View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

fcccfivw,c AsSh

iJ-.hj'1'•/

.

THE SEARCH FO R A TARGET FO R MONETARY POLICY

i

JORDAN:

I t is a pleasure to welcome so many fellow search party

members here. I'm going to throw the program a little bit of a curve by
saying th a t I'm not sure th at the question should be searching for a monetary
aggregate, or monetary target, but rather, searching for an objective for
monetary policy. A l m o s t d e c a d e s ago, K arl B runner hosted at UCLA a
couple of conferences called "Targets and Indicators for M onetary Policy.”
This was when Lee Hoskins and I were young students at UCLA. They
brought in all of the eminent people in the profession, some 50 or 60 of the
leading names in monetary theory and policy, both within the Federal
Reserve and the academic profession. It was a fascinating couple of days,
both times.
W hat I failed to appreciate at th at time was th at those conferences
were conducted in the context of more than a decade of under 2% inflation,
followed by a couple of years (1964 and 1965 ) when inflation jum ped into
the range of 3 or 4%. It was thought at that time th at the problem was only
one of choosing the appropriate targets and indicators - th at is, the levers, or
handles for monetary policy which would serve as guides to formulation and
implementation of our policy objectives. The mindset of the American
people at th a t time, I think, was that increases in inflation and interest rates
were tem porary, destined to go back down.
Now, after more than three decades of inflation, the mindset of the
American people is th at declines of inflation and interest rates are temporary.

2

*

#

People believe th at the permanent condition is for inflation and interest rates

i

to go back up.

Coming back to the Fed after some 15 years out of the System, I
initially thought the question was "which M " (monetary aggregate)? The
subject of this session might imply more debate about M l versus M2 and the
various measures of the monetary base. In fact, three months before
rejoining the Fed I did an analysis for the Fed of w hat was going on with M2
in 1991. Already there was concern about the extent to which it was or was
not giving a reliable indication of the thrust of policy actions. Now,
however, I've spent 15 months or so not only going to meetings, but worse,
living with the staff th at I inherited from Lee Hoskins. The staff persistently
said to me " It’s the objective, stupid - not the target, that is the real issue!"
And, I kept saying th at I thought the objective was quite clear - not only to
myself, but also to my colleagues. The staff kept telling me I was naive.
But after sitting with the Committee for 10 FOM C meetings,
watching the deliberations, the interaction between our decisions and the
financial m arket participants, the people's elected representatives, and the
media, I ’m now convinced too, it's the objective, not the target, th at is the
real issue of m onetary policy.
Some years back, I heard about "G oodhart’s Law." I think it was
Henry Wallich who first told me about it and wrote about G oodhart's Law.
The idea is th at once a Central Bank makes it known that it is using a certain
variable as an indicator or intermediate target because of some past
empirical relationship to a specific objective, th at variable ceases to be

3

reliably related to the objective.
The analogy was something from physics called the Hiesenberg
Principle, which states th at when you focus a high-powered microscope on
an electron, it alters the behavior of the electron. Therefore, you can never
see it behaving as it would behave if it was not being observed. The analogy
to m onetary policy is, once the Central Bank has a target which the people
know it is responding to, the behavior of the people is changed ~ traders in
the m arkets, such as bond markets, equity markets, foreign exchange m arket,
as well as real people. Then, because they change their behavior in
anticipation of w hat the Central Bank is going to do based on these
indicators, you d on't get the same outcome that you otherwise would.
So, we went through the silly season in the late 1970s when the
Thursday night money numbers would cause the interest rate futures
m arkets to do wild gyrations. There was a period some years back when the
m onthly merchandise trade balance was all the markets responded to. And,
unfortunately, during the last year or two it was nonfarm payroll
employment or the jobless claims numbers. The lack of theory or empirical
evidence relating these variables to ultimate objective seemed to be

K

irrelevant.
I think th at G oodhart's Law is/galitjjonly if monetary policy is not

^

anchored by clear, well-understood objectives. jG oodhartV iaw ^w ill-B ot-hol^-- ^
J^fev ery o n e understands and acts on the belief th at the objective will be
achieved^ Thrcf^lm tnfTr people will not care w hat the interm ediate targets
are. So, I'm going to assert what I will call the "Hoskins’ Corollary" to

4

^
^
^

G oodhart’s Law (after he kept hammering on this thing for four years while
he was in my job). That is, if the m onetary authorities have an objective of a
stable purchasing power of money which is known and believed , such th at
the actions of households and businesses reflect this knowledge and belief,
then it may be th at any target is adequate. It certainly becomes a secondary
issue at th at point.
I no longer think th at monetary targeting is a sufficient condition for
a satisfactory monetary policy. W hether or not it is a necessary condition is
still debatable. I do think th at having a clear objective of m onetary policy is
necessary. I'm still waiting for people to persuade me w hether it is or is not
sufficient.
Milton Friedman taught us th at having and hitting any monetary
growth target was superior to a pure discretionary policy. I think w hat was
left unsaid about this Friedman dictum was th at it was valid in the absence
of policy being anchored by a clear objective. I'm not so certain th at a
policy th at is anchored is flawed, just because implementation involves
discretion. All of the debate about rules vs. discretion that we saw in the
profession for a couple of decades, I think now implied an absence of an
unambiguous long-run objective.
Back in the 1960s —20 or 30 years ago when all of that lively debate
was going on -- we had the Friedman-Meiselman research relating money
growth to measures of income or output, and the response by AndoModigliani. Then we had Deprano, M ayor and Hester and a num ber of
others including economists at the St. Louis Fed, the New York Fed, and the

5

Board of G overnoriV I now think th at whole framework was wrong. That
fram ework related money growth to total spending-- nominal GDP ~ and
then we derived from th at implications for the rate of change of prices, and
the rate of change of output and employment, and so on. We in rt nf . <7
decomposed spending into its output and price components. I think there
was a fundam ental mistake in our research strategy at th at time, and all of
the rhetoric of "dem and management" was wrong because both the
research strategy and the rhetoric assumes knowledge that we did not have -­
knowledge about the economic structure and factors influencing the
aggregate supply of output.
A lot of w hat we did at St. Louis, including my own writing at th at
time, was in the context of demand management. Think about the message
th at is implied by that. It says that the role of monetary policy is to manage
demand or spending. That clearly is wrong.

It left a lot of people with the

idea th at you could pursue an activist discretionary monetary policy to hit
certain objectives in terms of output growth, the rate of inflation, levels of
employment, the unemployment rate, and so on. And, that it was
appropriate to use monetary policy to pursue "countercyclical stabilization
policy" to either offset real shock affects, or shocks emanating from the rest
of the government sector. And, we hear right up to present times references
to the idea of a monetary policy being adjusted based on what happens on
the fiscal side.
In one im portant respect, monetary and fiscal policy actions should be
thought of in concert. That is, that in a fiat money world, monetary policy is

6

a fiscal instrum ent —a way to finance government. No one can possibly

I

know the effects of proposed changes in explicit tax rates w ithout knowing
w hat is going to happen to the purchasing power of money. The C hairm an
of the House Ways and Means Committee should be ju st as interested in the
C entral Bank's intentions regarding future inflation as the Chairm en of the
Senate and House Banking Committees.
W hile I now think that most of the discussion about activist monetary
policies is nonsense, it did derive from the work th at I and others were
involved in back in the 1960s and '70s. We should not have allowed the
emphasis of our work at the time to sound to people as though monetary
targeting was about creating alternative levers for monetary authorities to
push and pull in order to achieve some sort of unspecified and frequently
changing objectives.
W hat we failed to make clear was a crucial underlying premise of
monetarism — the Hayekian principle th at a m arket system, based on
private property and using prices to allocate resources, is inherently resilient
and naturally gravitates toward full utilization of its productive resources in
the absence of various types of shocks emanating from the government
sector. That means th at the4wiPuitimate objective^of m onetary policy
would be to achieve the highest sustainable growth over time th at is
consistent, not only with the endowment of resources, but also with the ^
| ^C_
various fiscal policies and regulatory policies of g o v ern m en ^ n o t trying^to
compensate or offset th e ^ fic t^ o fth o se policies^

Oj a
^

wjflatinffwf all the rhetoric about hawks and doves,/Ifc«=cr*)
^

' ^ rHjy '-

id fir

^ m|Taaon^n^ti"trallO dimii^iiuuaiip.nrln i »nMli^rn[VTP-ln!liuUwwMiCff*>mr
ethtrrblH lig'tt d tflt. Lee Hoskins was often labeled a hawk, and the rhetoric
in the press was th at he was anti-growth because he was anti-inflation. T hat
was a conceptual mistake and one th at we need to clarify. It is a false
dichotomy; there is no choice between inflation and growth over time.
-atoo-H plics a rejection of.alM | the jargon-of "tightening'*- a n d ^ e a s in g ^ o f
M
-jmonetarv Dolicv^but-nrovH)f all, it.iin p liira''f eh.QhiiiJuf the Phillips curve /
- the notion u f some suit uf siuclal/Mlillial tradeoff between the rate of
change of prices and the level of employment, m uiiemployiuculi ■
.— y

The persistence of the idea of a social/political tradeoff is what's

behind tiifrvgcent Congressional efforts to alter the FOM C, toprtfc the
Presidents of the R&«cye Banks off the Committee, jtp'fnake them subject
to presidential appointm enfv0t^onfirm ation> y'the Senate. It is a line of
argum ent th at says, " because th isjs^ p o litical decision, it must be made by
people who are poIiticallv^eCountable." B ut> 4reject the premise, so the
conclusion does nj*Hollow. The basic inherent resiliency proposition implies
a role forjnt>netary policy in the economy such that there is mKpolitical
Iinfluence in the formulation and implementation of monetary
W e have other problems in our language with the terminology of
m onetary policy. Lee popularized this idea of "zero inflation."

Hom er

Jones at the St. Louis Fed once said, " If people see something often enough,
they come to believe it whether they understand it or not." I think th at was
a p art of Lee's operating strategy. Just keep saying "zero inflation" over and
over again until it became respectable to talk about price stability and zero

inflation.
I believe in "stable prices" in a certain m anner of speaking, but I
realized th at we had a problem with our language when I started meeting
people from the form er Soviet Republics and Eastern Bloc countries. These
are people th at had been listening to the messages from us as their regimes
were starting to break up, and we were telling them two things: "am ong the
various things you need to do, such as establish property rights and so on, is
th at you need to end all forms of price controls. Let prices be free to move,
create flexible prices, and achieve price stability at the same tim e!" And,
they'd say, "well wait a minute, which is it you want? Do you w ant prices to
be stable, or do you want them to be flexible?" W e'd say, "Well, both."
And, we had trouble explaining to them what price stability means. It
doesn't mean constant prices fixed by government, but rather some other
notion. I struggled with this idea of trying to explain to them w hat it was
th at we wanted to be stable.
I think th at we sometimes miscommunicate with our own people
when we talk in these terms. W hat we really want to be stable is in the minds
of the people - th e national standard of value, or unit of account. So
following B runner and Meltzer -- or Armen Alchian's piece on "W hy
M oney" -- "Society chooses to use as money, th at entity that economizes
best on the use of other real resources in gathering information about relative
prices and conducting transactions." The implication is that money belongs
in the production function because a less than optimally efficient money
means th at you are under employing real resources. Uncertainty about the

9

trend in the price level lowers the production possibility boundaries So,
ies.
given the state of knowledge and technologies, w hat we are trying to do is
create the condition th at achieves the highest production possibility
boundary obtainable over time, consistent with the other things th at th
- government is doing, such as reallocation of command over resources.. Any

K

/ A ^ r« * * e < & /Incertainty about relative prices - both current output prices and asset ,
y
p r ic e s ^ uid-thg n n m itM iw t of future consumntion — moves us to a lower
production possibility boundary. That condition of achieving the highest
....
production possibility boundary is what we really nrean
price stability. '

y & M

While fiscal and regulatory policies might be used by government, for
redistribution, ! d on't think it would be appropriate for m onetary policy to
either intend to cause redistribution, or to have an unintended consequence - neither interpersonal nor intertemporal -- of redistribution.
None of th at should be taken to mean that I think th at money
targeting is not desirable, or that it is not useful. W hen we are starting from
a position of past inflation, having conditioned people to act in the belief th at
the purchasing power of money will be eroded over time, then I think it is a
question of how do we proceed to restore credibility in the commitment to
w hat we call price stability, or sound money.
W hen Lee Hoskins took this job in 1987, he started talking in public
about M2 growth. He would explain to me that the Board staff had
Vn
^
developed model b ased on ail idea'of opportunity cus^uHiuldiiig M2------- -

M

—
balances th at allowed them to jiggle short-term interest rates, giving them
pretty good predictions of what M2 growth was going to be over subsequent

10

I

months. Furtherm ore, based on longer-term historical correlations, the so/
called "p -star model," they argued th at they had sjgne pretty good idea of

It**

the^implications. So, the idea was: use open m arket operations to influence
u

<

short-term interest rates, which influences M2 growth, which achieves
objectives for the price level, or the rate of inflation. The model seemed to
work pretty well, roughly from the time Lee got there until 1989 or 1990.
My dissents-jast year at Suuftttfi^meetings llmli I lilt* illlllll nl I ST
Ti

\

were ^enditnmed- by th at experience. The idea was that in order to achieve
credibility in the objective of monetary policy, we had to hit the targets that
we actually announced. And, if we wanted people to believe in the upper
limits of the target range, we had to hit the lower boundary of the target
range, (it-sim ply-wasn 't cr cdiblc-to-mfr-to gi ve-a

o r be mg

be|ew-thenntirimttin4tmitr'bttt*ROt>doti>g-airything40‘try>to^etit4>aek^pr
while-saying to4he^eppler>
'b~frrdqn>
t~worry7you-cairtrusfr'the-uppep»limitsy

Some weeks back, I was talking to a member of the Board staff about
w hat we call the Blue Book -- which is the input to policy and the tradeoffs of
various money growth and interest rates —and about the experience of the
last couple of years of their opportunity cost model of M2 growth; how lousy
the projections have been, and the velocity of M2 problem . The Board
staffer said, "Well, with no causality intended, it worked pretty well before
you got here."
It's pretty clear to most people that the linkages are broken. That the
link between open market operations and money growth, at least in terms of

11

the broader money measures, is not as reliable as once thought, and also the
/
linkage between some measure of money and the rate of inflation, or the
price level is not working.

^

Back in 1980, we had some legislation that was in p art intended to
enthrone M l. W e had had a couple of decades where the M l velocity (really
the variance around the trend of the M l velocity) was quite reliable, leading
us to believe th at if we conducted monetary policy (open m arket operations)
in such a way so as to hit an M l target, we had a pretty good idea of w hat
was going to happen to nominal GDP growth. So, the legislation was to
move toward w hat we called uniform and universal reserve requirements —
to have reserve requirements only on transactions liabilities, and have the
same reserve ratio on all transactions liabilities at all institutions, so as
balances bounced around from a credit union, to a savings and loan, to a
bank -- from a big bank to a little bank, and so on —we did not release and
absorb reserves in an erratic way, causing noise in the multiplier. The
legislation was intended to improve monetary control because we thought all
we needed to do was tighten up on our control of M l and everything would
be fine. Along about th at time, or maybe shortly after —
some people say it
was the 1982 G arn-St. Germain legislation -- M l got dethroned. While we
were trying to w ork on improving the monetary control side, the connection
to something we cared about seemed to go haywire as the velocity of M l
started to become less predictable.
The Committee then shifted to an emphasis on M2, based on the
historic relation between the broad measure of money and the rate of

12

*

inflation. But, we had the wrong institutional arrangem ents in order to h i t M 2 .
_____
I
TV*.
A u d k j cr*M, / " .
Ui
mfttwyvFiftwIg33n=gftt whon-wa defined-it-as a brood-mcnstire. That-fo-what «
gave rise to this opportunity cost model. If we still believed th a t we had a
reliable linkage between M2 growth and w hat we really care about, then-tb*^
i)^jyct^cs-s h ^ dP l^ to seek institutional arrangem ents — in the form of
s«^MgBg9BM-a change in reserve ratios ~ to stabilize the^multiplier/^VVe
could have one standard reserve ratio, say about 4 or 5% , on all noncapital
liabilities of all depository institutions, and th at would improve the link
between open m arket operations and M2# in-ordeHhat-MJ-growtb-wou:
) t hen p roduce r esults we-wanted. But, given the problems of velocity of the
last couple of years, I don't think that anvong ha« llp - r n n y tr t^ - t^nt we
could enthrone M2 through legislation and regulation, and achieve a better
result. So, there is no support for that at the moment.
In the past two years, after the Gulf W ar, there have been two rival
conjectures about monetary policy, and the role th at it played.

One view

was th at the recession was caused by an oil shock and w ar fears. According
to this view, all you had to do was remove this shock th at was depressing
economic activity, and you would naturally get a rebound. And, for a couple
m onths, in the spring of 1991, it appeared that was happening. Then the
economy simply went flat for about 4 quarters or so, and it's been kind of
bouncing around on a fairly low growth trend since. "Phat view says th at it

$Ifrv*
r

was a restrictive monetary policy, summarized by. M2 growth, th a t held
back total spending, causing subpar economic performance. According to
thrft view, all th at is necessary is to get our foot off the brake, to be more

13

U
rtL

generous in reserve supplying operations, and allow M2 growth to move up
into the middle of its target range, a n i the economy will be fine.
A second view, quite a different one, is th at various regional and
sectoral depressants have been at work over the last couple of years, and
these include such things as cuts in defense spending; the unwinding of the
commercial real estate overhang, and the redeployment of those resources to
something more productive; the corporate restructuring th at’s going on —
including all of the reengineering, or right sizing th at partly is technology
driven, the balance sheet restructuring related to the unwinding o^leverage
buyout and ju n k bond phenomenon of the '80s; the overhang of consumer
indebtedness; commercial bank recapitalization; and the demise of the
savings and loan industry. This set of depressants constitutes w hat Ala
Greenspan calls the "50 mile an hour headwinds." According to th atA
view,
m onetary policy has been very expansionary, as summarized by either the
exceptionally rapid growth of narrow measures of money (M l grew 14% in
1992, the highest rate for a year in the post-war period), or, the low level of
short-term interest rates, (fed funds being the lowest level in 30 years and
the steepest yield curve in post-war history). According to th at view, the
trick is to back off of the accelerator as these headwinds dissipate, so th at we
do not sow the seeds of soaring inflation.
The first view says it is a question of ''easing'' monetary policy. The
second view says it is a question of when is the appropriate time to begin to
"tighten". I think that both of those divergent movements of the money
measures can be reconcile

----- "

Nevertheless, w hat gives rise to my concerns is the lack of understanding on
the p a rt of the American people, and their elective representatives, as to
where w e're going.
If households and business do not believe the Central B ank has both
the intent and the ability to achieve and maintain price stability, they will
naturally look for evidence of changes in Central Bank policies. They will
assume th a t changes in the political party occupying the W hite House, or
controlling the U. S. Congress, will affect the objectives and results of
m onetary policy. They will assume that success or failure to achieve certain
fiscal policies will have implications for monetary policy. In short, they will
base their own plans and decisions on the assumption th at any society that
does not have the political will to constrain the growth of government
spending within the range of tax revenues will ultimately resort to debt
monetization and the consequent debasement of the currency. T hat means
they do not believe the stated objectives of the monetary authorities; there is
a lack of credibility in the commitment to price stability. O ur challenge is
one of trying to reinstitute a regime in which people once again believe that
any increase in inflation and interest rates is tem porary; the longer-term
trend is toward price stability.

M ORRIS: I remember th at debate in the 80s was at Berkeley with Jerry
Jo rd an where we had quite diverse opinions on the whole issue of monetary
targeting, and I now find th at there is very little difference between us. So
this session is going to be much less heated than the one at Berkeley was.

15

We have all learned much from the events of recent years. W hat I propose

j

to do is to take a look at the history of monetary policy since the Treasury Federal Reserve Accord in 1951, and try to identify some of the lessons we
should have learned from that 42 year experience.
I find it useful to break down the 1951-1993 experience into four
periods. Two of the periods were successful interest rate targeting periods:
one from 1951 through the mid-' 60s, the second from 1982 through 1993.
W e had one period of one policy failure in the '70s using interest rate
targeting, and one brief period of targeting non-borrowed reserves from *79
to ’82.
The first period , from 1951 through the mid -'60s, is widely viewed
as one in which the Federal Reserve conducted an extremely successful
m onetary policy. These years were the high w ater m ark for the U.S.
economy and, perhaps the high water m ark for U.S. monetary policy. It is
ironic th at the Chairm an during this period, William McChesney M artin,
Jr., who during his regime had the greatest success in controlling the
monetary aggregates, was a man who did not believe in targeting the money
supply. The conditions for monetary policy were generally pretty favorable
during those years. We had no supply shocks such as we had in the '70s.
We had a war, the Korean war, but it was a well financed war. There was no
sustained fiscal policy pressure on financial markets. Furtherm ore, we had
Regulation Q in place throughout the period.
Regulation Q had a great many faults, and nobody was terribly sorry
to see it go. But the fact remains that Regulation Q was a wonderful device

16

for monetary co n tro l.

All the Fed had to do to begin restraining the

economy was to move short term rates a little bit above the ceiling rate,
money would flow out of the banks and thrifts into the open m arket
instrum ents and there would be an immediate impact on all of the parts of
the economy th at were dependent on financing by the banks and the thrifts.
T hat meant th at with relatively small interest rate movements the economy
could be controlled quite well. Once Regulation Q died, there was no longer
the kind of credit rationing th at it produced. It took much larger interest
movements to have the same effect that small movements had in earlier
years.
There was no grand theoretical structure guiding m onetary policy in
those years. There was a very simple structure which M artin encompassed
in two sayings: one was "leaning against the w ind” and the other was "to
take away the punch bowl when the party was ju st getting going". They
m eant th at the Fed should "lean against the wind " whenever the economy
began to grow at a rate th at could not be sustained w ithout an acceleration in
the inflation rate. Conversely, policy should lean in the other direction
whenever the economy weakened. This very simple policy guideline
produced, as a by-product, modest and relatively stable rates of growth in
the monetary aggregates.
W hen the Accord initially took affect, the Fed decided to target free
reserves. There was no theoretical basis, whatsoever, for choosing free
reserves as a target for monetary policy. It was chosen because the Fed at
th a t time wanted to demonstrate to the market that it was no longer pegging

17

interest r a te s . As tim e passed and the m arket understood th at the Fed
would move interest rates to achieve its objectives, the Fed dropped free
reserves and moved to a policy of overtly targeting interest rates.
The leaning against the wind doctrine of the M artin years produced
four recessions in nineteen y e a rs, for which the Fed was much criticized by
the academic community. But all recessions were very mild, both the
inflation rate and the unemployment rate averaged out at very low levels
and the rate of productivity growth was at a historic high. We have not
succeeded since th at time in sustaining inflation rates and unemployment
rates as low as they were during that period.
W hen we turn to the second period, the period of failure with interest
rate targeting in the 1970s, we find that conditions were much more difficult
than in the 50s and 60s. We had the two oil price stocks, which were very
difficult to contend with. We had continued large budget deficits associated
with the Vietnam W ar, and we had the demise of Regulation Q.
The Federal Reserve moved vigorously in response to the 1973 oil
shock and produced the biggest recession we had seen since the 1930s. But
while the am plitude of the recession was about r i g h t , the recession was too
short. As a consequence, the inflationary expectations that were generated
by the Viet Nam w ar build- up and the 1973 oil shock emerged from the
recession of 1974-75 pretty much intact.
Probably the biggest problem for the policy maker in the 1970s was
the fixation on nominal interest rates. Remember th at the policy m aker had
been conditioned during the Regulation Q era to see very small interest rate

18

changes having big results. In the '70s, much larger interest rate increases
were needed to have the same results. ^Policy makers were impressed by the
fact th a t interest rates were extremely high by historical standards. But
those high rates were, in real terms, negative.

There was not much

emphasis on real rates either in the Fed or in the bond m arket. The bond
m arket was taken in by the same fixation on nominal rates. In retro sp ect,
it seems almost unbelievable that for ten consecutive quarters the real rate on
ten year Treasury notes was negative (based on the average inflation rate of
the preceding three years). During the entire period 1971-1980 the real rate
of return on ten year Treasuries was less than 1% in more than half of the
quarters.
The Fed failed to lean against the wind adequately during the years
1975- 77, when big trouble was building up. When they did begin to move in
1978- 79, driving the Federal funds rate up to 11.5% in S eptem ber, 1979, it
was too little and too late. W ith the second oil price shock, the horse was
out of the barn.
The turning point was the famous Saturday meeting of the FOM C in
October, 1979 when the Committee, by unanimous vote, decided not only
to move interest rate up sharply but to change the operating procedure -- to
quit targeting the Federal Funds rate and begin to target non-borrowed
reserves. The rationale was not that the whole Committee had suddenly
turned monetarist in their thinking, but that everyone had come to the
conclusion th at it was going to take extremely high interest rates to deal
w ith the powerful tide of inflationary expectations, and that, if the

19

Committee were overtly targeting interest rates, there would be political

I

difficulties in accomplishing those very high interest rates. It was felt th at
moving to the targeting of non-borrowed reserves would give the Fed a little
political shelter in this difficult time and, in fact, it worked.
When the Fed Chairm an met with the Congress and heard their
concerns about high interest rates, he could honestly say th at the Fed was not
controlling interest rates any more. They were controlled by the market; the
Fed was controlling the money supply. Congressman get a lot of letters about
interest rates, but they get few about the money supply. W hen the Fed had
to abandon targeting non-borrowed reserves in 1982, most Committee
members were reluctant to do so because they valued so much the political
sheltering th at targeting reserves rather than interest rates had given to the
Fed. B ut there was no choice. The Fed faced an exceptionally strong
demand for money, despite a very weak economy, and they had to meet that
demand or produce a financial crisis of vast proportions.
For obvious reasons, the FOM C did not want to announce that it was
again targeting interest rates, so the decision was made to target borrowings.
This was very similar to the decision after the Accord in 1951 to target free
reserves. As targets, both free reserves and borrowings have the same
characteristics as interest rate targeting. The manager, to hit these targets,
m ust give the m arket whatever non-borrowed reserves it demands. To do
otherwise would be to increase or reduce the level of borrowings from the
desired level.
The Federal Reserve learned a great deal from the experience of the

20

'70s. There is, I believe, widespread agreement th at the Fed should not allow
real short-term interest rates to fall to negative levels, except, perhaps, under
extremely depressed conditions. In the 1970s real short term rates were
negative for seven years (based on the trailing 12-month inflation rate). If in
the 1970s the Fed had been operating with a rule th at real short term rates
should not be negative, the recession of 1974-75 would have lasted longer,
the subsequent of the growth rate of the economy would have been more
subdued, and inflationary expectations would have been greatly diminished
by the time the second oil price shock arrived in 1979.
W here does th at leave us for the future? I've read papers recently
proposing th at the Fed target such variables as nominal GDP or commodity
prices or exchange rates or the yield curve. The fact is th at we cannot target
any of these in the same sense that we tried to target M l in 1979-82. The Fed
has only two instrum ents at its disposal. It can control short-term interest
rates or it can control the rate of growth of bank reserves. A target has to be
something th at is susceptible to reasonably precise control by one of these
two instrum ents.
M onetary policy-making would be simple, indeed, if by setting a
certain level of short- term interest rates or a certain rate of growth of bank
reserves , the Fed could determine with some precision the level of nominal
GDP. Unfortunately, there is no model which would perm it the Fed to
w ork such magic. I think the Fed's policy has got to be w hat it was during
the M artin years —one of "leaning against the w ind". The variables
proposed as targets for the Fed will provide useful information about the

21

w ind's direction and velocity.
.

i

-

W hat about the monetary aggregates? The aggregate against which
reserves m ust now be held is M l. U nfortunately, M l has become too
interest -sensitive to be a satisfactory target for monetary policy. The
problem lies in the pricing of interest- bearing M l accounts ,which has been
very sticky. As a consequence of this failure to adjust the rates paid on NOW
accounts to changing m arket conditions, there are wide swings in the spread
between NOW account interest rates and rates on CDs. W hen it becomes
very wide, such as it was back in early 1980 when the spread against the 6
month CD rate was 375 basis points, the growth in NOW accounts slows
sharply and dampens the growth of M l. In the past two years the spread
has dropped to between 65 and 80 basis points, the opportunity cost of
holding assets in NOW accounts has declined and we have seen very large
growth rates in M l.
W hat about M2 and M3? I find it surprising to read works written
by such normally sensible people as M artin Feldstein and Paul M cCracken,
arguing th at Federal Reserve policies are too tight because M2 is growing so
slowly. The slow rate of growth of M2 and M3 does not reflect monetary
policy. The non- M l deposits of M2 and M3 carry no reserve requirements.
They are not constrained by Fed policy; they are only constrained by the
capital position of the banks.
W ith the erosion of much bank capital due to losses in real estate,
many banks have had difficulty meeting the new capital requirements. They
And th at one way to improve their capital ratio is to reduce the volume of

22

their assets, and the best way to accommodate a decline in assets is to allow
their managed liabilities run off. The slow growth rates for M2 and M3
reflect those decisions. To get M2 to grow more rapidly in this context would
require an even faster rate of growth in M l than the historically high rate we
currently have.
Perhaps at some time in the future, when the banks become
comfortable about their capital positions, a stable relationship of M2 and M3
to the nominal GDP may again emerge. Until such time, the Federal Reserve
will have no option but to target the Federal Funds rate.

QUESTIONS AND ANSWERS:

JORDAN:

I have a couple of thoughts, and a question. Y ou're noting the

role of Reg Q as an instrum ent of policy in that period. W hatever we
thought about the efficiencies and the effectiveness of Reg Q, when you put
it into political economy or public choice context, it did involve identifying
"victim sectors" th at took the brunt of monetary policy actions.
M ORRIS: They always take the brunt of monetary policy actions.
JORDAN: But it was very, very focused when you created
disintermediation, which created a political reaction from the housing and
the auto sectors. I think^was the politics of it that killed it as much as any
argum ent in the profession about its desirabilities and efficiencies.
M ORRIS: I think w hat killed it was the money m arket fund.
—
JORDAN: That was an innovation th pt efln»csfr<mi«thnt environment. W hat

-fk tL

23

is intriguing to me is to think that back in the 1970s, C itibank wanted to
issue a VISA card with a credit balance on which they were going to pay a

'Jr

positive m arket rate of interest. The Board of Governors said , "If^quacks
like a deposit, it is a deposit, it's subject to Reg Q and Reg D, " therefore it's
uneconomic. Then M errill Lynch realized , "Gee, we can do that. We can
have a credit balance on a brokage account and pay a m arket rate of
interest."

So it was the Board of Governors' regulatory prohibition th at led

to the creation of an alternative vehicle and killed a p art of the banking
industry.
The other comment th at intrigues me is your emphasis on the role of
free reserves as a transition mechanism after the Accord to unhook political
and m arket sensitivities to interest rate targeting until the time was receptive
to go back, and then your characterization of the 1979-82 reserve targeting
as again, political shelter. Can you think about w hat you would do in the
current environment, with its highly politicized focus on short term interest
rates? After 5 years of cuts in the Funds rate, the first increase will be
headline news. Can you imagine using some other instrum ent as a transition
devise to unhook the markets and the political sensitivities from this
preoccupation with the funds rate, and if so what would it be?

FRANK M ORRIS: I would like to respond to two things. First, with
respect to Regulation Q. When Fidelity first decided to perm it customers of
their money m arket fund to withdraw their funds by means of a check, I
called up Ned Johnson and told him th at he was destroying the concept of

24

the money supply in the United States. He said th at he d id n 't have anything
like th a t in mind. They simply found ^that the cheapest way to handle
withdrawals was to give customers a special bank account and perm it them
to w rite checks on their balances. It was purely a dollars and cents
adm inistrative proposition.
The second issue is something the FOM C has got to be thinking about
now — how to handle the political flack when they have to start pushing
interest rates up. I don't think they can go before banking committees and
have as a credible response th at they are not controlling interest rates, th at
they're ju st controlling borrowings from the Fed. I don't think th a t will fly.
They’ll have to do what was done in the M artin years; do w hat is right and
stand up to the Congress.
HOSKINS: I am Lee Hoskins of the Huntington National Bank, Columbus,
Ohio. I'm ju st going to ask a question or two, and make an observation.
One observation - as many people in this room have established, is th at there
is a very large body of information and research about instrum ents and
targets, and it has always surprised me the paucity of research about the
value of price stability relative to that. So, I agree with Jerry. I think we
focused on the wrong thing through most of our academic efforts in terms of
looking at interest rate or money growth targets.
There are two central banks in large economies th at have a better
record than the US in terms of inflation in the last 25 years; one is
Germ any, the other is Japan. One country uses monetary targets, and the
other uses interest rates targets. I think all we are talking about is a very

25

built-in inflation around the mean in terms of choosing which target. In
terms of zero inflation, the question I w ant to pose is I think if the Central
B ank truly wants to get zero inflation, it can do it with either set of targets.
That is, interest rate or money growth targets.
The point is, we don’t have a clear cut objective about w hat we want
to do as a Central Bank. Talking about nominal GDP targeting seems to
miss the point. The Fed does not control nominal GDP. The Fed controls
one thing over time and that's the price level. Period. T hat's w hat it
controls. So why not target the price level, and how would you go about
doing it? That's my question. I never wanted to get to zero inflation
tomorrow because credibility is an im portant thing, but I d o n 't think
credibility has to come from an M2 target or an interest rate target.
Credibility can come from a clear statement of objectives, th at is, we want
price stability and a timeframe in which to achieve it. I picked five years
because contracts have a chance to adjust to it within th at timeframe. So the
question is, why not target the price level?
M ORRIS: The reason the Fed cannot be focused solely on the price level is
th at monetary policy is often the only flexible tool of economic policy in
W ashington. It would be nice if the Fed could take the position th at it was
going to focus on inflation and leave it to fiscal policy to take care of the real
economy. Unfortunately, we don’t live in th at kind of world.
I have never proposed a zero interest rate target. Realistically, if we
could get back to the kind of inflation rates we had during the Kennedy years
--1 to 1.5% -- where most of that inflation reflected the failure to adjust for

26

quality changes and where inflation was low enough th at it was not going to
distort investment decisions, I would consider th at a trium ph for the
Federal Reserve.
JORDAN: Now you know the difference between F rank and me. F rank
w ants to get back to the 1 to 1/2% inflation of the Kennedy years. 1 w ant to
get back to the 1 to 11/2% inflation of the Eisenhower years.
I think we can, and should, target the price level. To achieve this, I
would adopt a multi-year objective for the price level. This multi-year path
would be a long term objective, not an operating target. To make the price
level objective explicit and measurable, I recommend th at the FOM C choose
a tim e-path for the intended inflation rate that leads gradually to a
horizontal trend in the level of prices. For example, today 1 would choose a
p ath th a t achieves price stability in 1998. It would imply 2.5 percent
inflation in 1993,2 percent in 1994, 1.5 percent in 1995 and so on until the
intended inflation rate is zero in 1998.
In the short run, between meetings and within a year, the FOM C
would continue to use its best judgem ent in achieving this objective, based on
m onetary growth targets, its best models and a wide variety of other
indicators. But these judgements would be exercised within the context of
the m ulti-year objective that, in effect, would provide the benchm ark for
setting annual monetary targets. This is particularly im portant when
velocity is misbehaving.
I would tie our announcements as to what we are doing to the
Hum phrey-Haw kins requirements on the budget and try and put some teeth

27

in it. First, th at the adm inistration, the Office of M anagem ent and Budget,
and the Congressional Budget Office, sliould agree with the Fed on 5-yearout price levels* with a variance at an expected level, and tie together w hat
the monetary authorities are aiming for down the road -- sort of like long
term strategic planning in the corporate world. W ork backwards from th at
to interm ediate annual price levels (or rates of inflation — year-to-year
changes th at will produce th at price level). It is im portant th at the public
has in their mind th at five years or seven years down the road, plus or minus
two o r three percentage points -- here's what a dollar will buy.
Think about

4%inflation,

which a lot of people have in their minds

as low or moderate. Yesterday, John Taylor said th at he thinks th at the Fed
(his empirical results imply), has shifted up from 2% inflation to a 4% rate
of inflation. A 4% inflation to the parents of a baby born this year means
th a t when their kid is ready to be a college freshman, the purchasing power
of the dollar will be cut in half — or the cost of going to college, other things
the same, will double. That has to go into their planning. To me, th at is
unacceptable. W hat we ought to do is try and say to them, and effectively
communicate it through our actions (and this does involve reserve or money
targeting), th at the purchasing power of money is going to be thus and so
some years down the road and you can count on it.
M ORRIS: I think the problem is that it is very difficult for the
Adm inistration and the Congress to think in a long-term context. I
rem em ber when Jerry was on the Council of Economic Advisors and the
"rosy scenario" came out. The "rosy scenario" showed a 4% real growth

28

rate into the distant future accompanied by a steadily declining interest rate.
I asked Jerry how they came up with /his forecast. Jerry said it was
simple: the supply-siders did the real economy and the monetarists did the
inflation rate.

,
-r~

)h

.

I

'‘
JORDAN: It was worst- Ihoififlpiil because the eclectics did nominal GDP
and it turns out ¥-di4nlt-equaf-)C plus P.

/

T

FURLONG: I am Fred Furlong from the Federal Reserve B ank of San
Francisco. The fundamental problem is th at no one knows exactly w hat the
right level of interest rates is in some sense. Yet, you have to have something
th a t you are reacting to. When the FOMC is deciding, "Well, do we change
the Fed Funds rate or not?" they are reacting to something. I guess Jerry ’s
suggestion is th at the Fed should react to the price level. The question for
F ran k is, if it's not the price level, what's your alternative? W hat is it that
you are basing your decisions on in terms of how you move th at interest rate
around?
M ORRIS: At any FOMC meeting most members have a concept of what
the optimal realistic target for nominal GDP ought to be over the next
period. T hat is not likely to be zero inflation and a 4% real growth rate. It
will often be a higher inflation rate than one would like to see over the long­
term and a lower rate of growth than one would like to have over the long­
term . The issue is the optimum mix that can be reasonably atttained over
the next four quarters. That's what I always had in mind when approaching
a m onetary policy decision.
JORDAN: That response surprises me a little bit because in your initial

29

^

rem arks you focused on the level of real interest rates. The lesson of the '70s

!

was the problems of having persistent negative real interest rates, and you
said we would not make th at mistake again. I thought you would say th at in
today’s environment with inflation measured over some period, we do have a
negative short- term real interest rate. You would use as your target, raising

iP

the nominal rate to a level so th at it was non-negative. The problem with
responding to nominal GDP, or worse, real output, is that any pickup in
nominal GDP growth in the current context, given lags, is almost certainly
going to be the result of faster real output and employment growth. So, if we
use nominal GDP as the trigger to raise the funds rate it will cast the Fed as
being antigrow th. That gives us a real political problem. The reason I was
disturbed by last year's action to cut the funds rate on the day of, or the day
after, release of the nonfarm payroll employment, was th at it conditioned the
m arket to believe th at all we cared about was employment. They concluded
th at with the first sign of strength in employment we were going to raise the
funds rate. T hat casts the Fed as a scrooge, because we would be viewed as
anti-employment and anti-growth.

ODRISCOLL:
'

I am Jerry O'Driscoll from the Federal Reserve B ank of Dallas.

W e've got the objective, and in your presentation, Jerry, you very clearly set
out th at in order to attain th at objective you need credibility and you need a
nominal anchor. Now, I don't agree with Lee Hoskins here that you get
credibility by announcing something. In fact, I think you lose credibility
because we've announced so many things in the past that we haven't done,
th at another announcement w on't do it. It's got to be actions over sustained

30

period of time. If the actions are going to be successful it seems to me that,
you're right, we need the nominal anchor. Do you have a suggestion for a

* cM isurrO C^c^
j& i
.

nominal anchor th at will work in this environment? ___
/
^
n
~ ____ n C L a sJ tH
JORDAN: Not o»c=4bAtJ=e»n-s«lK/I n o rd erto reconcile the problem of

-

r

interpretation of current M l and various combinations of things to produce
a broader measure p^M 2,1 use the monetary base —adjusted for increases in
foreign holdings of U.S. currency. However, we do not have timely
inform ation.

'd I think th at the focus has to be on longer-term objectives in terms
fefty
of the price level. We should change our rhetoric and our focus. Instead of
describing our actions as "tightening" and "easing" -- all of those words
th a t are used about our actions implying stimulus or restraint —we should
instead, ju st talk in terms of achieving sound money. Plain and simple,
nothing else. T hat's all we are trying to do is stop the debasement of the
national currency.
QUAYLE: I am Phillip Quayle from Ball State University. The Fed's
prim arily a political institution, a creature of Congress. The reason why we
can 't agree upon targets is that there is no agreement among the general
populous. If the Fed follows a target that is not consistent with w hat we, the
public, w ant, the Fed is not going to last. Evolutionary mechanisms assure
th a t the Fed's not going to do anything extraordinarily painful, at least,
perceived to be painful. If they attem pt to do that, it's going to be changed.
So in essence, the shifting targets you get are explained by the fact that
there's no consensus among the American public on zero price inflation or

anything else th at monetary policy should do.
M ORRIS: The saving grace is that the Federal Reserve is given a lot of
leeway, in terms of time, because of the slowness of the legislative process. It
is tru e th at the Federal Reserve cannot follow indefinitely policies which are
not acceptable to the public. However, for a considerable period of time the
Fed can, and has, pursued policies th at were unpopular. I am not sure th at
the policies followed in 1979-82 would have been supported by a majority
1
;
'

vote of the public. Fortunately, the system gave the Fed the leeway to follow
those policies and to show some results before the rebellion got too hot and
heavy.
JORDAN: I agree with the thrust of your comment and it was the same th at
Lee made earlier about, in effect, m arketing the idea of price stability. In a
sense, the R&D has already been done as to the benefits of achieving stable
money, but we have not engaged in the successful marketing effort to ptnsiie
the American public or their elected representatives as to why th at's the case.
Nor have we invested enough in the techniques of implementation of
monetary policy to be persuasive th at we can, in fact, achieve it. And so, as a
research strategy I would shift resources toward how to achieve price
stability and persuade people that it is desirable to do so.
FAND: I am David Fand of the Center for the Study of Public Choice,
George Mason University.

I have one question for Frank M orris. Let's

look at ju st 1990, '91 and '92, to make it very simple. If the Fed had hit the
m idpoint of their own target, those three years, would we have been better
off, would we have had more income, more employment and more jobs than

32

Jh

we have now? And would it have been desirable?
M O RRIS: M y answer is no. I think tile only way the Fed could have hit the
m id-point for M2 would have been to increase M l enought to offset the ru n ­
off o f CDs.

Remember th at in the two years 1991 and 1992, b ank reserves

rose by 30% , an extremely rapid growth rate by historical standards. To
have increased reserves much, much more in order to meet an M2 target by
increasing M l does not seem to me to make sense. The whole concept of an
M2 target for monetary policy is completely outdated by the real facts of the
situation.
Remember, also, that we have a bond m arket th at is very different
th an in the early 1970s, and I think th at we are very lucky th at it is
different. W e have a bond market that has to be persuaded every day that
the Fed is sufficiently concerned about inflation. The kind of policy you’re
advocating here would have a very adverse impact on the bond m arket.
JORDAN: I would like to add th at it is this lack of credible commitment to
w here w e're going to come out at the end of the day, th at is our problem.
Rem em ber 1990. The CPI shot up to over 6% in the midst of w hat was
being touted as the third oil shock, and people remembered w hat happened
during the 1973-1974 and 1979-1980 oil shocks. Even aside from the shock
you were still averaging between 4 and 5% inflation. The M ichigan survey
and everything else you looked at showed an expectation of a persistent
inflation. You had the badgering by the Brady Treasury and others in the
Bush adm inistration about Fed policy being "too tight" and needing to "ease
u p". Because of this, more aggressive reserve supplying operations in an

33

effort to get up into the midpoint of the M2 target range would have resulted
in simply an "A h-H a" experience on tlie part of the American public;
namely th at the Fed had thrown in the towel and was sacrificing its
objectives for political expediency. The dollar would have plunged and bond
yields would have risen. If th at had happened, then the economy wouldn’t
be better off today. Your question was, would the economy be better off?
And I ’m saying, given the lack of a credible commitment to price stability I
don't know if we would have been better off or not. I wish I did.

34