Goold, and Tal Niv. Tech-
nical support was proved by the Berkeley Law School library and the Berke-
ley Electronic Press. Financial support and encouragement were provided
by The Kauffman Foundation, The Humboldt Prize, the Mercatus Center
and the Law and Economics Program at George Mason University, and the
Berkeley Law School
 Preface
After years of parish work, a Catholic priest sat up in bed one morning and
thought, “Maybe the Pope is wrong and Buddha is right.” After years of
teaching and writing about economic efficiency and the law, I sat up in bed
one morning and thought, “Maybe efficiency is wrong and innovation is
right.” The effects on human welfare from inventing the tractor far exceed
the effects from more efficient allocation of horses. Innovation causes com-
pound growth that swamps static inefficiency like a tsunami swamps a scow.
These facts compelled me to rethink previous work on law and economics,
including my own.

The state provides infrastructure on which the economy runs. The materi-
al infrastructure includes roads and the institutional infrastructure includes
laws. Smooth roads and good laws sustain the economy. Law and efficien-
cy economics explains how law improves resource allocations. Its practical
usefulness, systematic consistency, and even its intellectual beauty, have in-
fluenced America legal scholarship. By contrast, law and growth economics
must explain how law increases economic innovation. Its potential useful-
ness potentially exceeds law and static economics, but, in its current state of
underdevelopment, it lacks systematic consistency or intellectual beauty.

Law and growth economics is underdeveloped partly because growth eco-

nomics is underdeveloped. Growth economics explains innovation’s effects,
but not its causes. Mystery shrouds the causes of innovation because inno-
vation is intensively legal and growth economics is not. Economics exports
ideas to law and imports little, like contemporary China exports commod-
ities to the USA imports little. Importing some legal ideas into economics,
thus balancing trade in ideas, can dispel some of the mystery of growth.

Economic innovation usually requires combining new ideas and capital.

They naturally repel each other because the investor distrusts the innovator
with her money, and the innovator distrusts the investor with his ideas. The
“double trust dilemma” refers to the problem of inducing the innovator to
trust the investor with his ideas, and also inducing the investor to trust the
innovator with her money. The solution to this problem depends on law,
especially the law of property, contracts, corporations, finance, and bank-
ruptcy. This book dispels some of the mystery of growth by explaining law’s
contribution to solving the double trust dilemma.
Will law and growth economics fulfill its promise? Two shoe companies
sent representatives to Hawaii in the 19th century. One of them telegraphed
the message to the home office, “No opportunity. No one wears shoes.”
The other telegraphed the message,” Great opportunity. No one wears shoes
yet.” Instead of shoes, this book telegraphs the latter message about law and
growth economics.
 Robert Cooter

Chapter 1. Overtaking1
Robert Cooter and Aaron Edlin

Robert Lucas, the economist who won the Nobel prize in 1995, famously
commented, “Once one starts to think about economic growth, it is hard to
think about anything else.” Compared to sustained growth, other sources of
national wealth are insignificant. Compounded over a century, 2% annual
growth (roughly the growth rate of the US economy in the last century) in-
creases wealth by more than 7 times, and 10% annual growth (roughly the
growth rate of the Chinese economy from 1980 to 2010) increases wealth
by almost 14,000 times.2 People systematically underestimate the effects of
compound growth, like thinking that a 5% increase in height is the same for
a teenager and a toddler.

Sustained growth causes wealth in one country to overtake another faster

than most people can imagine. Assume that country x’s national income
begins twice as large as country y’s, but the former stagnates and the latter
grows. To overtake x’s national income, y will need 36 years if it grows at
2%, 15 years if it grows at 5%, and 8 years if it grows at 10%. To illustrate
concretely, most people are surprised to learn that South Korea, Mexico, and
Senegal had roughly equal income per capita in 1950, yet in 2008 South
Korea was more than twice as high as Mexico and ten times higher than Sen-
egal. Similarly, in 1870 Argentina’s per capita income was 33 percent higher

1 This chapter draws on Robert Cooter and Aaron Edlin, “Overtaking,” in The American Illness:
Essays on the Rule of Law, ed. Frank Buckley (Yale University Press, forthcoming), and “Maximizing
Growth vs. Static Efficiency or Redistribution,” working paper, Berkeley Economics Department.
2 Here’s a table of size reached by an economy that starts at 1 and grows at various rates and years.*

Growth
1 year 5 years 10 years 25 years 50 years 100 years
rate
0% 1.00 1.00 1.00 1.00 1.00 1.00
0.5% 1.005 1.01 1.05 1.13 1.28 1.65
1% 1.01 1.05 1.10 1.25 1.64 2.70
2% 1.02 1.10 1.22 1.64 2.69 7.24
5% 1.05 1.28 1.63 3.39 11.47 131.50
10% 1.10 1.61 2.59 10.83 117.39 13,780.61

*size of economy = (1+r/100)t, where r=percentage growth rate, and t=years of growth

1.1
Ver. 1.4 - Chapter 1: Overtaking

than Sweden’s, yet by 2004 Argentina’s per capita income had dropped to 43
percent of Sweden’s.

Angus Maddison estimated changes in wealth over millennia for regions in

the world. He concluded that Egypt was the world’s richest region two thou-
sand years ago with per capita income 50 percent higher than elsewhere in
the Roman Empire, China, or India. In the year 1000, Iran and Iraq under
the Abbasids were the economically most advanced countries with a per cap-
ita income 50 percent higher than in Europe or Asia. By 1500 Italy had the
lead with a per capita income 50 percent higher than in the rest of Western
Europe, double that of Asia, and three times that of Africa. In 1820 Western
Europe and the United States had the highest income per capita, twice as
much as in eastern Europe, Latin America, or Asia, and three times as much
as Africa. In the 19th century the rate of growth gradually accelerated, and
accelerated more in the 20th century. Constant compound growth such as
2% implies accelerating absolute growth, just as a 2% gain in a teenager’s
weight is absolutely more than a 2% gain in a toddler’s weight. From the
perspective of two centuries, the compound growth rate in the industrial-
izing countries accelerated modestly, which implies a large acceleration in
the absolute growth rate.3 The successful countries are getting absolutely
richer very fast,. However, economic growth in the world is uneven, so the
economic gap between fast-growing countries and slow-growing countries
is much larger today than ever before in history. The wealth of the fastest
growers has risen above the slowest growers like the Himalaya Mountains
rising above the Ganges Plain.

3 

1.2
 Robert Cooter

Figure 1.1. Ganges Plain and Himalaya Mountains

Behind these facts stands a mathematical generalization: an economy that

increases at an exponential rate will overtake an economy that increases at
a constant absolute rate. (“Exponential rate” is a mathematician’s term for
“constant proportional rate”.) Figure 1.2 depicts this fact.4 First consider
function A, whose value increases at a constant rate with time, as indicated
by A’s constant slope. An addition to the value of A shifts it up and yields
B. A multiplication of B’s value rotates it up and yields C. The additive and
multiplicative increases make C larger than A or B at each point in time.
Now contrast C to function D that increases at an exponential rate. D starts
behind C at time 0, and overtakes it at time t*. This book concerns exponen-
tial growth like function D.

4  Joke: “Economists need a graph to reach the wrong conclusions with certainty, whereas lawyer
can do so immediately.

1.3
Ver. 1.4 - Chapter 1: Overtaking

Figure 1.2. Exponential Growth Overtakes Additive and Multiplicative

Increases

An economy produces different goods like houses, clothes, books, and clean
water. The usual measure of economic growth is the increase in wealth (a
stock) or income (a flow). To measure wealth or income, economists com-
bine different goods into comprehensive aggregates, such as gross national
product (GNP). To combine heterogeneous goods, economists first weight
quantities of goods by their market prices. Economists ideally adjust market
prices for unpriced externalities such as the social cost of pollution. Com-
prehensive measures of wealth or income include externalities, and also pub-
lic goods such as national security and air quality. To get closer to human
welfare, economists measure consumption instead of wealth or income, and
they divide the total by the number of people.5 Later we consider more con-
troversial measures, specifically utility and social welfare.

The vertical axis in Figure 1.2 is unlabeled because the measure of growth
-- total or per capita wealth, income, consumption, utility, or welfare – mat-
ters little to our conclusions. For all economic measures, exponential growth

5  In the national accounts, income can be “consumed” or “invested.” Consumption includes

almost all the activities that people will pay to do, but not activities that people must be paid to do.
Thus people “consume” an opera, a golf lesson, or a holiday. Note that in the technical language of
economics, people “consume” many things without using them up, like playing a recoding, whereas
in ordinary speech “consumption” refers to using up something, like eating food or burning wood.

1.4
 Robert Cooter

overtakes static effects. The overtaking principle in economics is my name for

the proposition that rapid exponential economic growth quickly overtakes
the direct effects of static efficiency and redistribution on wealth and welfare.
Given fast growth, static efficiency and income redistribution effects are im-
portant for their contribution to growth, and unimportant in themselves.

Besides prioritizing dynamic over static effects, the overtaking principle pri-
oritizes some dynamic paths over others. Thus in Figure 1.3, the growth-
path of income or wealth represented by D is unambiguously better than
the growth-path represented by E, because D starts higher and grows faster
than E. The analysis is different, however, when comparing E and F. F starts
lower than E, grows faster, and overtakes E at t*. If overtaking occurs rapidly,
then t* is small and nothing matters except growth. If, however, overtaking
occurs slowly, then t* is large and static effects of efficiency and redistribution
matter. Thus F is better or worse than E depending on the rate of growth.
The measure of growth seldom changes these conclusions because different
measures have similar overtaking times.

Figure 1.3. Different Exponential Growth Rates

In contrast, the overtaking principle loses much of its normative appeal when
growth and overtaking is slow. Thus interpret E and F as two sequences of
consumption in an infinite number of generations. In the first sequence E,

1.5
Ver. 1.4 - Chapter 1: Overtaking

initial consumption is higher and growth is lower. In the second sequence

F, initial consumption is lower and growth is higher. If overtaking occurs
quickly (t* is small), any reasonable economic measure will yield a larger
value for F than E. Conversely, if overtaking occurs slowly (t* is large), some
reasonable economic measures will yield a larger value for F than E, and oth-
er reasonable measures will yield a large value for E than F. For example, util-
itarians measure economic value by summing the utilities of different people.
Theorists disagree over whether lawmakers should give similar weight to the
utility of future generations as to the present generation. If growth is fast
and t* is small, the sum of utilities in F overtakes the sum of utilities in E for
any reasonable discount rate. 6 Conversely, if growth is slow and t* is large,
whether F or E yields the larger sum of utilities depends on the rate for dis-
counting future utility.

Innovation
Economic theory ascribes long run growth to three broad factors: physical
capital (buildings and machines), human capital (work adjusted for the skills
people bring to their jobs), and “innovation”. An increase in these factors
increases income and wealth. MIT economist Robert Solow won the Nobel
Prize partly for showing that innovation was more important to growth in
the 1950s in the United States than increased physical or human capital.
Subsequent empirical work by Edward Denison, Robert Barro, and others
confirmed this finding for other developed economies.7 In the last 100 years,
innovation caused more economic growth than anything else, including us-
ing more resources.

Innovations use new ideas to produce goods cheaper or to make better goods.
The supply curve shows the cost of making a good,8 so a cost-reducing inno-

6  Utilitarians often hold that future utility should not be discounted, whereas future income
should be discounted by its marginal utility. In this view, utility is what discounts money, but there
is nothing to discount utility. For a justification of not discounting future utility or welfare, see
Derek Parfit & Tyler Cowen, Against the Social Discount Rate, in Justice Between Age Groups and
Generations (Peter Laslett & Fishkin eds., 1992).
7  Add citations.
8  The supply curve is a monotonic relationship between price and quantity. Inferring from a
price to a quantity, the supply curve shows the quantity that a firm or industry will supply at a given
price. Conversely, inferring from a quantity to a price, the supply curve shows the cost for a firm or
industry to supply a given quantity.

1.6
 Robert Cooter

vation shifts the supply curve down, as when the tractor replaced the horse
for plowing. In Figure 1.4, innovation shifts the supply curve from S to S`,
and the shaded area indicates the social value of the savings in cost from pro-
ducing the quantity x of the good.

Figure 1.4. Social Value of a Cost-Saving Innovation

Similarly, producing a better good increases its value to consumers, as when

waterproof fabrics were made to breathe (“Gor-Tex”). When quality im-
proves, people are willing to pay more. The demand curve shows how much
people are willing to pay for a good,9 so a quality-improving innovation shifts
the demand curve up. In Figure 1.5, the innovation shifts the demand curve

9  Like the supply curve, the demand curve is a monotonic relationship between price and quan-
tity. Inferring from a price to a quantity, the demand curve shows the quantity that consumers will
buy at a given price. Conversely, inferring from a quantity to a price, the demand curve shows the
price that consumers are willing to pay for a given quantity.

1.7
Ver. 1.4 - Chapter 1: Overtaking

from D to D`, and the shaded area indicates the social value of the improve-
ment in quality from producing the quantity x of the good.

Figure 1.5. Social Value of a Quality-Improving Innovation

When one person uses a scarce resource, others cannot use it, like taking a
bite from a sandwich. Scarce resources like capital and labor have rival uses.
In contrast, when one person uses an idea, just as much remains for someone
else to use. Thus architects have used the Pythagorean Theorem for over two
millennia and just as much remains as before. Using theorems, principles,
natural laws, designs, discoveries, expressions, and compositions does not use
them up. Economists call this property “nonrivalry”. Furthermore, when
one person uses an idea, preventing others from doing so is difficult, more
difficult than preventing someone from trespassing on your land or taking a
bite from your sandwich. Ideas spread like gossip in Puddletown. Econo-
mists call this property “non-exclusion.”10

10  In economics, non-rivalry and non-exclusion define “public goods” like national security and
air quality, which resemble ideas in these two traits.

1.8
 Robert Cooter

“Nonrivalry” and “non-exclusion” are good news about ideas, and here is
even better news: ideas are fertile like seeds, not dead like coal. 2,500 years
after its proof, generalizations of the Pythagorean Theorem continue to ex-
pand geometry’s power.11 Similarly, the number of transistors that can be
placed inexpensively on an integrated circuit has doubled approximately ev-
ery two years for the past fifty years (Moore’s law). Ideas about transistors are
so fertile that innovation grows exponentially.

With a continuing stream of fertile innovations, growth can continue indef-

initely and production can rise like the falcon’s gyre.12 This hypothesis is a
reasonable conclusion from the last century of experience. It is also a reason-
able conclusion from recent technological and scientific progress, especially
develops in computers, nanotechnology, robotics, genetics, and medicine.
In spite of wars and recessions, the United States and other Western capital-
ist countries have averaged growth of 2 to 3 percent per year in per capita
GDP for more than one hundred years. In 1930 after the stock market’s
crash triggered the Great Depression, John Maynard Keynes wrote his essay,
“Economic Possibilities for Our Children”, which argues that the economic
problem of the future is how to use abundance. Similarly, in 2012 Dia-
mandis and Kotler wrote Abundance: The Future Is Better Than You Think.
Or perhaps absolute growth will accelerate so much that the future is un-
imaginable to us, as Kurzweil argued in The Singularity is Near: When
Human Transcend Biology (2005).

Depletion
Although the signs are good, the future remains in doubt until it arrives. A
worrisome doubt concerns resource exhaustion and environmental degrada-
tion. Must depletion of resources eventually end growth? Some resources
renew like a forest, river, or wheat. Use does not necessarily reduce their stock
permanently because we replenish it. Instead of renewing, however, other re-
sources deplete, like oil and iron. Use depletes their stock because we do not

11  The theorem says, “The square of the hypotenuse of a right triangle is equal to the sum of the
squares on the other two sides.” The fascinating story of the proof of its generalization in n-dimen-
sions is told in Simon Singh’s book, Fermat’s Enigma: The Epic Quest to Solve the World’s Greatest
Mathematical Problem (New York, etc.: Doubleday, Anchor Book, 1998).
12  For a survey of technologies where that great abundance is the future, see Peter H. Diaman-
dis & Steven Kotler, Abundance:  The Future Is Better Than You Think (2012).

1.9
Ver. 1.4 - Chapter 1: Overtaking

know how to replenish it. Whereas fertile ideas increase with use, depletable
resources diminish with use.

However, even with non-renewable resources, sustainable depletion is possi-

ble in principle. “Sustainable depletion” sounds oxymoronic like ”unmar-
ried husband”, but it makes sense. In the 5th century BC, Xeno posed his
paradoxes of motion that resembled the following: “By traveling half of the
remaining distance to your destination each day, you never arrive (except in
the mathematical limit when time goes to infinity).” Similarly, if the stock of
oil falls by 50 percent in every period, the stock never reaches zero in finite
time. “Sustainable depletion” refers to a path of depletion that never reaches
zero, such as Figure 1.6.

Figure 1.6. Sustainable Resource Depletion

Depletion is sustainable if the rate of depletion decreases forever. Instead of

decreasing, the rate of depletion is increasing in the world for many non-re-
newable resources. This trend could be reduced by continually reducing
consumption. However, laws and policies to reduce consumption encounter
fierce political resistance. If sustainability requires reducing consumption,
humanity might destroy its future rather than diminish its present.

Fortunately, sustainable depletion is possible in principle through innova-

tion, without reducing consumption. Innovation slows depletion by using

1.10
 Robert Cooter

resources more efficiently, as when new automobile engines economize on

fuel. Innovation also slows depletion by substituting renewable resources
for non-renewable ones, as when hydro replaces coal to generate electricity.
To deplete sustainably without decreasing consumption, innovation must
get more output from fewer inputs of non-renewable resources, and innova-
tion must substitute renewable resources for non-renewable resources. Con-
sumption can increase forever if the gains from innovation outpace the losses
from depletion.

Freedom
Every youth who watches the Olympics fantasizes standing on the central
platform with a gold medal while the national anthem plays. In wealth as
in sports, individuals and nations hope to surpass others and fear being sur-
passed by them. Like the Olympics, the value of winning the race for wealth
or income is self-evident to many people. If you are one of them, facts cited
above provide ample motivation to study law and growth economics.

Other readers of this book, however, will want moral or political justification
for focusing on the growth of wealth. Wealth and income are means for buy-
ing goods, not ends in themselves like beauty, love, holiness, knowledge, or
self-fulfillment.13 Accumulating wealth only to misuse it is a labor of shame.
Is the nation that wins the growth race like the winner of the pie-eating
contest whose prize is another pie? Does studying growth turn wealth into a
fetish like falling in love with a shoe?
13  The champion of this view in development economics is Amartya Sen, as two quotes suggest.
“Economic growth cannot be sensibly treated as an end in itself. Development has to be more con-
cerned with enhancing the lives we lead and the freedoms we enjoy.” Development as Freedom
(New York: Knopf, 1999), 14. “The challenge of development . . . is to improve the quality of life.
Especially in the world’s poor countries, a better quality of life generally calls for higher incomes—
but it involves much more. It encompasses as ends in themselves better education, higher standards
of health and nutrition, less poverty, a cleaner environment, more quality of opportunity, greater
individual freedom, and a richer cultural life.” World Bank, World Development Report 1991
(New York: Oxford University Press, 1991). Note that governments supply many non-market
goods, and GDP measures their value by their cost (e.g. salaries paid to civil servants), not by their
benefits to the citizens. Cost-benefit analysis can measure some of these non-market values more
convincingly. To measure the value of non-market goods, economists try to find out how much peo-
ple would pay for them if they had to pay, given that they don’t have to pay. This can be a measure-
ment maelstrom, so national accounting limits its use of cost-benefit analysis. For the relationships
between happiness and money, see B. S.Frey and A. Stutter (2002), “What Can Economists Learn
from Happiness Research?” Journal of Economic Literature.

1.11
Ver. 1.4 - Chapter 1: Overtaking

One justification for focusing on wealth and income concerns freedom. Like
driving a car, most people think that they are good at spending money, in
spite of occasional lapses such as purchasing uncomfortable shoes. People
know what they want to buy, and they criticize what others buy. A liberal
education should help you to think critically about what is worth buying. To
critique the uses of wealth, intellectuals draw on traditions in philosophy and
religion concerning what is really good, as opposed to what seems good. Thus
intellectuals in the university often carp about strip malls, popular movies,
fried food, ostentatious furniture, pay-day loans, and television preachers.

Most economists, however, refrain from debate about what people ought to
want. By doing so, economists cultivate neutrality in controversies about the
values of consumers. Instead of embracing a particular conception of good-
ness, economists often accept “consumer sovereignty” – the right of consum-
ers to buy what they want. Consumer sovereignty implies the right to buy
things that are good or bad, just as free speech implies the right to say things
that are true or false. I may disagree with what you buy and defend your right
to buy it, just as I may disagree with what you say and defend your right to
say it.14

Philosophers distinguish two kinds of freedom that apply to consumer sov-

ereignty.15 First, the absence of prohibitions or “negative freedom” increases
when people have fewer restrictions on what they can do, including spend-
ing their money. Economists promote negative freedom of consumers by
defending free markets. Second, the presence of opportunities or “positive
freedom” increases when people have more money to spend. Economists
promote positive freedom of consumers by finding policies and laws that
increase wealth.

Besides consumers, the two kinds of freedom apply to entrepreneurs. Eco-

nomic innovation mostly occurs in business ventures that develop new ideas.
Negative freedom gives entrepreneurs the legal right to experiment wktn new
economic ideas. Positive economic freedom gives entrepreneurs the resourc-
es to launch innovative ventures. Negative and positive economic freedom
is the legal condition for business creativity, as discussed in the next chapter.

14  That’s how I felt when I drove through Los Vegas.

15  The classical source is Isaiah Berlin, Two Concepts of Liberty, in Four Essays on Liberty
118–172 (1969).

1.12
 Robert Cooter

Welfare
Increasing negative and positive freedom is an appropriate goal for econo-
mists in a democratic state. Increasing freedom, however, is not the usual
justification given by economists for increasing wealth. Economists usually
justify increasing wealth as the means to increase the welfare of people. The
normative branch of economics connects wealth to welfare, rather than con-
necting wealth to freedom. In the history of economics, the philosophy of
utilitarianism was central to connecting wealth to welfare, especially in 19th
and early 20th century England.16 Utilitarians equate individual welfare with
a person’s utility, and they equate social welfare with the sum of individual
utilities. In notation, SW= u1+u2+…+un. Instead of thinking of social welfare
as a sum of individual utilities, however, modern economists think of it as an
increasing function of individual utilities: SW = f(u1, u2,…, un).17 The func-
tion may be additive as with utilitarianism, or it may be non-additive, such as
multiplicative (SW= u1×u2×…×un) or logarithmic (SW= log u1 + log u2+…+
log un). From a mathematical viewpoint, the social welfare function general-
izes utilitarianism,18 and different forms of the function have different ethical
consequences for equality.19

16  Economic theory assumed its recognizably modern mathematical form when theorists
understood that equating marginal benefits and marginal costs maximizes utility. This recognition
joined Newton’s calculus and the philosophy of utilitarianism and. See the “marginalist revolution”
in economics as discussed in the classic by Joseph A. Schumpeter, History of Economic Analysis
(1986).
17  “Utility” is a notoriously elusive concept that pervades economic theory. For a discussion of its
various meanings, see…
18  The concept of social welfare as a function of individual utility was introduced by Abram Berg-
son, A Reformulation of Certain Aspects of Welfare Economics, Quarterly Journal of Economics
310–34 (1938). All forms of the social welfare function, additive or non-additive, assume that social
welfare increase with individual utility. For a recent discussion of social welfare that is deep but
challenging to read, see Matthew Adler, Well Being and Fair Distribution (2012).
19  Summing utilities gives the same weight to utility regardless of its distribution among persons.
Thus, in a society of 5 people, social welfare is the same as measured by the sum of utilities if 5
people each enjoy 10 utils, or if one person enjoys 50 utils and the other 4 get 0 utils. In contrast,
social welfare is higher for the multiplicative or logarithmic social welfare functions if 5 people enjoy
10 utils than if one person enjoys 50 utils and the other 4 get 0 utils. In general, the additive form
of SWF is indifferent about the distribution of a fixed amount of utility across people, whereas the
multiplicative and logarithmic forms favor its equal distribution across people. Note, however, that
all three forms of SWF favor a more equal distribution of a fixed amount of income across people, so
long as the marginal utility of money is decreasing for all individuals. The most sophisticated treat-
ment of this problem, which is also challenging to read, is found in Matthew Adler, Well Being
and Fair Distribution (2012).

1.13
Ver. 1.4 - Chapter 1: Overtaking

People agree about some connections between wealth and welfare. Thus life
expectancy at birth is 83 years in Japan and 66 years in Bangladesh,20 and
enrollment in secondary school is 98 percent among Japanese children of the
appropriate age and 42 percent in Bangladesh.21 Facts like these make almost
everyone agree that welfare is higher in Japan than Bangladesh. The poor in
Bangladesh need extra money to buy basic health care and primary education
more than the rich in Japan need extra money to buy cosmetic surgery and
theater tickets. In general, the poor need extra money for necessities more
than the rich need extra money for luxuries.

Economists often describe this fact as the “decreasing marginal utility of

money with respect to income”. 22 Progressive taxes and government expen-
ditures can shift consumption from less to more basic needs, which presum-
ably increases a nation’s welfare. Many economists believe that the poor need
extra money more than the rich, but they disagree about how much more.
After a century of trying, there is no generally accepted method for measur-
ing the marginal utility of income like the thermometer measures heat. Thus
economists have no uniquely correct measure of the difference in welfare
between Japan and Bangladesh.

Perhaps comparing the marginal utility of income of rich and poor is like
distinguishing between your face and the back of your head: the difference
is real but the boundary is imprecise. In this respect, “social welfare” may
be like other values that social scientists measure in controversial ways, such
as “happiness”23. If so, economists can talk meaningful about the declining
marginal utility of income, but they will never find a uniquely correct mea-
sure of it. According to this view, economics must encompass controversial

20  Life expectancy at birth, total (years)) in 2008, World Bank. http://data.worldbank.org/
indicator/SP.DYN.LE00.IN. Note that life expectancy is in the 40s in many African countries.
21  Data & Statistics on Education for 2007 and 2008, World Bank, http://web.worldbank.
org/WBSITE/EXTERNAL/TOPICS/EXTEDUCATION/0,,contentMDK:20573961~menuP-
K:282404~pagePK:148956~piPK:216618~theSitePK:282386,00.html.
22  The utility of a poor person increases more from an increase in wealth than the utility of a rich
person (decreasing marginal utility of money. Social welfare increases in the utility of individuals.
Therefore, redistribution of wealth from rich to poor (with no loss in the amount of wealth) increases
social welfare. In Adler’s theory (see preceding footnote), social welfare increases with individual
wealth, and social welfare increases with a more equal distribution of wealth or utility across individ-
uals.
23  See Bruno S. Frey & Alois Stutzer, What Can Economists Learn from Happiness Research?,
Journal of Economic Literature (2002). Also see Daniel Kahneman & Richard Thaler, Utility
Maximization and Experienced Utility, 20 Journal of Economic Perspectives 221–234 (2006).

1.14
 Robert Cooter

measures of imprecise facts. Thus the “material welfare” school of the early
20th century held that additional money benefits the poor more than the
rich, which can be shown by measures of the causes of welfare like health and
education that are reasonable, pragmatic, and incomplete.24

In contrast, another tradition in economics called “positivism” holds that

the marginal utility of money has no scientific meaning. Given enough data,
all scientific propositions can allegedly be proved true or false. In contrast,
propositions about “social welfare” or the “marginal utility of money” are val-
ue judgments or expressions of political commitments, not testable assertions
of fact. The question “How much higher is the marginal utility of income
of the poor than the rich?” has no more scientific meaning according to the
economic positivists than the question, “Is it morning or afternoon on the
sun?” Since value judgments do not make testable claims, according to the
economic positivists, “social welfare” cannot be part of scientific economics.

In any case, measuring social welfare and the declining marginal utility of
money is less important to growth economics than to static economics.
When progressive taxation and state spending shift consumption from lux-
uries to necessities, the static effects – whether measured in wealth, income,
utility, or welfare for the poor or everyone in society – correspond to the shift
in Figure 1.2 from A to B, or from B to C. The static effects are small relative
to the dynamic effects represented by D, regardless of how they are mea-
sured. When growth is exponential, the mathematics of overtaking applies,
regardless of the function’s interpretation in Figure 1.2. Measuring utility or
welfare is unimportant to policy conclusions when fast growth causes rapid
overtaking, whereas the measure is important when sluggish growth causes
slow overtaking.

To illustrate, assume that redistributive policies improve the health and edu-
cation of the poor, which directly increase their welfare. In addition, health-
ier workers with better education are more creative, so they may increase
economic growth. Economists who believe that better education and health
of poor Americans would increase U.S. growth point to Denmark and Korea
as examples where good education and health have contributed to robust

24  For an account of the material welfare school and its tension with scarcity economics, see Rob-
ert Cooter & Peter Rappoport, Were the Ordinalists Wrong About Welfare Economics?, 22 J.Economic
Literature 507 (1984).

1.15
Ver. 1.4 - Chapter 1: Overtaking

economic growth, whereas poor education and health of workers may partly
explains the economic struggles of the Philippines.

In the preceding hypothetical, redistribution increases the growth rate by

improving the welfare of workers. Redistributive policies can also decrease
the growth rate by undermining economic incentives. China’s Cultural Rev-
olution, which commenced in the 1960s and expired by 1975, enforced
strict economic equality and destroyed what was left of the private sector.
Economic decline immiserized many people. Reversing policies after 1980,
China dissolved agricultural communes, farmers were allowed to keep the
surplus from selling their crops, and agricultural productivity soared. Spec-
tacular success in agriculture prompted the Chinese leadership to allow the
development of private business in the new export sectors, where entrepre-
neurs could keep much of what they earned. Improved incentives unleashed
economic growth without historical precedent. Instead of trickling down to
the Chinese workers, the benefits of growth cascaded down like the Yangtze
River at the Three Gorges dam. Equality decreased, economic growth ex-
ploded, and poverty plummeted. The lowest wage earners in China bene-
fited greatly from a faster growth rate in national income after 1980. Even
unemployed poor people who depend on welfare payments and government
subsidies benefitted from faster growth, which increased state revenues avail-
able for transfer payments.

Conclusion
The first question of law and growth economics is, “Which laws increase the
pace of economic innovation?” Increasing the pace of innovation can lead
to sustained growth in wealth, income, utility, and welfare. When growth is
rapid, overtaking is the only normative standard required to choose among
many laws and policies. Balancing growth against static efficiency or equality
is unnecessary when growth is fast, which simplifies the analysis of many laws
and policies such as patents. Isaac Newton invented calculus and discovered
the laws of motion associated with gravity in 1666, which historians of sci-
ence call the “miracle year” (“annus mirabilis”). Civil engineers still use New-
tonian physics for ordinary objects, although it fails for very large objects
(the cosmos) and very small objects (sub-atomic particles). Similarly, rapid
growth is the domain of the overtaking principle, which is good for making

1.16
 Robert Cooter

law and policy in most of growth economics. In contrast, sluggish growth is

the domain of efficiency and equality.

The overtaking principle supplies the normative justification for prioritizing

growth economics over static economics. It challenges conventional law and
economics that treats static efficiency as a fundamental goal of law and poli-
cy. Which was more important to agricultural production, inventing a trac-
tor or using horses more efficiently? A better allocation of horses for plowing
the fields increases agricultural production marginally, whereas inventing
the tractor caused a jump in production. Once you appreciate exponential
growth, it’s hard to care about static efficiency for its own sake.

The overtaking principle also challenges ethical theories concerning the re-
distribution of wealth. Controversies about fair distribution, social welfare,
the marginal utility of income, and time-discounting do not matter when
growth is rapid. In much of political philosophy, fairness concerns distrib-
uting shares of fixed income, and economic equality is an end in itself. With
rapid growth, however, putative ends turn out to be only valuable as means.
Economic equality mostly affects welfare through growth, not in its own
right. When rapid growth is possible, progressive taxes and state expendi-
tures increase the welfare of the poor if they cause faster growth in wages and
subsidies.

Perhaps you think that static efficiency and growth align, with more efficien-
cy causing faster growth. This view is roughly two-thirds right and one-third
wrong. Efficiency and growth have a common cause: competition. To
maximize innovation, the law must create a framework of open competition
so innovators can develop their ideas. The core of economic growth comes
from entrepreneurs. Only the possibility of extraordinary profits can induce
entrepreneurs to bet big on risky ideas. However, extraordinary profits re-
quire market power, not competition. Thus patent law creates open compe-
tition to innovate and rewards the winners by giving them temporary market
power. As told in this book, the story of law and growth economics is open
competition to innovate and extra-ordinary profits for the winners.

Astronomers from the time of Aristotle saw the earth as the center of the
universe and the sun revolving around it, until the Copernican Revolution
put the sun at the center. In general, a revolution rearranges the central

1.17
Ver. 1.4 - Chapter 1: Overtaking

elements of a scientific theory.25 Such a revolution occurred in economics in

the 1930s. Before the revolution, economics was defined as the science of
material welfare, and then a famous essay proposed this alternative in 1932:
“Economics is the science which studies human behavior as a relationship
between given ends and scarce means which have alternative uses.”26 The
scarcity definition formulated a new scientific paradigm for economics that
displaced its predecessor.27 The scarcity definition characterizes contempo-
rary law and economics, which can be called “law and scarcity economics.”
This book replaces scarcity with growth as the paradigm, yielding “law and
growth economics.” Once one starts to think about law and growth eco-
nomics, it is hard to care about law and scarcity economics.

25  This is the thesis of the classic by Thomas S. Kuhn, The Structure of Scientific Revolu-
tions (1st ed. 1962).
26  Lionel Robbins, An Essay on the Nature and Significance of Economic Science. (London:
McMillan, 1932),page 16.
27  For a discussion of the paradigm shift, see Robert Cooter and Peter Rappoport, “Were the
Ordinalists Wrong About Welfare Economics?,” J.Economic Literature 22 (1984): 507.

1.18
 Robert Cooter

Chapter 2. Ventures1
Robert Cooter and Aaron Edlin

A bold ship’s captain in seventeenth-century England proposes to investors

that they finance a voyage to Asia for spices.2 The voyage is inherently risky.
Weather is uncertain, channels are uncharted, the Dutch prey on English
ships, the English prey on Dutch ships, and other pirates prey on both of
them. If the captain returns to the English port with a cargo of spices, how-
ever it will be worth a fortune. The ship’s captain needs a large ship outfitted
for two to five years of travel. He talks to investors and discloses secrets about
how to get to Asia and what to do when he arrives. The investors decide to
entrust him with a ship and supplies. Unlike so many other ships that sail for
Asia, this one returns safely after two years. The townspeople spot the vessel
sailing toward the harbor and the investors rush to the dock to divide the
cargo. The success of the venture attracts imitators. As years pass, the spice
trade eventually becomes a normal business with moderate risk and ordinary
profits.

Similarly, an engineer in Silicon Valley in 1985 has an idea for a new com-
puter technology. The engineer cannot patent the idea until he develops it.
Developing it requires more money than the engineer can risk personally. He
drafts a business plan, meets with a small group of investors, and explains
his idea to them. Developing the idea is inherently risky— someone might
steal the idea before it is patented, an unknown competitor might patent the
invention first, the invention might not work, or it might work but not sell.
If the plan succeeds, however, the innovators will make a fortune. The inves-
tors agree to form a company and develop the product. Unlike so many oth-
er start-ups, this one succeeds after two years and the firm acquires a valuable
patent. The company sells the product to an established firm and divides the
proceeds between the engineer and the investors. The success of the venture
attracts imitators. Production and sale of the invention eventually becomes
a normal business with moderate risk and ordinary profits.

1  This chapter draws on Chapters 4 and 5 of Cooter and Schaefer, 2011.

2  This characterization of the spice trade is based on Ron Harris, “Law, Finance and the First Cor-
porations”, in James J. Heckman, Robert L. Nelson & Lee Cabatingan, eds., Global Perspectives on
the Rule of Law ((2009, Routledge).

2.1
Ver. 1.4 - Chapter 2: Ventures

Seventeenth-century spice voyages and twentieth-century technology start-

ups involved up-front investment, high risk, and high return. Many business
ventures have these characteristics. Discovering new ideas and developing
them usually requires upfront investment. If development succeeds, the in-
novator has a temporary advantage over competitors until they catch up.
While the temporary advantage lasts, the innovator enjoys extraordinary
profits, which we call “venture profits.” In the end, the imitators catch up
and profits fall to the normal level. The life-cycle of a business venture is this
chapter’s subject.

Life-Cycle of a Business Venture

In efficiency economics, the production function and competitive equilib-
rium are fundamental tools of analysis. In growth economics, the business
venture and its life cycle are fundamental tools of analysis.3 A successful
business venture has numbers like Figure 2.1. The venture begins with the
development of a new idea in period 1, which costs 8. When the product is
developed, the innovator has a valuable secret or patent, or perhaps a cluster
of secrets or a portfolio of patents (more on that later). After development,
the innovation is launched and marketed to buyers. When launched in pe-
riod 2, the innovation has no competitors, so the innovator is a monopolist
who receives a payoff of 7. In period 3, imitators develop competing prod-
ucts that substitute for the innovation, which reduces the innovator’s payoff
to 4. In period 4, imitations improve and competition intensifies. Taking
competition to its logical extreme in period r, the imitations become perfect
in period 4, so the market is perfectly competitive and the innovator’s payoff
is zero. (“Zero profit” is economist’s code for “ordinary rate of profit”.4)
Summing over the life cycle, the venture’s net payoff equals +3.

3  A production function determines outputs from inputs, such as y=f(k,l) where the output is y
and the inputs are capital k and labor l. In a competitive equilibrium, the price of output y equals its
marginal social value. So the marginal value product of an input, say capital k, equals pf1. A business
venture at time t in its life-cycle has an expected profit function πt determined by the probable payoffs
as a function of future investment of k and l. As explain in Chapter 1, the social value of alternative
business ventures is compared by the contributions to growth. According to the overtaking principle,
one business venture that increases the sustained rate of growth relative to another business venture
is much more socially valuable, perhaps infinitely so.
4  Perfect competition drives the prices of all goods to their cost of production. Profits are zero
after including the cost of capital in the other costs of production. The cost of capital equals the
ordinary rate of profit in alternative uses.

2.2
 Robert Cooter

A successful innovation is like a horserace that pays handsomely to the win-

ner, moderately to the near-winners, and nothing to the rest of the pack. A
successful business venture earns extraordinary profits when launched, ex-
traordinary profits when imitated, and ordinary profits when perfectly sub-
stituted. The industry eventually settles into equilibrium, like the crowd
eventually stops yelling after horses cross the finish line.

Figure 2.1. Life Cycle of a Successful Venture

2.3
Ver. 1.4 - Chapter 2: Ventures

Figure 2.1 indicates the benefits and cost of the venture to its owners. What
about its benefits and costs to society? As conventionally measured, the net
social benefits from a successful business ventures equal the sum of innova-
tor’s profits, profits of other firms such as imitators’, and the consumer’s sur-
plus.5 The innovator’s profits are a small fraction of the net social benefits.
The richest innovators get less wealth for themselves than the value that their
innovations convey to consumers and other producers. Thus the wealth that
Apple investors obtained from the iPhone is less than the value of the iPhone
to consumers and other firms that imitate it or create applications for it. In
the appendix to this chapter, Tables 2.1 and 2.1 use numbers to illustrate the
venture depicted in Figure 2.1,

Production of the improved widget will continue beyond time 4 under con-
ditions approximating perfect competition until the product becomes obso-
lete. The product becomes obsolete when a new innovation destroys the old
one’s value and the industry begins a new cycle of innovation. When ventures
like the one in Figure 2.1 repeat themselves, one innovation follows another,
and the path of net social benefits traces a gyre like the falcon on this book’s
cover. In the appendix, Figure 2.4 traces the gyre of net social benefits from
a numerical example.

Imitators or Enthusiasts?
Figure 2.1 depicts a profitable venture, but most ventures fail and lose mon-
ey. Recent U.S. data suggests that 40% of new businesses survive and 60%
disappear within four years.6 Figure 2.3 depicts a losing venture. The in-
novator in Figure 2.2 spends 8 in period 1 to develop the product. Many
innovations fail before completing development and beginning production,
without recouping any development costs. The innovator in Figure 2.3,
however, is a little more successful and brings the product to market. When
the innovation is launched in period 2, the innovator has no competitors and
enjoys profits of 7. The only difference between Figures 2.1 and 2.3 is in pe-
5  See appendix to Chapter 1. Recall that the consumer surplus equals the difference between what
the good is worth to consumers and what they must pay for it, or the difference between the wili-
ness-to-pay and the market price.
6  44% of U.S. businesses started in the 1990s still existed 4 years afterwards. See Knap, A. E. (2005).
“Survival and longevity in the Business Employment Dynamics data.” Bureau of Labor Statistics of
the U.S. Department of Labor. The number given by Dun and Bradstreet is 37% as cited in “Some of
the Reasons Why Business Fail and How to Avoid Them,” Entrepreneur Weekly, Issue 36, 3-10-96.

2.4
 Robert Cooter

riod 3. In Figure 2.1 the innovation yields a payoff of 4 in period 3 because

it is better than the imitations. In Figure 2.2, however, the imitations in peri-
od 3 are just as good as the original – they are perfect substitutes. Perhaps the
innovator cannot keep a secret or it has no patent. In Figure 2.2, the market
reaches a competitive equilibrium in period 2, so the innovator’s payoff is 0
in period 2, as well as in period 3. Summing over the life cycle, the venture’s
net payoff equals -1. The innovator in Figure 2.3 cannot recover the cost of
developing the new idea.Figure 2.2. Life-Cycle of an Unsuccessful Venture

2.5
Ver. 1.4 - Chapter 2: Ventures

The innovator sometimes has the advantage as in Figure 2.1, and the imitator
sometimes has the advantage as in Figure 2.2. If the innovator in period 0
can foresee that the venture will have a net payoff of +3 as in Figure 2.1, it
will launch the venture. The expectation of positive venture profits causes the
development of innovations. If the innovator in period 0 can foresee that
the venture will have a net payoff of -1 as in Figure 2.2, it will not launch the
venture. The expectation of negative venture profits prevents the develop-
ment of innovations.

The difference between a profitable venture in Figure 2.1 and an unprofitable

venture in Figure 2.2 is the ease of imitation. Difficult imitation extends the
innovator’s period of extraordinary profits in Figure 2.1, and easy imitation
reduces the innovator’s period of extraordinary profits as in Figure 2.2. To
slow imitators, an innovator often tries to keep the innovation secret. Some
innovations reduce to explicit information that is easy to copy, like a recipe
or a formula. Being easily copied, explicit information is intrinsically hard
to keep secret. Thus developers of software need to understand parts of the
core code in order to write new software Microsoft Windows. Microsoft will
disclose parts of its core code to software developers, but not all of it.

Unlike a computer code, other innovations involve implicit information that

is irreducible to simple communication. Implicit information is often im-
bedded in a practice or organization, like judgment in mixing chemicals, art
in baking cakes, or methods to motivate salesmen. Information is implicit
when someone knows how to do something that is hard to communicate.
Because communication is hard, implicit information is easier to keep se-
cret than explicit information. To steal another company’s implicit infor-
mation, you need to hire its employees rather than readings its documents.
(See Chapter _.)

The law of trade secrets helps innovators to keep their secrets private. A per-
son who violates these laws by distributing proprietary information is liable
for the harm done to the firm with the secret. When firms in Silicon Valley
hire employees or discuss collaboration with other firms, they routinely sign
non-disclosure agreements (NDAs) in which they promise not to reveal any
secrets that they learn about the company. Sometimes trade secrets laws
work -- the recipe for Coca Cola has remained a secret for decades. More

2.6
 Robert Cooter

often, trade secret law is ineffective. Trade secrets laws are hard to enforce in
Silicon Valley and they are unenforced in much of the world.7

Instead of secrets, another route to extraordinary profits is patenting. To

patent an invention, the innovator must reduce it to explicit information
disclosed in the patent application. Any member of the public can read the
patent and obtain essential information for producing the invention. How-
ever, producing the invention infringes the patent, which is illegal without
the consent of the patent holder. The creator of a patentable invention must
decide whether secrecy or a patent is more profitable.

A patent applies to an invention, not to a market. However, some patent

portfolios industry standards that dominate markets, like the Windows oper-
ating system currently dominates desktop computing in business firms. The
owner of a portfolio of patents that becomes an industry standard can collect
royalties from an entire industry. Standards are a kind of natural monopoly
created by a coordination problem, as explained in Chapter _. With indus-
try standards as with pop stars, many volunteer and few are chosen.

An invention is likely to become a standard if its value for each user increases
as the number of its users increases. The uncompensated benefit that one
user conveys to another is a “positive externality.” The joint operation of the
positive externalities is a “network effect.”

The three sources of extraordinary profits for innovators are secrecy, intellec-
tual property, and natural monopoly. Some innovators enjoy extraordinary
profits from all three sources. Thus Microsoft cloaks the core code of Win-

7  Yuval Feldman, Confidential Know-how Sharing and Trade-secrets Laws: Studying the Interac-
tion Between Legality, Social Norms and Justice Among High-tech Employees in Silicon Valley (PhD
Thesis, Law School, University of California at Berkeley, 2004.
In addition, patents and copyright give innovators a temporary monopoly by law. An effective patent
gives the inventor monopoly profits until the patent expires or new inventions undermines its value.
To illustrate numerically, consider how A’s profits in Figure 1.5 might change if A holds an effective
patent to its innovation that lasts for periods 1 and 2. If A refuses to license the innovation to anyone,
it enjoys monopoly profits of 10 in time 1 as indicated in Figure 1.5. Assume that the patent extends
through time 2 and that competitors make little progress circumventing the patent by their own inno-
vations. In time 2, A enjoys profits much like in time 1, say, profits of 9. Thus the patent increases A’s
earnings net of development costs for all periods from 6 to 13. Later chapters contain more details
about intellectual property, including patents and copyright.

2.7
Ver. 1.4 - Chapter 2: Ventures

dows in secrecy, protects parts of it by patents and copyright, and enjoys

natural monopoly from network effects.

Rate of Innovation
The expectation of positive profits launches innovative ventures, and the ex-
pectation of negative prevents innovative ventures. Law affects the profit-
ability of all phases of a venture -- finance, development, marketing, and
competition. When better law makes innovations more profitable, the num-
ber of innovations increases. To see why, think of an array of new ideas that
differ according to the expected profitability of developing them. Innovators
develop the ideas that they expect to yield positive profits, and they do not
develop the ideas that they expect to yield negative profits. If better law
increases the expected profits for innovations, some ventures will tip from
negative to positive expected profits, so innovators will develop more ideas.

Figure 2.3 depicts this fact. The horizontal axis indicates an industry’s array
of possible innovations in order of profitability, with higher profit ventures
farther to the left and lower profit ventures farther to the right. Figure 2.3
distinguishes between profits under bad and good law. An innovation’s prof-
itability under bad law equals the difference between its revenues R and costs
C over the venture’s life. To the left of I, revenues R exceed costs C, so the
innovations will be developed. To the right of I, costs C exceed revenues R,
so the innovations will not be developed. The revenue curve R intersects
the cost curve C at point I where venture profits are zero (development costs
before launch equal profits after launch). I is the tipping point that indicates
the number of developed innovations under bad law.

2.8
 Robert Cooter

Figure 2.3. Development of More Innovations with Better Law that

Increases Venture Profits

Improving the law in this industry increases revenues from R to R* and

decreases costs from C to C*. With improved law, the revenue curve R* in-
tersects the cost curve C* at point I*. I* indicates the number of developed
innovations under good law. I* is larger than iI so better law causes higher
venture profits, and higher venture profits cause more ventures.

Conclusion
Chapter 1 posed the first question of law and growth economics: “Which
laws increase the pace of economic innovation?” Chapter 2 simply answers,
“The laws that increase venture profits.” Fleshing out this answer requires the
rest of this book. First we must look inside the business venture. The next

2.9
Ver. 1.4 - Chapter 2: Ventures

chapter explains how law enables individuals with different interests to give
their ideas and money to innovative ventures with shared goals.

2.10
 Robert Cooter

Appendix
Table 2.1 uses numbers to illustrate the business venture in Figure 2.1. As-
sume that the innovator develops an innovation that lowers the cost and
improves the quality of a particular consumer good – say, a better widget.
At time 0, before development of a better widget, widget makers earn zero
profit and consumers enjoy a surplus of So. At time 1, development of the
improved widget costs the innovator -8. At time 2 when the improved wid-
get is launched, the innovator has lower production costs, so the innovator
enjoys extraordinary profits of 7 temporarily. Informed consumers buy the
improved widget and enjoy its higher quality, so the consumer’s surplus rises
to So+10. At time 3, some imitators partially succeed in replicating the in-
novation. This competition drives the price down, the innovator’s profits fall
to 4 at time 3, and the imitators enjoy profits of 2. The fall in price causes
the consumer’s surplus to increase by 30, and the rise in average quality of
widgets causes the consumer’s surplus to increase by 3. At time 4, compe-
tition drives the price down to the cost of production, so the innovator and
imitators receive zero profits. The fall in price causes the consumer’s surplus
to increase by 70, and the rise in average quality causes the consumer’s sur-
plus to increase by 10. Summing over the life cycle in Table 2.1, the venture
profits are 3 and the consumer’s surplus increases by 123. Consequently, the
innovator enjoys less that 3% of the innovation’s total benefit to society.

Table 2.1. Net Benefits from Business Venture

Phase in Cycle of Innovation
time 0 time 1 time 2 time 3 time 4
equilibrium development market power imitation equilibrium
venture
profits
0 -8 +7 +4 0
imitators’
profits
... ... ... +2 ...
consumer’s
surplus
So So So+10 So+30+3 So+70+10
total net
benefits
So -8+So So+17 So+39 So+80

2.11
 The explanation of the following table shows that these numbers are reason-

Summary table of numbers underlying Figures 2.1 and 2.2.

time or phase
perfect oligopoly perfect
development monopoly
competition (imitation) competition
price 1 1 1 .97 .93
venture’s production 100 100 100 100 100
venture’s revenues 100 100 100 97 93
venture’s cost -100 -108 -93 93 -93
venture profits 0 -8 +7 +4 0
Ver. 1.4 - Chapter 2: Ventures

imitators’ costs ... ... ... .96 .93

imitators’ profits ... ... ... 2x.01x100 0
consumer’s surplus So+.03x1000 So+.07x1000
So So So+.01x100
(100 x (value-price)) +.01x300 +.01x1000
total profits +

able.
So -8+So +7+So +6+So+30 + 3 So+70 + 10

2.12
consumer’s surplus
 Robert Cooter

The industry consists of, say, 10 firms and each of them supplies 100 units
of the good for total production of 1,000. The unit cost of production un-
der the original technology is 1, the market price of the good is originally 1
under perfect competition, and the consumer’s surplus is So. The innovator
spends 8 to develop a new production method that reduces the cost to .93.
Since the good’s price is 1, the innovator earns .7 per item for 100 items in
the monopoly stage when the innovation is launched. The innovation also
improves the good’s quality. 100 consumers know that the innovator’s good
has higher quality, so their surplus increases by .1 x 100.

In the oligopoly phase, 2 imitators comprehend enough of the secret to low-

er their production costs from 1 to .96. Oligopolistic competition bids the
market price down to .97. Each of the 2 imitators produces 100 units and
earns profits of 1. When the price falls to .97, the increase in the consumer
surplus from the fall in price equals 1,000 x .03. In addition, the 300 con-
sumers who buy from the innovator or the imitators enjoy an increase in
surplus by .1x300.

In the final phase of perfect competition, all firms learn the new production
technique, the price falls to the cost of production .93, and all firms earn
zero profits. Consumers buy 100 units from each of 10 firms. The original
consumer surplus is So. When the price falls to .93, the consumer surplus
increases by 1,000 x .07. In addition, the consumer surplus increases from
the higher quality of the good by 1,000 x .01.

Notice that these calculations assume fixed market shares and perfectly in-
elastic demand, which eliminates the “welfare triangles” in a standard effi-
ciency analysis. A more complete analysis that relaxed these assumptions
would have larger total net benefits.

The product becomes obsolete when a new innovation destroys the old one’s
value and the industry begins a new cycle of innovation. The graph of the
net social benefits forms a gyre as depicted in Figure 2.4. The first cycle of
innovation begins at time 0 when social benefits equal So and ends at time 4
when social benefits equal So+80. The curve beyond time 4 suggests what the
next cycle of innovation looks like.

2.13
Ver. 1.4 - Chapter 2: Ventures

Figure 2.4. Growth Gyre

2.14
 Robert Cooter

Chapter 3. The Double Trust Dilemma

Robert Cooter and Aaron Edlin

When someone discovers a better way to make something or something bet-

ter to make, developing the new idea usually requires capital. One person of-
ten has ideas and someone else has money. To launch an innovative venture,
they must combine their assets, which require the innovator to trust the fi-
nancier with his idea and the financier to trust the innovator with her capital.
We call the problem of uniting capital and ideas the double trust dilemma of
innovation -- a new name for an idea with a rich economics pedigree.1

An economist who worked at a Boston investment bank received a letter

that read: “I know how your bank can make $10 million. If you give me $1
million, I will tell you.” The letter captures concisely the problem of buying
information: The bank does not want to pay for information without first
determining its worth, and the innovator fears to disclose valuable informa-
tion without first getting paid. In general, a person cannot evaluate an idea
until after its disclosure, and after its disclosure she may not need to pay for
it.2 If you know information then you have it, whereas you can know a car,
cow, or coffee without having it.

Combining this characteristic of information with investment creates the

double trust dilemma. A Berkeley mathematician named Richard Niles in-
vented bibliographic software called EndNote that many professors use on
1  The phrase “double trust dilemma” was introduced by Cooter and Schaefer in Chapter 3, “The
Double Trust Dilemma of Development,” Solomon’s Knot: How Law Can End the Poverty of
Nations. Contemporary (Princeton University Press, 2012). Theories of finance often begin with the
question, “How can an investor, who puts his money under the control of a manager, write a con-
tract so that the manager profits most when the investor profits most?” This is the “principal-agent
problem.” This is a single-trust problem because the investor must trust the manager with his money,
but the manager need not trust the investor. This single-trust problem is the building block for
analyzing double trust problems. A good introduction to this vast literature is Kenneth J. Arrow,
“The Economics of Agency: An Overview,” in Principals and Agents: The Structure of Business, ed.
John W. Pratt and Richard J. Zeckhauser, 1985. For a pioneering paper on secrecy and investment,
see Edmund W. Kitch, “The Law and Economics of Rights in Valuable Information,” J. Legal Studies
9 (1980). Two-sided trust problems have been investigated in game theory, such as two-sided moral
hazard or principals-agent with two principals. For an example of how modern financial institutions
combine ideas and capital, see Bernard Black and Ronald Gilson, “Does Venture Capital Require an
Active Stock Market?,” 11 Journal of Applied Corporate Finance 36 – 38 (2005).
2  Economists call this fact “Arrow’s paradox of information.” For an early exploration of this
problem of asymmetric information, see Arrow, Kenneth J. (1972) The Value of and Demand
for Information, in C. B. McGuire and R. Radner (eds.), Decision and Organization, New York:
North-Holland, Chapter 6.

3.1
Ver. 1.4 - Chapter 3: The Double Trust Dilemma

their computers. In the early stage of development, he hoped and feared

receiving a call from large firm asking for an explanation of EndNote. Once
the large firm understood the product, it might buy Niles’ company and
make him rich, or it might develop its own version of his program and bank-
rupt him. Niles eventually got a call from a large publisher (Thompson) who
bought the company.

A new idea collides with capital in Silicon Valley when an innovator presents
a business plan to potential investors. Here is a simple example of how such
a presentation might go:

“I have a new idea. I can use my time to develop it and sell it for $1 mil-
lion, or I can explain it to you. I have no effective legal protection for my
idea. Once you know it, you can use your capital of $3 million to develop it
yourself and sell it for $5 million, and I will gain nothing. However, once you
know the idea, I think that you will see that we can make a lot more money
by working together. If you supply me with $3 million in capital, I can use
my expertise to develop the innovation and sell it for $15 million. With $15
million, we can give you $5 million for not developing it yourself, and we can
split the remaining $10 million.”

Figure 3.1 depicts this proposal as a decision tree. In step 1 the innovator can
develop his idea without capital and earn 1, in which case the financier keeps
her capital of 3. Alternatively, in step 1 the innovator can explain his idea to
the financier and go to step 2. The innovator has no effective legal protection
of his idea. In step 2 the financier can use capital of 3 to develop the idea her-
self and sell it for 5. Alternatively, in step 2 the financier can supply capital to
develop the innovation in cooperation with the innovator and they can earn
15, with the innovator getting 5 and the financier getting 10. Each receives
what he could get without cooperation plus an equal share of the surplus
from cooperation. (This is the “Nash bargaining solution” in game theory.3)

3  The Nash Bargaining Solution is named after its inventor, John Nash, whose genius and insanity
inspired the popular Hollywood film, “A Beautiful Mind.” It is an axiomatic account of coopera-
tion. The Nash bargaining solution corresponds to the outcome of a non-cooperative game repeated
indefinitely. See Ariel Rubinstein,. “Perfect Equilibrium in a Bargaining Model”. Econometrica 50
(1): 97–109. doi:10.2307/1912531..

3.2
 Robert Cooter

Figure 3.1. Proposed Solution to Double Trust Dilemma

Many business ventures begin with plans resembling this one. The plan in
Figure 3.1 proposes payoffs of 5 to the innovator and 10 to the financier. To
implement this proposal, the innovator and financier must make 5 and 10
into the actual payoffs from cooperating. Successful implementation implies
that the innovator and financier will get the payoffs in Figure 2.1 if they
make the decisions depicted in the figure. If the plan can be implemented,
self-interest will motivate the parties to cooperate as envisioned in the pro-
posals. The financier will decide in stage 2 to supply capital to the innovator
to develop the idea, because the payoff from cooperation exceeds the payoff
from non-cooperation: F2>F1. Foreseeing this fact, the innovator will decide
to cooperate in stage 1, because the payoff from explaining his idea exceeds
the payoff from developing it on his own: I2>I0. These two inequalities ex-
press the conditions for cooperating in a venture structured like Figure 3.1.
Successful implementation of the plan satisfies these two conditions.

Implementing such a plan is remarkably difficult. Investing in a business

venture requires confidence that strangers will not appropriate its profits.
The innovator and financier will need protection against outsiders who might
steal 15 from them. All ventures require protection from predators such as
hoodlums, mafias, cheating accountants, Ponzi artists, conniving state reg-
ulators, oppressive tax collectors, and thieving politicians. Without effective
property protection, people fear theft, so resources flow to protectors, not to

3.3
Ver. 1.4 - Chapter 3: The Double Trust Dilemma

entrepreneurs. Families, clans, and gangs protected property historically, and

they continue to do so today in some countries. However, an effective state
is more reliable than private protectors. State protection of property is the
legal foundation for investment in the future, including business ventures by
families and friends.

If the innovator and financier are protected against outsiders, they still need
to trust each other. At the final stage when the firm’s has 15, the plan calls
for the financier to receive 10 and the innovator to receive 5. What prevents
the innovator from grabbing all 15? Perhaps the innovator will take all 15 as
salary and leave nothing for paying dividends to shareholders. Many busi-
ness ventures never launch because the financier cannot trust the innovator
to distribute the profits as promised.

If the final stage of Figure 3.1 cannot be implemented, then the whole pro-
posal collapses. Assume that if venture were to reach the final stage, the
innovator would grab all 15, as depicted in Figure 3.2. Foreseeing this fact,
if the venture reaches stage 2, the financier will develop the innovation her-
self, so the financier’s payoff will be 5 and the innovator’s payoff will be 0.
Foreseeing this fact, in stage 1 the innovator will develop the idea himself
and receive a payoff of 1, instead of explaining the idea to the financier. The
payoffs in Figure 3.2 fail to implement the plan in Figure 3.1 for sharing the
gains from cooperation, so cooperation unwinds and noncooperation stifles
an innovative business venture. (Notice that the payoffs in Figure 3.2 violate
the two conditions for cooperation: F2>F1 and I2>I0.)

3.4
 Robert Cooter

Figure 3.2. Unsolved Double Trust Dilemma

Relationships
How can the innovator and financier implement a business plan as depicted
in Figure 3.1, rather than collapsing as depicted in Figure 3.2? One way is
to deal with trusted relatives. Loyalty within families is partly irrational and
partly rational. Here’s how rationality prompts family members to cooperate
with each other. Kinship is ascribed, not chosen. Your uncle is your uncle
until death. Ascription provides a framework for repeat dealings. Through
repeat dealings, people come to know whom to trust and whom to distrust.
One reason to be trustworthy is fact that your relatives will find out if you
are not trustworthy.

The logic of relying on relatives is depicted in Figure 3.3 by repeating the

game in Figure 3.2 indefinitely many times. Each repetition is called a
“round” and the rounds are numbered consecutively. In each round, the
innovator can cooperate or appropriate. The financier (possibly a rich aunt)
decides to reward cooperation and punish appropriation. This strategy by
the financier gives the payoffs to the innovator as depicted in Figure 3.3. If
the innovator cooperates in round t, then the financier cooperates with him

3.5
Ver. 1.4 - Chapter 3: The Double Trust Dilemma

in the next round, and likewise for subsequent rounds.4 With cooperation,
the innovator receives 5 in each round of the game. Alternatively, if the inno-
vator appropriates 15 in any round, then the financier refuses future dealings
with him, so the innovator will have to develop his ideas without capital in
each subsequent round and earn 1.

Figure 3.3. Innovator’s Payoffs in Repeated Game

round t t+1 t+2 t+3 t+…
innovator’s cooperate 5 5 5 5 5
choice appropriate 15 1 1 1 1

Compare the payoffs from the two strategies. In round t, “appropriate” pays
the innovator 15 and the innovator gains 1 in each subsequent round. In
contrast, “cooperate” pays 5 in every round. Summing payoffs from round
t to round t+3, “cooperate” pays 20 and “appropriate” pays 19. Thus “coop-
erate” overtakes “appropriate” in round t+3, and the gap widens with each
subsequent round. A rational player who values future payoffs will cooperate
and not appropriate.5 Conversely, a rational player who highly discounts
future payoffs will appropriate rather than cooperate. Specifically, the inno-
vator in Figure 3.3 will cooperate who discounts payoffs by less than 40%
per round, and the innovator will appropriate who discounts payoffs by more
than 40% per round.6

4  The financier’s strategy is called “grim” or “grim trigger.” Once the innovator cheats the finan-
cier treats him as untrustworthy forever. Another possible strategy with similar implications for
cooperation is a variant of “tit-for-tat, in which the financier punishes an innovator for cheating and
then resumes cooperating. The payoff matrix looks like this:
round t t+1 t+2 t+3 t+3 t+5 t+6 t+…
appropriate 15 1 1 1 15 1 1 1
Experimental evidence indicates that tit for tat comes close to maximizing the party’s payoffs in a
variety of circumstances.
5  To be precise, the discounted sum across rounds is the game’s “present value” to the player. The
present value of “cooperate” exceeds the present value of “appropriate” if the game repeats itself
enough times and the discount.
6  Cooperation yields

and appropriation yields

Setting them equal yields 5+ = 15+5{1/n)}, whose solution is n=.4.

3.6
 Robert Cooter

Private Law
Family members can cooperate even if law cannot effectively enforce their
promises to each other. However, entrepreneurs do not have enough family
and friends to conduct business exclusively with them on the most profitable
scale. Successful entrepreneurs look beyond relationships to strangers for
money and ideas. Whereas kinship endures, strangers come and go. In busi-
ness, friendships are often chosen for particular needs, not ascribed. The pos-
sibility that a business relationship will end poses an obstacle to cooperation.

In Figure 3.2, cooperation continues forever unless the innovator behaves

badly and appropriates in one round. Even without bad behavior, the pos-
sibility that a relationship will end on its own can prevent cooperation. We
already explained that solving the problem of cooperation by infinite repeti-
tion requires the actors not to discount future payoffs too much. If the game
is likely to end, the players may discount future payoffs for uncertainty too
much to cooperate in the present. Thus when communism began to collapse
in Russia around 1990, production plummeted partly because producers
could not cooperate when they lost confidence in future dealings with each
other.

Whereas a possible end to a relationship increases the problem of coopera-

tion, a definite end to a relationship makes cooperation even more problem-
atic. Instead of assuming that the game in Figure 3.3 is repeated indefinitely,
assume that t+3 is the game’s last round in Figure 3.3. As before, the finan-
cier adopts the strategy of rewarding cooperation and punishing appropri-
ation, but the financier has no opportunity to punish appropriation that
occurs in the last round. Therefore, the rational innovator will appropriate
in round t+3 without fear of retaliation. “Cooperate until the last round
and then appropriate” pays the innovator more than “always cooperate,” as
shown in Figure 3.4.

3.7
Ver. 1.4 - Chapter 3: The Double Trust Dilemma

Figure 3.4. Innovator’s Payoffs in Repeated Game

Where Financier Cooperates Until the Innovator Appropriates
round t t+1 t+2 t+3
always cooperate 5 5 5 5
cooperate until the last round and then appropriate 5 5 5 15

The payoffs in Figure 3.4, however, do not reflect what each player will do if
he foresees what the other player will do (“Nash equilibrium”). If the finan-
cier foresees that the innovator will appropriate in round t+3, she will change
her strategy and not cooperate in round t+3. If the innovator foresees the
financier’s behavior in round t+3, he will change his strategy and appropriate
in round t+2. If the financier foresees the innovator’s behavior in round t+2,
she will change her strategy and not cooperate in round t+2. By repeating
this reasoning, cooperation unwinds all the way to the game’s beginning. 7
With a definite end in sight, repeating the game provides no basis for cooper-
ation. Every round looks like Figure 3.2, where the innovator and financier
do not launch a joint venture. This is the “end-game” problem.

In a game with a definite end, cooperation may unwind back to the begin-
ning. In a game with a likely end, cooperation may stop if uncertainty causes
high discounting of future payoffs. Unlike relatives, strangers have more dif-
ficulty solving the problem of non-cooperation by repeating the double trust
dilemma because they foresee an end to their relationship, or they discount
future payoffs by the probability that the relationship will end. If the inno-
vator and financier cannot cooperate, the business venture does not launch,
the innovation is undeveloped, and the economy does not grow.

Where repetition fails, private law may succeed in solving the double trust
dilemma. Here is one legal implementation often used in business ventures
like Figure 3.1. The innovator and financier form a corporation to develop
the idea. The financier loans the firm 3 for development (debt), and the
innovator contributes his expertise to the firm. The firm issues 7 shares (eq-
uity) to the innovator and 5 shares (equity) to the financier. Shares are claims
on profits left over after paying the firm’s debt. To implement the proposal,
the law must effectively enforce the firm’s repayment of debt and an equal
7  Experimental data shows that people are not so rational. In a repeated game with a definite
end, experiments often find that players cooperate in the beginning and they stop cooperating near
the end. Apparently people are partly logical and partly psychological.

3.8
 Robert Cooter

distribution of profits per share. If law is effective, the corporation earns 15

from developing the innovation, repays its debts of 3 to the financier, and
distributes dividends of 5 to the innovator and 7 to the financier, as depicted
in Figure 3.1.

This implementation requires effective contract law to enforce repayment of

the firm’s debts and effective corporate law to assure that profits get distribut-
ed equally to shares. The framework for finance by a small group of investors
mostly comes from the law of contracts that belongs to “private law,” and
the law of business organizations that belongs to “business law.” Ineffective
contract or corporate law prompts chicanery, not cooperation.

Here is one way that weak corporate law could allow the innovator to ap-
propriate all profits. Assume that the innovator is the corporation’s chief
executive officer (CEO) who controls the company without effective legal
constraints. After the firm earns 15, the innovator/CEO repays the firm’s
debts of 3 to the financier, and then the innovator/CEO pays himself a salary
of 12. Since all 15 go to pay debts and salary, no profits remain to distribute
as dividends to shareholders. The financier gets nothing beyond the repay-
ment of her debt of 3, and the innovator gets everything else. If the financier
foresees this ploy by the entrepreneur, cooperation will unwind.8 (Similarly,
another ploy transfers all profits to the financier instead of the innovator,
which also causes cooperation to unwind.9)

Public Law
In private finance of a business venture, the innovator and financier nego-
tiate with each other over price and non-price terms and then use contracts
and corporate law to implement their plans. In contrast, traders in public
markets like the New York Stock exchange buy and sell stocks and bonds at
8  In stage 2, the financier foresees that lending capital of 3 for the firm to develop the innovation
will result in repayment of the loan and nothing more for her. Foreseeing this fact, the financier
will choose to develop the idea without the innovator in stage 2, which pays the financier 5 and the
innovator 0. Foreseeing this fact, the innovator will choose to develop the idea at stage 1 and earn 1,
rather than explaining the idea to the financier and earning 0. Unless law can prevent the innovator’s
chicanery at the end of stage 2, cooperation will unwind.
9  Assume that the financier is the Chairman of the Board who controls the company without ef-
fective legal constraints. After the firm develops the idea, the financier/chairman fires the innovator/
CEO and sells the innovation for 15 to another firm that she owns. The selling firm uses 3 to repay
its debts, and the buying firm distributes its profits of 12 to the financier/owner.

3.9
Ver. 1.4 - Chapter 3: The Double Trust Dilemma

prevailing prices, without bargaining or negotiating. The stock exchange

approximates the economist’s ideal of perfectly competitive markets. In
perfectly competitive markets, the prices are public (the exchanges post the
prices at which stocks are selling), everyone pays the same price, and the
goods are standardized (one common share of Exxon is the same as another).

In developing innovations, relational and private transactions play the pri-

mary role, and public markets for stocks and bonds play a supplementary
role. Supplementary by public finance especially occurs as a business venture
matures. Thus assume that the innovator and financier in Figure 3.1 reach
a private agreement to develop the innovation. To enjoy extraordinary prof-
its, they must keep many details of their venture secret from competitors.
Developing the innovation in stage 2 creates a temporary advantage over
competitors, which might come from patenting the innovation or marketing
it first. Patenting or marketing reveals the innovation to the public. At the
venture’s end, the firm gets 15 by selling its stock in an “initial public offer-
ing” or by allowing a public company to acquire it.

The ability to sell securities to the public presupposes compliance with secu-
rities regulations that ideally protect buyers from chicanery. Members of the
public who buy stocks or bonds have no direct control over how the firm uses
their money. Instead, the firm’s managers and board of directors control it.
Insiders have many opportunities to appropriate outsiders’ investments. For
example, insiders may use accounting tricks to convert profits into salaries,
thus depriving stockholders of their dividends. Protecting outsiders from
insiders in public companies requires more than securing property and en-
forcing contracts. For public finance, the additional protection comes espe-
cially from corporate law and the law of securities. These laws ideally assure
that outside investors get their share of a firm’s profits, rather than insiders
appropriating all of it. Conversely, if these laws are ineffective, insiders will
appropriate all of a firm’s profits, and outsiders who foresee this fact will be
unwilling to invest.

If the firm gets 15 by selling itself to another firm or by selling its stock in an
initial public offering, the founders often take the cash from the initial public
offering and exit the business soon thereafter. A timely exit by the founders
makes a business venture more profitable, which increases innovative activ-

3.10
 Robert Cooter

ity.10 Profits increase because innovating and producing commodities are

different specialties requiring different talents. By exiting, the founders leave
others to produce commodities while they pursue the next innovation.

As explained in Chapter 2, a successful business venture begins with a dis-

equilibrium caused by developing a novel idea. To gain a decisive advantage
over competitors, development requires secrecy in order, which precludes
public markets. After development builds a decisive advantage, the need for
secrecy diminishes. The innovator has fewer secrets to hide and the double
trust dilemma attenuates. In these circumstances, the firm may conduct an
initial public offering of its stock or sell itself to a public company. Public
finance approaches the economist’s ideal of a perfectly competitive equilib-
rium, where no one has valuable private information and everyone earns the
same profit rate (“ordinary rate of return”). Like contentment, a perfectly
competitive equilibrium is approached and never reached. Perfect competi-
tion is the end of an innovation, its irresistible and final fate, not the process
of its creation.

Three Stages of Finance in Silicon Valley11

Relational transactions require effective property law, private transactions re-
quire effective contract law and corporate law, and public transactions require
securities regulations. The three stages of finance in Silicon Valley illustrate
how these three types of laws support innovation. Relationships are espe-
cially important for many startup firms in Silicon Valley. First, someone has
a new idea and obtains capital to develop it. At this point, the innovation’s
economic value has not been established. To establish its value, the innovator
needs funds from a few “angel investors.” Family and friends have confi-
dence in the innovator invest, so they invest without fully understanding the
innovation’s market value. Perhaps a few people invest who think that they
can evaluate the innovation without understanding it. According to a pop-
ular quip, initial funding comes from “the 3 Fs”: Family, friends, and fools.

10  For a recent application, see Bernard Black and Ronald Gilson, “Does Venture Capital Require
an Active Stock Market?” Journal of Applied Corporate Finance 11 (winter 1999): 35–38.
11  The three stages of finance in Silicon Valley are distinguished and related to stages in the de-
velopment of financial markets in The phrase “double trust dilemma” was introduced by Cooter and
Schaefer in Chapter 3, “The Double Trust Dilemma of Development,” Solomon’s Knot: How Law
Can End the Poverty of Nations. Contemporary (Princeton University Press, 2012).

3.11
Ver. 1.4 - Chapter 3: The Double Trust Dilemma

In the first stage, the entrepreneur develops the innovation sufficiently to

prove its value to knowledgeable strangers. Most innovators, however, have
too few personal relationships with wealthy people to finance an innova-
tion’s full development. They must eventually turn to financiers. In Silicon
Valley, the second stage of funding comes from “venture capitalists.” They
are experts in evaluating undeveloped innovations, but they are not family,
friends, or fools. Venture capital is private finance because it comes from a
small group of investors with information unknown to a broad public. Most
countries impose different laws on finance by a small group of investors than
on finance by many investors.

Innovators and entrepreneurs have goods reasons to distrust each other. Sil-
icon Valley innovators sometimes expropriate the investments of their finan-
ciers. John P. Rogers convinced some prominent California investors to give
him $330 million for a high-tech start-up named Pay By Touch that would
“transform how America pays its bills” by using “biometric authentication
technology” (e.g., fingerprints). In 2008 the company went bankrupt, and
investors contend in lawsuits that Rogers burned through $8 million per
month without producing anything of value.12

Similarly, the creative people who found a company often manage it badly.
When the founders prove to be bad managers, the venture capitalists must
replace them with good managers. In these circumstances, the venture cap-
italists seize the firm to increase its profitability. Alternatively, where the
founders prove to be competent managers, venture capitalists may seize the
firm to avoid sharing profits with the founders. Venture capitalists sometimes
want to remove good managers with large claims to the firm’s future profits.
The initials “v.c.” stand for “venture capitalists” and also “vulture capitalists.”

Innovators and venture capitalists use various legal devices to overcome their
mutual distrust. The founders of the firm commit to performance goals
implicitly or explicitly. If they fail to meet their goals, they may lose their
investments and their jobs. Specifically, the venture capitalists often hold
preferred shares of stock, unlike the common shares held by the founders.
Preferred shareholders may have special powers of governance that common
shareholders lack, as well as other advantages. The financing contract may say
12  Lance Williams, “How a ‘Visionary’ Raised—and Lost—a Fortune,” San Francisco Chroni-
cle, December 7, 2008, available at http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2008/12/07/
MNIK137QU3.DTL.

3.12
 Robert Cooter

that preferred shareholders can demand repayment of their investment after

three years. Such a contract reassures the venture capitalists that the founders
will do their utmost to perform as promised. The contract also reassures the
founders that the venture capitalists have an interest in keeping the firm’s
secrets. 13

Corporate governance provides another device to solve the double trust prob-
lem in Silicon Valley. The firm’s bylaws may stipulate that common share-
holders (founders) and preferred shareholders (venture capitalists) appoint
an equal number of directors to the company’s board, plus an independent
director accepted by both sides. If the founders and venture capitalists dis-
agree, the independent director holds the decisive vote. Thus the indepen-
dent director will decide whether or not the venture capitalists can replace
the founders with new management.

In the third stage of finance in Silicon Valley, a successful start-up sells itself
to the public, either directly through an initial public offering of its stock,
or indirectly when a publicly traded company acquires it. In order to sell
stock to the public in the United States, a firm must comply with disclosure
rules of the Securities Exchange Commission. Brokers disseminate the firm’s
disclosed information to potential investors. After disclosures, many people
understand the innovation sufficiently to decide whether or not to invest in
its further development. The investors in stock markets are a large group of
people – the “public.” Regulatory law, which belongs to “public law,” espe-
cially controls finance by a large group of investors.

Small and Large Firms

The double trust dilemma afflicts individual innovators and financiers. Does
it also afflict large firms? In a large firm, the innovators and financiers are
employed by one organization. They proceed something like this. Exec-
utives sift through proposals from the research divisions, choose the most
13  Instead of issuing preferred shares to venture capitalists, firms could create much the same
result by selling them bonds and common shares. However, preferred shares have a tax advantage.
The valuation of common shares affects the tax liability of managers and employees who are partly
paid in common shares or options to buy them. Since preferred shares are different from common
shares, selling preferred shares to venture capitalists does not necessarily affect the price at which the
firm can value its common shares for tax purposes. The firm can raise money at the relatively high of
preferred shares while claiming that common shares are worth much less.

3.13
Ver. 1.4 - Chapter 3: The Double Trust Dilemma

promising ones, and order the finance division to pay the research division to
develop the innovations. Perhaps you think that hierarchical authority solves
the problems of trust that afflict negotiations between individual innovators
and financiers.

If that is what you think, consider the following. When an employee has a
new idea, he understands its worth better than others. If he gives a valuable
idea to his employer, he may get a bonus and the firm may get richer. A typi-
cal employee, however, would rather enrich himself than his firm. Instead of
giving the idea to his firm, the employee may quit, form a startup company,
develop the idea on his own, and try to get rich. Various contracts and laws
restrain employees from quitting firms and taking creative ideas with them.14
The more large firms can claim legal ownership of employees’ discoveries, the
less people who are creative will work for them. Large firms often cannot
prevent employees from quitting and taking creative ideas with them, and
large firms often cannot hire the most creative innovators to work for them.
That is why many innovations in Silicon Valley begin in small, startup firms,
where innovators own a significant share of the firm.

The double trust dilemma persists in large firms. The employee in the re-
search division asks, “Can I trust the firm to reward me if I give it my in-
novative idea?” The employee in the finance division asks, “Can I trust the
research division to develop this innovative idea as planned?” Each one fears
that the other party is withholding crucial information. Perhaps the inno-
vator-employee keeps secret a potentially fatal obstacle to development, or
perhaps the financier-employee keeps secret his intention not to pay a bonus
to the innovator-employee.

The problem of trust favors innovation in small firms. Why do large firms
innovate? The amount of money required to develop an innovation often ex-
ceeds the innovator’s wealth. The innovator is reluctant to risk all his wealth
on an innovation with an uncertain future. In a startup firm, the innovator
shifts much of the financial risk to the financier. If they form a business
venture with the innovator in charge, the innovator risks another’s money,
not his own. This fact creates incentive problems that we discuss later. To
improve incentives, the innovator usually retains substantial risk of failure.

14  See Chapter __ trade secrets law and Chapter __ on non-competition clauses in employment
contracts.

3.14
 Robert Cooter

Compared to a startup firm, the innovator-employee in a large firm shifts

almost all of the risk of financial failure to the employer.

Creative people differ in their appetite for risk taking. Creative people who
are risk-lover form startup firms, and creative people who are risk-averse find
employment in the research division of large firms. The main advantage of
developing innovations in large firms is risk spreading. Different tastes for
risk contribute to the distribution of innovations by firm size.

Ontogeny Recapitulates Philogeny

Firms in Silicon Valley often pass through all three stages of finance – rela-
tional, private, and public. Similarly, the three stages of finance for a start-up
firm in Silicon Valley resemble three stages of historical evolution in capi-
tal markets for countries. The industrial revolution in England, which was
the world’s first, went through these stages. In the early eighteenth centu-
ry, inventors mostly relied on their personal assets and loans from family
and friends (relational finance). As industrialization proceeded, loans from
wealthy investors and banks became available more readily to new industries.
Finance of industrial companies by sales of stocks and bonds to the general
public came later. Public financing of industrial companies originally con-
cerned infrastructure like canals, docks, and railways, where private business
and the state intertwine. As the law became more reliable, public finance
spread to manufacturing firms.15

Like 18th century England, the poorest countries today have weak capital
markets, so businessmen mostly borrow from family and friends. Starting
from a condition of lawlessness, imposition of secure property rights can
cause a spurt of growth based mostly on relational finance, as in China’s new
industries after the 1980s. Some peoples, notably the Chinese and the Jews,
have family networks that extend business relationships beyond the usual
boundaries. However, the conditions of trust among relatives do not reach
the scale of modern businesses. Relational finance keeps business small and
local. No modern country became wealthy by relying exclusively on relation-
al finance.

15  See P. M. Deane, The First Industrial Revolution (Cambridge: Cambridge University Press,
1955), summary on 155–157. For financing of key eighteenth-century inventions by name, see 155.

3.15
Ver. 1.4 - Chapter 3: The Double Trust Dilemma

To increase the scale of business, an economy must augment relational fi-

nance with private finance, especially bank loans. In countries where banks
dominate, an elite of wealthy insiders often lend to business ventures based
on private information. Thus bank finance in some developing countries
performs a similar role to venture finance in Silicon Valley. As countries
become affluent, they increasingly augment private finance with public fi-
nance, which means selling stocks and bonds to the general public. Stocks
and bonds compete with banks and wealthy individuals to finance economic
growth.

Compared to debt, equity financing requires a superior legal framework.

Debt involves a fixed schedule of repayment specified in a contract, whereas
equity involves sharing future profits. Enforcing a fixed schedule of repay-
ment is easier for courts than enforcing profit sharing. Consequently, loans
and bonds often flourish in countries where stock markets languish. When
weak law shrinks local stock markets, businesses are deprived of equity capi-
tal needed to launch innovative ventures.

Biologists sometimes say, “Ontogeny recapitulates phylogeny,” which means

that a single organism’s development from conception to maturity resem-
bles an entire specie’s evolution16.17 The same applies to solving the double
trust dilemma. Startup companies in Silicon Valley and all companies in
lawless countries tend to rely on relationships more than formal law to solve
the double trust dilemma. With better law, finance expands from relational
to private, and from private to public. The expansion supplements earlier
forms without replacing them. All three forms of finance—relational, pri-
vate and public—remain important in the richest countries. The extent of
public finance varies significantly among countries, including rich countries.
Japan and northern Italy have achieved affluence mostly through relational
and private finance, with relatively little public finance, whereas the United

16  See Chapter 8, pages 115-120 of Cooter and Schaefer, “The Double Trust Dilemma of Devel-
opment,” Solomon’s Knot: How Law Can End the Poverty of Nations. Contemporary (Princeton
University Press, 2012)
17  Each animal begins life as a single cell containing genetic instructions for how to grow into a
complex organism. For animals in different species with a common evolutionary ancestor, the path
of individual growth suggests the older, evolutionary forms found in the fossil record. While the pat-
tern of individual growth does not strictly recapitulate the evolution of the species, comparing them
provides useful clues about the genes that control the development of individuals and species. For a
book that finds the origin of the human fetus in fish, see Neil Shubin, Your Inner Fish: A Journey
into the 3.5-Billion-Year History of the Human Body (New York: Pantheon Books, 2008).

3.16
 Robert Cooter

States and Great Britain rely mostly on public finance for mature industries.
Germany appears to be shifting from the former to the latter.18

Expanding the basis of finance requires effective law that controls behavior,
not aspirational law that expresses lofty ideals. What makes a law effective?
Not just writing it down. Written law in a poor country often resembles
written law in a rich country. Property and contract law-on-the-books in
India and Nigeria resemble English common law, and property and contract
law-on-the-books in Peru resemble the Spanish civil code. Writing down a
law, however, does not make it effective. The written laws are less effective in
Nigeria or Peru than in England or Spain.

Conclusion
New business ventures begin with secrecy, risk, and high profit expectations.
All three decrease as the venture matures. Sea routes from Europe to Asia
were eventually mapped and secured, trade between them became common-
place, and middle-class Europeans could buy spices. In Silicon Valley, com-
petitors work around patents and ferret out secrets, thus converting today’s
technological breakthroughs into tomorrow’s commodities. However, an in-
novative economy never settles into a permanent condition without secrecy,
risk, or extraordinary profits. For economic growth, new business ventures
must repeatedly confront and solve the double trust dilemma. To solve the
double trust dilemma, entrepreneurs, financiers, courts, and regulators have
worked out contracts, laws, and institutions over years and centuries. One
cannot understand these creations without studying them as legal scholars
do. That is how law dispels some mysteries of growth that baffle economists.

Where law is weak, innovators and financiers cannot rely on formal law for
much more than protecting property from predators. They must deal mostly
through relationships. This is true for most business in some undeveloped
countries, and it is true for startup firms in developed countries where no law
can protect fragile secrets and implicit financial agreements. In developing
countries and startup firms, good law cannot do all of the work of creating
trust. When the effective law of contracts and corporations improves, fi-
nance expands from relational to private. This is true in undeveloped coun-
tries that improve their legal institutions, and it is true in startup firms after
18  For details on finance in different countries, see...

3.17
Ver. 1.4 - Chapter 3: The Double Trust Dilemma

they can demonstrate an innovation’s promise. When institutions improve

further in underdeveloped countries, or when a startup firm develops an
innovation sufficiently to acquire a decisive advantage against competitors,
more effective securities laws allows, finance to expands from private to pub-
lic. In general, improved law allows more ideas to combine with more capital
so the economy grows faster.

3.18
 Robert Cooter

Chapter 4. Separation
Robert Cooter and Aaron Edlin

When a successful business venture develops an innovation, its launch yields

high profits temporarily, then profits fall towards zero as the innovation dis-
seminates and competition intensifies. Over its lifetime, the venture’s profits
equal a fraction of the innovation’s social value – usually a small fraction
– and the rest goes to other firms as profits and to consumers as “surplus.”1
Since the social value of an innovation far exceeds the innovator’s profits,
venture profits provide deficient incentives to innovate.

To increase innovation, law and policy should increase its profitability. Fig-
ure 4.1 expresses this idea. (It is essentially the same as Figure 2.4 in Chapter
2.). The vertical axis represents venture profits, which equal the discounted
present value of the stream of revenues from the innovation’s sales, minus
the discounted present value of its development and production costs. The
horizontal axis arrays ventures by profitability from high on the left to low
on the right. With open competition, investors will finance ventures until
profits fall to zero. Development stops when the present discounted value of
revenues equals the present discounted value of development and production
costs.

Figure 4.1 contrasts venture profits under two different legal regimes. Un-
der the original law, venture profits reach 0 at innovation I. Improved law
changes the situation by shifting the venture profit curve up as indicated by
the arrows, so venture profits reach 0 at innovation I*. Thus improved law
increases the number of innovations that get developed from I to I*. The fig-
ure depicts the increase in one period of time. If better law increases innova-
tions in every future period of time, then the sustained growth rate increases.
(Figure 4.1 depicts one of several ways that improved law increases growth by
increasing the profitability of innovation.2)

1  By definition, the consumer’s surplus equals the difference between the price that a consumer
would be willing to pay for a good and the price that he actually pays.

2  Figure 4.1 does not depict the speed with which an innovation is developed. In a race to in-
novate, the expectation of higher profits may cause an entrepreneur to develop an innovation faster.
Faster development affects infra-marginal innovations, not just the marginal innovation I*.

4.1
Ver. 1.4 - Chapter 4. Separation

Figure 4.1. Laws that Increase Venture Profits Cause More Innovations

How much should law ideally increase the profits of innovative ventures?
To maximize social welfare, according to the overtaking principle, law and
policy should increase venture profits as much as possible. You might doubt
the overtaking principle or the conditions that make it true, but when you
consider the historical record you cannot doubt that sustained growth in a
comprehensive measure of consumption dramatically improves social wel-
fare. Nor can you doubt that faster sustained growth would increase social
welfare faster. In any case, this book is playing the intellectual game of law
and growth economics whose objective is to maximize sustained growth.

Paradox of Growth
Law and policy rests on a paradox, like “This statement is false.” The paradox
of growth is that rapid innovation requires competition for extraordinary
profits. This is a paradox because “perfect competition” in economic theory
implies that all firms earn ordinary profits. If some firms earn extraordinary
profits, competition must be imperfect.

How can law and policy combine intense competition and extraordinary
profits for innovators? Patent law illustrates the solution. Chapter 2 distin-
guished a venture’s phases into development, launch, and imitation. Patent

4.2
 Robert Cooter

law opens competition to develop patentable innovations, closes competi-

tion when development succeeds and the inventor gets a patent, and restores
competition eventually when the patent expires.3 The formula for maximiz-
ing growth is open competition to innovate plus temporary market power
for successful innovators.

With open competition to innovate, everyone is free to launch a venture

and develop an innovation. Firms compete to launch business ventures and
earn extraordinary profits. The development of innovations is very risky, and
entrepreneurs take the risk in the hope of earning extraordinary profits. Mo-
nopoly is more profitable than competition. With higher profits, more ven-
tures launch. Ventures launch until the marginal venture expects to earn zero
profits, as depicted in Figure 4.1. Market power for innovators causes more
competition to innovate.

Subsequent chapters in this book apply this formula to different bodies of

law. Besides patents, many other kinds of law affect a business venture –
property, contracts, torts, corporations, finance, regulations, antitrust, and
taxation, to name a few. They should have the same aim as patents and use
the same formula to combine competition and market power. The aim is to
maximize the profitability of ventures in open competition to develop inno-
vations. The gain from growth in the long run will overtake the loss from
higher prices in the short run.

This is not the conventional approach in the economic analysis of law. The
conventional approach first observes that a venture incurs losses in develop-
ment, so overall profitability requires extra profits after development. How
much extra? The usual words are “sufficient,” “adequate,” “cost-recovery,” or
“break-even.” Second, the conventional approach notes that market power
for a firm increases the price of its product to consumers. The words are
“excessive,” “inefficient,” or “distorting.” Third, the conventional logic con-
cludes that law affecting market power should balance the gain from faster
innovation and the loss from higher prices for consumers. The words are
“tradeoff” and “optimal.” Unlike the conventional approach, our words are
“maximize innovation.”

3  A patent cycle begins with competition for a market, and ends with competition in a market.
The phrase “competition for a market” was used by Bill Baxter to analyze antitrust law.

4.3
Ver. 1.4 - Chapter 4. Separation

Producer’s Market Power

In microeconomics, perfect competition and monopoly are extremes on a
continuum of power to set a good’s price. At the extreme of perfect compe-
tition, each seller must take the market price, not set a price. No seller can
charge more than the market price because ach buyer has so many alternative
sellers. The demand curve is perfectly elastic. Furthermore, competition
among sellers drives the market price down to the cost of production, so
each seller receives ordinary profits. Moving along the continuum, as mar-
ket structure becomes less competitive, a seller can set a price. Each buyer
has imperfect alternatives among the sellers. The seller who charges a higher
price will make fewer sales, not zero sales as in perfect competition. The de-
mand curve is imperfectly elastic. Moving farther along the continuum, as
the seller’s market power increases, buyers have fewer choices. At the extreme
of monopoly, there is only one seller. Everyone who wants the good must
buy from the monopolist. The monopolist sets the price above the cost of
production and earns extraordinary profits.

When a firm uses market power to increase its price, part of the buyers’ losses
transfers to the seller as higher profits, and part of the buyers’ losses does not
transfer to the seller. “Deadweight” describes a loss without an offsetting
gain.4 To illustrate numerically, assume that a seller earns 0 profits under
perfect competition and 100 under monopoly. A change from competition
to monopoly transfers 100 from buyers to the seller. According to standard
economics, the seller gains less from monopoly than the buyers lose. Thus
if monopoly causes the seller to gain 100, buyers may lose, say, 130. The
difference of 30 is the deadweight loss.5 (Behind this problem of monopoly
lies a deeper theoretical problem of transaction costs.6) Firms with market
4  The deadweight loss occurs because the higher price causes buyers to purchase less of the good.
When buyers forego some purchases, the buyers lose the gains that they would have enjoyed from
these foregone goods, and the seller gains nothing from goods that buyers do not buy.
5  The standard analysis is somewhat more complicated than these remarks suggest because of the
difference between long run and short run effects. In the long run, all of the deadweight loss falls
on buyers, but in the short run sellers can also bear part of the deadweight loss. Note that in the
standard graphical analysis, the transfer is a rectangle and the deadweight burden is a triangle.
6  If there were no obstacles to bargaining, the monopolist and the buyers could write contracts
that eliminate the deadweight lost. The contracts would charge according to a price schedule, not a
single price. The schedule discriminates according to how many units of the good the buyer buys,
charging a higher price for the first unit and a lower price for the last unit. By definition, a perfectly
discriminating monopoly charges the reservation price to each buyer, so the marginal rate of substi-
tution equals the marginal rate of transformation as required for static efficiency. With zero transac-
tion costs, perfect discrimination eliminates the deadweight loss of monopoly. The deadweight loss

4.4
 Robert Cooter

power can raise prices to buyers and enjoy extraordinary profits. Static eco-
nomics condemns market power as inefficient because the deadweight loss
reduces wealth.

In contrast, growth economics condemns market power when it slows

growth, and absolves market power when it increases growth. When does
market power retard innovation, and when does it enhance innovation? As
we explain, the identity of the buyers determines whether market power re-
tards or enhances innovation.

Market power slows growth when a producer who does not innovate raises
prices to innovators, so the cost of developing innovations increases and the
profitability of innovation decreases. When profitability decreases, the mar-
ginal venture in Figure 4.1 shifts to the left as depicted by a move from I* to
I. To prevent this slowing of innovation and growth, law and policy should
suppress market power by non-innovating producers against innovators.

Besides raising the cost of developing innovations, market power of non-in-

novating producers has another bad effect. Prices guide entrepreneurs in
deciding which innovations to develop. Market power of non-innovating
producers has the bad effect of aiming innovation at the wrong target. Inno-
vators develop substitutes that are no better than the original good, instead
of developing better goods. We call this distortion of inventive activity the
“mistargetting effect”.

To illustrate, a cartel that controls diamond mines and raises diamond prices
spurs the development of manufactured diamonds. Manufacturers will prof-
it from developing diamonds so long as the cost does not exceed the cartel’s
price. Since mined and manufactured diamonds virtually indistinguishable,
the only gain to society from manufacturing diamonds is to undermine the
cartel’s price. The cartel’s price could be undermined directly by laws and
policies aimed at reducing the market power of the diamond monopoly,
without the cost of developing manufactured diamonds.

To illustrate abstractly, if the competitive price of good α is more than good

β, then society values α more than β. However, the producers of β might
have more market power than the producers of α, so the price of β might
occurs because transaction costs makes perfect price discrimination impossible. Chapter 5 discusses
these theoretical claims at length.

4.5
Ver. 1.4 - Chapter 4. Separation

exceed the price of α. Since the price of β exceeds the price of α, innovators
might foresee higher profits from developing substitutes for β rather than α,
even though society benefits more from the opposite.

When market power raises a good’s price, more innovators try to invent sub-
stitutes for it. The invention of a perfect substitute is costly. However, a per-
fect substitute is no better than the original good. The only consequence of
inventing the substitute is to reduce the distortion from monopoly pricing.
Law and policy could reduce market power without the expense of develop-
ing the substitute.

Economists sometimes make the same point in different language. By defini-

tion, a “rent” is income from owning something that is scarce. Compared to
perfect competition, market power makes a good relatively scarce. Thus mo-
nopoly profits are “rents” from having market power. High prices charged
by producers with market power causes innovators to invent substitutes. In-
venting substitutes for over-priced goods enables the inventors to get a share
of the rents for themselves. Thus market power causes innovators to “seek
rents” instead of inventing better products.

We have explained two reasons why market power of non-innovating pro-

ducers against innovators retards growth. The first and most important rea-
son is the transfer of profits from innovators to non-innovators. The sec-
ond and less important reason is mistargetting innovation. This analysis in
growth economics reinforces the usual condemnation of producer’s monop-
oly in static economics.

Innovator’s Market Power

If market power is condemned for non-innovating producers, what about
market power for innovators? Does market power for innovators enhance or
inhibit growth? Either effect is possible. This chapter describes the form of
market power among innovators that usually enhances growth, and the next
chapter describes the form that usually inhibits growth. The difference de-
pends on who buys from the innovator. Divide buyers into non-innovating
producers, consumers, and innovators. When the buyers are non-innovating
producers and consumers, the innovator’s market power increases the inno-

4.6
 Robert Cooter

vations price, which transfers wealth from producers and consumers to the
innovator. The transfer increases the average profits of innovative ventures,
which causes the marginal venture in Figure 4.1 to shift to the right as de-
picted by the move from I to I*. More resources are devoted to innovation
and fewer resources are devoted to production and consumption. Thus mar-
ket power for innovators against producers and consumers enhances growth

Besides transferring wealth, the innovator’s monopoly imposes a deadweight

loss on producers and consumers. If maximizing growth is the aim, the
deadweight loss does not matter in the long run. The benefits of faster
growth from an innovator receiving monopoly profits overtake the costs of
the deadweight loss to consumers and producers (the “overtaking principle
in Chapter 1).

To illustrate by our numerical example, assume that an innovator’s monop-

oly imposes costs of 130 on non-innovating producers and consumers. 100
transfers to the innovator as profits and 30 is deadweight loss. The transfer of
100 increases the profitability and rate of innovation. Faster growth creates
long run benefits for producers and consumers that overtake the temporary
loss totaling 130.

Consider new pharmaceutical drugs. Stronger monopoly powers for phar-

maceutical innovators immediately increase the prices of drugs paid by con-
sumers. The higher profits of pharmaceutical innovators cause faster devel-
opment of better drugs. Consumers lose from higher drug prices and gain
more from better drugs. Thus relatively cheap ulcer drugs replaced relatively
expensive ulcer surgery.

Similarly, cotton manufacturers buy ever-improving looms. Patents on

looms increase the price of cotton textiles in the short run, but patents in-
crease innovations that lower prices in the long run.

The three effects of market power are transfer, deadweight loss, and mistar-
getting. Analyzing these effects on growth implies a simple conclusion for
law and policy. Growth economics proscribes market power for (non-in-
novating) production against innovation and consumption, and growth
economics prescribes market power for innovations against consuming and
(non-innovating) producing. In both cases, the aim is the same: to increase

4.7
Ver. 1.4 - Chapter 4. Separation

the profitability of innovative ventures. We have arrived at the separation

principle: law and policy should treat market power differently for producing
and innovating, proscribing it for producing without innovating, and pre-
scribing it for selling innovations for production and consumption.7

The separation principle can be explained in terms of externalities. Inno-

vations build on each other. The value of each innovation spills over to
other innovations, with large effects on growth. Each innovator contributes
to economic growth that benefits others. Making an innovation increases
growth whose benefits spill over from the innovator to others. In contrast,
consuming an innovation or using it in production does not necessarily caus-
es benefits to spill over to others. Thus innovation has larger external benefits
than consuming or producing without innovating. The separation principle
calls for transfers in favor of innovative activities with large external benefits.

Conclusion
In static economics, monopoly causes buyers to pay more than the marginal
cost of goods, so too little is consumed and produced. The remedy is to abol-
ish market power and to restore competition whenever possible. In growth
economics, however, market power is necessary to increase the profitability
of ventures competing to innovate. Crossing the boundary between static ef-
ficiency and growth changes economics, like crossing the border from France
to Germany changes languages.

Market power of sellers has three consequences affecting growth: transfer,

deadweight loss, and mistargeting. When (non-innovating) producers have
market power in the sale of goods to innovators, all three effects reduce
growth and human welfare. Conversely, when innovators have market pow-
er against non-innovating producers, the overall effect increases growth and
human welfare. The transfer of profits from static to dynamic activities in-
creases growth, in spite of mistargetting, and deadweight loss.

7  Note that the three effects of monopoly align in static and growth economics to reinforce the
proscription of producer monopolies against innovators. The three effects do not align with respect
to the prescription for innovator’s monopolies against producers. The prescription for innovator’s
monopolies against producers rests on the transfers to innovators overtaking deadweight loss and
mistargeting.

4.8
 Robert Cooter

These facts imply the separation principle: to maximize growth, law and pol-
icy should separate innovation and (non-innovating) production. Growth
economics proscribes market power for non-innovating production, and
it prescribes market power for innovating against consuming and produc-
ing. The next chapter will analyze the effect of market power by innovators
against each other.

4.9
 Robert Cooter

Chapter 5. Fertility
Robert Cooter and Aaron Edlin

The social value of an innovation far exceeds the innovator’s profits. To in-
crease innovation, law and policy should increase the profitability of business
ventures in open competition with each other. Profitability increases when
the venture that wins the competition enjoys market power. A successful
venture obtains temporary market power by innovating first. Imitators even-
tually enter the industry and erode the venture’s market power. Law and
policy that slows imitators will extend the innovator’s market power, thus in-
creasing the venture’s profits. Chapters 4 and 5 concern general principles for
extending the innovator’s temporary market power, and subsequent chapters
discuss specific legal instruments such as intellectual property, secrecy, and
collusion.

According to the separation principle in Chapter 4, law and policy should

enhance the market power of the creator of an innovation who sells it for
consumption or production. The aim is to transfer resources from consum-
ing and producing (the “static economy”) to innovating (the “dynamic econ-
omy”). As defined in microeconomics, market power is the power to set a
good’s price above the cost of production and to earn extraordinary profits.
Thus innovator B develops innovation β and sells it to consumer C, as de-
picted in Figure 5.1. According to the separation principle, law and policy
should enhance innovator B’s market power against consumer C. If B has
market power and raises the price of β, then the dynamic economy as rep-
resented by B extracts resources from the static economy as represented by
C, and the rate of innovation increases. The separation principle commends
lawmakers to enhance the market power of innovators against consumers
and non-innovating producers.

5.1
Ver. 1.4 - Chapter 5. Fertility

Figure 5.1. Three-Party Straight-Line Transaction

What about market power of innovators against each other? Besides selling
to consumers and producers, innovators sell to other innovators. Today’s
new pharmaceutical molecule is discovered from yesterday’s new molecule;
today’s new operating system is discovered from yesterday’s new operating
system; today’s new power cell is discovered from yesterday’s new power cell.1
The subject of this chapter is redistribution within the dynamic economy.
The question is, “Should law and policy enhance market power of some in-
novators against other innovators?” Thus in Figure 5.1, A creates innovation
α and sells it to innovator B, who uses α to create innovation β and to sell β
to consumer C.

When one innovator has market power in the sale of an innovation to anoth-
er innovator, the seller raises the price to the buyer and profits transfer from
one innovator to another. In Figure 5.1, assume that A has market power
against B, so A can capture some of B’s profits from sales to C. Market power
for A against B redistributes within the dynamic economic, whereas market
power of B against C redistributes from the static economy to the dynamic
economy.

According to the separation principle in Chapter 4, market power of A against

B should have different legal treatment than market power of B against C.
Market power of B against C transfers resources from consuming to innovat-
ing, which increases growth and improves human welfare. In contrast, mar-
ket power of A against B transfers resources from one innovator to another
innovator, which might increase or decrease growth.

To increase growth, A’s innovation α must be more fertile than innovator B’s
innovation β. An innovation is fertile like a breeding horse if it can be used
to create another innovation. In Figure 5.1, innovation α is used to create
1  Brian Arthur describes technology as “self-creating” (autopoietic). ”…if new technologies were
constructed from existing ones, then considered collectively, technology created itself.” W. Brian
Arthur, The Nature of Technology (2009), page 2.

5.2
 Robert Cooter

innovation β, so α is fertile. Fertile innovations are often called “fundamen-

tal,” “basic,” ”pioneering”, “general,” “first-generation,” ”rising technology,”
and “shoulders to stand on”.2  Unlike innovation α in Figure 5.1, innovation
β is sterile like a mule because it cannot be used to create another innova-
tion. Sterile innovations are often called “applied”, “derivative,” “specific,”
or “stand alone.” For purposes of economic analysis, one innovation is more
fertile than another by definition if the costless transfer of profits from the
latter to the former increases the growth rate. (Mathematical notation makes
this characteristic more precise.3)

A costless transfer occurs without loss. Imagine that oasis Y in the desert fills
a one liter container with ice cream and sends it across the hot sands to oasis
X. Costless transfer implies that X receives one liter of ice cream. None of
it melts. Conversely, a costly transfer between oases implies that some ice
cream melts and X receives less than one liter. Similarly, a costless transfer
from B to A in Figure 5.1 means that B’s loss from the transfer equals A’s
gain. The burden of the transfer on B equals the benefit of the transfer to
A. If law and policy can transfer profits costlessly, then maximizing growth
requires redistributing profits from less to more fertile innovations. Redis-
tribution should proceed until all developed innovations are equally fertile.

This fact can be explained in terms of spillovers. The value of a fertile innova-
tion spills over to other innovations with large effects on growth, whereas the
value of an sterile innovation spills over to consumers and non-innovating
producers with small effects on growth. To maximize growth, some prof-
its from selling sterile innovations to consumers and producers should pass
through to fertile innovations. To pass profits through, the seller of a fertile
innovation like α should have market power against the buyer who uses it for
sterile innovation like β. (A footnote contains an example.4)
2  cite Tim Bresnahan, General Purpose Technolgies).cite Scotchmer
3  In notation, let i denote an innovation. The winner of an open competition to make innovation
i enjoys profits denoted πi. The sustainable growth rate g is a function of the distribution of profits:
g=g(π0,π1,π2,…,πI). By definition, if j is “economically more fertile” than i, then a costless transfer of
profits δ from i to j increases the growth rate:

g(π0,π1,π2,…,πI) < g(π0,π1,π2,πi-δ,...,πj+δ,..πI).

The characteristics of an innovation that make it economically fertile concern the way one idea
prompts the discovery of other profitable ideas.
4  Innovator A in Figure 5.1 creates innovation α and sells it to B. Assume that A’s only income
comes from selling α to B. If law and market structure deny A market power, then A must sell α

5.3
Ver. 1.4 - Chapter 5. Fertility

When innovators buy and sell innovations to each other, however, the trans-
fer is usually costly, not costless. Market power imposes a deadweight loss on
most transactions, including transactions between innovators. In Figure 5.1,
market power for A might cause an increase in the price of α that benefits A
by, say, 100 and burdens B by, say, 130. The difference of 30 is the “dead-
weight loss” as explained in Chapter 4. Whereas the transfer from less fertile
innovations to more fertile innovations causes growth to increase, the dead-
weight loss causes growth to decrease. The net effect determines whether or
not the overall growth rate increases. These considerations imply a general
principle. To maximize growth, law and policy should transfer profits from
less to more fertile innovations until the increase in growth from higher fertility
equals the decrease from a heavier deadweight loss. This is the fertility principle.
(The mathematical expression is in a footnote.5)

Figure 5.1 can illustrate this principle. Recall that A must develop α in order
for B to develop β. If A develops α and sells it to B at the cost of production,
then A suffers a loss equal to the cost of developing α. Foreseeing this fact,
A will not develop α. Alternatively, if A develops α and uses market power
to sell it to B at more than the cost of development and production, then A
makes a profit. Foreseeing this fact, A will develop α. Thus market power
of fertile innovators can benefit all innovators. (A footnote illustrates with
numbers.6)
at its production cost, so A cannot recover α’s development costs. If A foresees this fact, it will not
develop α, which B needs to develop β. To induce A to create α, B could pass some profits back to A
obtained from selling β to C. To increase the rate of innovation, law should facilitate the pass-back
of profits from sterile innovators like B to fertile innovators like A.
5  In footnote 2, if j is a “economically more fertile” innovation than i, then a costless transfer of
profits δ from i to j increases the growth rate:

g(π0,π1,π2,…,πI) < g(π0,π1,π2,πi-δ,...,πj+δ,..πI).

However, an actual transfer caused by law and policy creates a deadweight loss. The burden on i is
δ, the transfer to j is γδ, and the deadweight loss is (1-γ)δ. At the optimum as given by the fertility
principle, the benefit from the transfer to a more fertile innovation exactly offsets the cost from the
deadweight loss, so the growth rate does not increase or decrease:

g(π0,π1,π2,…,πI) = g(π0,π1,π2,πi-δ,...,πj+γδ,..πI).

6  Assume that A invests 75 to develop α and A spends 25 to produce α. If B buys α at 25, which
is the cost of production, then A loses 75. Foreseeing these facts, A will not develop α. If A does not
develop α, then B cannot develop β, and their joint profits are 0.

5.4
 Robert Cooter

Instead of the fertility principle, court cases refer to other principles such
as the principle of cost recovery (innovators should enjoy enough profits to
recover their development costs) and the principle of proportionality (suc-
cessful innovators should receive profits in proportion to their development
costs). The next chapter discusses the connection between the fertility prin-
ciple and the principles actually invoked by courts.

Applying the Fertility Principle

What determines the size of the deadweight loss caused by market power? In
the standard economic analysis of market power, the deadweight loss increas-
es with the elasticity of demand for the good in question, holding constant
the monopolist’s profits. The appendix to this chapter contains a graphical
explanation of these facts. Here is a verbal explanation. When a seller uses
market power to increase a good’s price, buyers respond a lot if demand is
elastic, and buyers respond a little if demand is inelastic. A larger response
indicates that buyers are making more costly adjustments to cut back more
on their purchases of the good. Thus if a patent on computer-driven printers
causes their price to rise, people will print less, which causes productivity
to fall. Productivity falls more when the elasticity of demand for printers is
higher.

When demand is inelastic, creating monopoly profits for innovators imposes

a small deadweight loss. The case for enhancing market power for fertile in-
novations is stronger when demand for the innovation by other innovators is
less elastic. Conversely, the case for enhancing market power for fertile inno-
Alternatively, assume that market power enables A to sell α to B for 200. A will make a profit of
100, which equals the sales price of 200 minus the cost of development production 75+25. Foresee-
ing this fact, A will develop α. B buys α at a price of 200, develops β, and sells β for 300 to C, so B
makes a profit of 100. The sum of the joint profits of A and B equal 200. Thus market power of A
against B benefits both and causes their joint profits to rise from 0 to 200.

There is, however, a deadweight loss caused by A’s market power against B, which we also illustrate.
As noted, when B buys α for 200, A and B each earn profits of 100. We assume that the state pays a
subsidy to A to develop α and then requires A to sell α to B at the cost of production. Specifically,
the state imposes a tax of 175 on B, which the state pays to A for developing α, and the state requires
A to sell α to B at 25, which is the cost of production. In sum, A receives a subsidy of 175, spend 75
to develop α, spends 25 to produce α, and sells α to B at a price of 25, for a net profit of 100 for A.
A’s profits remain constant at 100, and B’s profits rise, say to 130, because B does not have to econo-
mize on the use of α when it develops β. The replacement of market power with a subsidy eliminates
a deadweight loss of 30 and increases joint profits of A and B from 200 to 230.

5.5
Ver. 1.4 - Chapter 5. Fertility

vations is weaker when demand by other innovators is more elastic. Market

power of more fertile innovators should vary inversely with the elasticity of de-
mand for innovations by less fertile innovators.7 (The market power of innova-
tors, however, should not vary with the elasticity of demand by consumers
for innovations, according to the separation principle in Chapter 4.8)

Besides the elasticity, the pattern of transactions among innovators affects the
deadweight loss. Circular transactions among innovators cause an especially
large deadweight loss. Much that a venture buys to develop an innovation
was originally an innovation by another venture. When innovators sell inno-
vations to each other, some transactions circle within the dynamic economy,
instead of proceeding straight from innovators to consumers as in Figure 5.1.
In Figure 5.2 venture A makes innovation α and sells it to venture B, who
uses it to develop innovation β and sells it to venture A for the next round of
innovations, and so on. (In Figure 5.2, innovators A and B also sell innova-
tions to consumers C.)

Figure 5.2 Circular Transactions

In circular transactions, the deadweight loss accumulates and becomes espe-

cially large. Thus assume that innovator A in Figure 5.2 uses market power
7  Ramsey worked out the inverse elasticity rule for optimal taxes in 1927. See analysis and foot-
note in the appendix to this chapter.
8  When demand by consumers for an innovation is elastic, market power for the innovator against
consumer causes a large deadweight loss. This effect, however, primarily concerns static efficiency.
Growth will overtake the effects on human welfare. So the market power of innovators should not
vary much with the elasticity of demand for innovations by consumers.

5.6
 Robert Cooter

to raise the price of the innovations that it sells to B. This use of market
power gains profits of 100 for A and imposes costs of 130 on B. Similarly,
assume that innovator B uses market power to raise the price of the innova-
tions that it sells to A. This use of market power gains profits of 100 for B
and imposes costs of 130 on A. The deadweight loss of monopoly in this cy-
cle of transactions between A and B is 60. If market power were eliminated,
the increase in profits of A and B would total 60, with each one gaining 30.
Their profits would increase because each gains more from lower costs than
it loses from lower revenues.

The more innovators deal in circular transactions with each other, the larger
is the increase in growth from reducing their market power. Eliminating
market power in circular transactions is relatively easy when everyone gains.
For example, if patents were the source of market power between A and B,
then both of them would enjoy higher profits if neither had patents that
were good against the other. Later we explain how firms achieve this result
by pooling their patents.

Circular transactions among innovators are sometimes described as an inno-

vation “thicket.” In an innovation thicket, everyone needs the innovations
of others in order to make their own innovations. If several innovators have
the power to withhold innovations that are necessary to the success of an-
other venture, then the each of them can effectively veto the venture. As the
number of innovator’s with veto power increases, the probability of a veto
increases. Similarly, many barons built castles on the Rhine River in medi-
eval times to extract tolls from passing ships. One baron who held out for a
high toll could block most traffic on the river. If law allows holdouts, they
sharply reduce innovation in thickets and trade along rivers. Consequently,
growth economics proscribes market power in innovation thickets, as the
next chapter explains for patents.

Bargaining and Transaction Costs

In perfect competition, everyone takes the market price. In contrast, market
power gives the seller choice over the price. The seller names the price and
the buyers respond to it. Sellers with market power name prices in two dif-
ferent ways. First, the seller may name a firm price, as when the vendor in

5.7
Ver. 1.4 - Chapter 5. Fertility

a football stadium sets the price of a glass of beer. In the standard model of
monopoly, the monopolist maximizes profits by setting a firm price where
the marginal revenue from a small increase in production equals the marginal
cost.9

Second, instead of a firm price, the seller may name a flexible price, and each
buyer may make a counter-offer. With price flexibility, the parties bargain
with each other to reach an exact price. Bargaining takes time and effort.
In addition to time and effort, bargaining sometimes fails to reach an agree-
ment. Three major costs of transacting by bargaining are time, effort, and
the loss from failure to agree.

To illustrate, assume that a football player is willing to play for 100 and a
team values the player at 120. When they bargain over salary, the player
will not accept less than 100 and the team will not pay more than 120. If
bargaining succeeds, they will agree on a price in between 100 and 120, de-
pending on their bargaining power.10 Say they agree on the price 112, so the
player gains 12 and the team gains 8. Bargaining takes time and effort that
costs, say, 1 for the team and 1 for the player. The gain net of bargaining
costs thus equals 7 for the team and 11 for the player. In addition to the
time and effort, bargaining involves uncertainty about the result. When the
parties begin bargaining, they do not know whether or not they will reach an
agreement. If the probability of success in our example is, say, 90%, then the
probability of failure is 10%. Rational parties discount the expected value
9  A seller’s revenue equals the price p multiplied by the quantity q that he sells. For a monopolist,
the quantity is a decreasing function of the price. So the monopolist’s revenue R can be written
R=pq(p) where q`<0. The monopolists cost of production C increase with the quantity, which
decreases with the price, so C=C(p) where C`<0. Profits, which equal revenues R- C, are maximized
when marginal revenues equal the marginal cost: q+pq`= C`.
10  When two parties bargain with each other, their bargaining power depends on their alterna-
tives. In perfect competition, each party to a bargain has a perfect substitute. The buyer can buy
the same good from another seller at the market price, and the seller can sell the good to another
buyer at the market price. With perfect substitutes, neither party has any bargaining power against
the other. That is why they are price takers. Alternatively, in monopoly, the buyer cannot buy the
same good from anyone except the single seller, whereas the seller can sell to another buyer. With
close substitutes for the buyer and no substitutes for the seller, the seller has all of the bargaining
power and the buyer has none of it. Law and policy increases the market power of the seller of an
innovation by reducing the substitutes available to buyers. For example, a patent that effectively
stops infringements reduces the substitutes available to its buyers. A remarkably simple formulation
of the relationship between bargaining power and negotiated prices is the Nash Bargaining Solution,
which Chapter __ discusses.

5.8
 Robert Cooter

of the bargain by its probability of failure. The net gain discounted for the
probability of failure is .9(7) for the team and .9(11) for the player, which
sum to 16.2. The difference between the gross gain of 20 for the team and
player, and the discounted net gain of 16.2, is 3.8.

This number, 3.8, can be called the bargain’s “transaction costs,” and it can
also be called the bargain’s “deadweight loss.” The equivalence of “transac-
tion costs” and “deadweight burden” yields two formulations of the fertility
principle. According to the original formulation, to maximize growth, law
and policy should transfer profits from less to more fertile innovations until
the increase in growth from higher fertility equals the decrease from a heavier
deadweight loss. Equivalently, law and policy should transfer profits from
less to more fertile innovations until the increase in growth from higher fer-
tility equals the decrease from higher transaction costs.

To apply this principle, contrast bargaining between two parties and many
parties. When only two people must agree to a bargain (say, the buyer and
the seller), negotiation often takes little time or effort, and the parties can
usually reach an agreement. With two parties, transaction costs are low,
especially when the law enforces contracts between them.11 Thus the dead-
weight loss is negligible if A sells α to a single buyer B. Conversely, with
many individuals involved, negotiation takes more time and effort, and the
parties may not reach an agreement. Thus the deadweight loss is likely to be
high if A sells α to B where “B” stands for many individuals who must agree.

“Frictionless plane” describes an engineer’s theoretical ideal, not the real

world. Without friction, objects move forever in the same direction at the
same speed forever. In the real world, the amount of friction determines
the speed and distance that an object will move. So a frictionless plane pro-
vides a baseline for calibrating and predicting actual movements. Similarly,
“zero transaction costs” describes an economists’ theoretical ideal, not the
real world. Without transaction costs, exchange continues until all resources
are owned by the parties who value them the most – a state of “allocative
efficiency”. Applied to law, this is the most famous proposition in law and
economics, name the “Coase Theorem” after its discoverer. According to this
11  With good laws, many agreements can be implemented in enforceable contracts, but not all
agreements. Recall the double trust dilemma in Chapter 3 where the innovator and financier cannot
implement their agreement to develop an innovation because performance of the terms is unverifi-
able in court.

5.9
Ver. 1.4 - Chapter 5. Fertility

theorem, the allocation of legal entitlements does not matter to efficiency so

long as private bargains can reallocate them without transaction costs. Thus
intellectual property law can allocate ownership of an innovation to anyone
(the inventor or someone else) in any form (strong or weak patents), and pri-
vate bargains will re-allocate ownership to the party who values it the most,
provided that transaction costs are zero. By referring to this intriguing and
puzzling theorem, subsequent chapters will convey its meaning for intellec-
tual property.

Conclusion
Maximizing growth requires an open competition among innovators with
extraordinary profits as the prize. Innovators enjoy extraordinary profits
when law and policy allows them to exercise the market power that they have
or gives them market power that they lack. According to the separation prin-
ciple in the preceding chapter, law and policy should strengthen the market
power of innovators against consumers and producers who do not innovate.
According to the principle of fertility in this chapter, law and policy should
strengthen the market power of fertile innovators against sterile innovators.

Market power within the dynamic economy has two effects on growth:
transfer and deadweight loss. Transfers from less to more fertile innovations
increase growth, and deadweight loss in transactions among innovators de-
creases growth. Redistribution from less to more fertile innovators should
proceed until the decrease in growth from deadweight loss equals the in-
crease in growth from fertility. Intellectual property, which is the subject of
the next chapter, is one bodies of law for implementing the separation and
fertility principles.

5.10
 Robert Cooter

Appendix: Welfare Triangles

Figures 5.3 and 5.4 depict the relationship between the burden, the dead-
weight loss, and the elasticity of demand when the seller has market power
against the buyer. 12 The horizontal axis represents the quantity of innovation
α produced, and the vertical axis represents prices and costs. After A devel-
ops α, c equals the cost of producing a unit of it. If A sells α to B at the cost
of production c, then c equals the market price pc. Since production cost
equals price, A earns zero profits on production and cannot recover the cost
of development. Alternatively, if A has market power against B, then A will
raise the price to pm and B will buy the quantity qm. A’s profit equals the dif-
ference between the price and cost of production multiplied by the quantity,
or (pm-c)qm, which Figure 5.2 denotes T. The transfer T is the gain to A from
receiving the high price pm for the small quantity qm, instead of receiving the
low price pc for the large quantity qc.

In the numerical discussed in this chapter, T equals 100. However, the loss
to B of paying the higher price pm is more than the transfer T. Besides paying
more for the qm units of α that B buys, the higher price causes B to forego
the purchase of qc-qm units of α. The loss to B from foregoing this purchase
equals the difference between the amount B was willing to pay for them, as
given by B’s demand curve, and their cost of production c. In Figure 5.2, L
denotes this loss. B’s loss from A’s market power equals T+L, which is the
burden on B. A’s gain from its market power equals T. The difference L be-
tween B’s burden and A’s gain is the deadweight loss L.

12  Ramsey first worked out the mathematics for taxation, and he concluded that optimal taxes are
inversely proportional to the elasticity of demand. See Frank P. Ramsey, A Contribution to the Theory
of Taxation, 37 Economic Journal 47–61 (1927).

5.11
Ver. 1.4 - Chapter 5. Fertility

Figure 5.3. Burden and Deadweight loss of Market Power

Figure 5.4 shows the increase in the deadweight loss when demand becomes
more elastic. Figure 5.4 holds demand constant at the point (pm,qm) and
rotates the demand curve from Dinelastic up to Delastic. This change holds
A’s gain of T constant, but the deadweight loss increase from L to L+L’. The
deadweight loss increases because B responds to the price increase by making
a larger reduction in the good’s use. The larger reduction in use involves
more costly adjustments to get by with less of the good.

5.12
 Robert Cooter

Figure 5.4. Burden, Deadweight loss, and Elasticity of Demand

5.13
 Robert Cooter

Chapter 6. Separation and Fertility in Intellectual

Property Law
Robert Cooter

To speed innovation and increase human welfare, law should provide a

framework for open competition among innovators with high profits for the
winners. High profits especially come from the innovators’ market power,
which is the power to affect prices. The first firm to develop and market
an innovation often enjoys market power temporarily (the “first mover’s
advantage”), which is a form of natural monopoly. Secrecy often extends
the innovator’s market power by slowing the discovery’s dissemination to
competitors. Patents and copyright that shield innovators from competi-
tion create market power by legal fiat. These three sources of market power
-- natural monopoly, private information, and legal fiat -- especially relate to
three bodies of law: antitrust, trade secrets, and intellectual property. This
chapter concerns the third source of market power for innovators -- intellec-
tual property law.

Article 1 Section 8 of the U.S. constitution reads:

“The Congress shall have power…To promote the progress of

science and useful arts, by securing for limited times to authors
and inventors the exclusive right to their respective writings and
discoveries…”
The “useful arts” mentioned in Article 1 Section 8 form a heterogeneous col-
lection of activities. To measure “progress” requires combining the activities
into an index. A comprehensive measure of economic growth is often the
best index of progress in the useful arts. Thus Article 1 section 8 authorizes
Congress to promote economic growth through intellectual property law.

Congress asserted its constitutional power by enacting Title 35 Sections 100-

105 of the United States Code.1 The constitution and much of the code is
general and abstract. Regulations and court decisions increase the precision
and certainty of these laws. In the U.S. where courts justify their interpre-
tations by reference to prior court interpretations, much intellectual prop-

1  Section 101 [2] sets out subject matter that can be patented; section 102 [3] defines novelty and
loss of right to patent; section 103 [4] lists what constitutes non-obvious subject matter.

6.1
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

erty law is “judge-made”. Economic growth provides a useful standard for

interpreting and critiquing intellectual property law made by Congress and
courts. The separation from Chapter 4 and fertility principle from Chapter
5 contribute to interpretation and critique, especially where innovations can
grow exponentially.

This chapter applies the separation and fertility principles to innovations that
can grow exponentially, so the effects of growth on human welfare overtake
other effects. Patented inventions can grow exponentially. So can some
copyrighted expressions such as computer programs, as well as some innova-
tions covered by special statutes.2 In contrast, cultural expressions (novels,
songs, plays) evolve and develop, but these changes do not constitute expo-
nential growth on a generally accepted measure such as cost-benefit analysis
or market value. Compared to economic growth, any attempt to measure
progress in artistic expression immediately encounters deep disagreements
over cultural values.

The patent or copyright for inventing a machine, creating a molecule, ex-

tracting a vaccine, or writing a computer program can be weak or strong.
The separation principle distinguishes infringement by innovators from in-
fringement by consumers or producers. Patent protection should be strong
against infringers who consume the invention or produce with it. According
to the fertility principle, patent protection should be weak or absent against
users who innovate with it in the usual cases. In exceptional cases, however,
patents should be strong against other innovators where the patented inven-
tion is more fertile than the innovation for which it is used.

Strength
We will develop the theory and application of the optimal strength of intel-
lectual property law. The strength of intellectual property rights depends on
three attributes: duration, breadth, and remedy. The duration of a patent can
be measured in years. Thus U.S. patent duration is 20 years from the date
of filing the patent application.3 When filing, the applicant and her patent
attorney make claims about the innovation’s breadth. The patent office and
2  The Semiconductor Chip Protection Act of 1984, which legally protects the layouts of integrat-
ed circuits, exemplifies such special statues.
3  Before 1995, U.S. patent duration was 17 years from the date the patent issued.

6.2
 Robert Cooter

the courts ultimately decide the validity of these claims. Breadth is easier to
order than to measure. Thus a patent covering rain gear is broader than a
patent covering umbrellas, and a patent covering umbrellas is broader than a
patent covering automatically opening umbrellas. While one patent can be
broader or narrower than another, there is no natural measure of the differ-
ence in breadth.4

Taken together, the duration and breadth of an intellectual property right

define its scope. Acting within the scope of a patent or copyright without
a license from the owner infringes it. Law usually provides a remedy for
infringement. For past infringements, the owner can usually sue to recov-
er compensatory damages. Higher damages (“super-compensatory”) are al-
lowed sometimes. For future infringement, the owner can usually enjoin the
infringing conduct.5

4  Patent breadth resembles the alphabet. Letters A through N encompasses more letters than A
through M, but there is no natural measure of the distance between M and N. Similarly, no natural
measure exists to answer the question, “Is the difference in breadth between a patent on rain gear and
a patent on umbrellas larger or smaller the difference in breadth between a patent on umbrellas and
a patent on automatically opening umbrellas?” In this respect, patent breadth contrasts with height.
One person is taller than another and the difference can be measured in centimeters.

Fortunately, an ordering without a measure can support the mathematics of maximization. This
fact makes modern utility theory possible in economics. Economic models often maximize utility
functions that order states of the world by preference (“ordinal utility”), but do not measure differ-
ences in utility levels. Similarly, finding the breadth of patent that maximizes growth only requires
an ordering by breadth, not a measure.
5  Note that money measures damages like years measure duration, whereas breadth orders (but
does not measure) injunctions like patent claims order their breadth.

6.3
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

Figure 6.1 Future Innovations

In addition to increasing the strength of individual patents, the patent sys-

tem as a whole can strengthen against unpatented activities. Figure 6.1 di-
vides future innovations into “ownable” and “unownable.”6 A new computer
chip is ownable (patentable) and a new metaphor in the English language is
unownable (no patent, no copyright). The strength of all patents indicates
the patent system’s reach, as illustrated by the boundary between ownable
and unownable innovations in Figure 6.1. Thus shifting the boundary in
Figure 6.1 from R to `R increases ownable innovations and decreases un-
ownable innovations. To illustrate, U.S. patent law increased its reach when
inventors were allowed to patent business processes for the first time, such as
Amazon’s “one-click” ordering from its online catalogue. Expanding patent
law’s reach fences in more of the common land of innovation.7

6  This is a picture, not a graph, because the breadth of innovations has no natural measure. See
the discussion in this chapter of the difference between an ordering and an index.
7  Increases in the scope of a particular patent do not necessarily increase the scope of patents as a
whole. Rather, the scope of a particular patent can increase by decreasing the scope of another pat-
ent. Instead of fencing more of the common land of innovation, the fence between patent holders
moves to the advantage of one of them and to the disadvantage of the other, as with the boundary
between α and β in Figure 6.1.

6.4
 Robert Cooter

Besides strengthening the patent system against unpatented activities, patent

law can strengthen one patent by weakening another. The space of innova-
tions represented by the small circle in Figure 6.1 is subdivided into α and β.
Assume that firm A invents α and firm B subsequently invents β. A claims
that α and β are a single invention and A files for a patent covering both of
them. Later B files for a patent for β and claims that β is a different invention
from α. The patent office must decide whether to award a broad patent to A
that covers α and β, or to award a narrow patent to A that covers α and a nar-
row patent to B that covers β. For a given space of inventions, the state can
generally give few broad patents or many narrow patents, rather like a city
planner can divide undeveloped land into a few large lots or many small lots.8

How Strong?
What reach and strength (breadth, duration, and remedy) of intellectual
property rights maximizes innovation? The answer is remarkably simple.
Maximizing innovation requires transferring resources from consumption
and production to innovating. According to the separation principle (Chap-
ter 4), an innovator should have a strong patent against others copying the
innovation to sell to consumers, such as a smart phone application. An in-
novator should also have a strong patent against others producing consumer
goods by using the innovation, such as industrial robots that manufacture
cars. Applied to the strength of intellectual property, the separation principle
asserts that the innovator’s rights against others consuming it or producing with
it should be strong -- broad and long with a generous remedy for infringement. 9

Besides producing and consuming, some innovations need prior innovations

like a carriage needs wheels. When a prior innovation is owned, the owner
usually has market power against a subsequent innovator who needs a license
8  Once lots are surveyed in a new town, owners can sell a lot but not half of a lot. The owner
who cannot sell half of a lot can lease it for use by someone else. Much the same is true for selling
patents. An owner cannot sell half of a patent, but an owner can contract for someone else to use
half of it. Some theorists think that divisibility is the essential difference between property rights and
contract rights. See Thomas A. Merrill & Henry Smith, “Optimal Standardization in the Law of
Property: The Numerus Clausus Principle,” 110 Yale Law Journal (2000).
9  Economists may wish to compare broad patents and broad taxes. Economists favor broad
taxes because the demand for broad categories of goods is less elastic than the demand for narrow
categories of good, so the deadweight loss is smaller. Similarly, the demand for broad patents is less
elastic the demand for narrow patents, so the deadweight loss is smaller and the transfer from buyers
to sellers is larger.

6.5
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

to use it. Market power of innovators against each other imposes a dead-
weight loss on all of them, which slows innovation and growth. This fact
argues that innovators should have weak intellectual property rights against
each other, or none at all.

Chapter 4 described four forms of deadweight loss and their likely extent.
Creating market power for an innovator causes a large deadweight loss when
the patented good has close substitutes (high elasticity of demand),10 few
obstacles to resale (no price discrimination),11 the innovation is complex and
opaque (incomplete information for buyers),12 and many buyers purchase
small amounts of the good.13 These conditions argue that innovators should
have weak intellectual property rights against each other, or none at all.

To apply the separation principle, lawmakers and courts must distinguish the
uses of an innovation into consuming, producing, and innovating. In many
cases, the distinction requires non-technical judgments of the kind that
courts routinely make without statistics or mathematical theories. Common
sense usually suffices, but difficulties arise at the boundaries of these three ac-
tivities. In hard cases, economists can assist by drawing on their long history
of distinguishing these three activities. Economists can also assist by predict-
ing the effect of different distinctions on the goal underlying the separation
principle: maximizing innovation.

Legal application of the separation principle can mimic licensing practices by

some firms owning fundamental inventions like computer platforms. Apple
allows developers of applications to use the iPhone platform for free, whereas
consumers must pay a fee for using many of the resulting applications. Thus
Apple separates innovation from consuming, and asserts weak rights innovat-

10  Goods are substitutes when having more of one makes a person demand less of the other.
When a good has close substitutes, raising its price (while keeping other prices constant) causes
buyers to switch to the substitute. Switching costs are a deadweight loss.
11  With few obstacles to resale, the owner cannot charge different prices to different buyers for
the same innovation. Price discrimination is impossible. The more the seller can discriminate
among buyers, the less the buyers will switch in response to a price increase, so the deadweight loss of
switching falls. This is an application of the proposition found in microeconomics textbooks that a
perfectly discriminating monopolist is efficient.
12  As a good becomes more complicated and opaque, the amount of bargaining increases, and so
do the search costs of determining whether an innovation is already patented.
13  Transactions costs of purchases often have economies of scale. Consequently, the transaction
cost of market exchange is higher with many buyers make small purchases than when few buyers
make large purchases.

6.6
 Robert Cooter

ing and strong rights against consuming. Similarly, Sun Microsystems devel-
oped the Java programming language in the 1990s and made it available for
free to developers under a general-purpose license.14 Developers use Java to
write programs that they sell to consumers.15 Thus Java is free for innovators
and costly for consumers of its applications.

As explained, the overtaking principle and the deadweight loss from mo-
nopoly argue for weak intellectual property rights of innovators against each
other. However, an argument in the opposite direction applies when in-
novations differ significantly in fertility. By definition, costless transfers (no
deadweight loss) from less to more fertile innovations increase the rate of
innovation and growth. Stronger intellectual property rights for more fertile
innovations transfer wealth away from less fertile innovations. In transac-
tions among innovators, relatively greater fertility favors relatively stronger
intellectual property rights.

As explained, deadweight losses argue against market power for innovators

against each other, and greater fertility argues in favor of market power for
some innovators against others. Maximum growth requires balancing these
considerations. The fertility principle asserts that the ideal balance is struck
when further strengthening of property rights of fertile innovations causes growth
to decrease from greater deadweight loss by the same amount as it increases from
greater fertility. Equivalently, the innovator’s rights are ideally strong against
others using it for less fertile innovations when further weakening increas-
es growth from lower deadweight loss by the same amount as it decreases
growth from lower fertility.

Applications
To maximize progress in the useful arts, intellectual property law should ap-
ply the separation and fertility principles. The separation principle favors
giving innovators strong rights – broad, long, and with a generous remedy
for infringement -- against others consuming the innovation or producing
14  A reverse of this pattern sometimes occurs. Thus when the owner of a platform for running
computer games allows consumers to use it for free and charges developers who want to use it to
create new games to sell to consumers.
15  Legal disputes subsequently arose when Microsoft modified java to run exclusively on its
Windows program, thus necessitating the use of Windows to run programs written in the modified
java language.

6.7
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

with it. The fertility principle usually favors giving innovators weak rights or
no rights against others using the innovation to make subsequent innova-
tions. In unusual case of especially fertile innovations, an innovator’s rights
against others using the innovation to make subsequent innovations should
balance incentives for fertile innovations and the deadweight loss from stron-
ger property rights. We will apply this general theory to intellectual proper-
ty’s strength as indicated by reach, breadth, duration, and remedy

Reach: Invention or Discovery

The boundary between owned and unowned innovations defines the intel-
lectual property system’s reach as depicted in Figure 6.1. Courts and regu-
lators have attempted to draw the boundary in various ways. One line of
cases distinguishes inventing and discovering. To be patentable in U.S. law,
an innovation must be an invention, but inventions often involve discoveries
about nature. A fault-line of legal dispute concerns the difference between
patentable inventions and unpatentable discoveries about nature.

To illustrate, the inventors of the telegraph in the 19th century attempted

to patent the use of the electromagnetic spectrum to transmit messages. In
O’Reilly v. Morse, the U.S. Supreme Court decided that the telegraph’s in-
ventors could patent a particular machine to transmit messages by using the
electromagnetic spectrum, but not all ways of doing so. The electromagnetic
spectrum is unpatentable, like the laws of nature and observations of physical
phenomena.16 A similar conclusion concerns DNA sequencing. Does the
first person to observe a DNA sequence acquire the exclusive right to observe
it? In Myriad Genetics (2013) the U.S. Supreme Court unanimously decided
“No”. Inventions can be patented, but not discoveries of natural laws and
facts.

How does this legal distinction square with the separation and fertility prin-
ciples? The separation principle favors strong patent protection over static
activities (consuming and producing) and weak or no patent protection over
dynamic activities (innovating). The telegraph was the first device to use the
electromagnetic spectrum to transmit messages, and many others followed,
including telephones and computers. The patent for the telegraph gave its
inventor’s temporary market power over transmitting messages by using the
16  The court recently sustained this view in Bilski (2010).

6.8
 Robert Cooter

electromagnetic spectrum. The patent was strong against telegraph trans-

missions, which mostly concern consumption and production. However, the
patent was absent against subsequent inventions that transmit message by
using the electromagnetic spectrum like telephones. So the breadth of the
patent for the telegraph was consistent with the separation principle and it
created strong incentives to innovate, not obstacles to it.

The separation and fertility principles broadly favor the patentability of in-
ventions and the non-patentability of discoveries of natural laws and facts.
Easy cases can follow common-sense distinctions, and hard cases can look to
evidence about locating the boundary to maximize economic growth. An
example of a hard case concerns whether diagnostic devices in medicine are
inventions or observations of nature. Diagnostic devices are used to treat
patients (consumption) and to conduct medical research (innovation). The
separation and fertility principles commend a different standard of “infringe-
ment” for medical treatment and medical research.

To speed innovation, the separation and fertility principles argue for patent
coverage of patient treatments, which implies that unlicensed use of a diag-
nostic device to treat a patient infringes the patent. The patent will increase
the cost of treatment, and patients will lose from higher prices in the short
run, but, according to the overtaking theorem in Chapter 1, patients will
gain more from faster innovation in diagnostic devices. The conclusion is
different for research as opposed to treatment. To speed innovation, the
fertility principle argues against patent coverage for research, which implies
that unlicensed us of a diagnostic devices to make medical innovations does
not infringe the patent. The law follows these prescriptions imperfectly. In
Prometheus (2012) the U.S. Supreme Court disallowed the patenting of di-
agnostic devices (and also surgical methods) without distinguishing between
treatment and research. In pharmaceuticals, however, the law follows the
separation principle and allows a “research exemption,” as we will discuss
shortly.

The preceding discussion of “reach” mostly concerns the difference between

inventing and discovering. To be patentable in U.S. law, an innovation must
be an invention, not a discovery of a natural phenomenon. In addition, it
must be useful. For most patent applications, the authorities presume that
the invention is useful without weighing the arguments. In unusual cases

6.9
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

that weigh the arguments, some commentators doubt that an invention has
utility if its only use is to generate more inventions. This reasoning is con-
fused. A fertile innovation promotes “progress” in the “useful arts” even if
it is not used in consumption or production. Progress in the useful arts is
the constitutionally recognized purpose of intellectual property law. Thus
fertility is a constitutionally recognized form of utility, not a suspect category.

Besides being useful, U.S. patent law requires an invention to be new. An

invention is new if it is unknown to the public and not obvious. The most
important disputes about novelty concern whether one innovation is suf-
ficiently different from another to justify a separate patent. The breath of
patents is our next topic.

Breadth
In Figure 6.1 the patent authorities could issue one broad patent to A that
encompasses both α and β, or a narrow patent to A for α and a narrow patent
to B for β. Are α and β sufficiently similar to justify one patent, or are they
sufficiently different to require two patents? Common sense understanding
of it and engineering facts give the answer in most cases, but not in the hard
ones. Resolving hard cases requires connecting legal rules to their purposes.
The constitutional purpose of intellectual property law in the U.S. is promot-
ing progress in the useful arts, which economic innovation measures.

Connecting patent breadth and growth requires understanding how econo-

mists measure similarity in goods. Economists regard two goods as similar if
they are substitutes in markets, which implies that people will buy whichever
one is cheaper, and not buy the dearer one (high cross-elasticity of demand).
The breadth of patents affects innovation partly through the substitutability
of goods. If α and β are close substitutes for each other, and if no close substi-
tutes exist for them, then one broad patent will create a monopoly and two
narrow patents will create a duopoly. Compared to duopoly, a monopoly
can charge at least as high prices and extract at least as much wealth from
buyers.17 So one broad patent for A covering α and β transfers at least as
17  Duopolists sometimes cooperate perfectly with each other, or they merge. In that
case, they maximize their joint profits and earn the same profits as a monopolist. More
often, however, duopolists cooperate imperfectly with each other and do not monopolize
their joint profits. So the profits of a monopolist are at least as high as a duopolist for the
same product.

6.10
 Robert Cooter

much wealth from buyers to the owner A as two narrow patents transfer to
the owners A and B.

When goods are substitutes, their use determines whether transferring wealth
from buyers to sellers increases or decreases incentives for innovation. If in-
novations α and β mostly have static uses (consuming and producing), then
one broad patent will transfer more wealth from static to dynamic activities
than two narrow patents, which increases the incentive to innovate. Thus
the separation principle favors broad patents for innovations that substitute
in consumption or production. Conversely, if innovations α and β are sub-
stitutes and they mostly have dynamic uses (innovating), issuing two narrow
patents will impose less deadweight loss on innovating, which increases the
incentive to innovate.

In hard cases, the connection between similarity as measured by substitut-

ability and incentives for innovation can often provide definite answers about
the optimal patent breadth. The separation principle favors broad patents for
substitutes in consumption and production. The fertility principle usually
favors narrow patents for substitutes in innovation. However, in the excep-
tional case the fertility principle favors broad patents for very fertile substi-
tutes in innovation.

Demand for two goods is independent if the price of one does not affect
how much people will pay for the other (zero cross price elasticity). Inde-
pendent goods are dissimilar economically. Intellectual property law cannot
increase the market power or profitability of innovating by issuing broad
patents encompassing independent inventions. Thus when demand for α is
independent of demand for β, owning the patent on β does not increase the
profitability of owning the patent on α. Since independent goods are dissim-
ilar, broad patents do not increase the incentive to innovate. The breadth of
intellectual property makes little difference to market power created by pat-
ents for independent innovations. The patent authorities should issue narrow
parents for inventions of independent goods.

Table 6.1 summarizes these conclusions about use, substitution, and pre-
ferred patent breadth.

6.11
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

Table 6.1. Preferred Patent Breadth

Substitutability
close substitutes independent
consuming or strong patent weak patent
producing (broad preferred to narrow) (narrow preferred to broad)
Use
innovating, usual weak patent weak patent
case (narrow preferred to broad) (narrow preferred to broad)
innovating, very strong patent weak patent
fertile case (broad preferred to narrow) (narrow preferred to broad)

This reasoning applies to a series of suits that Apple brought against Samsung
beginning in 2011 for infringement of design patents for the iPhone and
iPad. For smart phones and tablets, Apple was the original innovator and
Samsung was the imitator. Samsung’s strategy was to build close substitutes
to the originals and to make small improvements. By imitating, Samsung
saved development costs, so Samsung built substitutes at lower cost and sold
them to consumers at lower prices. Consequently, Samsung’s share of the
smartphone market increased at Apple’s expense.18

The legal question is whether Samsung infringed Apple’s patents. Table 6.1
provides an answer by relating preferred patent breadth of substitutes to their
use. In so far as Samsung mostly used Apple’s innovations to make consumer
goods, the separation principle favors strong protection. Strong protection
will reward innovation and speed progress in the useful arts. In so far as
Apple’s innovations were fundamental and Samsung’s innovations were de-
rivative, the fertility principle also favors strong protection of Apple’s patents
against Samsung. Again, strong protection will reward fertile innovation
and speed progress in the useful arts. However, in so far as Samsung mostly
used Apple’s innovation to make its own innovations of similar fertility, the
fertility principle favors weak protection of Apple’s patents against Samsung.
Weak protection will reduce the deadweight loss on innovation and speed
progress in the useful arts.

Samsung’s liability to Apple ideally depends on the extent to which Samsung’s

smart phones and tablets embody its own fertile innovations. Samsung’s de-

18  Sang-Hun Choe, After Verdict, Assessing the Samsung Strategy in South Korea By CHOE
SANG-HUN ``, New York Times, Summer, 2012, http://www.nytimes.com/2012/09/03/
technology/companies/south-korea-reassesses-samsung-after-battle-with-apple.html?_
r=1.

6.12
 Robert Cooter

vices appear to use Apple’s designs mostly to increase consumer sales, and
Samsung’s innovations appear to be derivative. If these appearances are the
deeper truth, then Apple’s patent protection should be strong against Sam-
sung. Liability of Samsung to Apple will promote faster growth in the future
by causing Samsung to devote more of its formidable technological prowess
to fundamental inventions and less to imitations.

We have discussed the alignment between patent breath and the separation
principle. For the pharmaceutical industry, an important exception to U.S.
patent law aligns them closely. Firms that develop a new drug need regula-
tory approval before marketing it in the U.S. To gain regulatory approval,
the drug’s owner must prove that it is effective and safe. Proof often requires
research on other chemically related drugs, including rivals for the same mar-
ket. The owner of a drug has reason to refuse to license its use by a rival. A
legal exemption (“Hatch-Waxman”) allows firms to use patented drugs for
research on regulatory compliance without a license from the patent’s own-
er. Instead of obtaining permission to license the patented drug at the price
demanded by the owner, the innovator can use the drug and pay a royalty
set by legal authorities. In the language of property rights, the innovator
can “take” the patentee’s property right by paying an “objective price” set by
the court, rather than having to pay the “subjective price” demanded by the
patent’s owner.

The research exemption for pharmaceuticals discriminates between using an

innovation for innovating and using it for consumption or production, as
prescribed by the separation principle. The research exemption for phar-
maceuticals should ideally generalize and extend to other industries. Un-
fortunately, a recent case on research by universities went in the opposite
direction.19

Duration
Next, apply the separation and fertility principles to the duration of a patent.
The horizontal axis in Figure 6.2 indicates the length of patents in years, and
the vertical axis indicates the rate of innovation. The curves in Figure 6.2
depict some possible relationships between the rate of innovation and pat-
ent length. For curve A, the rate of innovation is an increasing function of
19  Universities were traditionally exempted from the legal requirement to license patented inven-
tions for use in research in laboratories. This traditional research exemption was eliminated in …

6.13
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

patent length. Curve A depicts an industry where innovators sell mostly to

consumers and producers, such as pharmaceutical drugs. For a curve like A,
optimal patent length is infinite for patents against consuming and produc-
ing an innovation.20

For curve C, the rate of innovation is a decreasing function of patent length.

Curve C depicts an industry where innovators of similar fertility sell mostly
to each other. In Curve C the dominant effect of longer patents is to transfer
wealth among dynamic users without regard to fertility. The deadweight loss
from the transfers slows innovation.

Curve B depicts the intermediate case in between the pure cases of Curve
A and Curve C. In the intermediate case, strengthening weak intellectual
property rights increase the rate of innovation at first, but a point is reached
where further strengthening decreases innovation. In Figure 6.2, the optimal
patent duration of patents is internal, not infinite as for curve A or zero as
for curve C. In the intermediate case, moderately strong intellectual property
rights maximize the rate of innovation.

20  Landes and Posner advocate patents of infinite length, which we favor for patents against con-
suming and producing, but not for patents against innovating. CITE;

6.14
 Robert Cooter

Figure 6.2. Rate of Innovation and Patent Length

Curves A, B, and C in Figure 6.2 depict three possible relationships that

often exist between the rate of innovation and the duration of intellectual
property rights. (We will not discuss other possible shapes.21) For curves
with these shapes, a simple rule leads to the maximum: Starting from 0,
increase the strength of intellectual property rights so long as venture profits
increase. The maximum occurs at infinity for curve A, zero for curve C, and
an interior point for curve B.

The ideal patent system grants patents of different length according to the
product’s growth profile as in Figure 6.2. German patent authorities issue
“petty” patents of short duration for minor inventions, and “full” patents
of longer duration for major inventions. In contrast, American patent law
issues full patents for all patented inventions. Does the German or Ameri-
can patent regime come closer to this ideal? For inventions mostly sold for
producing and consuming, full patents will transfer more wealth from static
to the dynamic uses. Consequently, American patent law is closer to the
ideal than German patent law for innovations mostly used for consuming
and producing. In contrast, for innovations mostly sold for innovating, the
ideal law issues strong patents to fertile inventions and weak patents to sterile
21  Curve B is a concave function with a unique maximum. Another possibility is a convex func-
tion with two local maxima, one for weak intellectual property rights and one for strong intellectual
property rights. Another possibility is multiple equilibria. For analyzing many activities, economists
assume a concave function with a unique maximum.

6.15
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

inventions. A stronger patent for fertile innovations and a weaker patent for
sterile innovations transfer wealth from the latter to the former in accordance
with the fertility principle. In so far as German patent authorities issue “full”
patents to fertile inventions and “petty” patents to sterile inventions, German
patent law is closer to the ideal than American law for innovations mostly
used for innovating.

The international default standard for patent duration, which the U.S. has
adopted, is twenty years from the date of the application for a patent. Quan-
titative research on the optimal duration of intellectual property rights has
barely begun. No compelling evidence indicates whether the pace of inno-
vation would increase from lengthening or shortening the duration, or from
distinguishing between petty patents and full patents. According to U.S. law,
copyright endures for the creator’s life plus 70 years.22 Many scholars be-
lieve that this is far too long. The shadow of dead writings apparently inhib-
its creativity. A U.S. statute on semiconductor protection possibly has a bet-
ter rule. The Semiconductor Chip Protection Act of 1984 (or SCPA) legally
protects the layouts of integrated circuits for 10 years from registration.23

Optimal Remedy
Patent law provides damages for past infringements. Three benchmarks for
damages are compensation, disgorgement, and treble damages. Compensa-
tion ideally puts the victim in the same position as if the infringement had
not occurred. For firms, this requires transferring profits from the injurer to
the victim until the victim’s profits equal what they would have been but for
the infringement. Disgorgement ideally puts the injurer in the same position
as if the infringement had not occurred. For firms this requires transfer-
22  In October, 1998, Congress passed the Sonny Bono Copyright Term Extension Act, which
lengthens copy- right protection for works created on or after January 1, 1978, to the life of the
author plus 70 years, and extends existing copyrights “created for hire and owned by corporations” to
95 years. Before the change, the 1976 Copyright Act had given protection for the author’s life plus
50 years. Whatever other reasons there may be for the Copyright Term Extension Act, one justifica-
tion is that it brings U.S. practice into conformity with Western European practice.
23  The three dimensions of strength of intellectual property protection for the layouts of integrat-
ed circuits, which the law calls a chip’s “mask”, mix standards from patents and copyright. Breadth
is determined by the substantial similarity” test of copyright law, duration is 10 years from registra-
tion, and the available remedies correspond generally to copyright and patent law. Protection covers
all embodiments of the chip’s layout, but does not cover the chip’s functionality that patents law
protects or the information on chips that copyright law protects (if they are protected at all). If the
defendant asserts reverse engineering as a defense and provides “probative evidence” of it, then the
standard shifts from “substantially similar” to “substantially identical.”

6.16
 Robert Cooter

ring profits from the injurer to the victim until the injurer’s profits equal
what they would have been but for the infringement. Treble damages simply
mean three times compensation.

For infringement of intellectual property rights, compensation is the usual

remedy in U.S. law. Disgorgement is seldom available except when infringe-
ment of an intellectual property right involves breach of a fiduciary duty
or unjust enrichment. When the law allows the disgorgement remedy, the
victim can choose between compensation and disgorgement. Treble damag-
es are the remedy in U.S. patent law for willful infringement of intellectual
property rights.

The following inequalities order these remedies by the usual size of damages,
or, equivalently, by their strength:

treble damages > choice of compensation or disgorgement > compensation.

In addition to damages for past harm, the owner can often obtain an injunc-
tion against future infringements.24 This possibility gives the owner a choice:
either prevents future infringements by enjoining them before they occur,
or deter future infringements by obtaining damages after they occur. If the
law allows the owner to defend his patent with an effective injunction, then
others must desist from using it or pay the owner’s asking fee for a license.

However, the law does not always give the owner a choice between injunc-
tion and damages. Sometimes the law allows damages but not injunction. If
the owner defends his patent by suing for damages, then, instead of licensing,
others may infringe and pay a price set by the court, not the owner’s asking
price.25 Thus in eBay v. Merc, the U.S. Supreme Court held that when a
24  In addition to patent law, U.S. firms have an additional remedy against infringing imports.
The U.S. firm can bring a case before the U.S. International Trade Commission asking it to ban the
import of all infringing goods. This powerful remedy against foreign infringement allegedly leads
to abuse by firms shielding themselves from foreign competition. See K. William Watson, Still a
Protectionist Trade Remedy:  The Case for Repealing Section 337, No. 708 Policy Analysis (Cato
Institute) (2012).
25  Legal scholars say that property rights cannot be taken without paying the owner’s subjective
price, whereas the law allows other rights to be taken by paying an objective price. An implication of
this fact is that sellers will not sell property for less than it is worth to them, and buyers will not buy
property for more than it is worth them. Consequently, voluntary sales move property from people
who value it less to people who value it more, which yields a surplus that can benefit buyer and seller.
By creating surpluses, sales allocate resources to their most productive uses and reward innovators.

6.17
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

patented invention is a small component in a large product, compensatory

damages can compensate sufficiently for future use of the patented good, and
no injunction will be given by the courts. Consequently, the owner receives
compensation instead of getting paid whatever she asks. The decision effec-
tively compels the owner of a patent on a small component in a large product
to license its use at a price close to compensatory damages.

The owner’s right to choose between damages and an injunction is at least as

strong a remedy as the right to damages without any choice of an injunction.
The following inequality orders these two remedies by strength:

compensation or injunction > compensation.

The preceding inequalities characterize different remedies for infringement

by their strength. To maximize growth, how strong should the remedy be?
The separation principle distinguishes between static (consumption and pro-
duction) and dynamic (innovation) activities. Infringements for static uses
should have strong remedies, in order to transfer more resources from static
to dynamic activities and increase the pace of innovation. For past infringe-
ments by consumers and producers, the separation principle favors treble
damages instead of a choice between disgorgement or compensation, and
it favors a choice between disgorgement or compensation instead of com-
pensation without a choice. For future infringements by consumers and
producers, the separation principle favors the choice of compensation or an
injunction instead of compensation without a choice.

The fertility principle distinguishes between the usual innovations and very
fertile innovations. Infringements for dynamic uses involving the usual in-
novations should have weak remedies in order to reduce the excess burden on
innovators. Consequently, the fertility principle usually ranks remedies the
opposite of the separation principle. For past infringements by innovators,
compensation without a choice is preferred to a choice between disgorge-
ment or compensation, and a choice between disgorgement or compensation
is preferred to treble damages. For future infringements by innovators, the
fertility principle favors compensation without a choice instead of the choice
of compensation or an injunction. In contrast to the usual cases, unusual
cases concern infringements of patents for fundamental innovations that are

6.18
 Robert Cooter

very fertile. Here the fertility principle ranks remedies the same as the sepa-
ration principle.

Table 6.2 summarizes these conclusions about the preferred remedy for eco-
nomic growth.

Table 6.2. Preferred Strength of Remedy for Infringement (“P”)

Time of Infringements
past infringements future infringements
consuming or strong strong
Infringing producing treble P choice P compensation injunction P compensation
Use innovating, usual weak weak
case compensation P choice P treble compensation P injunction
innovating, very strong strong
fertile case treble P choice P compensation injunction P compensation

Table 6.2 compares compensatory damages to the strength of two other rem-
edies. Compensatory damages also differ in strength according to their basis
of computation. Production costs and development costs often figure as
elements in compensation for infringement. Compensation can include the
cost of the invention’s production and development, which suggests this or-
dering of compensatory damage measures for patent infringement:

production and development costs > development costs > no compensation.

The same reasoning as in Table 6.2 motivates the relationship between the
infringing use and the preferred measure of compensatory damages that Ta-
ble 6.3 depicts.

Table 6.3. Preferred (“P”) Damages for Infringement

Damages for infringment

consuming or high
producing production & development P development P nil
Infringing
Use innovating, usual low
case nil P development P production & development
innovating, very high
fertile case production & development P development P nil

6.19
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

To illustrate the use of the tables, return to the case of eBay v. Merc, where
the court held that compensatory damages can compensate sufficiently for
infringing a patent on a small component in a large product. This decision
is apparently consistent with Table 6.2, which indicates that a patent owner
should receive compensatory damages for infringement by innovators in the
usual case. The “usual case” involves an innovator’s infringement of a patent
with low or average fertility, and a “small component in a large product”
presumably has low or average fertility. Furthermore, Table 6.3 indicates
that compensatory damages should preferably exclude development costs or
extraordinary profits. (Notice that the rule of compensatory damages repli-
cates the cost structure in a patent pool, which eliminates market power and
deadweight loss.26) In the exceptional case where one patented innovation is
much more fertile than others, however, Table 6.2 requires a strong remedy
– injunction or high damages.

Thickets and Holdup

Bell Labs developed the computer operating system Unix in the 1960s. Sub-
sequently, many different programmers created variants of Unix, including
the version powering Apple computers (BSD developed at the University of
California at Berkeley). Chapter 5 described circular transactions among in-
novators as an “innovation thicket.” In an innovation ticket, everyone needs
the innovations of others in order to make their own innovations, as with
variants of Unix. We will apply Tables 6.1, 6.2, and 6.3 to patent thickets.

Unix is “unowned” or “open access” or “in the public domain.” Its users are
unrestricted by patent or copyright, although it is trademarked.27 Since no
one privately owns Unix, using it does not require negotiating with anyone.
If someone owned Unix, however, anyone who wanted to improve it would
have to license its use. Negotiating a license takes time and effort, which re-
duces profits and slows innovation. An innovation thicket contains so many
circular transactions that bargaining over property rights slows growth. In-
novations in a thicket ideally remain in the public domain where anyone can
use them freely, like Unix. If not in the public domain, then innovations in a
26  In a patent pool, anyone is free to produce and use another’s patented good, without paying a
license fee. So anyone selling its patented good to another member of the pool must charge a price
that does not exceed the good’s production cost. If courts award compensation for production and
not development among innovators in a patent pool, costs replicate a patent pool.
27  Only users conforming to certain specifications can apply the name “Unix”.

6.20
 Robert Cooter

thicket should, according to the fertility principle, receive weak patent rights
(narrow breadth in Table 6.1, weak remedy in Table 6.2, low damages in Ta-
ble 6.3). In the unusual case of a pioneering patent in a thicket, however, its
high fertility warrants strong patent rights (wide breadth in Table 6.1, strong
remedy in Table 6.2, high damages in Table 6.3).

Three different kinds of private transactions among innovators can some-

times overcome the patent thicket problem. Large firms often accumulate
a rich patent portfolio. When firm A needs to use a patent in B’s portfolio,
sometimes B needs to use a patent in A’s portfolio, so they can negotiate cross
licenses. Or, instead of waiting until the need arises, cross licensing can be
negotiated in advance. Each member of a patent pool agrees in advance to
cross-license its patents for a given technology to everyone else.28 Or, instead
of forming a pool, owners of complimentary patents can merge into a single
firm that holds all of the patents required to develop an innovation.29

Each of these solutions – cross licensing, pools, and mergers -- has limited
success in solving the patent hold out problem. 30 In practice, firms cross
license or pool a small fraction of patents, and small firms often resist merg-
ing for fear that size will impede creativity.31 Private transactions among
individuals often fail so solve the challenge of innovation thickets because of
problem that game theory explains. If innovations in a thicket are patented,
then each venture to develop an innovation needs licenses from several other
innovators. When several patent owners have the power to withhold licenses
that are necessary to develop an innovation, then the each of them can ef-
fectively veto the venture. Given multiple vetoes over a venture, every veto
holder must consent for the venture to proceed. If everyone with veto power
consents except for one, the last hold out can demand most of the venture’s
value in exchange for his consent. Since a patent holder gains by consenting
last, no one wants to consent first. As the number of patent-holders with
veto power increases, the probability increases that someone will hold out for
a higher license fee.

28  A subsequent chapter on contracts discusses cross licensing and pools.

29  A subsequent chapter on antitrust discusses mergers among innovators.
30  Each firm wants to give the pool its least valuable patents, and each firm wants to use the most
valuable innovations of others in the pool. This is the problem of adverse selection.
31  Merging dilutes the concentration of ownership needed to solve the double trust problem, as
mentioned in Chapter 3 and examined further in a subsequent chapter on antitrust.

6.21
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

Behind this bargaining story lies a technical fact. When distributing the
value of a joint product, the usual rule in economics is for each participant
to receive his marginal product. The marginal product of a participant in a
cooperative venture equals the increase in the venture’s value when he par-
ticipates in it (holding the participation of others constant).32 In a patent
thicket, many people own essential patents for further innovation. Conse-
quently, the holdout’s marginal product is large, possibly equal to the ven-
ture’s entire value or nearly so. Furthermore, anyone can hold out who owns
an essential patent for the venture. In these circumstances, the sum of the
marginal products of the contributors to the venture exceeds the venture’s
total value. The venture does not have enough value for each contributor
to receive his marginal product. Demands that are reasonable individually
are unreasonable collectively. If the participants negotiate and each one de-
mands his marginal product, the venture is doomed.33 (For the same reason,
marginal product provides no basis for computing compensatory damages or
reasonable fees.34)

Given its intractability, private actors need law’s help to solve the hold out
problem. Law can help by weakening patents among innovators in a thicket.
Narrowing a patent’s breadth weakens its market power and lowers the price
that a license commands. In Apple v. Motorola, Judge Posner recently fol-
lowed this prescription by dismissing Apple’s infringement claims. Law can
also help to solve the holdout problem by weakening the remedy for patent
infringement. As mentioned above, in eBay v. Merc, the U.S. Supreme Court
held that the weaker remedy – damages instead of injunctions – applies to
small components in a complex innovation.

32  This is usually equivalent to the decrease in the venture’s value when he stops participating in it
(holding the participation of others constant).
33  In more technical language, increasing returns to the scale of a coalition implies that the sum of
the marginal contributions of each member exceeds the coalition’s total product, and the game’s “core
is empty”.
34  Damages for using another’s property often equal the marginal product that the owner lost
when the infringer used his property (compensatory damages), or else damages equal the marginal
product that the trespasser got from using another’s property (disgorgement damages). In a thicket,
“compensatory damages” or a “reasonable fee” for using a patented good without a license cannot
refer to its marginal product, because the sum of the marginal products exceeds the total product.
Marginal productivity theory is unreasonable for computing damages from infringement or negotiat-
ing shares in cross-licenses.

6.22
 Robert Cooter

Middlemen and Trolls

“Middlemen” refers to intermediaries who buy goods for resale. Throughout
history, people have viewed middlemen with suspicion because they make
money without making goods. Confusion abounds about their economic
role. Do they increase or decrease prices to consumers? Do they create cartels
or match buyers and sellers?35 The same confusion infects discussions about
patent middlemen, who buy patents to enforce them, not to produce or
innovate. According to the Norwegian fairy tale of the “Three Billy Goats
Gruff”, a troll living under a bridge gobbled up pedestrians who tried to
cross. Similarly, some law firms gobble up patents and sue others who need
them for production or innovation. “Patent troll” is the derogatory term for
“patent middlemen.”

Name calling aside, law requires analysis of patent middlemen. After a busi-
ness venture develops and patents an innovation, the innovation must be
marketed to users and enforced against infringers. Innovating, marketing,
and enforcing are different specialties that different people can perform.
With respect to enforcement, the innovator can directly defend her patent or
she can indirectly defend by selling the right to collect damages to someone
else. Thus firm A invents and patents α. A can enforce the patent on α, or
A can sell α to B, or A can sell to B only the right to sue and collect damages
from infringers of α.

By facilitating middlemen, the law gives patent owners more enforcement

choices. Innovators tend to choose the most profitable form of enforcing pat-
ents, whether by themselves or others. More enforcement choices strengthen
the patent owner’s rights de facto, but not de jure. Stronger patent rights
can stimulate or stifle innovation, as previous discussions show. According
to the separation principle, increasing a patent’s strength against consumers
and producers stimulates innovation. By prosecuting infringers who use an
invention to produce consumer goods, patent middlemen transfer resourc-
35  Middlemen often save buyers and sellers the cost of finding each other, but they sometimes
form cartels to separate buyers and sellers, and to extract higher prices. Thus middlemen try to pre-
vent buyers from obtaining information about sellers from any source but themselves. To illustrate,
the American Bar Association historically prohibited its members from advertising their specialties,
including printing a specialty on a business card. Consequently, a person who needed, say, a patent
attorney often had to ask the family attorney for a reference. This restriction on information gen-
erated referral fees for lawyers. Such predatory practices by middlemen cause hostility and resent-
ment among those who must pay for it. The U.S. courts eventually struck down such legal rules as
violations of “commercial free speech”.

6.23
Ver. 1.4 - Chapter 6. Separation and Fertility in Intellectual Property Law

es from consuming to innovating, which speeds progress in the useful arts.

To increase progress in the useful arts, law should facilitate sales of patent
rights to middlemen for innovations used in consumption or production.
Conversely, increasing a patent’s strength imposes a deadweight loss on inno-
vators in the usual case. By prosecuting infringers who use an invention to
innovate, patent middlemen impose a deadweight loss on innovators, which
slows progress in the useful arts. To increase progress in the useful arts, law
should inhibit sales of patent rights to middlemen for innovations used in
innovation. In the usual case of a very fertile patent, however, law should
facilitate sales of patent rights to middleman. Perhaps the derogatory word
“troll” should be reserved for patent middlemen who stifle innovation.

Conclusion
Intellectual property law, which includes patents and copyright law, establish-
es the ownership of innovations by people. It conveys a bundle of rights to
creators as determined by rules. The normative question of growth econom-
ics is, “Which ownership rules maximize innovation?” In order to increase
the pace of innovation, ownership rules should open competition to innovate
and increase the winner’s profits. This chapter poses the question, “Which
ownership rules maximize profits among ventures in open competition to
innovate?” Answering the normative question of growth economics requires
comparing two opposite effects of stronger intellectual property rights. The
first effect is a transfer from static activities (consumption and production)
to dynamic activities (innovation), which increases the profitability and pace
of innovation. According to the overtaking theorem explained in Chapter
1, transfers from static activities to dynamic activities increase the welfare of
people. So the first effect favors strong intellectual property against infringe-
ment by consumers and producers. The second effect is the dead-weight
loss of licensing dynamic activities, which decrease the profitability and pace
of innovation. So the second effect favors weak intellectual property rights
against infringement by innovators.

6.24