A Behavioral Economics experiment demonstrates People's Unselfishness
Behavioral Economics is a fast-emerging, cross-disciplinary alternative to neo-classical economics.
It seeks inspiration not from mathematical formulae, but from
studying how people actually behave.
Its results challenge the classical assumptions of homo economicus, that
people are 100% rational, perfectly informed self-maximisers.
|A typical experiment is the so-called 'Ultimatum Game', which
involves two people (A and B) and 10 coins.
Person A has 10 coins and places some of them in front of him,
and the rest in front of person B, who is
asked whether or not he accepts the proposed allocation.
If B says 'Yes', then both A & B keep the coins in front of them.
if B says 'No', then neither player gets any of the coins.
The standard tool for analysing this game is Game Theory.
The first step is to map out all the possible outcomes, creating a so-called game tree:
Each outcome is awarded a score, called a payoff, (A,B)
where the first number
reflects the player A's winnings, and the second player B's winnings. Once all the possible
game outcomes have been evaluated, the next step is to work up the tree from the bottom.
For example, look in detail at the (8,2) branch.
Person B has a choice between a payoff of (8,2) or (0,0). Saying 'Yes' gives him
2 coins (in which case A gets 8), whereas saying 'No' means neither player gets any.
Game theory assumes (as do most economic models) that individuals are self-maximising
- that irrespective of all else, they seek the most for themselves.
In this example, since 2>0,
the theory predicts that player B would say 'Yes', resulting in a payoff of (8,2).
Since player A can see this, he is able to simplify the game tree by ignoring
branches that the theory predicts that player B will never play.
This results in the simplified game tree shown below:
Also a self-maximiser, player A now chooses which of the available outcomes is best for him;
he places 9 coins in font of him, and just a single coin in front of player B:
This is a fine theory, but unfortunately it's not how real people play the game.
How do people actually play?
Player A usually splits the coins nearly evenly, averaging around 6:4, and player B
usually accepts this offer. Interestingly, autistic players are the only ones who consistently split the
coins 9:1, as game theory predicts.
Player B usually accepts the allocation when the coins are fairly evenly split.
However, with unfair allocations, especially when offered only a single coin,
very few people make the 'logical' decision to accept the offer.
Why don't people actually play as the theory predicts?
The answer to this question should be obvious to anyone who has
not been trained in game theory. If player B accepts a very unequal offer,
this would mean rewarding player A's greed, so instead of cooperating
he is prepared to suffer a loss himself.
Basically, people like things to be fair,
and if they don't perceive what is going on as being fair, they are prepared to
suffer in order to punish those they see as the source of the unfairness.
In a world in which the gap between the materially rich and materially poor is growing ever wider,
this phenomenon - referred to as 'altruistic punishment' - is of great relevance to us all.
In fact, it makes sense to ask the question the other way round:
"Why not use a theory which predicts how people actually play?"
In reality, people do care about the welfare of others,
so we are developing
altruistic economics to deal with this fact.
We're not trying to fit the people to the model, but fit the model to the people.
|Downloads||Title||Author(s)|| Date ||Reference|
| Myth: Homo economicus is a valid assumption of human behavior|| Steve Kangas|