CRITERIA FOR A WELFARE IMPROVEMENT

We often need to compare different economic states, none of which may be optimal. For example, if one is doing a cost–benefit analysis of a public motorway project, one wants to compare social welfare in two ‘states of the world’: state 0, where the motorway is not built, and state 1, where the motorway is built and the money to build it is raised in some specified fashion. This is called cost–benefit analysis because in comparing state 0 with state 1 we want to assess the additional benefits generated by the project minus the alternative benefits which would have been available but have been lost due to the project (i.e., the costs of the project). In principle one can do a cost–benefit evaluation of the effects of any action, be it private (getting married?) or public (subsidizing food?). How?

The action of shifting from state 0 to state 1 is to be judged by its effects on the happiness of all those affected. One can imagine two main possible types of outcome:

1. Someone gains and no one loses (or no one gains and someone loses).

2. Someone gains but someone else loses.

The Pareto criterion

Case 1 is illustrated by the move from P0 to P1 (figure 5.1 (a)). Whatever one's social welfare function (provided it is benevolent), this must be an improvement. The jargon describes such a change as a Pareto improvement.