Say vs. Keynes: Part 2

In the first part of this article I explained how Say’s Law is illogical and cannot function unless we assume away time itself. In this part I am going to explain how the alternative, Keynes’s Law, works and how it, unlike Say’s Law, is the realistic basis for economic analysis.

Since the key problem with Say’s Law is its reliance on logical time, in other words a de-facto elimination of time itself, Keynes starts with replacing logical time with historical time. Also known as sequential or discrete time, this concept of time is far from as fancy as it sounds. It simply means that we start our economic analysis by recognizing reality as it is:

  • Economic decisions are made in sequences where one precedes the other; and
  • Every economic decision today is based on information asymmetry, where we know less about the future than we know about the past.

It is this asymmetry that Say’s Law is trying to ignore, but Keynes recognizes. He made valuable contributions toward our understanding of the implications of historical time, but the most pertinent one is captured in his concept of uncertainty. Since the future has not happened yet, we have not way of knowing what it will look like unless we establish predictable conditions for future economic activity.

What does this mean? Suppose you wake up tomorrow on another planet. You know nothing about who you can trust, let alone how to interact with anyone. You don’t know how to get food, shelter or other necessities. Your future survival is genuinely uncertain; normally you could just extrapolate your past experiences, relying on the repetition of successful interaction from the past. But you can’t do that now.

This situation, while stylized and empirically unrealistic (I don’t think any of us will wake up on Hoth or the Klingon Homeworld any time soon), gives us a fair idea of Keynes’s concept of time and its inevitable companion: genuine uncertainty. But it can also be applied in a more realistic, everyday context. Here is how, and how this concept of uncertainty makes Say’s Law unworkable.

To reduce uncertainty about tomorrow, we build, so to speak, points of confidence in the future. We enter into agreements with people that hopefully will result in a gainful exchange of resources. I agree to do my neighbor’s taxes while he shovels snow off my driveway. If we are both happy with the exchange we decide to repeat it next year, or to agree on other gainful exchanges on a more frequent basis.

We have actually institutionalized the repetition of successful trade. Every retail store is an organization that reduces uncertainty and increases our confidence in the future. A mortgage payment is another example: we can set the terms we want for a regular expense for a considerable period of time into the future. An employment contract works the same way, setting terms on which a person can reasonably predict his future.

There are countless examples that in the same way make our future predictable: everything from street names to our Constitution serve the purpose of reducing uncertainty and improving the confidence with which we can go about our daily business. Taken together, all our economic, social and cultural conventions, habits, contracts and institutions create reference points in the future, thereby actually creating the future.

But how does all this relate to Say’s Law? It centers in on product prices. Under Say’s Law – and all mainstream microeconomics – prices are flexible, adjusting instantaneously to shifts in the market. When there is a slump in demand for a product, the price falls immediately until it brings supply and demand back in equilibrium, or vice versa for an increase in demand (with corresponding changes, obviously, in response to supply changes). The problem with this view is that it allows for a scenario where you wake up tomorrow morning and every single price that you encounter, from the gas pump on the way to work to the wage you earn, has changed radically overnight. Everything you thought you knew about how to plan your finances, be it as a household or as a business, is obsolete and irrelevant.

As mentioned, Say’s Law addresses this problem by simply assuming that you will know, the night before you go to bed, exactly what those prices will be. How you know it remains an enigma; it is just axiomatically stated that we all have perfect foresight of tomorrow.

This is, of course, a convenient way to protect a nice theory from a not-so-nice reality. It is, however, not a convenient way to start a foray into realistic economic analysis. For that, we have to rethink that market adjustment process in response to declining sales, where Say’s Law predicts a fall in the price. To do so, let us go back and remind ourselves of the single most important passage in Keynes’s General Theory of Employment, Interest and Money (1936, ch. 16):

An act of individual saving means – so to speak – a decision not to have dinner to-day. But it does not necessitate a decision to have dinner or to buy a pair of boots a week hence or a year hence or to consume any other specified thing at any specified date. Thus it depresses the business of preparing to-day’s dinner without stimulating the business of making ready for some future act of consumption. It is not a substitution of future consumption-demand for present consumption-demand, – it is a net diminution of such demand.

Say’s Law responds to this decline in demand by saying that restaurants and their patrons will agree on a price tomorrow, today. That price is known to everyone today, as they go about and plan their business today, for tomorrow. Say’s Law also prescribes that if there is a net decline in restaurant business, there will be an increase in the business of doing something else in the economy: the net reduction in dine-out spending comes with a direct promise by restaurant patrons to spend their money somewhere else in the economy.

At no point does the economy as a whole skip a full-employment beat.

Keynes, on the other hand, refutes the idea that the decline in spending in one market automatically means the increase in spending in another market. His refutation is based on his concept of time and of uncertainty, its inevitable companion. In the scenarion he lays out in the quote above, the decision not to have dinner today is motivated not by changing preferences where the consumer prefers to buy something else instead – as would necessarily be the case under Say’s Law – but by the consumer being less confident in his ability to manage his finances in the near future. He prefers to keep more cash on hand.

The reasons for his weakened confidence are important in a broader context, but not in order to understand the mechanics of Keynes’s Law and how it contrasts to Say’s Law. Given a weakening in consumer confidence, there is less money spent in the economy as a whole. Neither restaurant owners nor other businesses will respond by immediately cutting prices; their prices represent confidence in the sense that they can plan their finances from one day to the next, from one month to the next. A price cut upsets their short-term financial planning: their bills, from food supplies to wages, all come in fixed-price format.

A fixed price is also a reference point for those patrons that still frequent the restaurant. While they would certainly appreciate a price cut, they also appreciate a price that remains fixed in both directions: a price that changes frequently in response to changes in supply and demand eventually becomes unpredictable. Where fixed prices are a bulwark against uncertainty, flexible prices are a venom for confidence.

This is not to say prices rigidly remain unchanged. Marginal adjustments take place all the time, but they are minor exceptions to the major trend of price stability. However, it is also important to recognize that even if prices did fall in response to an uncertainty-driven decline in consumer spending, it is by no means a given that consumption would pick up again. As I demonstrated in my own study of consumer confidence and price stability, it is practically impossible to identify price adjustments that would accommodate a rise in uncertainty and still be palatable for price setters, i.e., businesses who have fixed-price bills to pay.

It takes a recession to set price changes in motion at any macroeconomically meaningful level. However, when those price changes occur the factor determining the rise in economic activity is whether or not employers have experienced a decline in the real wage. If their labor costs have fallen enough to raise revenue per hour worked, they will summon the confidence needed to again expand business activity and payroll.

In practice, a decline in the real wage does not have to appear as a strict adjustment of prices, i.e., per-unit sales revenue vs per-unit labor. All it takes is a rise in revenue from either higher prices or increased productivity. Contrary to what Say’s Law predicts, a rise in the real wage at the bottom of the recession would increase costs for businesses and thereby delay the recovery, or even aggravate the recession.

We discovered the flaw in Say’s Law when we sequenced it, finding quickly that it does not work under historical, or discrete time. By the same token, Keynes’s Law shows its prowess under sequencing:

  • A decline in consumer spending depressed business sales;
  • Businesses largely maintain prices in order to maintain, short term, the predictability of their own finances;
  • All other things equal, any price adjustments made have to result in a decline in the real wage before businesses are motivated to increase activity and expand payroll.

It is not possible to predict with any precision how long a recession will last when it begins. Generally, it is a good rule of thumb that an organic recession – not one caused by an artificial government-imposed economic shutdown – lasts for 18-24 months. However, they can last longer, driven exclusively by an adverse shock to confidence in the economy. The reason is to be found at the microeconomic level; contrary to what some of Keynes’s critics suggest, he was very well aware of microeconomics.

The problem for many of Keynes’s critics is that they adhere ot the fallacy that microeconomic behavior, i.e., decisions made by individual economic agents, is driven by a rationality that will always result in full employment general equilibrium at the macroeconomic level. This is not at all the case: as Keynes explains on at least three occasions in the General Theory, actions that are perfectly rational for the individual in pursuit of goal X, may very well result in non-X happening. If it is rational for all consumers to decide to cut spending today so they will be better off tomorrow, they may all find themselves worse off tomorrow.

Say’s Law proponents would suggest that if only they all had perfect foresight, they would know that it is better not to act in pursuit of X, or to anticipate non-X as the inevitable result of X. However, that assumption is premised on the existence of logical time as opposed to historical time. Tomorrow does not exist today; when our knowledge of tomorrow turns out to be wrong, we already have proof that we do not have, nor can we achieve, perfect foresight. But even more so: to realize that the pursuit of X will result in non-X, we all need to beat the cat-and-tail game I discussed in my book on prices and uncertainty. To get the knowledge that everyone else is going to pursue X, we need to gather information about what everyone else is doing; once we have done that, we adjust our economic plans accordingly. But since everyone else has done the same surveys, they now change their decisions as well. Our knowledge of their behavior is now obsolete and we have to do the same round of information gathering again.

The cat tries to catch its own tail, but every time he tries the tail moves.

In other words, some rational behavior leads to an irrational outcome, precisely because time is historical and not logical.

Keynes’s Law explains recessions, even depressions. At no point in his scholarly work – and certainly not in the General Theory – does Keynes say that government should play a regular role in the economy. The only role he assigns to government is to play an active role in moving the economy out of a depression – not a regular recession.

I wish that Keynes’s critics would be as interested in reading Keynes as they are in reading Hayek or Murray Rothbard. Maybe then we could actually summon some forces among conservatives and libertarians to get decent, workable economic reforms done to reduce the size of government, phase out the welfare state and restore the full force of economic freedom to our economy.

Click here for Part 1.