joint with Yudistira Permana and Nuttaporn Rochanahastin (University of York)

Way back in 1955 Herbert Simon made a call for a new kind of economics stating that:
“the task is to replace the global rationality of economic man with a kind of rational behavior that is compatible with the access to information and the computational capacities that are actually possessed by organisms, including man, in the kinds of environment in which such organisms exist”.
Since this call, many economists have tried to produce theories of ‘satisficing’ behaviour – as Simon called it – but most have failed. The trouble is that the expression ‘rational behaviour’ covers virtually all forms of behaviour, as long as it is motivated by some ‘rational’ objective function, and the decision-maker has all relevant information available to him or her, and the decision-maker can perform all the necessary calculations costlessly. A ‘rational’ objective function is one that is not internally inconsistent, though it may not be complete. In recent years economics has started seeing theories of behaviour with incomplete preferences; these seem to be a step in the right direction. There has also been an outburst of theoretical work on decision-making under ambiguity – which is a situation where the probabilities of future outcomes either do not exist or are not known to the decision-maker. Some of these make the decision-makers task very complicated – for example the Smooth Model of decision-making under ambiguity assumes that, while the decision-maker (DM) does not know the probabilities, he or she can specify the set of possible probabilities and can attach probabilities to each member of the set. The computational problems are enormous – but usually economics assume that the DM can solve any task, however complex.
But this is not true: numerous experiments have shown that there is noise in human behaviour: that given the same problem more than once, decisions change. Is this simple error in decision-making, or is it a sign that the decision-maker cannot or does not want to find the best decision? The problem may simply be too complicated. The economist would argue that the DM realises that it is not worth thinking about the problem, and that he or she is simply trading off costs and benefits. But we have to be careful here: if we want to argue that there are costs of working out the optimal solution, we need to admit that there must also be costs associated with working out whether it is worth working out the optimal solution, and that there must be also costs associated with working out whether it is worth working out whether it is worth working out the optimal solution, and so on ad infinitum. There is no end to this infinite regression.
But admitting that there are costs to thinking is a first step in the right direction. Such costs are the crucial component of a new theory advanced by Charles Manski in a working paper with the title “Optimise, Satisfice or Choose Without Deliberation?”. In this he incorporates the first cost – the cost of thinking, or, as he calls it, the costs of deliberation. It is just the first step as these costs are given, are exogenous; and he ignores ‘higher level’ costs – the costs of thinking about thinking, the costs of thinking about thinking about thinking, and so on. He considers three decisions that the DM can make: optimise (by incurring a large deliberation cost to learn the action that yields the maximum welfare); satisfice (with a smaller cost of thinking to reveal whether there is some action that passes some aspiration level) and try to achieve some given aspiration level; or take a decision with no deliberation. His theory shows that satisficing is the best decision if the cost of thinking is sufficiently smaller than the cost which reveals the action that yields the maximum welfare.
This theory shows that satisficing may be best for certain costs of thinking. This theory could change the way economists think about rational decision making as it incorporates the cost of thinking. It will help the economics profession move away from the standard neo-classical economic theory by modelling the concept of satisficing in a more satisfactory way.
We report on an experiment testing Manski’s theory.

VIEW PAPER