| Abstract: |
An electoral cycle created by governments is a phenomenon that seems to
characterise, at least in some particular occasions and/or circumstances, the
democratic economies. As it is generally accepted, the short-run
electorally-induced fluctuations prejudice the long-run welfare. Since the
very first studies on the matter, some authors offered suggestions as to what
should be done against this electorally-induced instability. A good
alternative to the obvious proposal to increase the electoral period length is
to consider that voters abandon a passive and naive behaviour and, instead,
are willing to learn about government’s intentions. The electoral cycle
literature has developed in two clearly distinct phases. The first one
considered the existence of non-rational (naive) voters whereas the second one
considered fully rational voters. It is our view that an intermediate approach
is more appropriate, i.e. one that considers learning voters, which are
boundedly rational. In this sense, one may consider neural networks as
learning mechanisms used by voters to perform a classification of the
incumbent in order to distinguish opportunistic (electorally motivated) from
benevolent (non-electorally motivated) behaviour of the government. The paper
explores precisely the problem of how to classify a government showing in
which, if so, circumstances a neural network, namely a perceptron, can resolve
that problem. |