Part 9 in a series of posts on modern macroeconomics. This post lays out a more philosophical critique of common macroeconomic models by drawing a parallel between the standard neoclassical model and the idea of “market socialism” developed in the early 20th century.
If an economist had access to all of the data in the economy, macroeconomics would be easy. Given an exact knowledge of every individual’s preferences, every resource available in the economy, and every available technology that could ever be invented to turn those resources into goods, the largest problem that would remain for macroeconomics is waiting for a fast enough computer to plug all of this information into. As Hayek pointed out so brilliantly 60 years ago, the reason we need markets at all is precisely because so much of this information is unknown.
Read any macroeconomics paper written in the last 30 years, and you’d be lucky to find any acknowledgement of this crucial problem. Take, for example, Kydland and Prescott’s 1982 paper that is widely seen as the beginning of the DSGE framework. The headings in the model section are Technology, Preferences, Information Structure, and Equilibrium. Since then, almost every paper has followed a similar structure. Define the exact environment that defines a market, and then equilibrium prices and allocations simply pop out as a result.
What’s wrong with this method of doing economics? To understand the issue, we need to take a step back to an earlier debate.
Mises’s Critique of Socialism
In 1922, Ludwig von Mises published a book called Socialism that remains one of the most comprehensive and effective critiques of socialism ever written. In it, he developed his famous “calculation” argument. Importantly, Mises’s argument did not depend on morality, as he freely admitted that “all rights derive from violence” (42). Neither did his argument depend on the incentives of social planners. “Even angels,” claims Mises, “could not form a socialist community” (451). Instead, Mises makes a far more powerful argument: socialism is practically impossible.
I can’t explain the argument any better than Mises himself, so here is a quote that makes his main point
Let us try to imagine the position of a socialist community. There will be hundreds and thousands of establishments in which work is going on. A minority of these will produce goods ready for use. The majority will produce capital goods and semi-manufactures. All these establishments will be closely connected. Each commodity produced will pass through a whole series of such establishments before it is ready for consumption. Yet in the incessant press of all these processes the economic administration will have no real sense of direction. It will have no means of ascertaining whether a given piece of work is really necessary, whether labour and material are not being wasted in completing it. How would it discover which of two processes was the more satisfactory? At best, it could compare the quantity of ultimate products. But only rarely could it compare the expenditure incurred in their production. It would know exactly—or it would imagine it knew—what it wanted to produce. It ought therefore to set about obtaining the desired results with the smallest possible expenditure. But to do this it would have to be able to make calculations. And such calculations must be calculations of value. They could not be merely “technical,” they could not be calculations of the objective use-value of goods and services; this is so obvious that it needs no further demonstration
Mises (1922) – Socialism p. 120
Prices are what allow calculation in a market economy. In a socialist economy, market prices cannot exist. Socialism therefore is doomed to fail regardless of the intentions or morality of the planners. Even if they want what is best for society, they will never be able to achieve it.
The Response to Mises: Market Socialism
Although Mises argument was effective, the socialists weren’t prepared to give up so easily. Instead, a new idea was offered that attempted to provide a method of implementing socialist planning without falling into the problems of calculation outlined by Mises. Led by Oskar Lange and Abba Lerner among others, the seemingly contradictory idea of “market socialism” was born.
Lange’s argument begins by conceding that Mises was right. Calculation is impossible without prices. However, he argues there is no reason why those prices have to come from a market. Economic theory has already demonstrated the process through which efficient markets work. In particular, prices are set equal to marginal cost. In a market, this condition arises naturally from competition. In a market socialist society, it would be imposed by a planner. In Lange’s view, not only would such a rule match a market economy in terms of efficiency, but it could even offer improvements by dealing with problems like monopoly where competition is unable to drive down prices.
We are left with one final problem, which is the pricing of higher order capital goods. Lange admits that these goods pose a more difficult problem, but insists that the problem could be solved in the same way the market solves it: trial and error. In other words, just as entrepreneurs adjust prices in response to supply and demand, so could social planners. When demand is greater than supply, increase prices and when supply is greater than demand, reduce them. At worst, Lange argues, this system is at least as good as a free market and at best it is far better since “the Central Planning Board has a much wider knowledge of what is going on in the whole economic system than any private entrepreneur can ever have” (Lange (1936) – On the Economic Theory of Socialism, Part One p. 67)
The Market Socialist Misunderstanding: Why do Markets Work?
Lange’s argument should be appealing to anybody that takes standard economic theory seriously. A Walrasian General Equilibrium gives us specific conditions under which an economy operates most efficiently. We talk about whether decentralized competition can lead to these conditions, but why do we even need to bother? We know the solution, why not just jump there directly?
But by taking the model seriously, we lose sight of the process that it is trying to represent. For example, in the model the task of a firm is simple. Perfect competition has already driven prices down to their efficient level and any deviation from this price will immediately fail. As Mises emphasized, however, the market forces that bring about this price have to come from the constant searching of an entrepreneur for new profit opportunities. He concedes that “it is quite easy to postulate a socialist economic order under stationary conditions (Socialism, 163).” Conversely, the real world is characterized by constantly shifting equilibrium conditions. The market socialist answer assumes that we know the equations that characterize a market. Mises argues that these equations can only come through the market process.
Hayek also made an essential contribution to the market socialism debate through his work on the role of the price system in coordinating market activity. For Hayek, prices are a tool used to gather pieces of knowledge dispersed among millions of individuals. An entrepreneur does not need to know the relative scarcities of various goods when they attempt to choose the most efficient production process. They only need to observe the price. In this way, Lange’s pricing strategy cannot hope to replicate the process of a dynamic economy. When prices are no longer set by market participants looking to achieve the best allocation of resources they lose almost all of their information content.
The final piece missing from Lange’s analysis is perhaps also the most important: profit and loss. In Lange’s depiction of a market economy, he seems to imagine one that is already close to an equilibrium. In particular, he assumes that the production structures in place are already the most efficient. Without this assumption, there is no way to determine the marginal cost and no way to use trial and error pricing effectively. Mises and Hayek instead view an economy as constantly moving towards an equilibrium but never reaching it. Entrepreneurs constantly search for both new products to sell and new methods to produce existing products more efficiently. Good ideas are rewarded with profits and bad ones driven out by losses. Without this mechanism, what incentive is there for anybody to challenge the existing economic structure? Lange never provides an answer.
For a longer discussion on the socialism debate, see a paper I wrote here.
Repeating the Same Mistakes
So what does any of this have to do with modern macroeconomics? Look back to the example I gave at the beginning of this post of Kydland and Prescott’s famous paper. Like the market socialists, their paper begins from the premise that we know all of the relevant information about the economic environment. We know the technologies available, we know people’s preferences, and we assume that the agents in the model know these features as well. The parallels between the arguments of Lange and modern macroeconomics are perhaps most clear when we consider the discussion of the “social planner” in many macroeconomic papers. A common exercise performed in many of these studies is to compare the solution of a “planner’s problem” to that of the outcome of a decentralized competitive market. And in these setups, the planner can always do at least as good as the market and usually better, so the door is opened for policy to improve market outcomes.
But because most macro models outline the economic environment so explicitly, competition in the sense described by Mises and Hayek has once again disappeared. The economy in a neoclassical model finds itself perpetually at its equilibrium state. Prices are found such that the market clears. Profits are eliminated by competition. Everybody’s plans are fulfilled (except for some exogenous shocks). No thought is given to process that led to that state.
Is there a problem with ignoring this process? It depends on the question we want to answer. If we believe that economies are quick to adjust to a new environment, then the process of adjusting to a new equilibrium becomes trivial and we need only compare results in different equilibria. If, however, we believe that the economic environment is constantly changing, then the adjustment process becomes the primary economic problem that we want to explain. Modern macro has heavily invested in answering the former question. The latter appears to me to be far more interesting and far more relevant. The current macroeconomic toolbox offers little room to allow us to explore these dynamics.