Econ 522
Economics of Law
Dan Quint
Spring 2011
Lecture 7
Last Wednesday…
 Established properties of an efficient property law system



Private goods are privately owned, public goods are not
Owners have maximum liberty over how they use their property
Injunctive relief used when transaction costs are low,
damages used when transaction costs high
 We also gave some thought to “testing Coase”


In-class experiment: can UW undergrads allocate poker chips
efficiently?
(Cost me $118)
1
Our experiment…
 Take 1: Full Information (values on nametags)
starting
allocation
efficient
allocation
actual final
allocation
12
red chip
purple chip
10
purple chip
purple chip
10
purple chip
purple chip
purple chip
red chip
purple chip
purple chip
60
56
8
red chip
8
6
fraction of potential
gains realized
purple chip
6
4
purple chip
4
purple chip
2
purple chip
32
24/28 = 86%
2
Our experiment…
 Take 2: Private Information (values hidden)
starting
allocation
efficient
allocation
actual final
allocation
10
red chip
red chip
8
purple chip
purple chip
8
purple chip
purple chip
6
red chip
6
4
fraction of potential
gains realized
purple chip
purple chip
purple chip
purple chip
4
purple chip
3
purple chip
3
purple chip
2
purple chip
24
48
44
20/24 = 83%
3
Our experiment…
 Take 3: Uncertainty
starting
allocation
efficient
allocation
actual final
allocation
chip
chip
12
12
100%
efficient
allocation
actual final
allocation
fraction of potential
gains realized
chip
chip
6
6
3 X die roll
(actually 12)
2 X die roll
(actually 8)
fraction of potential
gains realized
chip
8
 Take 4: Adverse Selection
starting
allocation
3 X die roll
(actually 6)
2 X die roll
(actually 4)
chip
4
100%
4
Our experiment…
 So…


Coase seems to work pretty well
My guess: if we redid the asymmetric information case a bunch of
times, we wouldn’t get trade very often
 Comparing “uncertainty” to “asymmetric info”…




Seller’s value was 2 X die roll, buyer’s value was 3 X die roll
If nobody knows die roll, no problem – they can trade based on the
expected value
But if seller knows die roll, problem
In strategic settings, information can have negative value –
the seller could be worse off for having information!
5
Sequential
Rationality
6
Dynamic games and sequential rationality
 Game theory we’ve seen so far: static games


“everything happens at once”
(nobody observes another player’s move before deciding how to act)
 Dynamic games


one player moves first
second player learns what first player did, and then moves
7
Dynamic games
FIRM 1 (entrant)
FIRM 2
(incumbent)
Don’t Enter
Enter
(0, 30)
Accommodate
(10, 10)
Fight
(-10, -10)
 A strategy is one player’s plan for what to do at each decision
point he/she acts at
 In this case: player 1’s possible strategies are “enter” and “don’t”,
player 2’s are “accommodate” and “fight”
8
We can put payoffs from this game into a
payoff matrix…
Firm 1’s Action
Firm 2’s Action
Accommodate
Fight
Enter
10, 10
-10, -10
Don’t Enter
0, 30
0, 30
 We can look for equilibria like before
 we find two: (Enter, Accommodate), and (Don’t Enter, Fight)
 question: are both equilibria plausible?
 sequential rationality
 firm 1 asks, “once I’ve entered, would he really choose to fight?”
9
Dynamic games
 In dynamic games, we look for Subgame Perfect Equilibria

players play best-responses in the game as a whole, but also in every
branch of the game tree
 We find Subgame Perfect Equilibria by backward induction

start at the bottom of the game tree and work our way up
FIRM 1 (entrant)
FIRM 2
(incumbent)
Don’t Enter
Enter
(0, 30)
Accommodate
(10, 10)
Fight
(-10, -10)
10
The key assumption behind subgame perfect
equilibrium: common knowledge of rationality
 Firm 1 knows firm 2 is rational
 So he knows that if he enters, firm 2 will do the rational thing
– accommodate
 So he enters, counting on firm 2 to accommodate
 This is the idea of sequential rationality – the assumption
that, whatever I do, I can count on the players moving after
me to behave rationally in their own best interest
11
Applications of
Property Law
12
Intellectual Property
 Intellectual property: broad term for ways that an individual,
or a firm, can claim ownership of information

Patents – cover products, commercial processes

Copyrights – written ideas (books, music, computer programs)

Trademarks – brand names, logos

Trade Secrets
13
Information: costly to generate,
easy to imitate
up-front investment: 1,000
monopoly profits: 2,500
duopoly profits: 450 each
 Example: new drug
 Requires investment of $1,000 to discover
 Monopoly profits would be $2,500
 Once drug has been discovered, another firm could also
begin to sell it
 Duopoly profits would be $450 each
14
Information: costly to generate,
easy to imitate
up-front investment: 1,000
monopoly profits: 2,500
duopoly profits: 450 each
FIRM 1 (innovator)
Don’t
Innovate
FIRM 2 (imitator)
(0, 0)
Imitate
(-550, 450)
Don’t
(1500, 0)
 Solve the game by backward induction:


Subgame perfect equilibrium: firm 2 plays Imitate, firm 1 plays
Don’t Innovate, drug is never discovered
(Both firms earn 0 profits, consumers don’t get the drug)
15
up-front investment: 1,000
monopoly profits: 2,500
duopoly profits: 450 each
Patents: one way to solve
the problem
 Patent: legal monopoly

Other firms prohibited from imitating Firm 1’s discovery
FIRM 1 (innovator)
Don’t
Innovate
FIRM 2 (imitator)
(0, 0)
Imitate
(-550, 450)
Don’t
(1500, 0)
450 – P
 Subgame perfect equilibrium: firm 2 does not imitate;
firm 1 innovates, drug gets developed
16
Comparing the two outcomes
up-front investment: 1,000
monopoly profits: 2,500
duopoly profits: 450 each
FIRM 1 (innovator)
FIRM 2 (imitator)
Innovate
Don’t
Without patents:

Drug never
discovered
(0, 0)
Imitate
(-550, 450)
Don’t
(1500, 0)
FIRM 1 (innovator)
 With patents:


Drug gets
discovered
But…
FIRM 2 (imitator)
Innovate
Don’t
(0, 0)
Imitate
(-550, 450 – P)
Don’t
(1500, 0)
17
Patents solve one inefficiency
by introducing another
up-front investment: 1,000
monopoly profits: 2,500
duopoly profits: 450 each
 Without patents, inefficient outcome: drug not developed
 With patents, different inefficiency: monopoly!
Monopoly
Net Surplus = 2,750
P = 50
Duopoly
Net Surplus = 3,950
CS
1,250
Profit
2,500
CS
4,050
P = 100 – Q
DWL
1,250
Q = 50
P = 10
Profit 450 x 2
DWL
50
Q = 90
 Once the drug has been found, the original incentive
problem is solved, but the new inefficiency remains…
18
Patents: a bit of history
 First U.S. patent law passed in 1790
 Patents currently last 20 years from date of application
 For a patent application to be approved, invention must be:



novel (new)
non-obvious
have practical utility (basically, be commercializable)
 Patentholder whose patent has been infringed can sue for
both damages and an injunction against future violations
 Patents are property – can be sold or licensed to others
19
Patent breadth
 Narrow patents might allow us each to patent own invention
 Broad patents might not

“Winner-take-all” race to be first
20
Patent breadth
 Does a patent on the “pioneering invention” cover the
application as well?
 Can you patent an improvement to an existing product?
21
Patent length
 Patent length




Need to last long enough for firms to recover up-front investment…
…But the longer patents last, the longer we have DWL from
monopoly
(Example from textbook: drug price drops from $15 to $1 per pill
when patent expires)
Tradeoff between ex-post inefficiency and ex-ante incentive provision
 U.S.: all patents last 20 years


Jeff Bezos (founder of Amazon) once suggested software patents
should last just 3 years
Germany: full-term patents for major inventions, 3 year “petty
patents” for minor ones, annual renewal fees
22
Do the details matter?
 Coase: without transaction costs, initial allocation of rights
irrelevant for efficiency
 But transaction costs may be high



Uncertainty on whether a patent is valid
Uncertainty of outcome of research
Many parties
23
Do the details matter?
 Coase: without transaction costs, initial allocation of rights
irrelevant for efficiency
 But transaction costs may be high



Uncertainty on whether a patent is valid
Uncertainty of outcome of research
Many parties
24
Do the details matter?
 Coase: without transaction costs, initial allocation of rights
irrelevant for efficiency
 But transaction costs may be high



Uncertainty on whether a patent is valid
Uncertainty of outcome of research
Many parties
25
Alternatives to patents for encouraging
innovation
 government purchase of drug patents
 prizes

Google $30 million prize for landing a rover on the moon
 direct government funding of research

~25% of research spending in U.S. is funded by government
26
patents
copyrights
trademarks
trade secrets
27
Copyright
 Property rights over original expressions

writing, music, other artistic creations
 Creations like this tend to fit definition of public goods



nonrivalrous
nonexcludable
so private supply would lead to undersupply
 Several possible solutions



government subsidies
charitable donations
legal rights to creations – copyrights
28
Copyright
 Copyright law less rigid than patent law

Unlike patent law, allows for certain exceptions
 Copyrights last much longer than patents

Current U.S. law: copyright expires 70 years after creator’s death
 No application process

Copyright law automatically applies to anything you’ve
written/created
 Copyrights more narrow than patents

Cover exact text, not general idea
29
Copyright
 Retelling of Gone With The Wind, from point of view of a
slave on Scarlett’s plantation, published in 2001



Margaret Mitchell’s estate sued to halt publication
Eventually settled out of court
Was there really any harm?
30
Copyright
 Retelling of Gone With The Wind, from point of view of a
slave on Scarlett’s plantation, published in 2001



Margaret Mitchell’s estate sued to halt publication
Eventually settled out of court
Was there really any harm?
31
patents
copyrights
trademarks
trade secrets
32
Trademarks
 Reduce confusion over who made a product
 Allow companies to build reputation for quality
 Don’t expire, unless abandoned
 Generic names can’t be trademarked
33
Trademarks – example
 WSJ article 9/17/2010: “Lars Johnson Has Goats On His
Roof and a Stable of Lawyers To Prove It”

Restaurant in Sister Bay WI put
goats on roof to attract customers

“The restaurant is one of the topgrossing in Wisconsin, and I’m
sure the goats have helped.”

Suing restaurant in Georgia

“Defendant has willfully continued
to offer food services from
buildings with goats on the roof”
http://online.wsj.com/article/SB10001424052748704285104575492650336813506.html
34
Trademark dilution
35
patents
copyrights
trademarks
trade secrets
36
Trade Secrets
 Protection against misappropriation
 But plaintiff must show…



Valid trade secret
Acquired illegally
Reasonable steps taken to protect it
37
patents
copyrights
trademarks
trade secrets
38