[An updated version of this article can be found at Inflation in the 2nd edition.]
Inflation is the loss in purchasing power of a currency unit such as the dollar, usually expressed as a general rise in the prices of goods and services. A classic example is the Great Inflation of the Roman Empire. Successive emperors replaced a steadily increasing fraction of the silver in their ancient currency, the denarius, with base metals like bronze or copper. As a result prices rose inexorably despite repeated attempts to restrain them through legislation. Diocletian, rather than taking responsibility for the debasement, attributed the rapid inflation of his day to the avarice of his subjects. His famous edict of
a.d. 301 threatened with death any vendor who charged prices exceeding official limits. But inflation ran along unhindered for another century until an alternative currency, an undepreciated gold coin known to Shakespeare as the bezant, became the customary unit of account, spreading throughout Europe and lasting well into the Middle Ages.
In modern times inflation continues to be blamed on private greed, and governments still seek to restrain it by decree, sometimes even devaluing their currencies as they do so. The United States has experienced much inflation during the twentieth century, especially since official efforts to maintain the gold price at thirty-five dollars an ounce ceased during the presidencies of Lyndon Johnson (in the world market) and Richard Nixon (through the "gold window" open only to foreign central banks). An annual inflation rate of 4 to 5 percent, once thought to be calamitous, has become routine.
We have many measures of inflation, but none provides a truly reliable gauge of inflation at any specific time. The most widely watched measure is the consumer price index (CPI), published monthly by the Bureau of Labor Statistics. Subindexes are available for different cities and for many different classes of goods and services.
One problem with the CPI is that the weight attached to each class of goods and services is held constant for years at a time. Therefore, when consumers lower their cost of living by buying more items whose relative price has fallen and fewer items whose relative price has risen, the CPI will not show a decline in the cost of living. Moreover, the difficult problem of allowing for changing quality has never been solved. Nor can the government inspectors who collect the data from retailers track down all the sales and discounts of which consumers are so keenly aware. As a result of these and other factors, the consumer price index reflects inflation trends only with a long delay and portrays an artificially smooth path for the inflation rate.
Other popular indicators of inflation include producer prices (formerly known as wholesale prices) and unit-value indexes for imports and exports. As we move back through the distribution chain from the consumer toward the supplier of raw materials, a more jumpy picture of inflation is revealed at each step. Commodities, whose prices can be monitored continuously on centralized exchanges, and which are easy to measure, are the most volatile indicators of all. An index of commodity prices, when plotted on a graph, looks much like an index of stock prices. But its ups and downs are significant; it provides warning one or even two years ahead of movements in the consumer price index.
In the news media, discussion of inflation often takes a "bottom up" view. Each month's change in the CPI can be, and is, split up into dozens of components, such as food, energy, and housing. It is tempting to see the sectors where prices rose the most as causes of the observed inflation. Sometimes policymakers speculate that if "price pressure" in those areas could be relieved, overall inflation could be reduced.
This way of looking at inflation is mistaken. The prices of some items always are rising or falling relative to others. This is a natural feature of the way a market economy adapts to changes in supply or demand. Rapid price increases within a single sector, though often labeled "sectoral inflation," are partly the result of an adjustment in relative prices and partly a manifestation of the overall inflation rate. They may have no causative significance whatever. When we watch the tide come in at the beach, we know that it is not caused by the waves, however forceful they may be. Inflation is not simply the sum total of a collection of independent price changes, as the arithmetic of the CPI implies. It is the degree to which all of those prices move in concert.
Another popular game is to sift out the more volatile items in the basket of goods and services—often energy and food—and focus on the remainder as a truer "underlying" or "core" rate of inflation. This exercise, though it succeeds in producing a less volatile index, is dubious. The least volatile components are not necessarily the most informative. Some of them appear to be unresponsive to economic forces because of pitfalls in measurement or stickiness in their speed of adjustment to market forces. The price of rental housing, for example, is fixed month-to-month by contract. At the other end of the scale, some of the most volatile items—such as precious metals—are highly informative, to the extent that their movements anticipate a broad range of sectors where price changes have not yet been perceived.
What does inflation cost? There are polar opinions here, and a lively debate. In and of itself inflation "costs" little or nothing because it consists of nothing more than a change in the units we use to measure prices throughout our economy. It is confusing and irritating to keep requoting prices, but that's something people get used to, as recent writers like Paul Krugman and Alan Blinder have emphasized. From this point of view it is possible to see a steady inflation rate as high as 10 percent annually as nearly costless. But dwelling on this problem misses the point of the debate.
Economists who view inflation as a very serious problem point to what they call the "inflation tax." By this they mean the reduction in the purchasing power of the cash balances held by the private sector—like a wealth tax. This tax is a drag on the economy—an "efficiency loss"—because it induces people and businesses to economize on cash balances, making it more difficult to participate in the money economy.
Economic losses associated with the inflation tax and other distortions are known as the "welfare cost of inflation." At one extreme of the debate, Harvard economist Martin Feldstein has claimed that the present value of the losses that result from unending inflation may be infinite! His argument is that each year the cost to the economy grows in proportion to society's money balances. Because the rate of growth of money balances exceeds the interest rate he uses to calculate the present value, the present value is unbounded.
But the force of the inflation-tax argument has been depleted in recent years by the increasing tendency to hold cash in the form of money market mutual funds and bank deposits that pay interest. The higher the expected rate of inflation, the higher the interest rate paid by mutual funds and banks. People have shifted their cash balances to these types of accounts only recently because government regulations used to prohibit the payment of interest on checking accounts.
Quite a different problem results from the collision of inflation with the U.S. tax system, particularly the federal taxes on personal income, corporate profits, and capital gains. Progressive rate structures were intended to shift tax burdens from low- to higher-income groups. But over the years they have instead imposed on the general income-earning population high tax rates that had originally been thought appropriate only for millionaires. A family with a constant real income of $50,000 (in 1978 dollars), for example, was pushed from the 28 percent bracket in 1965 to a 46 percent rate in 1978. Its average tax rate rose from 16 percent to 23 percent. Offsetting reductions in tax rates have been extremely slow to develop, in spite of across-the-board rate cuts during the Kennedy-Johnson and Reagan years. Instead, government spending has tended to absorb revenue unintentionally collected as a result of escalating tax brackets driven by inflation.
The increase in government spending could be claimed as either a cost or a benefit to the economy, depending on whether one wants more or less government spending. But there is a real cost that is not ambiguous. High tax rates on employment, on business investment, and on the accumulation of capital deter all these activities in favor of untaxed uses of the economy's resources and, therefore, impede output and growth.
The effects are visible in the lurching path that the economy has followed in the past few decades, coupling highs in inflation with lows in economic growth. Since 1953, as table 1 shows, there has been a consistently inverse relationship between inflation and growth in the U.S. economy. This is true not only for the 1973-74 and 1979-80 periods, when large increases in oil prices were partly responsible for both high inflation and low growth, but for other years as well. Such evidence undermines the widely held belief in the "trade-off" between inflation and unemployment.
Still more difficult than measuring inflation is the problem of identifying its root causes. In spite of its long and rich history, few subjects in the field of economics are more confused. Professional economists have still not reached broad agreement as to the origins of the inflation process. Two camps dominate the debate. Some see inflation as a malady of the currency (as was surely the case in the Roman Empire). In the words of Milton Friedman, "Inflation is always and everywhere a monetary problem." Others see nonmonetary forces at work, such as monopolies, union demands for higher wages, oil politics, or the "wage-price spiral."
Some nonmonetary ideas are illogical. The existence of monopoly power or union power might be argued to raise prices generally relative to what they otherwise would be. But a continuing price rise year-in year-out requires a continuing increase in the degree of monopoly or union power in the economy. This is neither plausible over long periods of time, nor consistent with evidence from recent decades for the United States.
Nonmonetary theories of inflation traditionally separate "demand-pull" sources from "cost-push" factors like oil, monopoly power, or wages. A surge in the demand for goods and services in general ("aggregate demand") is thought to "pull" prices up across the board, especially when "aggregate supply" is held back by inertia or capacity limitations. Skeptics rightly question how demand could constantly outstrip supply. Surely, demand must originate from purchasing power, purchasing power from wealth, wealth from income, and income from the ability to produce (and hence supply) goods and services. This contradiction was understood early in the nineteenth century by Jean-Baptiste Say and others.
Other logical objections to the idea of demand-pull inflation center on the importance of money. How could prices rise without a commensurate increase in the quantity of money in private hands? If such a thing happened, the purchasing power of the quantity of money would have declined involuntarily, and that would not be consistent with market equilibrium. Economists of the "monetarist" school emphasize the power and discretion of government to vary the money supply, causing private markets to bring the economy's price structure into conformity.
Finally, there is strong, though surprisingly little known evidence against the demand-pull view that excessively rapid economic growth ("overheating") is an important source of inflation. The evidence in table 1 shows that the reverse is nearer the truth for the United States in recent decades. Inflation has tended to increase in periods of slow growth or recession and decrease in periods of expansion. The idea that growth risks inflation is not on as strong a footing factually as the idea that inflation hurts growth.
Among those who attribute inflation to monetary causes, at least two quite different views exist. The monetarist view is that increases in the quantity of money cause inflation. Critics of this view point out that the quantity of money is difficult to define, especially when funds can be transferred electronically and credit cards can substitute for cash balances. It can also be argued that people have freedom to choose the quantity of money they want to hold rather than merely accept the quantity the government wishes to impose upon them.
The other monetary view, held historically by opponents of fiat (i.e., government) paper money, and by advocates today of restoring the gold standard, is that the quantity of money can take care of itself. What really is needed, according to this view, is a mechanism for keeping the price of the currency stable, for providing an anchor, so to speak.
Governments have been slow to accept the recommendations of either of these camps. That probably is because either a strict monetary rule or strict adherence to a gold standard or other price rule would place strict limits on discretionary government management of the economy.
David Ranson is president of H. C. Wainwright and Company, Economics, an investment research firm in Boston. He was formerly an assistant to the secretary of the Treasury in Washington.
Blinder, Alan. "The Efficiency Costs of Inflation: Myth and Reality." In Blinder. Hard Heads, Soft Hearts. 1987.
Feldstein, Martin. "The Welfare Cost of Permanent Inflation and Optimal Short-Run Economic Policy." Journal of Political Economy 87 (August 1979): 749-68.
Jones, A. H. M. "The Anarchy." In Jones. The Later Roman Empire. 1964.
Krugman, Paul. "Inflation." In Krugman. The Age of Diminished Expectations. 1990.
Laffer, Arthur, and David Ranson. "Inflation, Taxes and Equity Values." Report prepared for H. C. Wainwright and Company, Economics. September 20, 1979.
Wanniski, Jude. "Money and Tax Rates." In Wanniski. The Way the World Works. 1978.