Full text of Review (Federal Reserve Bank of St. Louis) : January/February 1988

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

January/February 1988

Vol. 70, No. 1




3 Are IWide Deficits a P rob lem ?
12 Protec tionist Trade Policies: A Survey
o f T h e o ry » Evidence and Rationale
30 T h e B orrow ed -R eserves O perating
P ro ced u re: T h eo ry and Evidence
55 F a rm Policy and M an d atory Supply
C ontrols — The Case o f Tobacco

THE
FEDERAL
A RESERVE
Jtk RANK of
A r ST. LOUIS

F e d e ra l R e se rv e Bank of St. L ouis
Review
January/February 1988

In This Issue . . .




T h ere is co n sid era b le c o n c e rn th at th e su b stan tial U.S. trad e d eficits in cu rred
in re cen t y ears are a sym ptom or, even w orse, a ca u se o f gen eral w eak n ess in the
U.S. eco n om y. In th e first article in th is R ev iew , K. Alec C hrystal an d Geoffrey E.
W ood an sw er th e qu estion , "Are T rad e D eficits A P ro b lem ?” T h ey first d em o n ­
strate th e relatio n sh ip b etw een trade d eficits an d cap ital su rp lu ses u sin g th e U.S.
b alan ce o f p ay m en ts a cco u n t. T h ey th e n d istin guish betw een circ u m sta n ce s in
w h ich trade d eficits are p ro b lem a tic — for exam p le, th o se p ro d u ced by inflation aiy d o m estic m o n etary p o licy — an d circ u m sta n ce s in w h ich they are not
— for exam ple, th o se p ro d u ced bv high private investm ent relative to the
d o m estic savings available to fund th em . T h ey co n clu d e by n otin g that the
p resen t trad e d eficits ap p ear to have resu lted from a co m b in a tio n o f large federal
d eficits and investm en t spen din g. In sofar as th e federal d eficit an d private savings
behavior are taken as given, th e ch o ice s that th e U nited States faces are co n tin u ed
high lev els o f private in v estm en t d em an d an d trad e d eficits o r b a la n ced trad e and
slow real grow th. In th e cu rren t case, red u cin g th e trade d eficit m ight well
p ro d u ce th e g reater p roblem .
*

*

*

P ro tectio n ist p ressu res have b ee n m o u n tin g w orldw ide d uring th e 1980s. In the
se co n d article in th is R eview , C letus C. C oughlin, K. Alec C hrystal an d Geoffrey E.
W ood survey th e theory ev id en ce an d ratio n ale co n cern in g p ro tectio n ist trade
p o licies. T h e au th o rs illu strate th e gains from free trade u sin g th e co n c e p t of
com parative advantage an d review re ce n t d evelop m en ts co n cern in g th e c o n s e ­
q u en ces o f in tern a tio n a l trade in im p erfectly com p etitive m arkets. T h ey argue
that, w hile p ro tectio n ist trade p o licies o cca sio n a lly m ay offset foreign m on op oly
p ow er or advantageously u se d o m estic m o n o p o ly pow er, trad e restrictio n s
g enerally re d u ce b o th th e co m p etitio n faced by d o m estic p ro d u cers and the
co n su m p tio n p o ssib ilities o f d o m estic co n su m ers.
T h e em p irical evid en ce is clea r-cu t. T h e co sts o f p ro tectio n ist trade p o licies
b o rn e by co n su m ers far e x ceed th e gains o f d o m estic p ro d u cers an d governm ent.
M oreover, th e adverse c o n su m e r effects are n o t short-lived. P ro tectio n ist trade
policies g en erate low er ec o n o m ic grow th rates th a n free trade p o licies. C o n se­
quently, n ation al in terests will b e served by th e red u ctio n o f trad e b arriers.
*

*

*

In the third article, en titled ' T h e B orrow ed R eserves O perating P roced u re:
T h eory and E v id en ce,” D aniel L. T h o rn to n analyzes th e Fed eral R eserve’s o p era t­
ing p ro ced u re as a m eth o d for co n tro llin g th e m o n ey stock o r th e fed eral funds
rate. T h o rn to n d em o n stra tes th at th e borrow ed reserves o p eratin g p ro ced u re is
n ot an effective m eth o d for co n tro llin g th e m o n ey stock. Ind eed , h e show s th at an
in terest rate targeting p ro ced u re w ould be m o re effective in co n tro llin g m oney.
Fu rtherm ore, h e show s th at th e borrow ed reserves op eratin g p ro ced u re is an
effective m eth o d o f co n tro llin g in terest rates in th e sh o rt ru n only w h en th e
variation in borrow ing is due solely to shifts in th e d em an d for total reserves; in
th e long run, it is an effective m eth o d o f co n tro llin g in tere st rates only w h en the
borrow ings fu n ctio n is stable.
T h o rn to n also investigates th e u se o f th e b orrow ing fu n ctio n s sin ce O ctob er
1982. His evid ence suggests th at th e borrow ings o p eratin g p ro ced u re h as b een
u sed to offset th e effect o f p erm an en t shifts in borrow ings on th e federal funds
rate.

1

In This Issue . . .


2


T h e farm se c to r o f the U nited States ex p e rie n ce d a severe d o w n tu rn in the
1980s d esp ite over $95 b illion in governm ent su p p ort. M any an alysts believe th at
su ch govern m en t su p p o rts an d o th e r tradition al U.S. farm p o licies have d on e
little to alleviate ag ricu ltu re’s problem and m ay have actu ally co n trib u te d to the
farm crisis. As a result, so m e have p ro p o sed m an d atory co n tro ls o n crop p ro d u c­
tion as an alternative to th e cu rren t policy o f voluntary acreag e red u ctio n and
d irect governm en t paym ents.
Su ch m an d atory co n tro ls have b een u sed in th e U nited States to su p p o rt the
p rice o f to b a c co sin ce the 1930s. In th e final article in th is R eview , “Farm Policy
and M andatory Supply C on trols — T h e C ase o f T o b a c c o ,” K en n eth C. C arraro
exam in es U.S. to b a c co policy to analyze th e likely co n se q u e n ce s o f exten d in g
m an d ato iy su p p ly co n tro ls to o th er m a jo r U.S. crop s. C arraro show s that the
United States w as able to u se m an d atory su p p ly co n tro ls su ccessfu lly for m any
y ears only b ec a u se b o th w orld su p p ly and w orld d em an d w ere in elastic. As th e
w orld elasticity o f b o th su pply an d d em an d in crea sed over tim e, how ever, the
effectiveness o f U.S. to b a c co policy erod ed . T h e a u th o r c o n c lu d e s that m a n d a to iy
su pply co n tro ls for o th e r crop s w ould n ot b e su cce ssfu l b ec a u se th e co n d itio n s of
elasticitv th at w orked well previously for to b a cco are n ot p resen t for o th e r m ajo r
crop s.

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Are Trade Deficits a Problem ?
K. Alec Chrvstal and Geoffrey E. W ood

I n 1986, the U.S. trad e deficit ex ceed e d $140 billion.
Su ch su bstan tial trade d eficits often are co n sid ere d a
sign o f w eak n ess in th e eco n om y. W hile this situ ation
is so m eth in g o f a novelty for th e U nited States, m any
o th e r co u n tries have h ad trad e d eficits off an d on
th rou gh o u t th e p ostw ar period.'
T h e p u rp o se o f th is article is to explain w hat is
m eant by trade deficits w ithin th e co n tex t o f th e b a l­
a n ce o f paym ents, to ou tlin e th e circ u m sta n ce s u n d e r
w h ich the state o f the b a la n ce o f pay m ents m ay be
sy m p to m atic o f a problem , and to co n sid e r w hat this
analysis im plies cu rren tly for th e United States. W ith
regard to th e last, w e will suggest that co n c e rn about
th e U.S. trad e d eficit h as b ee n overstated. Ind eed , a
trade d eficit can be indicative o f a h ealth y and strongly
grow ing eco n om y.

THE BALANCE OF PAYMENTS
ACCOUNTS
T h e b alan ce o f pay m ents a cc o u n ts are a reco rd o f
tra n sactio n s b etw een d o m estic resid en ts an d th e rest
o f the w orld over a sp ecific perio d o f tim e. Like any

K. Alec Chrystal is the National Westminster Bank Professor of Per­
sonal Finance at City University, London. Geoffrey E. Wood is a
professor of economics, also at City University, London. This article
was written while Chrystal was a professor of economics at the
University of Sheffield, Sheffield, England. Vincent T. Waletzki provided
research assistance.
'The United States did run deficits in the 19th century, but not quite
as big relative to GNP as are current U.S. trade deficits. See Mudd
and Wood (1978).



d ou ble entry bookkeeping system , th e b a la n ce o f pay­
m en ts a cc o u n ts m u st b alance.- T h e re is n o th in g m ys­
teriou s about this, n o r d o es it involve any statem en t
about h ow th e w orld w orks.
T h e sim p lest form in w h ich th e b a la n ce o f pay­
m en ts a cc o u n ts ca n b e ex p ressed is as follow s:
(1) CA + K + F = 0,
w h ere (.'A is th e cu rren t a cc o u n t b alan ce, K is net n o n ­
official cap ital flows an d F is official reserve financing.
T h e se item s are defined in su ch a w ay th at th ey m u st
sum to zero. Let u s co n sid e r ea ch o f th em in tu rn.

Current Account
T h e cu rren t a cc o u n t h as two m ajo r co m p o n e n ts.
T h e se are th e trad e b a la n ce an d th e seiv ices or “invisi­
b le s ” b alan ce. T h e form er, w h ich generally gets the
m ost atten tio n , is th e d ifference b etw een th e value of
goods exp o rted an d th e value o f good s im ported.
T h e se exp orts an d im p o rts are o f ph y sical o b jects
w h ich , in prin cip le, co u ld b e observed cro ssin g the
bord er. In co n tra st, “in v isib les” are services for w h ich
in tern atio n al p ay m en ts are m ad e b u t that do not

2Because of measurement errors, the actual accounts add in a
“statistical discrepancy” which when included in (1) ensures bal­
ance. The reason we say that they must balance, however, is not a
statement about the accuracy of the statistics. The current and
capital account (including official balance) are defined to be equal
and opposite. Think of the current account as the excess of income
over spending. The capital account is then merely net saving, which
is equal to income minus spending. If you measure saving as
negative and the excess of income over spending as positive, they
will obviously add up to zero.

3

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

involve th e d irect tran sfer o f a p h y sical p ro d u ct. For
exam ple, if a New York sh ip p in g co m p an y w ere to
in su re a cargo w ith Lloyds o f London, th e p u rch a se o f
th at in su ran ce co n tra ct w ould rep rese n t an invisible
im port for th e United States and an invisible exp ort for
the United Kingdom .
Invisibles take m any different form s. Tw o exam p les
are w o rth m entio n in g in ad d ition to su ch financial
send ees as in su ran ce an d banking. First, if a n ation has
eith e r a ssets or liabilities overseas, th e n et paym ent o f
in terest o r dividends is m easu red as an invisible im ­
p ort or export. A positive n et retu rn on foreign assets is
co u n te d as an invisible export, b eca u se it g en erates an
inflow o f pay m en ts into the eco n om y ju st as an export
o f good s d o es. Seco n d , in tern atio n al to u rism is
co u n te d as part o f th e invisible co m p o n e n t o f th e
cu rren t a cco u n t. If U.S. citizen s spend m ore on over­
seas trips th an foreigners sp en d on U.S. vacations, it is
m easu red as an invisible n et im port in th e U.S. b a la n ce
of p aym ents.

Non-Official Capital Account
T h e cap ital a cc o u n t o f th e b alan ce o f paym en ts
m easu res th e ch an g e in n et in d eb ted n ess b etw een th e
d o m estic eco n o m y and th e rest o f th e w orld. It is
im p o rtan t to get this clear, as th ere is so m etim es
co n fu sio n abou t w hat th e cap ital a cco u n t co n ta in s. It
d oes n ot involve im p o rts and exp orts o f cap ital goods,
su ch as m a ch in e tools an d co m p u ters. T h e se are all
p h y sical goods, and th e ir im port an d exp ort are th e re ­
fore co u n te d in th e trad e a cco u n t. T h e cap ital a cco u n t
involves th e tran sfer o f fin an cial claim s o f various
kinds. T h e se claim s are referred to as “ca p ita l” b e ­
ca u se th ey re p resen t claim s to in terest o r dividend
p aym en ts and, in th e ca se o f co m p any shares, do
involve ow nership o f u nd erlying real assets.
T h e term in ology co m m o n ly u sed to d escrib e the
cap ital a cc o u n t is rath er co n fu sin g w h en it is related
to th e way in w h ich cap ital a cc o u n t item s are m e a ­
su red. In th e cu rren t a cc o u n t o f th e b a la n ce o f pay­
m ents, goods leaving th e co u n try is m easu red as a
plus item . In th e cap ital acco u n t, how ever, w hat is
generally called a cap ital “outflow ” is m easu red as
negative. Only th e term in ology h ere is confusing, h o w ­
ever; a cc o u n ts are qu ite logical. W hat w e m ean by a
cap ital outflow is th at d o m estic resid en ts are buying
foreign assets. In o th e r w ords, th ey are “im p o rtin g ”
foreign shares, titles or secu rities. T h u s, all p u rch a ses
of foreign goods, secu rities (stocks, bon d s, bills) or any
o th e r asset are m easu red as negative (im ports), an d all
sales to foreigners are m easu red as positive (exports)
irrespective o f w h eth er they are goods sales or asset
sales.

http://fraser.stlouisfed.org/
4
Federal Reserve Bank of St. Louis

In prin cip le, th e cap ital a cc o u n t o f th e b a la n ce of
pay m en ts m easu res th e ch an g e in th e n et asset/
liability p o sitio n b etw een the h o m e eco n o m y an d the
rest o f th e w orld. W e say “in p rin cip le ” b ec a u se th ere
is o n e re sp ect in w h ich th is is n ot co rrect. T h e capital
a cco u n t m easu res th e value o f th e n et flow o f financial
in stru m en ts (stocks, bon d s, bills, etc.) th at p asses b e ­
tw een d o m estic and overseas resid en ts. But the ex ter­
nal in d e b te d n e ss o f an eco n o m y ch a n g es not ju st as
a ssets ch an g e h an d s. It also ch an g es as a result of
ch an g es in values o f a sse ts th at have n ot ch an g ed
h an d s. F o r exam ple, U.S. resid en ts m ay ow n sh ares in
Rolls Rovce w h ich rise in value. T h is cap ital gain (or
loss) elem en t o f th e extern al asset/liability p o sitio n is
n ot m easu red as part o f th e b a la n ce o f paym en ts
a cc o u n ts u ntil it is realized by an asset sale. Only th e
flow o f fin an cial claim s is in clu d ed .3

Official Balance
T h e final item in th e b a la n ce o f p ay m en ts a cc o u n ts
is the b a la n ce for official fin an cin g. T h is co m p rises
ch a n g es in the official foreign ex ch a n g e reserves o f th e
d o m estic eco n om y. T h e se reserves are m ainly claim s
against foreign governm ents lor cen tra l banks), for
exam ple, Fed h old ings o f D eu tsch e m arks. F o r m ost
co u n tries, reserves are h eld as a m e a n s o f in tervening
in foreign ex ch a n g e m arkets to su p p o rt th e value o f
the d o m estic cu rren cy .4 T h is item is a sp ecial official
se c to r co m p o n e n t of th e cap ital a cco u n t. It is treated
sep arately for h isto rical rea so n s a sso cia te d w ith the
fixed ex ch an g e rate system w h ich o p erated alm ost
w orldw ide from W orld W ar II u ntil 1973.3 U nd er a
freely floating ex ch a n g e rate regim e, th e official fin a n c­
ing b a la n ce is alw ays zero. If F in eq u atio n 1 is zero,
clearly, CA an d K m u st b e equal an d o f o p p o site sign.

U.S. Balance o f Payments
Table 1 show s the U.S. b a la n ce o f pay m en ts for 1986.
It show s a cu rren t a cco u n t d eficit o f a little over $141
billion. T h e cu rre n t a cc o u n t is m ad e up o f item s 1 and
2. T h e cap ital a cc o u n t su rp lu s o f $117 billio n is show n
in lin es 3 an d 4. C hanges in U.S. official reserves are
show n in lin e 5. T h ere w as a very sm all fall o f $0,312
billion in 1986 (a p lu s sign in d ica tes a d eclin e in
h old ings o f foreign assets). T h is in d ica tes th at th e U.S.
au th o rities in terv en ed little d uring 1986 as a w hole.

3Some have claimed that the United States has become a net debtor
vis-a-vis the rest of the world. This claim ignores the capital gains on
U.S.-owned foreign assets; in reality, the United States is likely still
to have positive net external assets.
4See Balbach (1978).
5See Batten and Ott (1983).

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

Table 1
U.S. Balance of Payments: 1986
(millions of dollars)___________________
1) Merchandise trade
2) Invisibles net
Balance on current account
(1 plus 2)
3) Change in U.S. assets abroad
(increase - )
4) Change in foreign assets
in U.S. (increase + )
Balance on capital account
(3 plus 4)
5) Change in U.S. official
reserves (increase - )
Statistical discrepancy
(1+ 2 + 3 + 4 + 5)

$-144,339
+2,987
$-141,352
-96,294
+213,387
+117,093
+312
$23,947

NOTE: The merchandise trade balance is exports minus imports.
The invisibles balance is the sum of: net military
transactions; net investment income; other service
transactions; net remittances, pensions and other
transfers; and U.S. government grants (non-military).
SOURCE: U.S. Department of Commerce.

sp en d on its exports. T h is will p ro d u ce an im b alan ce
in its foreign ex ch a n g e m arket.6 A ttem pted sales o f
d o m estic cu rre n cy (for foreign cu rren cy) will exceed
a ttem p ted p u rch a se s. T h e m arket value o f th e cu r­
ren cy will fall u ntil th e qu an tity o f th e cur ren cy d e­
m an d ed is equal to th at su p p lied . At th is point, eith e r
th e cu rren t a cc o u n t h as a d ju sted so th at it is no lon ger
in deficit; o r th e n et exp ort o f a sse ts (in d uced as a ssets
in the co u n try b eca m e ch ea p er, th rou gh d o m estic
cu rren cy devaluation, an d th u s m ore attractive to for­
eigners; an d p rices o f foreign assets b eca m e h ig h er
and h e n c e less attractive to U.S. citizen s! is just equal
to th e cu rren t a cco u n t deficit. T h u s, th e ex ch a n g e rate
will a d ju st to en su re th at th e cu rren t an d cap ital
a cc o u n ts are exactly offsetting.
T h ere is n o th in g m agical abou t this o u tco m e. T h e
en d resu lt is th e sam e for any individual. If y ou sp en d
m ore th an y o u r in co m e, y ou m u st b orrow o r sell th e
equivalent value o f y o u r assets to cover th e d ifference;
if y o u sp en d less th a n y o u r in co m e, y ou m u st inevita­
bly acq u ire in crea sed claim s on so m eo n e else. Sim i­
larly, a n atio n that ru n s a cu rre n t a cc o u n t d eficit m u st
eith e r borrow from abroad o r sell off som e o f its assets,
w h eth er th ese assets are d o m estic o r foreign. Like­
w ise, a cu rren t a cc o u n t su rp lu s m u st be a sso cia ted
w ith eith e r an in crea se o f claim s on foreigners o r a
re d u ctio n o f previous borrow ings.
A nother im p licatio n o f the defin ition of b a la n ce of
pay m en ts is th e follow ing identity:
(2) CA = GNP —GDE.

T h u s, th e d o m in an t p ictu re is on e o f U.S. resid en ts
buying m ore goods and services overseas th a n foreign
resid en ts are buying from th e U nited States an d of
foreigners in creasin g th e ir net hold ing o f claim s
against th e U nited States.
N otice, how ever, th at th ere is a fairly large statistical
d iscrep an cy . T h e p re se n c e o f th is d iscrep an cy in d i­
ca tes that th e data do not in clu d e som e trade and/or
cap ital flow s. W hile it is im p o ssible to say w h ere the
in a ccu ra cies arise, it is often p resu m ed that th e g reat­
est errors are likely to be in th e cap ital a cco u n t, p ri­
m arily b ec a u se asset tran sfers are m ore difficult to
keep reco rd s on. If th e d ata h ad n o o m issio n s, th en
the cu rren t and cap ital a cc o u n ts (includ ing official
flows) w ould add to zero.
It is n o t obvious at first g lan ce w hy th e cu rre n t and
cap ital a cc o u n ts m u st offset ea ch o th e r exactly. W hat
w ould h ap p en if they did no t? Su p p o se for exam ple,
th at at cu rren t ex ch a n g e rates a co u n try is ru n n in g a
cu rren t a cc o u n t d eficit bu t its p la n n e d n et cap ital
flows are zero. T h is m eans that th e co u n try is trying to
sp en d m ore on im p orts th an foreigners are w illing to



T h e cu rren t a cc o u n t su rp lu s (or m inu s th e cu rren t
a cc o u n t deficit) is equal to gross n atio n al p ro d u ct
m inu s gross d o m estic ex p en d itu re. T h is identity
show s th at th e cu rren t a cc o u n t o f th e b a la n ce of
p aym en ts is the difference b etw een th e value o f w hat
th e n ation p ro d u ces an d w h at it sp en d s. T h e form er
(GNP) ca n also be th ou gh t o f as th e value o f the
n a tio n ’s gross in co m e. Id entity (2) is useful b eca u se it
m akes cle a r th at any n a tio n th at sp en d s m ore th an it
p ro d u ces will have a trad e deficit. T h e in terestin g
qu estion , o f co u rse, is w h eth er su ch an im b alan ce is
good or bad.

WHAT MAKES THE CURRENT
ACCOUNT BALANCE A PRO BLEM ?
T h e n atu re o f w hat is u sually term ed a b a la n ce of
paym en ts problem varies co n sid erab ly, d ep en d in g

6See Chrystal (1984).

5

FEDERAL RESERVE BANK OF ST. LOUIS

upon w h eth er th e co u n try in q u estio n has a fixed o r a
floating exch an ge rate regim e. T h e p roblem p ro d u ced
by a deficit on the cu rren t a cc o u n t can be m ost a cu te if
th e n ation is m ain tain in g a fixed ex ch an g e rate re­
gim e.7 In th is case, "th e p ro b lem " is felt d irectly by the
cen tral bank.
M aintaining a fixed ex ch an g e rate vis-a-vis on e or
m ore co u n tries requ ires th e pegging nation s cen tral
ban k to h o ld foreign ex ch an g e reserves w ith w h ich to
intervene in the foreign ex ch an g e m arket. T h is in ter­
v en tion can be n ecessary to stop th e ex ch an g e rate
from m oving in eith e r d irectio n . Su ppose, for exam ple,
that th e co u n tiy has a cu rren t a cc o u n t d eficit an d no
d esired n et private cap ital flow s. In o rd er to m ain tain
th e existing ex ch an g e rate, th e cen tral bank m u st sell
foreign ex ch an g e for its d o m estic cu rren cy . W h eth er
the origin or so u rce o f th e n et supply o f d o m estic
cu rre n cy in foreign ex ch an g e m arkets is from the
cu rren t or cap ital a cc o u n t side o f the b alan ce o f pay­
m ents is irrelevant. T h e d o m estic cu rre n cy value of
reserves sold in a p articu lar period is th e official
fin an cin g b alan ce, F, in eq u atio n 1. B eca u se it involves
the sale to foreigners o f a d o m estically h eld asset, a net
loss of reserves is m easu red as positive in th e b a la n ce
o f p aym ents a cco u n ts.
U nd er a fixed exch an ge rate regim e, ex ch an g e rate
p ressu re p o ses a p roblem if th e cen tral bank in q u e s­
tion starts to run ou t o f foreign ex ch an g e reserves.
T h is possibility m akes th e p roblem w orse b eca u se
hold ers o f th e d o m estic cu rren cy , fearing a devalua­
tion, will tiy to bu y foreign cu rren cy . Speculative sales
of th e d o m estic cu rre n cy in foreign ex ch an g e m arkets
force th e cen tral bank to sell even m o re foreign ex ­
ch an g e reserves. Inevitably, th e n atio n m u st eith e r
devalue its cu rre n cy o r in tro d u ce m easu res to cu t
d o m estic sp en d in g (includ ing sp en d in g on foreign
goods). T h is actio n is u navoidable; o th erw ise, th e c e n ­
tral bank will run out o f foreign ex ch an g e reserves.
T h is d escrib es th e n atu re o f m ost b a la n ce o f pay­
m ents crise s ex p erien ced bv co u n tries attem p ting to
m aintain fixed ex ch an g e rates in th e 1950s an d 1960s.
It is w orth noting, how ever, that th e U nited States
u n d e r th e p ostw ar "B retto n W o o d s” regim e w as n ot
th e sam e as o th e r co u n tries.8All o th e r co u n tries in the
system pegged th eir cu rre n cies to the d ollar an d h eld

The exception to this is when a currency is depreciating at a fast
rate. This is a symptom of acute internal problems normally associ­
ated with hyperinflation.
8The system was named after the place in New Hampshire where the
final negotiations setting it up were held in July 1944.

http://fraser.stlouisfed.org/
6
Federal Reserve Bank of St. Louis

JANUARY/FEBRUARY 1988

d ollar reserves for th is p u rp o se. T h e lin ited States,
th erefore, did n o t n eed to su p p o rt its ow n ex ch an g e
rate and, in fact, did not h old significant reserves of
foreign cu rre n cy d uring th is period .0
S in ce th e spring of 1973, w h en all the m a jo r in d u s­
trial co u n tries m oved to a floating ex ch a n g e rate re­
gim e (the United K ingdom h ad floated in Ju n e 1972),
th e n atu re o f b a la n ce o f p aym en ts pro b lem s has
ch a n g ed .10 U nder a floating ex ch a n g e rate system , a
cen tral ban k d o es not have to u se its foreign ex ch an g e
reserves to fin a n ce a d eficit in th e n on -official part o f
th e b a la n ce o f p ay m en ts; in fact, th ere w ill b e n o n e ."
In eq u ation 1 above, the term F b e co m e s zero. In stead
o f cen tra l ban k in tervention, th e ex ch an g e rate m oves
to assu re th at th e cu rren t a cc o u n t an d th e capital
a cc o u n t su m to zero on th e ir own.

WHY WORRY ABOUT THE TRADE
BALANCE?
C o n cern about the state o f the trade b a la n ce lias a
long h isto iy . It is useful to put this c o n c e rn in h isto ri­
ca l co n text, as it lead s n atu rally to th e analysis o f w h en
su ch co n c e rn is justified.
In th e follow ing d iscu ssio n , w e take it as given that
trade itself is ben eficial, a p o in t not clearly esta b lish ed
until R ica rd o ’s fam ous d em o n stratio n p u b lish ed in
1817. T h ere w as, how ever, so m e co n n e ctio n h isto ri­
cally betw een the ca se against trade d e fic it s an d th e
u n d erstan d in g o f w hy trad e in gen eral w as a good
thing. Only w h en the gains from trad e w ere properly
u n d ersto o d cou ld p eop le begin to m ake sen sib le a s­
sessm en ts o f th e ca u se an d effect o f trad e d eficits.
T h e co n tex t in w h ich th e early d eb ates took p lace
w as an in tern atio n al eco n o m y in w h ich p ay m en ts for
extern al trad e w ere largely m ad e in p recio u s m etals,
esp ecially gold. T h e effect o f ru n n in g a trad e su rplu s
w as th at a n atio n w ould a ccu m u la te gold. In m any

9The U.S. authorities agreed to convert dollars into gold at $35 per
ounce. This commitment was abandoned for all but official holders
in March 1968 and for official holders in August 1971. See Batten
and Ott (1983) for evidence on exchange market intervention.
,0Note that even today the majority of small countries peg their
exchange rates to either a major currency or a weighted basket of
currencies. Reserve shortages still may cause acute problems for
them.
11In fact, none of the major currencies are floating freely. All the major
central banks have intervened from time to time to influence ex­
change rates. Intervention to support the dollar has been especially
heavy since the “ Plaza Accord” of September 1985.

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

p e o p le ’s m ind s, the accu m u latio n o f gold itself b e ­
cam e th e o b je ct o f trad e: trad e su rp lu ses w ere ‘g o o d ”
and trad e d eficits “b a d .” T radin g in o rd er to bu ild up
gold h old ings b eca m e know n as m ercan tilism .

flation .13 In d eed , it w as a perio d o f rapid an d p ro ­
longed e c o n o m ic grow th. T h ere is th u s at least one
co u n te rex a m p le — an d a m a jo r on e — to R ica rd o ’s
gen eralization . How can th is be exp lain ed ?

M ercan tilism w as criticized by several em in en t w rit­
ers, in clu d in g David Hum e (1752), w ho sh ow ed th at a
co n tin u in g trade su rp lu s w as u n attain able. An e x ist­
ing trad e su rplu s, he n oted , p ro d u ces an inflow of
gold. B ecau se gold is a form of m oney, th e qu antity of
m o n ey in th e co u n try rises. This, in turn, p ro d u ces a
rise in p rices, w h ich co n tin u es as long as m o re gold
flows in. As th e co u n try ’s goods b eco m e m ore e x p e n ­
sive relative to th o se p ro d u ced overseas, how ever,
few er will be bought, eventually elim in atin g the trade
su rp lu s.12

At th e tim e R icardo w as w riting, h is claim w as
d isp uted , m ost notably, bv H eniy T h o rn to n . T h o rn to n
argued that, alth ou g h p rio r excessive m o n ey ex p a n ­
sion w as in d eed su fficien t to p ro d u ce a trad e deficit, it
w as n o t a n ecessa ry co n d itio n for a trad e deficit.

Som e y ears later, David R icardo (1817) u sed th is
d em o n stratio n to sh o w w hy trade d eficits o ccu rred .
His an sw er to this q u estio n brings us d irectly to ou r
cen tra l point: trade d eficits can resu lt from a variety of
so u rces, not all o f w h ich are “b a d .”

RICARDO, THORNTON AND TRADE
DEFICITS'*
R icardo argued that a trade d eficit w as the inevita­
ble co n se q u e n ce o f p rices in th e d eficit co u n try bein g
"too h ig h .” T h e se p rices, in tu rn, w ere p ro d u ced by
excessive p rio r m o n etary ex p an sio n from d o m estic
so u rces th at w ere u nrelated to p rio r trad e su rp lu ses.14
He argued, in o th e r w ords, th at excessive m o n etary
ex p an sio n w as n o t only a su fficient co n d itio n for a
trade d eficit to o ccu r, it w as also a n e c e ssa iy co n d itio n
an d vice versa for trade su rp lu ses.
T his d escrib es w hat h ap p e n ed in m any co u n tries
d uring th e B retto n W oods regim e. W hile th is se ­
q u en ce o f events portrays a co m m o n ca u se o f trade
deficits, how ever, it is not th e only cau se. In th e 65y ea r perio d b etw een 1830 and 1895, th e U nited States
h ad a cu rren t a cc o u n t d eficit in alm ost eveiy year;
th ere w ere only 13 y ears in w h ich a su rp lu s was
reco rd ed . Yet this w as not a period o f su stain ed in ­

,2Hume, although dealing explicitly with the mercantilist argument,
dealt implicitly with the notion that an export surplus is necessary for
growth. Since a perpetual export surplus is impossible, if an export
surplus were essential for growth, growth would have stopped. It did
not, however, and to date has not.
,3An extensive discussion of the ground covered in this section can be
found in Perlman (1986).
,4The issue of domestic bank notes partially backed by gold was a
topic of controversy between the “ currency” and “banking” schools
through the 19th century in Britain.



T h o rn to n d istin g u ish ed b etw een trade d eficits a ris­
ing from real ca u ses an d th o se arising from excessive
m o n ey creatio n . T h e form er ca n o c c u r b ec a u se in d i­
viduals in a co u n try w an t to sp en d m ore th a n th e ir
cu rre n t in co m e, that is, th ey w ish to re d u ce th e ir net
fin an cial w ealth or in cre a s e th e ir n et in d e b te d n e ss.16
In term s o f eq u atio n 2 above, an yth in g th at ca u ses
d o m estic sp en d in g to ex ceed o u tp u t w ill p ro d u ce a
trade deficit.
Of co u rse, th e b a la n ce o f p ay m en ts d eficit from this
ca u se ca n n o t p ersist forever. It will d isap p ear w h en
individuals have re a ch ed th e ir n ew lo w er d esired
w ealth level; in th e sam e m an n er, a trade d eficit p ro ­
d u ced by ex cess m o n ey crea tio n w ill en d w h en the
e x c e ss m o n ey h a s b e e n d isp e rse d o v erseas (or
deflated by h ig h er p rice s).17
In sum m ary, a trad e d eficit ca n b e p ro d u ced n ot
ju st by ex cess m o n etary exp an sio n , bu t by dissaving.'8
Both o f th ese will p ro d u ce d eficits that are tem porary;
how ever, th e se d eficits will b e elim in ated eventually
by different m e ch a n ism s. D issaving and th e a sso ci­
ated d eclin e in fin an cial w ealth can be p ro d u ced by
several factors; exam in in g so m e m ajo r o n es h elp s to
u n d ersta n d the cu rre n t U.S. situ ation .

,5For more details on this, see Mudd and Wood (1978) and Friedman
and Schwartz (1963).
"This highlights the fact that a trade deficit can be a symptom of a
problem, but is not itself a problem. Alternatively, it may be a
symptom of something that is not a problem at all.
l7Note that, when we talk about a “ lower desired wealth level," we are
referring only to financial wealth. If financial assets are being con­
verted into physical capital, the composition rather than the level of
wealth is changing. If the physical capital offers a greater rate of
return than financial assets, this change actually will increase peo­
ple’s wealth. This distinction is central to the argument that a trade
deficit associated with high levels of domestic real investment could
lead to faster real growth, increased wealth and higher output in the
future.
18Monetary expansion need not always lead to a trade deficit. In a
classic paper, Robert Mundell (1963) showed that, with perfect
capital mobility, floating exchange rates and sticky goods prices,
monetary expansion causes capital outflows (purchases of foreign
assets). This causes the currency to depreciate and results in a
current account surplus. Similar results are found in the modern
“overshooting” literature.

7

FEDERAL RESERVE BANK OF ST. LOUIS

WHY SHOULD TH ERE BE DISSAVING?
In o rd er to d iscu ss th e p o ssib le so u rces o f dissaving
in th e d o m estic econ om y, it is co n v en ien t to set out
a n o th e r id en tity :1'1
13! CA = ( S - I ) + ( T - G l .
T his show s th at th e cu rren t a cc o u n t su rp lu s m u st be
equal to th e ex cess o f private saving over private in ­
vestm en t (S —I), plus th e govern m ent bu dget su rplu s
(T —G). In o th e r w ords, the su rp lu s for th e eco n o m y as
a w h ole can be broken dow n in to th e private se c to r
su rplu s plus th e p u blic se c to r su rp lu s. T h is cla ssifica ­
tio n suggests p o ssib le d irectio n s in w h ich to look for
ca u ses o f th e trade deficit: a fall in private saving, a rise
in private investm ent o r an in crea se in th e governm ent
budget deficit.
A fall in private saving m u st be a sso ciate d w ith an
in crease in co n su m p tio n relative to in co m e. T his
cou ld h ap p en if th ere w ere a tem porary fall in in co m e
due, for exam ple, to a crop failure o r a natu ral d isaster.
It is w ell estab lish ed that, at tim es w h en in co m e is
abn orm ally low, p eop le attem p t to m aintain th eir c o n ­
su m p tio n p attern s by dissaving. If th e n atio n as a
w h ole d o es this, it will n ecessarily involve a trade
deficit. It sh o u ld b e em p h asized that, w hile crop fail­
ures or o th e r n atu ral d isasters are u nfo rtu n ate, the
ability to a d ju st to th e se events by dissaving an d th u s
im portin g good s from abroad is preferable to red u cin g
d o m estic co n su m p tio n . In extrem e cases, th e ch o ice
m ay be b etw een ru n n in g a trad e d eficit an d starvation.
W hile natu ral d isasters can explain som e trade d e ­
ficits, it is unlikely to explain the U.S. d eficits in the
1980s. After all, this h as b een a period o f fairly steady
in co m e grow th.
T h e seco n d alternative su ggested by id entity (3) is a
rise in private investm ent, cau sed by an ex p ected rise
in th e productivity o f d o m estic cap ital (relative to th at
overseas). T his alternative is an extrem ely h ealth y sign
for th e d o m estic eco n om y. It in d icates th at th e ex ­
p e cted profitability o f in vestm en t w as su ch th at firm s
w ere p rep ared to borrow in o rd er to fin a n ce the
h ig h er investm ent. If private investm en t ex ceed s p ri­
vate saving (for a b a la n ced governm ent budget), the
private se c to r m u st borrow from overseas. W e have
seen alread y th at n et borrow ing from overseas im plies
a cu rren t a cc o u n t d eficit in th e b alan ce o f paym ents.

,9This can be derived as follows: GNP = C + I + G + CA from the
expenditure accounts. It is also true that GNP = C + S + T from the
income accounts. Sol + G + CA = S + T and CA = (S - l) + (T -G ).

http://fraser.stlouisfed.org/
8
Federal Reserve Bank of St. Louis

JANUARY/FEBRUARY 1988

If ov erseas-fin an ced grow th in private in v estm en t
lies b eh in d the trade deficit, w e have to be carefu l in
in terp retin g th e sta tem e n t th at th e trad e d eficit is
a sso cia ted w ith dissaving o r a red u ctio n in w ealth . It is
true that the private se c to r will be in creasin g its net
fin an cial liabilities (or red u cin g n et fin an cial assets). At
th e sam e tim e, how ever, it is converting th o se liabili­
ties into real cap ital. T h e retu rn on that real cap ital is
ex p ected to b e g reater th a n th e co st o f th e borrow ing.
H ence, th is provides th e b asis for in co m e an d w ealth
grow th in th e future and, presum ably, exp lain s w hy
th e U n ited S ta te s h ad s u s ta in e d tra d e d eficits
th rou gh o u t th e seco n d h alf o f th e 19th cen tu ry. R ap­
idly grow ing co u n tries th at attract cap ital from over­
seas typically will have trade deficits.
T h e final p ossibility is th at th e cu rren t a cco u n t
deficit reflects th e governm ent budget deficit. Obvi­
ously, if private saving an d in vestm en t w ere equal, th e
bu dget d eficit an d th e cu rre n t a cc o u n t d eficit w ould
be equal. We shall not p u rsu e th e q u estio n o f w h eth er
th e bu dget d eficit is “g o o d ” o r “b a d ” for th e eco n om y.
A ssum ing th at th e bu dget d eficit re p rese n ts th e d elib ­
erate ch o ice o f policym akers, how ever, it follow s th at
th e a sso cia te d trad e d eficit m u st be preferred to th e
alternatives.
T h u s, w e have seen th at a rising cu rren t a cc o u n t
deficit m u st b e a sso cia te d w ith eith e r a rise in invest­
m en t relative to saving (or fall in saving relative to
investm ent) or a rise in th e budget d eficit o f th e gov­
ern m en t. W e alread y h ad see n th at cu rren t a cc o u n t
d eficits co u ld resu lt from excessive m o n etary ex p a n ­
sion, a ca se th at is co n siste n t w ith id en tity (3): th e
attem p t to sp en d th e ex cess m o n ey will resu lt in
eith e r a fall in S —I (higher co n su m p tio n , low er saving
or h ig h er investm ent) or a fall in T —G (m ore govern­
m ent sp en d in g relative to taxes).

The Evidence f o r the United States
We n ow look at th e p o ssib le ca u ses o f th e U.S.
cu rren t a cc o u n t deficit. First, w e co n sid e r th e argu ­
m ent, favored by Ricardo, o f fast m o n etary grow th
a sso cia te d w ith high d o m estic inflation. At first sight,
th is ap p ears a likely possibility. M on etary grow th a c ­
celera ted after 1982 (chart 1) at th e sam e tim e as th e
cu rren t a cc o u n t p lu n g ed into d eficit (chart 2). How­
ever, U.S. inflation fell (chart 3) an d rem a in ed c o n s is t­
ently below th e OECD average d uring th is p eriod . Also,
b o th the real and effective ex ch a n g e rates a p p reciated
strongly u ntil 1985. T h e in flation an d ex ch a n g e rate
behavior are signs o f m o n etary tigh tn ess, n o t m o n e-

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

C hart 1

U.S. M o n e y G r o w th , Effective a n d Real Exchange Rates
Percent

Index

N O T E : The e f f e c t i v e e x c h a n g e r a t e is th e F e d e r a l R e s e r v e B o a r d o f G o v e r n o r s ’ t r a d e - w e i g h t e d e x c h a n g e r a t e , a w e i g h t e d
i n d e x (1973=100] o f t h e d o l l a r ' s v a l u e in t e r m s o f 10 i n d u s t r i a l c o u n t r y c u r r e n c i e s . The r e a l e f f e c t i v e e x c h a n g e r a t e is
o b ta in e d

b y d i v i d i n g th e n o m i n a l e f f e c t i v e e x c h a n g e r a t e b y the r a t i o o f c o n s u m e r p r i c e i n d e x e s (CPI] o f th e 10

i n d u s t r i a l c o u n t r i e s , ( t r a d e - w e i g h t e d , t h e s a m e a s t h e e x c h a n g e r a te s ] to t h e CPI o f th e U n i t e d S t a t e s ; a l l CPIs a r e
i n d e x e d t o 1 9 7 3 , 1973 =1 00.

taiy ease.2" Only the high U.S. m on ey grow th in 1986
looks co n siste n t w ith R icard o ’s ex p lan ation : b o th the
real an d n o m in al ex ch an g e rates fell during 1986. T h e
in crea se in th e trad e deficit in 1986, how ever, w as
sm all. H ence, little w eight can be a tta ch e d to the
m o n etary exp lan ation o f th e trade deficit. In d eed , w hy
th e rapid m oney grow th o f 1 9 8 2 -8 6 did not crea te

20lt is possible that the Mundell model referred to above is relevant
here. This predicts that monetary tightness causes capital inflows, a
currency appreciation and a current account deficit. We think this
unlikely to be relevant here. There is no clear evidence of sufficient
monetary tightening over the entire 1981-86 period to explain what
happened. More importantly the same outcome is predicted from the
Mundell analysis as resulting from fiscal expansion. Hence mone­
tary neutrality combined with fiscal expansion would be sufficient. It
is the latter which seems to us to dominate in this case.




inflation is still so m eth in g o f a m ystery. T h e re was,
over th is period, a sign ifican t d eclin e in th e velocity of
circu latio n , w h ich m ean s th at th e extra m o n ey b a l­
a n ce s w ere w illingly h eld ra th er th a n sp en t d o m esti­
cally.-'
A m u ch m ore plau sib le sto iy em erg es from a plot o f
th e private an d p u b lic se c to r su rp lu ses (chart 2). No­
tice th at w e sh o w h ere I —S ra th er th a n S —I, b ec a u se it
is ea sier to see its co rre sp o n d e n c e w ith T —G. Before
1982, th e relatio n sh ip b etw een th e p u blic se c to r d e­
ficit and th e private se c to r su rp lu s w as rem arkably
clo se. As a result, cu rren t a cc o u n t d eficits an d su r­
p lu ses generally w ere sm all. After 1982, how ever, the

21See Stone and Thornton (1987).

9

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

C h a rt 2

Relationships B e tw ee n Public a n d Private Sector Deficits a n d the
Current Account
Billions

of

dollars

p u b lic se c to r d eficit stayed high w hile private invest­
m ent rose relative to private saving. Bv 1986, th e p ri­
vate se c to r invested in ex cess o f its saving. H ence, the
co n tin u ed p u blic se c to r d eficit is n ecessarily m atch ed
by a cu rren t a cc o u n t d eficit o f equivalent size. Insofar
as th e governm ent bu dget d eficit is taken as given, the
ch o ice s th at the U.S. faces are high levels o f private
in vestm ent and a trade deficit or b alan ced trade and
slow real grow th.

Billions

of

dollars

general, a trade d eficit m u st b e a sso cia ted w ith som e
co m b in atio n o f private an d p u b lic se c to r d eficits. Until
1982, the budget d eficit w as approxim ately fin an ced
by private se c to r su rp lu ses. T h e p resen t situation,
how ever, is th e inevitable resu lt o f th e co m b in a tio n of
a budget d eficit an d high in v estm en t relative to private
saving.

REFERENCES
CONCLUSION

Balbach, Anatol B. “The Mechanics of Intervention in Exchange
Markets,” this Review (February 1978), pp. 2-7.

A trade deficit arises w h en a co u n try buys m ore
from overseas th an foreigners buy from it. T h e co u n ­
terpart o f a trad e d eficit in the b alan ce o f paym en ts
a cc o u n ts is an in crease in borrow ings (or red u ctio n in
net lending) from th e rest o f th e w orld. T rad e deficits
cou ld resu lt from inflationary d o m estic m o n etary pol­
icies; th ere is no evidence, how ever, th at su ch policies
are th e cau se o f the cu rren t U.S. trade deficit. In

Batten, Dallas S., and Mack Ott. “What Can Central Banks Do
About the Value of the Dollar?” this Review (May 1984), pp. 5-15.


http://fraser.stlouisfed.org/
10
Federal Reserve Bank of St. Louis

Chrystal, Alec K. “A Guide to Foreign Exchange Markets,” this
Review (March 1984), pp. 5-18.
Friedman, Milton, and Anna Jacobson Schwartz. A Monetary His­
tory of the United States 1867-1960 (Princeton University Press,
1963).
Hume, David.

Political Discourses (Edinburgh, 1752).

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

C hart 3

U.S. a n d OECD Inflation

1968

70

72

74

76

78

80

82

84

1986

Mudd, Douglas R., and Geoffrey E. Wood. “The Recent
U.S.
Trade Deficit — No Cause for Panic,” this Review (April 1978), pp.
2-7.

Perlman, Morris. “The Bullionist Controversy Revisited,” Journal of
Political Economy (August 1986), pp. 745-62.

Mundell, Robert. "Capital Mobility and Stabilization Policy Under
Fixed and Flexible Exchange Rates,” Canadian Journal of Eco­
nomics and Political Science (November 1963), pp. 475-85.

Stone, Courtenay C., and Daniel L. Thornton. “Solving the 1980's
Velocity Puzzle: A Progress Report,” this Review (August/Septem­
ber 1987), pp. 5-23.




Ricardo, David. The Principles of Political Economy and Taxation
(London: J. M. Dent & Sons, Ltd., reprinted 1948).

11

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

Protectionist Trade Policies: A
Survey of Theory; Evidence and
Rationale
Cletus C. Coughlin, K. Alec Chrystal and Geoffrey E. W ood

JL

ROTECTIONIST p ressu res have b een m ou n tin g
w orldw ide d uring th e 1980s. T h e se p ressu res are due
to various ec o n o m ic p ro blem s in clu d in g th e large and
p ersisten t b alan ce o f trade d eficits in th e United
States; th e h ard tim es ex p e rie n ce d by several in d u s­
tries; an d th e slow grow th o f m any foreign co u n tries.'
P ro p o n en ts o f p ro tectio n ist trade p o licies argue that
in tern atio n al trade h as co n trib u ted su bstan tially to
th e se p ro blem s an d th at p ro tectio n ist trad e p o licies
will lead to im proved resu lts. P rofessional eco n o m ists
in th e U nited States, how ever, gen erally agree that
trad e re strictio n s su ch as tariffs and qu o tas su b sta n ­
tially re d u ce a n a tio n ’s e co n o m ic w ell-being.2
T h is article surveys th e theory, evid en ce an d ra­
tio n ale co n cern in g p ro tectio n ist trad e p o licies. T h e
first sectio n illu strates th e gains from free trade using
th e co n c e p t o f com p arative advantage. R ecen t devel­
o p m en ts in in tern atio n al trad e th eory th at em p h asize
o th e r reason s for gains from trade are also reviewed.
T h e th eo retical d iscu ssio n is follow ed bv an ex am i­
n ation o f re cen t em p irical stud ies that d em o n strate
th e large co sts o f p ro tectio n ist trade p o licies. T h en,
th e ration ale for restrictin g trade is p resen ted . I’he
co n clu d in g sectio n su m m arizes th e p a p e r’s m ain
argum ents.

Cletus C. Coughlin is a senior economist at the Federal Resen/e Bank
of St. Louis. K. Alec Chrystal is the National Westminster Bank Profes­
sor of Personal Finance at City University, London. Geoffrey E. Wood is
a professor of economics at City University, London. This article ivas
written while Chrystal was a professor of economics at the University of
Sheffield, Sheffield, England. Thomas A. Pollmann provided research
assistance.
'See Page (1987) for a detailed examination of trade protectionism
since 1974.
2 This consensus was found in a survey published in the late 1970s
(Kearl et at., 1979). Recent developments in international trade
theory, which can be used to justify governmental intervention in
trade policy, have not altered the consensus (Krugman, 1987).

http://fraser.stlouisfed.org/
12
Federal Reserve Bank of St. Louis

THE GAINS FROM FR E E TRADE
T h e m o st fam ou s d em o n stra tio n o f th e gains from
trade a p p eared in 1817 in David R icard o ’s P rin c ip le s o f
P o litic a l E c o n o m y a n d T ax a tion . W e u se h is exam ple
involving trade b etw een E n glan d an d Portugal to d em ­
o n strate how b o th co u n tries ca n gain from trade. T h e
two co u n tries p ro d u ce the sam e tw o goods, w ine and
clo th , an d th e only p ro d u ctio n co sts are la b o r co sts.
T h e figures below list th e am o u n t o f la b o r (e.g.,
w orker-days) requ ired in e a c h co u n tiy to p ro d u ce one
bo ttle o f w in e o r o n e b o lt o f clo th .

E ngland
Portugal

W ine

Cloth

3
1

7
5

Sin ce b o th good s are m ore co stly to p ro d u ce in E n ­
gland th a n in Portugal, E n glan d is ab so lu tely less
efficient at p ro d u cin g b o th good s th an its p rospective
tradin g p artn er. Portugal h a s an ab so lu te advantage in
b o th w in e an d clo th . At first g lan ce, th is ap p ears to
rule out m u tu al gains from trade; how ever, as w e
d em o n strate below , ab so lu te advantage is irrelevant in
d iscern in g w h eth er trad e ca n ben efit b o th co u n tries.
T h e ratio o f th e p ro d u ctio n co sts for th e two goods
is different in th e tw o co u n tries. In England, a b o ttle of
w ine will ex ch an g e for 3/7 o f a b o lt o f clo th b e c a u se th e
lab or co n ten t o f th e w in e is 3/7 o f that for clo th . In
Portugal, a b o ttle o f w in e will ex ch a n g e for 1/5 o f a bolt
o f clo th . T h u s, w ine is relatively c h e a p e r in Portugal
th an in England and, conversely, clo th is relatively
c h e a p e r in England th an in Portugal. T h e exam ple
in d icates that Portugal h as a com p arative advantage in
w ine p ro d u ctio n an d E ngland h as a co m p arative ad ­
vantage in clo th p ro d u ctio n .
T h e different relative p rices provide th e b asis for

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

both co u n tries to gain from in tern atio n al trade. T he
gains arise from b o th ex ch an g e an d sp ecializatio n .
T h e gains from e x c h a n g e can be high lighted in the
follow ing m an n er. If a P ortu gu ese w in e p ro d u c e r sells
five b o ttles o f w ine at h om e, h e receives on e b o lt of
clo th . If he trades in England, he receives m ore than
two bolts of clo th . H ence, h e can gain by exp o rtin g his
w ine to England. E nglish clo th -p ro d u cers are w illing
to trade in Portugal; for eveiy 3/7 o f a bolt o f clo th they
sell th ere, they get ju st over two b o ttles o f w ine. T h e
English gain from exportin g clo th to (and im porting
w ine from) Portugal, an d th e P ortu gu ese gain from
exportin g w ine to (and im p ortin g clo th from) England.
E ach co u n try gains bv exp orting th e good in w h ich it
has a com p arative advantage an d by im portin g th e
good in w h ich it h as a com p arative disadvantage.
(rains from s p e c ia liz a tio n can be d em o n strated in
th e follow ing m an n er. Initially, e a c h co u n try is p ro ­
d u cin g so m e o f b o th good s. Su p p o se that, as a resu lt of
trade, 21 u n its o f lab or are sh ifted from w ine to clo th
p ro d u ctio n in England, w hile, in Portugal, 10 u n its of
lab or are shifted from clo th to w ine p ro d u ctio n . T his
reallo catio n o f lab or d oes n o t alter th e to tal am o u n t o f
lab or u sed in th e tw o co u n tries; how ever, it cau ses the
p ro d u ctio n ch an g es listed below .
B ottles o f W ine
England
Portugal
Net

-

7

Bolts o f Cloth
+ 3

+ 10

-

+ 3

+ 1

2

T h e shift o f 21 u n its o f lab or to th e English clo th
in d u stry raises clo th p ro d u ctio n by th ree bolts, w hile
red u cin g w in e p ro d u ctio n by seven bo ttles. In P ortu ­
gal, th e shift o f 10 u nits o f lab or from clo th to w ine
raises w ine p ro d u ctio n by 10 bo ttles, w hile red u cin g
clo th p ro d u ctio n by tw o bo lts. T h is reallo catio n of
lab or in cre a se s th e total p ro d u ctio n o f b o th goods:
w ine by th ree bo ttles and clo th by on e bolt. This
in crea sed o u tp u t will be sh ared by th e tw o co u n tries.
T h u s, th e co n su m p tio n o f b o th good s and th e w ealth
o f both co u n tries are in crea sed by the sp e cia liz a ­
tion brou ght abou t by trade b ased on com p arative
advantage.

lab or w as th e only re so u rce u sed to p ro d u ce th e two
good s in th e exam p le above; y et, la b o r is really only
o n e o f m any re so u rces u sed to p ro d u ce good s. T h e
exam p le also a ssu m ed th at th e co sts o f pro d u cin g
ad d ition al u n its o f th e good s are co n sta n t. F or ex a m ­
ple, in England, th ree u n its o f lab or are u sed to p ro ­
d u ce on e bo ttle o f w ine regard less o f th e level o f w ine
p ro d u ctio n . In reality, u nit p ro d u ctio n co sts cou ld
eith e r in crea se o r d ecrea se as m o re is p ro d u ced . A
third a ssu m p tio n w as th at th e goods are p ro d u ced in
p erfectly com p etitiv e m arkets. In o th e r w ords, an in d i­
vidual firm has no effect on th e p rice o f th e good that it
p ro d u ces. Som e in d u stries, how ever, are d om in ated
by a sm all n u m b er o f firm s, e a c h o f w h ich ca n affect
the m arket p rice o f the good by altering its p ro d u ctio n
d ecisio n . Som e o f th e se ex ten sio n s are d iscu sse d in
th e append ix.
T h e se th e o retica l d ev elo p m en ts g en erally have
stren g th en ed th e ca se for an o p en tradin g system .
T h ey suggest th ree so u rce s o f gains from trade. First,
as th e m arket poten tially served by firm s exp an d s
from a n ation al to a w orld m arket, th ere are gains
a sso cia te d w ith d eclin in g p e r u nit p ro d u ctio n co sts. A
seco n d so u rce o f gains resu lts from th e re d u ctio n in
th e m on o p o ly p o w er o f d o m estic firm s. D om estic
firm s, facing m ore p re ssu re from foreign co m p etito rs,
are forced to p ro d u ce th e o u tp u t d em an d ed by c o n ­
su m ers at th e low est p o ssib le co st.3 T h ird is th e gain to
co n su m ers from in crea sed p ro d u ct variety an d low er
p rices. G enerally speaking, th e gains from trade result
from the in cre a se in com p etitiv e p ressu res as the
d o m estic eco n o m y b e co m e s less in su lated from the
w orld econ om y.
T h e gains from free trad e ca n also b e illu strated
graphically. T h e sh ad ed in sert on pages 14 an d 15
exam in es th e gains from trade in perfectly com p etitive
m arkets u sin g su p p ly an d d em an d analysis. T h e in ­
sert also analyzes th e effects o f trad e restrictio n s, a
to p ic th at w e d iscu ss below .

COSTS OF TRADE PROTECTIONISM
P ro tectio n ist trade p o licies can take n u m ero u s
form s, so m e o f w h ich are d iscu sse d in th e sh ad ed
in sert on pages 16 an d 17. All form s o f p ro tectio n are

TRADE THEORY SINCE RICARDO
Sin ce 1817, n u m ero u s analyses have g en erated in ­
sights co n cern in g th e gains from trade. They chiefly
exam in e th e co n se q u e n ce s o f relaxing th e a ssu m p ­
tio n s u sed in th e p reced in g exam p le. F or exam ple,



3A profit-maximizing firm produces its output at minimum cost. When
firms are insulated from competition, costs are not necessarily being
minimized. This situation, which is called X-efficiency, has been
stressed by Leibenstein (1980). The increase in competitive pres­
sures due to international trade reduces the probability that costs
are not minimized.

13

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

A Supply and Demand Analysis of the Gains from Free
Trade and the Effects of a Tariff
A stand ard illu stration o f th e gains from free
trade an d th e eifects o f a tariff is p resen ted below .
T h e analysis assu m es perfectly com p etitive m ar­
kets th rou gh ou t.

Gains From Free Trade
In figure 1 th e lines Sus and Dus are th e U.S. supply
a n d d em an d curves for a h y p oth etical good. T h e ir
in tersectio n at B resu lts in th e equilibriu m values
for p rice and qu antity o f Pu, and Q us. A ssum ing th e
United States h as a com p arative disadvantage in th e
p ro d u ctio n o f th is good, th e p rice will be low er
abroad th an in th e U nited States. Let this low er
w orld p rice be Pw, and assu m e that U.S. p u rch a ses
do n o t affect th is w orld price. G raphically, this is
re p rese n te d bv th e h o rizo n tal w orld su pply curve
S„ . If o n e allow s for free trade, th is lo w er w orld p rice
h as tw o effects. First, U.S. co n su m ers will in crea se
th e ir co n su m p tio n to D 'us. Second , U.S. p ro d u cers
will co n tra ct th eir p ro d u ctio n to S 'us. T h e e x cess o f
U.S. c o n su m p tio n o v erp ro d u ctio n is U.S. p u rch a ses
from foreign p ro d u cers (that is, im ports).
T h e low er p rice sim u ltan eo u sly ben efits U.S. c o n ­
su m ers an d h arm s U.S. p ro d u cers, a fact that u n ­
d erlies th e re ce n t controversial d iscu ssio n s o f U.S.

F ig u re

1


http://fraser.stlouisfed.org/
14
Federal Reserve Bank of St. Louis

trade policy. T h e m agn itu d e o f th e se gains and
lo sses u sin g the co n c e p ts o f c o n su m e r an d p ro ­
d u ce r su rp lu s ca n b e se e n in figure 1. C on su m ers
gain in two w ays. Initially, co n su m e rs p u rch a sed
Q uk at a p rice p er unit o f Pus. W ith free trade, they
p u rch a se Q us at th e low er p rice p e r unit o f P„. T his
gain is rep rese n te d by th e rectan g le Pus BE P„. In
ad d ition, th e lo w er p rice in d u ce s co n su m e rs to
in crea se th e ir p u rch a se s from Q„s to D 'us. T h is gain
is re p resen ted by th e triangle BCE. T h e total gain to
co n su m ers is Pus BC P„ or, u sin g th e low er case
letters to rep resen t areas, a + b + c. Analogously,
p ro d u cers lo se due to the low er p rice th ey receive
for th e ir o utput, S 'us, an d d ue to th e ir c o n tra ctio n o f
p ro d u ctio n from Q„s to S 'us. T h e total loss to p ro ­
d u cers is Pus BF P,v o r a.
T h e n atio n as a w h o le gains b e c a u se th e c o n ­
su m er gains o f a + b + c ex ceed th e p ro d u cer
lo sses o f a by b + c. T h is analysis ca n also b e viewed
u sing a good th at th e U nited States exp orts. In o th e r
w ords, th e U nited States w ill have a com p arative
advantage in th e p ro d u ctio n o f a good. F o r the
exp ort good, th e ch a n g e to free trade w ill ca u se
p ro d u cer gains th at ex ceed c o n su m e r losses.

The Effects o f a Tariff
T o m ake th e analysis o f p ro tectio n ist trade policy
as straightfoi’w ard as p o ssib le, th e im p act o f a tariff
is analyzed . (One ca n view any p ro tectio n ist trade
policy, how ever, in an an alogous m anner.) For c o n ­
ven ien ce, th e free trad e resu lts in figure 1 are d u p li­
ca ted in figure 2. Given the free trad e w orld p rice o f
Plv, U.S. co n su m p tio n , p ro d u ctio n an d im p orts are
D 'us, S ' _ and S 'us D 'lls. A ssum e a tariff is im posed ,
cau sin g the p rice in th e United States to in crea se to
P ,. T h e p rice in th e U nited States n ow e x ceed s p rice
in th e w orld by th e am ou n t o f th e tariff, PW
PT.
T h e h ig h er U.S. p rice ca u se s c o n su m e r p u r­
ch a se s to d e cre a se from D'„, to D"us, d o m estic p ro ­
d u ctio n to in crea se from S 'us to S"us, an d im p o rts to
d ecrea se from S 'usD 'us to S"„sD"us. By im p osin g the
tariff, co n su m ers lo se th e area PTJCP„. o r d + e + f
+ g an d p ro d u cers gain th e area PTIFP„ o r d. Do-

FEDERAL RESERVE BANK OF ST. LOUIS

F ig u re 2

JANUARY/FEBRUARY 1988

m e stic p ro d u cers are p ro tected at th e ex p e n se of
d o m estic co n su m ers.
O ne co m p lica tio n stem s from tariff revenue. T a r­
iff revenue, w h ich ca n b e view ed as a gain for th e
governm ent, eq u als th e tariff, PW
PT, tim es th e q u a n ­
tity o f im ports, S"usD"us. T h is reven u e is equal to area
IJGH o r f.
Overall, th e n a tio n lo ses b e c a u se th e c o n s u m e rs’
lo sses o fd + e + f + g e x ceed th e p ro d u c e rs’ gains
o f d an d th e g overnm ent gains o f f by e + g. Area e is
called a “d ead w eight p ro d u ctio n lo s s ” an d ca n b e
viewed as a lo ss resu lting from in efficien t (excess)
d o m estic p ro d u ctio n , w hile area g is called a “d ea d ­
w eight co n su m p tio n lo s s ” an d ca n be view ed
as a loss resu ltin g from in efficien t (too little)
co n su m p tio n .

in ten d ed to im prove th e p o sitio n o f d o m estic relative
to foreign p ro d u cers. T h is can be d on e th rou gh po li­
cies that in crea se th e ho m e m arket p rice o f the foreign
prod u ct, d ecrea se th e p ro d u ctio n co sts o f d o m estic
firm s, o r so m eh o w restrict th e a c c e ss o f foreign p ro ­
d u cers to th e d o m estic m arket.
1'he sp ecific goal o f p ro tectio n ist trad e p o licies is to
expand d o m estic p ro d u ctio n in th e p ro tected in d u s­
tries, ben efitin g th e ow ners, w orkers and su p p liers of
re so u rces to the p ro tected in d u stiy. The governm ent
im posing p ro tectio n ist trade p o licies m ay also benefit,
for exam ple, in th e form o f tariff revenue.
T h e ex p an sio n o f d o m estic p ro d u ctio n in p ro tected
in d u stries is not co stle ss; it requ ires ad d ition al re ­
so u rces from o th e r in d u stries. C onsequently, ou tp u t
in o th e r d o m estic in d u stries is red u ced . T h e se in d u s­
tries also m ight be m ad e less com p etitive b ec a u se of
h ig h er p rices for im p o rted inp u ts. Sin ce p ro tectio n ist
trad e p o licies frequ en tly in crea se th e p rice o f th e
p ro tected good, d o m estic co n su m ers are h arm ed.
T h ey lose in two w ays. First, th e ir co n su m p tio n o f th e
p ro tected good is red u ced b eca u se o f th e asso cia ted
rise in its p rice. Seco n d , th ey co n su m e less o f o th e r
goods, as th e ir o u tp u t d eclin es and p rices rise.
T h e p reced in g d iscu ssio n highlights th e d o m estic
w in n ers and lo sers d ue to p ro tectio n ist trad e p o licies.
D om estic p ro d u cers o f th e p ro tected good and the
governm ent (if tariffs are im posed) gain; d o m estic c o n ­
su m ers and o th e r d o m estic p ro d u cers lose. Foreign
in terests are also affected by trad e restrictio n s. T h e
p ro tectio n o f d o m estic p ro d u cers will harm som e
foreign p ro d u cers; oddly enou gh, o th e r foreign p ro ­



d u cers m ay b enefit. For exam ple, if qu otas are p laced
o n im ports, som e foreign p ro d u cers m ay receive
h ig h er p rices for th eir exp o rts to th e p ro tected m arket.
T h ere have b ee n n u m ero u s stu d ies o f the c o sts of
p ro tectio n ism . W e b egin by exam in in g th ree re cen t
stu d ies o f p ro tectio n ism in th e U nited States, th e n
p ro ce e d to stu d ies exam in in g developed and, finally,
d eveloping co u n tries.

Costs o f Protectionism in the United
States
R ecen t stu d ies by T a rr an d M orkre (1984), Hickok
(1985) an d H ufbauer et al. (1986! estim ated the co sts of
p ro tectio n ism in th e U nited States. T h e se stu d ies use
d ifferent estim atio n p ro ced u res, exam in e different
p ro tectio n ist p o licies an d cover d ifferent tim e period s.
N on eth eless, th ey provide co n siste n t results.
T arr an d M orkre (1984) estim ate an n u al co sts to the
U.S. eco n o m y o f $12.7 billion (1983 dollars) from all tar­
iffs an d from qu otas on au tom ob iles, textiles, steel and
sugar. T h e ir co st estim ate is a n et m easu re in w h ich
the lo sses o f co n su m ers are offset partially by th e gains
o f d o m estic p ro d u cers an d th e U.S. governm ent.
E stim ates by Hickok (19851 in d ica te that trad e re ­
strictio n s on only th ree good s — cloth in g, sugar, and
a u to m o b iles — ca u sed in crea sed c o n su m e r e x p e n d i­
tu res o f $14 billion in 1984. Hickok also show s th at lowin co m e fam ilies are affected m o re th an h ig h -in co m e
fam ilies. T h e im port restrain ts on cloth in g, su gar and
a u to m o b iles are ca lcu la ted to b e equivalent to a 23
p e rce n t in co m e tax su rch arg e (that is, an ad d itional

15

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Form s Of Protectionism
P ro tectio n m ay b e im p lem en ted in n u m ero u s
w ays. All form s o f p ro tectio n are in ten d ed to im ­
prove th e p o sitio n o f a d o m estic relative to foreign
p ro d u cer. T h is ca n be d o n e by p o licies th at in ­
cre a se th e h o m e m arket p rice o f th e foreign p ro d ­
uct, d ecrea se th e co sts o f d o m estic p ro d u cers or
re strict th e a c c e ss o f foreign p ro d u cers to th e hom e
m arket in so m e o th e r way.

in crea se in th e h o m e m arket. T h is in d u ce s d o m es­
tic p ro d u cers to in crea se p ro d u ctio n an d c o n ­
su m ers to re d u ce co n su m p tio n . O ne d ifference b e ­
tw een a tariff an d a qu ota is th at th e tariff g en erates
revenue for th e governm ent, w h ile th e qu o ta g en er­
ates a revenue gain to th e o w n er o f im port licen ses.
C on sequen tly, foreign p ro d u cers m ight ca p tu re
som e o f th is revenue.

Tariffs

In re ce n t y ears, a slightly different version of
quotas, called eith e r ord erly m arketing ag reem en ts
o r voluntary exp ort restrain ts, h as b een used. In an
ord erly m arketin g ag reem en t, th e d o m estic govern­
m en t asks th e foreign governm ent to re strict the
qu an tity o f exp orts o f a good to th e d o m estic c o u n ­
try. T h e req u est ca n b e view ed as a d em an d , like the
U .S.-Japan au tom ob ile a g reem en t in th e 1980s, b e ­
ca u se th e d o m estic co u n try m akes it cle a r that
m ore restr ictive a ctio n s are likely u n less th e foreign
govern m en t "volu ntarily" co m p lies. In effect, th e
orderly m arketing a g reem en t is a m u tu ally agreedupon quota.

Tariffs, w h ich are sim ply taxes im p o sed on goods
en terin g a co u n tiy from abroad, resu lt in h ig h er
p rice s and have b ee n th e m o st co m m o n form of
p ro tectio n for d o m estic p ro d u cers. Tariffs have
b ee n p o p u lar w ith govern m ents b ec a u se it ap p ears
th at th e tax is bein g p aid by th e foreig n er w ho
w ish es to sell his good s in th e h o m e eco n o m y and
b ec a u se th e tariff revenue ca n be u sed to fin an ce
governm ent sen d ees o r re d u ce o th e r taxes.
In th e 20th cen tu ry, U.S. tariff rates peaked as a
resu lt o f th e Sm oot-H aw ley T ariff o f 1930. F o r ex am ­
ple, in 1932, tariff revenue as a p ercen tag e o f total
im p orts w as 19.6 p e rce n t. An id en tical calcu la tio n
for 1985 yield s a figure o f 3.8 p ercen t. T h e d eclin e
w as due prim arily to tw o reaso n s. First, sin ce m any
o f th e tariffs u n d e r Sm oot-H aw ley w ere set as sp e ­
cific d ollar am o u n ts, th e rising p rice level in th e
U nited States erod ed th e effective tariff rate. S e c­
ond, sin ce W orld W ar II, n u m ero u s tariff red u ctio n s
have b een n eg o tiated u n d e r th e G eneral A greem ent
on Tariffs an d Trade.
On th e o th e r hand, various o th e r form s o f p ro te c ­
tion, frequ en tly term ed n o n -tariff barriers, have b e ­
co m e in creasin gly im p ortan t. A few o f th e m ore
frequ ently u sed devices are d iscu sse d below .

Quotas
A qu ota seem s like a sen sib le alternative to a tariff
w h en th e in ten tio n is to restrict foreign p ro d u c e rs’
a c c e ss to th e d o m estic m arket. Im p o rters typically
are lim ited to a m axim u m n u m b er o f p ro d u cts that
th ey can sell in th e h o m e m arket over sp ecific
period s. A qu ota, sim ilar to a tariff, ca u se p rices to


http://fraser.stlouisfed.org/
16
Federal Reserve Bank of St. Louis

Regulatory Barriers
T h ere are m any o th e r w ays o f restrictin g for­
eign ers' a c c e ss to d o m estic m arkets. M un ger (1983)
h as n o ted th at th e tariff co d e itself te n d s to limit
trade. T h e 1983 T a r i f f S c h e d u le s o f t h e U n ited S ta te s
A n n o ta te d co n sists o f 792 pages, plus a 78-page
appen d ix. Over 200 tariff rates pertain to w a tch es
an d clo ck s. Sim ply a scerta in in g th e ap p rop riate
tariff classificatio n , w h ich requ ires legal a ssista n ce
an d ca n b e su b je ct to d ifferen ces o f o p in ion , is a
d eterren t.
P rod u ct sta n d a rd s are a n o th e r co m m o n regu la­
tory b arrier. T h e se stan d ard s ap p ear in various
form s an d are u sed for m an y p u rp o ses. T h e sta n d ­
ards ca n b e u sed to serve th e p u b lic in terest by
en su rin g that im p orted food p ro d u cts are p ro­
ce ssed acco rd in g to a ccep ta b le san itary stan d ard s
an d th at drugs have b ee n scre e n e d b efo re th e ir
in tro d u ctio n in th e U nited States. In o th e r ca ses,
the stan d ard s, so m etim es in ten tion ally, p ro tect d o ­
m estic p ro d u cers. An exam p le o f u n in te n d ed re-

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

strictio n s m ay be th e im p o sitio n o f safety o r p o llu ­
tion stan d ard s th at w ere not previously bein g m et
by foreign cars.

Exchange Controls

An alternative to restrictin g th e term s u n d er
w h ich foreigners can co m p ete in th e h o m e m arket
is to su bsid ize d o m estic p ro d u cers. Su bsid ies m ay
be focu sed u p o n an in d u stry in g en eral o r u p o n th e
exp ort activities o f th e in d u stiy . An exam p le o f th e
form er, d iscu sse d by M orici an d M egna (1983), is
th e co m b in atio n o f cred it program s, sp ecial tax
in cen tiv es and d irect su bsid y p ay m en ts th at benefit
th e U.S. ship bu ild ing ind u stry. An exam p le o f th e
la tter is th e fin ancial a ssista n ce to in cre a se exp orts
provided by th e U.S. E xp o rt-Im p o rt Bank throu gh
d irect loans, lo an g u aran tees an d in su ran ce, an d
d isco u n t lo an s. In eith e r case, p ro d u ctio n will
expand .

All o f th e above relate d irectly to th e flow of
g ood s. A final cla ss o f re strictio n s w orks by re strict­
ing a c c e ss to th e foreign m on ey requ ired to buy
foreign good s. F o r exam ple, a govern m en t that
w ish e d to p ro te c t its ex p o rtin g an d im p o rtco m p etin g in d u stries m ay try to h o ld its ex ch an g e
rate artificially low . As a result, foreign goods w ould
a p p ea r expensive in th e h o m e m arket w hile h o m e
good s w ould b e ch e a p overseas. H om e p ro d u cers
im plicitly are su b sid ized an d h o m e co n su m ers im ­
p licitly are taxed. T h is p o licy is norm ally h ard to
su stain . T h e ce n tra l bank, in h old ing th e ex ch a n g e
rate dow n h as to bu y foreign ex ch a n g e w ith d o m es­
tic cu rren cy . T h is new ly issu ed d o m estic cu rre n cy
in crea ses th e d o m estic m o n ey stock an d eventually
ca u ses inflation. In flation ary p o licies are n ot n o r­
m ally regard ed as a sen sib le w ay o f p ro tectin g d o ­
m e stic ind ustry.

An im p o rtan t d ifferen ce b etw een su bsid ies and
tariffs involves th e revenu e im p licatio n s for govern­
m ent. T h e form er involves th e govern m ent in p ay ­
ing out m oney, w h ereas tariffs g en erate in co m e for
th e governm ent. T h e effect on d o m estic p ro d u ctio n
an d w elfare, how ever, ca n be th e sam e u n d e r su b si­
dies as u n d e r tariffs an d qu o tas. In all cases, th e
p ro tected in d u stiy is bein g su bsid ized by th e rest o f
th e eco n om y.

T h e re is a n o th e r a sp e ct to ex ch a n g e co n tro ls.
T h e ju stifica tio n is th at preven tin g h o m e resid en ts
from investing overseas b en efits d o m estic grow th
as it lead s to g reater d o m estic real in vestm en t. In
reality, it co u ld do exactly th e o p p o site. R estrictin g
a c c e ss to foreign a sse ts m ay raise th e v arian ce and
lo w er th e retu rn to o w n ers o f d o m estic w ealth . In
th e sh o rt run, it also m ay a p p recia te th e d o m estic
ex ch a n g e rate and, thereby, m ake d o m estic p ro ­
d u cers less com p etitive.

Subsidies

tax ad d ed to th e n orm al in co m e tax) for fam ilies w ith
in co m es less th an $10,000 in 1984 and a 3 p ercen t
in co m e tax su rch arg e for fam ilies w ith in co m es ex ­
ceed in g $60,000.
H ufbauer et al. (1986) ex am in ed 31 ca se s in w h ich
trade volum es ex ceed e d $100 m illion an d the United
States im p o sed p ro tectio n ist trade re strictio n s.4 They
g en erated estim ates o f th e w elfare co n se q u e n ce s for
ea ch m a jo r group affected (see table 1). T h e figures in
th e table in d icate th at an n u al c o n su m e r lo sses ex ceed
$100 m illion in all bu t six o f th e cases. T h e largest
losses, $27 billion p e r year, co m e from p ro tectin g the
textiles and ap p arel in d u stiy . T h e re also are large

“While there were cases in which the industry adjusted to its new
competitive position and the protection was terminated, these cases
were more the exception than the rule. In far more cases, protection­
ist policies were maintained indefinitely or removed because of
favorable demand changes.



co n su m e r lo sses a sso cia te d w ith p ro tectio n in carb o n
steel ($6.8 billion), a u to m o b iles ($5.8 billion) an d d aily
p ro d u cts ($5.5 billion).
T h e p u rp o se o f p ro tectio n ism is to p ro tect jo b s in
sp ecific in d u stries. A u seful a p p ro a ch to gain som e
pei"spective on c o n su m e r lo sses is to exp ress th ese
lo sses on a per-job-saved basis. In 18 o f th e 31 cases,
th e co st p er-job-sav ed is $100,000 or m ore p e r y ear; the
co n su m e r lo sses p er-jo b -sav ed in b en z e n o id c h e m i­
cals, ca rb o n steel (two sep a ra te periods), sp ecialty
steel, an d bolts, n u ts an d screw s ex c e e d e d $500,000
p e r year.
Table 1 also reveals th at d o m estic p ro d u cers w ere
th e p rim a iy b en eficiaries o f p ro tectio n ist p o licies;
how ever, th e re are so m e n o tew o rth y ca se s w h ere for­
eign p ro d u cers realized relatively large gains. F o r the
U .S.-Japan ese voluntary exp o rt ag reem en t in a u to m o ­
biles, foreign p ro d u cers gain ed 38 p e rce n t o f w hat
d o m estic co n su m ers lost, w hile a sim ilar co m p u tatio n

17

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Table 1
Distribution of Costs and Benefits from Special Protection
Consumer Losses

Case
Manufacturing
Book manufacturing
Benzenoid chemicals
Glassware
Rubber footwear
Ceramic articles
Ceramic tiles
Orange juice
Canned tuna
Textiles and apparel: Phase I
Textiles and apparel: Phase II
Textiles and apparel: Phase III
Carbon steel: Phase I
Carbon steel: Phase II
Carbon steel: Phase III
Ball bearings
Specialty steel
Nonrubber footwear
Color televisions
CB radios
Bolts, nuts, large screws
Prepared mushrooms
Automobiles
Motorcycles
Services
Maritime industries
Agriculture and fisheries
Sugar

Totals
(million
dollars)
$

500
2,650

270,000

2,000

neg.

102

1,000

930

550

410

5

130

5,000

250

34

1,370

170
1,600

neg.
135

9
44

14
145

560

60,000
690/acre
220,000
1,800/cow
1,000/acre
160,000
225/head
21,000

200

170

177

15

6,900
67

160,000
30,000

2,0003
4

70
11

3,000
5

170
1,800

305
2,250

neg.
neg.

130
90
25
62
390
74
8,700
18,000
22,000
1,330
2,770
3,800
21
420
250
190
14
60
13
2,600
67

neg.
neg.
neg.
neg.
neg.
$
7
neg.
350
1,800
330
930
2,000
neg.
50
220
140
neg.
neg.
neg.
2,200
neg.

54
139
69
55
128
10
1,158
2,143
2,535
290
556
560
18
32
262
77
32
16
25
790
21

13
33
6
11
130
4
1,100
3,100
4,850
50
120
330
neg.
30
16
7
5
1
0.8
200
17

4,800
46

$

0
252

Efficiency
loss
(million
dollars)

3,000

Peanuts
Meat

$

Tariff
revenue
(million
dollars)

200
230
95
116
525
91
9,400
20,000
27,000
1,970
4,350
6,800
45
520
700
420
55
110
35
5,800
104

5,500

Mining
Petroleum
Lead and zinc

Totals
(million
dollars)

Welfare Costs of Restraints
Gain to
foreigners
(million
dollars)

$ 100,000
over 1
million
200,000
30,000
47,500
135,000
240,000
76,000
22,000
37,000
42,000
240,000
620,000
750,000
90,000
1,000,000
55,000
420,000
93,000
550,000
117,000
105,000
150,000

Dairy products

Fish

Per job
saved1
(dollars)

Producer
Gains

$

29
14

Neg. = negligible.
'Unless otherwise specified, figures are per worker.
Estimated duties collected on ship repairs performed abroad.
3ln this case, because of the way the quotas were allocated, the gains to importers accrued to domestic refiners rather than foreign
exporters.
SOURCE: Trade Protection in the United States: 31 Case Studies.


http://fraser.stlouisfed.org/
18
Federal Reserve Bank of St. Louis

FEDERAL RESERVE BANK OF ST. LOUIS

for the latest p h ase o f p ro tectio n for carb o n steel w as
29 p ercen t.
Finally, table 1 in d icates that the efficien cy losses
are sm all in co m p ariso n to the total lo sses b o rn e bv
co n su m ers. T h e se efficien cy lo sses, w h ich are defined
p recisely and illu strated in th e first sh ad ed insert,
result from the e x cess d o m estic p ro d u ctio n and the
red u ctio n in co n su m p tio n cau sed by p ro tectio n ist
trad e p o licies. In large ca se s su ch as textiles and
apparel, p etroleum , d aiiy p ro d u cts an d the m aritim e
industr ies, th ese lo sses equal o r exceed $1 billion. It is
likely th at th e se estim ates u n d erstate th e actu al co sts
b ec a u se th ey do n o t cap tu re th e seco n d ary effects th at
o c c u r as p ro d u ctio n and co n su m p tio n ch an g es in one
ind u stry affect o th e r in d u stries.’ In ad d ition, re s tric­
tive trad e p o licies g en erate ad d ition al co sts b e c a u se of
b u rea u cratic en fo rce m en t co sts and efforts by th e
private se c to r to in flu en ce th e se p o licies for th e ir ow n
gain as well as sim ply co m p ly w ith adm inistrative
regu lations.

Costs o f Protectionism Throughout the
World
In 1982, th e O rg a n iz a tio n for E c o n o m ic C o ­
o p eratio n and D evelopm ent (OECD) beg an a p ro je c t to
analyze th e co sts an d b en efits o f p ro tectio n ist p o licies
in m an u factu rin g in OECD co u n tries. T h e OECD
(1985) highlighted a n u m b er o f w ays that p ro tectio n ist
p o licies have g en erated co sts far in e x cess o f benefits.
Sin ce p ro tectio n ist p o licies in crea se p rices, th e report
co n clu d e s th at th e attain m en t o f su stain ed n o n in flatio n ary grow th is h in d e re d by su ch p ricein creasin g effects. M oreover, e co n o m ic grow th is p o ­
ten tially re d u ce d if th e u n certain ty create d by varying
trade p o licies d ep resse s investm ent.
W ood an d M udd (1978), and m any o th ers, have
sh ow n th at im p orts do n ot ca u se h ig h er u n em p lo y ­
m ent. Conversely, th e OECD stud y stresse s th e fact
that a red u ctio n in im p orts via trade re strictio n s d oes
not ca u se g reater em p loy m en t. A red u ctio n in th e
value o f im p o rts resu lts in a sim ilar re d u ctio n in th e

5Recent estimates of the costs of protectionist policies using general
equilibrium models suggest that the secondary effects, to the limited
extent they are measurable, are substantial. For example, Grais, de
Melo and Urata (1986) estimate that the elimination of quotas in
Turkey in 1978 would have caused a 5.4 percent rise in gross
domestic product, while Clarete and Whalley (1985) estimate that
the elimination of tariffs, quotas and export taxes in the Philippines in
1978 would have caused a 5.2 percent rise in gross national
product.




JANUARY/FEBRUARY 1988

value o f exports. O ne ratio n ale for th is finding is th at a
re d u ctio n in th e p u rch a se s o f foreign good s red u ces
foreign in co m es and, in turn, ca u se s red u ced foreign
p u rch a se s o f d o m estic goods.
W hile th e re d u ctio n in im p o rts in crea ses em p lo y ­
m ent in in d u stries th at p ro d u ce p ro d u cts sim ilar to
the previously im p o rted goods, th e red u ctio n in ex ­
p orts d ecrea se s em p lo y m en t in th e export in d u stries.
In o th e r w ords, w h ile so m e jo b s are saved, o th ers are
lost; how ever, th is e c o n o m ic reality m ay not be obvi­
ous to b u sin essm en , lab or u nion lead ers, p o litician s
an d o th ers. Luttrell (1978) has stressed that th e jo b s
saved by p ro tectio n ist legislation are m o re readily
observed th an th e jo b s lo st d ue to p ro tectio n ist legis­
lation. In o th e r w ords, th e jo b s that are p ro tected in,
say, th e textiles in d u stry by U.S. im p o rt re strictio n s on
foreign textiles are m o re readily a p p aren t (and p u b li­
cized) th an the jo b s in agricu ltu re an d high te c h n o l­
ogy in d u stries th at do not m aterialize b ec a u se o f the
im port restrictio n s. T h e se em p lo y m en t effects w ill net
to approxim ately zero.6
T h e OECD stud y also stresse s that d eveloping co u n ­
tries n ee d exp orts to offset th e ir d ebts. T h u s, p ro te c ­
tio n ist trad e p o licies by d eveloped co u n tries affect n ot
only th e e c o n o m ic activity o f th e developing c o u n ­
tries, but th e stability o f th e in tern atio n al financial
system as d eb to r n a tio n s find it in creasin gly difficult
to service th e ir d ebts.
Not only d o es a free trade policy by developed
co u n tries ben efit developing co u n tries, bu t a free
trade policy by d eveloping co u n tries b en efits d evelop­
ing co u n tries. A re ce n t W orld Bank stud y (1987) o f 41
developing co u n tries co m p a red th e p erfo rm an ce of
co u n tries follow ing a free trad e p o licy w ith co u n tries
follow ing a re stricted trad e p olicy.7 Table 2 lists the

6Recent evidence shows that protectionist legislation actually may
reduce employment. Denzau (1987) estimated that 35,600 manu­
facturing jobs were lost as a result of the September 1984 voluntary
export restraints that limited the level of U.S. steel imports. Despite
an increased employment for producers of steel (14,000) and pro­
ducers of inputs for steel producers (2,800), these increases were
more than offset by the 52,400 job losses by steel-using firms.
These losses are due to the higher steel prices that cause steelusing firms to be less competitive in export markets and subject
them to more foreign competition in the U.S. market.
The World Bank study divides trade strategies into two groups:
outward oriented and inward oriented. An outward-oriented strat­
egy, which we call a free trade policy, is one in which trade and
industrial policies do not discriminate between production for the
domestic market and exports, nor between purchases of domestic
and foreign goods. An inward-oriented strategy, which we call a
restricted trade policy, is one in which trade and industrial policies
are biased toward production for the domestic market relative to the
export market.

19

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Table 2
Annual Average Growth of Per Capita Real Gross National Product
Restricted Trade

Free Trade
Strongly

Period

Moderately

1963-73

Singapore
South Korea
Hong Kong

9.0%
7.1
6.0

Brazil
Israel
Thailand
Indonesia
Costa Rica
Malaysia
Ivory Coast
Colombia
Guatemala
Cameroon

1973-85

Singapore
Hong Kong
South Korea

6.5
6.3
5.4

Malaysia
Thailand
Tunisia
Brazil
T urkey
Israel
Uruguay
Chile

Moderately

Strongly

5.5%.
5.4
4.9
4.6
3.9
3.8
3.5
3.3
2.7
-0.1

Yugoslavia
Mexico
Nigeria
Tunisia
Kenya
Philippines
Bolivia
Honduras
El Salvador
Madagascar
Nicaragua
Senegal

4.9%
4.3
4.2
4.0
3.9
2.2
2.0
1.9
1.4
1.1
1.1
0.6

Turkey
Dominican Republic
Burundi
Argentina
Pakistan
Tanzania
Sri Lanka
Ethiopia
Chile
Peru
Uruguay
Zambia
India
Ghana
Bangladesh
Sudan

3.5%
3.4
3.2
3.1
3.1
2.7
2.3
1.9
1.7
1.5
1.5
1.2
1.1
0.4
-1 .4
-1 .9

4.1
3.8
2.9
1.5
1.4
0.4
0.4
0.1

Cameroon
Indonesia
Sri Lanka
Pakistan
Yugoslavia
Colombia
Mexico
Philippines
Kenya
Honduras
Senegal
Costa Rica
Guatemala
Ivory Coast
El Salvador
Nicaragua

5.6
4.0
3.3
3.1
2.7
1.8
1.3
1.1
0.3
0.1
0.8
1.0
1.0
1.2
3.5
3.9

Bangladesh
India
Burundi
Dominican Republic
Ethiopia
Sudan
Peru
Tanzania
Argentina
Zambia
Nigeria
Bolivia
Ghana
Madagascar

2.0
2.0
1.2
0.5
-0 .4
-0 .4
-1.1
-1 .6
-2 .0
-2 .3
-2 .5
-3.1
-3 .2
-3 .4

SOURCE: World Development Report 1987 and The Economist (1987).

an n u al average grow th in real p e r cap ita gross n a ­
tional p ro d u ct for e a ch o f th e 41 co u n tries for 1 9 6 3 -7 3
an d 1 9 7 3 -8 5 . T h o se co u n tries th at did n o t bias in d u s­
trial p ro d u ctio n tow ard th e d o m estic m arket by trade
restrictio n s grew at faster rates th a n th o se th at did.
F or exam ple, th e average an n u al grow th rate in real
p e r cap ita in co m e for 1 9 6 3 -7 3 w as 6.9 p e rce n t in the
eco n o m ies strongly o rien ted to free trade an d 1.6
p e rce n t in th e eco n o m ies strongly o rie n te d to re ­
stricted trade. F o r 1 9 7 3 -8 5 th e se grow th rates w ere 5.9
p e rce n t and —0.1 p ercen t, respectively.

ARGUMENTS FOR RESTRICTING
TRADE

T h e study p ro ceed s to identify th e m a cro e co n o m ic
reason for th e gen eral finding. A given am o u n t o f new

If p ro tectio n ism is so costly, w hy is p ro tectio n ism so
pervasive? T h is sectio n reviews th e m a jo r argu m en ts


http://fraser.stlouisfed.org/
20
Federal Reserve Bank of St. Louis

in vestm en t g en erated m o re ad d ition al o u tp u t in
co u n tries follow ing a free trad e p o licy th a n a re­
stricted trade p olicy. T h e rea so n is th at a free trade
en v iron m en t allow s cap ital to flow to its m o st highly
valued u ses, w hile a restricted trad e en v iron m en t d is­
torts ec o n o m ic in cen tives.

FEDERAL RESERVE BANK OF ST. LOUIS

for restrictin g trad e and provides ex p lan atio n s for the
ex iste n ce o f p ro tectio n ist trad e p o licies.

National Defense
T h e n atio n al d efen se arg u m ent savs that im port
barriers are n ecessary to en su re th e cap acity to p ro ­
d u ce cru cial good s in a n atio n al em erg en cy . W hile this
argum ent is esp ecially ap p ealin g for w eap o n s d uring
a war, th ere will likely be d em an d s from o th e r in d u s­
tries that d eem th em selves essen tial. F or exam ple, the
footw ear in d u stry will d em an d p ro tectio n b eca u se
m ilitary p e rso n n el n ee d co m b at boots."
T h e natio n al d efen se argum en t ignores th e p o ssib il­
ity o f p u rch a se s from friendly co u n tries d uring the
em erg en cy . T h e p o ssib ilities o f storage and d ep letio n
raise ad d itional d oubts abou t th e gen eral ap p licability
o f th e argum ent. If cru cial good s can b e stored , for
exam ple, the least co stly w ay to p rep are for an em er­
g en cy m ight be to buy th e goods from foreigners at the
low w orld p rice before an em erg en cy and store them .
If th e cru cia l good s are d ep letab le m ineral resou rces,
su ch as oil, th e n the restrictio n o f oil im p orts before an
em erg en cy will ca u se a m o re rapid d ep letio n o f d o ­
m e stic reserves. O n ce again, stockp iling m ight be a far
less co stly alternative.

In com e Redistribution
Sin ce p ro tectio n ist trade p o licies affect th e d istrib u ­
tion o f in co m e, a trade restrictio n m ight be d efen d ed
on th e g roun d s that it favors som e disadvantaged
group. It is unlikely, how ever, th at trade policy is the
best tool for dealing w ith th e perceived evils o f in co m e
inequality, b e c a u se o f its b lu n tn e ss an d adverse effects
on th e efficient allo catio n o f re so u rces. A ttem pting to
eq u alize in co m es d irectly by tax an d tran sfer pay­
m en ts is likely less co stly th an u sin g trad e policy. In
ad d ition, as H ickok’s (1985) stud y in d icates, trad e re ­
strictio n s on m an y item s in crease rath er th an d e­
crea se in co m e inequality.

Optimum Tariff Argument
T h e op tim u m tariff arg u m en t ap p lies to situ atio n s
in w h ich a co u n try h as th e ec o n o m ic pow er to alter
w orld p rices. T h is p ow er exists b ec a u se th e co u n try
(or a group o f co u n tries actin g in co n so rt like th e
O rganization o f P etroleu m Exp ortin g C ountries) is

8See Pine (1984).




JANUARY/FEBRUARY 1988

su ch a large p ro d u cer or c o n su m e r o f a good that a
ch an g e in its p ro d u ctio n o r co n su m p tio n p attern s
in flu en ces w orld p rices. F or exam ple, by im p osin g a
tariff, th e co u n try ca n m ake foreign good s ch ea p er.
Sin ce a tariff re d u ce s th e d em an d for foreign goods, if
the tariff-im p osing co u n try has som e m arket pow er,
the w orld p rice for th e good will fall." T h e tariffim p osin g co u n try w ill gain b e c a u se th e p rice p e r unit
o f its im p orts will have d ecreased .
T h e re are a n u m b er o f o b sta cle s that p reclu d e the
w id esp read ap p lica tio n o f th is argum en t. Few c o u n ­
tries p o sse ss th e n ecessa ry m arket p o w er and, w h en
th ey do, only a sm all n u m b er o f goods is covered.
Second ly, in a w orld o f shifting supply an d dem and ,
calcu latin g th e op tim u m tariff an d a d ju stin g th e rate
to ch an g in g situ atio n s is difficult. Finally, th e p o ssib il­
ity o f foreign retaliatio n to an a ct o f ec o n o m ic w arfare
is likely. Su ch retaliatio n cou ld leave b o th co u n tries
w orse off th an th ey w ould have b een in a free trade
environm ent.

Balancing the Balance o f Trade
M any co u n tries en a ct p ro tectio n ist trad e p o licies in
th e h o p e o f elim in atin g a b a la n ce o f trad e d eficit or
in creasin g a b a la n ce o f trad e su rplu s. T h e d esire to
in crea se a b a la n ce o f trad e su rp lu s follow s from the
m ercan tilist view th at larger trade su rp lu ses are b e n e ­
ficial from a n ation al p erspective.
T h is arg u m en t is su sp ect on a n u m b er o f g round s.
First, th ere is n o th in g in h eren tly u n d esirab le about a
trade d eficit o r d esirab le about a surplus.'" For ex a m ­
ple, faster ec o n o m ic grow th in th e United States th an
in th e rest o f th e w orld w ou ld ten d to ca u se a trade
deficit. In th is ca se, th e trade d eficit is a sign o f a
h ealth y eco n om y. Seco n d , p ro tectio n ist p o licies that
red u ce im ports will ca u se exp orts to d ecrea se by a
co m p arab le am o u n t. H ence, an attem p t to in crea se
exp o rts p erm a n en tly relative to im p o rts will fail. It is
doubtful th at th e trad e d eficit will b e red u ced even
tem porarily b ec a u se im port q u an tities do n ot d eclin e
quickly in re sp o n se to th e h ig h er im port p rices and
th e reven u es o f foreign p ro d u cers m ight rise.

9lf a country such as the United States has no market power, the
world price is fixed. Consequently, the price faced by U.S. con­
sumers and producers rises by the full amount of the tariff. In the
optimum tariff case, the price faced by U.S. consumers and pro­
ducers rises, but not by the full amount of the tariff. This must be the
case because the world price falls and the amount of the tariff is the
difference between the world price and the U.S. price.
,0See Chrystal and Wood (1988) earlier in this issue.

21

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

Protection o f J o b s

—

Public Choice

T h e p ro tectio n o f jo b s arg u m ent is clo sely related to
th e b alan ce o f trade argum ent. Sin ce a red u ctio n in
im p orts via trade re strictio n s will resu lt in a sim ilar
re d u ctio n in exports, th e overall em p lo y m en t effects,
as fou nd in the OECD (1985) study and m any oth ers,
are negligible. W hile th e o v e r a ll effects are negligible,
w orkers (and re so u rce ow ners) in sp ecific in d u stries
are affected differently.
A d o m estic in d u stry faced w ith in crea sed im ports
from its foreign co m p etitio n is u n d e r p ressu re to re ­
d u ce p ro d u ctio n and low er co sts. Productive re­
so u rces m u st m ove from th is in d u stry to o th e r d o m es­
tic in d u stries. W orkers m u st ch an g e jo b s and, in som e
cases, relo cate to o th e r cities. Sin ce th is ch an g e is
forced u p o n th e se w orkers, th e se w orkers b e a r real
co sts th at they are likely to r esist. A sim ilar sta tem en t
ca n be m ad e abou t th e ow ners o f cap ital in th e af­
fected industry.
W orkers an d o th e r re so u rce ow ners w ill likely resist
th e se ch an g es by lobbying for trade restrictio n s. T h e
pr eviously cited stu d ies on th e co sts o f p ro tectio n ism
d em o n strated th at trad e restrictio n s entail su bstan tial
real co sts as w ell. T h e se co sts likely e x ceed the a d ju st­
m ent co sts b eca u se th e a d ju stm en t co sts are o n e-tim e
co sts, w hile th e co sts o f p ro tectio n ism co n tin u e as
long as trad e re strictio n s are m aintain ed .
An obvious q u estio n is w hy p o litician s su pply the
p ro tectio n ist legislatio n d em an d ed by w orkers and
o th e r re so u rce ow ners. A b ra n ch o f ec o n o m ics called
p u blic ch o ice , w h ich focu ses on th e in terp lay betw een
individual p re fe ren ces an d p o litical o u tco m es, p ro ­
vides an answ er. T h e p u blic c h o ice literatu re views the
p o litician as an individual w ho offers voters a b u n dle
o f governm entally su p p lied good s in o rd er to w in
e le c tio n s ." M any argue th at p o litician s gain by provid­
ing p ro tectio n ist legislation. Even thou gh th e nation al
eco n o m ic co sts ex ceed th e ben efits, th e p o litician
faces different co sts and benefits.
T h o se h arm ed by a p ro tectio n ist trad e p o licy for a
d o m estic industry, esp ecially h o u se h o ld co n su m ers,
will incur- a sm all individual co st that is difficult to
identify. F o r exam ple, a co n su m e r is unlikely to p o n ­
d er h ow m u ch extra a sh irt co sts b e c a u se o f p ro te c­
tio n ist legislation for th e textiles an d ap p arel industry.

"The role of pressure groups, acting in their economic self-interest,
has been stressed by Stigler (1971) and Peltzman (1976). For
references, as well as an example of an international trade study
focused on the interaction of politicians and interest groups, see
Coughlin (1985).


22


Even thou gh th e aggregate effect is large, th e harm to
ea ch co n su m e r m a y b e sm all. T h is sm all co st, o f w h ich
an individual m ay n o t even b e aw are, an d th e co sts of
organizing co n su m ers d eter th e form ation o f a lobby
against th e legislation.
On th e o th e r h an d , w orkers an d o th e r re so u rce
ow ners are very co n c e rn e d abou t p ro tectio n ist legisla­
tio n for th e ir ind ustry. T h e ir ben efits ten d to be large
individually an d easy to identify. T h e ir voting and
cam paign co n trib u tio n s assist p o litician s w ho su p ­
port th e ir p o sitio n s an d p en alize th o se w ho do not.
T h u s, p o litician s are likely to resp o n d to th e ir d e ­
m an d s for p ro tectio n ist legislation.

Infant Industries
T h e p reced in g arg u m en t is c o u ch e d in term s of
p ro tectin g a d o m estic ind ustry. A slightly different
argu m en t, th e so -ca lled in fan t in d u stry ca se, is
co u ch e d in term s o f p r o m o t in g a d o m estic ind ustry.
Su p p ose an industry, alread y esta b lish ed in o th er
co u n tries, is b ein g esta b lish ed in a sp ecific cou n try.
T h e co u n try m ight n o t b e able to realize its co m p a ra ­
tive advantage in th is in d u stry b ec a u se o f th e existing
co st an d o th e r advantages o f foreign firm s. Initially,
ow ners o f th e fledgling firm m u st b e w illing to suffer
lo sses until the firm develops its m arket and low ers its
p ro d u ctio n co sts to th e level o f its for eign rivals. In
order' to assist th is en tran t, tariff p ro tectio n can be
used to sh ield th e firm from so m e for eign co m p etitio n .
After th is tem p o rary p eriod o f p ro tectio n , free trade
sh ould be resto red ; how ever, the rem oval o f tariff
p ro tectio n frequ ently is resisted . As th e in d u stry d e ­
velops, its p o litical p o w er to th w art o p p o sin g legisla­
tion also in creases.
A n o th er p ro b lem w ith th e in fan t in d u stry argum en t
is th at a tariff is n ot th e b est w ay to in terven e. A
p ro d u ctio n su bsid y is su p erio r to a tariff if the goal is
to exp an d p ro d u ctio n . A su bsid y will do th is directly,
w hile a tariff h as th e u n d esirab le side effect o f re d u c­
ing co n su m p tio n .
In m any cases, in terven tio n m ight not be ap p ro p ri­
ate at all. If th e infant in d u stry is a good can d id a te for
bein g com p etitive in tern ation ally, b orrow ing from the

,2Special interests benefiting from trade will likely resist the forces for
protectionist legislation. Destler and Odell (1987) identify exporters,
industrial import users, retailers of imported products, businesses
providing trade-related services, foreign exporters, and foreign gov­
ernments as interest groups capable of exerting some anti­
protection pressure. Decisions about protectionist legislation result
from the interaction of both pro-protection and anti-protection
forces.

FEDERAL RESERVE BANK OF ST. LOUIS

private cap ital m arkets can fin an ce th e exp an sio n .
Investors are w illing to abso rb lo sses t e m p o r a r ily if th e
p ro sp ects for future profits are su fficiently good.

Spillover Effects
T h e ju stificatio n for p ro tectin g an ind ustry, infant or
otherw ise, frequ en tly en tails a su ggestion th at the
in d u stiy g en erates spillover ben efits for o th e r in ­
d u stries o r individuals for w h ich th e in d u stiy is not
co m p en sated . D esp ite p aten t law s, on e co m m o n su g ­
gestio n is th at certain in d u stries are n o t fully co m p e n ­
sated for th e ir re sea rch and developm ent ex p en d i­
tures. T h is argum en t is frequ en tly d irected tow ard
tech n o log ically progressive in d u stries w here som e
firm s can cap tu re th e resu lts o f o th e r firm s’ re sea rch
an d d evelopm ent sim ply bv d ism antling a p ro d u ct to
see how it w orks.
T h e ap p licatio n o f this argum ent, how ever, en g en ­
ders a n u m b er o f problem s. Spillovers of know ledge
are difficult to m easu re. Sin ce spillovers are not m ar­
ket tra n sactio n s, th ey do n o t leave an obvious trail to
identify th e ir b en eficiaries. T h e lack o f m arket tra n sa c ­
tion s also co m p lica te s an a sse ssm en t o f th e value of
th ese spillovers. To d eterm in e th e ap p rop riate su b ­
sidy, o n e m ust be able to p lace a d ollar value on the
spillovers g en erated by a given re se a rc h an d develop­
m en t exp en d itu re. Actually, th e calcu latio n requires
m u ch m ore th an th e alread y difficult task o f re c o n ­
stru ctin g th e past, ft requ ires co m p lex estim ates o f the
spillovers' future w orth as w ell. Sin ce re so u rces are
moved from o th e r in d u stries to th e targeted ind u stiy,
th e governm ent m u st u n d e rstan d th e fu n ctio n in g of
the en tire econ om y.
Finally, th ere are political p ro blem s. An aggressive
ap p licatio n of th is argum en t m ight lead to retaliation
and a m utually d estructive trade w ar. In ad dition, as
in terest groups co m p ete for th e g overnm en tal assis­
tan ce, th ere is no gu aran tee th at th e right groups
will b e assisted o r th at they will u se th e assista n ce
efficiently.

Strategic Trade Policy
Recent th eo retical d evelop m ents have id entified
ca ses in w h ich so -called strategic trade policy is su p e ­
rior to free trade. As w e d iscu sse d earlier, d ecreasin g
unit p ro d u ctio n co sts and m arket stru ctu res th at c o n ­
tain m on o p o ly elem e n ts are co m m o n in in d u stries
involved in in tern atio n al trade. M arket im p erfectio n s
im m ed iately suggest the p o ten tial ben efits o f govern­
m ental intervention. In th e strategic trad e p o licy argu­
m ent, governm ent p o licy can alter th e term s o f c o m ­



JANUARY/FEBRUARY 1988

p etitio n to favor d o m estic over foreign firm s an d shift
the ex cess retu rn s in m o n o p o listic m arkets from for­
eign to d o m estic firm s.
Krugm an (1987) illu strates an exam p le o f th e argu ­
m en t. A ssu m e th a t th ere is only o n e firm in th e United
States, Boeing, an d o n e m u ltin atio n al firm in Europe,
Airbus, cap ab le o f p ro d u cin g a 150-seat p a sse n g e r
aircraft. A ssum e also th at the aircraft is p ro d u ced only
for export, so th at th e retu rn s to th e firm ca n be
id entified w ith th e n atio n al in terest. T his exp o rt m ar­
ket is p rofitable for e ith e r firm if it is th e only p ro d u cer;
how ever, it is u n p ro fitab le for bo th firm s to p ro d u ce
th e plan e. Finally, a ssu m e th e follow ing payoffs are
a sso cia ted w ith the four co m b in a tio n s o f p ro d u ctio n :
1) if b o th B oeing an d Airbus p ro d u ce th e aircraft, ea ch
firm lo ses $5 m illion; 2) if n eith e r B oeing n o r Airbus
p ro d u ces th e aircraft, profits are zero; 3) if B oeing
p ro d u ces th e aircraft an d A irbus d oes not, B oeing
profits by $100 m illion an d Airbus h as zero profits; and
4) if Airbus p ro d u ces th e aircraft an d B oeing d oes not,
Airbus profits by $100 m illion an d B oeing has zero
profits.
W h ich firm(s) w ill p ro d u ce th e aircraft? T h e ex am ­
ple d oes n ot yield a u n iq u e o u tco m e. A u niqu e o u t­
co m e can be g en erated if on e firm, say Boeing, has a
head start an d beg in s p ro d u ctio n before Airbus. In
th is case, B oeing will reap profits o f $100 m illion and
will have d eterred Airbus from en terin g th e m arket
b eca u se Airbus will lose $5 m illion if it en ters after
Boeing.
Strategic trade policy, how ever, suggests th at ju d i­
cio u s g overn m en tal in terven tion ca n alter th e o u t­
co m e. If th e E u ro p ea n g overn m en ts agree to su bsid ize
A irbus’ p ro d u ctio n w ith $10 m illion no m a tter w hat
Boeing does, th e n Airbus w ill p ro d u ce th e p lan e. Pro­
d u ctio n by Airbus will yield m o re profits th an not
prod u cin g , no m a tter w h at B oeing does. At th e sam e
tim e, B oeing w ill b e d eterred from p ro d u cin g b eca u se
it w ould lo se m oney. T h u s, Airbus w ill ca p tu re th e
en tire m arket an d reap profits o f $110 m illion, $100
m illion o f w h ich ca n b e view ed as a tran sfer o f profits
from the U nited States.
T h e criticism s o f a strategic trad e policy are sim ilar
to th e criticism s against p ro tectin g a tech n o log ically
progressive in d u stiy th at g en era tes spillover b en e fits.13
T h ere are m a jo r in form ation al p ro b lem s in applying a

13A recent volume edited by Paul Krugman (1986) examines the
jDolicy implications of the new trade literature. See Grossman's
article in that volume for a discussion of the information require­
ments.

23

FEDERAL RESERVE BANK OF ST. LOUIS

strategic trade policy. T h e govern m ent m u st estim ate
th e poten tial payoff o f ea ch co u rse o f actio n . E c o ­
n o m ic know ledge abou t th e behavior o f in d u stries
th at have m o n o p o ly elem en ts is lim ited . Firm s m ay
behave com p etitively or cooperatively and m ay co m ­
p ete by setting p rices or output. T h e behavior o f rival
governm ents also m u st be an ticip ated . Foreign retalia­
tio n m u st be view ed as likely w h ere su b stan tial profits
are at stake. In ad dition, m any in tere st grou p s will
co m p ete for th e governm ental assistan ce . Though
only a sm all n u m b er o f secto rs can be co n sid ered
poten tially strategic, m any in d u stries w ill m ake a case
for assistan ce.

Reciprocity and the “Level Playing Field”
Bhagw ati and Irwin (1987) n o te that U.S. trade policy
d iscu ssio n s in re ce n t y ears have frequ ently stressed
th e im p o rtan ce o f “fair trad e .” T h e co n c e p t o f fair
trade, w h ich is te ch n ica lly referred to as recip rocity ,
m eans different things to different people.
U nder th e G eneral A greem ent on Tariffs an d Trade,
n eg o tiation s to re d u ce trad e b arriers fo cu s u p on
m a tch in g c o n c e s s io n s . T h is form o f recip ro city ,
k now n as first-d ifference recip rocity , attem p ts to re­
d u ce trad e b arriers by requiring a co u n try to provide a
tariff re d u ctio n o f value co m p arable to on e provided
by th e o th e r cou ntry. In this case, recip ro city is
defined in term s o f m atch in g ch an g es.
R ecen t U.S. d em an d s, exem plified by th e G ephardt
am en d m en t to th e cu rren t trade legislation, reveal an
a p p ro ach that is called full recip rocity . T h is ap p roach
seeks recip ro city in term s o f th e level o f p ro tectio n
b ilaterally an d over a sp ecific range o f good s. R eci­
p ro city requires equal a c c e ss and this a c c e ss can be
d eterm in ed bv bilateral trad e b alan ces. A trade d eficit
w ith a trading p a rtn er is claim ed to b e p r im a f a c i e
ev id ence o f u neq u al a cc e ss. E xam ples abou n d . For
exam ple, U.S. co n stru ctio n firm s have not had a m ajo r
co n tra ct in Ja p a n sin ce 1965, w hile Ja p a n e se c o n ­
stru ctio n firm s did $1.8 b illion w o rth o f b u sin ess in th e
United States in 1985 alo n e. R ecen t legislation bars
Ja p a n e se p articip atio n in U.S. p u blic w orks p ro jects
until the Ja p a n e se offer recip ro cal privileges.
As th e n am e suggests, th e fu n d am en tal argum en t
for fair trade is o n e o f equity. D om estic p ro d u cers in a
free trade co u n try argue that foreign trade b arriers are
u nfair b e c a u se it p laces th em at a com p etitive d isad ­
vantage. In an extrem e version, it is asse rte d th at th is
unfair co m p etitio n will virtually elim in ate U.S. m a n u ­
facturing, leaving only jo b s th at co n sist prim arily o f
flipping ham bu rgers at fast food restau ran ts or, as
Bhagw ati an d Irw in have said, rolling rice cak es at
Ja p a n ese -o w n ed su sh i bars. W hile d o m estic pro­

http://fraser.stlouisfed.org/
24
Federal Reserve Bank of St. Louis

JANUARY/FEBRUARY 1988

d u cers a r e relatively disadvantaged, th e w isd om o f a
p ro tectio n ist re sp o n se is d oubtful. Again, th e co sts of
p ro tectio n ism ex ceed su b stan tially the b en efits from a
n atio n al p erspective.
In an attem p t to rein force th e arg u m en t for fair
trade, p ro p o n en ts also argue th at retaliatory threats,
co m b in ed w ith ch a n g es in tariffs an d n o n -tariff b a rri­
ers, allow for th e sim u ltan eo u s p ro tectio n o f d o m estic
in d u stries against u n eq u al co m p etitio n an d in d u ce
m ore o p en foreign m arkets. T h is m ore flexible a p ­
p ro a ch is view ed as su p erio r to a “o n e -sid e d ” free
trad e policy. T h e su ggestion th a t a fair trad e policy
p ro d u ces a trading en viron m en t w ith few er trad e re ­
strictio n s allow s p ro p o n e n ts to a ssert th at su ch a
policy serves to p ro m o te b o th equity an d efficiency. In
o th e r w ords, n o t only w ill d o m estic an d foreign p ro ­
d u cers in th e sam e in d u stry b e treated equally, b u t the
gains a sso cia ted w ith a freer trading en v iron m en t will
b e realized.
On th e o th e r hand , critics o f a fair trade policy argue
that su ch a p o licy is sim ply d isguised pr o te ctio n ism
— it sim ply achieves th e goals o f sp ecific in terest
groups at th e ex p en se o f th e n a tio n at large. In m any
cases, fair trad ers focu s on a sp ecific p ra ctice th at can
be portrayed as p ro tectio n ist w hile ignoring th e en tire
package o f p o licies th at are affecting a n a tio n ’s c o m ­
petitive p o sitio n . In th e se ca ses, th e foreign co u n try is
m ore likely eith e r n o t to resp o n d o r retaliate by in ­
creasin g ra th er th an red u cin g th eir trade barriers. In
th e la tte r case, th e esca la tio n o f trad e b arriers ca u ses
lo sses for b o th n atio n s, w h ich is exactly o p p o site to
th e alleged effects o f an activist fair trade policy.
C ritics o f fair trad e p ro p o sals are esp ecially b o th ­
ered by th e u se o f bilateral trad e d eficits as evid en ce of
u nfair trade. In a w orld o f m any tradin g co u n tries, the
tr ade b etw een two co u n tries n eed n ot b e b a la n ced for
th e trad e o f e a c h to b e in global b a la n ce. Differing
d em an d s an d prod uctive cap ab ilities a cro ss co u n tries
will ca u se a sp ecific co u n try to have trad e d eficits w ith
som e co u n tries an d su rp lu ses w ith o th e r co u n tries.
T h e se b ilateral im b a la n ce s are a n orm al resu lt of
co u n tries trading on th e b a sis o f com p arative advan­
tage.'4 T h u s, th e focu s on th e b ilateral trad e d eficit can
pro d u ce in ap p ro p riate co n clu sio n s abou t fairness
and, m o re im portantly, p o licies a ttem p tin g to elim i­
n ate bilateral trad e d eficits are likely to b e very co stly
b eca u se they elim in ate th e gains from a m u ltilateral
trading system .

,4Bergsten and Cline (1985) estimate an equilibrium U.S.-Japanese
bilateral trade deficit of $20-$25 billion annually.

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

CONCLUSION
T h e p roliferation o f p ro tectio n ist trade p o licies in
re cen t y ears provides an im p etu s to re co n sid e r th eir
w orth. In th e w orld of trad itional trade theory, c h a r a c ­
terized bv p erfect co m p etitio n , a definitive re co m ­
m en d atio n in favor o f free trad e ca n be m ade. T h e
gains from in tern atio n al trad e resu lt from a re a llo ca ­
tion o f productive re so u rces tow ard goods th at ca n be
p ro d u ced less co stly at ho m e th an abroad and the
ex ch an g e o f som e o f th e se good s for good s th at can be
p ro d u ced at less co st abroad th a n at h om e.
R ecen t d evelop m ents in in tern atio n al trad e theory
have exam in ed th e co n se q u e n ce s o f in tern atio n al
trade in m arkets w here th ere are m arket im p erfec­
tions, su ch as m on o p o ly an d tech n o lo g ical spillovers.
Do th e se im p erfectio n s ju stify p ro tectio n ist trad e p o li­
cies? T h e an sw er co n tin u es to be no. W hile p ro te c­
tio n ist trad e p o licies m ay offset m on o p o ly p o w er over­
seas o r ad vantageou sly u se d o m estic m o n o p o ly
pow er, trad e re strictio n s ten d to re d u ce th e co m p e ti­
tion faced by d o m estic p rod u cers, p ro tectin g d o m es­
tic p ro d u cers at th e ex p e n se o f d o m estic co n su m ers.
T h e em p irical evid ence is clea r-cu t. T h e co sts of
p ro tectio n ist trade p o licies far ex ceed th e benefits.
T h e lo sses suffered by co n su m ers ex ceed th e gains
reaped by d o m estic p ro d u cers an d governm ent. Lowin co m e co n su m ers are relatively m o re adversely af­
fected th an h ig h -in co m e co n su m ers. Not only are
th ere in efficien cies a sso ciate d w ith excessive d o m es­
tic p ro d u ctio n and re stricted co n su m p tio n , but th ere
are co sts a sso ciate d w ith th e en fo rce m en t o f th e p ro ­
tectio n ist legislation and attem p ts to in flu en ce trade
policy.
T h e prim ary reaso n for th e se co stly p ro tectio n ist
p o licies relies on a p u b lic c h o ice argu m ent. T h e desire
to in flu en ce trad e policy arises from th e fact that trade
policy ch an g es ben efit som e groups, w hile harm in g
o th ers. C on su m ers are h arm ed by p ro tectio n ist legis­
latio n ; how ever, ig norance, sm all individual co sts, an d
th e high co sts o f organizing co n su m ers prevent the
co n su m ers from bein g an effective force. On th e o th e r
hand , w orkers and o th e r re so u rce ow n ers in an in d u s­
try are m o re likely to be effective p o litically b e c a u se of
th e ir relative ease o f organizing an d th e ir individually
large an d easy -to-id entify ben efits. P olitician s in ter­
ested in re -e le ctio n will m ost likely resp o n d to the
d em an d s for p ro tectio n ist legislation o f su ch an in ter­
est group.
T h e em p irical evid en ce also su ggests that th e a d ­
verse c o n su m e r effects o f p ro tectio n ist trad e p o licies
are n o t short-lived. T h e se p o licies g en erate lo w er e c o ­
n o m ic grow th rates th an th e rates a sso ciate d w ith free
trade p o licies. In turn, slow grow th co n trib u te s to
ad d ition al p ro tectio n ist p ressu res.



In terest g roup p re ssu res from in d u stries e x p e rie n c­
ing difficulty an d th e g en eral ap p eal o f a "level playing
field " co m b in e to m ake th e re d u ctio n o f trade barriers
esp ecially difficult at th e p resen t tim e in th e United
States. N o n eth eless, n atio n al in tere sts will be served
b est by su ch an ad m itted ly difficult p o litical co u rse. In
light o f th e cu rre n t Uruguay R ou nd n eg o tiation s u n ­
d er th e G eneral A greem ent on Tariffs an d Trad e, as
well as n u m ero u s b ilateral d iscu ssio n s, this fact is
esp ecially tim ely.

REFERENCES
Bergsten, C. Fred, and William R. Cline. The United States-Japan
Economic Problem (Institute for International Economics, October
1985).
Bhagwati, Jagdish N., and Douglas A. Irwin. “The Return of the
Reciprocitarians — U.S. Trade Policy Today,” World Economy
(June 1987), pp. 109-30.
Brander, James A. “ Intra-Industry Trade in Identical Commodities,”
Journal of International Economics (February 1981), pp. 1-14.
Brander, James A., and Paul R. Krugman. “A Reciprocal Dumping'
Model of International Trade,” Journal of International Economics
(November 1983), pp. 313-21.
Chrystal, K. Alec, and Geoffrey E. Wood. “Are Trade Deficits a
Problem?" this Review (January/February 1988), pp. 5-13.
Clarete, Ramon L., and John Whalley. “ Interactions Between Trade
Policies and Domestic Distortions,” Center for the Study of Inter­
national Economic Relations Working Paper 8522C (London, On­
tario: University of Western Ontario, 1985).
Coughlin, Cletus C. “ Domestic Content Legislation: House Voting
and the Economic Theory of Regulation,” Economic Inquiry (July
1985), pp. 437-48.
Denzau, Arthur T. “ How Import Restraints Reduce Employment,”
Washington University Center for the Study of American Business,
Formal Publication #80 (June 1987).
Destler, I. M., and John S. Odell. Anti-Protection: Changing Forces
in United States Trade Politics (Institute for International Eco­
nomics, 1987).
Dixit, Avinash K., and Victor D. Norman. Theory of International
Trade (Cambridge University Press, 1980).
Dixit, Avinash K., and Joseph E. Stiglitz. “ Monopolistic Competition
and Optimum Product Diversity," American Economic Review
(June 1977), pp. 297-308.
The Economist.

July 4, 1987, p.74.

Grais, Wafik, Jaime de Melo, and Shujiro Urata. “A General Equilib­
rium Estimation of the Effects of Reductions in Tariffs and Quanti­
tative Restrictions in Turkey in 1978,” in T. N. Srinivasan and John
Whalley, eds. General Equilibrium Trade Policy Modeling (MIT
Press, 1986).
Grossman, Gene M. “ Strategic Export Promotion: A Critique," Stra­
tegic Trade Policy and the New International Economics in Paul R.
Krugman, ed. (MIT Press, 1986), pp. 47-68.
Helpman, Elhanan. “ International Trade in the Presence of Product
Differentiation, Economies of Scale and Monopolistic Competi­
tion,” Journal of International Economics (August 1981), pp. 30540.
Hickok, Susan. “The Consumer Cost of U.S. Trade Restraints,”
Federal Reserve Bank of New York Quarterly Review (Summer
1985), pp. 1-12.

25

FEDERAL RESERVE BANK OF ST. LOUIS
Hufbauer, Gary Clyde, Diane T. Berliner, and Kimberly Ann Elliott.
Trade Protection in the United States: 31 Case Studies (Institute for
International Economics, 1986).
Kearl, James R., Clayne L. Pope, Gordon C. Whiting, and Larry T.
Wimmer. “A Confusion of Economists?" American Economic
Review, Papers and Proceedings (May 1979), pp. 28-37.
Kierzkowski, Henry K. “ Recent Advances in International Trade
Theory: A Selective Survey,” Oxford Review of Economic Policy
(Spring 1987), pp. 1-19.
Krugman, Paul R. “ Is Free Trade Passe?” Journal of Economic
Perspectives (Fall 1987), pp. 131-44.
_________Strategic Trade Policy and the New International Eco­
nomics (MIT Press, 1986).
Lancaster, Kelvin. “ Intra-Industry Trade Under Perfect Monopolistic
Competition,” Journal of International Economics (May 1980), pp.
151-75.
________ Variety, Equity, and Efficiency (Columbia University
Press, 1979).
Leibenstein, Harvey.
Press, 1980).

Beyond Economic Man (Harvard University

Luttrell, Clifton B. “ Imports and Jobs — The Observed and the
Unobserved,” this Review (June 1978), pp. 2-10.
Mill, John Stuart. Principles of Political Economy, W. J. Ashley, ed.
(Longman, 1909).
Morici, Peter, and Laura L. Megna. U.S. Economic Policies Affecting
Industrial Trade: A Quantitative Assessment (National Planning
Association, 1983).
Munger, Michael C. “The Costs of Protectionism: Estimates of the
Hidden Tax of Trade Restraint,” Washington University Center for


26


JANUARY/FEBRUARY 1988
the Study of American Business, Working Paper #80 (July 1983).
Organization for Economic Co-Operation and Development
(OECD). Costs and Benefits of Protection (1985).
Page, Sheila. "The Rise in Protection Since 1974,” Oxford Review
of Economic Policy (Spring 1987), pp. 37-51.
Peltzman, Sam. “Toward a More General Theory of Regulation,”
Journal of Law and Economics (August 1976), pp. 211—40.
Pine, Art. "Footwear Industry Tells Congress Shoe Gap’ Threatens
U.S. Defense," Wall Street Journal, August 24, 1984.
Ricardo, David. The Principles of Political Economy and Taxation
(Penguin, 1971).
Stigler, George J. "The Theory of Economic Regulation," Bell Jour­
nal of Economics and Management Science (Spring 1971), pp. 3 21 .

Stolper, Wolfgang, and Paul A. Samuelson. “ Protection and Real
Wages,” Review of Economic Studies (November 1941), pp. 5873.
Tarr, David G., and Morris E. Morkre. Aggregate Costs to the United
States of Tariffs and Quotas on Imports: General Tariff Cuts and
Removal of Quotas on Automobiles, Steel, Sugar, and Textiles,
Bureau of Economics Staff Report to the Federal Trade Commis­
sion (December 1984).
Venables, Anthony, and Alasdair Smith. “Trade and Industrial Pol­
icy Under Imperfect Competition,” Economic Policy (October
1986), pp. 621-72.
Wood, Geoffrey E., and Douglas R. Mudd. “The Recent U.S. Trade
Deficit — No Cause for Panic,” this Review (April 1978), pp. 2-7.
World Bank. World Development Report 1987 (Oxford University
Press, 1987).

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Appendix
Developments in International Trade Theory and the
Gains from Trade
Sin ce 1817, n u m ero u s developm ents have taken
p la ce in in tern atio n al trad e theory. T h e c o n se q u e n ce s
o f m o re th an on e facto r o f pro d u ctio n , in creasin g and
d ecreasin g u n it p ro d u ctio n co sts, an d im perfectly
com p etitive m arkets are exam in ed in th is ap p en d ix.
Sp ecial atten tio n is focu sed on d evelop m ents in in ter­
n atio n al trade th e o iy in th e last d ecad e.

Increasing the N u m b er o f Factors o f
Production
A ssum e that, in th e U nited States, two reso u rces,
la b o r an d cap ital (e.g., m achin es), are u sed in the
p ro d u ctio n o f two goods, au to m o biles and airplan es.
T h e p rices o f th e se re so u rces will be affected differ­
en tly by trade. As trad e develops, d em an d for the
ex p o rted good (that is, th e good in w h ich th e United
States has a com p arative advantage) will in cre a se and
d em an d for U.S. p ro d u ctio n o f th e im p o rted good will
fall. T h is d em an d shift ca u ses th e p rice o f th e ex ­
p o rted good to rise relative to th e p rice o f th e im ­
p orted good. Sim ilarly, th e shift m ay also p ro d u ce
ch a n g es in th e p rices o f re so u rces; how ever, th ese
p rice ch an g es are n o t alw ays obvious.
Initially, assu m e that the re so u rces ca n n o t b e tra n s­
ferred a cro ss in d u stries. F o r exam ple, th e lab o r and
cap ital u sed to p ro d u ce au to m o biles, th e good im ­
p o rted in to th e U nited States, ca n n o t be u sed to p ro ­
d u ce airp lanes, the ex p o rted good. C on sequently, as
th e p rice o f airp lan es rises in th e U nited States, th e
co m p en satio n for lab o r an d cap ital in th e airplane
in d u stry w ill rise; m eanw hile, th e d eclin e in au to m o ­
bile p rices ca u ses a d eclin e in co m p en sa tio n for lab or
an d cap ital in th e in d u stiy. It w ould n ot be su rprisin g
if lab or an d ow n ers o f cap ital in th e in d u stry w ould
resist su ch ch an g es by asking for trad e p ro tectio n .



W hile re so u rces m ay n o t be easily tran sferred
a cro ss in d u stries in th e sh o rt run, w orkers ca n ch an g e
jo b s an d cap ital ca n b e m oved as tim e p asses. If
re so u rces are m obile, th e n th e lo n g er-ru n c o n s e ­
q u en ces for la b o r an d o w n ers o f cap ital are different
from th o se d escrib ed above. Even if lab or an d cap ital
are p erfectly m obile, how ever, on e set o f re so u rce
ow ners m ay ben efit w hile a n o th e r group is h arm ed by
tra d e .1
T h e real w orld is m ore co m p lica ted th an this d is­
cu ssio n h as allow ed. T h ere are m o re th an tw o factors
o f p ro d u ctio n an d varying d egrees o f m obility for
th e se factors. For exam ple, th e U.S. lab or force c o n ­
tain s scien tists an d en g in eers as w ell as sh o rt-o rd er
cooks. N on eth eless, th e u nd erlyin g analysis d o es sug­
g est so m e g en eralizatio n s. W h en trad e o ccu rs, ow ners
o f th e re so u rces th at are m ore sp ecialized in the
p ro d u ctio n o f export goods will ten d to b eco m e w ea l­

'Who wins and who loses? It depends on the U.S. endowment of
capital to labor relative to other countries. If the United States has
relatively larger amounts of capital to labor relative to other countries,
then owners of capital would benefit, while labor would be harmed.
This result follows from the Stolper-Samuelson Theorem (Stolper
and Samuelson, 1941). In the example, the United States is defined
to be capital-abundant. The example also implicitly assumes that
airplanes are produced by capital-intensive methods and automo­
biles by labor-intensive methods. Thus, the production of airplanes
requires the use of more capital relative to labor than automobiles.
Since the United States is relatively well-endowed with capital and
the production of airplanes is capital intensive, the United States will
have a comparative advantage in the production of airplanes. With
the elimination of trade barriers, the relative price of airplanes to
automobiles will increase. The Stolper-Samuelson Theorem shows
that an increase in the relative price of the capital intensive good will
increase the return to capital relative to the prices of both goods and
reduce the return to labor relative to the prices of both goods.

27

FEDERAL RESERVE BANK OF ST. LOUIS

th ier; th o se w ho ow n re so u rces m ore sp ecialized in
th e p ro d u ctio n o f im p o rt-co m p etin g goods will tend
to lose w ealth . People also gain or lose, how ever,
d ep en d in g o n w hat h ap p en s to the p rices o f th e goods
they buy. Individuals w ho chiefly co n su m e im ported
goods w ill ben efit, w hile th o se w ho p refer co n su m in g
th e exp o rted goods will lose. T hu s, th e net effect on
any individual d ep en d s on b o th th e gains or lo sses
a sso ciate d w ith th e p rice ch an g es on th e good s th at
h e co n su m es and th e effect o f trade on his w ealth (or
incom e).

Increasing Unit Production Costs
A se c o n d assu m p tio n u nd erlying th e R icardian e x ­
am ple o f th e gains from trad e is th at u nit p ro d u ctio n
co sts are co n stan t. If u nit p ro d u ctio n co sts rise as
m ore is p ro d u ced , how ever, th e gen eral co n clu sio n s
abou t th e gains from trad e rem ain essen tially u n ­
ch an g ed . T h e m a jo r difference is th at rising u nit p ro ­
d u ctio n co sts lim it th e ex ten t to w h ich sp ecializatio n
o ccu rs.

Decreasing Unit Production Costs and
Imperfect Competition
On th e o th e r hand , if u n it p ro d u ctio n co sts d e­
crease as p ro d u ctio n in creases, th e ex ten t to w h ich
a ctu al trad e p attern s ca n b e ex p lain ed by com parative
advantage b e c o m e s u n clear. It also fo rces trad e th eo ry
to deal w ith n u m ero u s ch a ra cte ristics o f in tern atio n al
trad e in th e real w orld. T h e m arket stru ctu re o f in d u s­
tries engaged in trad e is frequ en tly highly c o n c e n ­
trated. In o th e r w ords, th e individual firm s in an
industry, co n trary to th o se in a p erfectly com p etitive
industry, ca n affect th e m arket p rice o f th e ir good by
th e ir p ro d u ctio n and advertising d ecisio n s. In ad d i­
tion, trad e statistics sh ow th a t in tra-in d u stry trade
(i.e., th e sim u ltan eo u s exp ort an d im p o rt o f th e o u tp u t
o f th e sam e industry) a cc o u n ts for in creasin gly larger
sh ares o f w orld trade.
In th e last d ecad e, trad e th eo rists have developed
n u m ero u s m o d els to deal w ith th e se facts. An ex h a u s­
tive review o f th is rapidly exp an d in g literatu re is b e ­
y ond th e sco p e of th is append ix; how ever, a few illu s­
trative articles are d iscu sse d in o rd er to estab lish
som e key p o in ts. B ran d er (1981) and B ran d er an d
K rugm an (1983) developed m od els u sing a h o m o g en e­
o us good to highlight ho w im p erfect co m p etitio n ca n
ca u se in tra-in d u stry trad e an d h ow in tra-in d u stry
trade can arise in th e a b se n ce o f co st d ifferences.
A ssum e tw o co u n tries w ith o n e firm in e a c h co u n ­
try. T h e firm s are p ro d u cin g a h o m o g en eo u s good
u n d e r id en tical co st situ atio n s an d th ere are n o tra n s­

http://fraser.stlouisfed.org/
28
Federal Reserve Bank of St. Louis

JANUARY/FEBRUARY 1988

portation co sts. E a ch firm o p erates u n d e r w h at is
term ed a “C ou rn ot c o n je c tu re ,” m ean in g th at ea ch
firm a ssu m es its p ro d u ctio n d ecisio n w ill n o t affect its
rival's p ro d u ctio n d ecisio n . B efore in tern atio n al trade,
ea ch firm h as a m o n o p o ly p o sitio n in its h o m e m arket.
Allowing for free trad e in d u ce s e a c h firm to e n te r th e
o th e r firm ’s m arket, b eca u se p rice ex ceed s m arginal
co st in e a c h co u n try . T h u s, th e sam e good w ill flow to
an d from e a c h cou n try.
Kierzkow ski (1987) h as n o ted th at th e bulk o f in tra ­
in d u stry trad e involves d ifferentiated ra th er th a n h o ­
m o g e n eo u s g oo d s. Tw o a p p ro a c h e s , L a n c a s te r’s
(1979) ch a ra cte ristics a p p ro a ch an d Dixit an d Stiglitz’s
(1977) “love o f variety” ap p roach , have provided the
fou n d ation for trad e m o d els involving d ifferentiated
goods.
In th e ch a ra cte ristic s ap p roach , individuals have
p re fe ren ces for th e ch a ra cte ristic s o f good s rath er
th a n for co lle ctio n s o f th e good s th em selv es. A group
of goods is defin ed as good s p o ssessin g th e sam e
ch a ra cte ristics b u t in different p ro p o rtio n s. A diversity
in c o n su m e r p re fe ren ces ca u ses different co n su m ers
to p refer d ifferent p ro d u cts (i.e., varieties) o f a g roup of
goods.
H elpm an (1981) an d L a n ca ster (1980) u sed th e c h a r­
a cteristics a p p ro a ch to sh o w h ow in tra-in d u stry trade
resu lts from co m b in in g th e d em a n d for variety w ith
eco n o m ies o f scale. T h e ch a n g e from a u ta rch y to free
trad e enlarges th e m arket an d ca u ses o u tp u t o f the
existing varieties to in cre a s e an d th e p ro d u ctio n of
new varieties to begin. C on su m ers gain from th e p ro ­
d u ctio n o f m ore varieties an d low er p rices as e c o n o ­
m ies o f sca le are realized .
T h e so u rces o f g ains from trad e are id en tical using
th e love-of-variety an d ch a ra cte ristic s ap p ro a ch es. In
th e love-of-variety ap p ro a ch , w h ich is u sed by Dixit
an d N orm an (1980), co n su m ers have id en tical tastes
an d p refer to co n su m e as m any types o f th e d ifferenti­
ated p ro d u ct as p ossible.
T h e in tro d u ctio n o f im p erfect co m p etitio n an d d e ­
clin in g u n it p ro d u ctio n co sts suggest th ree so u rces o f
gain from free trade. As th e m arket poten tially served
by firm s ex p a n d s from a n atio n al to a w orld m arket,
th ere will b e gains d ue to d eclin in g u nit p ro d u ctio n
co sts. T h e se co n d is th e red u ctio n in m on o p o ly p ow er
o f firm s faced w ith foreign co m p etito rs. T h e th ird is
th e gain to co n su m ers from low er p rices an d in ­
creased p ro d u ct variety. G enerally speaking, gains
from trad e resu lt from th e in cre a se in com p etitive
p ressu res as th e d o m estic eco n o m y b e c o m e s less
in su lated from th e w orld eco n om y. N on eth eless, the
n u m ero u s m arket stru ctu res an d firm behaviors p o s­
sible u n d e r im p erfect co m p etitio n p reclu d e a d efin i­
tive sta tem e n t abou t th e optim ality o f free trade.

FEDERAL RESERVE BANK OF ST. LOUIS

F ig u re 3

JANUARY/FEBRUARY 1988

T h e d em a n d curve is Dlls an d th e m arginal co st curve is
MCUS. (The m arginal revenu e curve a sso cia te d w ith Dus
is om itted.) T h e m o n o p o ly p rice an d o u tp u t are Pus
an d Q lls.
T h e ch a n g e from au tarch y to free trad e tran sform s
th e n atio n al m o n o p o lies into a w orld duopoly. As­
sum ing th e firm s follow a C ou rn ot strategy, p rice
d eclin es from Pus to Plv, sales in th e U nited States
in crea se from Q us to D 'us, an d co n su m ers gain area Pus
RS P„ or h + i. Profits (ignoring fixed costs) in th e
U nited States d eclin e from area Pus RWM o r i + j + k to
area P„ SXM o r j + k + 1. T h e d o m estic firm h as o n eh alf o f th e d o m estic m arket, so its profits are j w ith k +
1 going to th e foreign firm. T h e d o m estic firm 's exp orts
allow it to ca p tu re o n e-h a lf o f th e foreign m arket. If th e
foreign m arket is id en tical to th e d o m estic m arket, th e
firm ’s profits on foreign sales w ill eq u al k + 1. T h e re ­
fore, th e n et re d u ctio n in th e d o m estic firm 's profits is
i — 1 an d th e overall w elfare gain to th e eco n o m y is
h + 1.

So m etim es th e ben efits o f exp an d ed co n su m p tio n
resu lting from free trad e are less th an th e co sts a s so c i­
ated w ith d istorted p ro d u ctio n . V enables and Sm ith
(1986) provide a g rap h ical illu stration, d u p licated in
figure 3, o f the p reced in g point using th e B rand erKrugm an d uopoly m odel. A ssum e the U.S. m arket and
th e m arket in the rest o f th e w orld for a sp ecific good
are m on op olies, the good is p ro d u ced at a co n sta n t
m arginal cost, an d th ere are n o tran sp o rtatio n co sts.




If th e a ssu m p tio n o f id en tically sized d o m estic and
foreign m arkets is d ropped, th e n a different c o n c lu ­
sion is possib le. If th e foreign m arket is sm aller than
th e d o m estic, th e profits o f th e d o m estic firm in the
foreign m arket will b e le ss th a n k + 1. A ssum ing zero
exp orts, th e d o m estic gains from trad e are h — k, an d
th e d o m estic eco n o m y co u ld lose from free trade. In
th is case, c o n su m e r gains ca n b e m ore th a n offset by
th e shifting o f profits from th e d o m estic to th e foreign
eco n om y. T h is shifting reflects th e co n tra ctio n o f an
activity th at is alread y to o little to an even sm aller level.

29

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

The Borrow ed-Reserves Operating
P roced u re: Theory and Evidence
Daniel L. Thornton

I n LATE 1982, th e Federal Reserve sw itch ed from a
n on borrow ed -reserves to a borrow ed -reserves o p era t­
ing p ro ced u re.' Analysts generally believe that the
a d op tion o f this p ro ced u re, w h ich involves th e use o f a
“borrow ings a ssu m p tio n ” sp ecified by th e Fed eral
O pen M arket C om m ittee (FOMC), rep rese n ts a p o licy
reversal tow ard th e settin g o f th e federal fu n d s rate
an d aw ay from d irect m o n ey stock con trol.
T h is p ap er d iscu sse s th e m erits o f th e borrow ed reserves op eratin g p ro ced u re as a m eth o d for m on ey
stock or in terest rate con trol, analyzes th e rela tio n ­
sh ip s b etw een th e borrow ings assu m p tio n , th e federal
funds rate and th e d isco u n t rate, an d provides som e
evid ence on h ow th e new p ro ced u re h as b een u sed
sin ce late 1982.

THE NEW OPERATING PROCEDURE
T h e co rn e rsto n e o f th e borrow ed -reserves o p era t­
ing p ro ced u re is th e borrow ings fu n ction , w h ich
reflects th e b asic e co n o m ic factors that in d u ce d ep o s­
itory in stitu tio n s to borrow from th e Fed eral Reserve.
It is u su ally argued th at th e level o f borrow ings (Borr)
from th e F ed eral Reserve is in flu en ced prim arily by
th e sp read betw een th e federal fund s rate (FFR) and
th e Fed eral R eserve's d isco u n t rate (DR). A ccordingly,
th e borrow ings fu n ctio n is
(1) Borr, = b„ + b,(FFR, — DR,) + v„
w h ere b0 an d b, are co n sta n ts (b, > 0). T h e rand om
erro r term , v„ cap tu res th e effect o f all o th e r factors
that d eterm in e d ep ository in stitu tio n s’ borrow ings. It
ca n be thou gh t to rep resen t “tran sitory " sh o ck s to the
borrow ings fu n ction , w hile ch an g es in b0 rep resen t

Daniel L Thornton is a research officer at the Federal Reserve Bank of
St. Louis. Rosemarie V. Mueller provided research assistance.
1For a discussion of this change, see Roley (1986), Wallich (1984)
Federal Reserve Bank of New York (1986) and Gilbert (1985).

http://fraser.stlouisfed.org/
30
Federal Reserve Bank of St. Louis

“p e rm a n e n t” shifts in th e fu n ctio n . T h is d istin ctio n
will b e u seful later.
T h e equilibriu m FFR, given DR, is d eterm in ed by the
d em an d for an d th e su pply o f total reserves. T h e
su pply o f total reserves is co m p o sed o f n o n b o rro w ed
reserves (NBR) an d reserves su p p lied th rou gh th e d is­
co u n t w indow . T h e d em an d for total reserves is co m ­
p osed o f th e d em an d for requ ired plus e x cess re­
serves. T h eory su ggests th at th e d em an d for reserves
is inversely related to th e federal fund s rate. Equating
th e d em an d for an d th e su pply o f to tal reserves resu lts
in an equilibriu m eq u atio n for th e fed eral fu n d s rate o f
the gen eral form
(2) FFR, = |x„ - jj.,NBR„

|x,„ |x, > 0.

T his eq u ation show s all p o ssib le co m b in a tio n s o f n o n ­
b orrow ed reserves an d fed eral fund s rates for w h ich
th e su pply o f an d th e d em an d for to tal reserves are
equal. E qu ation 1, p re sen ted in figure lb , an d eq u ation
2, p resen ted in figure la , can be u sed to illu strate how
the borrow ings p ro ced u re o p era tes an d to sh o w sim i­
larities and d ifferen ces b etw een a borrow ings o p era t­
ing p ro ced u re an d a fed eral fu n d s rate targeting p ro ­
ced u re. Su p p o se th e borrow ings assu m p tio n is set at
Borr", sh ow n in figure lb . T h e borrow ings a ssu m p tio n
rep rese n ts th e am ou n t o f reserves th at th e Fed w ish es
to in d u ce d ep ository in stitu tio n s to b orrow from the
d isco u n t w indow . T h is im plies FFR m u st eq u al FFR*.
Given th e d em an d an d su p p ly fu n ctio n s, th e F ed ca n
h it its borrow ings target by su pplying n o n b o rro w ed
reserves equal to NBR*. T h is esta b lish es an eq u ilib ­
rium FFR at FFR* co n siste n t w ith th e borrow ings
objective.
W hile figure 1 clarifies th e relatio n sh ip b etw een the
b orrow ings objective, th e fu n d s rate an d n o n b o r­
row ed reserves, it m ay b e u nfam iliar to m an y read ers.
C on sequen tly, from th is point on, th e analysis will be
illu strated in term s o f th e m ore fam iliar figure 2. (See
ap p en d ix A for a d iscu ssio n in term s o f figure 1 an d
th e d etails co n cern in g th e v arian ces o f th e fed eral
fu n d s rate an d borrow ings stated in th e text.) In figure

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

F ig u re 1

Equi l i bri um in the Reserv e Ma r k e t u nder a Bo r r owi n gs Pr oc edu r e
FFR

FFR

2, th e su pply o f total reserves is o btain ed by adding
borrow ed reserves (from eq u atio n 1) to th e d esired
level o f n on borrow ed reserves, NBR*, on th e a ssu m p ­
tion th at b„ > 0 (see th e sh ad ed insert). T h e eq u ilib ­
rium fed eral fund s rate, FFR*, is d eterm in ed by th e
in tersectio n o f th e su pply an d d em an d curves in
figure 2. As before, th e target level o f borrow ings is
ach ieved by providing th e ap p rop riate am o u n t o f n o n ­
borrow ed reserves.

F ig u re 2

E q u i l i b r i u m in the Re s e r v e M a r k e t u nd e r a
Borrowings Procedure

H o w the Fed Should R esp on d to
Shocks under a Borrowings Operating
Procedure
To u n d e rstan d p rop erly th e efficacy o f th e borrow ­
ings p ro ced u re as a m eth o d o f m o n ey stock o r in terest
rate co n tro l, it is im p o rtan t to see how th e F ed reacts
to sh o ck s w h en u sin g it. First, co n sid e r its re sp o n se to
an in crea se in th e d em an d for total reserves, illu s­
trated in figure 3a. O th er things co n stan t, an in crea se
in th e d em an d for to tal reserves ca u se s th e eq u ilib ­
rium funds rate to rise from FFR* to FFR *' and borrow ­
ings to rise from th e d esired level (TR* —NBR*) to
(TR*' —NBR*). T o bring borrow ings back to its d esired
level, th e supply of n o n b o r r o w e d reserves m u st be
in creased via an o p en m arket p u rch a se o f governm ent
secu rities. T h is re d u ce s th e federal fu nd s rate and



31

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

The Borrowings Function: A Supply o r Demand Curve?
T h e F ed d oes n o t set quantitative re strictio n s on
th e level o f aggregate borrow ings. C onsequen tly, it
is u n cle a r w h eth er th e qu antity o f reserves in je cted
throu gh th e d isco u n t w indow p rop erly reflects
“su p p ly " o r “d em a n d ” co n sid eratio n s. In o th er
w ords, d o es th e borrow ings p o rtio n o f the TRScurve
in figure 2 rep rese n t a su pply curve, as is co m m o n ly
assu m ed , o r a d em an d curve? T h ou g h th e an sw er
to th is q u estio n h as no bearin g on th e analysis
p resen ted in th e text, th e issu e is in terestin g in its
ow n right.
It is im p o rtan t to n o te th at d isco u n t w ind ow
borrow ing is a privilege for d ep ository in stitu tio n s,
not a right. N evertheless, u ntil th e m id-1960s, the
d isco u n t w ind ow w as co n sid ere d “o p e n " in the
sen se th at th o se w ho w an ted to borrow w ere m ore
or less au to m atically gran ted a loan . Only if a bank
h ad b ee n d eem ed to have m ad e too frequ en t or
in ap p ro p riate u se o f th e privilege w ould it b e “d is­
co u rag ed " from fu rth er borrow ing.
F u rth erm o re, for various reason s, so m e banks
m ay be “re lu cta n t to b o rro w ” from th e Fed. M any
e co n o m ists believe th at th e re lu cta n ce to borrow is
in d u ced by th e F ed 's ad m in istratio n o f th e d is­
co u n t w ind ow u n d e r w h ich th e am o u n t o f b o r­
row ed reserves available is d eterm in ed by a co m p li­
cated system o f p rice an d n o n -p rice rationing.
A ccording to th is view, in th e a b se n c e o f ad m in is­
trative restrictio n s, d ep ository in stitu tio n s w ould
borrow from th e Fed w h en ever the d isco u n t rate is
belo w th e fed eral fu nd s rate, an d obtain fu n d s in
th e federal fund s m arket w h en ev er th e federal
fund s rate is below th e d isco u n t ra te.1 T h u s, in the
a b se n ce o f any ad m in istratio n o f th e d isco u n t w in ­
dow, th e d isco u n t rate sh o u ld establish an effective
ceiling for th e federal fu nd s rate.
If true, th e su p p ly o f reserves sh o u ld b e perfectly
in tere st-elastic at o r below th e cu rre n t d isco u n t
rate. Sin ce d isco u n t borrow ings are sm all (usually
b etw een $1 billion an d $2 billion) an d th e federal
fund s rate is generally a b o v e th e d isco u n t rate, it is
often assu m ed th at th ere is n o n -p rice ration in g at
th e d isco u n t w indow . Presum ably, th e m arginal,
n o n -p ecu n iary rationing co st is th e am o u n t by
w h ich th e federal fund s rate ex ceed s th e d isco u n t
rate. T h is sp read is th e m arginal p rem iu m d ep o si­

'See Goodfriend (1983).

http://fraser.stlouisfed.org/
32
Federal Reserve Bank of St. Louis

tory in stitu tio n s w illingly pay to o b tain fund s from
so u rces o th e r th an th e Fed. C on sequen tly, a c c o rd ­
ing to this view, th e borrow ings fu n ctio n is e s s e n ­
tially a su p p ly curve th at ch a n g es as th e Fed
ch a n g es its n o n -p ecu n ia ry p rice o f fund s o btain ed
at th e d isco u n t w indow .
T h is view h as certa in d eficien cies. First, it im plies
th at borrow ings sh ould b e zero w h en ev er th e d is­
co u n t rate is at o r above th e fed eral fund s rate —
th at is, th e in tercep t term in eq u atio n 1, b„, sh ould
b e zero. E stim ates o f eq u atio n 1, how ever, typically
y ield a sign ifican t positive in tercep t term . A ppar­
ently, d ep ository in stitu tio n s borrow at th e d is­
co u n t w in d ow even w h en it is th e m ore costly
so u rce o f m arginal funds. F o r som e d ep o sito ry in ­
stitu tio n s at least, borrow in gs a p p ea r to b e d e te r­
m in ed by factors o th e r th an the relative c o s t o f
funds, at least as m e a su red by th e fed eral fund s
rate/discount rate spread .
Riefler (1930) argued th at th e positive in tercep t is
d ue to a "n e e d ” th at ca n n o t b e satisfied in the
m arket. At on e level, R iefler’s n ee d s th eory cou ld
result from a lack o f fin an cial so p h istica tio n . For
exam ple, so m e banks m ay seld om , if ever, b o rro w or
len d in th e fed eral fu n d s m arket. W h en th e se in sti­
tu tio n s ex p e rie n ce a reserve d eficien cy , for w h a t­
ever reason , th ey m ay go to th e F ed ra th er th a n th e
u nfam iliar fed eral fun d s m arket, even if th e d is­
co u n t rate is above the fed eral fun d s rate. Viewed
from th is p erspective, th e b orrow ing fu n ctio n is a
d em an d curve, at least to th e ex ten t that ch a n g es in
d em an d -related facto rs ca u se th e curve to shift.
Today, th is exp lan atio n lacks th e cred ib ility it m ight
have h ad at R iefler's tim e. T h e in crea sed so p h is tic a ­
tion o f d ep o sito ry in stitu tio n s an d th e w id esp read
use o f co rresp o n d en t banking have w eaken ed the
validity o f th is argum ent.
A different exp lan atio n for th e positive co n sta n t
term in th e borrow in gs fu n ctio n is th at th e average
federal fund s rate d o es n ot rep rese n t th e relevant
op p o rtu n ity co st for all d ep o sito ry in stitu tio n s.
Som e in stitu tio n s, for exam p le, m ay b e able to b o r­
row in th e federal fund s m arket on ly at a very high
federal fu n d s rate, if at all. F or th e se in stitu tio n s,
d isco u n t w in d o w borrow ings m ay b e attractive,
even w h en th e average level o f th e fed eral fu n d s rate
is w ell below th e d isco u n t rate. In th is ca s e too,
how ever, shifts in th e borrow in gs fu n ctio n reflect
ch a n g es in d em an d ra th er th a n supply.

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

F ig u re 3

The Effect of an Increase in the D em and for Total Reserves u n d er a B o rrow ing s P ro ce d u re
FFR

FFR

(o)

brings borrow ings back to Borr*; th at is, (TR* —NBR*)
eq u als (T R * '—NBR*’), as show n in figure 3b. C o n se­
quently, n eith e r borrow ings n o r th e fed eral fun d s rate
is ch an g ed ; in stead , th e d em an d for total reserves is
satisfied by an in crea se in NBR.
Alternatively, th e borrow ings fu n ctio n co u ld shift to
th e right, illu strated by th e rightw ard shift in th e TRS
fu n ctio n in figure 4a. O th er things th e sam e, borrow ­
ings will in crea se and th e equilibriu m federal fund s
rate will d eclin e from FFR* to F FR *'. If th e borrow ings
assu m p tio n is m ain tain ed at Borr*, th e su p p ly o f n o n ­
borrow ed reserves m u st b e in creased . T h is will put
fu rth er dow nw ard p re ssu re on th e fed eral fund s rate
until it re a ch es FFR*" (see figure 4b), in d u cin g borrow ­
ings back to Borr*. In th is in sta n ce , borrow ings are
u n ch a n g ed an d th e fund s rate falls.

The Borrow ings Procedure:
An Ineffective Tool f o r M oney
Stock Control
T h e above analysis su ggests th at strict ad h e re n ce to
th e borrow ings p ro ced u re w ill n ot provide effective
m o n ey stock co n tro l. If borrow ings are kept at the
a ssu m ed level, ch an g es in th e d em an d for m on ey and,
h en ce, reserves will be a cco m m o d ated bv c o m p e n sa ­
tory ch an g es in th e su p p ly o f reserves. T h is is illu s­



(b)
trated by th e u sual m o n ey supply/m oney d em an d
paradigm sh ow n in figure 5a. Here, th e m o n ey supply
is positively related to th e FFR an d is draw n h old ing
th e d isco u n t rate an d th e level o f n o n b o rro w ed r e ­
serves u n ch a n g e d .2As usual, th e d em an d for m o n ey is
negatively related to th e in terest rate. If th e borrow ings
p ro ced u re is u sed to co n tro l th e m o n ey stock, a
m o n ey target, M*, m u st b e estab lish ed . Given th e d e­
m an d for m o n ey an d th e d isco u n t rate, th is requires
achieving a sp ecific in terest rate, FFR*. Given th e b o r­
row ings fu n ctio n , th is im plies a target level o f b o rro w ­
ings (Borr*) an d target settin g for n o n bo rro w ed re ­
serves (NBR*).

2The supply of money is positively sloped on the assumption that
borrowings are positively related to the funds rate and that the
demand for other reservable components of the monetary base is
negatively related to the funds rate. See Thornton (1982b) for a
model that incorporates these assumptions.
The federal funds rate is not used commonly as the representa­
tive interest rate for money demand. But it is used commonly in
money supply models as well as those that incorporate both money
supply and money demand functions. This may not be desirable,
especially if the relationship between the funds rate and the true
representative rate in the money demand equation is either highly
variable or affected by changes in policy or policy-related variables.
It is adequate, however, if there is a fixed proportionate relationship
between these rates.

33

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

F igu re 4

The Effect of an Increase in the Borrowings Function under a Borrowings Procedure
FFR

NBR

FFR*
FFR*

T R * TR*'

(a)

Su p p o se th e d em an d for m oney in crea ses from Md
to M',,.3 O th er th in gs th e sam e, th e resu ltin g rise in the
fu nd s rate w ould in cre a se borrow ings. In o rd er to
red u ce borrow ings back to Borr*, th e su pply o f n o n ­
borrow ed reserves m u st in crease from NBR’ to NBR*',
shifting the m on ey supply sch e d u le to the right as
sh ow n in figure 5a. B eca u se borrow ings d ep en d only
on th e level o f th e federal fund s rate (given th e d is­
co u n t rate and assu m in g the fu n ctio n is oth erw ise
stable), th e d esired level o f b o rro w in g s ca n he
achieved by supplying th e requ isite qu an tity o f n o n ­
b orrow ed reserves. H ence, all shifts in th e d em an d for
m on ey are a cco m m o d a ted bv co rresp o n d in g shifts in
th e m oney su pply if Borr* rem ain s u n ch an g ed . In this
case, no differ e n c e b etw een m o n ey stock co n tro l u n ­
d er a borrow ings p ro ced u re and a fed eral fund s rate
targeting p ro ced u re exists.

3Total reserves demand is composed of the demand for required and
excess reserves. The demand for required reserves can be thought
of as a derived demand, derived from the demand for money via the
relationship between checkable deposits and required reserves.
Because the demand for money generally is estimated to be
interest-inelastic, the demand for required reserves should also be
interest-inelastic. Of course, during the lagged reserve accounting
(LRR) period prior to February 1984, the demand for required
reserves should be perfectly interest-inelastic. (See Thornton
(1983) for a discussion of the differences between lagged and
contemporaneous reserve accounting.) The demand for excess
reserves usually is found to be relatively interest-insensitive as well.

http://fraser.stlouisfed.org/
34
Federal Reserve Bank of St. Louis

(b)
Now, su p p o se th e borrow ings fu n ctio n tem porarily
in crea ses, th at is, v, > 0. T h is p ro d u ces a tem p orary
rightw ard shift in th e m o n ey su p p ly from M s to M 's, as
illu strated in figure 5b.4 O th er things th e sam e, the
feder al fund s rate falls an d borrow ings in crea se. To
bring borrow ings b a ck to Borr*, n o n b o rro w ed reserves
m u st b e in crea sed , shifting th e m o n ey su p p ly s c h e d ­
ule still fu rth er to th e right from M 's to M"„. As a result,
the m o n ey stock is fu rth er aw ay from its targeted level,
M*. Str ictly en fo rced , th e borrow ings o p eratin g p ro c e ­
dure yield s less sh o rt-term co n tro l over th e m o n ey
stock th an a straight forw ard fed eral fund s rate target­
ing p ro ced u re as long as th e borrow ings fu n ctio n is
su b je ct to som e variability an d th e Fed m akes no
allow ance for the shift.5

The Borrowings Procedure as a Federal
Funds Rate Target
T h e borrow ings p ro ced u re p ro d u ces resu lts that
are id en tical to th o se u sin g a fed eral fu n d s rate target­
ing p ro ced u re if all sh o ck s em a n a te from th e d em an d s
for m o n ey o r reserves. T h e two p ro ced u res y ield dif-

4The assumption here is that the discount window is assumed
“open,” given a set of unchanged administrative constraints.
5This point has also been made by VanHoose (1988).

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Figu re 5

The Effect of an Increase in the Demand for M oney and the Borrowings Function on the M oney
Supply under a Borrowings Procedure

feren t resu lts, how ever, w h en th ere are shifts in th e
borrow ings fu n ctio n . T h e borrow ings p ro ced u re e x a c­
erb ates the effect o f sh o rt-ru n flu ctu atio n s in the b o r­
row ings fu n ctio n on th e funds rate.1’ If an y part o f su ch
shifts is offset bv ch an g es in NBR, th e borrow ings
p ro ced u re will p ro d u ce g reater variability in th e fed ­
eral fund s rate and less variability in borrow ings th an
a d irect fund s rate targeting p ro ced u re. Ind eed , u n d e r
a borrow ings p ro ced u re, th e variability o f borrow ings
will b e less th an th e variability o f th e fund s rate u n d er
fairly general con ditions/ N evertheless, if th e bo rro w ­
ings fu n ctio n is stable (in th e sen se th at all flu ctu a ­
tio n s are transitory), flu ctu atio n s in th e fed eral fund s

rate will n et out over tim e [E(vt) = 0]; th erefore, the
borrow ings o p eratin g p ro ced u re ca n b e u sed to
achieve a fed eral funds rate target over a som ew h at
lon g er-term h o rizo n .8
O f co u rse, th e borrow ings fu n ctio n also co u ld ex ­
h ibit p e rm a n en t shifts a sso cia te d w ith ch a n g es in b0 in
eq u atio n 1. In th is in sta n ce, th e a ssu m ed level of
borrow ings w ould be ach ieved only w ith a su bstan tial
ch a n g e in th e fed eral fund s rate.‘J F or exam ple, if
borrow ings are m ain tain ed at a p red eterm in ed level
d esp ite a p e rm a n en t d ecre a se in th e borrow ings fu n c­
tion, n o n b o rro w ed reserves m u st b e re d u ce d u ntil th e
fed eral fu n d s rate rises en o u gh to retu rn borrow ings
to th eir form er level. On th e o th e r hand , if th e fed eral
fu n d s rate is kept at its fo rm er level, th e borrow ings
assu m p tio n m u st b e low ered .

6Roley (1986) argues that, while the Fed moves quickly to offset
changes in the demand for total reserves, it does not do so for
changes in the borrowings function. Thus, he argues that the federal
funds rate will vary more in the short run under a borrowed-reserves
procedure than under a federal-funds-rate targeting procedure.
Roley’s assertion implies the Fed can distinguish between shifts in
these two functions.

8The two procedures are equivalent if the shocks to both the demand
for total reserves and borrowings exhibit no persistence and if no
attempt is made to offset such temporary shocks. See appendix A
for details.

7The relative variability of borrowings will be less if the slope of the
total reserves supply function in figure 2 is flatter than that of the
demand for total reserves — a condition that is likely to hold — or if
the Fed is reasonably successful in offsetting the effect of shifts in
the borrowings function. See appendix A for details.

9Wallich (1984), p. 26, notes that the borrowings function is unstable.
Therefore, he contends the borrowings procedure cannot be re­
garded as a form of "rate-pegging,” because the " . . . chosen level
of borrowing is consistent with any range of values of the funds
rate.”




35

FEDERAL RESERVE BANK OF ST. LOUIS

In sum m ary, th e borrow ings o p eratin g p ro ced u re is
a useful su rrogate for a sh o rt-ru n federal fu nd s rate
targeting p ro ced u re o n ly if all ch an g es in aggregate
borrow ings are p ro d u ced by shifts in th e d em an d for
total reseives. It is a useful su rrogate for a longer-ru n
fed eral fu nd s rate targeting p ro ced u re only if th e b o r­
row ings fu n ctio n is stable, th a t is, su b je ct only to
tem porary, ran d o m shifts. It is u n su ited for a federal
fund s rate target w henever there are p erm an en t shifts
in th e borrow ings fu n ctio n , u n less th e borrow ings
assu m p tio n is ch a n g ed sufficiently.

The Policy Implications o f a Change in
the Borrowings Assumption
Usually, m onetary aggregates or in tere st rates are
ch o se n as in term ed iate policy targets. W hy is the
borrow ings p ro ced u re u sed w h en th e m o n ey sto ck or
th e fund s rate co u ld be m o re d irectly co n tro lled by
o th e r p ro ced u res? W hat are th e policy im p licatio n s o f
a ch an g e in th e borrow ings assu m p tio n ? W ithout p re ­
cise in form ation abou t th e in term ed iate p o licy target,
it is difficult to an sw er th e se q u estio n s definitively;
n ev ertheless, so m e g en eralizatio n s ca n be m ad e.
If th e borrow ings fu n ctio n is stable, an in crea se in
th e borrow ings assu m p tio n can be in terp reted as a
m ove tow ard restrain t in th at it red u ces th e su p p ly of
reserves relative to d em an d . Conversely, a d e cre a se is
a m ovem ent tow ard ease. If th e borrow ings fu n ctio n is
u nstable, in th e sen se th at p erm an en t shifts o ccu r,
how ever, ch an g es in th e borrow ings a ssu m p tio n m ay
reflect th e F e d ’s aw aren ess o f th ese shifts an d its
d esire to m itigate th e ir effect o n th e funds rate. A
failure to ch an g e th e borrow ings assu m p tio n , on th e
o th e r hand , co u fd be in terp reted as a m ovem en t to ­
w ard ease o r restraint, d ep en d in g on th e d irectio n of
th e shift o f th e borrow ings fu n ctio n .

The Relationship Between Changes in
the Borrowings Assumption and
Changes in the Discount Rate
C hanges in th e borrow ings a ssu m p tio n an d th e
d isco u n t rate can be view ed as su b stitu tes. B eca u se it
d ep en d s on th e d isco u n t rate, th e T B Scurve shifts w ith
a ch an g e in th e d isco u n t rate. F o r exam ple, a d isco u n t
rate in crea se shifts th e slop ed p o rtio n o f th e TRS curve
to th e left at all levels o f th e fu nd s rate. If th e b o rro w ­
ings a ssu m p tio n is u n ch an g ed , th e qu antity o f n o n ­
b orrow ed reserves m u st be re d u ce d u ntil th e funds
rate rises en o u gh to resto re borrow ings to th e ir d e­
sired level. On average, th e fed eral fu n d s rate will

http://fraser.stlouisfed.org/
36
Federal Reserve Bank of St. Louis

JANUARY/FEBRUARY 1988

ch an g e p o in t-fo r-p o in t w ith a ch a n g e in th e d isco u n t
rate u n d e r a strictly en fo rced borrow ings p ro ced u re.
T h e sam e ch a n g e in th e eq u ilibriu m fed eral fu n d s
rate co u ld b e o b tain ed by ch an g in g th e borrow ings
assu m p tio n in stead . C on sequen tly, ch a n g es in th e
borrow ings assu m p tio n an d ch a n g es in th e d isco u n t
rate are su b stitu tes in th e ir effect on th e fed eral fund s
rate u n d e r a strictly en fo rced borrow ings p ro ced u re."’
Table 1 rep o rts ch a n g es in th e d isco u n t rate an d the
borrow ings a ssu m p tio n from O cto b er 1982 th rou gh
D ecem b er 1986. T e ch n ic a l d isco u n t rate ch an g es, re ­
p orted ly m ad e s o le ly to keep th e d isco u n t rate in lin e
w ith m arket in terest rates, are d en o te d by a T; th o se
m ad e for o th er, policy -related , rea so n s are d en o ted by
a P ." As th e table show s, ch a n g es in th e borrow ings
assu m p tio n an d th e d isco u n t rate gen erally o ccu rred
aro u n d th e sam e tim e: five of th e 11 ch a n g es in the
d isco u n t rate cam e w ith in abou t o n e w eek o f a ch an g e
in th e borrow ings assu m p tio n , w hile tw o w ere w ith in
tw o w eeks. M oreover, all ch a n g es th at o cc u rre d clo se
to g eth er w ere in th e sam e d irectio n , in d icatin g c o n ­
sisten t m ovem en ts in b o th th e borrow ings a ssu m p ­
tio n an d th e d isco u n t rate.
T h e table also show s altern atin g p eriod s o f ea se and
restrain t. From O cto b er 1982 th rou gh th e en d o f the
year, th e borrow ings a ssu m p tio n an d th e d isco u n t
rate w ere red u ced . W hile ch a n g es in th e borrow ings
a ssu m p tio n w ere m o d est (even cum ulatively) an d on e
d isco u n t rate ch a n g e w as te ch n ica l, p o licy eased
m od erately d uring th is period.
From spring 1983 to sprin g 1984, p o licy m oved to ­
w ard restrain t. T h e borrow in gs a ssu m p tio n w as
raised bv $600 m illion from M arch th rou gh August
1983, low ered in O cto b er 1983 by $150, th e n in crea sed
by $350 m illion in M arch 1984. T h e last in cre a se w as
follow ed clo sely by a 50 b a sis-p o in t te ch n ica l in crea se
in th e d isco u n t rate.
Policy w as ea sier d uring th e fall o f 1984. T h e b o rro w ­
ings assu m p tio n w as red u ced by $700 m illion from
early O cto b er to late D ece m b er an d two policy-related
cu ts in th e d isco u n t rate re d u ce d it by a full p e rc e n t­
age point. T h e re w ere no large, co n siste n t m ovem ents
in th e borrow ings a ssu m p tio n d uring 1985 an d n o n e
after early Febru ary 1986, d esp ite fou r cu ts in th e
d isco u n t rate (three o f w h ich w ere p olicy-related ).

'“Because a one percentage-point change in the discount rate is
associated with about a $420 million change in borrowings over this
period, a $420 million change in the assumed level of borrowings
should have an effect on the funds rate equal to a one percentagepoint change in the discount rate. See Thornton (1986).
"See Thornton (1986,1982a) for a discussion of the classification of
discount rate changes into technical and non-technical changes.

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

Table 1
Changes in the Discount Rate and the
Borrowings Assumption
Change in the
Change in the
discount rate borrowings assumption
(in basis points) (in millions of dollars)
October 6,1982
October 12,1982
November 17, 1982
November 22, 1982
December 14, 1982
December 22,1982

$ -5 0
-5 0 T
-5 0
-5 0 P
- 50 P
-5 0
50
100
100
250
100
-1 5 0

March 30,1983
May 25, 1983
June 24, 1983
July 14, 1983
August 24, 1983
October 5, 1983
March 28, 1984
April 9, 1984
October 3,1984
November 8,1984
November 23,1984
December 19, 1984
December 24, 1984
February 14, 1985
March 27, 1985
May 20, 1985
May 22,1985
August 21, 1985
October 2, 1985
November 6, 1985
December 18,1985
February 13, 1986
March 7, 1986
April 21, 1986
July 11, 1986
August 21, 1986

350
50 T
-2 5 0
-1 7 5
-5 0 P
-2 7 5
- 50 P
50
50
- 50 P
-5 0
75
75
-5 0
-1 0 0
-5 0
-5 0
-5 0
-5 0
-5 0

P
T
P
P

tion in borrow in gs abou t its m ean level.'2 F or th e
b orrow ings p ro ced u re to b e u sed effectively as a fed ­
eral fund s rate target over th e lo n g er run, th e borrow ­
ings fu n ctio n m u st b e stable. It is im portan t, therefore,
to d eterm in e w h eth er th ere have b ee n p erm an en t
shifts in th e borrow ings fu n ctio n . If so, th e key issu e is
h ow th e borrow in gs a ssu m p tio n w as ch a n g ed in re ­
sp o n se to th e se shifts.
To ex am in e this issu e, eq u atio n 1 w as estim ated
using rand om co efficien t regression , w h ere b o th the
co n sta n t term , b„, an d th e slop e coefficien t, b„ are
allow ed to vaiy th rou gh tim e .13 C hart 1 p resen ts
ran d o m -co efficien t-reg ressio n estim a tes o f th e c o n ­
stant term an d a b an d re p resen tin g plus or m inu s one
stan d ard error. T h e vertical lin es sh ow the d ates on
w h ich th e borrow ings a ssu m p tio n w as raised o r low ­
ered, as in d icated .
T h e in tercep t show s co n sid era b le variability. W ith
b u t th ree sign ifican t ex cep tio n s, th e borrow in gs a s­
su m p tio n w as ch a n g ed in th e d irectio n co n siste n t
w ith m itigating th e effect o f shifts in th e borrow ings
fu n ctio n on th e fed eral fu n d s ra te.14 T h e th ree e x c e p ­
tio n s o cc u rre d in O cto b er 1984, August 1985 an d O cto ­
b e r 1985. In O cto b er 1984, w h en th e borrow in gs fu n c­
tion sh ifted upw ard, th e borrow ings a ssu m p tio n w as
red u ced from $1 b illio n to $750 m illion. In b o th August
a n d O cto b er 1985, th e borrow ings a ssu m p tio n w as
raised, d esp ite th e d ow nw ard shift in th e borrow ings
fu n ctio n . B oth in crea s es w ere relatively sm all ($75
m illion each ), how ever, an d b o th w ere co m p letely
offset by th e m id -D ece m b er d ecrea se.
T h e se resu lts are co n siste n t w ith m ovem en ts in
borrow ings, th e borrow ings assu m p tio n an d th e fed ­
eral fund s rate p re sen ted in ch art 2. T h e O cto b er 1984
ch a n g e in th e borrow ings a ssu m p tio n p reced ed
m ovem en ts in b orrow ings; how ever, th is a ctio n fol­
low ed a 100 b a sis-p o in t drop in th e fund s rate from its

T Indicates changes made solely to keep the discount rate in line
with market interest rates
P Indicates changes made for other, policy-related, reasons.
12This is for the period from October 1982 through June 11,1986. See
Thornton (1986). This same function estimated for the 222 weeks
prior to October 6,1979, has an R2of about .70.

EMPIRICAL EVIDENCE ON THE USE
OF THE BORROWINGS PROCEDURE
An im p o rtan t fa cto r in d eterm in in g th e im p a ct of
th e borrow ings p ro ced u re on th e fu nd s rate and the
m o n ey sto ck is th e stability o f the borrow ings fu n c ­
tion . H istorically, th e borrow ings fu n ctio n has b een
s u b je ct to co n sid erab le ran d o m variation: by itself, the
sp read betw een th e fed eral funds rate an d the d is­
co u n t rate exp lains less th a n 50 p e rce n t o f th e varia­



13The procedure used here is suggested by Garbade (1977). The
equation was first estimated allowing only the constant term to vary.
It was then reestimated allowing both the constant and slope coeffi­
cients to vary; this was done to determine whether variation in the
slope coefficient was being inappropriately attributed to the constant
term. The results presented in chart 1 are from the latter estimation.
The qualitative interpretation of the relationship between changes in
the borrowings assumption and shifts in the borrowings function
was not affected by the different estimation procedures.
,4There were two other exceptions: they occurred on October 6,1982,
and May 22, 1985. In both instances, however, these changes
predate the shift by only one week. Including these in subsequent
statistical tests does not affect the results.

37

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

C h a rt 1

V a r y i n g P a r a m e t e r Intercept

.11________________

Jan .'82

The b lu e v e r tic a l lin e s
The g re y v e rtic a l line s re fe r to p o s itiv e c h a n g e s in th e b o rro w in g s assu m p tio n .

F u rth er evid ence on th e effects o f using th e borrow ­
ings p ro ced u re ca n be o btain ed by analyzing sca tte r
p lots o f borrow ings an d th e federal fund s rate during
p erio d s in w h ich b o th th e borrow ings assu m p tio n

an d th e d isco u n t rate w ere u n c h a n g e d .“ T h e previous
resu lts suggest th at th ere are co n sid era b le tem p orary
shifts in th e d em an d for bo rro w ed reserves. If m o st o f
th e effect o f sh o rt-ru n variation in th e borrow ings
fu n ctio n on th e federal fu n d s rate w ere offset quickly,
th ere w ould be little variation in th e fed eral fund s rate
bu t co n sid era b le variation in borrow ed reserves. In
th e extrem e, if th e effect o f all su ch shifts on th e fund s
rate w ere quickly an d co m p letely offset, all observa­
tio n s w ould lie along a v ertical lin e rep resen tin g th e
average o f th e fed eral fund s rate in a sc a tte r plot of
borrow ings an d th e fed eral funds rate. On th e o th er
hand , if borrow ings w ere kept clo se to th e a ssu m ed

15The FOMC meeting was held on October 2,1984. The federal funds
rate had fallen to 9.84 percent on September 26, though it averaged
10.73 percent for the week ending Wednesday, September 26. The
weekly peak was 11.77 percent for the week ending Wednesday,
August 22; the daily peak occurred on August 1, when the federal
funds rate was 12.04 percent.

>6This procedure was suggested to me by R. Alton Gilbert. It is
interesting to note that the variability of borrowings could be reduced
by simply “tying” the discount rate to the federal funds rate. This
point was made by Thornton (1982b) and more recently by
VanHoose (1987).

cy clical peak for th e w eek end ing August 2 2 .15 Nearly
all o th e r ch an g es in th e borrow ings assu m p tio n w ere
p re ce d e d by m ovem ents in borrow ings an d th e fed ­
eral fu n d s rate in th e sam e d irectio n .

A Com parison o f the Variability o f
Borrowings and the Federal Funds Rate


http://fraser.stlouisfed.org/
38
Federal Reserve Bank of St. Louis

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

C h a rt 2

B o r r o w i n g s , B o r r o w i n g s A s s u m p t i o n , a n d F e d e r a l F unds R a t e
Perce nt

M i l l i o n s of d o l la r s

5000

4000
rate
CALE^

3000

2000

Borrow ings
^S C A LE

1

1000

Borrow ings assumption
ISCALE

-1000

-2000

level, the funds rate sh o u ld vary relatively m ore th an
borrow ings. In this case, th e o bserv ations sh o u ld be
clu stered abou t a h o rizo n tal lin e at the assu m ed level
for borrow in g s.17
D uring th e p o st-O cto b e r 1982 period, th ere w ere six
p eriod s w ith 10 or m ore w eeks in w h ich b o th the
borrow ings a ssu m p tio n and th e d isco u n t rate w ere
u n c h a n g e d .S c a tt e r plots of borrow ings and the fed ­
eral fund s rate for th ese p erio d s are p re sen ted in

l7The variability of borrowings and the funds rate depend on the
slopes of the TRSand TRd curves and the extent to which random
shocks are offset. If more than 50 percent of such shocks are offset
during the period, however, there will be more variability in the funds
rate than in borrowings regardless of the slopes of these curves.
,aPlots for the omitted periods show no pattern. They consist, how­
ever, of very few observations.



ch a rts 3a th rou gh 3f.19 T h e data u sed in th e se ch arts
have b een n o rm alized . T h e actu al level o f borrow ings
w as n o rm alized by dividing it by the level o f the
borrow ings a ssu m p tio n for th e resp ectiv e per iod. T h e
fed eral fund s rate w as n o rm alized by dividing it by its
average rate for th e p erio d .20 All ch a rts have id en tical
sca les for b o th variables to m ake it easy to co m p a re th e
relative variability. T h e solid h o rizo n tal an d vertical
lin es d en o te w h ere th e n o rm alized variables are equal

19These data exclude outliers such as the “window-dressing” borrow­
ings during the final reserve period of the year and the unusually
large borrowings associated with Continental Bank of Illinois. See
Thornton (1986) for a discussion of the latter episode.
“ Because the mean of the normalized rate spread equals one, the
rate spreads will be scattered symmetrically about the vertical line.
In contrast, the data points will be scattered asymmetrically above
(below) the vertical line depending on whether the borrowings
assumption is below (above) the average level of borrowings for the
period.

39

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

C hart 3

Selected Scatter Plots of Normalized Borrowings
(11=12)

Decem ber 29. 1982 - March 23, 1983

October 12, 1983 - M arch 21, 1984
(N=23)

M illio n s of d o lla rs

M i l l i o n s of d o l l a r s

P e rc e n t

P e rc e n t

(a)

(b)

M ay 29, 1985
(N= 12)

Ap ril 18, 1984 - September 26, 1984
(N=20)

-

August 14, 1985

M i l l i o n s of d o l l a r s
J IOU
2930
•

2680

•

24 30
21 80
19 40

—

•

•

1690
1440
1190

-

•

••

940 700 —

•

45 0

0.85

0.8 8

0.91

0.94

0.97

0 .0 0

P e rc e n t

1.03

1.0 6

1 .0 9

"1
200
0.85

1 1 1 1 1 1 1 1 1 1 1 1 1
0.88

0.91

0.94

1 1 1 1 I I

0.97 0 .0 0
P e rc e n t

1.03

1.0 6

1 I 1
1 .0 9

(d)

The vertical re fe re nc e lines refer to the n o r m a l i z e d me an of the f e d e r a l funds rate; the
h o r i z o n t a l reference lines refer to the n o r m a l i z e d m e an of borrowi ngs.

http://fraser.stlouisfed.org/
40
Federal Reserve Bank of St. Louis

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

and the Normalized Federal Funds Rate
August 27, 1986 - December 24, 1986
(N=18)

April 30, 1986 - July 9, 1986

(N=11)

0.85

0.88

0.91

0.94

0.97 0 .0 0
P e rc e n t

1.03

1.06

1 .0 9

0.85

0.88

0.91

0.94

0.9 7 0 .0 0
P e rc e n t

1.03

1.06

1. 0 9

(f)

to on e. Som e d escrip tive sta tistics for th e raw d ata are
p re sen ted in table 2.

December 27, 1978 - July 18, 1979
(N=30)

Finally, ch art 3g is a sc a tte r plot from late D ecem b er
1978 to late Ju ly 1979, w h en it is gen erally ack n o w l­
edged th at th e Fed w as targeting th e fed eral funds
rate. D uring this period , a 75 b asis-p o in t target range
for the fed eral fu n d s rate w as sp ecified .21
W ith th e ex cep tio n p erh a p s o f ch art 3c, no period
su ggests a rapid a d ju stm en t to m ain tain borrow ings
at th e a ssu m ed level. In co n trast, tw o p eriod s (charts
3a an d 3e) sh ow relatively little variability in th e fed ­
eral fund s rate. In d eed , a co m p a riso n o f th e se ch arts
w ith ch art 3g show s th at th e fu n d s rate flu ctu ated less
aro u n d its m ean d uring th e se p eriod s th a n it did
aro u n d th e m id p oin t o f th e Fed ’s narrow range for the
federal fund s rate in early 1979.
T h ere sh o u ld b e less variability in borrow ings an d
m ore variability in th e fed eral fund s rate u n d e r a

0.85

0.88

0.91




0.94

0.97 0 .0 0
P e rc e n t

(g)

1.03

1.06

1 .0 9
21T h is w a s th e on ly e x te n d e d pe riod in w h ic h th e federal fu n d s rate
b a nd w a s bo th n a rro w an d u n chang ed.

41

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Table 2
Descriptive Statistics for Borrowings and the Federal Funds Rate
Borrowings
Period
Dec. 27,1978July 25, 1979

Federal funds rate

Mean1

SD2

CV

Mean

SD

CV

$1,190.47

$367.0

.3083

10.16%

.1549

.0153

Dec. 29, 1982March 2 3 ,19832

326.3

165.9

.5084

8.56

.1176

.0137

Oct. 12,1983March 21, 1984

688.0

214.9

.3124

9.51

.1863

.0196

April 18,1984Sept. 26, 19843

951.3

301.8

.3173

11.15

.4935

.0443

May 29, 1985Aug. 14,1985

559.0

213.6

.3821

7.74

.2618

.0338

April 30, 1986July 9,1986

284.9

79.6

.2794

6.89

.0525

.0076

Aug. 27, 1986Dec. 24, 1986

343.8

184.7

.5372

5.98

.1742

.0291

NOTE: SD denotes the standard deviation. CV denotes the coefficient of variation, i.e., SD/Mean.
'In millions of dollars.
2Excludes the “window-dressing” borrowings for the week ending January 5, 1983.
3Excludes the four weeks of unusually large borrowings (May 16-June 6) associated with the problems
of Continental Bank of Illinois.

borrow ings target th an u n d e r an in terest rate target.
T able 3 p re sen ts the m ean, stan d ard deviation ISD)
and a m easu re o f relative variability, the co efficien t o f
variation (CV), for w eekly d ata d uring th e p eriod o f the
borrow ings o p eratin g p ro ced u re and d uring an equal
n u m b er o f w eeks u n d e r an in terest-rate-targ etin g re ­
gim e. T h e resu lts are gen erally co n siste n t w ith th o se
d iscu ssed above. T h e variability o f borrow ings differs
little in eith e r abso lu te o r relative term s b etw een th e
tw o p eriod s. T h e variability o f th e fed eral fu nd s rate,
how ever, fell consid erably; its SD d eclin ed nearly 30
p ercen t, w hile its CV d eclin ed n early 50 p e rce n t.22

The Im pact o f Changes in the
Borrowings Assumption on the Federal
Funds Rate
If ch an g es in th e borrow in gs assu m p tio n w ere
m ad e prim arily to offset shifts in th e borrow ings fu n c­
tion, th ere sh o u ld be no sign ifican t relatio n sh ip b e ­

22This result is only marginally affected by the switch from a one-week
to a two-week reserve accounting period. If only reserve period data
are used for the CRR period, the standard deviation of the funds rate
is 1.60 percent and the coefficient of variation is .19.

http://fraser.stlouisfed.org/
42
Federal Reserve Bank of St. Louis

tw een ch a n g es in th e borrow in gs a ssu m p tio n and
m ovem en ts in th e federal fund s rate. If ch a n g es in the
borrow ings a ssu m p tio n are m ad e for o th e r reason s,
th ey sh o u ld p ro d u ce a sign ifican t effect on th e federal
funds rate.2:1

23This is true only if a discount rate change shows significant direct
effect on the federal funds rate; Thornton (1986) argues such an
effect should be small and insignificant. Indeed, this may provide an
expectations-effect-free method of assessing the direct effect of a
discount rate change on market interest rates. See Thornton (1986)
for discussion of three potential effects of a change in the discount
rate on the federal funds rate.
Because the Federal Reserve makes a public statement when it
changes the discount rate, it is difficult to separate the direct and
announcement effects. In contrast, the levels of the borrowings
assumption for the previous calendar year are made public in the
Spring issue of the Federal Reserve Bank of New York Quarterly
Review. Because appropriately scaled changes in the discount rate
and the borrowings assumption have equivalent announcementfree effects on the supply of credit in the market, the direct effect of
discount rate changes can be gauged by investigating the effect of
changes in the borrowings assumption if the Federal Reserve
moves quickly to stabilize the level of borrowings at the new as­
sumed level. If changes in the borrowings assumption are made to
offset the effect of shifts in the borrowings function on the federal
funds rate, they will not produce a significant effect on the federal
funds rate and will not provide an announcement-effect-free test of
the direct effect of a discount rate change.

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

To investigate this, th e follow ing eq u atio n w as
estim ated :
L
(3)

AFFR, =

ot„ +

S

c^

A F F R ,., +

p„ADRT,

i= l
K
+ P.ADRN T, +

S

iXiABA,., + e,.

i= 0
Here, AFFR d en o tes th e ch an g e in th e fed eral funds
rate, ADRT and ADRNT d en o te "te c h n ic a l” and “n o n ­
te ch n ica l'' ch an g es in the d isco u n t rate, ABA d en o tes
th e ch an g es in th e borrow ings assu m p tio n an d e
d en o tes a rand om erro r term . A ch an g e in th e b o rro w ­
ings assu m p tio n w as assu m ed to be effective th e day
after th e d ecisio n w as m ad e.24 T h is eq u atio n w as e sti­
m ated u sing ordinary least squ ares (OLS) for th e p e­
riod from O cto b er 1 ,1 9 8 2 , throu gh D ece m b er 3 1 ,1 9 8 6 ;
how ever, th e eq u atio n w as estim ated sep arately for
the period o f lagged reserve requ irem en ts, LRR (up to
Febru ary 1, 1984) an d co n tem p o ra n e o u s reserve re­
qu irem en ts, CRR.23 T h e eq u atio n w as estim ated using
daily, w eekly an d reserve-p eriod data (one w eek b e ­
fore Febru ary 1, 1984, and two w eeks thereafter).
B eca u se it is n o t know n h ow quickly ch an g es in th e
borrow ings assu m p tio n are im p lem en ted or how rap ­
idly th e fed eral fun d s rate m ight resp o n d , lags o f ABA
w ere in clu d ed ; how ever, th e F -statistic for in clu d in g
lagged values o f th e ABA, Faoalacs, show s th at th e rela­
tio n sh ip b etw een th e AFFR and ABA is co n te m p o ra n e ­
ous regard less o f w h eth er daily, w eekly or reserveperiod data are u sed .28
T h e resu lts are rep o rted in table 4. T h ey in d ica te a
statistically sign ifican t p o s itiv e relatio n sh ip betw een
ch a n g es in th e funds rate an d ch an g es in th e b o rro w ­
ings assu m p tio n only for w eekly data during th e CRR
period. A fu rth er investigation o f this relatio n sh ip ,
how ever, show s it to be qu ite fragile (see ap p en d ix B

24AII changes in the borrowings assumption but one were made at
regularly scheduled meetings of the FOMC.
25The equation was estimated for separate periods for several rea­
sons. First, it would be inappropriate to estimate the equation using
reserve-period data for the entire sample period with OLS because
the error terms of one-week and two-week average data will be
different and OLS would not reflect the heteroskedasticity induced
by the change in the reserve accounting period. Second, the coeffi­
cients do not appear to be stable overthe entire period as the results
of table 4 suggest. Third, there is a pronounced quarterly seasonal
spike during the LRR period (as is readily evident in chart 2) that is
not statistically identifiable during the CRR period. Finally, there is
low-order autocorrelation in the error term during the CRR period
which is not evident during the LRR period.
26The exception was for daily data during the CRR period; however,
the sum of the coefficients was not significantly different from zero.



Table 3
Descriptive Statistics on Borrowings and
the Funds Rate under Interest Rate and
Borrowings Operating Procedures_____
Statistic

Borrowings

Federal funds rate

July 9 ,1 9 7 5 - October 3,1979
Mean
SD
CV

$567
533
.94

6.88%
2.09
.30

October 6,1982 - December 31,1986
Mean
SD
CV

$660
571
.87

8.60%
1.42
.17

NOTE: SD denotes the standard deviation. CV denotes the
coefficient of variation, i.e., SD/Mean.

for details). H ence, th ere is n o strong, statistically sig­
n ifican t relatio n sh ip b etw een ch a n g es in th e borrow ­
ings a ssu m p tio n an d ch a n g es in th e fu n d s rate. T h ese
resu lts are c o n siste n t w ith th e previous o n es th at
ch a n g es in th e borrow ings a ssu m p tio n w ere m ad e
prim arily to a cco m m o d a te shifts in th e borrow ings
fu n ctio n .27

SUMMARY AND CONCLUSIONS
T h is p ap er a sse sses th e u sefu ln ess o f th e b o rro w ­
ings op eratin g p ro ced u re in co n tro llin g th e m oney
stock o r th e in terest rates. T h e borrow ings p ro ced u re
is a p o o r m eth o d for achieving m o n ey sto ck co n tro l. In
fact, a fed eral fund s rate targeting p ro ced u re is su p e­
rior for b o th m o n ey stock an d in tere st rate con trol.
I’h e borrow ings p ro ced u re is an effective m ean s of
targeting th e fed eral fu n d s rate in th e sh o rt run only
w h en the variation in borrow in gs is due solely to shifts
in th e d em an d for total reserves. It is an effective
m ean s o f targeting th e fed eral fu n d s rate over lo n g er
p erio d s only w h en th e borrow in gs fu n ctio n is stab le. If
th ere are p e rm a n en t shifts in th e borrow ings fu n ction ,

^Alternatively, these results could be interpreted as evidence that the
announcement-free, “direct effect” of a discount rate change on the
federal funds rate is nil.

43

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Table 4
Estimates of Equation 3
October 1,1982February 1,1984

February 2,1984December 31,1986

Reserve
period’

Daily

Weekly

-0.01
(0.46)

0.04
(1.02)

0.01
(0.44)

0.01
(0.26)

0.03
(0.63)

ADRT,

1.32*
(2.08)

0.49
(0.55)

0.13
(0.19)

0.23
(0.30)

-0 .0 5
(0.08)

ADRNT,

0.85*
(2.09)

1.59*
(2.65)

0.72
(1-82)

0.69
(1.78)

0.80*
(1.96)

ABA,

0.0004
(0.50)

0.0005
(0.50)

0.0019*
(2.38)

0.0004
(0.58)

Daily
Constant

p

'AFFRLAG S

2

8.44*

15.36*

-0.0018*
(2.12)
9.82*

7.01*

Reserve
period

0.34

F
3
r ABALAGS

1.27

1.80

2.56*

1.53

1.91

R2

0.1854

0.5302

0.1188

0.1640

0.0397

SEE

0.2860

0.2942

0.4709

0.3815

0.3247

Indicates statistical significance at the 5 percent level; two-tailed test.
'During the LRR period, the reserve period was one week.
2F-statistic for the lagged values of AFFR. A quarterly seasonal was included for weekly data for the LRR
period.
3F-statistic for lagged values of ABA. The reported results are for an equation that did not include lagged
values of ABA if they were not significant.

th e fed eral fu n d s rate w ill vary w ith shifts in th e
borrow ings fu n ction , an d th e borrow ings p ro ced u re
ca n be u sed to target th e fed eral funds rate only if
co m p en sato ry ch an g es in th e borrow ings assu m p tio n
are m ade.
Evidence in d icates that the borrow ings fu n ctio n is
u n stab le. Also, it suggests that generally th e b o rro w ­
ings assu m p tio n h as b e e n ch an g ed in th e d irectio n
th at offsets th e effect o f p e rm a n en t shifts in th e b o r­
row ings fu n ctio n on th e federal funds rate.

"The Federal

Dickey, David A., and Wayne A. Fuller. “ Distribution of the Estima­
tors for Autoregressive Time Series With a Unit Root,” Journal of
the American Statistical Association (June 1979), pp. 427-31.
Engle, Robert F., and C. W. J. Granger. “Co-Integration and Error
Correction: Representation, Estimation, and Testing,” Econometrica (March 1987), pp. 251-76.

http://fraser.stlouisfed.org/
44
Federal Reserve Bank of St. Louis

Garbade, Kenneth. “Two Methods for Examining the Stability of
Regression Coefficients,” Journal of the American Statistical Asso­
ciation (March 1977), pp. 54-63.
Gilbert, R. Alton. “Operating Procedures for Conducting Monetary
Policy,” this Review (February 1985), pp. 13-21.
Goodfriend, Marvin. “ Discount Window Borrowings, Monetary Pol­
icy, and the Post-October 6, 1979 Federal Reserve Operating
Procedure,” Journal of Monetary Economics (September 1983),
pp. 343-56.
Plosser, Charles I., G. William Schwert and Halbert White. “ Dif­
ferencing as a Test of Specification,” International Economic Re­
view (October 1982), pp. 535-52.

REFERENCES
Board of Governors of the Federal Reserve System.
Reserve Discount Window,” 1980.

Federal Reserve Bank of New York. “ Monetary Policy and Open
Market Operations in 1985,” Quarterly Review (Spring 1986), pp.
34-53.

Polakoff, Murray E. “ Reluctance Elasticity, Least Cost, and Member-Bank Borrowing: A Suggested Integration,” Journal of Finance
(March 1960), pp. 1-18.
Riefler, Winfield. Money Rates and Money Markets in the United
States (Harper and Brothers, 1930).
Roley, V. Vance. “ Market Perceptions of U.S. Monetary Policy
since 1982,” Federal Reserve Bank of Kansas City Economic
Review (May 1986), pp. 27^t0.

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS
Thornton, Daniel L. "The Discount Rate and Market Interest Rates:
Theory and Evidence," this Review (August/September 1986), pp.
5-21.

VanHoose, David D. “A Note on Discount Rate Policy and the
Variability of Discount Window Borrowing," Journal of Banking &
Finance (December 1987), pp. 563-70.

_________ “ Lagged and Contemporaneous Reserve Accounting:
An Alternative View,” this Review (November 1983), pp. 26-33.

_________ “ Discount Rate Policy and Alternative Federal Reserve
Operating Procedures in a Rational Expectations Setting,” Board
of Governors of the Federal Resen/e System, Finance and Eco­
nomics Discussion Series, 12 (February 1988).

________. “The Discount Rate and Market Interest Rates: What’s
the Connection?” this Review (June/July 1982a), pp. 3-14.
_________ “Simple Analytics of the Money Supply Process and
Monetary Control,” this Review (October 1982b), pp. 22-39.

Wallich, Henry C. “ Recent Techniques of Monetary Policy,” Fed­
eral Reserve Bank of Kansas City Economic Review (May 1984),
pp. 21-30.

Appendix A
Complete Results for a Simple Model of the
Reserves Market
T h is ap p en d ix develops the resu lts stated in th e text
in term s o f a sim ple m odel o f the m oney stock. T he
m od el co n sists o f the follow ing eq u atio n s:
(All TKtl = a„ - a , I I P, + u
(A2I B o rr = b„ + b,(FFR —DR) + v
(A3! TR„ = NBR + Borr,

and
(A4) TR„ = TR„,

w h ere TR d en o tes total reserves an d th e su b scrip ts
“d " and “s" d en o te "d e m a n d ” an d “su p p ly .”' Borr
d en o tes th e am o u n t o f borrow ings an d NBR th e su p ­
ply o f n onborrow ed reserves, w h ich is assu m ed to be
co n tro lled by the Fed. FFR an d DR d en o te th e fed eral
fund s and d isco u n t rates, respectively, an d u an d v are
ran d o m errors su ch that E(u) = E(v) = E(uv) = 0.
E qu atio n s A1 - A4 can be co m b in ed to yield the
ex p ressio n for th e eq u ilibriu m fed eral funds rate
(A5) FFR = - X - ' I N B R + (b0- a „ ) - b,DR + (v - u ) ],

w h ere X = (a, + b ,). Figure A l-a show s th e ex p ected
value o f th is eq u ilibriu m eq u atio n .2 Given th e d isco u n t
rate an d th e stru ctu ral p aram eters, it show s all p o ssi­
ble co m b in atio n s o f FFR an d NBR su ch that the re ­
serve m arket is in equilibriu m . Figure A l-b reflects the
ex p ected value o f th e borrow ings fu n ctio n , equation
A2.

'The “time” subscript, t, is dropped for convenience.
2The curve slopes downward on the assumption that the interest rate
intercept is positive. A sufficient condition for this is that a0 > b0.



If th e F ed esta b lish es a borrow ings objective, Borr*,
th e fed eral fu n d s rate m u st equal FFR*, given the
d isco u n t rate. T h e eq u ilibriu m trad e-o ff curve in d i­
ca tes th at th e target level o f borrow ings ca n b e hit by
providing n o n b o rro w ed reserves equal to NBR*. T his
illu strates th e rela tio n sh ip b etw een a borrow ings o p ­
erating p ro ced u re an d a fed eral fund s rate targeting
p ro ced u re. If th e Fed d oes n ot resp o n d to sto c h a stic
sh ocks, th e varian ce o f borrow ings will be id en tical
u n d e r eith e r p ro ced u re, as will th e varian ce o f the
federal fu n d s rate.
D ifferen ces betw een th e two p ro ced u res em erge
w h en th e F ed a cts to offset d istu rb a n ces in b o rro w ­
ings, v. T h e resu lts d ep en d on th e tim e perio d over
w h ich th e d istu rb a n ces are operative an d th e a ssu m p ­
tion m ade abou t th e d istrib u tio n s of u an d v. For
exam ple, if sh o ck s o c c u r e a c h day an d if v an d u are
w h ite noise, su ch shifts essen tially will b e im p ossible
n
n
to offset. F u rth erm o re, b eca u se 2 u/n an d S v/n
i= 1
i= 1
a p p ro a ch zero as n g ets large, th ere is no n eed to offset
th ese shifts if th e p lan n in g h o riz o n is fairly long. Over
sh o rte r p erio d s su ch as a reserve p eriod (one w eek
before Febru ary 1984 an d tw o w eeks thereafter), th ese
errors will seld om “average o u t;” th erefore, it m ay be
d esirab le to offset part o f th e se sh o ck s. Also, th ese
sh o ck s m ay exh ibit p e rsisten ce, e.g., u, = cp,u,_, + e,
and v, = tp,v,_, + T|t, w h en e, an d r|t are w h ite n oise. In
th is case, th e Fed m ay also find it advantageous to
offset som e shifts d uring th e reserve period (or, for
that m atter, over a so m ew h at sh o rte r or lo n g er period)
d ep en d in g on th e m ag n itu d e o f tp, an d ip,.

45

FEDERAL RESERVE BANK OF ST. LOUIS

The M odel with Complete Adjustment to
Shocks
T h e borrow ings o p eratin g p ro ced u re can be differ­
en tiated from an in terest rate targeting p ro ced u re by
co m p arin g the ap p rop riate resp o n se to sh o ck s in
eith e r borrow ings o r th e fed eral funds rate u n d e r ea ch
p ro ced u re. Initially, th is is d on e u n d er th e a ssu m p ­
tion th at the Fed co m p letely offsets all sh o ck s.
U nder the borrow ings p ro ced u re, th e ap p rop riate
re sp o n se to sh o ck s is to ch an g e n o n bo rro w ed re ­
serves in a cc o rd a n ce w ith th e rule:
(A6) dNBR = u + (a,/b,)v.3
T hu s, no n bo rro w ed reserves sh o u ld ch an g e d o llar for
d ollar w ith a sh o ck to th e d em an d for to tal reserves
an d by a larger o r sm aller am o u n t (d epend ing on th e
relative m ag n itu d es o f a, and b,) for a sh o ck to borrow ­

JANUARY/FEBRUARY 1988

ings. T h e se ca s e s are illu strated in figures A2 an d A3.
In figure A2, a fully a n ticip a te d in crea se in th e d em an d
for total reserves shifts th e m arket eq uilibriu m curve
by u; th e borrow ings fu n ctio n rem ain s u n affected by
th is sh ock. C on sequen tly, th e target level o f th e fed eral
fund s rate is u n ch an g ed , b u t NBR is in crea sed bv u.
In figure A3, a positive value o f v shifts the bo rro w ­
ings fu n ctio n to th e right by v an d th e m arket eq u ilib ­
rium curve to th e left by v. As a result, th e level o f the
fund s rate th at is co n siste n t w ith th e borrow ings o b ­
jective is lower- an d n on borrow ed reserves m u st be
ex p an d ed by (a,/b,) to bring th e fund s rate dow n
en ou gh to m ain tain borrow ings at th e target level. If
th e Fed fully offsets shifts in th e d em an d for total
reserves, n e ith e r borrow ings n o r th e fu n d s rate will
ch an g e. If th e borrow ings fu n ctio n shifts, how ever,
borrow ings w ould rem ain at th e ir target level b u t th e
fed eral fu n d s rate w ill ch an ge.
U nder a fed eral fu n d s rate targeting p ro ced u re, the
ap p rop riate ru le for ad ju stin g n o n b o rro w ed reserves
w ould b e

3This rule is obtained by substituting A5 into A2, totally differentiating
the result and setting it equal to zero. Technically, the result is dNBR
= du + (a,/b,)dv; however, since the results are presented about the
expected value, du and dv have been replaced with u and v.

http://fraser.stlouisfed.org/
46 Bank of St. Louis
Federal Reserve

(A7I dNBR = u - v.
Note th at th e resp o n se to a sh o ck in to tal reserves
d em an d is th e sam e as u n d e r th e borrow ings op erat-

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

F ig u re A 2

The Effect of a Rise in v u nde r a B o r r o w i n g s O p e r at i n g P r oc edu r e

F ig u re A 3

The Effect of a Rise in v u nder a Bo rr o w i n g s Pr oc e d u r e




47

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Figu re A 4

The Ef fect of a Rise in v u nder a F ed er a l Rat e T ar get i ng P r o c e d u r e

ing p ro ced u re, eq u atio n A6. T h e d ifference in th e two
p ro ced u res co m es in th e re sp o n se to shifts in the
borrow ings fu n ctio n , in th e ca se of an in terest rate
target, th e Fed offsets th e effect o f an in crea se in the
borrow ings fu n ctio n by re d u c in g n o n borrow ed re ­
serves by v (illustrated in figure A4), w hile u n d e r a
borrow ings op eratin g p ro ced u re, th e Fed in crea ses
n o n bo rro w ed reserves by (a,/b,) o f th e shock.
U nder an in terest rate target, if th e Fed offsets all
shifts in th e d em an d for total reserves, n eith e r b o rro w ­
ings n o r th e fund s rate will deviate from th e ir target
levels (as u n d e r th e borrow ings p ro ced u re). If th e Fed
offsets shifts in th e borrow ings fu n ction , th e fund s rate
will not vary; how ever, th ere will be variability in
borrow ings.

The M odel with Incom plete Adjustment
to Shocks
T h e above analysis is b ased on the assu m p tio n that
th e Fed h as p erfect foresight an d co m p letely offsets
sh o ck s to total reserves or borrow ings. Now assu m e
th at th e Fed only offsets part o f th e sh o ck s. T h at is,
eq u atio n A6 can be rew ritten as
(A8I dNBR, = 8u + 8(a,/b,)v,

http://fraser.stlouisfed.org/
48
Federal Reserve Bank of St. Louis

w h ere 8 rep rese n ts th e p ro p o rtio n o f sh o ck s w h ich
th e Fed offsets over a given p lan n in g h orizo n , 0
8 *£
1. 8 = 1 is th e co m p lete a d ju stm en t m odel, 8 = 0
rep rese n ts a m od el in w h ich th e F ed m akes n o a t­
tem p t to offset sh o ck s. 8 w ould likely in crea se w ith the
len gth o f th e plan n in g h orizon .
T h e varian ce o f borrow ings an d th e fun d s rate u n ­
d er a borrow ings op eratin g p ro ced u re can b e ex ­
p ressed as
(A9I Var( B o rr |Borr*) = b 2(l - S)2 \~2 a 2
+ [1 —b, X~'(l + 8a,/b,)l2 ct;

and
(A10) VarlFFR |Borr*) = (1 - 8)2 \ “2 of, + X“-(l + 8 a./b^cr;,

respectively.4 Note th at VarlBorr |Borr*) eq u als zero
if 8 = 1, an d X.“2(b;CT;i + a;cr;) if 8 = 0. Also,
VarlFFR |Borr*) equals (of/b;) if 8 = 1, an d
+ &;) if

8 =

0.

T h e v ariance o f borrow ings an d th e fu n d s rate u n ­
d er a fund s rate target ca n b e ex p ressed as

These expressions are obtained by applying the definition of the
variance, e.g., E[Borr - E (Borr)]2, and replacing NBR - E(NBR)
with equation A8.

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS
(A ll) VariB orr |FFR*) = b 2(l - 8)2 \ - 2ct2

o f th e fu n d s rate u n d e r an in terest rate targeting
p ro ced u re.

+ (1 - I ) , \ - ' l l - 8 ) P o f

and
(A12I Var(FFR|FFR*) = \ “211 - 5 i2Ict2 +orl,

respectively. T h e VariBorr |FFR*) eq u als cr; if 8 = 1 and
\ "2(b2(r2 + a,of) if 8 = 0 , w hile th e VarlFFR |FFR*) equals
zero if 8 = 1 and \_~lerf, + erf) if 8 = 0.
A co m p ariso n o f eq u atio n s A9 and A l l show s that
th e variance o f borrow ings will be s m a lle r u n d e r a
borrow ings p ro ced u re th a n u n d e r an in terest rate
targeting p ro ced u re for 8 > 0 and equal for 8 = 0. T h e
v ariance o f th e funds rate will be la r g e r u n d e r a b o r­
row ings p ro ced u re th an u n d e r an in terest rate target­
ing p ro ced u re for 8 > 0 and equal for 8 = 0.
Also, it is p o ssible to estab lish co n d itio n s u n d e r
w h ich th e v ariance o f borrow ings wall be sm all relative
to th e varian ce o f th e fed eral fund s rate u n d e r a
borrow ings target. Solving eq u atio n A10 for II —8)2
A."’erf, and su b stitu tin g th e resu lt into A9, y ield s

W hile it m ay seem od d that the ex p ressio n s for the
relative v ariance do n ot d ep en d on ct2, th is result is
qu ite intuitive. V ariation in th e d em an d for total re ­
serves affects th e varian ce o f borrow ings only throu gh
its effect on th e variation o f th e federal fund s rate, not
d irectly th rou gh th e borrow ings fu n ctio n . C o n se­
quently, variability in th e d em an d for total reserves
only p ro d u ces variability in th e m arket in tere st rate;
given th e borrow ings fu n ctio n , th is tra n sla tes in to an
equal am o u n t o f ap p rop riately sca led variability in
borrow ings. T h is resu lt is illu strated in figure A5 u n d er
the assu m p tio n th at 8 = 0.
T h is also ex p lain s w hy co n tro l errors, i.e., NBR =
NBR* + a), w h ere w re p rese n ts a ran d om co n tro l error,
in crea se the variability o f b o th borrow ings an d the
fed eral fu n d s rate, b u t do n o t affect th e variability of
borrow ings relative to th e fu n d s rate. T h is is illu s­
trated in figure A5 alternatively as NBR above (NBR') or
below INBR") th e target level INBR*).

What I f u and v Are Correlated?
(A13) VariBorr | Borr*)

=

b; Varl FFR | Borr* l

+ [1 —b, \"M1 + 8a,/b,)]2 a —b2

11 + 8a,/b,)2 of.

Sin ce the term b2Varl FFR |Borr*) is m erely th e varian ce
of th e in terest rate exp ressed in u n its co m p arab le to
VariBorr |Borr*), after so m e sim plification, th e vari­
a n ce o f borrow ings relative to th e fed eral fund s rate
u n d e r a borrow ings op erating p ro ced u re ca n be w rit­
ten as
(A14) VariB orr |Borr*) — b 2 Varl FFR [ B o rr’ )
=

II

IA16) FFR = —\_l INBR + lb0- a „ ) - b.DR -

II —g) u].

0 12 erf — 0 2ct2,

w h ere 0 = b, \ “'ll + 8a/b,l. 0 is a m o n o to n ic in c re a s­
ing fu n ctio n o f 8. T h e righ t-h an d side o f A13 is negative
if 0 > 1/2. T h is co n d itio n will hold if b, > a, o r if 8 2= 1/2.
H ence, u n d e r so m e fairly gen eral co n d itio n s, th e vari­
a n ce o f borrow ings w ill b e less th a n th e v arian ce o f th e
federal fund s rate u n d e r a borrow ings op eratin g
p ro ced u re.
Likew ise, eq u atio n A12 ca n b e solved for \_2I1 —Sp a2
a n d th e resu lt su b stitu ted into eq u atio n A l l . T his
yield s
IA15) VariBorr |FFR*) - b 2 Varl FFR |FFR*)

= 11 —4*)- ct; — i|r a 2,
w h ere i|i = b, \ “'(1 —8). i|i is a m o n o to n ic d ecreasin g
fu n ctio n o f 8. T h e righ t-h and side o f eq u atio n A15 will
b e negative if i|/> 1/2. T h is w ill b e satisfied if b, > a, o r if
8 > 1/2. C on sequently, if th e Fed is able to offset m ore
th a n h a lf o f th e sh o ck s over its p lan n in g h orizon , th e
variance o f borrow ings will b e larger th a n th e variance



O ne p ossibility th at d eserves co n sid era tio n is the
ca se w h ere u an d v are co rrelated , th at is, sh o ck s to the
d em an d for total reserves, u, p ro d u ce a ch an g e in the
d em an d for b o rro w ed reseiv es, v. To see h o w this
affects th e resu lts, co n sid e r first th e sp ecial ca se in
w h ich th e sh o ck s are p erfectly co rrelated , e.g., v = £u.
A ssum e th at £ is positive, alth o u g h th is assu m p tio n is
n o t critica l to th e resu lts. Given th e se assu m p tio n s,
eq u atio n A5 ca n b e rew ritten as

Note th at (1 —£) is positive if 0 =£ £ < 1, zero if £ = 1 an d
negative if £ > 1. Given th is assu m p tio n , no shifts in th e
borrow ings fu n ctio n are in d ep en d en t o f shifts in th e
d em an d for to tal reseiv es. H ence, th e d ifference that
th e co rrela tio n b etw een th e erro r term s m akes ca n be
see n by co m p arin g th e effect o f a ch an g e in u u n d e r
b o th a ssu m p tio n s. In th e m o d el th a t a ssu m ed in d e­
p e n d e n ce , th e equ ilibriu m in tere st rate curve shifted
to th e right by u w h ile th e borrow ings fu n ctio n did n ot
shift, as in figure A2. U nd er p erfect positive co rrelatio n ,
th e m arket eq u ilibriu m curve shifts by (1 —£)u, w hile
th e borrow ings fu n ctio n shifts by £u. T h e se shifts
d eterm in e th e ex ten t to w h ich o p en m arket o p era ­
tion s m u st b e u n d ertak en to stabilize borrow ings at
the target level.5 It also ca n be sh o w n th at th e assu m p -

5lf i < 0 and equal to - b, a,, nonborrowed reserves will not have to
change to stabilize borrowings at the target level. In this case, the
leftward shift in the borrowings function just cancels the effect of the
rightward shift in the equilibrium curve on nonborrowed reserves.

49

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

F ig u re A 5

A n Il l ustrati on of W h y the Var i a n ce of Bo r r o w i n g s Rel at i ve to the Fed er al F un ds Rat e is U na f f e ct e d by 6

FFR

N B R "N B R *

NBR'

NBR

(a)

tion o f p erfect co rrelatio n has n o effect on co n clu sio n s
abou t th e variability o f borrow ings and/or th e federal
fund s rate u n d e r th e alternative o p eratin g p ro ced u re.
W hat if th e sto c h a stic d istu rb an ces are not p erfectly
co rrelated ? F or exam ple, assu m e th at v = £11 + t ) ,
w h ere T| is id en tically and in d ep en d en tly d istribu ted
w ith a m ean zero an d a variance cr2. Given th is a s ­
su m p tio n , th e erro r term o f eq u atio n A5 is sim ply
—
+ (1 —£;) u]; th e sam e as th at o f A5 ex cep t th at u
is rep laced by (1 —£)u and T| re p laces v. C onsequently,


50


all o f th e previously stated resu lts h old .6

6The intuition for this is straightforward. The variability of borrowings
under a borrowings operating procedure relative to that under a
federal funds rate operating procedure depends only on the variabil­
ity of the borrowings function. Since variability of the borrowings
function is the same under any of these assumptions, i.e., v = f;U or
even v = £u + r), for both the borrowings and federal funds rate
targets, the assumption made does not affect the general conclusion
about the variability under these procedures. This is also the reason
the general conclusions about the variance of borrowings relative to
the federal funds rate under the borrowings operating procedure are
unaffected by this assumption.

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

Appendix B
A More Detailed Analysis of the Effect of a Change in the
Borrowings Assumption on the Federal Funds Rate
T h e p u rp o se o f th is ap p en d ix is to p re sen t d etailed
resu lts on th e effect o f a ch an g e in th e borrow ings
assu m p tio n on th e federal fund s rate. O ne w ay to
calib rate su ch effects is to estim ate a red u ced form
eq u atio n for th e le v e l o f th e fed eral fu n d s rate (FFR):
( B l)

FFR,

=

a

+ PDR, +

jjiB A ,

+

e„

w h ere DR and BA d en o te the level o f th e d isco u n t rate
and borrow ings assu m p tio n . U nder a strict borrow ed reserves op eratin g p ro ced u re, (3 sh ould be positive
an d equal one. OLS estim ates o f eq u ation B l are re ­
ported in th e top h a lf o f table B l for th e LRR and CRR
period s. T h ree significant a sp e cts o f th e se resu lts d e­
serve p articu lar atten tio n . First, th e h yp o th esis th at (3
= 1 is re je cte d at th e 5 p e rce n t level d uring b o th
period s. Seco n d , th e Q statistic d o es not in d icate loword er serial co rrelatio n during th e LRR period, but
d oes in d icate it during th e CRR period. N evertheless,
the resid u als sh ow a p ro n o u n ced qu arterly season al
spike d uring th e LRR period (clearly evident from
ch art 2 o f the text). Third , th e stand ard erro r o f th e
eq u ation in crea ses d ram atically during the CRR p e ­
riod, in d icatin g in creased variability o f th e FFR u n d e r
CRR. (This is tru e w h eth er w eekly o r reserve period
data are used.)

B eca u se o f th e sea so n a l spike d uring th e LRR period
an d serial co rrelatio n o f th e resid u als d uring th e CRR
period, th e eq u a tio n s w ere reestim a ted in clu d in g
lagged d ep e n d en t variables. T h e resu lts are rep o rted
on th e b o tto m h alf o f table B l. (Four lags o f FFR are
in clu d ed d uring the CRR period ; in ad d ition, FFR,_13 is
in clu d ed d uring th e LRR period.) D uring the LRR
period, th e co efficien t on BA in crea sed som ew hat,
alth ough its t-ratio d eclin ed . Also, th e estim a te o f (3
d eclin ed su b stan tially an d th e h yp o th esis that p = 1
is re je cte d at very low sig n ifican ce levels. F or th e
CRR period, th e estim ated co efficien t on BA d e­
clin e d by n early tw o-th ird s an d th e t-ratio d eclin ed
dram atically.
T h ere are several rea so n s for q u estio n in g th e e sti­
m ates from the level eq u atio n s. T h e first rea so n relates
to the tim e-series p ro p erties o f th e individual series
th em selves. BA is highly au to co rrelated , as table B2
in d icates. T h e fact th at th e levels o f BA an d FFR are
highly a u to co rrela ted affects th e relatio n sh ip b etw een
th em . T h is is evident in th e sim p le co rrelatio n co ef­
ficien ts given in table B3. T h e sim p le co rrela tio n o f FFR
an d BA is h ig h er th a n th at o f FFR an d a ctu a l a d ju st­
m en t plus seaso n al borrow ing, Borr, d uring th e LRR
period ; how ever, th e co rrela tio n co efficien t o f first

Table B1
Estimates of Equation B1
Constant

DR

BA

October 1,1982February 1, 1984

2.23*
(2.37)

.72*
(6.78)

February 2, 1984December 31, 1986

0.81*
(2.22)

October 1, 1982February 1,1984
February 2,1984December31,1986

Period

DL

Q3

SEE

.0017*
(9.60)

—

4.69

.3041

.80*
(13.4)

.0029*
(10.95)

—

38.05*

.3763

5.39*
(5.23)

.42*
(2.02)

.0022*
(6.64)

5.71*'

5.95

.2616

0.53
(1.76)

.28’
(2.29)

.0011*
(2.90)

10.13*2

0.78

.3072

’ Indicates statistical significance at the 5 percent level.
’Test that FFR,_, - FFR,_„ and FFR,_13are jointly zero.
2Test that FFR,., - FFR,_4are jointly zero.
3Test for white noise residuals, distributed x2(6).




51

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Table B2
Autocorrelations of Time-Series Variables for Reserve-Period Data
Lag
1

2

3

4

5

FFR

.55

BA
Borr

.98
.37

.53
.95
.37

.48
.92
.39

.46
.88
.25

.37
.85
.27

Variable

6

7

8

9

10

11

12

13

14

15

.17
.60
.17

.17
.55

.12
.49

.34
.44

.16

.14

.42

.00
.39
.15

.01
.34
.16

.48
.40
.12

.43
.33
.10

.38
.26
.09

.33
.20
.04

.28
.14
.07

.23
.09
.00

October 1,1982 - February 1,1984
.36
.81
.31

.32
.76
.36

.23
.71
.27

.23
.66
.09

February 2,1984 - December 31,1986
FFR

.97

.94

.89

.84

BA
Borr

.97
.50

.93
.32

.89
.25

.83
.27

.78
.76
.19

.72
.69
.17

.66
.61
.09

Table B3
Simple Correlations of Time-Series
Variables for Reserve-Period Data
Variables

BA

Borr

ABA

ABorr

October 1,1982 - February 1,1984
FFR
AFFR

.605*
—

.540*
—

—
.166

—
.533*

February 2,1984 - December 31,1986
FFR
AFFR

.903*
—

.492*
—

—
.175

—
-.0 2 4

* indicates statistical significance at the 5 percent level.

d ifferences o f FFR an d BA is d ram atically different
from that o f th eir levels. T h is is not true, how ever, of
th e co rrelatio n b etw een B o rr an d FFR an d A Borr and
AFFR. F o r th e CRR period, w h en th e ir au to co rre la ­
tio n s m a tch closely, th e co rrelatio n b etw een FFR an d
BA is high. Yet in first-d ifference form , th e co rrelatio n
is essen tially the sam e as d uring th e LRR period an d is
n ot statistically significant.
A se c o n d reaso n to be cau tio u s o f th e level eq u ation
resu lts h as to do w ith th e lon g -ru n stability o f th e
b orrow ings fu n ctio n itself. T h e borrow ings a ssu m p ­
tio n d o es n o t rep rese n t an exog en ou s su p p ly o f b o r­
row ings; m o re precisely, it is an exog enou s target level
th at th e Fed attem p ts to in d u ce d ep ository in stitu ­
tio n s to hold by altering th e su pply o f n on borrow ed

http://fraser.stlouisfed.org/
52
Federal Reserve Bank of St. Louis

.60
.54
.07

.54
.47
.10

reserves. C on sequen tly, actu al borrow ings can , and
do, deviate from th e d esired level. N evertheless, over a
lo n g er tim e period, th e average level o f borrow in gs ca n
b e clo se to th e d esired level. T h is is esp ecially likely if
a d ju stm en ts are m ad e to n o n b o rro w ed reserves or if
th e borrow ings a ssu m p tio n itself is ch a n g ed to keep it
in lin e w ith a ctu al borrow ings levels.
T h erefore, w h en th e le v e l o f th e fu n d s rate is re­
g ressed on th e le v e l o f th e borrow in gs assu m p tio n ,
th ere is a te n d en cy to retrieve this lon g -ru n rela tio n ­
ship to a g reater or sm aller degree, d ep en d in g on
h ow clo sely th e borrow ings a ssu m p tio n m im ics a c ­
tual borrow in g s.1
In o rd er to m ore clo sely ca p tu re the effect of an
exo g en o u s c h a n g e in th e borrow ings assu m p tio n on
th e fu n d s rate, first d ifferen ces o f th e fu n d s rate are
reg ressed on first d ifferences o f th e borrow in gs a s­
su m p tio n . T h is sh o u ld yield c o n siste n t estim ates of
th e im m ed iate re sp o n se o f th e fed eral fu n d s rate to an
exo g en o u s ch a n g e in th e borrow ings assu m p tio n ,
even if th e level sp ecifica tio n is co rrect.- M oreover, it

'Augmented Dickey-Fuller tests for stationarity applied to borrowings
and the funds rate indicate that both series are integrated of order
one, i.e., 1(1) for the LRR period. When the test is applied to the
residuals from OLS estimates of equation 1, however, the results
indicate that borrowings and the funds rate are cointegrated in the
Engle-Granger (1987) sense. The augmented Dickey-Fuller test
indicates that BA and FFR are l(2) over the CRR period. Yet the test
indicates that the residuals from equation 1 estimated over this
period are stationary.
The OLS estimate of b,, of equation 1 from the text for the LRR
period is 471. This yields an implied coefficient estimate of (3 of
equation B1 equal to .0021 (1/471). The implied estimate of 0 for the
CRR period using resen/e-period data is .0038 (1/260).
2See Plosser, Schwert and White (1982).

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Table B4

Table B5

t-ratios for ABAj from equation B1

Estimation of the Funds Rate Equations
with ABA Partitioned as in Table B1

October 1,1982February 1,1984

February 2,1984December 31,1986

Reserve-Period'

Weekly

Reserve-Period

— 2.78 AN
1.23 A
-1 .9 8 AN
-1 .6 6 AN
-2 .1 4 AN
-0 .4 8
0.42
-0.56
0.56
0.34
-0.81
0.52
0.96
-1 .1 7 AN
-0.3 5
- 1.76 AN
0.89

0.03
1.00
-0 .3 7
2.69 A
1.13
2.12 A
0.22
0.31
-0 .6 4
0.27
-0 .3 4
0.83
0.26
-0 .3 8
0.37
-0 .5 4
-1 .0 0
0.65
0.53
-0 .3 5
0.13

0.64
0.66
1.80 A
2.51 A
0.75
1.57 A
-0 .8 7
-0 .1 5
-0 .2 0
-0 .2 0
0.43
-0.21
-0 .1 7
-0 .3 9
0.06
0.19
0.06
0.33
0.35
-2 .0 9 AN
0.35

(A,
M-2
M-3
M-4
fJ-5
M"6
M"7
M-8
M-9
M-10
M-12
M-13
M-14
ms
M"16
M-17
M-18
M-19
M-20
M*21

NOTE: A denotes those that are positive and significant at a 25
percent level. AN denotes those that are negative and
significant at a 25 percent level.
'During the LRR period, the reserve period was one week.

sh o u ld avoid sp u rio u s co rrelatio n often ex p erien ced
w h en th e levels o f n o n statio n ary series are used.
Finally, b eca u se th e borrow ings assu m p tio n is
ch an g ed infrequently, ch an g es in th e borrow ings a s­
su m p tio n can be p artitio n ed into th o se th at do have a
significant effect on th e fed eral funds rate and th o se
th at do not. T his is d on e bv estim atin g th e eq u ation :
K
1B2) AFFR, = a„ +

2

a,AFFK,_, + p„ ADRT, + (3, ADRNT,

i= 1

October 1,1982February 1,1984
Reserve-Period1

February 2,1984December 31,1986
Weekly

Reserve-Period

Constant

0.04
(0.98)

0.015
(0.47)

0.033
(0.87)

ADRT

0.20
(0.22)

0.247
(0.33)

0.082
(0.15)

ADRNT

2.280*
(3.60)

0.755*
(1.99)

0.891*
(2.36)

ABA-A

0.0087
(1.16)

0.0055*
(3.10)

0.0058’
(3.21)

ABA-B

0.0007
(0.77)

0.0011
(1.25)

0.0002
(0.26)

-0.0160*
(2.53)

—

-0.0076’
(2.37)

R2

0.5737

0.1875

0.2046

SEE

0.2803

0.3761

0.2951

6.57*

0.65

ABA-AN

F AFFRLAGS

15.81*

'indicates statistical significance at the 5 percent level.
'During the LRR period, the reserve period was one week.

borrow ings a ssu m p tio n over th e sam p le p eriod .3 Th e
estim ated stan d ard erro r from th is eq uation, cr;, m ea ­
su res th e co n d itio n a l varian ce o f AFFR for p eriod s
w h en ABA = 0. H ence, th e t-ratio for th e jth ABA
in d ica tes h ow m u ch th e fed eral fund s rate m oved
d uring th is period relative to period s w h en th e b o r­
row ings assu m p tio n w as u n ch an g ed .
T h e t-ratios for ea ch ABA are rep o rted in table B4.
T h e resu lts in d icate that, o f th e 17 ch a n g es in the
borrow ings a ssu m p tio n d uring th e LRR period, 10
w ere in v erse ly related to ch a n g es in the fed eral funds
rate. D uring th e CRR period , eith e r seven or eight of
th e 21 ABAs w ere inversely related to th e fu n d s rate,
d ep en d in g on w h e th e r w eekly or reserv e-p eriod data
are u sed . T h e resu lts in table B4 ca n be u sed to
partition ABA in to th o se th at have a positive signifi-

L
+

X

(JLiABA;, +

e ,,

j= l

w h ere ABA,: takes on th e value o f th e jth ch an g e in the
borrow ings assu m p tio n d uring th e perio d an d is zero
o th erw ise. L d en o tes th e n u m b er o f ch an g es in the



3lf daily data were used, L is equal to the number of changes on the
borrowings assumption over the sample period. When weekly or
reserve-period data are used, the data are averaged on a pro-rata
basis. Consequently, L denotes the number of weeks or reserve
periods that are affected. This is usually larger than the number of
changes in the borrowings assumption itself.

53

FEDERAL RESERVE BANK OF ST. LOUIS

can t effect on th e fund s rate, ABA-A, th o se th at have a
significant negative effect, ABA-AN, and all others,
ABA-B. T h is w as d one by in clu d in g in th e A o r AN
groups all ABAs th at are significant at th e 0.25 p e rce n t
level u sing stand ard analysis. T h o se ch an g es in BA
th at are in th e A or AN groups are d esignated co rre ­
spond ingly in table B4.
E stim ates o f th e sam e b asic eq u atio n w ith th e p arti­
tioned data are p resen ted in table B5. In all cases,


http://fraser.stlouisfed.org/
54 Bank of St. Louis
Federal Reserve

JANUARY/FEBRUARY 1988

excep t th e single observ ation in th e A group d uring
th e LRR period, th e co efficien ts on th e A an d AN
partitio n s are sign ifican t at th e 5 p e rce n t level. M ore
im portantly, th e co efficien ts on th e ch a n g es in th e
borrow ings assu m p tio n in th e B p artition , w h ich a c ­
co u n t for th e vast m ajo rity o f ch a n g es in th e borrow ­
ings assu m p tio n , w ere uniform ly in sign ifican t at th e 5
p e rce n t level. T h is evid en ce in d ica tes th at th e link
b etw een ch a n g es in th e borrow ings assu m p tio n and
th e fed eral fund s rate is, at best, w eak.

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Farm Policy and Mandatory Supply
Controls — The Case of T obacco
Kenneth C. Carraro

F
ROM 1980 throu gh 1986, th e U nited States sp en t
$43.9 billion in d irect p ay m en ts to farm ers and $52.3
billion on o th er p rice su p p o rt program s.1D esp ite su ch
exp en d itu res, th e U.S. farm se c to r has ex p erien ced a
severe d ow nturn. Falling exports, d eclin in g farm land
values, high rates of farm lo an d elin q u en cie s an d
in creasin g d e p e n d en ce on governm ent su p p o rt p ay ­
m en ts w ere visible sym ptom s o f th e farm s e c to r’s
difficulties.
B eca u se o f th e great ex p en se an d th e ap p aren t
failure o f farm program s, som e policym akers have
called for the use o f m an d ato iy su pply co n tro ls to
lim it crop p ro d u ctio n an d raise prices.- Advocates
assert th at su ch co n tro ls co u ld gu aran tee farm ers a
“fair” p rice an d im prove th eir in co m es, w hile d rasti­
cally cu ttin g th e co st o f farm program s and elim in a t­
ing farm co m m o d ity su rp lu ses.

Kenneth C. Carraro is an economist at the Federal Reserve Bank of St.
Louis. Dawn M. Peterson provided research assistance.
'U.S. Department of Agriculture, History of Budgetary Expenditures of
the Commodity Credit Corporation, Book 2, and Agricultural Outlook
(December 1987), p. 53, table 32.
2The Harkin-Gephardt “ Save the Family Farm Act” is the most
prominent domestic example of mandatory supply control legisla­
tion currently being debated in Congress. In 1986, Congress allo­
cated $10 million for the study of mandatory controls and the polling
of farmers. Mandatory supply controls have recently been proposed
in the European Economic Community to limit milk production.



T h is article ex am in es th e effects o f m an d atory su p ­
ply co n tro ls. T h e an alysis beg in s w ith a th eo retica l
d iscu ssio n o f th e effects o f m an d ato ry su pply co n tro ls
o n eco n o m ies th at are clo sed to in tern atio n al trade
an d th o se th at engage in in tern atio n al trade. Next, the
ex p e rie n ce o f th e U.S. to b a c co in d u stry an d its m a n d a ­
tory supply co n tro ls is exam in ed .3 Finally, th e key
p o in ts from th e th e o retica l d iscu ssio n an d th e U.S.
to b a cco in d u stry 's e x p e rie n ce are co m b in ed w ith sp e ­
cific facts abou t U.S. cro p s to suggest th e likely c o n s e ­
q u en ces o f th e su p p ly legislatio n cu rren tly u n d e r c o n ­
sid eration .

THE ECONOMICS OF SUPPLY
CONTROLS IN A CLOSED ECONOMY
Supply co n tro l p rogram s are d esign ed to in crea se
th e p rice o f a good above its free m arket p rice by
restrictin g th e qu an tity o f th e good th at re a ch es the
m arket. T h e su pply re strictio n s typically are e sta b ­
lish ed by a governm ent ag en cy or a co n so rtiu m of
p ro d u cers. T h e O rganization o f P etroleu m Exporting
C ou n tries (OPEC) is o n e exam p le o f a group o f p ro ­
d u cers w ho agree (usually) to re strict p ro d u ctio n as a
m ean s o f secu rin g a h ig h er p rice for cru d e oil.

3U.S. tobacco policy has used mandatory supply controls since the
1930s.

55

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

p rice. If th e qu an tity o f a p ro d u ct d em an d ed ch a n g es
p ro p o rtio n ately less (in a b so lu te value) th a n th e
ch an g e in p rice, th e d em an d is referred to as in elastic.

Figure 1

Supply Controls in a Closed Ec o n o m y
Sin ce a 1 p e rce n t in crea se in p rice ca u ses le ss th a n a
1 p e rce n t d ecrea se in quantity d em an d ed w h en d e­
m and is in elastic, th e p rice in cre a s e ca u se s total reve­
n u e to in crea se. Conversely, a p rice d ecrea se ca u ses
total revenue to d ecrea se w h en d em an d is in elastic.
T h e effects on total revenue o f p rice ch a n g es are
reversed w h en d em an d is elastic. E la stic d em an d ex ­
ists w h en th e qu an tity o f a p ro d u ct d em an d ed
ch a n g es p ro p o rtio n ately m o re (in a b so lu te value) th an
th e ch an g e in p rice. Sin ce a 1 p e rce n t in crea se in p rice
ca u ses a m o re th a n 1 p e rce n t d ecrea se in qu an tity
d em an d ed w h en d em an d is elastic, th e p rice in crea se
ca u ses total revenue to d ecrea se. Conversely, a p rice
d ecrea se ca u ses total revenue to in crea se w h en d e­
m an d is elastic. A final possibility, know n as u n ita iy
elasticity, is that a 1 p e rce n t ch a n g e in p rice lead s to a
1 p e rce n t ch a n g e in q u an tity d em an d ed , w h ich h as no
effect on total revenue.

Supply
Limit

Quantity

Figure 1 d em o n strates how p rices are d eterm in ed
in an eco n o m y that is clo sed to in tern atio n al trade
an d ho w su p p ly co n tro ls ca n in crea se th e p rice o f a
good above its free m arket level. T h e su p p ly curve,
lab eled S, rises upw ard an d to th e right, ind icatin g
th at p ro d u cers w ill su pply larger qu an tities o f a good
as its p rice is in creased . T h e sh o rt-ru n d em an d curve,
labeled Ds, slop es d ow nw ard to show th at co n su m ers
will buy sm aller q u an tities o f a good as its p rice rises.
In a free m arket, the in tersectio n o f the supply and
d em an d curves at p o in t A d eterm in es that th e p rice
w ould be P, w hile th e qu an tity su p p lied w ould equal
th e qu antity d em an d ed , at Q,. Sin ce th e qu an tity o f the
good su p p lied to th e m arket at th at p rice exactly
satisfies co n su m e r d em and , n eith e r p ro d u cers n or
co n su m ers have an incen tive to ch an g e th e ir p ro d u c­
tio n o r co n su m p tio n p attern s.
By im p osing a su p p ly lim it at Q ,, th e p rice ca n b e
in crea sed from P, to P,. T h is w ou ld b en efit p ro d u cers,
how ever, only if it in crea sed th e ir profits. Sin ce p ro ­
d u ctio n d eclin es, th e total co sts in cu rred by p ro ­
d u cers w ill d eclin e also. As long as total revenue is n o t
re d u ce d by an am o u n t larger th a n th e re d u ctio n of
total co sts, profits will rise.
T h e ch an g e in total revenue resu lting from a p rice
ch an g e d ep en d s u p on th e p rice elasticity o f dem and .
T h e p rice elasticity o f d em an d m e a su res th e re sp o n ­
siveness o f th e qu antity d em an d ed to a ch a n g e in

http://fraser.stlouisfed.org/
56
Federal Reserve Bank of St. Louis

In figure 1, th e su p p ly con trol, w h ich re d u ce d th e
qu an tity su p p lied from Q, to Q,, ap p ears to have
ca u sed th e p rice ap p roxim ately to d ou ble from P, to P,.
T h e qu an tity d em an d ed , how ever, a p p ears to have
d ecrea sed m u ch less. In o th e r w ords, th e d em an d is
co n sid ered to b e in elastic in th at p rice range. W hen
th e d em an d for a p ro d u ct is in elastic, a su pply co n tro l
program in crea ses th e total revenue o f p ro d u cers.
Sin ce total co sts will have fallen also, profits m u st
in crease.
W h en th e d em an d for a p ro d u ct is elastic, a su pply
co n tro l program w ould re d u ce th e qu an tity d e ­
m a n d ed p ro p o rtio n ately m o re th a n th e p rice in ­
crea se. T h e re d u ctio n in total revenue m akes it p o ssi­
ble th at th e su p p ly co n tro l program co u ld lead to
re d u ce d profits. In general, a su p p ly co n tro l program
is b en eficial to p ro d u cers facing an in ela stic d em an d .
A variety o f factors in flu en ce th e elasticity o f d e ­
m an d for a p ro d u ct. O ne o f th e m ost im p o rtan t of
th e se is th e availability o f su b stitu tes for th e p ro d u ct. A
p ro d u c t’s d em a n d is m o re likely to b e elastic if a c c e p t­
able su b stitu tes fo r th at p ro d u ct exist. F o r exam ple,
th e p rice elasticity o f b e e f likely ex ceed s th a t o f g aso ­
lin e b e c a u se th ere are n u m ero u s su b stitu tes for b e e f
w hile th e re are few su b stitu tes for gasoline.
A n oth er extrem ely im p o rta n t in flu en ce on d em an d
elasticity is tim e. In th e sh o rt run, a p ro d u c t’s d em an d
is generally less ela stic th a n over th e lon g ru n b eca u se
co n su m ers find su b stitu tes or learn to co n serv e on the
co n su m p tio n o f th e p ro d u ct over tim e. D em an d b e ­
co m e s m o re ela stic th e lo n g er th e tim e p erio d as

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

F igu re 2

Supply Controls in a W orld Economy W ith Trade
Panel A

Panel C

Panel B

Domestic Eco nom y

Rest of

the

World

Entire

World

Supply
limit
Price

Price

'DA UDA«C

SA

SA
Quantity

co n su m ers read ju st th e ir co n su m p tio n p atte rn s.4
Figure 1 portrays th e effect o f ch an g es in d em an d
elasticity over tim e. T h e curve D Lportrays th e lon g-run
d em an d curve for th e p ro d u ct an d is m u ch flatter th an
th e sh o rt-ru n d em an d curve Ds. T h is reflects the
g reater elasticity th a t is co m m o n over th e lon g run.
T h e su p p ly co n tro l th at resu lted in th e d oublin g of
p rices from P, to P, in th e sh o rt ru n is m arkedly less
ben eficial to p ro d u cers over th e long ru n. In th is case,
th e im p o sitio n o f th e su p p ly restrain t h as a relatively
sm all effect on th e p rice, raising it only to P.,. F u rth er­
m ore, it ap p ears that th e total revenue h as d eclin ed
th rou gh th e u se o f th e co n tro ls. T h e sh o rt-ru n strategy
th at ap p eared to in crea se profits m ay lead to low er
future profits if th e lon g-run d em an d b e co m e s elastic.

THE ECONOMICS OF SUPPLY
CONTROLS IN AN OPEN ECONOMY

Quantity

Quantity

d u ced in co u n tries o u tsid e o f th e co u n try (or group of
cou n tries) a ttem p tin g to in crea se retu rn s th rou g h a
su pply co n tro l policy, an d 2) th at th e co n tro lled good
ca n b e trad ed b etw ee n co u n tries. In a clo sed e c o n ­
omy, a p ro d u ct ’s p rice is d eterm in ed solely by d o m es­
tic su p p ly an d d em an d . W ith th e ad d ition o f trade,
p rice d eterm in atio n o cc u rs in th e w orld ra th er th an
d o m estic m arket.
Figure 2 p ortrays p rice d eterm in atio n in th e w orld
m arket. Panel A re p rese n ts th e d o m estic m arket for a
good. Panel B re p rese n ts th e su pply an d d em an d of
th e p ro d u ct for all o th e r co u n tries in th e w orld. F i­
nally, p an el C is th e w orld econ om y, w h ich is derived
by horizon tally co m b in in g th e su pply an d d em an d
curves o f th e d o m estic an d rest-of-th e-w orld e c o n o ­
m ies.

So far, w e have fo cu sed o n a sim p le eco n o m y w ith ­
out in tern atio n al trad e to illu strate fu n d am en tal
p o in ts abou t su pply co n tro l program s. T h is sectio n
exp an d s th at analysis to in clu d e su p p ly co n tro ls in a
w orld eco n o m y w ith trade. T h e ad d ition o f trad e to
th e analysis im p lies: 1) th a t a p ro d u ct m ay b e p ro ­

Ignoring tra n sp o rta tio n co sts, th e equilibriu m p rice
for b o th th e d o m estic eco n o m y an d th e rest o f the
w orld is Pw. In th is ca se, th e eq u ilibriu m p rice is above
w hat the d o m estic p rice w ould have b een in a clo sed
eco n om y. A ccord in g to p a n el A, at th e w orld p rice,
d o m estic p ro d u cers su p p ly a larger qu an tity (QSA)
th a n d o m estic co n su m ers are w illing to p u rch a se
( Q da ). T h e d ifference b etw ee n th e se tw o is ex p o rted to
th e re st o f th e w orld w h ere, at Pvv, co n su m ers d em an d
a larger qu an tity (QDB) th a n p ro d u cers in th e rest o f the
w orld are w illing to su p p ly (Q SB) as sh o w n in p a n el B.

4For example, Houthakker and Taylor (1966) estimated the long-run
price elasticity for gasoline at - .7, while the short-run elasticity was
estimated to be much more inelastic at - .2.

T h e d o m estic eco n o m y in figure 2 is p ortrayed as
th e d o m in an t w orld su p p lier o f a p ro d u ct for w h ich
th e d em an d is in elastic. T h e im p o sitio n o f a su pply




57

FEDERAL RESERVE BANK OF ST. LOUIS

co n tro l in th e d o m estic eco n o m y at th e qu an tity Q (:
ch an g es th e w orld su pply from Su to Sc. T h is shift, in
turn, ca u ses th e w orld p rice to ju m p from Pu to P ,.
B eca u se o f su pply co n tro ls in th e d o m estic econom y,
th e qu antity su p p lied falls from Q SAto Q ,. At th e h ig h ­
er p rice o f P( , foreign p ro d u ctio n in crea ses from Q sli to
Q 's„; w hile foreign co n su m p tio n falls from Q„„ to Q 'UB.
As a resu lt o f th e se ch ang es, th e level o f exp o rts from
th e d o m estic eco n o m y to th e rest o f th e w orld d e­
clin es. T h e sh ares o f w orld trad e and w orld p ro d u c­
tio n h eld by th e d o m estic eco n o m y also d eclin e.
T h e lo ss o f sh ares o f w orld p ro d u ctio n an d trad e is a
p red ictab le o u tco m e o f a su pply co n tro l m easu re.
W hile an exporting co u n try m ight p refer n ot to lose its
sh ares o f w orld p ro d u ctio n and trade, it is m ore likely
to a cc e p t th e se lo sses if th e su pply co n tro ls result in
hig h er retu rn s to p ro d u cers. In figure 2, it appeal's that
retu rn s w ould b e in crea sed in th e sh o rt ru n b ec a u se
th e in e la stic w orld d em an d curve an d th e in elastic
foreign su pply curve resu lt in h ig h er total revenue for
d o m estic p rod u cers.'1
T h e se sh o rt-ru n retu rn s will erode, how ever, b e­
ca u se th e p rice elasticities o f bo th d em and an d supply
in crea se over tim e. A given d o m estic su pply co n tro l
resu lts in a sm aller p rice in crea se in th e long ru n th an
in th e sh o rt ru n. T h is effect is even m o re p ro n o u n ced
w ith in tern atio n al trad e b e c a u se th e elasticity o f for­
eign, as w ell as d o m estic, su pply generally in crea ses
over tim e. In th e sh o rt run, p ro d u cers are u n ab le to
resp o n d fully to a p rice in crease b ec a u se the capital
b ase u sed for p ro d u ctio n is fixed. Over a lo n g er period,
p ro d u cers can in cre a se o u tp u t by add ing p ro d u ctio n
cap acity, im proving tech n o lo g y and ad op tin g n ew
technology. T h is long-run foreign su p p ly resp o n se
co n trib u te s to th e d eclin e in th e sh are o f w orld p ro ­
d u ctio n and trad e o f th e d o m estic co u n try by in c re a s­
ing foreign p ro d u ctio n and, in th e p ro cess, red u cin g
th e d em an d for th e d o m estic c o u n tiy ’s exp orts. T h e
foreign su pply re sp o n se b e c o m e s in creasin g ly m ore
im p ortant b ec a u se o f th e grow ing foreign sh are of
w orld p ro d u ctio n .

5The example of OPEC is instructive at this point. When OPEC
reduced production as a means of increasing the price of crude oil, it
was logical to expect that its share of both oil exports and production
would fall. While its share fell, it was able to greatly increase its
returns because of the elasticities of world demand and supply. With
a lack of acceptable energy sources as substitutes, the world
demand for crude oil was extremely inelastic. The world supply of oil
also was extremely inelastic because of the small share of world
production held by non-OPEC countries and the difficulty, expense
and time required to find and tap new oil reserves. If non-OPEC
countries had been able to expand production easily and quickly in
response to higher prices, the price increases would not have been
as great.

http://fraser.stlouisfed.org/
58
Federal Reserve Bank of St. Louis

JANUARY/FEBRUARY 1988

In sum m ary, th e in tro d u ctio n o f in tern atio n al trade
m akes th e d ecisio n to u se su pply co n tro l m easu res
d ep en d en t on th e elasticity o f w orld d em a n d an d
w orld supply. It is im p o rtan t to n o te that, w hile the
elasticity o f foreign (rest-of-the-w orld) su p p ly is im ­
portant, it is th e en tire w o rld ’s elasticity o f su pply that
d eterm in es if a d o m estic su pply co n tro l program will
be effective. F or exam ple, foreign su pply m ay be very
elastic over a sm all range; b u t if foreign p ro d u ctio n
re p rese n ts only a sm all sh are o f to tal w orld p ro d u c­
tion, th e d o m estic su pply co n tro l program m ay still be
very profitable. T h is is tru e b e c a u se th e foreign supply
resp o n se, w h ile very elastic, m ay have only a sm all
effect on th e total qu an tity su p p lied in th e w orld if
d o m estic p ro d u ctio n dwarf’s foreign p ro d u ctio n . A
natu ral c o n se q u e n ce of d o m estic su pply co n tro ls and
foreign su pply elasticity, how ever, is an in crea se in the
foreign sh are o f w orld p ro d u ctio n an d a resu lting
in crea se in th e w orld su pply elasticity.

THE ORIGINAL TOBACCO PROGRAM
T h e cu rren t to b a cco program has its roots in the
farm legislation o f th e 1930s know n as th e Agricultural
A d ju stm en t Act (AAA). T h is legislatio n u sed p ro d u c­
tion co n tro ls on m o st agricultu ral p ro d u cts as a
m ean s o f in crea sin g p rices. Of th e n u m ero u s supply
co n tro l program s p ro p o sed in th e original AAA legisla­
tion, on ly th e to b a cco an d p ea n u t program s have
m a in tain ed d irect p ro d u ctio n co n tro ls.
T h e to b a c co program fu n ctio n ed , an d co n tin u es to
fu n ction , by first estab lish in g a su p p o rt p rice .6 In i­
tially, farm ers w ere assig n ed allo tm en ts th a t in d ica ted
th e n u m b er o f a cres o f to b a c co e a c h farm er cou ld
c u ltiv a te . In th e 1 9 6 0 s an d 1 9 7 0 s , th e a c r e a g e a llo t­
m en ts w ere su p p le m en ted w ith m arketin g q u o tas that
lim ited th e n u m b er o f p o u n d s o f to b a cco ea ch farm er
co u ld sell. T h e se q u o tas w ere b a se d on estim a tes of
th e qu an tity th at co u ld be sold at th e su p p o rt price.
T h e p rice su p p o rt m e ch a n ism h a s ch a n g ed only
slightly over tim e. Initially, if a farm er did n o t receive
an offer g rea ter th a n th e su p p o rt p rice, th e govern­
m en t p u rch a se d th e farm er’s to b a c co an d h eld it u ntil
it co u ld b e sold at th e su p p o rt p rice. In th e 1940s, a
system o f g ro w ers’ coop erativ es w as organized to p u r­
ch a se an d h o ld th e su rp lu s to b a cco . T h e coop erativ es
received, a n d co n tin u e to receive, govern m en t fin a n c­
ing.

6From its inception in the 1930s until 1985, the tobacco support price
was based on a “parity index" which measures the ratio of prices
received by farmers to prices paid by farmers. The parity ratio is
typically criticized for having no relationship to market prices.

FEDERAL RESERVE BANK OF ST. LOUIS

F or a long period, th e to b a cco program w as c o n sid ­
ered extrem ely su ccessfu l. T h e p rice o f U.S. to b a cco
co n tin u ed to rise, an d th e program w as ru n at little
co st to th e g overnm ent. In ad d ition, th e qu o ta rights to
grow an d sell to b a c co w ere m arketable; in fact, they
g en erated as m u ch as $800 m illion p e r y e a r in in co m e
for qu ota o w n ers.7 It is, in part, b ec a u se o f th e ap p a r­
en t s u c c e ss o f th e to b a c co program th at in tere st in
supply co n tro ls h as resu rfaced for o th e r crop s.
T h e to b a c co p ro g ram ’s ability to en d u re w hile g en ­
erating su bstan tial w ealth th rou gh th e sale an d le a s­
ing o f qu otas w as attribu table to th e in e la stic n atu re of
b o th w orld d em an d an d su pply o f to b acco . T h e m ajo r
rea so n for th e in e lastic su pply re sp o n se w as th at th e
United States h eld a large sh are o f th e w o rld ’s p ro d u c­
tion an d sales o f p articu lar varieties o f to b a c co .8 As
recen tly as th e 1950s, th e United States p ro d u ced
m ore th an 80 p e rce n t o f th e w o rld ’s bu rley to b acco .
It is im p o rtan t to n o te th at th e U.S. d o m in an ce in
to b a c co p ro d u ctio n an d th e in elasticity o f w orld su p ­
ply w ere even g reater w h en o ne co n sid ers th e im p o r­
tan t d istin ctio n o f to b a cco quality. Owing to sp ecial
soil an d clim atic co n d itio n s and grow ing ex p erien ce,
U.S. to b a c co generally w as regarded to be o f u n ­
m a tch ed quality.3 T h is fu rth er d ifferentiated it from
to b a cco grow n in o th e r co u n tries. If o th e r co u n tries
w ere u nab le to grow su p erio r quality to b a c co even as
its p rice in creased , th e su pply o f th at to b a c co w ould
be co n sid ere d p erfectly in elastic. Perfectly in elastic
supply m ean s th at th e qu antity su p p lied w ould not
ch an g e w h en th e p rice changed .
T h e d em an d for to b acco , in general, w as also in e la s­
tic. O ne so u rce estim ated th e in term ed iate-ru n d e ­
m an d elasticity o f to b a c co at — .1 and th e long-run
elasticity at —.5.'" T h e m ajo r reaso n for th e in elastic
n atu re o f to b a c co d em an d is th e lack o f su bstitu tes.
T h e addictive n atu re o f to b a c co fu rth er re d u ce s se n s i­
tivity to p r i c e ch an g es. F u r t h e r m o r e , to b a c co p u r ­

7Sumner and Alston (1985), p. 13. The U.S. General Accounting
Office study found that, although farmers were the intended benefi­
ciaries of the tobacco program, 68 percent of quota owners were not
active farmers. U.S. General Accounting Office (1982), p. 18.
8There are numerous varieties of tobacco. Two varieties, flue-cured
and burley, account for more than 90 percent of the tobacco grown
in the United States. There are other varieties used in the blending
of cigarettes that are not grown in this country, such as Oriental
tobacco.

JANUARY/FEBRUARY 1988

ch a se s gen erally re p rese n t only a sm all sh are o f a
c o n su m e r’s budget, a fact th at u sually re d u ces the
elasticity o f d em an d . W hile to b a c co u sers ca n sw itch
from U.S. to foreign to b a c co (or cigarettes), th ere are
few su b stitu tes for to b a c co in g eneral.
By using su pply co n tro ls, U.S. to b a cco p ro d u cers
initially earn ed h ig h er in co m es. W hile th e qu an tity of
to b a cco m arketed fell, th e resu ltin g p rice in crea se w as
large en ou gh to ca u se th e total revenue received by
q u o ta ow ners an d to b a c co grow ers to in crea se. B e­
ca u se o f th e h ig h er p rice, U.S. exp orts fell as foreign
co n su m ers red u ced th e am o u n t o f to b a cco p u rch a sed
at th e h ig h er p rice. Foreign su p p liers resp o n d ed to
th e h ig h er p rice by p ro d u cin g larger q u an tities of
to b a cco .

SOME LONG-TERM TRENDS
T h e su pply an d d em a n d analysis su ggested th at the
ad o p tio n o f a su p p ly co n tro l p o licy w ould lead to both
a re d u ctio n in U.S. p ro d u ctio n an d a sm aller U.S. share
o f w orld trad e an d w orld p ro d u ctio n . An exam in ation
of to b a cco p ro d u ctio n an d qu ota tren d s d o cu m en ts
th e lon g -term p ro c e ss o f red u cin g th e d o m estic to ­
b a c co in d u stiy as a m e a n s o f m ain tain in g th e p rice
su p p o rt m ech a n ism . C hart 1 tracks th e p ro d u ctio n of
to b a cco in th e U nited States against th e p ro d u ctio n of
to b a cco in th e rest o f th e w orld over th e p ast 30 years.
It show s th at d o m estic p ro d u ctio n , th o u gh variable,
h as b een tren d in g d ow nw ard w h ile foreign to b a cco
p ro d u ctio n h as grow n steadily. Sin ce 1966, d o m estic
to b a c co p ro d u ctio n h a s fallen by 38.8 p e rce n t, w hile
foreign p ro d u ctio n h as grow n by 56.5 p ercen t.
A lo n g er-term p ersp ectiv e on th e im p act o f th e
to b a cco program re strictio n s ca n b e gain ed by ex a m ­
ining acreag e data. T h e to b a cco program initially a t­
tem p ted to co n tro l p ro d u ctio n solely by restrictin g
th e n u m b er o f a cres th at farm ers cou ld grow. C hart 2
show s th e lon g-term tren d o f falling acreag e a llo t­
m e n ts."
As y ield s in creased , acreag e lim itatio n s b eca m e less
effective in co n tro llin g p ro d u ctio n an d w ere aug­
m e n ted by m arketing qu otas th at lim ited th e n u m b er
o f p o u n d s o f to b a c co farm ers co u ld m arket. C hart 3
show s th e tren d o f falling m arketing qu o tas for fluecu red and bu rley to b a cco s, th e two varieties th at a c ­
c o u n t for 90 p e rce n t o f all d o m estic to b a c co p ro d u c-

9Starkey (1985), p. 50 and U.S. General Accounting Office (1982),
p. 18.
,0Tweeten (1970), p. 201. These measures of demand elasticity are
interpreted to mean that a 1 percent increase in price would lead to
only a .1 percent decrease in quantity demanded in the intermediate
run and to a .5 percent decrease in the long run.



11Although not shown in the graph, tobacco acreage in 1986 was at its
lowest point since 1874 as a result of the supply control program.

59

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

C hart 1

U.S. a n d Foreign Tobacco Prod uction
Index

Index

1970=100

60
1959

1970=100

62

65

68

S o u r c e : U .S . D e p a r t m e n t o f A g r i c u l t u r e

tion. T h e ch art sh ow s that, after initially rising,
po u nd ag e qu otas for th e se tw o to b a c co s generally
have b e e n d ecreasin g in th e 1980s.
As in d icated earlier, a red u ctio n in th e U.S. sh ares o f
w orld to b a cco p ro d u ctio n and total ex p o rts is an
ex p ected resu lt o f th e su pply restrictio n . T able 1 d o cu ­
m ents th ese sh are lo sses. For exam ple, in th e 1 9 5 5 -5 9
period, th e United States a cc o u n te d for m ore th an 80
p e rce n t o f th e w o rld ’s p ro d u ctio n o f burlev to b a cco .
By 1985, th e U.S. share o f bu rley p ro d u ctio n h ad fallen
to 38 p e rce n t. Sim ilar tren d s are evident for flu e-cu red
to b a c co and for th e category lab eled “all to b a c c o .”
Not only have th e U.S. sh ares o f w orld p ro d u ctio n
an d trad e fallen, but th e u se o f im p o rted to b a cco has
risen su bstantially (see table 2). Until th e 1970s, th e u se
o f im p o rted bu rley and flu e-cu red to b a cco w as n eg li­
gible. In 1969, less th an 1 p e rce n t o f all burley to b a cco
u sed in th e United States w as im p orted. Bv 1985,
im p orts a cco u n ted for m ore th an 24 p e rce n t o f all
bu rley u se. O th er varieties n o t p ro d u ced in th is c o u n ­

http://fraser.stlouisfed.org/
60
Federal Reserve Bank of St. Louis

try, su ch as O riental to b a cco , co n tin u ally have b een
im p o rted for b len d in g p u rp o ses.
A n oth er im p o rtan t tren d is th e re d u ctio n o f th e
quality advantage th at U.S. to b a c co h old s over foreign
to b a cco . N um erous so u rces a ssert th at th e quality gap
b etw een foreign an d d o m estic to b a c co is n arrow in g.12
T his reflects th e fact that attem p ts to in crea se the
p rice o f high-qualitv to b a cco have provided foreign
p ro d u cers w ith an in cen tiv e to im prove th e quality of
th e ir to b a cco . T h e resu lt o f a sm aller quality advantage
an d risin g p rice s h a s led, p red ictably, to th e lo ss of
b o th d o m estic an d foreign m arkets for U.S. to b a cco .
Over tim e, th e d em an d for U .S.-p ro d u cea to b a cco
h as b e c o m e m o re ela stic as o th e r so u rces o f su pply
from th e rest o f th e w orld have ap p eared . T h e elastic-

12Starkey (1985), p. 50 and U.S. General Accounting Office (1982), p.

18.

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

C ha rt 2

H a r v e s t e d A c r e a g e o f B u rle y a n d F lu e -C u r e d T o b a c c o
M illions

of

acres

M illions

1.7

of

acres

1.7

.5
1920

25

30

35

40

45

S o u r c e : U .S . D e p a r t m e n t o f A g r i c u l t u r e

ity o f su pply also h as in creased . In th e sh o rt run,
foreign to b a c co p ro d u cers w ere lim ited in th e ir re ­
sp o n se to h ig h er p rices by th e ir lan d b ase an d o th e r
facto rs su ch as th e know ledge an d tech n o lo g y n eed ed
to p ro d u ce h igher-qu ality to b a cco . W ith tim e, h o w ­
ever, foreign p ro d u cers have acq u ired th ese a d d i­
tional inp u ts. T h e result h as b een a d ram atic in crea se
in th e qu antity o f to b a c co su p p lied by th e rest o f the
w orld. As a co n se q u e n ce , th e im p act o f U.S. to b a cco
policy on w orld to b a cco m arkets has d eclin ed .
A lthough th e lon g -ru n ben efits o f su pply co n tro l
p o licies m ay b e in q u estio n for U.S. to b a cco farm ers,
b en efits for foreign p ro d u cers are obvious. T h e se
b en efits are co n ferred in tw o w ays. First, by restrictin g
th e su pply o f U.S. to b a c co initially th rou gh qu otas and
la ter th rou gh th e m a in ten a n ce o f th e lo an stocks by
th e grow ers' cooperatives, a h ig h er w orld p rice is
m aintained . Second , the program crea te s a strong
in cen tive for foreign p ro d u cers to im prove the quality
of th e ir to b a cco by m aintain in g a h ig h er p rice in the
m arket for high-quality to b a c co th an w ould oth erw ise
result.



N one o f th e se lo n g -term tren d s o f d ecreasin g p ro ­
d u ctio n , falling q u o tas o r falling U.S. sh ares, how ever,
w ere ca u se for co n c e rn . T h e p u rp o se o f su p p ly c o n ­
trols w as to raise th e co m m o d ity ’s p rice and, m ore
im portantly, to raise th e net revenue o f farm ers. For
m any y ears, th e to b a cco program w as su cce ssfu l in
th is resp ect.
Over a re ce n t period, how ever, the program led to
low er revenu es for to b a c co grow ers. From 1982 to
1985, th e p ou nd age a llo tm en ts for bu rley to b a c co fell
by 30.4 p e rce n t. Over th is sam e period , how ever, the
average p rice p aid to grow ers for bu rley fell by 11.9
p ercen t. T h e co m b in a tio n o f lo w er o u tp u t an d low er
p rice tran slated into a 38.7 p e rce n t d eclin e in to b a cco
re ceip ts for bu rley farm ers.

RECENT PROGRAM DEVELOPMENTS
AND CHANGES
In the 1980s, the to b a c co p rice su p p o rt m ech an ism
led to m a jo r p ro b lem s. T h e to b a c co p rice su p p o rt w as,

61

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

C hart 3

P o u n d a g e A llo tm e n ts for B u rle y a n d F lu e -C u re d T o bacco
Millions

of

M illions

pounds

2 .7

of

pounds

------------------------12 . 7

2 .5

2 .3

2.1

1.9

1.7

1.5

1.3

1.1
1971

73

S o u r c e : U. S. D e p a r t m e n t o f A g r i c u l t u r e

Table 1
U.S. Percentage of World Tobacco Production and Exports
Burley
Year
1955-59
1960-64
1970
1975
1980
1981
1982
1983
1984
1985

Flue-Cured

All tobacco

Production

Exports

Production

Exports

Production

Exports

82%
80
62
52
44
51
49
33
42
38

60%
57
33
27
27
21
30
22
18
26

41%
40
30
28
20
18
13
13
12
10

60%
52
46
33
29
26
23
22
24
22

23%
25
19
18
16
16
13
11
12
10

35%
30
28
20
20
18
18
17
17
18

SOURCE: U.S. Department of Agriculture, Tobacco Outlook and Situation Report (December 1986), p. 42.


http://fraser.stlouisfed.org/
62
Federal Reserve Bank of St. Louis

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

th e stored to b a c co in crea sed . T h e a sse ssm en ts for
1986 w ere estim a ted at 30 ce n ts p e r p o u n d o r $600 p e r
acre.

Table 2
Percentage of Imported Flue-Cured and
Burley Tobacco Used Domestically
Year beginning
July 1
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985

Flue-Cured

Burley

0.9%
1.6
1.7
1.9
2.8
3.4
3.5
4.6
8.3
9.3
13.1
11.7
11.5
17.7
17.6
20.9
24.1

0.6%
0.6
0.9
1.6
5.4
8.4
8.4
7.2
14.7
15.1
18.6
22.3
19.1
24.1
25.8
28.9
24.5

SOURCE: U.S. Department of Agriculture, Tobacco Outlook and
Situation Report (April 1987), p. 14.

an d still is, ad m in istered by g ro w ers’ cooperatives,
w h ich p u rch ase d su rp lu s to b a cco an d h eld it u ntil it
co u ld be sold at a p rice above th e su p p o rt level. Any
lo sses on th e su rp lu s stocks w ere ab so rb ed by the
C om m od ity C redit C orp oration (CCC), w hile gains
w ere red istribu ted to th e cooperative's m em b ers. As
th e stocks h eld by coop eratives co n tin u ed to grow bu t
p ro sp e cts for selling th e se stocks at a gain seem ed
rem ote, th e p o ten tial co st to th e governm ent in ­
crea se d greatly.

Legislation in 1985, how ever, relieved grow ers o f the
p o ten tial liability for lo sses on th e stored to b a cco . T h e
CCC took title to th e su rp lu s stocks an d sold th e m at
d isco u n ts o f u p to 90 p ercen t, resu ltin g in a n et loss o f
ap p roxim ately $373 m illion . T h is loss w ill n o t b e re­
covered th rou gh th e No Net C ost Act.
In ex ch a n g e for th e g ov ern m en t’s rescu e, to b a cco
farm ers a cc e p te d lo w er su p p o rt p rice s. B e ca u se o f th e
low er p rices, th e a sse ssm en ts fell to only 2 ce n ts p e r
p ou nd . T h e U nited States D ep artm en t o f A griculture
(USDA) also w as given in crea sed freedom to red u ce
to b a c co p rices fu rth er if n eed ed an d w as p erm itted to
u se a m ore m ark et-o rien ted m eth o d o f calcu latin g
su p p o rt p rices an d settin g q u o ta s.13
T h e n ew to b a c co program has resu lted in su b sta n ­
tially low er p rices. T h e average to b a c co p rice paid to
grow ers fell from $1.80 p e r p o u n d in 1985 to $1.45 p e r
po u n d in 1986. As a result, to b a cco exp o rts ro se in
1987. Im p orts also fell an d n o w re p rese n t a sm aller
sh are o f th e to b a c co u sed in th e U nited States. M arket­
ing qu otas also have b e e n in crea sed in an ticip a tio n of
grow ing sales.

CAN SUPPLY CONTROLS BE USED
EFFECTIVELY ON OTHER CROPS?
T h e initial s u c c e s s o f th e to b a c co p ro g ram ’s u se of
su p p ly co n tro ls c a n b e a ttrib u ted to su pply an d d e ­
m an d ch a ra cte ristic s th at are n o t p re sen t for o th e r
m a jo r crop s. T h e to b a c co program ben efited from the
fact th at th e d em an d for U.S. to b a c co w as in elastic
b ec a u se o f a lack o f a good su b stitu te. A dditionally, th e
w orld supply w as in e la stic b ec a u se th e U nited States
held a d om in an t sh a re o f th e w o rld ’s p ro d u ctio n .

In resp o n se, the No Net C ost T o b a cco Program Act
o f 1982 w as p assed . T h is a ct stip u lated that the to ­
b a cco program be run at no n et co st to th e govern­
m en t o th e r th an adm inistrative co sts. U nder th is law,
a sse ssm en ts w ere levied on grow ers an d bu yers to
su p p o rt lo sses in cu rred by th e program . In 1985, b o th
buyers an d p ro d u cers o f flu e-cu red to b a c co w ere re ­
qu ired to pay a sse ssm en ts o f 7 ce n ts p e r p o u n d to
cover p rogram co sts. T h is am o u n t w as roughly equiv­
alen t to $140 p e r acre for farm ers.

,3The support price formerly had been determined by a combination
of the parity index and limits set by the Secretary of Agriculture.
Tobacco support prices currently are determined by a formula using
five-year moving averages of tobacco prices and year-to-year
changes in costs of production. This approach is substantially more
“market-oriented” than the previous method, which was driven by
costs of numerous products unrelated to the open market for to­
bacco.

U.S. to b a c co su rp lu ses grew as th e gap b etw een th e
su p p o rt p rice an d th e w orld p rice w id ened and im ­
p o rts gain ed a larger sh are o f U.S. to b a c co m arkets.
W ith less d o m estic to b a c co bein g sold on th e m arket,
th e cooperatives p u rch ase d m ore su rp lu s to b acco . As
a result, th e grow ers' p o ten tial liability for lo sse s on

The USDA determines tobacco quotas based on three factors.
The first factor is the intended purchases of tobacco by cigarette
manufacturers based on the support price. Cigarette manufacturers
must provide these estimates and purchase a minimum of 90
percent of their stated intentions or face a penalty. The remaining
two factors are the average tobacco exports of the past three years
and an estimate of the quantity of tobacco needed to maintain
tobacco stocks at desired levels.




63

JANUARY/FEBRUARY 1988

FEDERAL RESERVE BANK OF ST. LOUIS

M ost, if not all, o th e r m ajo r cro p s do not en jo y th ese
ch a ra cte ristics.
F or exam ple, if th e U nited States w ere su cce ssfu l in
restrictin g th e p ro d u ctio n o f co rn an d raising its m ar­
ket price, co n su m ers w ould m o st likely sw itch to any
o f th e n u m ero u s co a rse grains su ch as barley, so r­
ghum , m illet o r oats, w h ich are a ccep ta b le su b stitu tes
for m an y o f th e feed u ses o f co rn . On an in tern atio n al
level, th e U.S. sh are o f th e w o rld ’s co arse grains is
sm all. If it w ere to im p o se su pply co n tro ls on corn , it
w ould be n ecessa ry to re strict greatly th e im p ortation
of foreign grain th at w ould o c c u r in re sp o n se to h ig h er
U.S. p rice s. Su ch trad e restrictio n s m ight negatively
affect th e ability to exp ort o th e r U.S. co m m o d ities.
In so m e crop s, th e U nited States d o e s have a large
sh are o f th e w o rld ’s p ro d u ctio n . B eca u se of th e avail­
ability o f su bstitu tes, how ever, su pply restrictio n
w ould be ineffective. T h e U nited States, for exam ple,
p ro d u ces m ore th an h alf o f th e w o rld ’s soybean s.
U nfortu nately for ad vocates o f su pply co n tro ls, o th e r
cro p s like co rn , co c o n u t an d co tto n seed can b e su b ­
stitu ted for so ybean s as in p u ts for ed ible oil p ro d u c­
tion.
An ad d ition al facto r restrictin g th e p o ten tial u se of
su pply co n tro ls for o th e r crop s is th e w orld elasticity
o f su pply o f th e se crop s. M ost cro p s for w h ich supply
co n tro ls have b ee n co n sid ere d in th e U nited States
ca n b e p ro d u ced th rou gh o u t th e w orld. W heat, for
exam ple, is p ro d u ced in m ore th an 100 co u n tries. If
th e U nited States w ere su cce ssfu l in raising w h eat
p r ic e s b y r e d u c in g p r o d u c tio n , o th e r w h e a t-

elasticities an d th e foreign sh are o f w orld p ro d u ctio n
have in crea sed . T o a large exten t, th e d eclin e in m ar­
ket p ow er ca n be a ttrib u ted to U.S. p o licy a ctio n s. In
re sp o n se to this d eclin e, th e su pply co n tro l program
h as b ee n altered to b e m ore m ark et-o rien ted in settin g
su p p o rt p rices. T h e o th e r m a jo r cro p s for w h ich su p ­
ply co n tro l legislation h as b ee n p ro p o sed do n o t have
th e n ecessa ry su pply an d d em an d ch a ra cte ristics
n eed ed to su ccessfu lly im p o se a su pply co n tro l p ro ­
gram , even in the sh o rt term .

REFERENCES
Belongia, Michael T. “The Dairy Price Support Program: A Study
of Misdirected Economic Incentives,” this Review (February
1984), pp. 5-14.
Galston, William A. A Tough Row to Hoe: The 1985 Farm Bill and
Beyond (Hamilton Press, 1985).
Grise, Verner N. “The Tobacco Program: Recent Changes and
Issues in the Eighties,” Tobacco Outlook and Situation Report, U.S.
Department of Agriculture, Economic Research Service (U.S.
Government Printing Office, March 1984).
Houthakker, H. S., and Lester D. Taylor. Consumer Demand in the
United States-1929-1970 (Harvard University Press, 1966).
Johnson, Paul R. The Economics of the Tobacco Industry (Praeger
Publishers, 1984).
Pugh, Charles R. “County Lease Rates for Flue Cured Tobacco:
Revisited,” Economics Special Report No. 66 (North Carolina
State University, May 1981).
Starkey, James H. “An Industry Viewpoint of the Tobacco Program
and its Effects,” in Removal of Price Supports and Supply Controls
for U.S. Tobacco (National Planning Association, Washington,
D.C., 1985).
Sumner, Daniel A., and Julian M. Alston. Removal of Price Supports
and Supply Controls for U.S. Tobacco: An Economic Analysis of
the Impact (National Planning Association, Washington, D.C.,
1985).

p ro d u cin g co u n tries w ould b e able to resp o n d quickly
by in c re a sin g p ro d u c tio n w h ile th e n o n -w h ea tp ro d u cin g co u n tries w ould have incen tives to b egin to
p ro d u ce w heat.

Tomek, William C., and Kenneth L. Robinson.
Prices (Cornell University Press, 1981).

SUMMARY

U.S. Department of Agriculture, Agricultural Marketing Service.
Annual Report on Tobacco Statistics, various issues.

C ontrolling th e su pply o f ag ricultu ral p ro d u cts has
received atten tio n recen tly as a p o ssib le so lu tio n to
th e p ro blem o f falling farm p rice s and grow ing co m ­
m odity su rp lu ses. T h e original to b a c co program p ro ­
vides an insigh t into th e likely effects o f su ch farm
p o licy ch an g es. T h e to b a cco program en jo y ed initial
su cce ss b eca u se o f u niq u e ch a ra cte ristic s o f th e su p ­
ply o f an d d em an d for to b acco . T h e m arket p ow er of
th e U nited States in th e w orld to b a cco m arket, h o w ­
ever, h as d ecrea se d over tim e as su p p ly an d d em an d


http://fraser.stlouisfed.org/
64
Federal Reserve Bank of St. Louis

Agricultural Product

Tweeten, Luther. Foundations of Farm Policy (University of Ne­
braska Press, 1970).

_________ Agricultural Stabilization and Conservation Service.
History of Budgetary Expenditures of the Commodity Credit Corpo­
ration, Fiscal Year 1980-1986 Actual, December 29, 1986.
_________Economic Research Service. Tobacco Outlook and Sit­
uation Report, formerly the Tobacco Situation Report, various is­
sues.
_________ Tobacco, Background for 1985 Farm Legislation, Agri­
culture Information Bulletin Number 468 (National Economics Divi­
sion, September 1984).
U.S. General Accounting Office. “Tobacco Program's Production
Rights and Effects on Competition,” Report by the Comptroller
General of the United States, CED 82-70 (April 23, 1982).

F e d e r a l R e s e r v e B a n k o f S t. L o u is
Post O ffice Box 442
St. L ouis, M isso u ri 63166

T h e R e v ie w is p u b lis h e d six
tim es p e r y e a r by th e R e s e a rc h
a n d P u b lic In fo rm a tio n
D e p a rtm en t o f th e F ed era l
R es erv e R a n k o f St. L ou is.
S in g le-co p v s u b s c rip tio n s a re
available to th e p u b lic f r e e o f
c h a r g e . Mail r e q u e s t s f o r
s u b s c rip tio n s , b a c k is s u e s , o r
a d d r e s s c h a n g e s to: R e s e a rc h
a n d P u blic In fo rm a tio n
D ep a rtm en t, F ed e ra l R es erv e
R an k o f St. L ouis, P.O. Rox 4 4 2 ,
St. L ou is, M isso u ri 6 3 1 6 6 .
T h e views e x p r e s s e d a r e th o s e
o f th e in d h id u a l a u th o rs a n d d o
n o t n ece ss a rily re fle c t official
p o s itio n s o f th e F e d e ra l R es erv e
R a n k o f St. L o u is o r th e F e d e ra l
R es erv e S ystem . A rticles h ere in
m ay b e r e p rin t e d p ro v id e d the
s o u r c e is c re d ite d . P lea se p ro v id e
th e R a n k ’s R e s e a rc h a n d P ublic
In fo rm a tio n D e p a rtm en t with a
c o p y o f r e p rin t e d m aterial.