In the marginalist theory of value and distribution, capital and labour are seen as 'factors of production' with supply and demand schedules elastic with respect to the correspond ing 'factor prices'. This theory necessarily implies full employment of 'factors' given sufficient flexibility of their prices. In recent years, it has been argued by Garegnani, Eatwell and Milgate1 (hereafter GEM) that Keynes, in his General Theory, retained demand schedules for labour and investment derived along these marginalist lines, and could therefore explain unemployment only by invoking rigidities such as sticky money wages—a possibility fully recognised by Pigou and others. As a result, The General Theory is easily absorbed by the 'neoclassical synthesis' as a special case, and the capitalist econ omy's fundamental tendency towards full employment (in the absence of such rigidities) is reasserted. This outcome is, of course, very different from Keynes's own view of his work. GEM extend the Cambridge critique of the marginalist theory of value to the corre sponding theory of output and employment. They argue that the surplus approach to value and distribution provides an alternative theoretical framework into which the principle of effective demand can be fitted, and within which value and distribution are determined separately from the level of output. This rules out the market-induced adjust ment of output to a level compatible with full employment, for 'the mechanisms of demand-and-supply theory are just not there' (Eatwell and Milgate, 1983, p. 5). Keynes's really revolutionary contribution, his demonstration that the economy is normally in a state of less-than-full employment, from which market forces cannot shift it, is thereby vindicated. This state is established not as a temporary disequilibrium or aberration resulting from market imperfections—which is how many economists treated it2—but as a true long-period position of the economy. (The term 'equilibrium' is deliberately avoided by GEM because of its supply-and-demand connotations.) This long-period