Why do we need a theory of value?

The theory of value and distribution is at the heart of economics. To be clear, when I say that it is at the center, it means that discussions of almost any topic in economics, in one way or another, depend on a certain theoretical position about the theory of value and distribution. However, most economists have no clue about it, about the centrality of value. Not only they don't understand the original and now infamous labor theory of value (LTV), that dominated between Petty and Ricardo (and Adam Smith too, even though that tends to surprise and puzzle most economists),* but also they misunderstand the dominant marginalist paradigm. Some economists actually think that you don't need a theory of value at all, and some don't even understand that they use a conventional (some vulgar form of supply and demand) theory of value. Hence, the reason of this post is to try to help clarify some very basic issues related to the necessity of a theory of value for proper theorizing in economics.

In a sense, this topic was discussed here before, in my post on Sraffa, Marx and the LTV. But it is worth revisiting, and thinking in broader terms, beyond the LTV, to understand why a theory of relative prices is needed in general, to understand almost everything in economics.

Let me start with the authors of the surplus approach. In fact, a bit earlier with the economists that would eventually be known as Mercantilists (if you can talk about a school). If we are allowed to generalize and simplify, the latter believed that the wealth of nations depended essentially on maintaining trade surpluses, and accumulating precious metals. Profits were essentially the result of buying cheap and selling dear, or profits upon alienation, which indicates that, for Mercantilists, profits were generated in the exchange process.

Classical political economy authors, starting with William Petty, emphasize the determination of profits in the process of production, as a residual of output, once the conditions for the reproduction of the productive system were satisfied. So profits are not the result of selling high and buying low, something that could result from the mere fluctuation of market prices, but from the ability to produce beyond what was needed for the simple material reproduction of society. Note that to obtain profits, part of the residual, the surplus over and beyond reproduction requirements, one needs to know the prices of the means of production. That is, one needs to be able to account for the normal prices of the goods that went into the production of all commodities. And these prices would include a normal profit. Again, not the extra gain that might occur from a high market price. So the normal rate of profit is needed to determine prices, and prices are needed to determine the normal rate of profit. This was well understood by both Ricardo and Marx.

Value (the relative prices of commodities) and distribution (the normal rate of profit) are intertwined. Smith knew that the simple LTV (amounts of labor incorporated) was not correct other than in very rudimentary economic systems, with essentially no produced means of production. His solution was to adopt the idea of labor commanded (more on that on my post on Sraffa, and the one on the standard commodity). Ricardo solved this problem, in his corn essay, by assuming that the surplus and the means of production advanced to produce output where all in physical quantities of corn, hence profits could be determined independently from relative prices, as a physical quantity. And Marx adopted the simple labor theory of value in volume one of Capital. Both believed, for slightly different reasons, that their main arguments would hold even if the LTV was not precisely correct.

I am not concerned with the problems with the LVT in Ricardo and Marx (worth noticing that the mathematical solution was not known in their time, and was essentially developed in the late 19th and early 20th centuries) or Sraffa's solution. It is worth insisting that the LTV does have an analytical solution that is unique, and stable (see my post on the standard commodity for the former, which suggests a Smithian, i.e. labor commanded, version of the LTV is perfectly fine).** That's good, btw. It suggests that the classical political economy notion that there are prices that guarantee the reproduction, and, beyond the the expansion (or accumulation), of the economic system do exist.

Here I want to emphasize the importance of the LTV for the analysis of other aspects of the economy. Ricardo saw the problems of the Smithian adding up theory. That's the notion that prices were composed by the sum of natural wages, profits and rent and that prices would go up if one of its components went up.  In order to determine the rate of profit properly, Ricardo noted the explanation of value was essential. The rate of profit was central because in his view the processes of accumulation depended on the rate of profit. Hence, proper discussion of accumulation and growth depends on a proper theory of value and distribution. Btw, all classical authors assumed that real wages were exogenously determined by institutional and historical circumstances (so there was a role for history and institutions in their theory; also, for accumulation that was seen as too complex to be theorized in the same level of abstraction that value). But even if one is less keen than Ricardo on the role of profits in accumulation, it is undeniable that distribution affects accumulation, and, hence, a proper theory of value and distribution is needed.

Note also, that other things that depend on relative prices are crucially affected by the theory of value and distribution. Classical authors assumed that the process of competition, by which they meant only free entry and not the size or the number of firms in an industry, would lead to a uniform rate of profit. In that sense, the forces of competition were central in forging the structure of production, and, hence, the determination of technological change or to understand the patterns of trade specialization, which cannot be understood without the determination of relative prices. In fact, perhaps the most famous and the most controversial issues coming out of Ricardian economics dealt with international trade and the effects of technical change (the so-called machinery question), and are directly connected to the theory of value.

Even the most crucial macroeconomic problem, the question of output determination (and employment, for a given technique) is affected by the theory of value. Note that classical political economists assumed output as given for the determination of the surplus. And Ricardo accepted Say's Law as a way of determining output and employment (not Marx, btw, so it's NOT a requirement of the surplus approach). But as much as for accumulation understanding of distribution is central for the determination of the level of output, as it is explicit in the Kaleckian effective demand model. the classical long term prices are compatible with levels of output that do not guarantee full employment. And the parametric role of distribution in affecting the size of the multiplier is crucial for output and employment determination. So unemployment is possible in the long run, as a regularity of market economies.

In other words, for a coherent theory of output, accumulation, international trade, technological change and more (taxation, etc.) you need a theory of value and distribution. That is also the case in the mainstream. Marginalism developed in the last quarter of the 19th century, both as a result of the lack of analytical solution in that period for the problems of the LTV and as a reaction to radical revival of the theory (Marxism). The important distinction is that while classical political economy authors dealt only with objective factors, and considered demand as given when determined value and distribution, marginalism incorporated subjective preferences as central for the explanation of long term normal prices, and prices and quantities were determined simultaneously.

Beyond the problems with the marginalist solution for the existence of long term prices (see this on the capital debates) and their switch to the intertemporal approach, which basically only deals with short term prices, their theory is also central for almost everything in economics. In a sense, given that in marginalist analysis distribution is determined by supply and demand, and by the relative scarcity of factors of production, the theory of value and distribution is even more central for other parts of their theory than in the surplus approach. Here the theory of distribution does not affect indirectly the level of output and the process of accumulation. Here the level of employment and, for a given technology, output determination is the same as the theory of distribution. Real wages and the level of employment are determined in the labor market simultaneously. Everything derives from that.

Before getting to the reason why the theory of value and distribution, central for everything, is often ignored, let me note briefly the possibility of a third alternative to value and distribution, beyond the surplus approach and marginalism. That would be the markup theories of pricing. Note that theories of markup pricing essentially describe how firms determine prices. Most of these theories were developed as a result of the imperfect competition literature sparked by Sraffa's famous (1926) critique of Marshallian price theory (see an old post on that here).

First, as it would be known for the readers of this blog (at least the ones that have been reading for a long while), markup pricing is actually dealing with a different set of issues, and Franklin Serrano suggested here that they are different than the classical political economy normal long term prices (the Marxist prices of production or Sraffa's prices), and that Fred Lee and Marc Lavoie were right about that. He argued that some Sraffians (I won't name names), and I would add probably Fred too, thought that Sraffian prices were compatible with the full cost pricing tradition, and I could have included myself in this group.*** Note that what I mean by that is simply that the behavior of firms must be compatible in the real world with the logic of gravitation in classical analysis. In other words, if prices of production imply a normal profit over the full cost for a given technique, then firms somehow must be trying to do that.

But it is clear that the full cost pricing of a particular firm might not be the long run equilibrium price around which market prices gravitate, with free mobility, that is, with competition in the classical sense. In a way, the same circularity suggested above reapers, costs depend on prices (and that involves the profit related to the markup), and prices depend on costs. The firm's individual prices might not be the prices that are required for the reproduction of the economy as a whole. In that sense, markup theories must be grounded on some surplus approach understanding of value and distribution, and they are essentially theories about market prices, meaning short run behavior. In that sense, they run into the same problem than the intertemporal marginalist models, the Arrow-Debreu type, that became more popular after the capital debates, and that led to what Garegnani famously referred to as the change in the notion of equilibrium (that is the abandonment by the mainstream of the notion of long run equilibrium). Some heterodox groups see this as a positive development, but again it implies that they cannot say anything clear about distribution and relative prices, and that has implications for almost any other theory.

I might add here, which is more concerning for some heterodox groups, is that many of these theories are also compatible with marginalist interpretations of the theory of value and distribution. Many imperfect competition theories just suggest simple inverse relations between markups and the price elasticity of demand. This again fall into the type of situation I discussed recently regarding Karl Polanyi, of well-meaning critics of the marginalist mainstream, using marginalist or neoclassical concepts w/o knowing they are doing it (if it's conscious acceptance of the mainstream model, then it's something different).

One last thing in this regard, while markup theories must be grounded on some theory of value and distribution, and my take is that the surplus approach is where it would make sense, the opposite is not true. There is no need for a theory of the firm, of individual behavior, to understand long term prices. Classical political economists certainly discussed behavior, but that essentially entailed some notion related to class, to general social norms, not about what is going on in someone's brain. Even Smith that was certainly concerned with the issue of the role of self-interest in determining the equilibrium outcomes in the market, cannot be assumed to be a precursor of the rational maximizing agents of the mainstream, or of methodological individualism. The same could be said of utilitarian views and Ricardo, who was, to some degree, close to many utilitarians including Bentham. Here too, many heterodox economists think that an alternative theory of behavior is central for economics, and that is why many see behavioral economics as somewhat heterodox.

Finally, getting, even if briefly, to the point of why most economists remain oblivious to the relevance of value and distribution. I would suggest that this is a recent phenomenon. It is the result of what I have discussed here before, the return of vulgar economics (for example, here or here), and that the mainstream has abandoned the long run, and provides only a theory of short run prices. But at the same time the mainstream must revert to the old model in order to promote economic policy. Note that only in that model you can guarantee that markets provide efficient allocation of resources (w/o imperfections), and the price system signals the direction of adjustment. It is often missed by the heterodox groups that resist old classical political economy (often for incorrectly assuming that it is a precursor of marginalism) that their theory of value and their long term prices provide something completely different, an understanding of the conditions for the reproduction of society. That notion, btw, is alive and well in other social sciences (see here or here). Not in economics.

* It survived in the fringes and it was rediscovered by Marx and then much later Sraffa, who actually provided a coherent solution to some of its logical limitations. But after Ricardo, the LTV was never dominant again.

** On the gravitation of market prices towards normal prices see the work by Bellino and Serrano here.

*** My fondness for the subject in part derived from having worked for Wynne Godley at the Levy for two years, who was a disciple of P. S. W. Andrews one of the key authors of the Oxford Economists' Research Group (OERG) behind full cost pricing theories.