Central banks in most advanced economies are committed, often by law, to maintaining price stability. Such banks generally implement this commitment by announcing a target inflation rate, commonly 2 percent per year, and then using monetary policy tools (generally open market operations, that is, buying or selling government securities) to adjust the supply of money in the economy to attain the target rate. Most major central banks, including the European Central Bank, have long maintained a target inflation rate of 2 percent. The United States Federal Reserve long resisted disclosing a target inflation rate, but it finally announced in 2012 that it too would target a 2 percent inflation rate.
Writing at Public Discourse, Samuel Gregg argues that such a policy “raises many questions that have as much to do with justice as with their impact on economic life.” To be clear, Gregg is all in favor of price stability; he does not recommend (who could?) the erratic monetary policies that have long plagued the economies of, for example, many Latin American countries. Rather, Gregg suggests that perhaps “constancy in a given currency’s average purchasing power is the best we can aim for”—that is, price stability with zero inflation. I am happy whenever intelligent and responsible scholars like Gregg turn the conversation to monetary policy, but I have a more sanguine (and conventional) view than Gregg does of positive target inflation rates, and I write today to explain why.
Let’s start with why central banks set low but positive inflation targets. Gregg suggests that central banks adopt such policies because “they provide a small, permanent stimulus for short to medium-term employment without letting the full-blown inflation genie out of the bottle.” I don’t think this is the reason; it is certainly not the stated rationale of the Federal Reserve. Nor, in my opinion, is this reason supported by the macroeconomic theories on which central bankers rely in setting monetary policy. Rather, following the foundational work of Milton Friedman and Anna Schwartz in their monumental A Monetary History of the United States: 1867-1960, the available empirical evidence tends to show that central bank monetary actions can indeed stimulate economic activity but only to the extent that changes in the money supply are unanticipated by economic actors. Inflation consistent with a publicly announced targeted inflation rate will, of course, be anticipated and thus be useless for stimulating economic activity.
Why does it matter whether the inflation is anticipated? The most commonly accepted explanation relies on Friedman’s price misperception model. In this model, increases in the money supply cause all prices to rise (that is, cause inflation), but workers, although they see their own wages rising, do not fully appreciate that all other prices are rising proportionately; they thus believe (mistakenly) that working produces greater rewards in terms of real goods and services than it previously had, and so they work more, thus increasing overall output in the economy. Once workers realize that all prices have risen proportionately and that work is not really more remunerative than it had been in the past, they work less, and labor inputs fall back to the previous level. As Harvard economist Robert Barro puts it,
The effects of an increase in the nominal quantity of money, M, on these real economic variables [such as the real wage rate and the quantity of labor input] are only temporary. In the long run, an increase in M leaves the real variables unchanged.
This idea is known as monetary neutrality. Assuming it’s correct, a creditable announcement by a central bank that it will target a certain inflation rate will have no effect on real economic variables like employment. Hence, when central banks set a positive inflation target, they are not attempting to stimulate economic activity.
So why have positive inflation targets? The primary reason is to prevent the economy from slipping into deflation, a situation in which the overall price level is falling. You may think that falling prices are good, and they are—if the reason prices are falling is that efficiency gains (generally resulting from technological change) are making it cheaper to provide goods and services. But this is not deflation. Deflation is the opposite of inflation, a monetary phenomenon in which the ratio of the money supply to economic output is falling.
In a deflationary episode, prices fall, but so do wages, generally producing sharp reductions in real economic output that are extremely difficult to arrest and reverse. For example, the Great Depression included a major deflationary period from 1930 to 1933 in which deflation was about 10 percent per year and unemployment hit 25 percent. By keeping inflation low but positive, a central bank guards against the possibility of deflation, giving itself ample opportunity to stave off deflation if weakening economic conditions or other economic shocks start generating deflationary pressures.
So a low but predictable inflation rate is sound policy. Is it also just? In my view, yes, for I doubt that it results in the injustices that Gregg and others think it does.
To begin with, Gregg argues that “even small levels of inflation help governments—and other debtors—to reduce the real value of their debts.” That is, if the inflation rate is positive, then when the debtor borrows $100 today and pays it back next year, the $100 returned to the lender is worth less, in terms of its real purchasing power, than the $100 the lender originally lent to the debtor. This is true as far as it goes, but this story assumes that the lender is unaware of inflation and has not acted to protect himself against it.
In reality, rational lenders factor inflation into the interest rates they charge borrowers. Indeed, modern financial theory holds that interest rates are the sum of three components: a rate reflecting the pure time-value of money, the expected inflation rate, and a risk premium. In other words, interest rates include the expected inflation rate, and so ex ante lenders are fully compensated for inflation. Yes, if the actual inflation rate is higher than the expected rate (that is, there is unanticipated inflation), lenders will lose and borrowers will win, but the opposite is also true: if the actual inflation rate is lower than the expected rate (that is, there is an unanticipated lack of inflation), borrowers will lose and lenders will win. On the reasonable assumption that deviations from the expected rate are random, it all nets out in the long run.
Gregg points out that Great Britain inflated away much of its national debt after World War II. This is certainly true, but the British did this by unexpectedly increasing the inflation rate, which is totally different from setting and sticking to an announced inflation rate. If the Bank of England had announced a target inflation rate and adhered to it, none of Great Britain’s bondholders would have been harmed by inflation.
Gregg also argues that people on fixed incomes lose because of inflation. That too is true, but there is almost no one on a fixed income in the relevant sense. For example, retirees receiving Social Security see their benefits increased to reflect inflation—so called COLA or Cost of Living Adjustments. Gregg is fully aware of this, but he argues that people on un-indexed pensions, since they don’t get such increases, are hurt by inflation. This assumes, however, that the persons who contracted to receive such benefits (often unions, who are financially sophisticated and bargain on behalf of individual employees) somehow failed to factor inflation into their calculations in bargaining for a fixed-sum payment, which strikes me as unlikely. If the inflation rate is predictable, anyone agreeing to receive a stream of fixed payments should be able to compute their true value and bargain accordingly. Once again, it is only unanticipated inflation that causes harm.
Even putting aside this point, any harm caused by inflation to individuals receiving un-indexed pensions should be addressed by switching from such defined benefit plans to defined contribution plans, under which employees can invest in a diversified portfolio of securities that naturally protects them against inflation. In fact, over the last few decades, for this reason (and many others) defined benefit plans have almost completely disappeared in favor of defined contribution plans. Determining large questions of monetary policy affecting the whole economy by their effects on a small group of people holding an obsolete type of retirement investment would surely be to let the tail wag the dog.
Gregg also says that “a salaried worker who receives a 3-percent wage increase in a given year will find the value of his raise halved by an inflation rate as low as 1.5 percent in the same year.” But this does not show that the worker was harmed by inflation; it shows, rather, that the worker’s raise was due in part to inflation and in part to other factors, such as productivity increases. If the inflation rate had been zero, the worker’s raise would have been 1.5 percent, not 3 percent, and his purchasing power would have been exactly the same. “Over the medium to long term,” Gregg continues, “this makes it difficult for such income groups to maintain (let alone increase) their salaries’ purchasing power.” But this assumes that inflation affects the prices of all goods and services in the economy except wages, which contradicts not only fundamental conclusions of macroeconomics but also virtually all of the empirical evidence. Inflation affects all prices, especially wages. Hence, inflation does not erode the purchasing power of salaries.
Gregg also says that the ability of a government to manage the money supply through its central bank creates an enormous temptation for governments to manipulate monetary policy to benefit incumbent politicians—for example, by pressuring the central bank to increase the money supply ahead of elections to generate an illusion of an improving economy. About this Gregg is right, which is why central banks are usually organized—albeit with only limited success—to be independent of the elected branches of government. But although monetary policy can be manipulated for political ends, this has nothing to do with targeting a positive inflation rate.
If the target inflation rate is zero, the central bank still has to manage the money supply to ensure that it increases (or, as the case may be, decreases) with economic output. The only difference between a target rate of zero and a target rate of 2 percent is whether the central bank aims to adjust the rate of change of the money supply to be exactly equal to the rate of change in output or to be 2 percent above that rate. The possibilities of manipulation are the same in both cases.
The danger of manipulation arises from having a discretionary monetary policy rather than one that operates according to fixed rules. It’s a serious question which kind of system is better, but it’s a different question from whether, within a discretionary system, we should set the target inflation rate at zero or at some small positive value.
Gregg concludes by saying that the most serious problem related to monetary policy is that most people outside the worlds of finance and politics do not understand the issues involved. Here, I think Gregg is too generous to the politicians—I’m pretty sure that most members of Congress don’t understand the first thing about monetary policy—but his larger point is very well taken. Macroeconomics is difficult, far removed from common experience, and often counterintuitive. I sometimes tell my students that if they don’t understand what the Federal Reserve does and how it does it, they are not entitled to an opinion on the state of the economy. That’s hyperbole, of course, but a republican form of government obviously requires an informed citizenry, and as the world becomes more complex, the epistemic demands on citizens grow. Understanding monetary policy is one of the most pressing, and least fulfilled, of these demands.