Dr. Selgin is Assistant Professor of Economics at the University of Georgia and the author of The Theory of Free Banking.
Lately a consensus has formed among policymakers that a stable price level or zero inflation should be the goal of monetary policy. Federal Reserve Chairman Alan Greenspan has recently expressed sympathy for this view, long favored by some other members of the Fed Board of Governors. A few years ago the House Subcommittee on Domestic Monetary Policy introduced a Joint Resolution that would have required the Fed to achieve and maintain zero inflation within five years, Today, Congress is considering amending the 1978 Humphrey-Hawkins Act so as to make achievement of zero inflation the sole object of monetary policy.
Zero inflation is, to be sure, a more realistic goal for monetary policy than such things as “full employment” or economic “fine-tuning.” Nevertheless, it is far from being the ideal policy its advocates proclaim it to be.
Price Stability or Stability of Spending?
The zero inflation norm goes back to classical economics and has inspired countless monetary-reform proposals during the last 100 years. One would think that such a longstanding ideal must be solidly grounded in theory. But the truth is otherwise. In fact the zero inflation ideal is largely dogma, founded upon the unrealistic assumption of a stagnant or stationary economy where the productivity of labor and capital never changes.
In such a stationary economy, price stability goes hand-in-hand with stability of total spending, or “aggregate demand,” measured in dollar terms. Economists generally favor stability of money spending because it allows the typical producer to just recover his money costs of production, avoiding depression on one hand and overexpansion of industry on the other. Thus stability of “aggregate demand” avoids deviations of real output from its “natural” level. But zero inflation implies stability of spending only in a stationary economy. In a growing economy with more to be bought, stability of money spending requires falling prices. In an economy where productivity is declining, stability of spending requires that prices generally rise. Because they overlook the reality of changing productivity, proponents of zero inflation wrongly conclude that the benefits of stable spending can be had by keeping the price level constant.
A popular argument for zero inflation is that unanticipated price-level movements lead to unfair transfers of wealth. When prices rise unexpectedly, debtors gain at the expense of creditors because loans are repaid in dollars having less purchasing power than when the loans were originally made. When prices fall, creditors profit at the expense of debtors. Zero inflation, it is claimed, would prevent such unjust transfers.
Although the argument is valid for a static economy, a zero inflation policy enforced in the face of changing productivity would itself lead to unjust redistribtions of wealth. Any overall change in productivity implies a change in real income for the economic community taken as a whole. Distributive justice then becomes, not a matter of avoiding “windfall” transfers of wealth, but one of deciding how an increase or decline in overall wealth should be shared. Imagine the consequence of an unanticipated, all-around doubling of productivity in the United States. A halving of product prices would, here as when productivity is constant, double the real burden represented by each dollar of debt. But most debtors would be compensated by a doubling of their real earnings. Creditors, in turn, would enjoy a higher real return on their loans. But their gain would merely reflect a pro rata share of similar gains being enjoyed by the rest of society. To deprive creditors of their share by stabilizing the price level would be arbitrary at best.
Moreover, a zero inflation policy that would be arbitrary when productivity is improving could lead to disaster were productivity to fall significantly. Zero inflation would then require a forced contraction (via tight money) of spending to offset the normal tendency for the prices of scarcer goods to rise. Debtors would find their real income reduced, but the amount of real income needed to repay each dollar of debt would be unchanged. Few people would call the resulting rash of defaults and bankruptcies “just.”
Helping Prices Do Their Job
Another argument for zero inflation is that price-level changes interfere with the price system’s ability to allocate resources. Because it takes time and effort to make money-price adjustments, changes in the relative values of different goods should be signaled with as few money-price changes as possible. Otherwise the risk is great that incomplete or incorrect price adjustments will lead to economic waste. Proponents of zero inflation claim it would allow the price system to do its job with a minimum of money-price changes by eliminating any need for general price changes to offset changes in the supply of or demand for money.
This argument, too, is only valid for a static economy: When productivity changes, a change in the general price level is not only consistent with, but essential to the efficient working of the price system. Such a price-level change merely reflects changes in real costs of production. Suppose for instance that the cost of producing computers falls to half its former value, while the cost of producing other goods remains unchanged. The relative price of computers needs to fall below its former level. If the public continues to spend the same amount of money on computers, buying proportionately more units as the money price falls, the needed relative price adjustment is easily made by halving the money price of computers, leaving other money prices unchanged. Money spending would remain stable with no need for any increase in money supply. A zero inflation policy, in contrast, to keep the average level of prices constant, would require a monetary injection to enhance spending so that the price of computers falls by less than one-half and all other prices rise slightly. The zero-inflation policy clearly places a greater burden upon the price system, with greater opportunities for misdirection of resources.
Likewise, the best way to handle a fall in output, like an OPEC-inspired cut in oil production, is to allow the fall to be reflected in higher prices. Manipulating the money stock to keep prices from rising would reduce the price-system’s ability to convey useful information about the true state of resource scarcity.
Many economists recognize the need to allow money prices to reflect changing output conditions in particular industries. Yet they refuse to extend their logic to situations involving general changes in productivity. They imply that doing so involves a fallacy of composition. But where lies the fallacy? They do not say.
If Not Zero Inflation, What?
Instead of aiming for zero inflation, a desirable monetary policy would stabilize total dollar spending, as measured by the money value of GNP or (more appropriately) sales of final goods and services. Money GNP stability is automatically approximated by a regime of free banking with a fixed stock of bank reserves. If a central bank controls the money supply, it could be assigned the goal of stabilizing money GNP. Stable spending would achieve all the really desirable ends sought by proponents of zero inflation, and would do so regardless of the state of productivity. Under stable spending, sustained improvements in productivity would necessitate falling prices; but these would not involve the bad side-effects usually associated with deflation. On the other hand, falling output would cause “inflation” of the price level, but without the pernicious effects of inflations stemming from excessive money injections. Finally, because money supply changes are more closely related to spending changes than to price-level changes in the short run, a spending target would be easier to enforce than a price-level target.
Zero inflation has its merits as a rough-and-ready policy goal. It is certainly better than the inflationary chaos that prevails in most fiat money regimes. But as a policy ideal it leaves a lot to be desired, and is plainly inferior to a goal of stable spending.