4 – MODELS – THE STYLE SHOW
Economists like to create ‘models.’ These are simplified schemes or representations of how the world works, economically. The models have to be simple because reality is impossibly complicated. Some are stated in words and some use mathematical formulas. Economic models fall into two broad types, one based on ‘equilibrium’ and the other on so-called called ‘stocks and flows.’ Both gave rise to important variations over the years. Both were proposed early-on but the equilibrium models dominated economic thought for most of its history. Economic equilibrium meant that production and consumption, buying and selling, were steady and prices unchanging. If the economy were slightly out of equilibrium it was presumed to be merely shifting to another equilibrium. Economic variables would be nearly stationary with time. The goods sought by buyers would equal the amounts produced by sellers.
Equilibrium economic models assume that people act both rationally and independently when buying and selling and that the entire economy is based upon such exchanges between them. Supply and demand models, in particular, are equilibrium models and the most familiar today. They have shaky foundations because they often predict wrong results. Dubious or not, those instances where they’re nearly right are studied and something may be learned. It’s a good way, really the only way, to begin. We must start from where we’re at.
Beginning textbooks introduce students to an extremely simple model, the so-called circular flow in a village economy. Money passes from hand to hand in one direction around a circle while goods and services go around the other way. It’s called “the great wheel of money” and shows how money and goods move through an economy in opposite directions. One man’s sale is always another man’s purchase. But one realizes that, in real life, money doesn’t go around in a circle – it moves in every direction through a network, from one hand to many other hands. Goods and services also flow through the network, opposite the money flow. Thus, the proper topology of even a small economy is a network with many nodes, each having connections to, (i.e., transactions with), many other nodes. Everyone buys and sells with many others. One man’s purchase is still another man’s sale but it doesn’t go in a circle. It goes every which way. Even in a village economy, every network node is connected to many others.
The idea of a network economy helps explain some puzzles. Keynes once outlined the process of producing a Chinaman’s shirt. It doesn’t start at a shirt factory but with the plowing of a field in South Carolina. Cotton seeds are planted and tended. The cotton is harvested and shipped by sea to England where mills weave cotton into cloth and factories make it into shirts. These are then shipped to Shanghai, bought by a merchant, sent inland by oxcart, and finally sold to the Chinaman. This is only one of the paths through the world’s network of manufacture and trade. South Carolina also ships lumber to New York. China ships silk to England. Branches of the network come together during shipping, and they diverge again in distribution.
Unless we make such networks the base of our economic thought, we won’t get a grip on some economic facts that stare us in the face. For example: if everyone buys and sells from everybody else it’s obvious that economic growth will help everyone: “a rising tide lifts all boats.” But in the real world, why does economic growth help some people much more than others? A network model of an economy helps to explain this. Nodes representing individuals can be multiply connected within a local group but may be connected to only a few nodes outside the group. That is, the in-group can be well connected to each other but poorly connected externally. Such groups are economically almost isolated from others. Examples are ethnic neighborhoods and countries with little external trade. Further, large economic networks may have groups of nodes arranged in hierarchies. In such an economy, money and goods will flow up and down tree-like structures, going to and from their many destinations. Commodities, for example, move in a hierarchically connected economic network: producers to refiners, to suppliers, manufacturers, distributors, retailers and to consumers. Each level has many branches, finally going out to the twigs of separate users.
In our modern economy persons cannot possibly trade with all of those who make daily life possible, so groups of nodes – businesses, governments, etc. – have evolved to simplify the interconnections. How these structures are organized within themselves and with others has a lot to do with the stability and resilience of an entire network. When the structure of an economic network is understood, the nodes and flows between them can be modeled and the behavior of the network under various conditions can be figured out. For large networks, this has to be done by computer since so many parts must be tracked simultaneously.
The alternative stocks and flows model originated with the so-called Physiocratic theory of the economy proposed in 1758 by a Frenchman, Francois Quesnay. It was an attempt to understand and explain economic growth by making a detailed diagram, called the tableau economique or economic table of who produced what, who bought and spent what, and with whom. In the physiocrats’ formulation, the great sectors of the economy – farmers, merchants, artisans, and large landowners – bought and sold from each other. Goods and money passed between them as described in the economic table. Remarkably, the physiocrats were able to show that reinvestment of capital was necessary for economic growth and that investments had diminishing returns. It was a very good start but was superseded by the influence of Ricardo and others in the later 1800s. Supply and demand ruled thereafter. It was too bad that such a promising beginning was forgotten for so long. The stocks and flows model reappeared only much later, evolving from practical problems in manufacturing.
Stock and flow models are familiar and useful to factory owners and retailers who are concerned with inventory turnover and efficiency. They consider their inventory to be a stock, an asset which costs money every day they hold it. They also think of production and sales as flows into and out of inventory, generating expenses and income in proportion to their rates of flow. Their economic model, when drawn on paper, shows boxes representing stocks connected by arrows representing flows. An example would be a stock of iron-bearing rock buried in the ground that flows from a mine into a stockpile of ore, which then flows via a railroad to another heap in Pittsburgh. It then flows via a conveyor belt into a blast furnace then flowing out into ingots of steel. These are shipped, i.e. they flow, to mills and factories. Parts are manufactured, stocked in warehouses and then flow to more stocks elsewhere. At every stage – factories, warehouses, distributors and retailers – stocks are held and flows are passed between them via trucks, railroads, and ships. This kind of model is very complicated but has big advantages: it is definite, can be specified precisely, is easily handled by computers, and can be tracked as it develops through time. On its face it says nothing at all about how supply matches up with demand or how markets operate.
Both kinds of models have theoretical problems. Supply and demand types consider money to be unimportant since it’s treated merely as a medium of exchange. Stock and flow models treat money both as a stock, (an asset or store of value when held) and a flow, (a medium of exchange when moving from place to place). Economic activity is always changing among the sectors of the economy and throughout the seasons. Suppliers try to match demands but never quite do; someone is always over or under the target. In supply and demand models this continual shifting is hard to analyze so economists deal with it by postulating that the economy is at equilibrium, i.e., by assuming that supply always equals demand and everything is settled down. This hypothesis is absolutely necessary to supply-demand models. It may be a good approximation in some real life situations but not in others. So by assuming the economy to be at equilibrium, economists can form models that get results some of the time. The stocks and flows model, on the other hand, does not require the equilibrium assumption. The shifting tides in the economy can be followed as they happen, even when way out of balance. Credit and debt are not considered in either model so the interactions of goods, time and money never appear. Some argue that any realistic economic models would be so complex as to be impossible to handle, but this isn’t so. Models need to be developed to the point where they reasonably correspond to the real world, but then further elaboration can stop. They may be far too complicated for mathematical solutions but that’s where computer simulations step in.
Equilibrium and stock-flow theories each have their own domains of usefulness. In equilibrium models the intuitive self-correcting balance of supply and demand is apparent. Mathematical calculations can link economic variables to each other during equilibrium and small departures from it. Stock-flow models, on the other hand, have no need to balance supply with demand. They can model non-equilibrium states, that is, how economic networks change as goods and money move from place to place. Varying demands and shortages shift flows to alternate suppliers. Prices change. Slack periods cause forced sales and bankruptcies to develop. Stock-flow models can also deal with new technologies which are an important factor in economic growth. New competitors may arise in the network. A cheaper way to produce something or the invention of a Next Big Thing will add new nodes and branches to the economic network. Other branches of the network will die away, obsoleted, victims of creative destruction. This shifting, growing, declining, ever-moving nature of our economy is normal today – it is what we live with – and it’s never in equilibrium. This kind of model, it seems to me, best fits the amazingly complex and fluid economy that we live in today.
Recently, because computers have become cheap, Australian economist Steve Keen hearkened back to Quesnay’s tableau economique and created a stock-flow computer model of a multi-sector economy without assuming that its variables would come to equilibrium. But then he added a credit mechanism: he added a banking sector which could loan to and take deposits from the other sectors. [1] With this in his model, money flows and trading volumes between the sectors could be advanced or delayed. Business cycles, booms, and busts became possible. Keen found they indeed happened in his simple computer simulations. Incomes rose and fell. Some profited over others. Sometimes bankruptcies occurred. These changes were not seen, were not even possible, in any equilibrium model. His model economy didn’t shift from one equilibrium position to another – it could go along steadily but then move abruptly. His work was an advance into new economic territory. Keen had made a real breakthrough. For his unorthodoxy he was pushed out of his position at the University of West Sydney: the entire department of economics was abolished in order to accomplish it. But I guess it’s the old story; the old guard stamps out innovation.
Another variation on a stock and flow model was proposed by Treval Powers in 1996. [2] He wrote a book out of dissatisfaction with textbook macroeconomics, which he studied after a career in the cement business. He approached macro with a new model based on a Composite Producer and a Composite Consumer which together form the economy as a whole. He realized that these composite entities are effectively immortal, cannot plan ahead, have no intentions, accomplish contradictory things simultaneously, but yet have exact and definite values for all variables like Output, Wages, Purchasing Power, etc., at every moment in time. Government statistical publications, quite accurate since WW-II, allow him to dissect and test his ideas.
Powers based his model on a circular flow – the Composite Consumer buys (and provides funds plus a profit) to the Composite Producer, who returns goods and services to the Composite Consumer. By itself, this is only an uninteresting closed loop. But he added in leakages to the flow of various kinds. On the inflow side are industrial raw materials and energy. The important leakages are on the outflow side: industrial waste, losses of money due to individual hoarding, mal-distribution of income, undistributed corporate profits, destruction in fire, flood and war, bank failures, bad loans, market crashes, fractional bank reserves, monetary restraint by the Fed, and funds flowing overseas into Eurodollar and other markets. Anything that subtracts from the flow of funds from the Composite Consumer to the Composite Producer will cause growth to slow.
Powers’ model violates Say’s Law, in which supply and demand are always equal. It therefore allows for involuntary unemployment or excess capacity. He did not consider additional cases such as widespread outsourcing of jobs to foreign countries, credit failures or runaway inflations. He did not mention Keynes. His model also told him that prior to 1961 money withdrawn by the Fed to cool the economy was rapidly restored during recessions, speeding up recovery. After 1961, this process seems to have failed. He conjectured that the money withdrawn found its way into investments other than the national cash reserve, and wasn’t available to the Fed for quick injection into the economy. It’s interesting that this amateur model foresaw the policy of Quantitative Easing (QE) that the Fed used during the recession of 2008.
Economists love to express their theories in mathematical form for analysis (and sometimes are accused of ‘mathiness’), but I don’t know if any of them have tried to ‘mathify’ Powers’ approach.
It’s easy to build models on stated assumptions but hard to find assumptions that are relevant to reality. Few economists see that equilibrium ideas take no consideration of time. They cannot deal with time-dependent things like debts, expectations, accumulations of inventory and plant, or durable vs. non-durable goods. Using money as a store of value as well a medium of exchange decouples producers from users and lets speculators into the circular-flow of money and goods.
The economy is a sea of conflicting interests, and its history matters. Most economists don’t want to discuss this. Its ultimate flaw is the classical assumption of perfect information about the past and future. The past is unalterable and is often forgotten, while the future is unknown and uncertain.