Bounded rationality: A parable
Often people accuse economists of relying on an imperfect view of human nature – a view that is often disparagingly called “homo-economicus“. However, our view of the world does not rely on the individuals being able to perfectly calculate the satisfaction they get from things or perceive every opportunity.
Ultimately all we require is that individuals respond to incentives (derived from the relative costs and benefits of things) and as a result try to set themselves up such that the benefits of their actions exceed the costs (include opportunity costs). Such a situation can occur when individuals are bounded rational.
A good example of bounded rationality comes from the parable of the centipede and its 100 legs. In the parable there is a centipede walking along. Now it co-ordinates its 100 legs without thinking and without calculating – it does so just because it has become part of its nature. Effectively, it is following a rule of thumb which allows it to make the optimal decision.
When asked about how it moves its legs it gets confused and trips – at this level anyone watch the centipede would take this as a failure of rationality, it obviously can’t deal with the complex calculation required to come up with its optimal solution. However, is this really fair? Even though the centipede couldn’t consciously decide how to move its legs, over time it had evolved a rule of thumb that had allowed it to achieve the optimal solution.
Economists “believe” (this is our value judgment here) that individuals create rules of thumb that allow them to achieve the optimal outcome – without having to consciously calculate.
Understanding how these rules of thumb evolve and change in a highly complex and changeable world is a big deal (in the parable the rule of thumb does not shift favourably to the introduction of information!), and is something behavioural and neurological economists are currently studying. However, it is clear from this individuals can still make optimal decisions, even when they aren’t expert calculators with complete information.
Economists “believe” (this is our value judgment here) that individuals create rules of thumb that allow them to achieve the optimal outcome – without having to consciously calculate.
Matt, perhaps I misunderstand what you’re saying, but surely this is a factual question not a value judgment. How do values come into describing the rules of thumb people use in market transactions?
Thought you might be interested:
FRIEDMAN – BERNANKE
@ben
Sorry that I have been unclear.
The value judgment is that these rules of thumb evolve and move in a way that leads to optimal outcomes. We are describing the purpose for the rule in a way which requires making a value-laden statement – specifically because the evidence is still being working on (both for and against).
While rules of thumb can sometimes create optimal results, there are no guarantees. The centipede’s walking problem isn’t a particularly difficult one. It is also well suited to simple mechanical control that evolution is good at mastering.
What is more interesting to me is where the centipede chooses to walk. If it goes in the wrong direction, it gets picked off by birds and other predators. This is a much harder problem. The rules of thumb a centipede uses to decide where to go work sometimes, but not always.
The centipede clearly doesn’t optimize its decision of where to go.
@Matt Nolan
Isn’t the value judgment in deciding the value of the purpose? I’m pretty sure the relationship between a set of rules and the distribution of outcomes they produce can be described without reference to anyone’s values.
@ben
But we can’t observe the rules – so the value judgment is the belief that the rules are formed to take us from the individuals decision process to an optimal outcome.
There is no room for multiple equilibrium in beliefs or systematic biases in formed rules – both of which might be relevant.
@Stephen Monrad
I’m not sure that the centipedes walking decision can’t be seen as formed from a rule of thumb that is effectively optimal (given the adjustment cost of re-evaluating the rule).
Simply put, the centipedes that have a better rule will not get eaten – and as a result will breed. If rules are genetic or taught then the optimal rule will eventually dominate.
What about some real life examples? In 7th Form Biology we did experiments with slaters (Porcellio scaber) which showed that they had “rules of thumb” (we called them taxes and kineses) to, for instance, prefer dark places (to hide from predators), damp places (so they don’t dry out) and prefer to walk downhill (so they don’t end up at the top of trees, etc).
I found http://www.tvhe.co.nz very informative. The article is professionally written and I feel like the author knows the subject very well. http://www.tvhe.co.nz keep it that way.