DECISION-MAKING THEORY AND RESEARCHof the commons.’’ The tragedy was that individuals
tried to maximize what they could get from the
common and, or the ‘‘commons,’’ which resulted
in the commons being overused, thereby becom-
ing depleted. If each individual had only used his
allotted share of the commons, then it would have
continued to be available to everyone.
Prisoner’s Dilemma. The best-known social
dilemma is the prisoner’s dilemma (PD), which
involves two individuals (most often, although
formulations with more than two people are possi-
ble). The original PD involved two convicts’ deci-
sion whether or not to confess to a crime (Rapoport
and Chammah 1965). But the following example is
functionally equivalent.
Imagine you are selling an item to another
person, but you cannot meet to make the ex-
change. You agree to make the exchange by post.
You will send the other person the item, and
receive the money in return. If you both do so,
then you each get 3 units (arbitrary amount, but
amounts received for each combination of choices
is important). However, you imagine that you
could simply not put the item in the post, yet still
receive the money. Imagine doing so results in you
getting 5 units and the other person −1 units.
However, the other person similarly thinks that
not posting the money would result in getting the
item for free, which would result in 5 units for the
other person and −1 for you. If you both do not put
anything in the post, although you agreed to do so,
you would be at the status quo (0 units each).
Do you post the item (i.e., cooperate) or not?
Regardless of what the other person does, you will
get more out of not cooperating (5 v. 3, when the
other person cooperates, and 0 v. −1, when other
does not). However, if you both do not cooperate,
that produces an inferior group outcome, com-
pared to cooperating (0 [0+0] v. 6 [3+3], respective-
ly). Thus, the dilemma is that each individual has
an incentive to not cooperate, but the best out-
come for the group is obtained when each person
cooperates. Can cooperation develop from such a
situation?
Axelrod (1984; cf. Hofstader 1985) investigat-
ed that question by soliciting people to participate
in a series of PD games (social dilemmas are often
referred to as games). Each person submitted a
strategy for choosing to cooperate or not over a
series of interactions with the other strategies.
Each interaction would result in points being award-
ed to each strategy, and the strategy that generated
the most points won. The winning strategy was Tit
for Tat. It was also the simplest strategy. The Tit
for Tat strategy is to cooperate on the first turn,
and then do whatever the other person just did
(i.e., on turn x, Tit for Tat will do whatever its
opponent did on turn x−1).Axelrod suggested that four qualities led to Tit
for Tat’s success. First, it was a nice strategy, be-
cause it first cooperates, and Tit for Tat will coop-
erate as long as the other person cooperates. But
when the other person does not cooperate, then it
immediately retaliates. That is, it responds to non-
cooperation with noncooperation, which illustrates
its second good quality. Tit for Tat is provocable,
because it immediately reacts to noncooperation,
rather than waiting to see what will happen next,
or ignore the noncooperation. However, if the
opponent goes back to cooperating, then Tit for
Tat will also go back to cooperating. That is quali-
ty three: forgiveness. Tit for Tat will not contin-
ue punishing the other player for previous
noncooperations. All that counts for Tit for Tat is
what just happened, not the total amount of non-
cooperation that has happened. Finally, Tit for Tat
has clarity, because it is simple to understand. A
complex strategy can be confusing, so it may be
misunderstood by opponents. If the opponent’s
intentions are unclear, then noncooperation is
best, because if or when a complex strategy is
going to be cooperative cannot be predicted.Thus, a cooperative strategy can be effective
even when there are clear incentives for noncoop-
eration. Furthermore, Axelrod did another com-
puter simulation in which strategies were reward-
ed by reproducing themselves, rather than simply
accumulating points. Thus, success meant that the
strategy had more of its kind to interact with.
Again, Tit for Tat was best, which further suggests
that a cooperative strategy can be effective and can
flourish in situations that seem to be designed for
noncooperation.