I have been meaning for a while now to explore potential connections between Marx's theory of value and Keynes or Kalecki-influenced approaches to macro. This is a tentative testing of the waters. It may be the first in an indefinite series of posts, sprinkled throughout the future, on correspondence between the two theoretical traditions. Then, again, it might not be. At this stage it is not clear to me how far the exercise can be taken, or how useful it might be. I know that there has been some exploration of the connections between Keynesian and Marxian approaches in the academic literature. Massimo De Angelis (here is a sample paper) and Andrew Trigg are two names that come to mind. Any posts I do here will be more exploratory and elementary by comparison. The emphasis will be on connections of a macro nature between MMT and Marx's value categories. One point of entry appears to be the 'monetary expression of labor time' (MELT), introduced by Alejandro Ramos Martinez (see chapter 5 of this link), and its connection with the Modern Monetary Theorists' 'value of the currency'.
Since readership at heteconomist is diverse, an ultra brief rundown of Marx's theory of value is probably in order. Only as much detail as is necessary for present purposes will be included. I'll let details leak out progressively over future posts, if there are any. In these discussions, I will be applying the so-called 'temporal single-system interpretation' (TSSI), which is one of the competing interpretations of Marx. Chief proponents of this approach include Andrew Kliman and Alan Freeman. The other main approach – the dominant one traditionally – is sometimes referred to as the 'simultaneist dual-system interpretation'. Robert Vienneau has provided an excellent FAQ that primarily focuses on that approach.
In confining my attention to the TSSI, I will make no effort to defend this interpretation or evaluate its merits relative to the alternatives. There are others far more qualified to attempt that task. My focus will simply be on explaining Marx in terms of the TSSI and linking this understanding, where possible, to Post Keynesian and/or MMT macroeconomics.
Even so, I can offer a couple of motivations for my choice, which do not concern the theoretical merits of the competing approaches but simply reflect my present purpose in considering Marx. One is that, like much of Post Keynesianism, the TSSI emphasizes the importance of historical time (the approach is dynamic or temporal) and is not confined to equilibrium or steady-state scenarios. To the contrary, it emphasizes non-equilibrium analysis. The other motivation is that Post Keynesianism, Modern Monetary Theory and the TSSI all seem particularly suited to the study of a monetary production economy.
Values and Prices
Marx's theory of value applies to commodity production. A commodity is a good or service that is produced for the purpose of being exchanged in a market. The theory does not apply to goods or services that have not been produced as commodities, which may instead be valued according to normative criteria, cultural norms, and so on.
Value, in Marx, is a quantity of socially necessary labor time or its monetary equivalent. Labor time is only socially necessary if it involves producing commodities at the prevailing, average level of productivity. This ensures that the value of a commodity is not increased simply by taking longer to produce it through inefficiency or the application of backward production techniques. The time taken must be consistent with prevailing technology, know-how and worker effort of the average kind in that line of production.
Price is distinguished from value, in Marx, in at least two different ways. In one usage, 'value' refers to the 'value created' in production, whereas 'price' refers to the 'value received' in exchange. In a second usage, price refers to value expressed in money rather than labor-time terms.
Unless otherwise stated, I will be employing the first usage. So value will refer to what is created in the sphere of production. Price will refer to what is realized in the market when output is sold. In this usage, both price and value can be expressed in labor time or, equivalently, in money, and often will be.
Marx's Theory of Value in a Nutshell
We can keep this as elementary as possible by considering the 'total value' and 'total price' produced in the economy as a whole. This is similar to considering the GDP of an economy without regard to sectoral breakdowns of the economy. Sectoral breakdowns are important, but will be left for possible future posts.
Total value, for Marx, is the sum of three components. First, there is the value outlayed for 'constant capital', which is the portion of plant, machinery and raw materials used up in the production period. In Marx's theory, constant capital represents value that existed prior to the current production period. It is 'dead labor', in the sense that it was produced by labor in a previous period. The magnitude of constant capital does not change during the current period. Instead, it is simply passed on to the final value of the output.
Second, there is 'variable capital'. This is the amount of value outlayed to employ workers in the current period. The employment of 'living labor' creates new value. Importantly, if workers are required to work for longer than is necessary to produce the equivalent of their wages, there will be 'surplus value' left over for capitalists. This surplus value can be realized in exchange as gross profit, prior to its distribution into various parts (e.g., corporate retained earnings, interest, rent).
In practice, capitalists may fail to realize in exchange all the surplus value created in production. If so, realized gross profit will fall short of produced surplus value. The difference will accumulate as unsold inventories. For simplicity, in this post (but not necessarily in future ones) it will be assumed that all surplus value created in production is successfully realized in exchange. That is, the important possibility of a realization crisis is abstracted from to bring out some basic relationships.
A Simple Numerical Example
Suppose that the equivalent of $50 in constant capital, C, is used up in the production period and that capitalists outlay $50 of variable capital, V, to employ workers for the period. The workers are required to work a total of 100 hours, which is the amount of total employment, L. We can imagine that 50 hours of the workers' labor time is spent producing a value equivalent to their wage bill, W (assumed to be equal to variable capital). The other 50 hours of the workers' time is spent producing surplus value, S, for the capitalists. Since half of the workers' labor time goes toward the production of surplus value and the other half produces value equivalent to variable capital of $50, this implies surplus value of $50 is produced. Suppose, lastly, that the general price level, P, is 1.
We can infer from the above numbers that the net product of the entire economy, equal to V + S, will be $100. This is the total new value produced in the period. Given our assumption that all value created in production is realized in exchange, this net product will correspond to nominal GDP, PY, so real output, Y, must be 100.
The table below summarizes the situation. The dollar figures are values and prices expressed in money terms. The figures in parentheses are values and prices expressed in terms of labor time.
__________________________________________________________________ C V C + V S Net Product Total Value (= W) (= V + S) (= Total Price) (= PY) (= C + V + S) __________________________________________________________________ $50 $50 $100 $50 $100 $150 (50) (50) (100) (50) (100) (150) __________________________________________________________________
In passing, we can note that the economy-wide rate of profit is calculated as S/(C + V) = 50%. The rate of exploitation (or rate of surplus value) is defined as S/V = 100%.
The Monetary Expression of Labor Time (MELT)
More relevant for our present purposes is the MELT. In this post, we can assume that productivity does not change over the period. Doing this keeps our calculation of the MELT ultra simple. Once productivity is allowed to change from one period to the next, we will need to resort to a more general formula. I'll leave this to a future possible post. Those who wish to jump ahead can consult the chapter by Ramos Martinez, linked to earlier.
The MELT is one hour of socially necessary labor time expressed in monetary (let's say dollar) terms. It is the dollar value of one hour of labor time. Equivalently, it is the reciprocal of the amount of labor commanded by $1. Or, expressed yet another way, it is the economy-wide ratio of total price to total value.
In our simple example, the MELT is equal to $1/hr. This says that each hour of socially necessary labor time creates $1 of value in money terms.
One other thing before moving on. Value is usually expressed in terms of 'simple' labor time. This is the simplest form of labor performed in the economy. More complex labor is then considered as a multiple of simple labor.
In relation to the above example, we could suppose one of two things. Either all labor is simple. Or, more likely, our above calculations were in terms of the average kind of labor. I say "more likely" because if we were trying to glean these numbers from national accounts, we'd be relying on total employment figures, then trying to work out how much simple labor time that translated into.
In our example, the average wage rate is $0.50/hr. We can see this from the fact that variable capital and the wage bill are $50 and total employment is 100 hours. Suppose that, in addition, we also knew that the minimum wage happened to be one-fifth of the average (i.e. $0.10/hr). If we interpret minimum-wage labor as simple labor, we might infer from this, imperfectly, that one hour of average labor is the equivalent of five hours of simple labor. Let's do this. That then implies that total employment is the equivalent of 500 hours of simple labor time.
This modifies our figure for the MELT. Whereas 150 hours of average labor (both dead and living) corresponds to the total value of $150, it is actually 750 hours of simple labor that corresponds to this $150 of total value. In other words, the MELT can be modified to $0.20/hr.
Connection Between the MELT and MMT's Value of the Currency
In MMT, the value of the currency is defined as what is required to obtain it. Currency value can therefore be defined in terms of labor time (see this old post for more details). In our example, the minimum wage is $0.10/hr. This implies that it takes 10 hours to obtain $1. The value of a dollar, in other words, is 10 hours of simple labor time.
So we have two definitions. Denote the MELT by m and the value of the currency by z:
Net Product PY m = ------------------ = ---- Total Employment L
Total Employment L z = ------------------ = --- Wage Bill W
This implies a simple relationship between the MELT and value of the currency that makes it easy to switch back and forth between the two:
WPY m = z . ----- L2
L2 z = m . ----- WPY
So, in our example, knowing that the MELT is $0.20/hr, we can easily determine the value of the currency as:
(500 hrs)2 z = $0.20/hr . ---------------- = 10 hrs/dollar ($50)($1)(100)
Or, knowing the value of the currency is 10 hrs/dollar, we can determine the MELT:
($50)($1)(100) m = 10 hrs/dollar . ---------------- = $0.20/hr (500 hrs)2