2. From Money Targeting to Inflation Targeting
Until the 1990s, central bankers typically thought about monetary policy in terms of the rate of money growth. A
central bank chose a rate of money growth that corresponded to its inflation target, and worried about how closely
the growth of monetary aggregates matched its goals. But it turns out that money demand does not have a stable
relationship with any of the monetary aggregates. As a result, inflation does not have a tight relationship with the
growth of any particular monetary aggregate. Faced with this reality, most central banks now think about
monetary policy in terms of the nominal interest rate. A central bank tries to achieve its target inflation rate by
changing the interest rate.
In the short run, an inflation target might seem to reduce the central bank’s flexibility to stabilize output. In
theory, this is not the case. If the central bank sees a recession arising from an adverse demand curve, it knows
that inflation will fall. Thus, it should pursue a monetary expansion to increase inflation. This policy of course
will also tend to return output to its natural level. Indeed, if the central bank manages to achieve a constant rate of
inflation, the accelerationist Phillips curve implies that unemployment is always at its natural rate. In practice,
central banks will not always be able to achieve their inflation targets in the short run. If inflation turns out to be
higher than the target, it is not clear that the central bank should try to reduce it, since to do so would require
higher unemployment. Moreover, the Phillips curve does not hold exactly. Sometimes inflation will increase
even when unemployment is at the natural rate. A strict inflation target would require the central bank to
undertake a monetary contraction and increase unemployment above the natural rate.
Given an inflation target for the medium run, a central bank must decide how to achieve it and under what
conditions to relax the target to respond to short-run fluctuations. John Taylor argues that a central bank should
think in terms of a target interest rate, since the interest rate affects spending, and suggests an implementation
rule. Under the Taylor rule, central banks set short-term interest rates according to
i = i* + a(π-π*) – b(u–un),
where π* is the target rate of inflation, i* is the interest rate associated with this target, and un is the natural rate of
unemployment. The coefficients a and b reflect how much the central bank cares about inflation versus
unemployment. Note that the coefficient a should be greater than one, because the real interest rate affects
spending. Thus, when inflation increases, the central bank must increase the nominal rate by a greater amount to
increase the real interest rate and reduce spending.
Basically, the Taylor rule says that the central bank should respond to short-run fluctuations to the extent it cares
about them. Taylor acknowledges that events could justify changing the nominal interest rate for reasons not
included in the rule. But he argues that the rule provides a useful way of thinking about monetary policy.
Evidence suggests that the Taylor rule describes quite well the actual behavior of the Fed and the Bundesbank
over the past 20 years, however they conceived of their policy decisions. Today it seems that most central banks
think in terms of interest rate rules rather than in terms of nominal money growth. Indeed, the growth rate of any
particular monetary aggregate seems to be of little concern to central banks or financial markets.
3. The Optimal Inflation Rate
The optimal medium-run inflation rate depends on the costs and benefits of inflation. The costs of inflation vary
with the level of inflation. Very high rates of inflation disrupt economic activity; low rates of inflation are much
less costly. Inflation creates four costs, all of which increase with the rate of inflation.
First, higher inflation leads to higher nominal interest rates, which cause people to economize on holdings of
money. This effort requires resources, referred to as shoe leather costs. Such costs can become quite large during
hyperinflation, but are unlikely to be very important when inflation is low.
©2017 Pearson Education, Inc. Publishing as Prentice Hall
23-107