The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
FRE:
CLEVETAND. ADDRESSES.
,l-nosxrus. lls.
ç'l
qc,
(:t
.<
t.;
,i
ts¡
\_n
LJ
a
ObJectlves and Strategy for
ilcnetary Pol I cy
l{. Lee Hosklns, Presldent
Federal Reserve Bank of Cleveland
Seattle Buslness Economlsts
Seattle, l{ashlngton
Aprll 15, 1988
ObJectlves and Strategy
for
l.lonetary Pollcy
It ls my pleasure to have thls opportunlty to address the Seattle Buslness
Economlsts. l,ly remarks today rlll focus on the obJectlves and strategy for
ls a complex network of rules, procedures
and lnstltutlons. Ideally, lt prorrctes the efflclent allocatlon of resources
by reduclng transactlon and lnformatlon costs. Our nonetary system provldes a
unlt of account, an efflclent medlum of exchange and a stable store of value.
pollcy.
monetary
How
vel
I
The monetary system
speclflcally on rhat obJectlves
lt
Balanced Growth Act
of
to achleve, and lts
seeks
achlevlng them. The Federal Reserve Act
of
Full
1913, the
1946, and subsequent amendments
glven the Federal Reserve the responslblty
The Baslc ObJectlve
my
the prtce
of
success ln
Employment and
to those Acts
have
for multlple obJectlves lncludlng:
stablllty ln the purchaslng porer of the dollar, stablllty
economy, and a hlgh level of employment.
In
--
the monetary system operates depends on the Federal Reserve
and grorth
of
the
l¡lonetary Pollcv
vlew, the baslc obJectlve of rcnetary poltcy should be to stablllze
level.
The varlables
re as a natlon care about nþst,
output and lncomes, for example, cannot be controlled
goods and servlces
employment,
dlrectly.
The supply
avallable to consumers depends on the quantlty of
productlve resources and how they are used. ìficnetary pollcy can do
affect the total quantlty of land, labor,
avallable.
of
The Federal Reserve can
and phystcal
llttle
to
capltal that ls
control the prtce level and can encourage
lnvestment and real economlc growth by provldlng a stable prlce envlronment.
It
cannot, except through controlllng the prlce level, affect other obJectlves
-2dlrectly. But, lf
other
the Federal Reserve provldes a stable envlronment, the
economlc obJectlves stand
The
a better
of
chance
belng met.
role of mney, and therefore, the role of nnnetary pollcy,
provlde transactlons servlces and lnformatlon.
If
the Federal
ls
to
Reserve
stablllzes the prlce level, then transactlon and lnformatlon costs ln the
economy
rlll
be reduced, and we
rlll
for
have an optlmum cllmate
allocatlon. If the Federal Reserve falls to
an lnflatlon-free envlronment, lt wlll obscure relatlve prlce slgnals,
declslon-maklng and resource
achleve
ralse transactlon costs and add to uncertalnty. By lncreaslng uncertalnty
about future
lnflatlon
and
lts
orn pollcles, the Federaì Reserve adds to the
lnstabtllty of the economy.
The Role
If
of
ìlonev
monetary
pollcy eannot control real varlables rhy do people belleve
lt
ls the ratlonale for stablllzlng the prlce level? Accordlng to
economlc theory, people attaln the hlghest posslble level of welfare ln a
competltlve economy rlth perfect lnformatlon and no transactlon costs. itrney
can and rhat
has no role to play ln such a slmple textbook economy. In the real world,
mney and mnetary pollcy do have a role to play because there are transactlon
costs and people do not have perfect lnformatlon. l,lacroeconomlc models wlth
flxed
the
nomlnal contracts
consequences
of
to represent transactlon costs can be used to
lmperfect lnformatlon and role of money
analyze
ln the economy.
For example, the exlstence of flxed vage contracts can generate a short-run
trade-off betreen lnflatlon and unemployment.
lnto
account rhen enterlng
People take expected
lnto these contracts.
Once rages
then choose output and employment levels to maxlmlze
lnflatlon
are flxed, flrms
proflts. If lnflatlon ls
-3hlgher than orlgtnally expected uhen the contracts rere slgned, the real
rate
wlll fall.
Flrms
rlll
lncrease output and employ¡¡ent
vage
to take advantage
proflt marglns. After the adJustments are complete, output and
employment ulll be no dlfferent than they rould have been rlthout lnflatlon
of
hlgher
because people are
I
not llkely to
make systematlc
errors ln predlctlng
nfl atlon.
Because
lt ls costly and dtfflcult to renegotlate contracts, there can be
a temporary perlod ln rhlch unexpected lnflatlon affects output
employment. ibnetary pollcy should be deslgned
ln the prlce level and thereby
and
and
to prevent unexpected
keep the problems assoclated wlth
prlce contracts to a mlnlmum. In short, lnflatlon reduces
performance, holdlng output, employment and lncome belor
thelr
changes
flxed
vage
economlc
longer-term
sustalnable levels.
Another lrnportant role
for
money and
mnetary pollcy ts to provlde
lnformatlon. For example, people face uncertalnty rhen chooslng rhether to
save
or consume. l{ot knoulng what lnflatlon
rlll
be
ls a slgnlflcant
problem. People do not knor the real lnterest rate, rhlch represents the
return to savlngs ln terms of future consumptlon. The real lnterest rate ls
slmply the nomlnal lnterest rate mlnus the expected rate
people could predlct
lnflatlon accurately the problem rould dlsappear.
because people are uncertaln about
But
future lnflatlon and the real lnterest
rate, they are unable to plan optlmally for current
The monetary
of lnflatlon. If
and
future
consumptlon.
authorltles can reduce the problem by maklng the prlce level
predl ctabl e.
Buslnesses also face
depend on
the cost
thls sort of uncertalnty.
of capltal
and on the expected
Investment declslons
return.
Expected returns
-4current lnterest rates reflect future
depend lmportantly on hor accurately
lnflatlon. tlot havlng thls lnfornatlon ls costly.
He knor
flrms paylng for lnsurance ln flnanclal markets.
see
ln flnanclal
markets
l,{any
lt ls because we
of the developments
ln the last 20 years represent an attempt by the prlvate
to protect ltself from the uncertalnty about lnflatlon. Even lf
prlce level cannot be made plrfectly certaln, the costs assoclated wlth
econcmy
lnflatlon uncertalnty
lf
can be reduced
the Federal Reserve focuses
level. These examples lllustrate why I
pollcy ls to produce a stable prlce level.
sharply on a stable prlce
that the best monetary
more
conclude
l{hlle the Federal Reserve ls glven the responslblllty to malntaln
stable prlces and
full
employment,
to stablllze the prlce level.
and therefore
zero
both
belleve the way to achleve both goals ls
prlce stablllty,
I
mean
that people expect
act as though prlces rlll be stable. By prlce stablllty,
lnflatlon. I
constant.
By
I
Each
don't
mean
that
the
all
the dlfferent prlce lndexes
ulll
I
be
prlce lndex has lts orn pecullar characterlstlcs. There
alrays be non-n¡onetary factors and
measurement
varlatlon ln the lndexes should be Just that
mean
rlll
problems. But the short-term
--
short-term varlatlons around a
zero trend.
The Strateqy
Thls brlngs
me
to
rny second
polnt, the strategy.
Havlng chosen
a stable
prlce level as the approprlate obJectlve for monetary pollcy, rhat ls
senslble strategy for achlevlng
years.
ît?
a
Indeed we have made much progress ln
ln that progress for us to cons'lder.
ïhe flrst, and most lmportant, part of a successful strategy must be to
enllst the support of market expectatlons. Thts ls done by announclng clear,
expllclt goals and actlng ln a credlble manner to achleve them. lr{hen
recent
There are some lessons
-5lnflatlon
uas
at double dlglt rates at the end of the 1970s, people dld not
belleve that lnflatlon vould stop
rlslng.
The
often pronlsed end to lnflatlon
ras not dellvered. In that envlronment, statlng pollcy goals slmply
was not
credl bl e.
Today, ye have gone through
a S-year perlod rlth lnflatlon fluctuatlng ln
a 2 to 4 percent range. l'larkets seen to belleve that the
lnf I atlon rl se above 4 percent
nrarkets expect
wl
lnflatlon to rlse
Fed
rlll
thout takl ng correctlve actlons.
above
that range, markets
seem
not let
hlhen
to expect
the
Federal Reserve to adopt a tlghter nonetary pollcy and lnterest rates tend to
rlse ln antlclpatlon. Because the Federal Reserve has been credlble ln
flghtlng lnflatlon ln the 1980s, lt should be posslble to contaln lnflatlon by
uslng very small pollcy changes as long as these changes are lnltlated ln
advance
I
of rlslng lnflatlon.
thlnk re can lmprove our economlc performance by announclng a goal of
zero lnflatlon to be achleved over
some
relatlvely short tlme perlod
--
3 to
5
years. If, as I belleve,4 percent ls the rate of lnflatlon today, then I
suggest the acceptable upper ltmlt should be 3 percent a year from now. lllth
zero as the ìower llmlt, the upper and lower llmlts would converge by 1992.
If clearly announced, I thlnk thls ls an acceptable goal.
Thls ls the flrst part of an acceptable strategy, announclng a clear
expllclt goal. The second part ls demonstratlng a determlnatlon to achleve
It. The goal of prlce stablllty must be the focal polnt of our pollcy
dlscusslon. The Federal Reserve can enhance lts credlblllty and reduce the
cost of achievlng thls goal by explalnlng lts pollcy declslons wlthtn the
context of the plan to achleve prlce stablllty over the stated perlod of tlme.
-6lle must be able to recognlze when pollcy should be changed and re
change
lt
accordlngly.
The Federal Reserve nay need support and asslstance ln
followlng pollcles to achleve a stable prlce
lt
lndependent, but
soclal
and
must
has a strange sort
of
level.
The Federal Reserve ls
lndependence. l{e
exlst rlthln
a polltlcal compact, and that compact must clarlfy our role
encourage us
to
do our
Job.
The Fed should be held accountable
a
and
for provldlng
prlce stabl I lty.
ls partlcularly lmportant because of the
breakdown ln the rellablllty of the monetary aggregates. The lack of a
Havlng a clear obJectlve
rellable llnkage
between the monetary aggregates and the
prlce level adds to
the dlfflculty of knowlng what ts the approprlate monetary pollcy to brlng
lnflatlon down. It also adds to the dlfflculty of others outslde of
the
Federal Reserve ln Judglng whether the Federal Reserve's actlons are lndeed
conslstent wlth the deslred outcome. Hlthout a well understood pollcy gulde
and a
rellable pollcy lnstrument, whlch connects pollcy to obJectlve, the only
proof
of the puddlng wlll
and varlable
lag,
Federal Reserve
be able
to form
lt
ls
be
becomes
ln the eatlng.
Because
pollcy works
rlth a long
rþre dlfflcult than ever to knor rhether the
uslng a good reclpe or a bad
Judgments about how
rell
one. Externally, markets must
ye are dolng.
Adopt Conslstent. Svstematlc Operatlno Procedures
The Fed should adopt short-run guldes
approprlate. Just as
or operatlng targets that
are
lt ls lmportant to choose obJectlves that can be
lt ls lmportant that these operatlng rules or targets be connected
meanlngfully and rellably to the ultlmate obJective. It ls also lmportant to
achleved,
choose short-run operatlng targets
efflclency of the nonetary system.
that
can be controlled wlthout reduclng the
-7 The framework
rhlch re adopt nust allow for uncertalnty and mlstakes ln
human Judgment. Any successful procedure should
take accot¡nt of that
uncertalnty and, rhenever posslble, mlnlmlze the costs assoclated wlth
nlstakes ln Judgment. Xlthtn that context
others because
some
mlstakes are less
lnflatlon ls very costly to deal rrth
some
costly.
rlsks are mre acceptable
One lesson
than
of the 1970s ls that
once underray and eúedded ln
expectatlons, contracts and resource al locailon declslons.
l{e are
thumb
ln a quandary on procedures today largely
or poltcy
slgnlflcantly
guldes
rhlch
connected
because the
rules of
to rork ln the past are not rellably or
today to the pollcy outcome we wlsh to seek. over the
seemed
long run, the Federal Reserve can dlrecily control only the quanilty of
hlgh-powered rcney, the monetary
base. Nevertheless, ln the short run
operate by flxlng elther the prlce or the quantlty
of the nnnetary
base
re control dlrecily.
It
can
of
bank reserves,
follor
lt
can
the part
an operailng
procedure vhlch flxes the lnterest rate on reserves, the federal funds rate.
At the other extreme lt
can
follor
some
can
flx the quantlty of the total reserves. Or, lt
lnter¡mdlate pollcy of establlshlng a relatlonshlp betreen the
federal funds rate
total reserves, allowlng the federal
and fall ylth changes ln reserves.
and
Before 0ctober 1979, the Federal Reserve operated
rlth
funds rate
to rlse
expllclt target
for the federal funds rate. The open market desk kept the rate ln the
targeted range by enterlng the market to buy or sell, sometlmes both ln the
same
an
day, whenever the rate threatened to move outslde the speclfled range.
At each Federal Open I'larket
Conmlttee (FOtlC) meetlng, the comnlttee
members
-8evaluated the state
of the economy and used thelr best Judgment to declde
whether the lnterest rate target ras conslstent
pollcy.
Durlng much of
rlth
thls perlod, the F0-lC had many obJectlves,
whlch may have been unclear and unstated. Perhaps
gotten out of control
obJectl ve
of
lf
of
lnflatlon would not have
prl ce stabl I I ty.
operatlng procedure.
have
If
be achleved uslng an
lnterest rate
lnflatlon expectatlons have rlsen, the federal funds
to be lncreased
enough
to reverse the lncrease ln lnflatlon
expectatlons. The problem wlth thls approach ln practlce
have
some
there had been a clear, overrldlng, publlcally stated
In prlnclple, prlce stablllty could
rate would
the obJectlves of
a rellable connectlon betveen the funds rate
and the
ls that we do not
prlce level.
A Total Reserve Taroet
The Federal Reserve mlght operate
level of total
at the other extreme,
lts
Judgment
to
choose
demand
the
a path for total reserves that
was
narket lnterest rates would automatlcally rlse
fall lf
flx
the
thought to be conslstent wlth the deslred prlce
above the target path and
could
metlng
bank reserves between Fü.IC meetlngs.
commlttee rould use
lt
for
At each
level.
lf
the
FO,IC
Between meetlngs rþney
demand
for
reserves grer
reserves grer below the target
path. Shlfts ln the reserve/npney multlpller, as rell as shlfts ln the demand
for nomlnal rcney balances, would lnduce changes ln the federal funds rate.
The change
ln lnterest rates rould ¡nove total reserves,
back toward the
deslred path, slmllar to vhat was done to durlng the lg7g-82 eplsode.
As
lt
does today, the Fed would have
to monltor the factors that affect
the relatlonshlp betveen total reserves and the prlce level. The FOl.lC would
-9have
to
use Judgment
uas assoclated
to declde whether an lncrease ln the demand for
rlth real factors or tlth
an lncrease
reserves
ln lnflatlon
expectatlons. At each F0[ meetlng, the cormlttee rould estlmate changes ln
the relatlonshlp betreen total reseryes and the prlce level obJectlve and then
make
the approprlate changes ln the total reserve path. The problem wlth thls
approach
ts that
we do
monetary base and the
not have a rellable short-run connectlon betreen the
prlce level.
A l¡llddle Course
There are other problems assoclated
wlth relylng solely on elther the
total reserves. Uslng an lnterest rate target, as ras done ln
the 1970s, made lt very dlfflcult for the Federal Reserve to make changes ln
funds rate or
pollcy
adequate
to control the acceleratlon of lnflatlon. Inflatlon
expectatlons rose faster than the
FOl,lC
ralsed the lnterest rate
though lnterest rates rose, monetary pollcy
lnflatlon to
become embodled
target. Even
ln effect ras eased, allowlng
ln expectatlons
and
ln contracts,
addlng further
to the dlfflculty of preventlng a further lncrease ln lnflatlon.
l{hlle the Federal
Reserve has never used the other extreme
the level of total reserves
between 1979 and
short-term
between meetings,
lt
pollcy, flxlng
dld approxlmate such a pollcy
1982. Thls perlod ras characterlzed by a great deal of
volatlllty ln the bond and rcney markets.
Thls
volatlllty
attrlbuted to short-run varlatlons ln nnney demand, to operatlng
rhlch dld not provlde
sometlmes,
adequate cushlonlng agalnst those
to confuslon about what the Federal
Reserve
technlques
varlatlons
and
ras trylng to
accompllsh. As a practlcal matter the very short-run varlatlons ln
demand should be accommodated by an
a
need
was
money
elastlc supply of reserves. Thls suggests
to reexamlne the operatlng procedures ln order to flnd a mlddle
course
_t0_
-- an operatlng scheme that allows the federal funds rate to rlse and fall
rlth the level of total reserves, but also accorrnodate the short-run
transltory shlfts ln the
demand
for
reserves wlthout
allorlng the accumulatlon
of reserves needed to support rlstng lnflatlon expectatlons. l,lhat we do not
knor yet ls how to accompllsh both of these obJectlves: comblnlng some
measure of short-term stablllty rlth an lnflatlon free long-term trend.
Developlng such a mlddle course
ls an lmportant task.
Ooncluslon
Twenty years ago,
Assoclatlon,
done,
lt
l.ll
lton
ln hts presldentlal address to the Amerlcan Economlc
Frledman
sald, "If
takes lnterest rates or the current unemployment percentage as the
lnedlate crlterlon of pollcy,
a
flx
on the ïrong
star.
lt rlll
l,lonetary pollcy cannot hope
lt
but
be
can provlde an
ylll
occur, but
for
vlll
problems
of a market
efflclent nonetary systen rlth a stable prlce
resource allocatlon
be more
lts
go astray."
to solve all the complex
level. In such an economy, markets wlll
lnformatlon
llke a space vehlcle that has taken
llo matter hor senslttve and sophlsttcated
guldlng apparatus, the space vehicle
economy,
, as the (monetary) authorlty has often
be better able
declslons.
to provlde the best
Buslness cycles
attrlbutable to real events and not to
wlll stlll
unexpected
ln the lnflatlon rate. It seems to me that our terrlble lnflatlon
experlence of the 1970s amply dermnstrates the ulsdom of that concluslon.
Equally, lt seems to me that the substantlal costs lncurred ln reduclng
lnflatlon and lnflatlonary expectatlons ln the 1980s ls a very porerful
argument for avoldlng an acceleratlon ln lnfìatlon ln the future. Flnally, I
bellve that there are lmportant beneftts for the performance of our economy ln
changes
reduclng lnflatlon further
¡
-- to zero.
@FedFRASER