4 SOCIAL SCIENTIST labour value is redundant, and yields no insights which cannot be obtained from the theory of 'competitive prices'. I Technology and Profit Rate Samuelson's first point is quite simple, and completely unassailable. For Marx technical progress has two effects. By creating and sustaining the industrial reserve army of the unemployed, it provides "the pivot on which the law of demand and supply of labour works"4 to hold down real wages. And by inducing a continuous increase in the organic composition of capital, it generates a tendency for the rate of profit to fall. Though his prediciions are hedged around with qualifications and reservations, there is no doubt that Marx expected a falling rate of profit simultaneously with stagnant or even falling real wages.9 Even economists friendly to Marx have been unhappy with this conclusion. Joan Robinson, for example, noted as early as 19426 that if technical progress raised the productivity of labour7 while the real wage remained constant, the rate of exploitation would rise. Now the organic composition of capital is the ratio of constant to variable capital, c/v (==k); and the rate of exploitation is the ratio of surplus value to variable capital, s/v (a==e). On Marx's assumptions technical progress increases both k and e. But the rate of profit (r) is s/c+v, which is equal to e/k-+-1.8 Thus the net effect of technical change on the rate of profit appears to be indeterminate.9 Samuelson's argument10 is that, given a constant real wage, the effect is determinate and operates in the opposite direction to that predicted by Marx. With a constant real wage, technical progress will increase r; conversely, ifr remains a constant or falls, the real wage must increase. Given a constant real wage, capitalists will introduce a new invention only if it raises r. If it does not, then they have no incentive to adopt the invention. And if a mistake is made, so that a "technical disimproveme.nt" which reduces r is introduced, capitalists will have an obvious incentive (once they realize their mistake) to scrap the invention and return to the old technology. The argument may be illustrated by reference to the "wage-profit frontier" analysed by Sraffa.11 For any given technology, the curve shows the highest rates of profit consistent with alternative levels of the real wage (w). Very little can be said a priori about its shape, except that it is monotonically decreasing.12 One possible frontier is F^ F^ in figure 1. For a given real wage (Ow1), A shows the highest available rate of profit (s=0r1). Samuelson's point is simply that capitalists will introduce an invention only if it permits them to move northeastwards from A to a new and higher frontier such as F^Fg. They would not introduce a technical change which forced them to move to a lower frontier such as FgF,, since here (at B) the same real wage would yield a lower rate of profit (=0r11). From the capitalists' point of view—and in capitalism it is their point of