ABSTRACT
CV12 and CV21, operation of the project producing the damaging spillover effects.
3 The cost-benefit criterion introduced in Part I of this book, that V > 0, was identified as the Kaldor-Hicks test, sometimes also referred to as a potential Pareto improvement, It may now be more precisely expressed as CV12 > 0. This is properly interpreted as requiring for its fulfilment that everyone in the community could be made better off by a costless distribution of the gains in moving from state 1 to 2. Yet, it is no less compelling to employ, instead, the alternative criterion,
CV21 < 0, which is properly interpreted as requiring for its fulfilment that everyone in the community could be made worse off by a costless distribution of the losses that are incurred in moving from state 2 back again to state 1. Admittedly, a superficial reflection would suggest that, if CV12 > 0, then
indeed CV21 < 0 and vice versa. After all, if it is true that everyone can indeed be made better or worse off by a movement from state 1 to state 2, then the return to state 1 must be able, respectively, to make everyone worse or better off. Yet, it is easy to show that having regard now that in absolute magnitude, CV21 can exceed CV12 or vice versa, for each person affected, this superficial reflection referred to is far from certain. Granted that the choice of the calculation CV12 rather than the calculation of
theCV21 or vice versa can make a crucial difference, the question arises: which of these alternative criteria should the economist adopt? On economic grounds alone, there can be no convincing answer.4 It follows that if, for any reason, the political decision makers were to require the economist to employ theCV12 > 0 criterion rather than the alternative CV21 < 0 criterion, or the reverse of this, the economist would have no grounds for demurring. He may accept the decision as a valid political constraint. It is, perhaps, unnecessary to remark that one cannot altogether rule out the
possibility that, for every person affected, CV12 is (ignoring the sign) exactly equal to the magnitude of CV21, in which case the CV12 calculation is exactly equal (save for the sign) to theCV21 calculation and, if the one criterion ismet, so will be the other. But once the magnitude of CV12 and CV21 differs for each person, as they generally would, the magnitude of the CV12 calculation will differ from that of the CV21 and, which is more important, it becomes possible for the CV12 > 0 criterion to fail, while the CV21 < 0 criterion to succeed. It also becomes possible for both CV21 < 0 and CV12 < 0.
