Stable prices, money, and the cost of living

American Journal of Economics and Sociology, The, Jan, 2005 by Martin Shubik

I

The Macroeconomic Design of Indexed Units of Account

THE QUESTION OF the desirability of stable prices belongs to the realms of political economy. Political economy in its old-fashioned sense covered both theory and practice and involved or at least was openly qualified by political, administrative, sociological, and psychological considerations.

Currently, as the social sciences have developed, economics and political science are more or less separate disciplines and micro- and macroeconomic theory and practice are often far apart.

From the viewpoint of applied macroeconomics and in the belief of most economists, the proposition that the economy would benefit from a more or less stable price system is well accepted. Commonsense wisdom also coincides, in this instance, with the beliefs of most economists. The implications of wage stickiness and the possibility of a money illusion have been discussed at least since Keynes (1936).

The discussions of Robert Hall (1997) and Robert Shiller (1998) deal with the problem of how to construct an indexed unit of account via a new financial instrument. In both instances they note the successful utilization of the unidad defomento in Chile as an example of where a small country has designed a bearer security linked to the peso, whose peso content changes daily, tracking the cost of living as closely as possible. Shiller in particular outlines various policy choices concerning the linkage of such an instrument to a consumer price index or to a per capita income index.

It is well known in medicine that one may stumble across a practical means to treat a disease without knowing why it works. I suggest that much of macroeconomic engineering falls in the same category. A further complicating feature in economics is that often we do not know how fatal is the disease. It can be argued that "a little inflation" is a good thing; but "too much inflation" is bad for the economy. My intuitive feeling is that an inflation rate of 1 or 2% may be a salutary "little inflation." Much above that level I suspect that in some broad sense inflation is not desirable, although if it were constant and cleanly predicable then larger rates might be acceptable.

In much of microeconomic theorizing Pareto optimality is discussed and attempts are often made to devise criteria for fair division, usually in a static model. In political reality neither optimality nor fairness are easy to define in a dynamic context. Billions of dollars are at stake, and large political constituencies influence and are influenced by choices of indices such as a cost-of-living index.

I believe that there are an array of new instruments of the variety indicated by Hall and Shiller that can be designed now. I am less sure how badly they are needed.

Recognizing that practical policy issues must be dealt with here and now on an ad hoc basis by the macroeconomists, the possible value-added contribution of the microeconomic theorist might be in raising some questions concerning what is being controlled and why?

II

What Are the Questions and Are There Microeconomic Answers to Them?

1. DO WE KNOW that a fixed high inflation (say 10% per annum) is necessarily bad for the economy? It is not clear to me that we do.

I suspect that the problem is in the variation and unpredictability of the inflation rate, not the size of the rate. If it were feasible to control the variation, the mean inflation might not be too important, although changing prices might be a nuisance.

2. Is the exercise in the design of an indexed unit of account essentially an ad hoc local fix we deem to be worth doing under the criteria of political feasibility and local, topical economic optimality, or can it be justified in terms of global properties?

I suspect that the problem is primarily political and administrative. The solution suggested is ad hoc, and the problem of controlling inflation could be treated better by the control of the money supply and the near money substitutes. My guess is that in much of South America the political and bureaucratic structure is such that a partial indexing serves to make it easier to write long-term contracts without having to worry as much about the politicians and bureaucrats.

3. Do we have a theory of economic dynamics that we believe in that at least enables us to say how important indexing is? Do we believe that what is at stake is of the order of 0.1% or 1% or 5% of GNP, or do we have some measure of efficiency?

We do not have a universally accepted dynamics, and I am not aware of a consensus on indexing and on its relative efficiency over other forms of monetary policy.

4. Is there any need whatsoever to engage in indexing? Could it be that the solution lies in money and credit control? Ricardo's reason for supporting the utilization of a gold or gold-backed currency was not that he regarded gold as a necessity, but that he regarded it as a better alternative than utilizing paper, which would be subject to manipulation by those who issued it.