The Effects of Boundary Rules on Collective Action

1
The Effects of Boundary Rules on Collective Action

• T.K. Ahn
• R. Mark Isaac
• Timothy C. Salmon
• Florida Sate University

2
BOUNDARY RULES

• Ostrom, 2005
• Entry and Exit Rules
• determining their own membership is the first
  step toward developing greater trust and
  reciprocity.

3
Research Question

• How do different entry/exit rules affect
• Group Size
• Cooperation/Contribution
• Efficiency
• Congestion
• Group Stability

4
Experimental Design

• 12 subjects in each session
• Playing a repeated public good provision game (up
  to 25 rounds)
• Fully endogenous group formation
• Group size and group composition
• Under three different entry/exit rules
• Free Entry Free Exit (FEE) 4 sessions
• Restricted Entry and Free Exit (RE) 4 sessions
• Free Entry and Restricted Exit (RX) 4 sessions

5
Experimental Environment 1A Congestible Public
Good

• Buchanan (1965)
• Club good
• Some degree of publicness / partial rivalry of
  consumption

6

• Subjects divide 15 tokens between public and
  private account.
• Payoff when contributing Xi tokens to public
  account is

7
Experimental Environment 2Voluntary Provision

• The collective action problem is not completely
  solved when groups are formed
• c.f. Buchanans consumption ownership-membershi
  p arrangement
• c.f. Tiebouts voting with the feet

8

• Economists since Buchanan
• Calculate optimal group size given the degree of
  publicness
• Calculate optimal level of public good given
  group size
• Devise incentive compatible mechanism for
  preference revelation
• But, when a collective good is provided by the
  efforts of group members after the group is
  formed

9
Experimental Environment 3Endogenous Group
Formation

• In period 1, each (of 12 subjects) belong to a
  group of size 1.
• In all subsequent periods
• They are asked if they wanted to switch groups
• They were able to view a screen showing the
  average contribution and group size of each group

10
Experimental Environment 4Boundary Rules

• FEE (free entry, free exit) -4 sessions
• RE (restricted entry, free exit) 4 sessions
• Entry into a group requires approval of a
  majority of the members
• RX (free entry, restricted exit) 4 sessions
• Exit from a group requires approval of a majority
  of the members

11
Benchmarks and Conjectures

• Individual and Group Optimal Contribution

12
Benchmark 1Game Theoretic Equilibrium

• One Period Game Nash equilibrium
• Partition (3,3,3,3) or (4,4,4)
• Contribution 3 each
• Finitely Repeated Game Subgame perfect
  equilibrium
• Any outcome paths consisting of one period game
  equilibrium in each period
• Some other subgame perfect equilibria with
  contribution greater than 3

13
Benchmarks 2 Group and Social Optimum

• Group optimum
• 9 members, 11 tokens each
• Social optimum
• A (3,9) partition
• But very risky
• And the difference is small between (3,9), (4,8)
  ..(6,6)

14
Conjectures

• Contribution REgtFEE, RX
• Group size ???
• Efficiency REgt FEE, RX
• Congestion RE lt FEE, RX
• Stability RE gt FEE, RX

15
Result 1 Group Size

• Measured at the individual level
• FEE 4.70
• RE 4.49
• RX 4.77

16
(No Transcript)
17
(No Transcript)
18
Result 2 Contribution ( of group optimal given
actual group sizes)

• FEE 4.32 (64)
• RE 4.92 (74)
• RX 4.34 (63)

19
(No Transcript)
20
Result 3 Earnings (Efficiency)

• FEE 52.91 ECUs
• RE 65.11 ECUs
• RX 55.60 ECUs

21
(No Transcript)
22
Result 4 Congestion

• Congestion
• m - n, where
• m optimal group size given contribution
• n actual group size

23
(No Transcript)
24
(No Transcript)
25
Result 5 Group Stability
26
(No Transcript)
27
Additional Results 1

• Determinants of Contribution and Voting

28
(No Transcript)
29
(No Transcript)
30
Votes
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
Which Institution Favors Contributors?
35
(No Transcript)
36
Discussion

• Many real world collective action involve
  congestible public goods
• Enforceable contract is not always feasible
• Individuals still face collective action problem
  after groups are formed
• Optimal group size is a function of
  cooperativeness of the members
• Entry and Exit rules matter

37
Discussion

• When entry is restricted
• Applicants tend to raise their contributions
• Voters use the boundary rule to carefully select
  new members
• Average earning is higher (c.f., Ahn, Isaac,
  Salmon, 2005)
• Less congestion

38
Discussion Evolution of institutions and types

• Restricted entry benefits all on average (when
  public good is congested)
• Restricted entry favors contributors over free
  riders contributors (cooperators) can evolve
  under this rule

39
Discussion

• But even under restricted entry
• The outcome is fare less efficient than social
  optimal (on average)
• TO what extent contractual provision is feasible
  and necessary?