ABSTRACT
Is it possible in a changing world to de®ne net social income in such a way that it is both welfare relevant and ideally computable?
Earlier writers, like Marshall, Fisher and Pigou were inclined to answer the question in the af®rmative. In contrast, Samuelson (1961: 57) argued that in general `the only valid approximation to a measure of welfare comes from computing wealth-like magnitudes, not income magnitudes of the Haig, Fisher or any other type.' More recently, Weitzman (1976: 156) has advanced the moderating view that `¼it is not really a question of choosing between a conventional but inappropriate current income concept and an impractical but correct wealth-like magnitude, because in principle they are merely different sides of the same coin.' More speci®cally, he has shown that, under several special assumptions, conventionally de®ned net national product can serve as `a proxy for the [wealth-like] present discounted value of future consumption'. In particular, to avoid `the index number problem', Weitzman assumed that there is a single consumption good (with the implication that primary factors like labour are inelastically supplied), that all utility functions are linear, and that both utility functions and the technology of the economy are stationary ± that is, independent of time. Finally, Kemp and Long (1982) have demonstrated that, even without the ®rst two of Weitzman's assumptions (a single consumption good, linear utility functions), our question can be answered in the af®rmative. In particular, it was shown that net national product or net social income, rede®ned in terms of utility, can serve as a proxy for the wealth-like present discounted value of the stream of future utility; indeed, it is the constant equivalent of that stream. Moreover, it was shown that non-stationarity can be overcome, either by treating those states of knowledge which result from learning or research activities as capital goods which can be accommodated in a stationary technology or, in the case of exogenous changes in knowledge, by a suitable extension of the notion of net national product.
