FRBSF Economic Letter
2001-03 | February 2, 2001
Inflation: The 2% Solution
By the beginning of the 1980s, double-digit rates of inflation had become so pervasive among industrialized economies that they were viewed as a major deterrent to global economic growth. Since then, an explicit policy goal of low inflation has become a mantra for policymakers, and many countries, such as the U.K., New Zealand, Australia, Japan, Sweden, and the eleven countries under the European Central Bank (ECB), have enacted fundamental reforms to achieve that goal.
- The Friedman Rule and the benefits of 0% inflation
- Are there benefits to inflation?
By the beginning of the 1980s, double-digit rates of inflation had become so pervasive among industrialized economies that they were viewed as a major deterrent to global economic growth. Since then, an explicit policy goal of low inflation has become a mantra for policymakers, and many countries, such as the U.K., New Zealand, Australia, Japan, Sweden, and the eleven countries under the European Central Bank (ECB), have enacted fundamental reforms to achieve that goal. Specifically, they have made their central banks more independent and thus insulated them from the temptations of inflationary finance; furthermore, in most of these cases, as well as in the U.S., central banks have practiced a greater degree of openness or transparency about monetary policy decisionmaking to give the private sector a better opportunity to monitor their activities.
Today, these countries can claim considerable success in reducing both inflation and inflationary expectations. For example, despite the run-up in energy prices in 2000, consumer price inflation rates from 1999.Q3 through 2000.Q3 fell to 3.5% in the U.S., to 3.2% in the U.K., to 1.6% in the EMU countries, to 2.7% in Canada, to 0.9% in Sweden, and to 3.0% in New Zealand. Japan, with an inflation rate of -1.2%, is something of a special case, as it is just beginning to emerge from a prolonged recession.
With inflation rates now in the low single digits, attention has become more narrowly focused on the problem of determining quantitatively what the “optimal” inflation rate should be. Evidence to date suggests that policymakers’ views have coalesced, however tentatively, around a “2% solution” to this question. For example, consider these explicit inflation targets: 2.5% for the Bank of England, less than 2% for the ECB, 1-3% for the Riksbank in Sweden, 1-3% for the Bank of Canada, 0-3% for the Reserve Bank of New Zealand, and 2-3% for the Reserve Bank of Australia. While the Federal Reserve has no explicit inflation target, the U.S. inflation rate has averaged 3.0% since the end of the last recession in 1991.Q2. This Economic Letter addresses the question: How well is a “2% solution” to the inflation problem supported by the voluminous academic studies that have been conducted recently on the optimal rate of inflation?
Milton Friedman (1969) described the optimal inflation rate as one that would not penalize households for holding monetary assets that bear no interest. This would require a zero nominal interest rate, such that the real return on money, which is the negative of the inflation rate, would exactly equal the real return on real assets. This so-called Friedman Rule has resurfaced in recent theoretical models. These models examine in an internally consistent manner how inflation leads to the inefficient use of the economy’s resources, and they quantify the significance of these inefficiencies by taking the models to the data.
Using the Friedman Rule as a benchmark, the standard measure of welfare losses associated with inflation is computed to be the percent change in the household’s consumption (or income) flow that the household would require to be indifferent between two inflation rates. The majority of the large number of theoretical studies devoted to this issue have produced estimates of the welfare costs of moderate inflation, say, of increasing the inflation rate from 0% to 10% that fall in the range of approximately 1/3% to 1-1/2% reduction in the household’s consumption (or income) flow.
Economists have estimated the magnitude of these welfare costs in several ways. Early studies (for example, Cooley and Hansen 1989) focused on inflation as a tax on consumption expenditures associated with monetary transactions; they found these costs to be toward the lower end of the range reported above. However, inflation induces additional distortions that raise the welfare costs. For example, the avoidance of inflation taxes also can divert resources away from production and into the payment system in order to increase the velocity of money. This response to higher inflation can take many forms, from households making more frequent trips to the bank or ATM machine, to firms devoting greater quantities of capital and labor toward enhancement of the efficiency of their cash management practices, to the emergence of nonmonetary systems (often relying on electronic funds transfers) as a means of effecting final settlement of transactions. As inflation rises, more of the economy’s resources flow into transacting via these channels and less is available for production, thus reducing output and consumption and lowering welfare. In addition, higher expected inflation induces higher nominal interest rates. When firms must borrow against current sales revenues to finance their working capital expenses, i.e., their wage bill and/or their gross investment, or when households borrow against current income to finance the purchase of durable goods, then the higher financing costs can reduce the amount of working capital employed by firms and the volume of consumer durables purchased by households, and again reduce consumption and lower welfare.
From a different perspective, Feldstein (1999) attempted to measure systematically the costs of inflation with respect to its effect on the rate of economic growth. He was particularly interested in the costs of low inflation and took the 2% solution as his benchmark against which he compared an environment of price stability, or zero inflation. He found relatively small direct effects of forgone interest income associated with holding noninterest-bearing monetary assets, but instead focused on the interaction of inflation with other distortionary measures in the tax code, including capital and labor income taxes and the home mortgage deduction. After accounting for the lost government revenue from seigniorage, he obtained hefty estimates of an approximately 1% reduction in the flow of income (GDP) associated with an increase in the inflation rate from 0% to 2%.
Other researchers also have advanced the argument that a long-run average inflation rate of zero would lead to the most efficient allocation of the economy’s resources. Their argument is premised on the view that ours is a dynamic economy in which the types of products that are produced and consumed, as well as the means of production, are constantly changing. For the “invisible hand” of the marketplace to work in this environment, many decisions must be made at the microeconomic level of the firm and the consumer. When the long-run average inflation rate is zero, the aggregate price level is fixed, so any change that is observed in individual goods prices can be attributed with assurance to a change in relative prices. These relative price changes are the signals that firms must rely on to make their production, inventory accumulation, and factor employment decisions, and that households must rely on in selecting their consumption bundles or investment portfolios. Aggregate price inflation confuses this relative price signal and creates uncertainty that can result in suboptimal decisions by firms and households that induce inefficiencies in the transformation of the economy’s resources into valued output goods. The result is a decline in welfare.
The above arguments suggest that the “2% solution” to the inflation problem is not optimal. They support the view that central banks’ long-run inflation goals should be either price stability (zero inflation) or some mild deflation. Yet some economists argue that there is a rationale for adopting policies that support positive inflation in the long run.
When the economy experiences an adverse supply shock, such as an oil shortage, that reduces the productivity of labor, the efficient economic response in the labor market, when households and firms are fully informed and make fully rational decisions, is a decline in both the level of employment and the real wage rate. Some economists have argued that nominal wages are inflexible downward, and that this rigidity can inhibit the real wage from falling when prices are stable. As a result, the response, say, to a negative supply shock would be a larger decline in employment than would be deemed optimal. In this case, a policy of positive inflation would allow real wages to fall, even though the nominal wage did not, thereby enabling employment to absorb less of the shock. Akerlof, et al., (2000) have argued further that in an environment of low inflation, nominal wages also may be inflexible upward in the short run, as households enter into labor contracts that have no provisions for inflation. In this case, real wages erode with inflation over the life of the contract and economy-wide employment is higher under moderate rates of inflation than it would be under a stable price environment.
Critics of these views have countered that these stories are inconsistent with rational behavior on the part of households, since they imply that households are more willing to accept real wage declines through higher inflation than through a lower nominal wage, when rationality would dictate indifference. They also imply that firms lack the ability to adjust labor costs without altering the nominal wage. Nonetheless, to the extent that this short-run rigidity in nominal wages is immune to these criticisms, benefits to mild inflation may exist, although no studies currently reveal whether these benefits are sufficiently large to offset the additional welfare losses from higher inflation as described above.
Finally, the optimal deflation implied by the Friedman Rule would require a zero nominal interest rate, which in practice is a lower bound with which economies have had little experience. Japan is perhaps a notable exception. However, the severity of Japan’s banking crisis with an extraordinary volume of bad loans, and the realization of the need for major structural reforms within the financial sector that have accompanied the Bank of Japan’s “zero interest rate” policy produces a cloud over any conclusions that one might be tempted to draw from that experience. Nonetheless, if the benefits of reducing the swings in the business cycle are perceived to be large, then central banks would need the flexibility to lower the interest rate to avoid recession. For example, if the real (inflation-adjusted) interest rate in the economy is 3%, and if the maximum amount by which the central bank would anticipate ever having to cut interest rates (below their long-run averages) is 2%, then the goal of keeping inflation as low as possible would suggest a 1% deflation on average. If the real interest rate is believed to be lower, or if the monetary authorities need greater flexibility to cut interest rates in times of economic weakness, then the optimal long-run average inflation rate would increase accordingly.
Over the past 20 years, nearly all industrialized economies around the world have concluded that high inflation is detrimental to their economies and to the welfare of their citizens. They have taken concrete steps to lower the inflation rate and to ensure that it remains low. The question that many are grappling with is how low should they go? The evidence suggests that there is a perhaps tentative coalescence of views around the choice of 2%. Research to date would support further reductions in the inflation rate to zero or to a mild deflation. However, further research is needed both on the functioning of economies at near-zero nominal interest rates, as emphasized recently by Lucas (2000), and on the nature and influence of various types of contractual arrangements, including financial as well as labor contracts, around which an important share of economic activity is organized. The results of this research may enhance the arguments for an outcome closer to the “2% solution,” but in any case is likely to suggest that a refinement of these inflation goals is called for.
Senior Economist, FRBSF, and
Professor of Economics,
Florida State University
Akerlof, George A., William Dickens, and George Perry. 2000. “Near Rational Wage and Price Setting and the Long Run Phillips Curve.” Brookings Papers on Economic Activity 1, pp. 1-44.
Cooley, Thomas E., and Gary D. Hansen. 1989. “The Inflation Tax in a Real Business Cycle Model.” American Economic Review 79 (September) pp. 733-748.
Feldstein, Martin. 1999. “Capital Income Taxes and the Benefit of Price Stability.” In The Costs and Benefits of Price Stability, ed. Martin Feldstein, pp. 9-43. Chicago: NBER.
Friedman, Milton. 1969. The Optimal Quantity of Money and Other Essays. Chicago: Aldine.
Lucas, Robert E., Jr. 2000. “Inflation and Welfare.” Econometrica 68 (March) pp. 247-274.
Please send editorial comments and requests for reprint permission to
Attn: Research publications, MS 1140
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120
Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.