Monetary policy rules that focus on smoothing out fluctuations in nominal income or
nominal output have been advocated by many economists and policy analysts. But new
research by Glenn Rudebusch, published in the latest issue of the Economic Journal,
decisively rejects such nominal income targeting. His analysis shows that policy rules
that respond separately to inflation and real output - such as the Taylor rule - generally
do much better than a nominal income rule in stabilising the economy.
Intuitively, Rudebusch notes, nominal income targeting appears to be a desirable
strategy for monetary policy because it automatically takes account of movements in
both prices and real output, which in practice are the two macroeconomic variables that
central banks seem to care about most.
Furthermore, as many have noted, both monetary targeting and nominal output
targeting should produce similar outcomes if there are no large shifts in the velocity of
money. Thus, for example, the European Central Bank's (ECB) announced strategy of
a reference value for money growth provides support for consideration of nominal
output targeting. (Indeed, the ECB has explicitly derived its 4.5% reference value for M3
growth from a desired growth rate for nominal output.) Conversely, the ECB's strategy
also obtains some further legitimacy from any research results that reflect favourably on
nominal output targeting.
But earlier research shows that policy rules that respond separately to inflation and real
output, such as the Taylor rule, generally do much better than a nominal income rule in
stabilising the economy. This is because a policy rule that reacts to nominal output
growth responds in an identical fashion to fluctuations in real output growth and inflation
and so does not perform well when the timing of real output and inflation responses to
monetary policy actions are quite different - as appears to be the case in the real world.
In response, some have argued that policy rules that respond to the growth rate of
nominal income perform better than other policy rules after taking account of the limited
information about the economy that policy-makers actually possess in real time. In
particular, nominal income rules may yield reasonably good outcomes across the wide
range of plausible macroeconomic models (even though these rules do not provide the
best performance in any particular model). In addition, nominal income rules do not rely
on measures of potential output (and the output gap) or the natural rate of
unemployment, which are subject to great uncertainty and are crucial for, say, the
Taylor rule.
Rudebusch’s research decisively rejects these arguments for nominal income growth
rules. The nominal output rules that are often advocated perform poorly relative to a
Taylor rule even after taking account of the range of model uncertainty and data
uncertainty that policy-makers appear to face.
A further implication of this result is that real output growth adds little to the formulation
of optimal monetary policy beyond the information contained in the level of real output
and the inflation rate. Thus, the very fast output growth of an economy emerging from
recession may have no implications by itself for the appropriate stance of monetary
policy, given that policy-makers have some information, however imperfect, about the
size of the output gap.
ENDS
Notes for Editors: ‘Assessing Nominal Income Rules for Monetary Policy with Model
and Data Uncertainty’ by Glenn Rudebusch is published in the April 2002 issue of the
Economic Journal.
Rudebusch is Senior Monetary Policy Advisor, Economic Research, MS 1130, Federal
Reserve Bank of San Francisco, 101 Market Street San Francisco, CA 94105.
For Further Information: contact Glenn Rudebusch on +1-415-974-3173 (fax: +1-415-
974-2168; email: Glenn.Rudebusch@sf.frb.org; website:
http://www.frbsf.org/economists/grudebusch); or RES Media Consultant Romesh
Vaitilingam on 0117-983-9770 or 07768-661095 (email: romesh@compuserve.com).