Why is it so hard to do the intellectual history of economics?-John Womack To do intellectual history, one must possess the depth and breadth of knowledge to judge others. May be most of us are not well prepared for this. To illustrate the issue in keeping with the threads topic, let us consider one of the most commonly used instances of equilibrium. It is the idea that the equality of demand and supply denotes a state of equilibrium. Now, suppose that I was to decide if economists before us did make intellectual progress in respect of this equilibrium concept. I have a sad report: we inherited failures. The likely reader is party to the same failings. To overcome the failings, we have to work much harder, and there is resistance to that. It is just too challenging to do hard work for intellectual history to discover our own ignorance. There are three sections below: preliminaries, the context of equilibrium, and an extremely brief sketch of intellectual history. 1. Preliminaries. This is not an attempt to reconstruct economics, but an exercise in historical scholarship that tries to go beyond the dictionary and to chase deeper meanings. We are trying to find out what meanings our predecessors wished to convey, and whether we are able to understand those concepts about equilibrium. Let us put a few basic things out of the way. First, the idea of balance on a scale or in an accounting book has very superficial similarity to the notion of equilibrium in science. Equilibrium is not simple equality (say of weights in a scale or values in an account), but a name for a state of causal determination. It must refer to a causal process which terminates opposing forces. Thus the mere equality of demand and supply is not the focal point in the analysis: the equality determines some variable, and the notion of equilibrium is a tool to convey a constructed sense of causality regarding that variable. Whether there is intellectual something is to be found in the construction of sense. Have we articulated what the REALITY seems to MEAN in a causal sense? Secondly, only a specialized expression of equilibrium refers to a state of rest, because the general causal meaning includes a dynamic process. Thus we need to know not just what happens when demand is equal to supply, but also what happens when they are not equal. A scientist must consider the concept of stability of equilibrium, and must be prepared to narrate the dynamic process of change that leads back to the same original state of rest or finds a new one. In short, the deeper definition of equilibrium must make a clear contrast with the non-equilibrium condition. We must be able to narrate how a state of non-equilibrium makes a transition to a state of equilibrium. Thirdly, the economist must be able to articulate the worldview, that is, describe the economic universe within which the particular event of equilibrium is to occur. My skeletal review concerns this element. The shallow and narrow worldview we have inherited makes it hard for us to perceive the deeper and wider meaning. We end up discovering our own lack of intellect below. 2. The context of equilibrium: What is determined when demand is equal to supply? Obviously, demand and supply are here the determining variables, and when these two are equal, they together determine some other variable, which is neither demand nor supply. The sorry story is that there is much confusion regarding what is determined by this equality. Micro says it determines price and macro says it determines output (albeit at the aggregate level) and nothing says what all these may mean. To clarify this and to achieve a deeper meaning, we must fully specify the context to define the event which we are describing. As I see it, the full description must include at least four different sets of conditions affecting four different sets of variables in the event of exchange in which the demand for one good is matched by its supply. In other words, equality of demand and supply is one of four equalities, all of which must be present to constitute the equilibrium. These four concepts are: output, price, value, and income. To make the points sharp and clear, let us define these four terms and connect the actions of production, consumption, buying (demand), selling (supply), paying, receiving, earning and spending. This preliminary may be tedious but it is necessary to explicate the meaning. Output: Output is the quantity of a real good. Demand is the quantity of the output the buyer is able and willing to buy at a given price and income. Supply is the quantity of output the seller is able and willing to sell at a given price, cost, and resource endowment. Suppose that the good in question is X. Then let Xd denote its demand and let Xs denote its supply. When Xd=Xs, the equilibrium OUTPUT X^ is determined. Please note: It is absolutely out of question for demand and supply of an output to determine price, because price must enter as a predetermined parameter in the demand and the supply functions. Price along with income determines demand, and not the other way round. And then again price along with cost determines supply and not the other way round. Price: The price of x is the quantity of its payment y per unit of x. Thus suppose that the buyer of x buys a quantity Xd and pays with a quantity Ys. Then the (real) price of x is Ys/Xd. It must be a ratio between two different goods that pay for each other. In the same transaction, the (real) price of y must be given by the ratio Xs/Yd. For convenience, and without any loss of meaning, we will convert the prices into units of numeraire, and call them nominal price. There is something extremely serious about the definition of real price. It must establish its connection with payment, namely, real price must be defined as the quantity of payment. It must not be confused with an arbitrary and imaginary numeraire. For example, if the buyer pays for meat with bread, the real price of meat must be quoted as the quantity of bread per unit of meat and nothing else. However, to compare values of several different kinds of things, one good may be taken as the numeraire, and then the prices be used to convert all goods into units of the numeraire good. The key point: a genuine price defined independently of the numeraire must exist before the numeraire can be used to compute values. We must learn to separate price from value. That is nominal price is not really price, but is nominal unit value. We should not use the term nominal price any more, but call it nominal value. We should reserve the term price for the ratio Py/Px where Py is the nominal unit value of y etc. Now, when demand and supply are equal, the equality determines output and does not determine price. Let us see how price is determined then. We begin by defining value. Value: The standard definition of the value of a real good x is the (nominal) price of x multiplied by the quantity of x. To arrive at this, we must agree that the buyer must pay the seller. One of the fundamental institutional rules of exchange is that the value of the payment must be precisely equal to the value of the good. Thus if the buyer buys Xd, then the value of what he buys is given by Vx=Px*Xd. Suppose that he pays with Ys, whose price is Py. Then it is necessary that the value of payment Vy= Py*Ys is precisely equal to the value of the good he buys. We see that Vx=Vy, or Px*Xd. =Py*Ys. The equivalence relation Vx=Vy, determines the (relative) price of x in terms of y. The buyer and seller must negotiate a price which makes the proposed payment equal in value to the proposed sale. It is not equality of demand and supply that determines price, but the equality of value of the good and the value of the payment that determines price. The most dramatic difference between the first equality Xd=Xs, and the second equality Px*Xd. =Py*Ys is that the (relative) price (Py/Px) may very well allow entrepreneurs to exist: the market price can be higher than the (relative) marginal cost and lower than the (relative) marginal benefit. In a proper sense, the buyer of x must be also a seller of y. Using this concept, we see that we must have Px*Xd. =Py*Ys as well as Px*Xs. =Py*Yd. Both equivalence relations refer to the identical outputs if and only if Xd=Xs, and Yd=Ys. That is, we can say that both x and y are payments for each other, and the exchange requires equivalence of payments. Our inability to understand the second equality (of values) trapped us into the endless loop of Says Law. Income: Lastly, we need to see that the buyer of x spends what the seller of x earns, and vice-versa. If the market had just this one transaction between x and y, the income of the seller of x would be Vx, and the income of the seller of y would be Vy. If there are more than two agents, the income is the sum of values of all payments earned by the seller (including sales to himself). This income must be equal to expenditure first for each individual and then for all of them together, requiring two independent equalities. Equalizing income and expenditure at the individual level must involve the use of money precisely when the income from sale of real goods is just equal to the expenditure on real goods, if there is indirect exchange. To achieve equality of aggregate income with aggregate expenditure, the goods on sale must now be broadened to include future and past goods for which payments are made in the current period. The third and fourth equalities respectively involve money and bonds (debts and credits). If one is unaware of three out of four equalities in this equilibrium, how would one do an intellectual history? 3. History of Our Intellectual Progress (?????): Now, if we look back at our intellectual history, we can see that we have made a soup of all four concepts mixed together, habitually confusing price with output, output with value, and value with income. And by God, we have just no idea that indirect exchange requires a transfer of claims on real payment by using an artificial payment such that ones customer of real good compensates ones supplier of real good, all strictly within the current period, and precisely when there is no debt at all. As it turns out, we have no idea that the buyer actually pays the seller, and that the payment may involve money, which is different from debt. Did we make any progress? Consider just four famous authors: Smith: What impression do we get from Adam Smith about this demand=supply equilibrium? In a rather vague way, we suppose that if there is say excess supply, it pushes the price down until demand is equal to supply. If there is excess demand, it pushes the price up. But what is there behind the change in the price? Price is not an external variable that changes autonomously: it is chosen by buyers and sellers through some bargain. The whole matter of intermediation is missing from Smith. My suspicion is that Smith did not have a clear view of the distinction between the different conditions to determine output, price, value, and income. Here are the main reasons for my suspicion. First, the notion of unintended consequence arose because Smith forgot all about payment. The butcher surely does care only about his own profit, and does not intend to give any benefits to the meat consumer. That is, the buyers benefit is unintended by the seller. But that is not all. The consumers benefit is intended by the buyer, and intended so strongly that he actually pays for it. No, it is certainly not unintended by the sole person to whom it is a matter of concern. The seller does not consume the good and hence his lack of intention of deriving the benefit of consumption is entirely irrelevant. Unintended consequence is certainly at odds with the notion that human beings intend something through their economic actions, but this notion takes a central place in the Mengerian construction of the spontaneous origin of institutions. Secondly, we see no clear sign that Smith has made a sharp distinction between autarky and trade. The price is set by visible people who must be seen as entrepreneurs (buyers, sellers), even if they are also optimizers (producers, consumers). Neither Smith, nor anybody after him (including Menger and Mises) ever made the necessary distinction between the optimizer who has a given budget and cannot gain anything, and the entrepreneur, who creates new wealth by way of exchanging something of lower value for something of higher value. If there is trade between x and y, it must mean that the seller of x sells x and buys y because he can get more of y by exchanging his x with somebody than if he himself were to substitute production of y for x. Thirdly, neither Smith nor anybody after him ever clearly distinguished direct trade (barter) from indirect trade. Amasa Walker is the only author known to me who had an idea of indirect exchange, and J.B. Say surely did not. If there was explicit consideration of indirect exchange, Smith could have seen the necessary role of the pure intermediary, who sells but does not produce, and buys but does not consume. The intermediation process must be portrayed precisely to describe the equilibrium. I see that neoclassical economics has not learned how to derive a demand for something by entrepreneurs who do not consume it. Nor is there a supply by an entrepreneur who does not produce it. Do we really know what demand is when an intermediary is able and willing to buy without an intention to consume it? Does the merchant affect the price when he buys the good, and again when he sells it? Since the merchant is not the producer, why would he care about marginal cost? And since he is not the consumer, why would he care about marginal benefit (marginal utility converted into numeraire units)? When we talk about price, do we exclude the merchant from his acts of buying and selling? Smiths neglect of payment disabled him from seeing either price or money because these are aspects of payment: price is the quantity of payment while money is a kind of payment (from among four different kinds of payments). This disability has stayed with us. Fourthly, individual demand in a single transaction between one buyer and one seller, total demand of all customers from one seller, and aggregate demand of all buyers from all sellers are not the same things, and they require different sets of equalities to achieve systemic equilibrium of the market. The same for supply. Do classical and neoclassical treatments make enough distinction to precisely identify the nature of demand and supply? Do we know what demand we are talking of? Apparently, we do not. If we did, we would see money as a necessary element in computing total demand facing one seller, and also see credit in computing aggregate demand, where money is absolutely different from credit. In short, we did not understand what was being discussed under the name of equilibrium. Alfred Marshall: What did we learn from Alfred Marshall? Consider the following indictment. The standard diagram in which a single good is shown in one axis and the nominal price is shown in the other axis is completely missing the all important issue of payment. The demand curve in the standard diagram is based on a predetermined income and price. The supply curve is based on predetermined technology and endowments (giving marginal cost) and price. Then the intersection shows where the demand is equal to supply. It shows the equilibrium output of one good, but cannot show the price. The price is the quantity of the other good y which pays for x, and this other good is absent in the picture. In a partial equilibrium model, the price is a parameter, not a choice variable. Missing an essential part in partial equilibrium model is a route to intellectual degradation. This degradation allowed us to swallow the idea that when demand for x is equal to supply of x, somehow this equality determine the price y/x, although both demand and supply themselves are based on predetermined price, and there is just nothing said about y. Let us remember that the optimizer is a price taker and cannot possibly change the price. Who then changes the price? Is it possible for demand to increase and lead to a higher price? If the buyer buys more and also pays more per unit, where is his larger income coming from? In the same way, the seller cannot possibly increase the supply and cut down the price. How is he going to bear the loss from such an action? Neoclassical economics has no excuse to forget that demand is an optimal quantity chosen under given prices and incomes, and it cannot possibly change without a change in either price or income or both. The confusion can be removed if we make the following observation. An extraneous shock of new knowledge or discovery of new endowment may reduce marginal cost such that the supplier can expand supply at the old price or reduce price at the old supply level or a combination of both. This would be optimal if the income of the supplier increases. Thus the real income of the producer of x may increase by virtue of a technological innovation or equivalently an increase in resource endowment. And with added income, the buyer of y (the seller of x) can buy more at the old price, or pay a higher price at the old level of demand or a combination thereof. In short, an increase in demand cannot occur independently on its own. An increase in income can increase both demand and price. Also, supply cannot change autonomously at all. But an autonomous change in marginal cost resulting from technological progress can change both the supply and the price. In short, no change in price is possible without a prior autonomous change in marginal cost through technological innovation or change in endowment (working through an enlarged supply, then an enlarged income of the supplier, successively in a multiplier fashion.) That is, the idea that excess supply reduces price is utterly unclear: what can it possibly mean? The meaningful statement is that lower marginal cost can enlarge supply and cut down price. In the event of an error, the seller may have to absorb a loss by cutting down the price. Says Law: Says Law has always been unintelligible, because it confuses output with its value. Suppose that the supplier of x wants to buy something y whose value must be equal to the value of x. The equality of the values of two goods cannot mean equality of demand and supply. Indeed, it is plainly absurd to equalize the supply of x with the demand for y. It is however not absurd, but is essentially tautological if the supply of x is to mean the creation of demand for x, in which case, the buyer is identically the seller, and there is no market at all. Outside autarky, it is plainly absurd to suppose that the supply of x by John creates the demand for x by Paul (when John is NOT Paul). So what did we understand by a statement such as supply creates its own demand? Did we mean that the supply of x creates the demand for x? Clearly, the aggregate value of all goods supplied must be equal to the aggregate value of all goods demanded. But this statement says absolutely nothing about market clearing at all, and is not about either supply or demand, but only about value. The all important issue of whether the seller of y wants to buy x when the seller of x wants to buy y is entirely left unexamined. Says Law holds under barter, which is extremely rare. But it must fail in indirect exchange. Thus if the seller of x wants to buy y, but the seller of y does not want to buy x, barter is not possible between x and y. However if the seller of z wants to buy x, and the seller of y wants to buy z, then an indirect exchange is possible if and only if a tool to transfer claims exists. The transfer of claim is necessary because the seller of x sells it to someone from whom he does not buy anything (of z) back, and wants to buy y from someone to whom he cannot sell x. In that case, he is connected to two different people: a supplier (of y) and a customer (of x). His customer (seller of z) must compensate his supplier (buyer of z). The tool that accomplishes this transfer is a device called money. Says Law fails if money is missing in indirect exchange. Equality of demand and supply is not enough; it is further necessary to achieve coincidence of offer and acceptance of payment. Sadly, under the crippling misguidance of Smith, Say and Mill, the otherwise brilliant Jevons drowned into the depth of error about double coincidence. The term double coincidence is an expression that specifies that the buyer pays the seller in every transaction, whether or not money is used. Suppose that the seller of x wants payment in money, then the buyer of x must pay with money and nothing else, that is, there must be double coincidence between the good x and money. Then one would see that procuring money is a real problem under indirect exchange. It is too complicated, and I must ask the reader to download an article on market clearing as noted below. Marshalls famous pupil desperately tried to define demand properly, resorting to an essentially unintelligible notion of effective demand. Keynes needed to make a sharp distinction between ability to buy as given by the real good x that the buyer of y possesses, and the ability to pay, as given by his command of z with which he can pay for y, because x is not accepted in payment for y. Money is possible and necessary only under indirect exchange, and this was never made clear by anybody: Smith, Say, Mill, Marshall and Keynes all failed to consider the matter of payment. Frisch: Of the many terms coined by Ragnar Frisch, macroeconomics has become a major hit. The division occurred because people just could not differentiate price from output. Microeconomics says that equality of demand and supply determines price, while macroeconomics says that the same equality determines output (albeit in the aggregate). Says Law must get the most credit for this disaster: it confused the value of demand (price times output, Px*X) with demand (output X only). The confusion is created by a monumental failure to connect the individual to the market. This connection must be made through the payment relation. Once payment is included, microeconomics is identical with macroeconomics, and trade theory is identical with monetary theory. There are four separate and independent equalities in the equilibrium of the market, each determined by a separate set of determining conditions. Equality of demand and supply is one of them. The second equality is that of the values of the two goods that pay for each other in each transaction. The third equality is the budget balance of each individual. The fourth equality is between aggregate income and aggregate expenditure. All of these must be neatly put in one compact model. It is easier done than said, if we ever care to define our terms. Now, would we want any more of going beyond dictionary to understand the deeper meaning of the terms we use? Apology: A deep history cannot be done in shallow articles like the present one. The readers may check the following articles for more elaborate treatment of the same set of issues. (Best if read in order.) 1. On Autarky and Trade "The meaning of gainful trade [1]http://ideas.repec.org/p/wpa/wuwpit/0405007.html 2. On Market Clearing: "Money in Market Clearing, [2]http://ideas.repec.org/p/wpa/wuwpma/0410009.html 3. On Equilibrium: "Equality of Demand and Supply Neither Determines Price Nor Clears the Market," [3]http://ideas.repec.org/p/wpa/wuwpmi/0405007.html 4. On Micro-Macro distinction: "Micro Takes Over Macro," [4]http://ideas.repec.org/p/wpa/wuwpma/0404012.html Mohammad Gani