The name is absent



probabilities under the conjecture that every player follows the sanctioning behavior
specified by the social norm; the other is derived under the conjecture that one player
deviates from the sanctioning behavior, once.

Case 1. Off-equilibrium, everyone sanctions. Consider a player who currently selects
Z. Let ρ=(ρ1,ρ2,ρ3,ρ4) with ρd being the probability that he meets a cooperator given
that
d =1,L, 4 players currently select Z. Clearly, the probability that he meets someone
who chooses
Z is 1 - ρ. Recall that each player can be paired to three other players, with
equal probability. Therefore, we have

ρ=(1,2, 3,0)

Here, the transition matrix is:

123

4

0

2/3

1

1


010

0 1/3 0

000

000

The bold numbers in the rows (columns) indicate the number Dt (Dt+1 ) of players who
currently (next period) play
Z. Each cell represents the corresponding conditional
probability Pr
[ Dt+11 Dt ]. Clearly, Pr [2|1] = Pr [4|3] = 1 since if an odd number of players
plays
Z today, then at least one of them is paired to a cooperator. The latter will choose Z
in t + 1 . Also, Pr [4|4] = 1, since the social norm does not specify reversion to
cooperation. To see why Pr [2|2] = 1/3 and Pr [4|2] = 2/3 recall that there are three
possible pairings. One of those involves the two players who currently choose
Z. So, with
probability 1/3 the sanctioning behavior does not spread further. If that pairing is not
realized, then
Z will be necessarily seen by the remaining two cooperators. So, with
probability 2/3, next period everyone will choose the sanction, Z.

Case 2. Off equilibrium, one player does not use the prescribed sanction.

Suppose, off-equilibrium, a player who observed a deviation in the past chooses to
deviate from the sanctioning rule and plays
Y this period. Instead, everybody else follows
the social norm. Consider this player. Again,
ρd is the probability that he meets a

34



More intriguing information

1. The name is absent
2. ISSUES IN NONMARKET VALUATION AND POLICY APPLICATION: A RETROSPECTIVE GLANCE
3. The name is absent
4. Restricted Export Flexibility and Risk Management with Options and Futures
5. Three Policies to Improve Productivity Growth in Canada
6. The name is absent
7. School Effectiveness in Developing Countries - A Summary of the Research Evidence
8. The name is absent
9. Improvement of Access to Data Sets from the Official Statistics
10. WP 48 - Population ageing in the Netherlands: Demographic and financial arguments for a balanced approach