The Second Best

Abstract

Pareto’s principles are designed to show whether or not society is in a position of optimum welfare, and, if not, what are the necessary conditions to reach such an optimum. But suppose that for some reason this ideal position is unattainable and it is known that a Paretian optimum cannot be reached. Is the best policy in these circumstances to fulfil as many of the Paretian conditions as possible in order to get nearly to a Pareto optimum, or does the second-best position lie in an entirely different direction? This controversy was the subject of a paper by Lipsey and Lancaster [1957] which attempted to clarify the general position in such cases; however, before we examine their conclusions in detail, let us consider how relevant the problem of second best is likely to be to practical decision-making.