changes in tastes or constraints on the supply side8 or as a consequence
of other consumers’ decisions.
Each of these assumptions is far from being a matter of course. It is in
the first place not self-evident that households are engaged in utility maximi-
sation. Also unlikely is the prevalence of a constant utility function over all
goods and representative for all households serving as a precise description
of consumer decisions in quantitative terms.
Ad 1: The assumption amounts to postulating an arguably idealised (ex-
treme), infinitely flexible and active homo oeconomicus. The two Irish au-
thors [25]Murphy and Garvey (2005) found out that contrary to their most
plausible expectations, poor households were less adversely affected by infla-
tion than the higher income groups. The received wisdom is that households
who are better off have a greater range of choice among goods which are
not only necessities so that they are supposed to be more flexible and more
inclined to substitute goods that became relatively dearer. However, surpris-
ingly their consumption pattern exhibited less flexibility than the one of less
affluent consumers. The reason for this paradox seems to be that the rich may
have been discouraged by the disutilities and (psychic) costs of search activ-
ities involved in constantly rationally rearranging ones consumption. They
may be interested in avoiding inconvenience and they are in the comfortable
situation that they can afford it. On the other hand the relatively poor may
be forced to be more flexible due to their disadvantaged economic situation.
Hence preferences and flexibility may in actual fact be a product of one’s
economic situation rather than exogenously determined. Not only prefer-
ences but also the propensity to actively rearrange consumption in response
to price movement can be income or class determined.
Ad 2: In order to transform a preference order into a real valued utility
function и = f (q) according to which utility u depends only on the vector q
of quantities of market purchased goods we have to assume a metric variable
u and that marginal changes of quantities can be made that will sensibly
affect и. This implies that the function f (q) has got to be a continuous,
8For example a good is no longer available or offered only with some new features not
wanted by the consumer. Ideally none of the phenomena that characterises a dynamic
market should take place: goods disappear and new goods emerge on the market, quality
changes etc. For all those reasons consumers may be forced to involuntarily change their
demand induced by activities on the supply side of the market.