Unequal Exchange, Assets, and Power: Recent Developments in Exploitation Theory

The literature is too vast for a comprehensive list of references, but recent contributions include van Donselaar [32], Ypi [44], Fleurbaey [9], Steiner [31], Vrousalis [37].

For all x, \(y \in \mathbb{R}^{n}\), \(x \geqq y\) if and only if \(x_{i} \geqq y_{i}\) (i = 1, …, n); x ≥ y if and only if \(x \geqq y\) and x ≠ y; x > y if and only if x _i  > y _i (i = 1, …, n).

For a slightly different, but related approach based on the notion of “abstract labour”, see Fleurbaey [8, section 8.5].

It is worth noting in passing that the vector \(c_{e}^{\nu ^{}}\) in LE may be a function of \(\left (p,w\right )\) and that once \(c_{e}^{\nu ^{}}\) is identified, the existence of \(\alpha ^{c_{e}^{\nu ^{}}}\) is guaranteed by A2 and A3.

Based on Flaschel’s [7] notion of actual labour values, another definition can be derived which satisfies LE. Instead, the subjectivist notion of labour exploitation based on workers’ preferences proposed by Matsuo [15] does not satisfy LE.

We are grateful to two anonymous referees for bringing this issue to our attention.

In later writings, Roemer himself has acknowledged the limits of purely distributive definitions. See, for example, Roemer [26] and, for a discussion, Veneziani [34].

This argument does not apply to the WL definition: the existence of a general monotonic relationship between initial wealth and WL exploitation at a RS where agents save is an interesting issue for further research.

Okishio [20] also shows that in a dynamic capitalist economy with neither population growth nor technical change, competition among capitalists may drive profits and exploitation to zero. According to Okishio [20], this profit squeeze derives from the increase in the real wage rate due to capital accumulation. Okishio’s [20] results, however, are based on simulation methods and only hold for a specific choice of parameters.

It might be objected that WL exploitation does not disappear, even if ρ = 1, and the relationship between initial wealth and WL exploitation status is preserved. Thus, from a mathematical viewpoint, the model may be interpreted as a generalisation of Roemer’s theory under the WL definition. Yet, this does not affect our main conclusions. First, given the theoretical relevance of the WP definition, Marxian exploitation should arguably be micro-founded as a persistent WP phenomenon. Second, not only is the tendential disappearance of WP exploitation disturbing per se; it also implies that ceteris paribus, WL exploitation, too, is lower in the dynamic model with agents living for T periods than in the T-fold iteration of the static model.

This point is also raised by Bowles and Gintis [2].