I have a new article out — — in the , which is co-authored with (at Queen Mary University of London) and (at UMass Amherst). The article explores the relationships between exploitation and standard measurements of income and wealth inequality in light of recent discussions of taxes on wealth. This paper is also a product of my ongoing collaboration with Veneziani and Yoshihara, which develops a computational approach to studying exploitation and classes in accumulating economies. We have an earlier paper from this project, in , with more in progress. The most recent paper extends our earlier work and is exciting, primarily, for two reasons: (1) it establishes clear theoretical connections between the Marxian notion of exploitation and standard approaches to measuring inequality; and (2) it allows us to explore some questions around the types of wealth taxes that have been popular topics of discussion lately and how they might affect exploitation. All of this is facilitated by the computational simulation environment we develop. Below I provide a little background on our project and this paper in particular to highlight some our key findings.
Our framework for analyzing exploitation and class originates from John Roemer’s general equilibrium models of exploitation and class, in which the existence of exploitation and classes is driven by the differential ownership of wealth. If one owns wealth it can be invested as capital and used to hire labor-power at a prevailing wage, and the owner of capital then derives an income (profit) from their investment. The wage paid to someone selling their labor-power is labor income and the profit earned by owners of capital is capital income. A person’s overall income will be the sum of their labor and capital income — if they receive the latter from owning wealth, it is not necessary that everyone own wealth. Someone can be considered exploited if they put more labor into the economy than they get out of it in terms of the labor embodied in their income, and they are an exploiter if they get more out than they put in themselves. The way to get more out of the economy than one puts in — in terms of their own labor — is to own capital. The differential ownership of wealth allows some people to use wealth to hire others to perform work and to perform more labor than is embodied in the wage. Thus, differential ownership of wealth entails the existence of exploitation and a clearly defined set of classes.
One problem that emerges with original models is that they are for static, single-period economies and differential ownership of wealth guarantees exploitation only if capital is scarce and labor is abundant. If one looks at accumulating economies — as done by — then over time as capital accumulates it eventually becomes abundant and the profit rate falls to zero. Abundance of capital means that labor is scarce and can command the maximum wage as though everyone has a claim on existing wealth. Thus, in order for exploitation to persist something (or things) needs to ensure the continued abundance of labor relative to capital; things like technical change and population growth being key factors in the persistence of exploitative social relations and class structures, as we explore in our previously mentioned . Understanding the persistence of exploitation and classes is important since Marxian political economy views these as characteristic features of capitalist economies.
Our latest paper takes a slightly different focus from studying the persistence of exploitation and classes. Instead, it examines some normative questions around exploitation and recently popularized — at least frequently discussed — notions of wealth taxes. The computational approach we take is well-suited to exploring different hypothetical redistribution schemes and studying the dynamics of exploitation and inequality under the redistribution of wealth, especially the relationships between inequality in exploitation and inequality in wealth and income. Thus, we are able to bring Marxian analyses of exploitation closer to recent conversations about income and wealth inequality, particularly those prompted by the work of Piketty, Saez, and Zucman.
Our computational framework consists of simulations with many agents who have highly different skill levels and different initial wealth. The initial distributions of skills and wealth are constructed such that the initial state of income and wealth inequality is close to empirical estimates for the United States. Comparisons of exploitation to income and wealth inequality are made through our proposed index of exploitation intensity. We measure the intensity of exploitation one experiences as the ratio of how much labor they perform to the labor embodied in their income, i.e. how much someone puts into the economy versus how much they get out. The intensity of exploitation can vary depending on skills, demand for labor, wages, and wealth. It is then possible to measure inequality in exploitation intensity using a fairly standard tool like a Gini coefficient and compare the dynamics of the Gini coefficient of exploitation intensity to Gini coefficients of the distributions of income and wealth to see how different notions of inequality evolve together, or are unrelated.
Comparing inequality in exploitation, income, and wealth allows us to consider a few interesting questions. (1) Since the existence of exploitation depends upon the unequal ownership of wealth, what happens to exploitation if wealth is redistributed through wealth taxes? If everyone’s labor (or skills) were equal then a uniform distribution of wealth would eliminate exploitation — everyone gets out exactly what they put into the economy in terms of their labor. However, agents in our simulations have very different skills, thus a uniform wealth distribution could lead to significant inequality in exploitation, and similarly, a distribution of wealth that eliminates exploitation could lead to significant wealth and income inequality. (2) What kind of wealth distribution is then desirable, should wealth be distributed equally or should it be distributed to eliminate exploitation; and (3), related to (2), if wealth is redistributed to eliminate wealth inequality, what kind of exploitation exists, and if wealth is redistributed to eliminate exploitation inequality, what happens to income and wealth inequality?
These are the kinds of questions explored in the paper. We first develop a baseline simulation where there are no wealth taxes and accumulation progresses as normal. The steady march of accumulation allows agents who begin the simulation with wealth (sometimes a lot of wealth) to amass even more, while those who begin with no wealth remain propertyless and experience exploitation. The rich get richer.
We then examine what happens to inequality in exploitation, income, and wealth (as measured using Gini coefficients) under three different redistribution schemes. First, under the kind of small taxes on wealth proposed by Piketty; the cumulative effect of which is dramatic reduction in all three dimensions of inequality. Second, under wealth taxes designed to achieve an equal distribution of wealth, which reduces overall inequality but does not do away with exploitation. Third, under wealth taxes designed to achieve what can be called a socialist allocation, where no exploitation exists, i.e. everyone gets out of the economy exactly what they put into it. Given agents’ significant differences in skills, the elimination of exploitation requires that wealth be distributed such that the highest skilled agents have the most wealth and the lowest skilled have the least wealth — or wealth is proportional to skills. This results in not insignificant income and wealth inequality, although it is still much lower than initial inequality or under the baseline simulation with no redistribution.
Exploring these kinds of scenarios provides a detailed account of how egalitarian-minded aims (wealth equality) may potentially be at odds with socialist-minded aims (elimination of exploitation), and how different styles of wealth taxes affect exploitation. All of this — and more — is discussed in detail in the paper.