Arrows Impossibility Theorem:
Pareto Optimality in the context of welfare economics occurs at that point where the welfare of one individual cannot be increased without reducing the welfare of the other individuals. On the other hand the movement towards Pareto Optimal Point leads to increase in welfare because in the process the welfare of at least one individual can be increased without reducing the welfare of the other individuals. For simplifying the analysis we can start with a society of two individuals. Then Pareto- improvement means that (i) there is increase in welfare of one individual while other individual stays in the original position or (ii) there is increase in the welfare of both the individuals. The analysis is based on the "Ordinalistic Approach" and it fails to explain the situation where the welfare of one individual increases at the cost of the other individuals basing it on value judgment. Subsequently the New Welfare Economics attempted to resolve this type of problem.
Through compensation criteria New Welfare Economics attempts to ascertain whether any public policy leads towards welfare gain or not without encountering the problem of value judgement comparison unlike the Pareto principle. Approaches of different economists in the context of Social Welfare Function towards public policy are also discussed such as Samuelson, Rawls and Bentham. K.J.Arrow has formulated some axioms to define rational collective choice and attempted to derive a social welfare function based on it translating individual preferences in terms of collective choice for its satisfaction but unfortunately fails stating his Impossibility Theorem. At last, from the point of view of the political aspects it is hinted that in a democratic country public policy determined by political behaviour tends to represent public choice being guided by the notion of politically sensitive individuals' welfare at the one extreme and exhibiting undue generosity towards the special interest groups (pressure groups) on the other extreme.