Liability Experiments:
Seller’s or Buyer’s?
March 19, 2002
Takao Kusakawa and Tatsuyoshi Saijo
kusakawa@iser.osaka-u.ac.jp saijo@iser.osaka-u.ac.jp
Research Institute of Economy, Trade and Industry
(RIETI), Tokyo, Japan,
Global Industrial and Social Progress Research Institute
(GISPRI), Tokyo, Japan; and
ISER, Osaka University, Osaka, Japan
1
1. Introduction and Summary
Basic Question : Is Emissions Trading really Cost-Effective?
Our Previous Experiments
Experiment 1 : (13 sessions, 78 subjects, 1998)
• Reversible and No Time Lag Investment
• Seller’s Liability
=> Extremely High Efficiency
Experiment 2 : (12 sessions, 72 subjects, 1999)
• Irreversible and Time Lag Investment
• Seller’s Liability
=> Two Cases:
(1) Success Case and (2) Bubble Case
2
Our Focus is
Experiment 3: (18 sessions, 90 subjects, 2001)
• Irreversible and Time Lag Investment
• Seller’s vs. Buyer’s Liability
Two Buyer’s Liability Systems: (the order is VERY important!)
• Kyoto-First: Retire Permits to Compliance Committee
=> Settle promises among countries
• Country-First: Settle promises among countries
=> Retire Permits to Compliance Committee
3
2. Emissions Trading
Efficiency
0%
$100
$10
Japanese
Russian
Reduction 1 Reduction1
Total Reductions = 2
4
Japanese
Benefit
$40
Efficiency
100%
Russian Payment
Profit
from JPN
10
to RU
$50
$60
1 $20
Russian Reductions = 2
Total Reductions = 2
Efficiency
50%
JPN
Benefit
$20
RU $50
Profit 10
Payment
from JPN
to RU
$25
$30
$15
JPN
Russian
Reductions = 1.5 = 0.5
Total Reductions = 2
5
Marginal Abatement Cost Curve
Marginal
Abatement
Cost and
Price
P
Marginal Abatement
Cost Curve
Payment to
Emission
Permits
0
Assinged Amount
Amounts of
Emissions
Status Quo
(Business as usual)
6
MAC
MAC
Country
1's MAC
<sell>
Supplier's
Surplus
Country 2's
MAC
P*
P*
1's Status Quo
(BAU)
01
Assigned Amount
02
Emissions
Demander's
Surplus
2's Status
Quo (BAU)
<buy>
Emissions
Assigned Amount
P
Country 2's
MAC
Country
1's MAC
<sell>
P*
<buy>
S
Assigned Amount
D
7
3. Point Equilibrium
Excessive reduction
P
Country
1's MAC
Country 2's
MAC
Price drops
Position after
Domestic Reduction
P*
P**
<sell>
<buy>
S
D
Assigned Amount
Point Equilibrium Price:
Market clearing price at each point of time
8
4. Experimental Design for Experiment 3
Common features to all sessions
• Ten student subjects in each session
• Used realistic marginal abatement cost curves
• Every subject could be a buyer and a seller depending on
the prices. Bohm (1997)
• We paid subjects money that was proportional to the
earnings in experiment.
9
Experimental Controls: Trading Methods and Information
• Bilateral Trading: A pair negotiates the price and quantity
vs.
• Double Auction:
Buyers’ Bids
Sellers’ Asks
(3) $56, 20 units (6)$104, 15 units
(1) $86,13 units (4)$92, 20 units
(2) grabs (4)’s ask
:
:
Trading Methods
Information of contracts
(subject #, p and q)
Bilateral
Trading
Double
Auction
Open
Closed
-------------10
5. Experimental Control: Liabilities
Seller’s Liability vs. Buyer’s Liability
Country A
Country B
Sold to
Country B
Assigned
Emissions
amount
22
10
Emissions
20
Assigned
amount
15
5
11
Seller’s liability
All the units purchased are absolutely valid for the buyer.
Country A
Country B
Sold to B
8
10
Emissions
20
12
8 units of
non-compliance
Purchased
10
Emissions
15
5
Exactly
complied
12
Buyer’s liability (Kyoto-First)
Some units purchased may be invalid for the buyer.
Country A
Surplus
after A's
2
compliance
Used in A's
compliance
Country B
Invalid
8
20
8
8
Valid 2
15
5
Exactly
complied
8 units of
non-compliance
13
2012
2008
Compliance
Committee
Seller's
Liability
C. C.
A.B C.
Permits
Trade
Permits
A.B C.
A.B C.
Permits
Annex B
Countries
Time
Buyer's
Liability
A.B C.
Trade
Promises
C. C.
Country Redeem
Permits
First
Promises
A.B C.
C. C.
Kyoto
First Permits
A.B C.
A.B C.
Redeem
Promises
C. C.
Permits
A.B C.
14
Other Rules
Default:
No monetary compensation
Non-compliance:
Penalty of $250 per unit => No Borrowing
Over-compliance:
Surplus has no value
=> No Banking
15
• Is Over-Selling beneficial?
A.B C.
Country-First
Redeem
Promises
Default =>
No Compensation
C. C.
Permits
Kyoto-First
A.B C.
Non-compliance
=> Penalty
When a country sold more bonds than her assigned amount,
Default =>
No compensation
Default =>
No compensation
Country-First
Non-compliance
Not quite
=> Penalty
Kyoto-First
Compliance
Yes!
=> No penalty
16
5. Results
Seller’s Liability: Two Cases
17
Country-First Buyer’s Liability: Three Cases
18
Kyoto-First Buyer’s Liability: Four Cases
19
Success Case
Seller's Liability, Bilateral Trading
Information Open, Second Session
Price
250
Contract
200
Point Eq. Price
Competitive Eq. Price
150
Expectation of Price
100
50
0
0
20
40
60
Minutes
80
100
20
Bubble Case
250
Seller's Liability, Double Auction, Second
Session
Price
200
150
100
50
0
0
20
40
60
Minutes
80
100
21
Anti-Bubble Case
Buyer's Liability (Country-First) , Bilateral
Trading Information Closed, Second Session
250
Price
200
150
100
50
0
0
20
40
60
Minutes
80
100
22
Intentional Bankruptcy Case
Buyer's Liability (Kyoto-First), Bilateral
Trading Information Closed, Second Session
250
Contract of
Russian Bond
Price
200
150
100
50
0
0
20
40
60
Minutes
80
100
23
24
7. Concluding Remarks
(a) Four Cases:
(1) Success Case
Seller's
(2) Bubble Case
(3) Anti-Bubble Case
(4) Intentional Bankruptcy Case
Country
First
Kyoto
First
(b) Country-First is better than Kyoto-First (Incentives)
(c) Which is better between Seller's and Country-First?
(1) Statistically, no difference (need more experiments!)
(2) If we can design some mechanism to eliminate the failure
case, it seems that Seller's is better than Country-First (?)
25