cf. Paul Ormerod _The Death of Economics_ .
Ormerod argues that a system of hundereds of equations that econometricians use to describe an economy do not have a unique solution (which represents the "equilibrium" of that system in econo-babble); the system is solved by iterations i.e. trial and error approximations that solve the equations within a certain margin or error. It is, however, possible that another solution of that system, perhaps even within a more narrow margin of error exists, but it cannot be found through that particular iterative process (which involves a rather small number of iterations that stop as soon the error reaches an apriori specified margin). Looking at it from a different angle, there are a set of numbers of which we do not know whether they are or are not solutions to a given set of equations.
The practical implication of the Ormerod's argument is that the neoclassical sacred cow, Pareto's optimum (meaning finding a unique solution of a system of multiple utility functions) is a fiction - there are many possible, yet unknown, solutions.
wojtek