Uncertainty and Liquidity

Last week I pointed to the most important paragraph ever written in macroeconomics. Today's article starts with the most important paragraph in microeconomics. It is from Prices and Quantities, probably the best piece that Arthur Okun ever wrote. He is, after all, one of the most under-appreciated economists of the 20th century. 

Okun’s inspiration for the book was the stagflation era of the 1970s. However, the timeless message in his book has to do with prices, reasons for sticky prices and the relation between price stickiness and money, or liquidity. As a background to his analysis, it is important to remember that mainstream textbooks in microeconomics are filled to the brim with the Walrasian narrative, according to which:

• all prices are flexible; 
• information is attainable at zero cost; 
• buyers and sellers can enter and exit a market at zero transaction costs; 
• every product can be traded for another at zero transaction costs; and 
• there are no significant scale advantages or disadvantages in production. 

These premises create the perfect market, where prices set in a one-time process by the auctioneer, under the interaction of buyers and sellers. Needless to say, this institution does not exist in real life, perhaps with the exception of a floor-traded equity market. No consumer market, labor market or market for productive capital complies with the premises of the Walrasian model. 

Which, of course, creates a problem. We spend hours, weeks, months, even years training the next generation of economists into thinking along the lines of an archaic abstract, created for purposes entirely different than those for which it is now being used. Walras never intended for his model to be an empirically valid representation of the real world, yet that is precisely what it has become. 

There are many ways to criticize the Walrasian model, and its uselessness in training students into understanding economics is one of the more serious of its indictments. Equally problematic, and closely related, is the fact that the model does not help students once they become economists and are charged with the thankless job to try to explain the economy of the real world to politicians and other important decision makers. 

When the legislator asks whether it is good or bad to eliminate a budget deficit by means of tax increases or spending cuts, the modern economist who dares to try to answer the question - many would not - resorts to one of two conventional-wisdom solutions: 

• If he leans left in his political beliefs, he will recommend tax increases on wealthy individuals as a means to maintain economic redistribution and pursue egalitarianism;  
• If he leans right he will recommend spending cuts because it reduces the size of government and leaves more of the economy to free enterprise. 

Neither answer is founded in good economic theory. Economic redistribution is often harmful as it destroys sound economic incentives; a reduction of government spending that is not accompanied by tax cuts actually cements the destructive role of big government in the economy. But the fact that these two answers tend to be the standard meal served to legislators hungry for the economist's input, leaves the world a less-well governed place than it otherwise could be. 

The path to better practice of economics does not run through the further complication of macro- or microeconomic models, as those models are often, one way or another, inspired by Walrasian theory. Instead, it requires a rebooting, of sorts, of our way of thinking about the economy. 

This is not as hard as it may seem; on the contrary, we already have many elements, even major building blocks, of a completely new approach to the teaching and the practice of economics. For example, we have Okun's essential contribution. In his challenge to Walrasian economics, he points to the fact that in this the standard microeconomic mode of thinking, there is no room for money. Since, he explains, the Walrasian world has no costs for information, transactions, transportation and storage of products, it also has no transaction costs in the way of barter trade. 

Already here, the lack of realism in this approach to teaching economics should be apparent even to the most stalwart microeconomics professor. Yet even in the Walrasian world, barter has its limits. These limits actually indicate that the very paradigm itself is flawed. Arthur Okun (1981, p. 4): 

If the cheese auctioneer is offered one numeraire’s worth of various goods, it is readily understandable that a nod will be given to the bidder who offers to pay with a product that the auctioneer wants to use; so the double coincidence of wants enters the picture. If it is to be resold, the auctioneer will prefer an item with low costs of storage and transportation, or one of homogenous quality and neat divisibility.

The last part holds the key to the paradigmatic alternative, namely to a replacement of the Walrasian pricing mechanism. That replacement begins with, as Okun points out, the preference for a means of payment that is "of homogenous quality and neat divisibility". 

A Walrasian market has no set price on a product until the very moment buyers and sellers trade. This means that buyers and sellers go into the market with no prior knowledge of the price they are going to face. They cannot even ballpark it, making it practically impossible for them to bring any means of payment - or any reasonable quantity of the product - to the market. 

To get around this problem, the Walrasian model removes the cost for obtaining information; buyers and sellers are perfectly informed about what to expect on the market as they plan their participation. This perfect information package allows them to accurately predict the price, or else they would face a transactions cost associated with bringing the adequate amount of their means of payment to the market. 

The problem is that perfect information does not exist. Sellers of cheese do not know with perfect foresight whether or not they will be able to trade the products they are offered by buyers, simply because perfect foresight does not exist. On the contrary, we know nothing about the future unless we actively build that knowledge up to a level where we feel confident enough to be able to predict it. 

Yet even at that point, our prediction of the future is limited, and will always be limited. Okun's solution to this problem is, as mentioned, that sellers give preference to buyers who bring a means of payment that is homogenous and neatly divisible. Its homogenous property allows the buyer to use it universally, i.e., for purchases of any kind of good or service he may desire. Furthermore, the homogeneity property stretches out through time, allowing the seller to store his earnings from the market for purchases both now and at any point in the future when he so desires. The divisibility property allows for further application of the homogeneity property, both in space and in time. 

As Okun explains, there is only one commodity that meets the homogeneity and divisibility requirements, namely money. The universal usefulness of money in both space and time de facto turns it into liquidity; it is theoretically possible to find cases where money is not liquidity, but those would be rare and, for the purposes of this article, pointless examples. 

In other words, we have now moved away from pure barter, and done so based on the dismissal of perfect foresight. Buyers and sellers face uncertainty in entering the market, and their first step toward managing that uncertainty is to adopt a universal means of payment, a.k.a., money. However, money is as useless as any other commodity as a means of payment unless buyers can predict, within reasonable boundaries, how much of it they need to bring to the market. 

Enter sticky prices. In exchange for a universally useful means of payment, the seller sets a price tag on his products, allowing buyers to predict, with reasonable accuracy, what the price will be of a product in good time before they enter the market. From both sides of the market, transaction costs are minimized - costs that would remain infinite in an uncertain world - as the seller can reasonably predict how much he will sell, and the buyer no longer has to come to the market with (figuratively speaking) infinite amounts of cash in his pocket. 

Okun elegantly explains the origin of the price sticker: it is better for both buyers and sellers than the flexible auction-market price in Walrasian theory. Since ours is a world of price stickers, it would be far better for the economics profession to begin the teaching of microeconomics with the explanation of why we have price stickers in the first place. 

I realize that what follows from that explanation is nothing short of a complete clean-slate redesign of microeconomics textbooks, but - quite honestly - that is long overdue. The next step in reasoning about interaction between buyers and sellers is summarized in the concept of "confidence". Davidson (1972) and Larson (2002) have explained the step from the price sticker, via confidence, to the macroeconomic level. Many others have made significant contributions to the same issue, which we will return to in a later article. 

--

Davidson, Paul: Money and the Real World; The Economic Journal, Vol. 82, No. 325 (March 1972), pp. 101-115. 

Larson, Sven: Uncertainty, Macroeconomic Stability and the Welfare State; Ashgate, Aldershot, UK 2002. 

Okun, Arthur: Prices and Quantities; The Brookings Institution, Washington, DC 1981.