Transcript Lecture 3 - Faculty Directory | Berkeley-Haas

Outline
An overview of trading institutions
Research opportunities
An in-class experiment
Smith’s first experimental economics paper (1962)
Excess-Demand (Walrasian) versus Excess-Rent
Hypotheses
Monopoly and Market Contestability
February 3 , 2005
Experimental Economics
Teck H. Ho
1
History
 Chamberlin (1948)
 Smith (1962)
 Smith (1964)
 Plott and Smith (1978)
 Bellman et. al (1957)
 Hoggatt (1959)
 Sauermann and Selten (1959)
 Siegel and Fouraker (1960; 1963)
 Rapoport and Chammah (1963)
February 3 , 2005
Experimental Economics
Teck H. Ho
2
Dimensions of Trading Institutions
 Simple  Complex
 Seller posted price (A single price offer and a yes or no response)
double auctions (no restrictions on the timing and order of price
messages and response)
 Three dimensions
 No. of sellers / No. of buyers
 Who makes price proposals
 How decisions are made (sequential or simultaneous)
February 3 , 2005
Experimental Economics
Teck H. Ho
3
Laboratory Market Trading Institutions
S/No
1
Name
8
Posted offer
auction
Posted bid
auction
Discriminative
auction
Competitive
auction
Clearinghouse
auction
Cournot quantity
choice
Walrasian
auction
Dutch auction
9
English auction
2
3
4
5
6
7
No. of Sellers /
No. of Buyers
-/-
Who Makes
Price Proposals
Sellers
Decisions
Sequential or Simultaneous
Offers posted simultaneously
How Contracts
Are Confirmed
Buyers shop in sequence
-/-
Buyers
Bids posted simultaneously
Sellers shop in sequence
1/-
Buyers
Bids posted simultaneously
1/-
Buyers
Bids posted simultaneously
-/-
Auctioneer
Price adjusted sequentially
1/-
Seller clock
Price lowered sequentially
1/-
Buyers
Price raised sequentially
Highest N bidders
pay own bid prices
Highest N bidders
pay N+1st price
Intersection of
bid and offer arrays
Intersection of
total quantity and demand
Confirmation when
excess demand is zero
Buyer confirmation
stops clock
Sale to highest bidder
-/-
Buyers and sellers Bids and offers simultaneously
-/-
Endogenous price Seller quantities simultaneously
10
Bid auction
-/-
Buyers
Prices raised sequentially
Sellers
11
Offer auction
-/-
Sellers
Prices lowered sequentially
Buyers
12
Double auction
-/-
Both types
Both types
13
Decentralized
Negotiation
List/discount
-/ -
Both types
Bids raised and offers lowered
sequentially
Sequential but decentralized
-/-
Sellers
Simultaneous (list), sequential
(discounts)
Buyers
14
February 3 , 2005
Experimental Economics
Both types
Teck H. Ho
4
Posted Prices (1,2,3)
(buyers pay different prices)
The market for this commodity is organized as follows: we open the market for each trading
period. Each seller decides on a price offer, which he or she will write on one of the cards provided.
The sellers will be given two minutes to submit their prices. After all sellers have chosen prices, the
cards will be collected and the prices written on the blackboard.
Buyers will then be free to make bids to purchase whatever quantities they desire and to specify
the seller from whom they wish to buy. Bids will be made as follows: a buyer will be chosen using
random numbers, and will indicate the seller and a desired quantity. The designed seller will then
accept any part of the buyer’s bid by stating the quantity he or she wishes to sell. However, a seller
who posts a price must be prepared to sell at least one unit at that price. If the first seller selected
will not sell all units the buyer wants to purchase, the buyer is free to choose a second seller, and so
on.
When the first buyer has made all desired purchases, another buyer will be selected at random
and will make bids in the same manner. The process will be continued until all buyers have had a
chance to make purchases. This completes the trading period. We will reopen the market for a new
trading period by having sellers submit new prices, and the process will be repeated. Except for the
bids and their acceptance, you are not to speak to any other subjects. Are there any questions?
(1)
(3)
February 3 , 2005
One seller and one buyer ultimatum game
One prize  first-price auction
Experimental Economics
Teck H. Ho
5





Rules: Customer bids, waits up to 1
hour. If bid is accepted, ticket must
be purchased. If not accepted,
customer must wait 7 days to reenter bid.
Restrictions: Cannot specify time
of flight, connection city or airline.
Refund Policy: No
Refunds/Exchanges.
Hotels, Rental Cars: Yes
Loyalty Miles or Points: No
February 3 , 2005





Rules: Mid-October launch.
Customer can see fare and has 30
minutes to accept offer. Customers
will be locked out for 72 hours if
they reject offer.
Restrictions: Cannot specify time
of flight, connection city or airline.
Refund Policy: No
Refunds/Exchanges.
Hotels, Rental Cars: Hotels by later
this fall, car rentals by early 2001.
Loyalty Miles or Points: No
Experimental Economics
Teck H. Ho
Source: USAToday, 9/29/00
6
Laboratory Market Trading Institutions
S/No
1
Name
8
Posted offer
auction
Posted bid
auction
Discriminative
auction
Competitive
auction
Clearinghouse
auction
Cournot quantity
choice
Walrasian
auction
Dutch auction
9
English auction
2
3
4
5
6
7
No. of Sellers /
No. of Buyers
-/-
Who Makes
Price Proposals
Sellers
Decisions
Sequential or Simultaneous
Offers posted simultaneously
How Contracts
Are Confirmed
Buyers shop in sequence
-/-
Buyers
Bids posted simultaneously
Sellers shop in sequence
1/-
Buyers
Bids posted simultaneously
1/-
Buyers
Bids posted simultaneously
-/-
Auctioneer
Price adjusted sequentially
1/-
Seller clock
Price lowered sequentially
1/-
Buyers
Price raised sequentially
Highest N bidders
pay own bid prices
Highest N bidders
pay N+1st price
Intersection of
bid and offer arrays
Intersection of
total quantity and demand
Confirmation when
excess demand is zero
Buyer confirmation
stops clock
Sale to highest bidder
-/-
Buyers and sellers Bids and offers simultaneously
-/-
Endogenous price Seller quantities simultaneously
10
Bid auction
-/-
Buyers
Prices raised sequentially
Sellers
11
Offer auction
-/-
Sellers
Prices lowered sequentially
Buyers
12
Double auction
-/-
Both types
Both types
13
Decentralized
Negotiation
List/discount
-/ -
Both types
Bids raised and offers lowered
sequentially
Sequential but decentralized
-/-
Sellers
Simultaneous (list), sequential
(discounts)
Buyers
14
February 3 , 2005
Experimental Economics
Both types
Teck H. Ho
7
Uniform Price Auctions (4,5,6,7)



Uniform price
 everyone pays the same N+1th price
 appearance of fairness
Competitive auction (4) (e.g., COE in Singapore)
http://www.aas.com.sg/coe/coe4.htm
Call market or clearinghouse (5)
 Both buyers and sellers submit price bids, offers, quantity limits in a
systematic manner
 Market clearing price is determined by the intersection of the demand
and supply functions obtained by arraying bids and offers in order
 Used to provide the daily opening price on NYSE

Cournot quantity choice / Walrasian auction
February 3 , 2005
Experimental Economics
Teck H. Ho
8
Laboratory Market Trading Institutions
S/No
1
Name
8
Posted offer
auction
Posted bid
auction
Discriminative
auction
Competitive
auction
Clearinghouse
auction
Cournot quantity
choice
Walrasian
auction
Dutch auction
9
English auction
2
3
4
5
6
7
No. of Sellers /
No. of Buyers
-/-
Who Makes
Price Proposals
Sellers
Decisions
Sequential or Simultaneous
Offers posted simultaneously
How Contracts
Are Confirmed
Buyers shop in sequence
-/-
Buyers
Bids posted simultaneously
Sellers shop in sequence
1/-
Buyers
Bids posted simultaneously
1/-
Buyers
Bids posted simultaneously
-/-
Auctioneer
Price adjusted sequentially
1/-
Seller clock
Price lowered sequentially
1/-
Buyers
Price raised sequentially
Highest N bidders
pay own bid prices
Highest N bidders
pay N+1st price
Intersection of
bid and offer arrays
Intersection of
total quantity and demand
Confirmation when
excess demand is zero
Buyer confirmation
stops clock
Sale to highest bidder
-/-
Buyers and sellers Bids and offers simultaneously
-/-
Endogenous price Seller quantities simultaneously
10
Bid auction
-/-
Buyers
Prices raised sequentially
Sellers
11
Offer auction
-/-
Sellers
Prices lowered sequentially
Buyers
12
Double auction
-/-
Both types
Both types
13
Decentralized
Negotiation
List/discount
-/ -
Both types
Bids raised and offers lowered
sequentially
Sequential but decentralized
-/-
Sellers
Simultaneous (list), sequential
(discounts)
Buyers
14
February 3 , 2005
Experimental Economics
Both types
Teck H. Ho
9
One-sided Sequential Auctions (8,9,10,11)

Dutch Auctions
http://www.landsend.com

English auctions
http://auctions.samsclub.com/?n=0

Bid auctions

Offer auctions (e.g., airline post prices in a computer system))
February 3 , 2005
Experimental Economics
Teck H. Ho
10
Laboratory Market Trading Institutions
S/No
1
Name
8
Posted offer
auction
Posted bid
auction
Discriminative
auction
Competitive
auction
Clearinghouse
auction
Cournot quantity
choice
Walrasian
auction
Dutch auction
9
English auction
2
3
4
5
6
7
No. of Sellers /
No. of Buyers
-/-
Who Makes
Price Proposals
Sellers
Decisions
Sequential or Simultaneous
Offers posted simultaneously
How Contracts
Are Confirmed
Buyers shop in sequence
-/-
Buyers
Bids posted simultaneously
Sellers shop in sequence
1/-
Buyers
Bids posted simultaneously
1/-
Buyers
Bids posted simultaneously
-/-
Auctioneer
Price adjusted sequentially
1/-
Seller clock
Price lowered sequentially
1/-
Buyers
Price raised sequentially
Highest N bidders
pay own bid prices
Highest N bidders
pay N+1st price
Intersection of
bid and offer arrays
Intersection of
total quantity and demand
Confirmation when
excess demand is zero
Buyer confirmation
stops clock
Sale to highest bidder
-/-
Buyers and sellers Bids and offers simultaneously
-/-
Endogenous price Seller quantities simultaneously
10
Bid auction
-/-
Buyers
Prices raised sequentially
Sellers
11
Offer auction
-/-
Sellers
Prices lowered sequentially
Buyers
12
Double auction
-/-
Both types
Both types
13
Decentralized
Negotiation
List/discount
-/ -
Both types
Bids raised and offers lowered
sequentially
Sequential but decentralized
-/-
Sellers
Simultaneous (list), sequential
(discounts)
Buyers
14
February 3 , 2005
Experimental Economics
Both types
Teck H. Ho
11
Double Auctions

The experiment in the first week was a double oral auction

At least 1000 sessions had been conducted

It is the most efficient trading mechanism so it is often used as
a benchmark
February 3 , 2005
Experimental Economics
Teck H. Ho
12
Comparisons of Market Efficiency (%)
Study
Double Auction
Posted Offer
Davis and Williams (1986)
96
82
Ketcham, Smith, and
Williams (1984)
97
94
Davis, Harrison, and
Williams (1993)
97
66
Davis and Williams (1991)
98
92
Smith, Williams, Bratton,
and Vannoni (1982)
95
Friedman and Ostroy (1989)
96
Negotiated Prices Posted Price With
Negotiations
89
90
92
Hong and Plott (1982)
97
Davis and Holt (1994)
84
February 3 , 2005
Clearing-House
Experimental Economics
83
Teck H. Ho
13
The Most Significant Results
There are no experimental results more important or more
significant than that the information specifications of traditional
competitive price theory are grossly overstated. The experimental
facts are that no double auction trader needs to know anything
about valuation conditions of other traders or have any
understanding or knowledge of market supply and demand
conditions, or have any trading experience (although experience
may speed convergence) or satisfy the quaint and irrelevant
requirement of being a price “taker” (every trader is a price
maker in the double auction)
www.google.com
February 3 , 2005
Experimental Economics
Teck H. Ho
14
Laboratory Market Trading Institutions
S/No
1
Name
8
Posted offer
auction
Posted bid
auction
Discriminative
auction
Competitive
auction
Clearinghouse
auction
Cournot quantity
choice
Walrasian
auction
Dutch auction
9
English auction
2
3
4
5
6
7
No. of Sellers /
No. of Buyers
-/-
Who Makes
Price Proposals
Sellers
Decisions
Sequential or Simultaneous
Offers posted simultaneously
How Contracts
Are Confirmed
Buyers shop in sequence
-/-
Buyers
Bids posted simultaneously
Sellers shop in sequence
1/-
Buyers
Bids posted simultaneously
1/-
Buyers
Bids posted simultaneously
-/-
Auctioneer
Price adjusted sequentially
1/-
Seller clock
Price lowered sequentially
1/-
Buyers
Price raised sequentially
Highest N bidders
pay own bid prices
Highest N bidders
pay N+1st price
Intersection of
bid and offer arrays
Intersection of
total quantity and demand
Confirmation when
excess demand is zero
Buyer confirmation
stops clock
Sale to highest bidder
-/-
Buyers and sellers Bids and offers simultaneously
-/-
Endogenous price Seller quantities simultaneously
10
Bid auction
-/-
Buyers
Prices raised sequentially
Sellers
11
Offer auction
-/-
Sellers
Prices lowered sequentially
Buyers
12
Double auction
-/-
Both types
Both types
13
Decentralized
Negotiation
List/discount
-/ -
Both types
Bids raised and offers lowered
sequentially
Sequential but decentralized
-/-
Sellers
Simultaneous (list), sequential
(discounts)
Buyers
14
February 3 , 2005
Experimental Economics
Both types
Teck H. Ho
15
Decentralized Negotiation /
List/Discount

Decentralized negotiation
 A dominant mode of selling in some countries

List / Discount
 Common in retail markets
 “Marked-down” accounts for a majority of sales
February 3 , 2005
Experimental Economics
Teck H. Ho
16
Research Opportunities
?
February 3 , 2005
Experimental Economics
Teck H. Ho
17
Outline
An overview of trading institutions
Research opportunities
An in-class experiment
Smith’s first experimental economics paper (1962)
Excess-Demand (Walrasian) versus Excess-Rent
Hypotheses
Monopoly and Market Contestability
February 3 , 2005
Experimental Economics
Teck H. Ho
18
Outline
An overview of trading institutions
Research opportunities
An in-class experiment
Smith’s first experimental economics paper
Excess-Demand (Walrasian) Hypothesis versus ExcessRent Hypothesis
Monopoly and Market Contestability
February 3 , 2005
Experimental Economics
Teck H. Ho
19
Demand and Supply Schedules
February 3 , 2005
Experimental Economics
Teck H. Ho
20
Symmetric Case
Trading Period
1
2
3
4
5
February 3 , 2005
Predicted
Exchange
Quantity (x0)
6
6
6
6
6
Average Actual Coefficient of
Predicted
Actual Exchange Exchange Price Exchange Price Convergence [a =
Quantity (x)
(¯P¯)
(P0)
(100s0)/(P0)]
5
2
1.80
11.8
5
2
1.86
8.1
5
2
2.02
5.2
7
2
2.03
5.5
6
2
2.03
3.5
Experimental Economics
No. of
Submarginal
Buyers Who
Could Make
Contracts
No. of
Submarginal
Buyers Who
Made Contracts
No. of
Submarginal
Sellers Who
Could Make
Contracts
No. of
Submarginal
Sellers Who
Made Contracts
5
5
5
5
5
0
0
0
1
0
5
5
5
5
5
0
0
0
1
0
Teck H. Ho
21
Coefficient of Convergence
n

2
(
P

P
)
 i o
i 1
P0
100%
P0  Predicted Equilibriu m Price
February 3 , 2005
Experimental Economics
Teck H. Ho
22
Rate of Convergence
and Steady State Price Bias
Excess Demand (Walrasian) Hypothesis
Speed of convergence is proportional to excess demand
Excess Rent Hypothesis
Speed of convergence is proportional to excess rent
Difference in Buyer and Seller Rents
The difference affects speed of convergence
Psteady state – P0 is proportional to buyer rent – seller rent at
time t
February 3 , 2005
Experimental Economics
Teck H. Ho
23
Excess-Demand, Excess-Rent, and
Difference in Buyer and Seller Rents
Price
S
D
0
Area = A t
P
P0
Area = x2t
Pt
x1t
D
0
Area = Bt
S
0
A0t - B t = x3t
Quantity
February 3 , 2005
Experimental Economics
Teck H. Ho
24
Excess-Demand, Excess Rent, and
Convergence
Trading Period
1
2
3
February 3 , 2005
Predicted
Exchange
Quantity (x0)
15
15
15
Predicted
Actual Exchange Exchange Price
Quantity (x)
(P0)
16
3.425
15
3.425
16
3.425
Average Actual
Coefficient of
Exchange Price Convergence [a =
(¯P¯)
(100s0)/(P0)]
3.47
9.9
3.43
5.4
3.42
2.2
Experimental Economics
No. of
Submarginal
Buyers Who
Could Make
Contracts
No. of
Submarginal
Buyers Who
Made Contracts
No. of
Submarginal
Sellers Who
Could Make
Contracts
No. of
Submarginal
Sellers Who
Made Contracts
4
4
4
2
2
2
3
3
3
1
1
0
Teck H. Ho
25
Excess-Demand, Excess Rent, and
Convergence
Trading Period
1
2
3
4
Predicted
Exchange
Quantity (x0)
16
16
16
16
February 3 , 2005
Actual Exchange
Quantity (x)
17
15
15
15
Predicted
Exchange Price
(P0)
3.5
3.5
3.5
3.5
Average Actual
Coefficient of
Exchange Price Convergence [a =
(¯P¯)
(100s0)/(P0)]
3.49
16.5
3.47
6.6
3.56
3.7
3.55
5.7
No. of
Submarginal
Buyers Who
Could Make
Contracts
No. of
Submarginal
Buyers Who
Made Contracts
No. of
Submarginal
Sellers Who
Could Make
Contracts
No. of
Submarginal
Sellers Who
Made Contracts
5
5
5
5
1
0
0
0
6
6
6
6
2
1
0
0
Experimental Economics
Teck H. Ho
26
Difference in Buyer and Seller Rents
(Buyer Rent > Seller Rent)
1
2
3
4
Predicted
Exchange
Quantity (x0)
10
10
10
10
1
2
3
8
8
8
Trading Period
February 3 , 2005
Average Actual
Exchange Price
(¯P¯)
9
9
9
9
Predicted
Exchange Price
(P0)
3.1
3.1
3.1
3.1
3.53
3.37
3.32
3.32
Coefficient of
Convergence [a =
(100s0)/(P0)]
19.1
10.4
7.8
7.6
8
7
6
3.1
3.1
3.1
3.25
3.30
3.29
6.9
7.1
6.5
Actual Exchange
Quantity (x)
No. of
Submarginal
Buyers Who
Could Make
Contracts
No. of
Submarginal
Buyers Who
Made Contracts
No. of
Submarginal
Sellers Who
Could Make
Contracts
No. of
Submarginal
Sellers Who
Made Contracts
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
None
Experimental Economics
Teck H. Ho
27
Difference in Buyer and Seller Rents
(Seller Rent > Buyer Rent)
Trading Period
1
2
3
4
5
6
Predicted
Exchange
Quantity (x0)
9
9
9
9
9
9
February 3 , 2005
Actual Exchange
Quantity (x)
8
9
9
8
9
9
Predicted
Exchange Price
(P0)
3.4
3.4
3.4
3.4
3.4
3.4
Average Actual
Exchange Price
(¯P¯)
2.12
2.91
3.23
3.32
3.33
3.34
Coefficient of
Convergence [a =
(100s0)/(P0)]
49.1
22.2
7.1
5.4
3
2.7
Experimental Economics
No. of
Submarginal
Buyers Who
Could Make
Contracts
No. of
Submarginal
Buyers Who
Made Contracts
No. of
Submarginal
Sellers Who
Could Make
Contracts
No. of
Submarginal
Sellers Who
Made Contracts
3
3
3
3
3
3
1
0
1
0
0
0
None
None
None
None
None
None
None
None
None
None
None
None
Teck H. Ho
28
Experience and Monetary Incentive
1
2
3
4
Predicted
Exchange
Quantity (x0)
10
10
10
10
1
2
12
12
Trading Period
February 3 , 2005
Average Actual
Exchange Price
(¯P¯)
11
9
10
9
Predicted
Exchange Price
(P0)
3.125
3.125
3.125
3.125
3.12
3.13
3.11
3.12
Coefficient of
Convergence [a =
(100s0)/(P0)]
2
0.7
0.7
0.6
12
12
3.45
3.45
3.68
3.52
9.4
4.3
Actual Exchange
Quantity (x)
Experimental Economics
No. of
Submarginal
Buyers Who
Could Make
Contracts
No. of
Submarginal
Buyers Who
Made Contracts
No. of
Submarginal
Sellers Who
Could Make
Contracts
No. of
Submarginal
Sellers Who
Made Contracts
7
7
7
7
0
1
1
0
7
7
7
7
0
0
0
0
4
4
0
0
3
3
2
0
Teck H. Ho
29
Imbalance between no. of intra-marginal
sellers & no. of intra-marginal buyers
1
2
3
4
Predicted
Exchange
Quantity (x0)
12
12
12
12
1
2
12
12
Trading Period
February 3 , 2005
Average Actual
Exchange Price
(¯P¯)
12
12
12
12
Predicted
Exchange Price
(P0)
10.75
10.75
10.75
10.75
5.29
7.17
9.06
10.90
Coefficient of
Convergence [a =
(100s0)/(P0)]
53.8
38.7
21.1
9.4
11
6
8.75
8.75
9.14
--------
11
--------
Actual Exchange
Quantity (x)
Experimental Economics
No. of
Submarginal
Buyers Who
Could Make
Contracts
No. of
Submarginal
Buyers Who
Made Contracts
No. of
Submarginal
Sellers Who
Could Make
Contracts
No. of
Submarginal
Sellers Who
Made Contracts
5
5
5
5
3
3
2
0
None
None
None
None
None
None
None
None
4
4
1
1
None
None
None
None
Teck H. Ho
30
Trading Volume versus Number of
Periods
1
2
3
Predicted
Exchange
Quantity (x0)
18
18
18
1
20
Trading Period
February 3 , 2005
18
18
18
Predicted
Exchange Price
(P0)
3.4
3.4
3.4
20
3.8
Actual Exchange
Quantity (x)
Average Actual
Coefficient of
Exchange Price Convergence [a =
(¯P¯)
(100s0)/(P0)]
2.81
21.8
2.97
15.4
3.07
13.2
3.52
10.3
Experimental Economics
No. of
Submarginal
Buyers Who
Could Make
Contracts
No. of
Submarginal
Buyers Who
Made Contracts
No. of
Submarginal
Sellers Who
Could Make
Contracts
No. of
Submarginal
Sellers Who
Made Contracts
6
6
6
3
2
2
None
None
None
None
None
None
4
3
2
0
Teck H. Ho
31
Excess-Demand versus Excess-Rent
Price
S
D
0
Area = A t
P
P0
Area = x2t
Pt
x1t
D
0
Area = Bt
S
0
A0t - B t = x3t
Quantity
February 3 , 2005
Experimental Economics
Teck H. Ho
32
Walrasian versus Excess-rent
 Modified Walrasian (excess-demand)
Pt  Pt 1  Pt     13 x1t    x3t
 Modified Excess-rent
Pt  Pt 1  Pt   4   24 x2t  34 x3t
February 3 , 2005
Experimental Economics
Teck H. Ho
33
Results
February 3 , 2005
Experimental Economics
Teck H. Ho
34
Results
February 3 , 2005
Experimental Economics
Teck H. Ho
35
Smith (JPE, 1965)
Price
D
4.20
S
P0= 3.10
Excess
supply
Quantity
February 3 , 2005
Experimental Economics
Teck H. Ho
36
Experimental Design
11 buyers
February 3 , 2005
Experimental Economics
Teck H. Ho
37
Results (e=2)
February 3 , 2005
Experimental Economics
Teck H. Ho
38
Results (e=8)
February 3 , 2005
Experimental Economics
Teck H. Ho
39
Excess Supply and Excess Rent
Excess
Supply = e
Excess
rent = (Pt – P0) e
Regression:
Pt 1  Pt 1  Pt
     ( Pt  P0 )e  e
February 3 , 2005
Experimental Economics
Teck H. Ho
40
Regression Results
February 3 , 2005
Experimental Economics
Teck H. Ho
41
The main conclusion
The odds favoring excess rent are over 300 to 1
February 3 , 2005
Experimental Economics
Teck H. Ho
42
Outline
An overview of trading institutions
Research opportunities
An in-class experiment
Smith’s first experimental economics paper (1962)
Excess-Demand (Walrasian) Hypothesis versus ExcessRent Hypotheses
Monopoly and Market Contestability
February 3 , 2005
Experimental Economics
Teck H. Ho
43
Monopoly
 Index of Monopoly Effectiveness
Actual Profit - Competitiv e Profit
M
Monopoly Profit - Competitiv e Profit
February 3 , 2005
Experimental Economics
Teck H. Ho
44
Double Auction versus Posted-offer Auctions
February 3 , 2005
Experimental Economics
Teck H. Ho
45
Monopoly Effectiveness in
Posted-Offer Auctions
Study
Subject Experience
Buyer Type
Cost Function
Monopoly M Value
?
Human
Increasing
1
Isaac, Ramey, and Williams (1984)
Inexperienced
Human
Increasing
0.45
Coursey, Isaac, and Smith (1984)
Experienced
Human
Decreasing
0.56
Design Experience
Simulated
Decreasing
0.72
Harrison, McKee, and Rustrom (1989)
Inexperienced
Simulated
Decreasing
0.44
Harrison, Mckee and Rustrom (1989)
Design and role
experience
Simulated
Decreasing
0.78
Smith (1981) (1 session)
Harrison and McKee (1985)
February 3 , 2005
Experimental Economics
Teck H. Ho
46
Empirical Regularities on Monopoly




Pricing in posted-offer monopolies is higher
than in double auction monopolies
Prices in poster-offer monopolies is above
competitive levels, but on average, profits are
significantly below theoretical monopoly
levels
Experience helps
Monopoly pricing is facilitated by constant or
decreasing costs.
February 3 , 2005
Experimental Economics
Teck H. Ho
47
Potential Competition as a Regulator:
Market Contestability

Clark (1887) emphasized the role of latent
competition

Would it be as effective as actual competition
in restraining monopoly pricing?

The importance of the absence of sunk costs
February 3 , 2005
Experimental Economics
Teck H. Ho
48
A Contestable Market
(Baumol, Panzar, and Willig, 1982)

There is at least one potential rival with the same cost
structure

Potential entrants evaluate the profitability of entry at the
incumbent firm’s prices

There are no barriers to entry or exit.
February 3 , 2005
Experimental Economics
Teck H. Ho
49
A Contestable Market
(Baumol, Panzar, and Willig, 1982)

The result is that a contestable market can only be in
equilibrium if the prices and quantities of the incumbent firm
are sustainable, which in turn requires that no new firm with
the same cost function as the incumbent can earn a profit by
charging a lower price and earning a profit by selling all or
part of the demand at that lower price.

For the decreasing cost structure, any equilibrium must
involve the incumbent choosing the price equals average
cost.
February 3 , 2005
Experimental Economics
Teck H. Ho
50
Experimental Test: Coursey, Issac, and Smith
(1984, Journal of Law and Economics)



Four monopoly and six duopoly sessions
The market lasts for 18 periods
Identical cost and demand conditions in both the
monopoly and duopoly conditions
February 3 , 2005
Experimental Economics
Teck H. Ho
51
Results
February 3 , 2005
Experimental Economics
Teck H. Ho
52
Zero Sunk Cost

The “competitive price” is the lowest price that yields
nonnegative earnings for the incumbent

In six experimental contested markets conducted by
Coursey, Isaac, and Smith (1984), four yielded competitive
price outcomes, and the other two exhibited downward
trends in price, with price deviations from the competitive
level being 50% less than the monopoly price deviations (at
period 18)
February 3 , 2005
Experimental Economics
Teck H. Ho
53
New Experimental Design : Coursey, Issac,
Luke, and Smith (1984; RJE)







Introduce sunk cost in the nature of a five-period operating permit.
The market lasts for 25 periods
A coin flip determines whether a seller is of Type A or Type B.
Seller A is the incumbent who was required to begin the experiment
by purchasing seller permits for periods 1-5 and in periods 6-10.
Seller A is the protected monopolist in periods 1-5
Seller B could observe prices and her own cost technology for periods
1-5, and was allowed to contest the market (by purchasing the permit
in any period beginning with period 6).
Seller A could choose to continue to contest the market by purchasing
a new permit in period 11 or any period thereafter.
February 3 , 2005
Experimental Economics
Teck H. Ho
54
Positive Sunk Cost

12 sessions (6 with simulated buyers and six with human
buyers)

Prices were closer to the benchmark than to the natural
monopoly level.

Prices converge to the competitive level in half of the sessions
February 3 , 2005
Experimental Economics
Teck H. Ho
55
An Experimental Study of
Three Internet-Based Market Institutions
February 3 , 2005
Experimental Economics
Teck H. Ho
56
Three Internet Pricing Mechanisms
Sellers posting prices (e.g., www.amazon.com)
Buyers posting prices (e.g., www.priceline.com)
Sellers posting minimum acceptable prices
(MAP) and buyers posting prices (a modified
version of www.priceline.com)
February 3 , 2005
Experimental Economics
Teck H. Ho
57
Performance Criteria
 Social Welfare Metrics
 Seller surplus
 Buyer surplus
 Inefficiency
 Marketing Metrics
 Average price
 Sales volume
 Price Posting Behavior
 Seller posted price
 Buyer posted price as a function of WTP
February 3 , 2005
Experimental Economics
Teck H. Ho
58
A Hypothetical Market
WTP
Buyer 1
Buyer 2
Buyer 3
Buyer 4
Buyer 5
Buyer 6
Seller 1
c=0
Seller 2
c=0
40
20
15
90
60
50
 Two sellers and six buyers
 Each seller has 2 units to sell and each buyer has a WTP for
a single unit of the product
February 3 , 2005
Experimental Economics
Teck H. Ho
59
Maximum Possible Surplus (MPS)
WTP
Buyer 1
Buyer 2
Buyer 3
Buyer 4
Buyer 5
Buyer 6
Seller 1
c=0
Seller 2
c=0
40
20
15
90
60
50
MPS = 40 + 90 + 60 + 50 = 240
February 3 , 2005
Experimental Economics
Teck H. Ho
60
Seller Posted Prices: An Example
Price
Seller 1
c=0
Buyer 1
Buyer 2
Buyer 3
Buyer 4
Buyer 5
Buyer 6
45
40
40
Seller 2
c=0
WTP
40
20
15
90
60
50
Seller surplus
= 45 + 40 + 40 = 125 = 52.08% of MPS
Buyer surplus
= 45 + 20 + 10 = 75 = 31.25% of MPS
Inefficiency
= 40 = 16.67% of MPS
Average Price
= 41.67
Sales Volume
=3
February 3 , 2005
Experimental Economics
Teck H. Ho
61
Experimental Design
Nine experimental sessions
Eight subjects per session; 2 sellers and 6 buyers
Standard experimental economics methodology (i.e.,
induced value/WTP, monetary incentives)
Seller’s earnings = sum of (price – cost) for each unit sold
Buyer’s earning = WTP - price
Subjects made $20 on average, the actual payoffs ranged
from $8 to $39 (a typical experiment lasted for 75 minutes)
February 3 , 2005
Experimental Economics
Teck H. Ho
62
Design of Amazon Market
Determining the role of each subject
Determining each buyer's WTP for the product
Determining each seller's price
Determining each buyer's buying sequence
Determining whether there is a trade between each
buyer and seller
 According to the buying sequence, each buyer is sequentially
asked to indicate whether she would buy from each of the
sellers
Determining the earnings for everyone
February 3 , 2005
Experimental Economics
Teck H. Ho
63
Design of Priceline Market
Determining the role of each subject
Determining each buyer's WTP for the product
Determining each buyer's price
Determining each buyer's buying sequence
Determining whether there is a trade between each
buyer and seller
 According to the buying sequence, sellers are asked whether
they would sell a unit to each buyer
Determining the earnings for everyone
February 3 , 2005
Experimental Economics
Teck H. Ho
64
Design of Priceline Market with MAP
 Determining the role of each subject
 Determining each buyer's WTP for the product
 Determining each seller’s minimal acceptable price (MAP)
 Determining each buyer's price
 Determining each buyer's buying sequence
 Determining whether there is a trade between each buyer and
seller
 Buyer prices that are below all MAPs are rejected
 According to the buying sequence, seller(s) are asked whether they would
sell a unit to each buyer whose price is above their MAPs
 Determining the earnings for everyone
February 3 , 2005
Experimental Economics
Teck H. Ho
65
Results: Social Welfare Metrics
Amazon Priceline
Seller
50.60% 45.09%
Surplus
Buyer
33.23% 42.58%
Surplus
Inefficiency 16.17% 12.33%
Priceline
with MAP
45.70%
42.33%
11.97%
Inefficiencies can result from a rule that limits the ability of
Experimental Economics
February 3 traders
, 2005
to deviate from posted
prices (Plott and Smith, 1978)
Teck H. Ho
66
Results: Social Welfare Metrics
from Prior Studies
Double Auction Posted Offer
Trading Institutions
Clearing House
Negotiated Prices Posted Price with
Subsequent Negotiations
Davis and Williams(1986)
96
82
Ketcham, Smith, and Williams (1984)
97
94
Davis, Harrison, and Williams (1993)
97
66
Davis and Williams (1991)
98
92
Smith, Williams, Bratton, and
Vannoni (1982)
95
89
Friedman and Ostroy (1989)
96
90
Hong and Plott (1982)
87
Davis and Holt (1994)
94
February 3 , 2005
Experimental Economics
92
83
Teck H. Ho
67
Results: Marketing Metrics
Price
Sales
Volume
Amazon Priceline Priceline
with MAP
28.43
43.00
33.00
3.00
3.75
3.33
• Red is statistically different from black or green
• Green is statistically different from black
February 3 , 2005
Experimental Economics
Teck H. Ho
68
Average Prices Over Time
Prices Over Time
60
50
Price
40
Amazon
30
Priceline
20
10
0
1
2
3
4
5
6
7
8
9
10
Time
February 3 , 2005
Experimental Economics
Teck H. Ho
69
Amazon: Bertrand Competition with
Capacity Constraints
Amazon: Price Frequency Plot
50.0%
45.0%
40.0%
35.0%
30.0%
Frequency 25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
< 38
38-45
>46
Price Range
It can proven that there is a unique mixed strategy equilibrium (Bechmann, 1965):
90% of the probability weights are assigned to prices between 34 and 37
February 3 , 2005
Experimental Economics
Teck H. Ho
70
Seller Posted Price (Amazon) versus
MAP (Priceline with MAP)
Seller Posted Price versus MAP Over Time
70
60
Posted Price
or MAP
50
40
MAP
Posted Price
30
20
10
0
1
2
3
4
5
6
7
8
9
10
Time
February 3 , 2005
Experimental Economics
Teck H. Ho
71
Summary
 Social Welfare Metrics
 Seller surplus: amazon > priceline and priceline with MAP
 Buyer surplus: amazon < priceline and priceline with MAP
 Inefficiency: amazon > Priceline and priceline with MAP
 Marketing Metrics
 Average price: priceline < priceline with MAP < amazon
 Sales volume: priceline > amazon and priceline with MAP
February 3 , 2005
Experimental Economics
Teck H. Ho
72
Other Extensions
groups.haas.berkeley.edu/simulations
Effects of non-price variables (e.g., product
quality)
Effects of brand loyalty
February 3 , 2005
Experimental Economics
Teck H. Ho
73