Thus the really basic issue is that we continue, in much of macroeconomic
analysis, to dismiss the aggregation problem and to treat economic
aggregates as though they correspond to rational economic individuals
although this is theoretically unjustified. It is this simple observation
that makes the structure of the models, however sophisticated, which
macroeconomists build, unacceptable. But what is worse is that in the
anxiety to preserve the scientific foundations, macroeconomists also dismiss
the questioning of the soundness of those foundations. Such questioning
comes from behavioural economics and explicitly casts doubt on
the idea that individuals themselves do behave according to the standard
‘sound microfoundations’. This is very far from new; Pareto devoted a
whole chapter of his Treatise on Sociology to what he described as
‘non-rational behavior’ (Pareto 1916). From Simon (1957) onwards,
many have suggested that individuals reason in a limited and local way.
Experiments, observation, and examination of the neural processes utilized
in making decisions all suggest that homo economicus is not an
accurate or adequate description of human decision making.6 But then
one might reasonably ask why macroeconomists insist on their ‘representative
agent’ models if the representative agent is assumed to be more
rational than those he represents.