Uncertainty, Macroeconomic Stability and the Welfare State

Sven Larson

• Publisher: Taylor & Francis
• Year: 2019
• Print ISBN: 9781138724945
• eBook ISBN: 9781351754132
• Edition: 1
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

This title was first published in 2002: This monograph sets out to model a macroeconomy that is inherently unstable because of qualitative - or Keynesian - uncertainty. By modelling a macroeconomic theory, this approach to fixed or sticky prices also investigates the link between uncertainty, sticky prices, and macro-stability - by suggesting that such prices improve economic activity rather than impeding it.

1 Uncertainty and Prices

That economic agents prefer predictability to uncertainty is as elementary a statement as a statement that they are willing to buy more of a commodity the lower the price is. Nevertheless, it points at a property of the economic system that has just as profound effects upon how the system works as has the price statement. Once we acknowledge that economic agents actively choose predictability over uncertainty, we can see profound effects on how economic activity is organized: as we will see in this chapter, unless any other form for interaction between consumers and entrepreneurs is reliable as a means to manage uncertainty, historically successful economic interaction will be repeated to assure as far as possible that consumers and entrepreneurs can experience the same success as they did before.

This don’t-mess-with-success principle does, among other things, mean the repetition of a price. If a buyer and a seller have successfully exchanged on a Monday it would be only logical that they repeated all the premises of that exchange on a Tuesday, unless they had an explicit reason not to. If the price was a loose end in their expected Tuesday exchange it would only invoke uncertainty in both parties — uncertainty that would proliferate into other exchanges where the buyer and seller are involved. A central point in this chapter is that in their struggle to manage in a world of uncertainty, consumers and entrepreneurs prefer sticky prices to flexible prices. This preference will be a crucial element in our development of a model of macroeconomic stability. A key point we will make is that sticky prices interact with the confidence of economic agents to inspire a higher propensity to consume if interaction sets a positive spiral in motion — successful uncertainty management leads to more spending per unit of income and time — or a lower propensity to consume if the spiral turns negative.

The statement that sticky prices are better — in the eyes of economic agents — than flexible prices is by no means original to this thesis. It is a general trend in results of recent empirical studies that economic agents exhibit a preference for sticky prices over flexible prices, not by explicitly stating that they are better off when prices remain in place, but by showing that agents normally use fixed, not flexible prices. It is reasonable to read such actual behavior as an expression of a sticky price preference, although theory does not always make this interpretation. So far, theory has given three different responses to the challenges of empirical studies — a fourth is added here. The first of these responses is the most reluctant to the preference interpretation: it refers to a sticky price regime as making an inferior contribution to employment and production as compared to flexible prices Barro and Grossman 1971; Lindbeck and Snower 1988). According to this response, obviously price stickiness — whether on product or labor markets — ought to be done away with.

A second response provides some theoretical support for the empirical findings, but the formal model is absent (Okun 1981; Hicks 1989). It is implicitly suggested that modelling would lead to formal proof that it is better for an economy to operate with sticky prices than with flexible prices.

A third response move towards outlining foundations of formal models within a more explicit Post Keynesian framework (Eichner 1976; Davidson 1978; Lee 1998). The fourth response, which is in essence the contribution of this book, is part of the same breed, as it shares the same Post Keynesian platform. It differs in focus, however, since it emphasizes macroeconomic uncertainty and its relation to consumption decisions. It is a separate response because of its emphasis upon consumption as the entry of uncertainty; empirical evidence is presented in section 1.2.

In what follows we shall be preoccupied predominantly with theory to build our argument; empirical evidence showing the dominant use of sticky prices is abundantly available. At the same time, that evidence does not ask whether sticky prices are preferred to flexible prices by economic agents. In order to show that our argument on uncertainty and consumption is valid, results will be presented of a study where consumer attitudes to sticky and flexible prices are revealed. This contribution is the first to explicitly inquire about uncertainty management based on a Keynesian definition of uncertainty. We present its results in the context of other studies and treat it as one verification of the overall theoretical argument — one component of a gradual process of verification or falsification — and not a definitive achievement.¹

In the first section of this chapter we define uncertainty in more detail and discuss what consequences the definition has for economic analysis. Our focus is on the economic agent and her role in the dialectical process in which she has to participate to make uncertainty manageable. The relevance of incorporating uncertainty into economic analysis is stressed, asis the significance of this incorporation for how we define the economic agent and the system she is part of.

Section 1.2 presents empirical results from earlier studies and adds the results of a new study. The common denominators of the studies are analyzed, as are the possible theoretical implications.

Section 1.3 gives a more formal shape to the unfolding theoretical argument by suggesting how uncertainty and consumption can actually be related to each other.

1.1 Uncertainty and uncertainty management

The two different interpretations of uncertainty that were mentioned above have in common that they claim the agent is unable to foresee perfectly what the future will bring to her in terms of yields on economic activities. That, however, is where the similarity stops; from hereon we shall explore the unique character of Keynesian uncertainty and its distinct axiomatic foundations in our perceptions of such fundamental properties of reality as time and space. We will also put this kind of uncertainty in relation to the alternative definition suggested by mainstream theory.

First we define Keynesian uncertainty. Then we discuss what this definition implies in terms of defining the economic agent, before we can proceed with a formalized analysis.

1.1.1 Time and organic interdependency

Fundamentally our society stands and falls with our collective ability to handle uncertainty. In societies and economic systems that prevail, uncertainty of the Keynesian kind has been eliminated from our daily economic activities. What remains in our every day life is another kind of imperfect foresight, a kind that in literature more loyal to mainstream economics is referred to simply as uncertainty:

[The household] has a whole variety of important decisions to take. Among these are decisions over whether, where and how much to work, over the allocation of income to present and future consumption, over the allocation of any saved or inherited wealth among alternative saving opportunities, and over the allocation of expenditure on goods and services of various kinds. (Hey 1979, p. 3)

We may refer to this kind of uncertainty situations as constituting a microeconomic layer of uncertainty. In contrast, there is a macroeconomic layer which, unlike its microeconomic counterpart, is characterized by organic interdependency. The key difference between the two layers of uncertainty is this: while it is perfectly possible for the individual economic agent to evade the negative consequences of microeconomic uncertainty, it is not possible to do that in the macroeconomic layer. In the macroeconomy our own activities to evade a possibly uncertain situation will instead affect the very conditions of that situation. Consider, e.g., consumer cash flow planning under uncertainty. Suppose that one day we decide that prices of food, clothes, daily transport etc. should be set anew every day by a market maker in an auction-style procedure — that, in other words, price stickers be banned. The motive is to improve the functioning of markets in compliance with standard microeconomic theory of free competition. This means we are not able to use historic information to form expectations as to what the prices will be tomorrow. (Even after a number of days with flexible prices we are, strictly speaking, unable to foresee tomorrow’s prices.) Being genuinely uncertain as to the cost of consumption tomorrow — what do we do? The most prudent reaction would (as this book will show) be to cancel as many planned transactions as possible, whereupon the activity level of the economy is adjusted downward, not upward as was the intention behind the reform. In order to attract buyers, sellers then have to promise predictable, non-flexible prices — again actions that contradict the intentions behind the reform.

By changing their behavior in order to meet an uncertain situation, consumers change the situation itself. They interact with their economic environment and change the conditions of their own decisions. This interaction is the trademark of macroeconomic theory and its component in macroeconomic uncertainty is crucial to our following analysis.

Davidson (1994) and Dow (1996) provide good explanations of what difference it makes to incorporate both layers of uncertainty in economic analysis. Davidson’s distinction between ergodic and non-ergodic processes draws a line of demarcation that resembles the line between micro- and macroeconomic layers of uncertainty. The category of ergodic processes belongs to the microeconomic layer — its processes…

…yield averages over time which provide good estimates of space or statistical averages at a point in time, and vice versa; statistical averages then are good predictors. (Dow 1996, p. 36)

In other words, we can study ergodic processes to get good historical information about how certain microeconomic systems work, and then we can use the statistical results from those processes to extrapolate a pattern into the future as a forecast. Since an ergodic process is closed and deterministic, such forecasting generally bears fruit; we can quantify the probabilities of success in forecasting and tailor the processes for formal logic (Dow 1995). By contrast, in non-ergodic processes of the macroeconomy…

…average values vary over time. It follows from Davidson’s vision of the economy in terms of processes which unfold irreversibly in historical time that most processes are non-ergodic and thus not amenable to representation as ergodic, stochastic processes. (Dow 1996, p. 36)

A constitutional difference between ergodic and non-ergodic processes is the nature of time. In ergodic processes we do not take time into consideration as a determining variable — which is equivalent to saying that time is logical. In non-ergodic processes, on the other hand, time is a relevant determinant, being a series of sequences which “unfold irreversibly”. In a similar fashion time is a distinguishing component in our two-layer definition of uncertainty. When dealing with microeconomic uncertainty we are (in principle) independent of time in the sense that our acts today to reduce uncertainty will not affect the premises of that act — they will, in other words, not interact with the structure where they are performed. On the other hand, macroeconomic uncertainty is time-dependent in the sense that our acts — through their effects on our economic environment — irreversibly affect our economic environment. The irreversibility is guaranteed by the nature of time: once yesterday is gone, we can no longer return to its state of affairs to make the changes we have caused undone.

It is therefore of central importance that we recognize the role of time in economic activity as well as analysis:

Time is a device that prevents everything from happening at once. The production of commodities takes time; and the consumption of capital goods and consumer durables takes considerable time. Decision makers recognize that the outcomes in terms of profit and/or utility to be obtained from engaging in production and exchange activities require a significant amount of historical time to pass between the point of time a decision is made and the time the consequences are experienced. (Davidson 1994, p. 88)

There is, in other words, a substantial benefit from incorporating time into economic analysis, given that it is defined as a series of one-way causal sequences. This benefit is the ability of analysis to take both layers of uncertainty into account as meaningful components. What this means in practice is that when sequence 1 is “happening”, the legacy for sequence 2 is in a gestation process, being created by the acts and inter-acts of agents during sequence 1; the aggregate result of these acts — the total legacy package that sequence 1 leaves behind — is not clear until sequence 1 has come to an end. Therefore we cannot foresee this result by collecting a “sufficient” volume of information but instaed we have to try to create that result already at the outset of sequence 1.²

In an ergodic process the possibility for each agent to obtain prefect foresight is provided by the limited boundaries of the process. Every agent can — if given the proper means to do so — obtain enough information to foresee the price of shoes tomorrow despite the fact that the price is flexible and therefore free to move anywhere demand and supply will take it. The impediment to the agent’s prediction lies in her having access to the proper means of information retrieval — and in this context a sticky price may lead her into the wrongful conclusion of trading at a price that is not the equilibrium price of the market. The information on what the future equilibrium price will be is there, on the other side of the veil of ignorance.

In a non-ergodic process the flexible price is no longer a surmountable problem as in the ergodic process. Since the future does not exist until it becomes the present, it cannot be foreseen on the macroeconomic level where all economic activities are organically interdependent. With a future that is genuinely uncertain, economic agents will have to create a fixed point that will serve as a reference value for their future planning - one such fixed point is a sticky price. In absence of such fixed points economic agents who are reluctant to exposing themselves to contingencies will choose to abstain from committing the limited resources they command. In a world of macroeconomic uncertainty, “he who hesitates is saved to make a decision another day” (Davidson 1978, p. 16).

This difference between ergodic and non-ergodic processes in terms of what role sticky prices can play can be viewed in from a different perspective. In an ergodic process agents have accomplished a state of foresight where every resource they employ is used in the best possible way, every possible outcome is known and either certain or calculable by probability. The entire system of ergodic processes is perfectly predictable until some single factor wedges in between the agent and the future, causing the veil of ignorance to fall before her face. It is important to notice that one single variable can cause that to happen — a single disturbance to the agent’s information retrieval is in other words a sufficient condition for the ergodic process to gear down. As a result, this process will yield a use of resources inferior to the one previously experienced. Such a disturbance is, e.g., the fixing of a price which disrupts the market maker’s trial-and-error process of identifying the equilibrium price.

It is not the case, however, that an economic singularity of this kind will bring an ergodic process to a stand-still. Since the singularity is the only impediment to a fully certain pattern of resource distribution in an ergodic process, the possible alternatives for the future are few enough to easily let the agent keep track of them. The uncertainty injected into the process by an economic singularity is of such a kind that estimations of probability will be enough to guide the individual agent to a prudent decision as to what to expect of the future.

Uncertainty and time as features of theory are thus closely connected; another feature of theory that, in turn, is closely connected to time and uncertainty is causality. The relation between cause and effect when time is logical is different from what it is under a regime of sequential time. Logical time ...

Table of contents

• Cover
• Half Title
• Dedication
• Title Page
• Copyright Page
• Table of Contents
• Preface
• Introduction
• 1 Uncertainty and Prices
• 2 Demand, Supply and Stability
• 3 Collective Stability
• Concluding Words
• Appendix
• Bibliography
• Index