Michele Boldrin Confused About Marxian Exploitation, Marginal Productivity

Michele Boldrin has written a paper in which supposedly Marxian themes are treated in a Dynamic Stochastic Equilibrium Model (DSGE). He writes:

...there is 'exploitation of labor' also in standard neoclassical models of production. Within each firm, almost all workers (i.e. everyone but the marginal worker at the marginal plant) are 'exploited' in the sense that they are being paid a wage smaller than their marginal productivity. -- Michel Boldrin (2009) "Growth and Cycles, in the Mode of Marx and Schumpeter. Scottish Journal of Political Economy, V. 56, N. 4 (p. 432)

The above is mistaken on at least two points:

• The notion of exploitation is Joan Robinson's neoclassical idea from the period in which she developed the theory of imperfect competition; this idea is not Marx's.
• The above passage seems to take the value of the marginal product of labor as defined prior to prices, including wages. If so, Boldrin follows the mathematically mistaken teaching of some not-so-bright orthodox economists.

Boldrin continues his mistaken line:

From a Marxian view point, labor-saving innovations are the means through which capitalist exploitation can be perpetrated and maintained over time... -- Michel Boldrin (2009)(p. 435)

The above might or might be true of Marxian exploitation, but Boldrin is using a different definition. And again:

Asymptotically, all existing firms use the same, best, technology and the market wage corresponds to the marginal productivity of labor in the marginal technology, which is also the one everybody uses. Hence, exploitation of the workers has ceased. -- Michel Boldrin (2009)(p. 440)

John Roemer describes a source of profits in a model which could exhibit perfect free entry, constant returns to scale, and individual profit maximization:

"...the existence of postive-profit equilibria ... is to associated with the necessity of time in production, that capitalists must advance the costs of production before they receive the revenues from production. It is this temporal structure of production that gives rise to the economic necessity of a capital constraint, whether or not funds for production are limited to internal finance or are available on a capital market." -- John E. Roemer (1981). Analytical Foundations of Marxian Economic Theory, Cambridge University Press (p. 84)

As those familiar with Frank Hahn's critique of the "neo-Ricardians" know, this sort of model is consistent with every valid marginal productivity principle holding.

Man-Seop Park's criticizes new growth theory (from, for example, Paul Romer) on a number of grounds in a number of papers. One of these grounds is that such models ignore the presence of time in production, even when they depict a number of stages in production. I think Boldrin's paper may be weak on this ground.

Update: I thought a little more about this. Bouldrin considers the case in which the marginal plants in both the investment and consumer goods sector are both operated at less than capacity. In this case, the value of the marginal product of labor is positive (in the competitive case), and the marginal product of the capital good is zero.

I would rather consider the case in which both labor and the marginal plants are binding in both sectors. In this case, an additional unit of either labor or plant would contribute nothing to production. On the other hand, a marginal unit decrease in either labor or capital would decrease production by some specified amount. Thus, the value of the marginal product of both labor and the capital good (in the competitive case) would be an interval from zero to some positive amount. I think this case results in the familiar Sraffian wage-rate of profits curve in which wages can only be larger if the rate of profits would be smaller.