Full text of Review (Federal Reserve Bank of St. Louis) : January/February 1989

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

Vol. 71, No. 1




January/February 1989

3 P a y m e n ts System Risk: W h a t Is It
an d W h a t Will H app en If W e T ry
T o R ed u ce It?
18 F e d e ra l B u d g et T re n d s and th e
1 9 8 1 R eagan E co n o m ic Plan
32 An In tro d u ctio n to N on-T ariff
B a rrie rs to T ra d e
4 7 Can a C en tral Bank In flu en ce Its
C u r r e n c y ’s Real Value?
T h e S w iss C a se

THE
FEDERAL
A RESERVE
^ 0 1 5 V \ k o t

A r ST.IJOI LS

1

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 i s
R e v ie w
Ja n u a ry / F e b ru a ry 1 9 8 9

In T h is Issu e . . .




B oth co m m ercial b a n k s and th e Fed eral R eserv e assu m e risk by p a r­
ticip atin g in th e pay m en ts system . In re c e n t y ea rs, th e F ed era l R eserv e
has tak en actio n s to lim it su ch risk, and fu rth e r actio n s a re u n d e r co n ­
sid eration. In th e first a rticle in th is R ev iew , "P aym en ts System Risk:
W h at Is It and W h a t W ill Happen If W e T ry to R ed uce It?” R. Alton
G ilbert u ses sim ple b a la n ce sh eet en trie s to d escrib e h ow pay m en ts a f­
fe c t risk. He also exam in es th e likely e ffe c ts o f possible actio n s to
re d u ce paym en ts system risk. Am ong th e actio n s co n sid ered a re ch a rg ­
ing fe e s on th e daylight o v erd ra fts o f b a n k s’ re s e rv e a cco u n ts at Fed eral
R eserv e b an k s and req u irin g b a n k s th a t o v erd raw th e ir re se rv e a c ­
co u n ts to hold additional re se rv e b alan ces. T h e illu stration s also c o n ­
sid er th e o p eratio n o f CHIPS—th e p riv ate system fo r e lectro n ic
p ay m en ts—u n d e r a p ro ced u re th a t en su re s th e ex ecu tio n o f pay m en ts
m essages p ro cesse d b y th a t system .
* * *
In th e seco n d a rticle in th is R ev iew , "F ed era l B ud get T re n d s and th e
1 981 R eagan Eco n o m ic P lan ,” K eith M. C arlson assesses th e su ccess and
failu re o f fed era l bu d g et policy d uring th e eigh t-y ear R eagan ad m in istra­
tion. T h e a rticle co m p ares th e 1981 R eagan b u d g et plan, along w ith its
eco n om ic assum ptions, w ith th e actu al p e rfo rm a n c e o f th e se bu d g et
fig u res o v er th e 1 98 1 -8 8 period.
C arlson co n clu d es th a t th e R eagan b u d g et policy w as su ccessfu l in
sev eral re sp ects, nam ely, in crea sin g n atio n al d efen se spending, red u cin g
th e g ro w th o f th e n o n in tere st p o rtio n o f n o n d efen se spending and
red u cin g th e o verall ta x b u rd en . T h e m a jo r excep tio n to th e R eagan
b u d g et plan w as th e rise in n e t in te re s t p ro d u ced by a fa ilu re to
fo re c a s t th e 1 981-82 re cessio n , w h ich , in tu rn , h ad a com p ou nd ing in ­
te re s t e ffe c t on outlays.
* * *
R estrictio n s on in tern a tio n a l trad e, prim arily n o n -ta riff b a rrie rs , have
m ultiplied rapidly in th e 1 980s. In th e th ird a rticle in th is R ev iew , "A n
In tro d u ctio n to N on -T ariff B a rrie rs to T ra d e ,” Cletus C. Coughlin and
G eo ffrey E. W ood provide a p rim er on th e se b a rrie rs. T h e a u th o rs
beg in b y id entifyin g n u m ero u s n o n -ta riff b a r r ie r s and d o cu m en t th e ir
p ro liferatio n . T h e g en era l e ffe c ts o f n o n -ta riff b a rrie rs , like th o se o f
ta r iff b a rrie rs, a re to in cre a s e th e d om estic p rice s o f th e p ro tected
goods and to im pede tra d e to b e n e fit selected p ro d u cers at th e exp en se
o f d om estic p ro d u cers. N um erous rea so n s fo r th e in crea sin g u se o f nonta r iff instead o f ta r iff b a rrie rs a re provided. Am ong th e re a so n s a re
th e ir m o re ce rta in p ro tectiv e e ffects, th e possibility th a t som e b en e fits
can b e cap tu red by fo reig n p ro d u cers and d om estic politicians and th e
fa c t th a t th e ir ad verse e ffe c ts a re g en erally less obvious to co n su m ers.

JANUARY/FEBRUARY 19S9

2

T o date, attem p ts th ro u g h th e G en eral A g reem en t on T a riffs and T ra d e
(GATT) to co u n te ra c t th e exp an sion o f n o n -ta riff b a rrie r s h ave m e t w ith
little su ccess. A b r ie f h isto ry o f th e se attem p ts co m p letes th e paper.
***
T h e large flu ctu atio n s in ex ch a n g e ra te s o b serv ed in th e 1 9 8 0 s have
p rom p ted m any ce n tra l b an k s o f in d u strialized n ation s to d iscuss e x ­
ch an g e m a rk et in terv en tio n . T h e se d iscussions have ex p lo red w h e th e r it
is ap p ro p riate to u se m o n eta ry actio n s to in flu en ce ex ch a n g e ra te s and,
if so, how su cce ssfu l su ch e ffo rts m ight b e. In th e final a rtic le in th is
issue, “Can a C en tral B an k In flu en ce Its C u rre n cy ’s Real Value? T h e
Sw iss C ase," M ich ael T . Belongia and W e r n e r H erm an n an aly ze th e
d istin ct ex p e rie n ce o f o n e c e n tra l b a n k pu rsu in g an ex ch a n g e-ra te
objective.
In th e first p a rt o f th e ir article, Belongia an d H erm an n rev iew th e
eco n o m ic th e o ry th a t re la tes m o n eta ry actio n s to m ov em en ts in th e real
ex ch an g e rate. A fter con clu d in g th a t th e e ffe c ts, if any, w ill b e sh o rt­
lived, th e y inv estigate h ow a ctio n s b y th e Sw iss N ational Bank, relativ e
b o th to th e F ed era l R eserv e and G erm an B un d esb an k, h ave a ffe cte d th e
Sw iss franc/dollar and Sw iss franc/DM re a l ex ch a n g e ra tes. T h e ir re su lts
in d icate th a t a ce n tra l b a n k ca n in flu en ce re a l ex ch a n g e ra te s only fo r a
perio d o f m on th s; m o reo v er, a p r e d ic ta b le resp o n se w ill o c c u r only w ith
re g a rd to th e on e b ila te ra l ra te th a t re ceiv es h ig h est p rio rity as a policy
objective.


http://fraser.stlouisfed.org/
Federal ReserveFEDERAL
Bank of St.
Louis
RESERVE
BANK OF ST. LOUIS

3

R. Alton Gilbert
R. Alton Gilbert is an assistant vice president at the Federal
Reserve Bank of St. Louis. Dawn M. Peterson provided research
assistance.

Paym ents System Risk: What Is It
and What Will Happen If We Try
To Reduce It?
O
OTH co m m ercial banks and th e Fed eral Re­
serve assu m e a certain am ou n t o f risk in p a rtici­
pating in th e p ay m en ts system . T h is p a p e r p ro ­
vides an in tro d u ctio n to p ay m en ts system risk and
th e p u blic p o licy issu es involved in lim iting the
risk. Using sim ple b alan ce sh eet en tries to illu s­
trate, th e p ap e r will exam in e how p o licies in ­
ten d ed to re d u ce pay m ents system risk w ould
affect banks and ban k cu sto m ers.

PAYMENTS SYSTEM RISK: WHAT IS
IT?
M any banks overdraw th eir reserve a cc o u n ts at
th e Fed eral Reserve d uring part o f e a c h b u sin ess
day as they p ro cess p ay m en ts w ithin th e pay­
m en ts system . T h e F ed eral Reserve is co n c e rn e d
about th e exten t o f th is intrad ay cred it for several
reason s. First o f all, sin ce it d oes not ch arge in ter­
est on th e in trad ay cred it it exten d s, it is providing
th is overdraft facility at no co st to banks and, thus,
may be overused by banks. Second , and m ore im ­
portant, it is p ossible, thou gh unlikely, th at a bank
co u ld fail w hile its reserve a cco u n t is overdraw n.
In th is event, th e Fed eral Reserve w ould b e c o m e a
general cred ito r o f th e failed bank. Finally, th e Fed
is co n c e rn e d w ith th e risk th at banks assu m e
th rou gh th e ir p articip atio n in private w ire tran sfer
system s. C urrent Fed eral Reserve p o licy is d e­



signed to lim it th e risk assu m ed by Reserve Ranks
as w ell as co m m ercial banks w ho p articip ate in
private system s for th e ir e lectro n ic paym en ts. (See
ap p en d ix 1 for a d escrip tio n o f th at policy.)

Federal Reserve Daylight Overdraft
Risk and the Operation o f Fedwire
W hile various types o f tra n sa ctio n s affect the
reserve b a la n ces o f banks, daylight overdrafts g en ­
erally reflect large tra n sa ctio n s th rou gh Fedw ire,
th e w ire tran sfer system op erated by th e Federal
Reserve System . In stitu tio n s w ith reserve o r cle a r­
ing a cc o u n ts at a Reserve Bank m ay tran sfer th e ir
reserve b a la n ces to o th e r in stitu tio n s th at have
sim ilar a cco u n ts. T h e se tran sfers, w h ich averaged
$605 b illion p er b u sin ess day in 1987, are
p ro cessed electro n ically th rou gh Fedw ire.
Fed eral Reserve B anks tran sfer reserves to re ­
ceiving banks even if th e reserve b a la n ce o f th e
sen d in g ban k is in su fficien t to cover th e tran sfers.
T ran sfers over Fedw ire are “fin a l” w h en th e receiv­
ing banks are notified o f th e tran sfers. T h u s, if a
sen d in g bank sh o u ld fail w hile its reserve a cco u n t
w as overdraw n, th e Fed eral Reserve w ould have
n o claim on banks th at received reserves from th e
failed bank over Fedw ire.
U.S. Treasu ry an d ag en cy secu rities also are
tran sferred am on g banks over Fedw ire. O w nership

JANUARY/FEBRUARY 1989

4

reco rd s o f th e se secu rities are m aintain ed in ea ch
F ed eral Reserve B an k ’s co m p u te r system . Banks
ca n tran sfer secu rities h eld in th e ir n am es to o th er
in stitu tio n s throu gh th ese co m p u ters, a system
called “book-en try.” A tran sfer o f secu rities in
b ook-en try form ca n be arranged eith e r in c o n ­
ju n c tio n w ith a tran sfer o f reserves o f equal value
or as a sep arate tran sactio n . Su ch secu rities tra n s­
a ctio n s co n trib u te to daylight overdrafts, sin ce
typically th e reserve a cc o u n ts o f banks are d ebited
w h en th eir b ook-entry secu rities a cc o u n ts are
cred ited . T ran sfers o f b o o k -en tiy secu rities over
Fedw ire averaged $312 billion p e r day in 1987.
T h e F ed eral Reserve m easu res its exp o su re to
paym ents system risk by sim ply su m m in g the
m axim um daylight overdraft e a c h day a cro ss all
banks. In 1987, th e F e d ’s ex p o su re to daylight
overdrafts averaged $112 billion, approxim ately 53
p e rce n t o f w h ich can be attribu ted to tran sactio n s
involving book-entry governm ent sec u ritie s.1 Som e
sp ecific featu res o f this risk m easu re sh o u ld be
n oted. First, unlike co n v en tion al risk m easures,
th e Fed eral R eserve’s m easu re d oes n o t in c o rp o ­
rate th e p robability th at a bank w ill fail w hile in an
overdraft p o sitio n o r th e p robability o f F ed lo sses
in su ch situ atio n s.2 Sin ce th e Fed eral Reserve has
never in cu rred a loss on daylight overdrafts, the
p robability o f lo sses in th e future are qu ite low.
Seco n d , it ex ceed s th e actu al su m o f reserve
a cc o u n t overdrafts at any p o in t d uring th e day;
th e m axim u m overdrafts o f individual banks typi­
cally o c c u r at different tim es during th e day.
Third, it rep rese n ts th e loss th at th e F ed eral Re­
serve w ould in c u r on a given day if all banks w ith
overdraw n reserve a cc o u n ts failed w h en th e ir
overdrafts w ere at m axim u m levels an d th e F ed ­
eral Reserve recovered nothing.

Systemic Risk and the Operation o f
CHIPS
T h e C learing H ouse In terban k P aym ents System
(CHIPS) is an electro n ic p aym ent system o p erated
by th e New York Clearing H ouse. It cu rren tly is the
only private electro n ic paym ent system in o p era ­
tion in th e United States. CHIPS h as abou t 140
m em bers, w h ich in clu d e U .S.-chartered banks an d

1Daylight overdrafts attributed to transactions in book-entry
securities are calculated as follows. A bank is in a net credit
position on book-entry securities transfers if the value of securi­
ties transferred to the bank’s book-entry securities account
exceeds the value of securities transferred out of that account
to other banks. The book-entry overdraft of a bank for each day
equals its largest net credit position on securities transfers that
occurs while the reserve account of the bank is overdrawn.

http://fraser.stlouisfed.org/
Federal Reserve
Bank of St.
Louis BANK OF ST. LOUIS
FEDERAL
RESERVE

foreign banks. M em bers o f CHIPS sen d an d receive
p aym en t m essages d uring th e day; no fund s are
actually tran sferred to cover th e se p aym en t m e s­
sages, how ever, until th e en d o f th e day. Net ob li­
gations are settled at d ay’s en d th rou gh Fedw ire
transfers in th e reserve a cc o u n ts o f CHIPS p a rtici­
p an ts. B anks in n et d ebit p o sitio n s on CHIPS at
th e en d o f th e day (value o f p aym en t m essages
sen t ex ceed s th e value o f p ay m en t m essages re ­
ceived) tran sfer fund s from th e ir a cc o u n ts at Re­
serve B anks to a reserve a cc o u n t m a in tain ed by
th e clearin g h o u se at th e Fed eral Reserve B ank of
New York, w hile banks in n et cred it p o sitio n s re ­
ceive reserve tran sfers from th at a cco u n t. T he
value o f p ay m en t m essages p ro cesse d by CHIPS
averaged $555 billion p e r day in 1987.
System ic risk refers to th e risk th at th e failure o f
on e bank will ca u se o n e o r m ore o th e r banks to
fail. O ne w ay th at th is co u ld h a p p en is throu gh
p articip atio n in CHIPS. If a ban k fails w hile in a n et
d ebit p o sitio n on CHIPS, o th e r CHIPS p a rticip a n ts
co u ld suffer lo sses as well, d ep en d in g on th e p ro ­
ce d u res in force for dealing w ith su ch a default.
P aym ents over Fedw ire, in co n trast, involve no
sy stem ic risk. T h e Fed eral Reserve w ould absorb
any lo sses resu lting from failures by banks w ith
overdraw n reserve a cco u n ts.
T h e Fed eral Reserve m easu res th e pay m en ts
system risk a ssu m ed by CHIPS p articip an ts as th e
sum o f th e ir m a x im u m n et d ebit p o sitio n s d uring
th e day on CHIPS. T his m easu re averaged $43.7
b illion in 1987.
To relate this m easu re to sy stem ic risk is dif­
ficult, how ever; u n d e r cu rren t CHIPS rules, pay­
m en t m essages do n o t reflect intrad ay ex ten sio n s
o f cred it am on g banks b u t provisional pay m en ts
w h ich m ay b e u nw o u n d at th e en d o f th e day. If a
b an k co u ld n o t cover its n et d ebit p o sitio n on
CHIPS at th e en d o f th e day, all p ay m en t m essages
to an d from th at ban k w ould b e ca n c e le d ; n ew n et
d ebit an d cred it p o sitio n s w ould th e n b e c a lc u ­
lated for th e rem ain ing CHIPS p articip an ts, an d
p aym en ts w ould b e m ad e to cover th e se revised
p o sitio n s. Unw inding CHIPS pay m en ts b ec a u se o f
a defaulting bank, how ever, co u ld ex p o se th e re ­
m aining CHIPS p articip an ts to lo sses if th e ir de-

2ln conventional definitions, risk is specified in terms of the
probability distribution of returns on an investment. Under one
definition, risk may be measured as the variance of the distribu­
tion of returns. See Rothschild and Stiglitz (1970).

5

p ositors h ad w ithdraw n b a la n ces cred ited to th e ir
a cco u n ts d uring th e day b ased on pay m ent m e s ­
sages from th e defaulting bank. T h e se banks in
tu rn m ay b e u nab le to recover th e fund s w ith ­
draw n by th e ir d ep o sito rs during th e day.3

Federal Reserve Policy on Payments
System Risk
In re cen t y ears, th e Fed eral Reserve B oard h as
taken actio n s to lim it its ow n risk an d th e sy stem ic
risk involved in CHIPS. T h e Fed eral Reserve in ­
d u ced CHIPS to requ ire e a c h bank in its system to
establish bilateral n et d ebit lim its w ith ea ch o th er
CHIPS particip an t, beg in n in g in 1984. U nd er a n ­
o th e r program th at w en t in to effect in M arch 1986,
th e Fed eral Reserve req u ires banks to set lim its on
th e ir daylight overdrafts a cro ss Fedw ire an d
CHIPS. (See ap p en d ix 1 for d etails o f th e se p o li­
cies.) T h e Fed is cu rren tly studying p ro p o sals to
estab lish an exp licit o r im p licit p rice for daylight
overdrafts o f reserve acco u n ts.

HOW PAYMENTS AFFECT RISK
T h is sectio n u ses sim p le b alan ce sh ee ts o f hy­
p o th etical banks to illu strate ho w tran sactio n s
throu gh th e pay m ents system affect th e exp osu re
o f th e Fed eral Reserve and co m m ercial banks to
p o ten tial lo sses. T h e illu stration s involve federal
funds tran sactio n s an d tra n sa ctio n s am ong CHIPS
p articip an ts. A ppendix 2 illu strates ho w th e pay­
m en t p ractices o f banks that serve governm ent
secu rities d ealers an d th o se th a t issu e an d red eem
co m m ercial p ap e r affect th e ir reserve overdrafts.

Federal Funds Transactions
Banks th at borrow federal fund s overnight are
co n c e rn e d prim arily abou t th e ir reserve b alan ces
as o f th e en d o f th e day, rath er th a n d uring th e
day, for two reason s. First, th e Fed eral Reserve is
m ore to leran t o f daylight overdrafts o f reserve
a cco u n ts th an o f negative reserve b a la n ces at th e
clo se o f b u sin ess. Secon d , th e intrad ay reserve
b a la n ces do n o t co u n t tow ard m eetin g reserve
req u irem en ts; only th o se b alan ces held at th e en d
of th e b u sin ess day do.
Banks th at borrow overnight federal funds typi­
cally receive reserves from th e lend ing bank s over
Fedw ire late in th e day; they retu rn th e requ isite
reserve b a la n ces th e follow ing m ornin g. Su ch
3The legal status of claims by the banks against their depositors
in such situations is currently unclear. See Mengle (1989).



tran sfers ca n ca u se th e b orrow ing banks to over­
draw th e ir reserve b a la n ces d uring th e day.
T h e b a la n ce sh ee t en tries in table 1 illu strate
h ow fed eral fu n d s tra n sa ctio n s affect th e risk
b o rn e by th e Fed eral Reserve. E a ch ban k begins
th e day w ith d ep o sits o f $100 an d reserves o f $10.
W ith a 10 p e rce n t reserve ratio, e x cess reserves are
zero. D uring th e previous b u sin ess day, B an k A
b orrow ed $25 from Bank B th rou gh th e fed eral
fun d s m arket. B efore th e en d o f b u sin ess on the
previous day, Bank B tran sferred $25 over Fedw ire
from its reserve a cc o u n t to th e a cc o u n t o f B ank A.
T h is tra n sa ctio n crea te d a liability for Bank A (fed­
eral fund s p u rch ased ) an d sh ifted $25 o f th e a ssets
o f Bank B from reserve b a la n ces to fed eral funds
sold.
T h e first tra n sa ctio n by Bank A in th e cu rren t
day is a tran sfer o f $25 from its reserve a cc o u n t to
th e reserve a cc o u n t o f Bank B, retu rn in g th e fund s
it h ad borrow ed overnight; th is elim in ates th e
liability o f fed eral fund s p u rch a se d by Bank A.
Sin ce th e b a la n ce in th e reserve a cc o u n t o f Bank A
w as only $10 at th e start o f th e day, th e tran sfer o f
$25 m akes its reserve a cc o u n t overdraw n by $15.
T h is p re sen ts n o p ro b lem for Bank A, how ever,
sin ce it p lan s to b o rro w $25 th rou gh th e federal
fund s m arket later in th e day to elim in ate its re ­
serve overdraft an d m eet its reserve requ irem en t
o f $10.
If Bank A borrow s th e $25 in th e fed eral funds
m arket, th e len d in g bank(s) w ill tran sfer th e re ­
serves to th e a cc o u n t o f Bank A in th e afternoon.
Given th e tim e gap b etw een th e tran sfer o f funds
to len d in g banks in th e m orn in g an d th e transfer
o f reserves to Bank A in th e aftern oon , th e Fed eral
Reserve effectively len d s $15 to Bank A during part
o f th e b u sin ess day by p erm ittin g th e reserve
overdraft.
T h e F ed is a g en eral cred ito r o f Bank A w hile its
reserve a cc o u n t is overdraw n. To illu strate th e risk
it a ssu m es in p erm ittin g daylight overdrafts, su p ­
p o se th at p a rticip a n ts in th e fed eral fund s m arket
find ou t th at th e value o f Bank A s assets have d e ­
clin ed by $15 ju s t after Bank A tran sfers $25 to
Bank B. After th is in form ation b e co m e s know n,
Bank A will b e u n ab le to borrow reserves in th e
fed eral fund s m arket at prevailing m arket rates.
T h e ag en cy th at ch a rtered B an k A m u st d ecid e
w h eth er it is solvent. If B ank A is d eclared solvent
an d h as assets to pled ge as collateral, it cou ld

6

Table 1
Risk Created by the Transfer of Reserve Balances in Overnight Federal Funds
Transactions
Balance sheets at start of day:

Bank A

Reserves
Other
assets

$10 Deposits
Federal funds
125 purchased
Net
worth

Bank B

$100

Reserves

$ 10 Deposits

25

Federal funds
sold

Net
25 worth

10

Other
assets

75

$100
10

Bank A sends $25 of its reserve balances to Bank B over Fedwire:
Bank A

Reserves
Other
assets

- $ 15 Deposits
Federal funds
125 purchased
Net
worth

Bank B

$100
0
10

Reserves
Federal funds
sold
Other
assets

$35 Deposits
Net
0 worth

$100
10

75

Value of other assets at Bank A reduced by $15:
Bank A

Reserves
Other
assets

-$ 1 5

Deposits

Federal funds
110 purchased
Net
worth

Bank B

$100
0
-5

receive a lo an from th e Fed eral Reserve to cover its
reserve overdraft. If th e supervisory ag en cy d e ­
clares Bank A insolvent, it w ill be clo sed . If Bank A
is clo sed an d liqu id ated , th e d ep o sito rs get first
claim on th e $110 o f “o th er a sse ts.” In th is case,
th e Fed eral Reserve w ill receive $10 against th e $15
overdraft o f th e reserve a cco u n t and, thu s, will
lo se $5.
If th e Fed eral Reserve h ad know n th at Bank A
w as in p o o r fin an cial co n d itio n , it w ould have
required th e bank to pled ge collateral against its
overdrafts.4 By requiring collateral, th e Fed shifts
th e risk to o th e r parties. Su ppose, in this case, th at
Bank A h ad pledged $15 o f its riskless assets to th e
Fed eral Reserve to cover its overdrafts. W hen the
b an k fails, th e Fed w ould hold th e $15 in collateral

"Task Force (1988), pp. 65-69.

http://fraser.stlouisfed.org/
Federal Reserve
Bank of St.
Louis BANK OF ST. LOUIS
FEDERAL
RESERVE

Reserves
Federal funds
sold
Other
assets

$35 Deposits
Net
0 worth

$100
10

75

to cover any lo sses. T h e loss o f $5 w ould b e b o rn e
by u n in su red d ep o sito rs o r th e Fed eral D eposit
In su ra n ce C orp oration (FDIC). T h u s, requiring
collateral against reserve overdrafts d oes n o t n e c ­
essarily p ro tect th e p u b lic secto r; it m ay sim ply
shift th e lo ss from th e Fed eral Reserve to th e FDIC.

Transactions Am ong CHIPS
Participants
In th e ca se illu strated in table 1, th e Fed eral
Reserve a ssu m es th e risk. B anks also assu m e risk
by p articip atin g in CHIPS. T h e in terb an k risk ex p o ­
su res crea ted th rou gh th e p ro cessin g o f p aym ent
m essages th rou gh CHIPS are illu strated in table 2.
In th e first tra n sa ctio n o f th e day, a d ep o sito r o f
Bank A sen d s $25 to a d ep o sito r o f Bank B in th e

7

Table 2
Risk Created by the Transfer of Funds over CHIPS
Balance sheets at start of day:

BankB

Bank A

Reserves
Other
assets

$10 Deposits
Net
100 worth

$100
10

Reserves
Other
assets

$10 Deposits
Net
100 worth

$100
10

Depositor at Bank A transfers $25 to depositor of Bank B, transaction over CHIPS:
BankB

Bank A

Reserves
Other
assets

$75

Reserves

Reserves
100 payable

25

Reserves
receivable

Net
worth

10

Other
assets

$10 Deposits

$10 Deposits
Net
25 worth

$125
10

100

Depositor at Bank B transfers $25 to depositor of Bank C, over CHIPS:
BankC

BankB

Reserves
Reserves
receivable
Other
assets

$ 10 Deposits
Reserves
25 payable
Net
100 worth

$100
25
10

Other
assets

form o f a w ire tran sfer over CHIPS. Bank A debits
th e d ep o sit a cc o u n t o f th at cu sto m e r for $25. B e­
ca u se banks do n o t rep ort th e ir b alan ce sh e e ts on
an intrad ay basis, th ere is no official term for the
offsetting liability en try in th is tran sactio n . In this
case, w e will call it "reserves p ay able.” F o r Bank B,
d ep o sit liabilities and an asse t item called "re ­
serves receivable" e a c h in crease by $25.
In th e next tran sactio n , a d ep o sito r o f Bank B
d irects it to sen d $25 to a cu sto m e r o f Bank C.
After th e seco n d tran sactio n , Bank B is even w ith
CHIPS. If th ere w ere n o m o re tran sactio n s over
CHIPS th at day involving B ank B, th e settlem en t
for CHIPS tran sactio n s w ould have a zero im p act
on th e reserve a cco u n t o f Bank B. Bank A, in c o n ­
trast, w ould have its reserve a cco u n t d ebited for
$25, w hile B ank C w ould have its a cc o u n t cred ited
by $25. Bank A w ould have to in crea se its reserve
b a la n ce before th e tim e for settlem en t o f CHIPS
paym ents to facilitate settlem en t.



Reserves
Reserves
receivable

$10 Deposits
Net
25 worth

$125
10

100

Su p p o se that, before th e en d o f th e day, adverse
p u b licity prevents Bank A from borrow ing $25 in
th e fed eral fund s m arket. T h is situ atio n co u ld
crea te a liquidity p roblem for Bank B. If Bank A
ca n n o t o b tain su fficient reserves to cover its n et
d ebit p o sitio n o n CHIPS, cu rren t rules call for
u nw in din g all tra n sa ctio n s involving Bank A and
settlin g th e tra n sa ctio n s am on g th e rem aining
CHIPS p articip an ts. T h is settlem en t w ould involve
a tran sfer o f $25 in reserves from Bank B to Bank C.
Su ch a n et settlem en t ca n n o t take p lace, how ever,
b ec a u se Bank B h as only $10 in its reserve a c ­
co u n t. T h u s, u n less th e Fed eral Reserve len d s $25
to Bank A o r Bank B, all CHIPS tra n sa ctio n s for th e
day w ould be ca n celed .
Sim ulation ex e rcises in d icate th at th e u n w in d ­
ing o f tra n sa ctio n s w ith o n e large CHIPS p a rtici­
p an t th at ca n n o t m eet its p aym en t obligations
w ould m ake a high p ercen ta g e o f o th e r p a rtic i­
p an ts u n ab le to m eet th eir co m m itm en ts on

JANUARY/FEBRUARY 1989

8

CHIPS w ithou t ad d ition al reserves.5 In th e se ex e r­
cises, som e ban k s th at b eco m e illiquid have no
d irect tran sactio n s w ith th e defaulting bank. Thu s,
as illu strated in table 2, a default by Bank A keep s
Bank C from receiving its p ay m en ts over CHIPS,
b e c a u se th e default by Bank A m akes Bank B
illiquid.
As th e cen tral bank, th e Federal Reserve is re ­
sp o n sib le for p reventing su ch a liquid ity crisis. In
o u r exam ple, th e F ed co u ld len d reserves eith e r to
Bank A o r Bank B. If it co n sid ers Bank A to b e so l­
vent, it co u ld len d th e $25 and take collateral. T h e
$25 ad d ed to th e reserve a cc o u n t o f Bank A facili­
ta tes th e n et settlem en t on CHIPS. If Bank A tu rn s
out to b e insolvent, th e co llateral p ro tects th e F ed ­
eral Reserve from loss, tran sferring it in stead to the
general cred ito rs and th e FDIC.
Alternatively, th e Fed eral Reserve co u ld prevent
a liquid ity crisis b y len d in g $25 to B ank B, allow ing
B ank B to m e et its requ ired reserves and CHIPS
obligation to B ank C. Even if th e F ed prevents a
liquidity crisis by len d in g $25 to Bank B, th e d e ­
fault o f B ank A co u ld m ake Bank B insolvent. T his
is an exam p le o f sy stem ic risk involved in th e o p ­
eration o f th e p ay m en ts system . Su p p o se th at the
tran sfer o f $25 from Bank B to Bank C is initiated
by th e d ep o sito r o f Bank B w ho received $25 from
B ank A. Bank B m akes th is tran sfer before discov­
ering th e default by B an k A. At th is tim e, it is n ot
clea r w h eth er th e co u rts w ould p erm it Bank B to
regain th e se funds from its d ep o sito r.6 If Bank B ’s
loss ex ceed s $10, it is bankrupt.
Su ppose, instead , that this d ep o sito r o f Bank B
h old s th e extra $25 in its d em an d d ep o sit a cco u n t
at Bank B u ntil th e en d o f th e day. T h e tran sfer of
reserves from Bank B to Bank C w as initiated by a
different d ep o sito r o f Bank B. W h en Bank A’s d e­
fault is d iscovered, Bank B co u ld c a n c e l th e $25 in
reserves receivable an d reverse th e $25 cred it to its
dem an d d ep o sit liabilities. In th is case, th e u n ­
w inding o f th e CHIPS tra n sa ctio n h as n o adverse
effect on th e n et w orth o f Bank B.
5Humphrey (1986).
6Mengle (1989).
7For discussions of these possible changes from Federal Re­
serve sources, see Belton, et al. (1987), Corrigan (1987),
Johnson (1988), Task Force (1988) and Mengle, et al. (1987).
For discussions of these issues by those in the private sector,
see Flannery (1987), Faulhaber, et al. (1989) and Large Dollar
Payments System Advisory Group (1988). Governor Wayne D.
Angell of the Federal Reserve Board has proposed another
approach to revising policy on payments system risk. Under

http://fraser.stlouisfed.org/
BANK OF ST. LOUIS
Federal Reserve FEDERAL
Bank of St.RESERVE
Louis

THE EFFEC TS OF POSSIBLE
CHANGES IN POLICY
C hanges in p o licy on p ay m en ts sy stem risk are
b ein g d iscu sse d w ith in th e F ed eral Reserve Sys­
tem an d th e private secto r. T h is sectio n illu strates
th e effects o f tw o p o ssib le p o licy ch a n g es: exp licit
fees on reserve a cc o u n t overdrafts an d in terestearn in g reserve b a la n ces req u ired to cover p art or
all o f daylight overdrafts.7
Fed eral Reserve p olicym akers have in d icated
th at s u ch ch a n g es w ould b e ad o p ted only after
CHIPS h as developed a rran g em en ts for en su ring
th e e x e cu tio n o f p ay m en ts on th a t system th at
th ey co n sid e r a ccep ta b le .8 T h is sectio n also illu s­
trates th e im p licatio n s o f su ch an arran g em en t for
banks.

Explicit Pricing o f Daylight Overdrafts
o f Reserve Accounts
O ne w ay to re d u ce F ed eral Reserve risk w ould
b e to ch arge a fee on daylight overdrafts. If th e fee
w ere high enou gh, banks w ould re d u ce th e size of
th e ir overdrafts by ch an g in g th e ir p ra ctice s for
m aking paym en ts.
R e s p o n s e s o f B a n k s to P ricin g D aylight O v er­
d r a fts — P erh ap s th e ea siest an d le a st expensive
ch an g e for m o st o f th e relatively large banks w ould
involve routing m ore o f th e ir w ire tran sfers o f
fund s throu gh CHIPS ra th er th a n Fedw ire. T h ere
are o th e r w ays for b an k s to re d u ce th e ir reserve
a cc o u n t overdrafts. T h ey co u ld p u rch a se m o re of
th eir federal fu n d s as term federal fu n d s o r u n d e r
rollover arran g em en ts th at involve paying a daily
rate b u t elim in atin g th e daily tran sfer o f reserve
b a la n ces. Pricing total daylight overdrafts o f r e ­
serve b a la n c e s (including b o o k-en try overdrafts)
w ould im p o se co sts on th e clearin g banks, w h ich
they w ould p ass on to th e governm ent secu rities
d ealers th ey serve. T h e d ealers co u ld red u ce booken try daylight overdrafts by bu ilding sm aller in ­
ven tories o f secu rities d uring th e day o r hold ing
larger in ven tories overnight. B anks th at a ct as
agents in issu in g co m m ercia l p a p e r co u ld ch arge
the Angell proposal, the Federal Reserve would prohibit day­
light overdrafts. Transfers of reserves that would make the
reserve balance of a bank negative would be funded as dis­
count window loans. To provide banks incentives to hold
enough reserves to prevent overdrafts, the Federal Reserve
would pay interest on excess reserves, but at a rate below the
discount rate. See VanHoose (1988).
8Johnson (1988), p. 15.

9

issu ers for th e fees on overdrafts o r delay pay­
m en ts to issu ers u ntil th ey receive pay m ents from
pu rch asers.
E ffe c ts in F in a n c ia l M a r k e ts — P ricing daylight
overdrafts co u ld have a variety o f in d irect effects in
th e fin an cial m arkets. Banks th at lend in th e over­
night federal fu nd s m arket co u ld find that th eir
reserves are bein g retu rn ed la ter th e follow ing day.
T h e tim e value o f intrad ay reserves m ight lead to
th e d evelopm ent o f an intrad ay federal funds m ar­
ket, w ith len d ers m aking reserve b a la n ces available
to b orrow ers for only p art o f th e b u sin ess day.
Som e analysts th ink this co u ld lead to greater
variability in an overnight federal fund s rate and
o th er in terest rates.9
Banks cou ld lim it th e size o f th e ir daylight over­
drafts by delaying w ire tran sfers o f fund s for d e ­
p ositors th at do not d em an d im m ed iate delivery
of fund s; or, they m ight charge an extra fee to d e­
positors th at d em an d im m ed iate delivery.
Clearing banks w ould ch arge governm ent se c u ­
rities d ealers for th e co st o f th e fee on daylight
overdrafts. G overnm ent secu rities dealers, in turn,
w ould in crea se th e tran sactio n co sts o f buying
an d selling governm ent secu rities. In terest rates
on governm ent secu rities w ould rise som ew hat
relative to yield s on alternative investm ents, in ­
creasin g th e T reasu ry ’s co st o f servicing the n a ­
tional debt.
How banks react to daylight overdraft fees cou ld
affect m arket yield s on o th e r fin an cial in stru ­
m ents. F o r in stan ce, th e fee on overdrafts w ould
in crease th e co sts to banks actin g as agen ts for
firm s th at issu e co m m ercial p ap er. T h e resp o n ses
by th e agent ban k s co u ld in crea se th e co sts to
firms o f raising fund s by issuing co m m ercial
p ap er.10

Supplemental Balance R equirem ent
T h e Fed eral Reserve cou ld im p ose an im plicit
p rice o n daylight overdrafts by requiring th e banks
9Task Force (1988), pp. 103-14.
10To illustrate the potential effects on the cost of issuing commer­
cial paper, suppose the Federal Resen/e charges 100 basis
points at an annual rate on the maximum daylight overdraft of
each bank. See Mengle, et al. (1987) for the basis for such a
rate. If an agent bank continues the timing of payments de­
scribed in appendix 2 in issuing and redeeming commercial
paper, the overdrafts fee would cost $54.79 per $1 million of
commercial paper issued and redeemed. If the banks pass this
cost on to the issuers, the annual cost of raising funds by
issuing commercial paper every 30 days would rise by 7 basis
points.
"A risk-based capital ratio is calculated as a measure of capital
divided by weighted assets, with weights assigned as approxi


th at overdraw th e ir reserve a cc o u n ts to h o ld su p ­
p lem en tal reserve b a la n ces. T h e se requ irem en ts
w ould b e set to cover part o r all o f th e ir daylight
overdrafts. T h e suggested in terest rate to b e paid
on th e su p p lem en tal b a la n ces w ould b e slightly
below th e federal fund s rate, th u s creatin g an o p ­
p ortun ity co st o f h old ing su p p lem en tal reserves.
T h is co st w ould have th e sam e im p licatio n s for
bank behavior an d fin an cial m arkets as an equal
explicit fee on daylight overdrafts.
T h e im p licatio n s o f a su p p lem en tal reserve r e ­
q u irem en t ca n b e exam in ed by ad ju stin g th e b a l­
a n ce sh ee t en tries in table 1. In th is case, B ank A
w ould b e requ ired to in crea se its average end -ofday reserve b a la n ce by $15. A reserve b a la n ce o f
$25 at th e start o f th e day w ould elim in ate th e risk
o f Fed eral Reserve loss b ec a u se Bank A’s reserve
b a la n ce w ould n ot fall belo w zero after th e $25
transfer.
T h e m eth o d by w h ich Bank A raises th e $15
su p p lem en tal b a la n ce affects th e d istribu tion o f
p o ten tial lo sses am ong p a rticip a n ts in th e banking
industry. Su ppose, for exam ple, B ank A sold som e
a ssets to obtain th e $15 in ad d ition al reserves. T h is
resp o n se w ould raise the risk-ad ju sted cap ital
ratio o f Bank A, u n less it shifted th e rem ain ing
$110 o f o th e r a sse ts into categ ories w ith h ig h er
risk w eights. A rise in Bank A’s risk-ad ju sted ca p i­
tal ratio w ould red u ce th e FD IC ’s poten tial
lo sse s."
Suppose, instead , th at Bank A raises th e $15 in
su p p lem en tal reserves by in creasin g federally
in su red d ep o sits from $100 to $115. T his resp o n se
w ould in crea se th e p o ten tial lo sses faced by th e
FDIC.12
B ank A also co u ld raise th e ad d ition al $15 in th e
term fed eral fu n d s m arket. T h e claim s o f th o se
selling term fed eral fu n d s to B ank A w ould be
su bord in ate to th e claim s o f Bank A’s d ep o sito rs.
Thu s, th e su p p lem en tal b a la n ce requ irem en t
w ould shift risk to th o se banks supplying th e term
mations to relative risk. Reserves have a weight of zero. See
“ Proposals for International Convergence” (1988).
l2Assume that these additional federally insured deposits have a
zero reserve requirement. To illustrate the implications for
FDIC risk, suppose that after Bank A transfers $25 to Bank B,
there is a public announcement of events that reduce the value
of the assets of Bank A by $15. Bank A fails and the FDIC
becomes the receiver. As receiver, the FDIC obtains assets
worth $110 and assumes liabilities of $115, for a net loss of $5.
In this case, therefore, the supplemental balance requirement
shifts risk from the Federal Reserve to the FDIC.

IA Ml IA D V / C C D D I I A D V 1 QQQ

10

federal funds, in creasin g th e system ic risk in th e
banking system .
Of co u rse, su p p lem en tal b a la n ce requ irem en ts
also w ould give banks an in centive to re d u ce th e
size o f th e in trad ay m ovem ents in th eir reserve
b alances, sin ce th e in terest rate paid on th e b a l­
a n ce s w ould be below th e m arginal retu rn on
o th e r assets an d belo w th e in terest rate on federal
funds. T h e su p p lem en tal b alan ce requ irem en t
w ould b e red u ced to th e ex ten t th at a b an k kept
its reserve b alan ce positive th rou gh ou t th e b u si­
n ess day. Su ppose, for in stan ce, th at Bank A
ch an g es its intrad ay p attern o f p ay m en ts so that,
w ith th e su p p lem en tal req u irem en t o f $15, its
reserve b alan ce never falls b elo w $5. T h e Fed eral
Reserve m ight red u ce its su p p lem en tal b alance
requ irem en t to $10, th u s red u cin g th e opp ortu n ity
co st o f Bank A.

Provisions f o r Settlement Finality o f
Payments over CHIPS
Settlem en t finality w ould involve p ro ced u res for
ensu ring th e ex ecu tio n o f pay m en ts (avoid u n ­
w ind ing p ay m en ts involving a defaulting bank)
an d th e allo catio n o f lo sses in th e event o f a d e­
fault b y a CHIPS p articip an t.13 If lo sses are sp read
w idely am ong CHIPS p articip an ts, th e failure o f a
CHIPS p articip an t to m eet its p aym ent obligation
w ould probably n ot ca u se o th er banks to fail.
T h e im p licatio n s o f settlem en t finality arran g e­
m en ts for p ay m en ts system risk are illu strated
u sin g th e b a la n ce sh ee t en tries in table 2. In th is

13Discussions of the finality of payments on private wire transfer
systems mention three aspects of finality. Sender finality
makes each message over the payments system final when
sent. Payment messages cannot be canceled later in the day.
The rules for payment messages on CHIPS include sender
finality.
Settlement finality refers to procedures that would ensure the
settlement of payments if a participant defaults on its net debit
at the end of the day. CHIPS does not have settlement finality
procedures in place at this time. Under current procedures,
CHIPS would cancel all payments by the bank that defaults, as
well as all payments to that bank, and calculate new net debit
or credit positions for the remaining participants. This section
illustrates the implications of adopting a form of settlement
finality.
Under receiver finality, credits to the deposit accounts of the
customers of CHIPS participants would be final when the
receiving banks receive payments messages over CHIPS. If a
sending bank defaults, the receiving bank would have no
recourse to its depositors. CHIPS rules do not include receiver
finality. For additional discussion of these aspects of the finality
of payments, see Humphrey (1986) and Belton, et al. (1987).
t4CHIPS has considered developing a bankers’ bank to ensure
that payment obligations over CHIPS would be treated as net
rather than gross obligations in the case of a default by a
CHIPS participant. See Kantrow (1988). To illustrate the signifi


illu stration, CHIPS is p resu m ed to have form ed a
ban kers' bank, w h ich is a cooperative venture th at
perform s banking services for CHIPS m em bers.
T h is in stitu tio n p ro c e sse s p ay m en t m essages for
its m em b ers as d ebit an d cred it en tries to th e ir
dem an d d ep osit a cc o u n ts at th e b a n k ers’ b an k .14
T h e illu stration is b a sed on som e gen eral p rin ci­
p les o f settlem en t finality arran g em en ts th a t have
b ee n co n sid ere d for several y ea rs.15
T h e h y p o th etica l arran g em en t requ ires m e m ­
b ers o f CHIPS as a group to pled ge en o u gh co lla t­
eral w ith th e ir b a n k ers’ ban k to cover th e largest
net debit p o sitio n o f any o n e p articip an t. T h is is
b a sed o n th e id ea th a t a default by o n e large p a r­
ticip a n t w ould d isru p t th e o p eratio n o f CHIPS.
Sin ce th ere h a s never b e e n a default by a CHIPS
p articipan t, how ever, a default by on e large p a rtic­
ipant is an unlikely event. C ollateral requ irem en ts
for CHIPS p articip an ts in e x cess o f th e largest n et
d ebit o f an individual CHIPS p a rticip a n t co u ld b e
in terp reted as an excessive d egree o f p recau tion .
In table 2, th e largest n et d ebit p o sitio n is $25.
To cover this p o sitio n (and to allow som e m argin
for error), CHIPS req u ires e a c h o f th e th ree banks
to pledge $10 o f th e ir in terest-ea rn in g assets w ith
CHIPS in th e form o f T reasu ry secu rities.
Su ppose th at after CHIPS p ro c e sse s th e tra n s a c ­
tion s d escrib ed in table 2, an a n n o u n c e m e n t in d i­
ca tes a $15 lo ss in th e value o f B an k A’s assets.
U nder th e settlem en t finality arran gem en t, CHIPS
w ould u se th e co llateral p o sted b y its p articip an ts
to raise $25, eith e r by selling p art o f th e co llateral

cance of the distinction between gross and net obligations,
suppose a bank fails while it is in a net credit position on
CHIPS payments. If CHIPS obligations are treated legally as
net obligations, CHIPS participants would make a payment to
the receiver of the failed bank for the amount of the net debit
position. The receiver of the failed bank might sue CHIPS
participants based on gross obligations. Under a successful
suit by the receiver, those that had sent payment messages to
the failed bank would have to pay the gross amount of those
payments, and those who received payment messages from
the failed bank would become its general creditors for the
amount of the gross transfers from the failed bank. This treat­
ment of CHIPS participants would increase the recovery rate of
the failed bank’s other general creditors. There have been no
such cases to indicate whether the courts would uphold pay­
ments to the receiver based on gross payments.
Suppose, in contrast, that CHIPS payments are processed
through demand deposit accounts at the bankers’ bank for
CHIPS. Under that arrangement, the only claim of the receiver
of the failed bank would be for the positive balance of the failed
bank in its demand deposit account at the bankers’ bank.
15Mengle (1989).

11

or u sing th e secu rities as co llateral for a lo an at
th e Fed eral Reserve d isco u n t w indow . CHIPS
w ould th e n tran sfer th e $25 to th e reserve a cco u n t
of Bank B, facilitating th e p ay m en t from Bank B to
Bank C. In tu rn, th e b an k ers’ bank o f CHIPS w ould
h o ld th e $10 in collateral p o sted by Bank A and
have a $15 claim against Bank A as a general cred i­
tor. L osses on th e $15 claim against Bank A w ould
th u s be sp read b etw een Bank B an d Bank C. Nei­
th e r bank w ould b e fo rced in to b an k ru p tcy by a
co m p lete loss on th e $15 claim .
From th e Fed eral R eserve’s p erspective, th is
settlem en t finality arran g em en t is b e tte r th an th e
p ro ced u re th at cu rren tly w ould b e u sed to deal
w ith a default by a CHIPS p articip an t — u n w in d ­
ing pay m ents involving th e bank. If th is settlem en t
finality arrang em ent w ere in p lace, th e u nw inding
o f p aym en ts, w h ich w ould d isru p t th e flow of
p aym en ts in th e econ om y, co u ld be avoided. If a
d isco u n t w ind ow loan w as n ecessary to avoid a
liquidity crisis in th e banking system , th e co llat­
eral w ould b e available throu gh th e CHIPS organi­
zation. T h e Fed eral Reserve w ould n o t have to
d ecid e w h ich banks sh o u ld receive d isco u n t w in ­
dow loans.
By m aking th e risk to CHIPS p articip an ts m ore
explicit, th e arran g em en t w ould give CHIPS p a rtic­
ip an ts strong er incentives to ex clu d e banks in
relatively p o o r fin an cial co n d itio n from th e ir sys­
tem . B anks th at are exclu d ed w ould ro u te th eir
w ire transfers throu gh Fedw ire, th u s red u cin g
system ic risk. Finally, th e spread ing o f p o ten tial
lo sses w ould lim it th e c h a n c e s o f th e failure o f one
ban k cau sin g o th ers to fail. It is n o t p o ssible to
d eterm in e w h eth er th e risk o f ban k failure is low er
u n d er cu rren t CHIPS p ro ced u res o r u n d e r this
p ro p o sed p ro ced u re for settlem en t finality. Su ch a
co m p ariso n d ep en d s on th e ex ten t to w h ich d e ­
p ositors o f CHIPS p articip an ts draw dow n th e
intrad ay cred its to th eir d em an d d ep o sit acco u n ts
a n d th e su cce ss th at banks w ould have in co lle c t­
ing from th o se d ep o sito rs in case o f a default by a
CHIPS p articip an t.

CONCLUSIONS
All banks assu m e som e risk by p articip atin g in
th e paym en ts system . T h e p aym ent p ra ctice s that
g en erate th is risk w ere developed in an environ­
m ent in w h ich th ere w as n o in terest ch arge on
intrad ay cred it and, u ntil recently, no co n strain ts
on th e m agnitu d e o f intraday cred it. T h ere have
b een no lo sses to th e Fed eral Reserve o r to m e m ­



b ers o f private w ire tran sfer system s resu lting from
th e daylight cred it ex p o su res. T h e F ed eral Re­
serve, how ever, h as ad o p ted a p o licy on pay m en ts
system risk w h ich in clu d es lim its on th e daylight
overdrafts o f individual banks.
T h e F ed h a s b ee n co n sid erin g p o ssib le ch an g es
in its p o licy to re d u ce its ow n risk an d provide
in cen tiv es for ban k s to ch an g e th e p ay m en t p ra c­
tice s th at ten d to crea te th e intrad ay risk ex p o ­
su res. O ne p ro p o se d a p p ro a ch involves a fee on
daylight overdrafts o f reserve a cco u n ts. A seco n d
a p p roach , w h ich involves an im p licit p rice on
daylight overdrafts, requ ires ad d ition al reserve
b a la n ces at th e banks w h ich regularly overdraw
th eir reserve a cc o u n ts during th e day. T h e Fed eral
Reserve w ould pay in tere st on th e se su p p lem en tal
reserve b a la n ces at a rate ju s t b elo w th e federal
fund s rate. U nd er eith e r ap p roach , CHIPS w ould
b e required to w ork out an arran g em en t th at is
satisfactory to th e F ed eral Reserve to en su re th e
finality o f its paym en ts.
T h e objective o f ch an g in g th e p o licy on pay ­
m en ts system risk is to re d u ce th e risk o f th e F e d ­
eral Reserve w ith ou t creatin g a large in crea se in
sy stem ic risk — th e risk th at th e failure o f one
bank will ca u se th e failure o f o th e r banks, th u s
disrupting th e o p eratio n o f th e p ay m en ts system .
T h e type o f settlem en t finality arran g em en t d e­
sired by th e Fed eral Reserve w ould en su re the
ex ecu tio n o f pay m en ts over CHIPS in th e event of
a default by a CHIPS p a rticip a n t an d sp read any
lo sses so w idely am ong o th e r CHIPS p articip an ts
th at on e ban k failure is unlikely to lead to th e fail­
ure o f o th e r CHIPS p articip an ts.

REFERENCES
Association of Reserve City Bankers. Report of the Working
Group of the Association of Reserve City Bankers on BookEntry Daylight Overdrafts (June 1986).
Belton, Terrence M., et al. “ Daylight Overdrafts and Payments
System Risk,” Federal Reserve Bulletin (November 1987), pp.
839-52.
Corrigan, E. Gerald. Financial Market Structure: A Longer View
(Federal Reserve Bank of New York, January 1987).
Faulhaber, Gerald R., Almarin Phillips, and Anthony M. Santomero. “ Payment Risk, Network Risk and the Role of the
Fed,” in David B. Humphrey, ed., U.S. Payment System:
Efficiency, Risk and the Role of the Federal Reserve System
(Kluwer, 1989).
Flannery, Mark J. “ Payments System Risk and Public Policy,”
Mimeo, University of North Carolina at Chapel Hill, November
30,1987.
Humphrey, David B. “ Payments Finality and Risk of Settlement
Failure,” in Anthony Saunders and Lawrence J. White, ed.,
Technology and the Regulation of Financial Markets (Lexing­
ton Books, 1986).

JANUARY/FEBRUARY 1989

12

Johnson, Manuel H. “Challenges to the Federal Reserve in the
Payments Mechanism,” Issues in Bank Regulation (Summer
1988), pp. 13-16.

Mengle, David L., David B. Humphrey, and Bruce J. Sum­
mers. “ Intraday Credit: Risk, Value and Pricing,” Federal
Reserve Bank of Richmond Economic Review (January/
February 1987), pp. 3-14.

Kantrow, Yvette D. “ Big NY Banks May Spin Off Chips Net­
work,” American Banker (June 27, 1988), pp. 1, 23.

“ Proposals for International Convergence of Capital Measure­
ment and Capital Standards.” Issues in Bank Regulation
(Winter 1988), pp. 3-12.

Large-Dollar Payments System Advisory Group. A Strategic
Plan for Managing Risk in the Payments System, Report to the
Payments System Policy Committee of the Federal Reserve
System, Board of Governors of the Federal Reserve System,
August 1988.
Mengle, David L. “ Legal and Regulatory Reform in Electronic
Payments: An Evaluation of Finality of Payment Rules,” in
David B. Humphrey, ed., U.S. Payment System: Efficiency,
Risk and the Role of the Federal Reserve System (Kluwer,
1989).

Rothschild, Michael, and Joseph E. Stiglitz. “ Increasing Risk: I.
A Definition,” Journal of Economic Theory (September 1970),
pp. 225-43.
Task Force on Controlling Payments System Risk. Controlling
Risk in the Payments System, Report to the Payments System
Policy Committee of the Federal Reserve System, Board of
Governors of the Federal Reserve System, August 1988.
VanHoose, David. “The Angell Proposal: An Overview,” staff
paper, Board of Governors of the Federal Reserve System
(June 1988).

Appendix 1
C urrent Federal Reserve Policy on Paym ents
System Risk
C urrently, th e F ed eral Reserve u ses sp ecific
lim its on daylight overdrafts o f reserve a cc o u n ts
an d n et d ebit p o sitio n s on private w ire tran sfer
sy stem s to re d u ce pay m ents system risk. T h e lim ­
its on n e t d ebit p o sitio n s apply to any private w ire
tran sfer system th at settles th e net p o sitio n s o f its
p articip an ts th rou gh tran sfers o f b a la n ces in re ­
serve o r clearin g a cc o u n ts at Reserve Banks. Sin ce
CHIPS is th e only su ch system in operation, the
follow ing d escrip tio n refers only to it, bu t w ould
apply to any su ch system d eveloped in th e future.1

Bilateral Net Credit Limits on CHIPS
T h e Fed eral Reserve requires e a c h p articip an t
on CHIPS to set a lim it on its n et cred it p o sitio n on
m essage transfers w ith ea ch o f th e o th e r p a rtici­
pan ts in th e system . Fund s tran sfer m essages that
violate th e se bilateral n et cred it lim its are re je cted
by th e co m p u te r system th at p ro c e sse s p aym en t
m essages. CHIPS p articip an ts have h ad bilateral
cred it lim its sin ce O ctober 1984.

S end er Net Debit Caps on CHIPS
T h e Fed eral Reserve requires CHIPS to estab lish
lim its on th e n et d ebit p o sitio n s o f ea ch p a rtici­

'For an analysis of the effects of these credit limits on daylight
overdrafts and the operation of the payments system, see Belton, et al. (1987).

http://fraser.stlouisfed.org/
FEDERAL
RESERVE
Federal Reserve
Bank of St.
Louis BANK OF ST. LOUIS

p an t w ith all o th e r p articip an ts on th e system ;
CHIPS sets th is lim it for ea ch p a rticip a n t at 5 p e r­
ce n t o f th e sum o f all bilateral cred it lim its for that
p articip an t ex ten d ed by all o th e r CHIPS p a rtici­
pan ts.2 CHIPS estab lish ed th e se s en d e r n et d ebit
cap s in O ctob er 1985.

Cross-System Caps
E ach bank th at o cca sio n a lly h as daylight reserve
overdrafts is required to adopt a cap on its c ro ss ­
system daylight overdraft. C ross-system refers to
th e daylight overdraft p o sitio n on Fedw ire an d
CHIPS. T h e relevant overdraft p o sitio n for th is cap
is th e su m o f a b a n k ’s fu n d s-related overdraft o f its
reserve a cc o u n t an d its n et debit p o sitio n on
CHIPS at e a c h m o m en t d uring th e day. E ach bank
sets its cap by p lacin g itse lf in o n e o f th e possible
categ ories in d ica ted in tab le A l; ban k s are d irected
to co n sid er th e ir cred itw orth in ess, cred it p o licies
an d o p eratio n al co n tro l an d p ro ced u res. E a ch
p o ssib le rating h a s co rresp o n d in g ca p s for b o th
th e o n e day an d tw o-w eek average m axim u m day­
light overdraft, ea ch as a p e rce n ta g e o f prim ary
ad ju sted cap ital. T h e se p e rcen ta g es have b een

2There are additional details involved in determining these limits,
See Belton, et al. (1987).

13

Table A1
Caps on Daylight Overdrafts Across Payments Systems
(multiples of adjusted primary capital)
Selfassessment
category

High

Above
average

Average

Limited

Period caps in effect

Cap
applied
to

March 27,1986 to
January 13,1988

January 14,1988
to May 18,1988

May 19,1988
to present

Two-week
average

2.000

1.700

1.500

Single day

3.000

2.550

2.250

Two-week
average

1.500

1.275

1.125

Single day

2.500

2.125

1.875

Two-week
average

1.000

0.850

0.750

Single day

1.500

1.275

1.125

Two-week
average

0.500

0.425

0.375

Single day

0.500

0.425

0.375

NOTE: Adjusted primary capital for U.S.-chartered banks is the sum of primary capital less all intangible
assets and deferred net losses on loans and other assets sold.
SOURCE: Federal Reserve Bulletin (November 1987), p. 843.

red u ced over tim e to m ake th em m o re effective in
co n strain in g overdrafts.

Book-Entry Securities Transfers
In calcu latin g th e relevant m easu re o f overdrafts
for th e cro ss-sy stem cap s, th e Fed eral Reserve n ets

o u t th e value o f b o o k -en tiy secu rities cred ited to
th e a cc o u n t o f th e bank. T h is step ex em p ts day­
light overdrafts g en erated th rou gh secu rities
tra n sa ctio n s from th e lim its im p o sed by th e cap s.
T h e Fed eral Reserve h as allow ed th is d istin ction
to avoid d isru p tin g th e m arket for U.S. governm ent
secu rities.

Appendix 2
Additional Illustrations of Paym ents and Risk
Transfers f o r Depositors Over
Fedwire
W ire tran sfers o f fund s for d ep o sito rs m ay cau se
banks to overdraw th e ir reserve acco u n ts, as table
A2 illu strates. A d ep o sito r in stru cts Bank A to pay



$25 to a d ep o sito r o f Bank B in th e form o f a w ire
transfer. Sin ce th e initial reserve b a la n ce is only
$10, th e $25 tran sfer m akes th e reserve a cc o u n t of
Bank A overdraw n by $15. As in table 1 in th e text,
an a n n o u n cem e n t o f a $15 d eclin e in th e value of
th e a sse ts o f B an k A w ould force th e F ed eral R e­
serve to ab so rb a $5 loss.

JANUARY/FEBRUARY 1989

14

Table A2
Risk Created by Transferring Depositor’s Funds over Fedwire
Balance sheets at start of day:

Bank B

Bank A

Reserves
Other
assets

$10 Deposits
Net
100 worth

$100
10

Reserves
Other
assets

$10 Deposits
Net
100 worth

$100
10

Bank A sends $25 of depositor’s money to Bank B over Fedwire:
Bank B

Bank A

Reserves
Other
assets

-$ 1 5

Deposits

Net
100 worth

$ 75
10

Reserves
Other
assets

$35 Deposits
Net
100 worth

$125
10

Value of other assets at Bank A reduced by $15:
Bank A

Reserves
Other
assets

- $ 15 Deposits
Net
85 worth

Bank B

$ 75
-5

Securities Transfers
A few ban k s in c u r large daylight overdrafts b e ­
ca u se o f th e tra n sa ctio n s th ey co n d u ct for c u s ­
to m ers th at deal in U.S. governm ent secu rities.
T h e se tran sactio n s w arran t sp ecial exam in ation . A
few large banks (called clearin g banks) sp ecialize
in serving governm ent secu rities d ealers; th ese
banks g en erate a large sh are o f th e to tal daylight
overdrafts o f ban k reserve a cco u n ts. In th e seco n d
q u arter o f 1988, for exam ple, fou r clearin g banks
a cco u n ted for abou t 70 p e rce n t o f th e daylight
overdrafts attributable to tran sactio n s in booke n tiy secu rities.

B u sin ess P ra c tic e s of D ealers an d C learing
Banks — G overnm ent secu rities d ealers w ho buy
an d sell secu rities for th e ir cu sto m ers have no
d irect a c c e ss to th e b o o k -en tiy system for tra n sfer­
ring ow nership o f governm ent secu rities. Instead ,
th ey m ain tain b o o k -en tiy secu rities a cc o u n ts and
d em an d d ep o sit a cc o u n ts w ith co m m ercial banks

'For a more complete discussion of the practices of clearing
banks and dealers, see Association of Reserve City Bankers
(1986).

http://fraser.stlouisfed.org/
FEDERAL
BANK OF ST. LOUIS
Federal Reserve Bank
of St.RESERVE
Louis

Reserves
Other
assets

$35 Deposits
Net
100 worth

$125
10

that serve as th eir clearin g banks for secu rities
transfers.
Daylight overdrafts o f th e clearin g b a n k s’ reserve
a cc o u n ts reflect th e p ra c tice s o f th e governm ent
secu rities d ealers in m anaging th e ir in ven tories o f
governm ents secu rities. D ealers h o ld large inven­
tories o f secu rities d uring th e day to m eet the
a n ticip a ted d em an d s o f th e ir cu sto m ers. T o m in i­
m ize th e co st o f h old ing th e inventories, th e d eal­
ers sell m o st o f th e ir secu rities by th e en d o f th e
day th rou gh re p u rch a se ag reem en ts. T h e inves­
tors w ho e n te r in to th e se ag reem en ts "o w n ” th e
secu rities overnight an d “re se ll” th em to d ealers
early th e n ex t day. T h u s, th e d ealers b u ild th eir
inventories o f governm ent secu rities in th e m o rn ­
ing o f e a c h b u sin e ss day by receiving secu rities
retu rn ed by th e overnight repo investors an d bu y­
ing ad d ition al secu rities offered for sale.1
T h e follow ing featu res o f th e b u sin ess p ra ctice s
o f governm ent secu rities d ealers exp lain w h y th ey
gen erally w ait u ntil early aftern oo n to b egin ru n -

15

ning dow n th e ir inventory o f secu rities. Salesm en
for a d ealer m ake co m m itm en ts to deliver sp ecific
secu rities to its cu sto m ers by th e en d o f th e day.
T h e d ealer is th en vu lnerable to lo sses if it can n o t
fulfill th e se co m m itm en ts. T h e cu sto m ers receive
in terest on th e p ro m ised secu rities for th at day,
even if th e d ealer d o es n o t m ake deliveiy. T h e
cu sto m ers, how ever, m ake p ay m en ts to th e d eal­
ers only w h en the secu rities are delivered. T h e
d ealer w ould fail to m ake deliveiy if it co u ld not
lo cate the d esired secu rities in its inven toiy o r in
th e m arket, or if it sen t th e w rong secu rities to a
cu sto m e r an d h ad th em retu rn ed . E a ch d ealer
attem p ts to m inim ize th e p robability o f su ch
“fails” by w aiting u ntil early aftern oo n to d irect its
clearing bank to send its secu rities to the bookentry acco u n ts o f th e banks th at serve th e c u s ­
tom ers th at have bo u g h t th em .
A no th er re a so n th e d ealers h o ld th e ir secu rities
until early aftern oo n involves p o ten tial profits
from sp ecial ord ers. On som e days, certain issu es
o f governm ent secu rities are in relatively high
d em an d. T h e d ealers ca n m ake larger profits if
th ey have secu rities available to m eet th e se sp ecial
ord ers. In co n trast to th e sp ecific req u irem en ts for
sp ecial orders, d ealers m ay su b stitu te a w ide vari­
ety o f secu rities as a ccep ta b le collateral for repos.

Effects on Intraday R eserv e B a la n ce s —
T h e se d ealer p ra ctice s affect th e in trad ay p attern s
o f th e ir d em an d d ep o sit b a la n ces an d th e reserve
b a la n ces o f th e clearin g ban k s th at serve them .
W hen a repo investor retu rn s th e secu rities to the
dealer, th ere is an in crease in th e secu rities a c ­
co u n t o f th e d ealer at its clearin g b an k and an
equal red u ctio n in its d em an d d ep o sit acco u n t.
On th e books o f th e Fed eral Reserve, th ere is an
in crease in th e secu rities in th e book-en try a c ­
co u n t o f th e clearing bank and a red u ctio n in the
reserve a cc o u n t o f th e clearing bank. T h e sam e
tra n sactio n s o c c u r w h en th e d ealer buys se c u ri­
ties to hold in its inventory th at day. T h e d ealer
bu ilds its inventory o f secu rities by overdraw ing
its d em and d ep osit a cc o u n t during th e day. T h e
d ealers do not co n tro l th e tim ing o f th ese inflow s
o f secu rities to th e ir a cc o u n ts and th e outflow s
from th e ir d em an d d ep o sit acco u n ts, sin ce the
party th at h old s th e secu rities initiates th e tran sfer
o f secu rities and reserves th rou gh th e Fedw ire
system .

a cc o u n ts o f th e clearin g banks rise as th e booken try secu rities are tran sferred to th e a cc o u n ts of
o th e r banks an d reserve b a la n ces are sim u ltan e­
ously tran sferred to the a cc o u n ts o f th e clearing
banks. T h e tim ing o f tra n sa ctio n s in book-en try
secu rities for th e d ealers ca u ses th e reserve a c ­
co u n ts o f th e clearin g banks to b e overdraw n by
b illions o f dollars during part o f th e day.

Im p lication s for Risk — T h e clearin g banks
exten d cred it to governm ent secu rities d ealers
during th e day by allow ing th em to overdraw th eir
d em an d d ep o sit a cco u n ts. T h e banks lim it th e ir
risk by obtain in g a lien against th e secu rities h eld
for th e a cco u n t o f th e d ealers. T h u s, a clearin g
bank co u ld claim th e secu rities cred ited to the
a cco u n t o f a d ealer to cover any lo sses on its d e ­
p o sit overdraft.2
T h e F ed eral Reserve h a s co n sid ere d various
m eth o d s o f estab lish in g lien s against th e se c u ri­
ties in th e book-en try a cc o u n ts o f banks bu t h as
n ot in itiated su ch collateral arran g em en ts. Thus,
th e Fed is vu lnerable to lo sses on th e full am ou n t
o f a b an k 's reserve overdraft, w h eth er th e overdraft
w as g en erated th rou gh fund s tran sfers o r tra n sa c­
tio n s in book-en try secu rities.3
T h e risk im p licatio n s o f book-en try overdrafts
ca n b e illu strated by exam in in g th e b a la n ce sh eet
en tries in table A2. B ank A is a clearin g ban k for a
governm ents secu rities dealer. T h e d ealer receives
$25 in b ook-en try secu rities an d h a s its d em an d
d ep osit a cco u n t d ebited by $25, leaving it over­
draw n at th at tim e. Su p p o se th e d ealer goes ban k ­
ru p t after th is tra n sa ctio n is co m p leted . Bank A
claim s th e $25 in secu rities th a t w ere cred ited to
th e secu rities a cc o u n t o f th e d ealer to cover any
p o ssib le lo sses on th e d ep o sit overdraft. T h e bank
is sp ared any lo sses, an d th e Fed eral Reserve suf­
fers no losses.

T h e p ro c e ss o f overdraw ing reserve and d ep osit
a cc o u n ts is reversed later in th e day as th e d ealers
sell th e ir inventories o f secu rities. T h e reserve

T h is b ook-en try daylight overdraft, how ever,
d oes leave th e F ed eral Reserve vu lnerable to a loss
on th e reserve overdraft. Su p p o se th at after th e
d ealer receives th e $25 in b o o k -en tiy secu rities,
th ere is an a n n o u n cem e n t th at im plies a $15 loss
in th e value o f th e o th e r assets o f Bank A, as in th e
o th e r illu stration s. U nd er cu rre n t arran gem en ts,
th e Fed has no claim on th e $25 in book-en try
secu rities th at h ad b ee n tran sferred to Bank A, to
offset its $5 loss. Thu s, collateral ag reem en ts b e ­
tw een clearin g ban k s an d th e d ealers m ake F e d ­
eral Reserve lo sses due to defaults by governm ent
secu rities d ealers unlikely, b u t th e daylight reserve

2Task Force (1988), p. 69.

3Task Force (1988), p. 70-72.




JANUARY/FEBRUARY 1989

16

Table A3
The Effects of Issuing Commerical Paper on the Balance Sheet of an Agent Bank
Balance sheets at start of day:

Bank A

Reserves
Other
assets

$10 Deposits
Net
100 worth

Bank B

$100
10

Reserves
Other
assets

$ 10 Deposits
Net
100 worth

$100
10

Bank A transfers $25 to Bank B, credited to the account of the firm that issues commercial paper:
Bank A

Reserves
Reserves
receivable
Other
assets

-$ 1 5

Deposits

Bank B

$100

Other
assets

25
Net
100 worth

Reserves

$35 Deposits
Net
100 worth

$125
10

10

Bank A receives $25 from purchaser of commercial paper:
Bank A

Reserves
Reserves
receivable
Other
assets

$10 Deposits

Bank B

$100

0
Net
100 worth

$ 10 Deposits

Other
assets

Net
100 worth

$100
10

10

overdrafts of the clearing banks expose the Fed to
potential losses in the event of large, u n an tici­
p ated declines in the value of the assets of the
clearing banks them selves.
A lien by th e Fed eral Reserve against th e booke n tiy secu rities in th e a cc o u n ts o f th e clearin g
banks m ight have little p ractical sign ifican ce in
lim iting Fed risk. Su p p o se th e p u b lic learn s during
th e day th at a clearing ban k m ay b e bankrupt.
W ould th e Fed eral Reserve su dd en ly seize th e
book-entry secu rities in th e a cco u n t o f th e cle a r­
ing bank? D oing so w ould disrupt th e b u sin ess of
th e governm ent secu rities dealers served by th e
clearing bank and, given th e high co n cen tra tio n of
b u sin ess am ong clearin g banks, w ou ld d isrupt
trading in th e w h ole governm en t secu rities m ar­
ket. T h e Fed an d th e o th e r fed eral supervisory
a u th o rities have b een relu ctan t to clo se large co m ­

“For a discussion of how daylight overdrafts reflect transactions
in commercial paper and other financial instruments, see
Large-Dollar Payments System Advisory Group (1988).

http://fraser.stlouisfed.org/
BANK OF ST. LOUIS
Federal Reserve FEDERAL
Bank of St.RESERVE
Louis

Reserves

m ercial ban k s b ec a u se o f th e ir effects o n o th e r
d epository in stitu tio n s an d th e fin an cial m arkets
in general. A lien on th e b o o k -en tiy secu rities of
banks m ight m ake th e supervisory au th orities
m ore relu cta n t to clo se a large ban k th at also
serves as a clearin g ban k fo r g overnm ent secu rities
dealers.

Issuing and R edeem ing Commercial
Paper
T h e tim ing o f pay m en ts by banks involved in
issu in g an d red eem in g co m m ercia l p a p e r crea tes
reserve overdrafts.4 Several banks act as agen ts for
firm s th at issu e co m m ercial p aper. T h e agent
banks co lle ct fu n d s from th o se p u rch a sin g th e
co m m ercial p ap er an d tran sfer th em to th e a c ­
co u n ts o f th o se firm s issu in g th e p ap er. W h en the
p a p er m atures, th e agent banks co lle ct from th e

17

p ap er issu ers an d m ake pay m en ts to the hold ers
of th e paper.
W h en a firm issu es co m m ercial paper, th e agent
bank generally pays th e firm befo re it receives
paym ent from th o se buying th e paper. D uring th e
period betw een th e pay m en t to th e issu er an d th e
receip ts from th e p u rch ase rs, th e reserve a cco u n t
o f th e agent bank falls by th e am ou nt o f th e funds
raised by issuing th e co m m ercial paper. T h e re ­
serve b alan ce o f th e agent bank also falls by the
face am ou n t o f th e issu e w h en th e p ap er m atures;
th e agent bank generally m akes p aym ent to th ose
holding the p ap er b efo re receiving pay m en t from
th e issuer.
T h e effects o f th ese tran sactio n s on th e b alan ce




sh eet o f th e agent bank are illu strated in table A3.
A firm raises $25 by issuing co m m ercia l paper.
Bank A is th e agent bank, an d b o th th e issu e r and
p u rch a se r o f th e p a p er have th e ir d em an d d eposit
a cco u n ts at Bank B. Early in th e day on w h ich th e
co m m ercial p a p e r is issu ed , Bank A tran sfers $25
to Bank B, to b e cred ited to th e d em an d d eposit
a cco u n t o f th e issuer. After th at tran sactio n , the
reserve a cc o u n t o f Bank A is overdraw n by $15. In
th is exam ple, th e offsetting tra n sa ctio n is a $25
in crea se in an a cc o u n t called “reserves receiv ­
a b le.” Later th at day, th e p u rch a se r o f th e p ap er
arranges for Bank B to sen d $25 to Bank A over
Fedw ire, elim in atin g th e reserve overdraft by the
en d o f th e day. As in th e o th e r b a la n ce sh eets, th e
Fed eral Reserve is a gen eral cred ito r o f Bank A
w hile its reserve a cc o u n t is overdraw n.

JANUARY/FEBRUARY 1989

18

Keith M. Carlson
Keith M. Carlson is an assistant vice president at the Federal
Reserve Bank of St. Louis. Thomas A. Pollmann provided re­
search assistance.

Federal Budget Trends and the
1981 Reagan Econom ic Plan

I n EARLY 1981, a new ly inau gu rated Ronald
R eagan a n n o u n ced an e co n o m ic p lan w h ich in ­
clu d ed goals o f “an im m ediate, su bstantial, and
su stain ed red u ctio n in th e grow th o f federal ex ­
p en d itu res [and] a significant re d u ctio n in federal
tax rates . . After two term s in office, it seem s
tim e to exam ine th e original Reagan bu d g et plan
in light o f th e actu al p erfo rm an ce over the 1980s.
A lthough th e bu d g et plan h ad far-reach in g e c o ­
n o m ic an d so cial co n se q u e n ce s, th is article fo­
cu ses on th e ex ten t to w h ich th e initial bu dget
p ro jectio n s w ere realized.2

First, th e 1981 e co n o m ic setting, w h ich provided
th e underlying rationale for th e Reagan plan, is
su m m arized . T h en , b ec a u se th e bu dget and e c o ­
n o m ic co n d itio n s are interrelated , the 1981 e c o ­
n o m ic assu m p tio n s are exam in ed in retro sp ect.
T h is is follow ed by a co m p ariso n o f p lan n ed and
realized ch an g es in fed eral outlays and receip ts.
T h e article co n clu d e s w ith an evaluation o f the
1981 bu d g et plan.

1America’s New Beginning . . . (1981), pp. 1-3. For further
details, see Office of Management and Budget (1981a) and
Carlson (May and November 1981).
2For more extensive analyses of the Reagan years, see Boskin
(1987), Mills and Palmer (1984), Modigliani (1988) and
Niskanen (1988).

http://fraser.stlouisfed.org/
Federal Reserve FEDERAL
Bank of St.RESERVE
Louis
BANK OF ST. LOUIS

THE 1981 ECONOMIC FORECAST IN
RETRO SPECT
W hen th e Reagan ad m in istratio n began p re p a r­
ing its budget in late 1980 an d early 1981, th e U.S.
eco n o m y w as recovering from a b rief re ce s sio n in
th e first h a lf o f 1980. O utput w as grow ing slug­
gishly for a recovery p h ase o f th e b u sin ess cycle,
u n em p loy m en t w as w ell above 7 p e rce n t o f th e
lab or force an d productivity, as m easu red by o u t­
put p er hour, w as d eclining. P rices generally w ere
in creasin g at d ouble-digit rates an d in terest rates
reflected th e high rate o f inflation. T h e federal
budget d eficit for fiscal 1980 w as $60 billion and
th e outgoing ad m in istra tio n ’s estim ate for 1981
w as about $55 billion.
T h e in co m in g p resid en t d escrib ed th e situ ation
as “the m ost serio u s set o f e co n o m ic pro b lem s
sin ce th e 1930s.”3 T h e m o st im p o rtan t ca u se of
th e se problem s, h e suggested, w as th e govern­
m en t itself: th rou gh taxes, spending, regulatory
p o licies an d m o n etary p o licies, it h ad sacrificed

3America’s New Beginning . . . (1981), p. 4.

19

Figure 1
Actual Movements vs. Reagan Forecasts of
Key Econom ic V ariables
Gross National Product (percent change)

Gross National Product Deflator (percent change)




JANUARY/FEBRUARY 1989

20

long-term grow th and p rice stability for ep h em eral
sh o rt-term goals. To co m bat th ese problem s, the
ad m in istration p ro p o sed a program th at w as in ­
ten d ed to:
restore fiscal integrity; increase incentives for
saving, investment, and production; attain m one­
tary and financial stability; and en hance the role of
the m arketplace as the principal force in the allo­
cation of resou rces.4

An im p o rtan t p art o f every bu dget program is
th e set of und erlying eco n o m ic assu m p tio n s.3
Figure 1 show s th e ad m in istratio n ’s 1981 forecasts
for a variety o f key e co n o m ic variables along w ith
th e ir actu al p erfo rm an ce. As th e top tier show s,
th e ad m in istratio n overestim ated th e grow th in
n o m in al GNP from 1980 to 1986 by a su bstan tial
am ou nt.6 In particu lar, it did n o t fo recast th e
1 981-82 re cessio n n o r did it foresee th e sharp
red u ctio n in n om inal GNP grow th after 1984. By
1986, the cum ulative erro r for GNP w as over $800
billion, o r alm ost 20 p e rce n t o f th e actu al level o f
GNP in 1986. T h is erro r reflected an actu al GNP
grow th rate o f 7.8 p e rce n t over th e 19 8 0 -8 6 period,
qu ite a bit low er th an th e assu m ed grow th rate of
11 p e rce n t.7
T h e overestim ate o f n om inal GNP reflected over­
estim ates o f b o th real grow th (secon d tier o f figure
1) an d inflation (third tier o f figure 1). T h e cu m u la ­
tive erro r in forecasts o f real GNP by 1986 w as 7
p e rce n t w hile th e GNP deflator w as overestim ated
by 11 p e rce n t. T h e 1981 adm in istration fo reca st for
inflation for th e 19 8 0 -8 6 p eriod w as a 7.1 p e rce n t
an n u al rate; th e actu al inflation rate d uring th is
p eriod w as 5.1 p e rce n t.8
T h e fourth tier o f figure 1 in d icates th at th e u n ­
em p loy m en t rate w as u n d erestim ated in ea ch of
th e y ears from 1981 to 1986. T h e ad m in istratio n
forecast th at th e u n em p lo y m en t rate w ould rise in
1981, th e n fall to 5.6 p e rce n t by 1986; th e actu al
1986 rate w as 7.1 p ercen t.
4Ibid., p. 9.
5Although such assumptions are absolutely necessary to project
outlays and receipts, economic conditions themselves are
influenced by congressional and legislative decisions that
affect the budget. This was the administration’s reasoning in
1981; its budget programs were designed to have a favorable
effect on the economy. In fact, its economic assumptions were
so optimistic, it felt compelled to say:
Indeed they do represent a dramatic departure from the
trends of recent years — but so do the proposed policies. In
fact, if each portion of this comprehensive economic pro­
gram is put in place — quickly and completely — the eco­
nomic environment could improve even more rapidly than
envisioned in these assumptions. [Ibid., p. 25.]

Generally, from this point on, all references to years are to
fiscal years, i.e., the 12-month period ending September 30.

http://fraser.stlouisfed.org/
BANK OF ST. LOUIS
Federal Reserve FEDERAL
Bank of St.RESERVE
Louis

Finally, as in d ica ted in th e b o tto m tier o f figure
1, th e Treasu ry bill rate w as also u n d erestim ated .
T h e B eagan a d m in istratio n fo reca st a steady d e­
clin e in th e Treasu ry bill rate from m ore th a n 11
p e rce n t in 1980 to 5.7 p e rce n t in 1986; th e actu al
rate rose sharply in 1981, b efo re falling to 6.4 p e r­
ce n t in 1986.
T h e se key eco n o m ic variables generally m oved
unfavorably during th e 1 9 8 0 -8 6 p erio d in term s of
th e ir effect on th e fed eral b u dget. T h e slow erth an -forecast grow th o f n o m in al GNP slow ed the
grow th o f receip ts an d co n trib u ted to a larger
deficit. Although slo w er-th a n -exp ected inflation
h elp ed to red u ce th e grow th o f b u d g et outlays,
slow er real GNP grow th an d h ig h er-th an -fo recast
u n em p loy m en t rates in crea s ed outlays, p a rticu ­
larly for u n em p loy m en t in su ra n ce. M eanw hile, the
h ig h er-th a n -ex p ected T reasu ry bill rate also
b o o sted outlays, esp ecially w h en th e governm ent
w as borrow ing m ore th a n p lan n ed . T h u s, m o st of
the errors in th e ad m in istra tio n ’s forecast w ere
o n es th at in crea sed th e d eficit m o re th a n
p ro jecte d .9

THE BUDGET TOTALS:
REALIZATIONS VS. THE REAGAN
PLAN
As figure 1 in d icated , th e Reagan a d m in istra­
tio n ’s 1981 eco n o m ic assu m p tio n s w ere erro n e­
ous. A related q u estio n is to w h at ex ten t w ere th e
bu dget p ro jectio n s also erro n eo u s? An obvious
m easu re o f th is p a rticu la r erro r is th e d ifference
b etw een the p lan n ed an d th e actu al surplus/
deficit. Figure 2 show s th e size o f this d iscrep an cy .
T h e Reagan plan p ro je c te d a stead y m ove tow ard
a b a la n ced bu d g et by 1986; th e a ctu al d eficit for
1986 w as $221 b illio n .10 To b e tte r u n d erstan d w hy
th e 1981 bu dget p la n 's p ro jectio n s w ere in error,
individual bu dget categ o ries are exam in ed below .”
7Such a projection was not unusual in early 1981. For example,
the Congressional Budget Office projected a 1980-86 nominal
GNP growth rate in excess of 11 percent. See CBO (1981).
8By comparison, the CBO projected a 2.8 percent rate of real
GNP growth and an 8.5 percent rate of inflation.
9For a statistical investigation of bias in government economic
forecasts, see Belongia (1988).
'“Throughout this article references to the “ Reagan plan" are to
the spending program that excluded what they called “ unallo­
cated savings.” These were cuts in spending for which detail
was to be provided later.
"The results of an alternative analysis using a small model of
budget determination appears in appendix A.

21

Figure 2
Federal Surplus/Deficit
Billions of dollars

Fiscal Years

Billions of dollars

50 -------- ------------------------------ -------------------

50

-100

-1 0 0

-1 5 0

-1 5 0

-2 0 0

-200

-2 5 0
1955

60

65

70

75

85

-2 5 0
1990

NOTE: Reagan plan does not include “ unallocated savings.’

Outlays
O ne m ajo r objective o f R eagan’s eco n o m ic p ro ­
gram w as to
reduce the rate at w hich governm ent spending
increases. . . . Thus, the badly needed effort to
"cut” the budget really refers to reductions in the
am ount of increase in spending requested from
one year to the n ext.12

T h e 1981 program for red u cin g th e grow th of o u t­
lays w as su b je ct to som e confu sion , how ever, b e ­
cau se a target ceilin g w as set w h ich in clu d ed su b ­
stantial “u n allo cated savings” th at w ere to be
sp ecified later. In th e follow ing d iscu ssio n , th ese
u n a llo cated savings are ignored.

^America's New Beginning .. . (1981), p. 10.
13This irregular movement reflects, among other factors, the
business cycle as it affects both GNP and total outlays.



Figure 3 show s th e R eagan plan for real federal
outlays along w ith a ctu al real outlays. T o tal o u t­
lays in real term s clearly did n o t slow as m u ch as
plan n ed . From 1976 to 1980, th e average grow th
rate o f real fed eral outlays w as 3.5 p e rce n t. T h e
actu al rate o f in crea se from 1980 to 1988 w as 2.9
p ercen t, only slightly slow er th an from 1976 to
1980 an d w ell in ex cess o f th e 1.1 p e rce n t rate that
th e ad m in istratio n h ad p ro jecte d in 1981.
As figure 4 show s, a n o th e r w ay to su m m arize
bu dget tren d s is to exam in e th e ratio o f outlays to
GNP. From 1955 to 1980, th e ratio o f total outlays
to GNP rose, albeit irregularly.13 A lthough th e
Reagan plan in ten d ed to reverse this tren d sharply
after 1981, th is did n o t'o c c u r.14

14The discrepancy in the 1980 ratio between the Reagan plan
and the realized outcome reflects the upward revisions of GNP
that have occurred since 1981.

JANUARY/FEBRUARY 1989

22

Figure 3
Total Outlays (constant 1982 dollars)
Ratio Scale
Billions of dollars

Fiscal Years

900

Ratio Scale
Billions of dollars

900

300
1990

300
1955
NOTE: Reagan plan does not include “ unallocated savings.”

Receipts
A nother key part o f th e 1981 econom ic; program
w as a set o f tax prop osals th at w as in ten d ed "to
im prove th e after-tax, after-inflation rew ards to
work, saving, an d in v estm en t.”15 Am ong th e se p ro ­
posals w ere red u ctio n s in m arginal tax rates for
individuals o f 10 p e rce n t a y ea r for th ree years
starting Ju ly 1, 1981. F or co rp oratio n s, th e ch ie f
featu re o f th e p ro p o sed tax ch an g es w as an a c c e l­
erated recovery rate for th e co st o f m ach in ery an d
eq u ip m en t and certain stru ctu res to b e p h ased in
over five y ears. In general, th e effect o f th e p ro ­
p o sed tax ch an g es w as to slow th e grow th o f fed ­
eral receip ts by red u cin g th e role o f individual
in co m e taxes and co rp orate in co m e taxes in th e
revenue stru ctu re.

15America's New Beginning . . . (1981), p. 9.

http://fraser.stlouisfed.org/
Federal ReserveFEDERAL
Bank of St.
Louis
RESERVE
BANK OF ST. LOUIS

Figure 5 show s th e Reagan p lan for to tal re ­
ceip ts along w ith actu al receip ts, b o th converted
to co n sta n t 1982 dollars. Clearly, th e tren d o f real
total receip ts slow ed after 1981 an d w as m u ch
slow er th an p lan n ed . Real re ceip ts p lu m m eted in
1982 an d 1983 due b o th to th e red u ctio n in tax
rates an d th e 1 9 8 1 -8 2 re cessio n . Sin ce th en , re ­
ceip ts have grow n faster th an in th e 1981 Reagan
forecast; b eca u se they fell so m u ch in 1982 and
1983, how ever, th e ir 1986 level w as still below that
p ro jecte d by th e ad m in istratio n in 1981.
W hen to tal receip ts are ch a rted relative to GNP
(figure 6), th e difference b etw een w h at w as
p la n n ed in 1981 an d w hat actu ally h a p p e n ed is
qu ite p ro n o u n ced . In an alternative way, th is dif­
feren ce show s th e in flu en ce o f th e re cessio n an d
h ow it su p p ressed fed eral re ceip ts relative to GNP.

23

Figure 4

Total Outlays Relative to Gross National Product

.16
1955

60

65

70

75

.16
1990

NOTE: Reagan pian does not include “ unallocated savings.”

Figure 5

Total Receipts (constant 1982 dollars)
Ratio Scale
Billions of dollars
800

300
1955




Fiscal Years

Ratio Scale
Billions of dollars
800

300
1990
I

JANUARY/FEBRUARY 1989

24

Figure 6
Total Receipts Relative to Gross National Product
22 ---------------------------------------------------------------------------------------------------------------------------------------------------------- .22

.21

A

Reagan
plan

.21

.16 I---------------------------------------------------------------------------------------------------------------------------------------- ---------1 .16
1955
60
65
70
75
80
85
1990

THE COMPOSITION OF THE
BUDGET: REALIZATIONS VS. THE
REAGAN PLAN
T h e 1981 Reagan p lan called for b o th a slow ing
in th e grow th of governm ent outlays and a ch ange
in th e co m p ositio n of spending and receip ts. T h e
chang e in the co m p ositio n of outlays w as in ­
ten d ed to:
shift Federal budget priorities so that Federal
resources are spent for purposes that are truly the
responsibility of the national government . . . our
budget plans reflect the increased im portance
attached to national defense, m aintain the Federal
Government’s support for the truly needy, and
fulfill our responsibilities for interest paym ents on
the national debt. The spending reductions will
restrain Federal involvement in areas that are
properly left to State and local governm ents or to
the private secto r.16

'elbid., p. 11.

http://fraser.stlouisfed.org/
RESERVE
BANK OF ST. LOUIS
Federal ReserveFEDERAL
Bank of St.
Louis

T h e p ro jecte d co m p ositio n of total receip ts
reflected th e two m a jo r tax ch an g es: tax relief for
individuals an d g reater tax incentives for invest­
m ent by b u sin esses.

Outlays
Table 1 show s th e m ajo r co m p o n e n ts o f outlays
relative to total outlays.'7 T h e first co lu m n show s
that the Reagan ad m inistration p lan n ed to in ­
crease n ation al d efen se outlays from 22.9 p e rce n t
of total outlays in 1980 to 35.7 p e rce n t in 1986.
Although d efen se outlays did rise, th e in crea se fell
short of th e p lan ned level; by 1986, d efense outlays
w ere 27.6 p e rce n t of total outlays. Looking at it in a
different way, th e p lan n ed grow th o f real d efense
outlays w ere p ro jecte d to grow at an 8.6 p e rce n t
annual rate from 1980 to 1986; th e ir actu al rate of
in crease w as 6.2 p ercen t. T h e actu al defense
build-up, w hile slow er than plan ned , did m ark a
reversal of th e previous trend.

,7See appendix B for additional detail on these components.

25

Table 1
Composition of Federal Outlays (percent of total)
National defense
Year

1955
1960
1965
1970
1975
1980
1981
1982
1983
1984
1985
1986
1987
1988

Actual

Reagan plan

62.4%
52.2
42.8
41.8
26.0
22.7
23.2
24.8
26.0
26.7
26.7
27.6
28.1
27.3

22.9%
23.9
26.5
29.3
31.0
34.0
35.7

Payments
to individuals
Actual

20.9%
26.2
28.0
33.1
46.2
47.0
47.7
47.8
48.9
46.9
45.0
45.4
46.7
46.9

Reagan plan

45.7%
46.8
47.4
47.3
47.6
46.9
46.4

All other grants
Actual

2.3%
4.9
6.1
7.9
10.0
10.1
8.5
6.7
6.3
6.2
6.1
6.0
5.2
5.1

T h e n o n d efen se p o rtio n o f th e bu dget w as re ­
d uced , but, again, n o t to th e ex ten t th at w as
plan n ed . T h e p lan called for n o n d efen se outlays
to fall to 64.3 p e rce n t o f th e total by 1986; th e a c ­
tual p ro p o rtio n w as 72.4 p e rce n t. Table 1 provides
fu rth er detail on n o n d efen se outlays. W hile the
plan for pay m ents to individuals, relative to total
outlays, seem s clo se to th e mark, th e grow th rate
co m p ariso n show s a different story. Individual
p aym en t outlays ro se faster th an p lan n ed in real
term s; th e p lan n ed in crease w as a 1.6 p e rce n t
average an n u al rate from 1980 to 1986 co m p ared
w ith th e actu al 2.8 p e rce n t rate o f in crease.
For the category o f “all o th er g ra n ts” (the third
co lu m n o f table 1), th e p lan n ed d eclin e w as real­
ized in th e first tw o y ears, bu t not afterw ard. Al­
th ou gh grants in real term s fell rath er d ram atically
at a 4.8 p e rce n t rate from 1980 to 1986, this w as
still less th an th e 10.7 p e rce n t rate o f d eclin e
p lan n ed by th e 1981 ad m inistration.
T h e fou rth co lu m n o f table 1 show s th e m ost
d ram atic d ep artu re from th e 1981 plan. Net in ter­
est outlays w ere forecast to d eclin e sharply; in ­
stead, how ever, they rose sharply. B ecau se this
co m p o n en t o f outlays cu ts a cro ss all factors th at
affect th e bu dget and reflects th e general in te ra c­
tion o f th e budget w ith th e econom y, this forecast
error serves as a sum m ary m easu re o f th e a c c u ­
racy o f b o th th e bu dget p lan an d th e eco n o m ic

Reagan plan

9.6%
8.1
6.6
5.9
5.4
4.9
4.7

All other

Net interest
Actual

7.2%
7.5
7.3
7.4
7.0
8.9
10.1
11.4
11.1
13.0
13.7
13.7
13.8
14.3

Reagan plan

8.8%
9.4
9.6
8.9
8.2
7.3
6.5

Actual

7.2%
9.2
15.8
9.9
10.8
11.4
10.4
9.2
7.7
7.1
8.5
7.3
6.3
6.5

Reagan pi

13.0%
11.8
9.9
8.6
7.8
6.9
6.8

fo reca st.18 B eca u se outlays grew faster th an
p la n n ed w hile re ceip ts ro se m o re slowly, n et in ­
terest outlays w ere tw ice as large as p lan n ed in
1981. E rrors in re ceip ts (overestim ated) an d o u t­
lays (und erestim ated ), co m b in ed w ith an u n d e re s­
tim ate of in terest rates, p ro d u ced th ese large
errors.
T h e final “all o th e r” category o f outlays show s a
d eclin e very clo se to, b u t generally som ew h at less
th a n p lan n ed .

Receipts
Table 2 show s th e co m p o n en ts o f receip ts rela­
tive to th e total. T h e first colum n , individual in ­
co m e taxes, reflects th e am bitiou s n atu re o f the
1981 tax proposal. T h e ad m in istratio n p ro p o sed a
30 p e rce n t re d u ctio n in m arginal tax rates for in d i­
viduals over a th ree-y ear period b egin n in g Ju ly 1,
1981. M arginal rates w ere to b e red u ced from an
existing range o f 14 p e rce n t to 70 p e rce n t to a
range o f 10 p e rce n t to 50 p e rce n t by Ja n u a ry 1,
1984. T his p ro p o sal w as ex p ected to red u ce in d i­
vidual in co m e taxes from n ea r 47 p e rce n t o f total
receip ts in 1980 to 43.9 p e rce n t in 1983; th e p e r­
cen tag e w as th e n forecast to rise to 46.7 in 1986
b eca u se o f its ex p e cted stim u lu s to activity via
in cen tives to w ork an d invest.
T h e gen eral m ovem en t o f individual in co m e
taxes relative to th e total w en t acco rd in g to plan;

18See appendix A.



JANUARY/FEBRUARY 1989

26

Table 2
Composition of Federal Receipts (percent of total)
Individual
income taxes
Year

Actual

1955
1960
1965
1970
1975
1980
1981
1982
1983
1984
1985
1986
1987
1988

43.8%
44.0
41.8
46.9
43.9
47.2
47.7
48.2
48.1
44.8
45.6
45.4
46.0
44.1

Reagan plan

46.9%
46.2
44.3
43.9
44.1
45.1
46.7

Corporation
income taxes
Actual
27.3%
23.2
21.8
17.0
14.5
12.5
10.2
8.0
6.2
8.5
8.4
8.2
9.8
10.4

Reagan plan

12.4%
10.8
9.6
9.7
9.3
8.4
7.7

Social
insurance taxes
Actual
12.1%
15.9
19.0
23.0
30.3
30.5
30.5
32.6
34.8
35.9
36.1
36.9
35.5
36.8

th e tim ing, how ever, w as su bstan tially d ifferent for
several reaso n s. O ne o f th ese w as th e tim ing o f the
a ctu al legislation. W h at’s m ore, an u n a n ticip a ted
re ce ssio n o ccu rred , an d th e a n ticip ate d bo o m in
ec o n o m ic activity th at w as ex p e cted to follow on
th e h ee ls o f th e tax program failed to develop.
T h e se co n d co lu m n o f table 2 su m m arizes c o r­
p o rate in co m e taxes. Again, th e Reagan p lan w as
bro ad ly realized . C orp orate taxes w ere red u ced
an d th e ir role in th e tax system w as red u ced , at
least th rou gh 1986. T h e p lan n ed an d th e actu al
p ercen tag es w ere qu ite clo se in 1986, although the
a ctu al p ath o f arrival from 1981 to 1986 w as so m e­
w h at different th a n p lan n ed . C orp orate in co m e
taxes w ere severely affected by th e 19 8 1 -8 2 r e c e s ­
sion, d ropping as a p ercen tag e o f total receip ts in
1982-83. D espite th e erro n eo u s e co n o m ic forecast,
how ever, th e gen eral co n to u rs o f th e Reagan c o r­
p orate tax p lans w ere realized. T h is p attern has
b een reversed sin ce 1986, how ever; th e Tax Reform
A ct o f 1986 tig h ten ed provisions for accelera ted
d ep reciatio n o f p lan t an d eq u ip m en t and repealed
th e investm ent tax cred it. T h e se resu lts have c a n ­
celled , to som e extent, th e effects o f th e 1981 tax
act.
T h e evolving role o f so cial in su ran ce co n trib u ­
tio n s in th e tax sy stem is sh o w n in th e th ird
co lu m n o f table 2. T h e actu al ratio follow ed the
plan very clo sely th rou gh 1982, bu t m oved w ell
above th e forecast after that. T h is divergence
reflected m ainly th e 1983 so cial secu rity am e n d ­

http://fraser.stlouisfed.org/
RESERVE
BANK OF ST. LOUIS
Federal ReserveFEDERAL
Bank of St.
Louis

Reagan plan

30.9%
31.1
33.0
33.8
34.2
34.8
34.8

Excise taxes
Actual
13.9%
12.6
12.5
8.1
5.9
4.7
6.8
5.9
5.9
5.6
4.9
4.3
3.8
3.9

Reagan plan

4.7%
7.3
8.6
8.0
7.6
7.0
6.4

Other
Actual
2.9%
4.2
4.9
4.9
5.4
5.1
4.8
5.4
5.1
5.2
5.1
5.2
4.9
4.8

Reagan plan

5.1%
4.7
4.5
4.6
4.9
4.6
4.3

m en ts th at a ccelera ted co lle ctio n s to k eep th e
so cial secu rity program afloat.
T h e fou rth co lu m n o f table 2 show s th e p ro p o r­
tio n o f ex cise taxes to to tal re ceip ts. T h e 1981
Reagan ad m in istratio n fo reca st a sh arp in crea se in
1981 an d 1982, follow ed by a stead y d eclin e. T h is
gen eral p attern o ccu rred , ex cep t th at th e peak
w as in 1981 an d w as at a m u ch low er level th an
forecast. T h e d iscrep a n cy b etw ee n w h at w as
p lan n ed an d w hat actu ally o ccu rred w as m ainly
th e resu lt o f m u ch sm aller th a n ex p e cted gains
from th e w indfall profits tax; oil p rice forecasts
w ere erro n eo u s.
Finally, th e “all o th e r” category, w h ich is u n im ­
p o rtan t relative to th e total, w as u n d erestim ated .
T h e m a jo r taxes in th is category are esta te an d gift
taxes, cu sto m s d u ties an d F ed eral Reserve d e­
p osits. T h e d ollar am o u n t o f all o th e r receip ts w as
forecast accu rately ; b e c a u se th e to tal w as o veresti­
m ated (figures S an d 6), how ever, “all o th e r” re ­
ceip ts as a p ro p o rtio n o f th e to ta l w as
u n d erestim ated .

SUMMARY EVALUATION OF 1981
REAGAN BUDGET PLAN
T able 3 su m m arizes th e 1981 Reagan budget
plan an d co m p a res its individual co m p o n en ts
w ith tren d s p rior to 1981 an d w hat actually o c ­
cu rred after 1981. R ates o f ch an g e for bu d g et totals

27

Table 3
Federal Budget Trends and the Reagan Plan (constant fiscal year 1982 prices,
annual rates of change)
1976-80
Trend

1981 Reagan plan
1980-86

1980-88
Trend

Consistent
with Reagan
plan?1

Outlays
Total
Defense
Nondefense
Payments for individuals
All other grants
Net interest
All other

3.5%
1.7
4.1
2.7
1.6
9.6
8.7

1.1%*
8.6
-1 .8
1.6
-1 0 .7
-5 .5
-9 .2

2.9%
5.6
2.0
2.6
-5 .5
9.1
-3 .9

No
Yes
No
No
Yes
No
Yes

Receipts
Total
Individual income
Corporate income
Social insurance
Excise
All other

5.8
7.6
3.0
5.8
0.8
2.3

2.6
1.7
0.3
5.0
0.2
1.9

Yes
Yes
No
Yes
No
No

3.0
2.9
-4 .8
5.1
8.5
0.2

11f 1980-1988 trend is closer to Reagan plan than to 1976-80 trend, yes; otherwise, no.
including “unallocated savings,” the rate of change was 0.3 percent.

an d th e ir m ajo r co m p o n e n ts are calcu lated from
th e co n stan t d ollar m easu res. A bro ad ju d g m en t is
re a ch ed on w h eth er actu al p erfo rm an ce w as c o n ­
sisten t w ith th e Reagan p lan d ep en d in g on
w h eth er th e actu al 1 9 8 0 -8 8 tren d w as clo se r to the
Reagan p lan th an th e p rio r 1 9 7 6 -8 0 trend .
T h e 19 8 0 -8 8 total outlay p erfo rm an ce w as in ­
co n siste n t w ith th e 1981 p lan. A lthough th e a n ­
nual grow th rate o f total real outlays slow ed from
a 3.5 p e rce n t rate to a 2.9 p e rce n t rate, th is w as
still su bstantially above th e Reagan estim ate o f 1.1
p ercen t. T otal real receip ts, on th e o th er hand,
grew at a rate co n siste n t w ith th e 1981 p lan; th ey
actually slow ed m ore th a n p lan n ed b ec a u se of the
19 8 1-82 recessio n .
An exam in ation o f th e grow th o f th e co m p o ­
n en ts o f real outlays show s that som e m oved in a
d irectio n co n siste n t w ith 1981 plan. Real d efen se
outlays did n o t rise as m u ch as p lan n ed ; however,
th eir grow th accelerated su bstan tially from the
1 9 7 6 -8 0 period. Although real n o n d efen se outlays
grew m u ch m ore slowly, th e Reagan p lan called
for a d ecline. T h e co m p o n e n ts o f real n o n d efen se
outlays show ed m ixed results. G rowth in real pay ­
m en ts for individuals an d n et in terest slow ed only
slightly. T h e o th e r two categories, however,



sh ow ed a sharp reversal from th e p rio r fou r years,
although n o t as m u ch as w as plan n ed .
Th ough real total re ceip ts m oved co n sisten tly
w ith th e Reagan plan, th e co m p o n e n ts o f th e total
sh ow ed m ixed resu lts. Real individual in co m e
taxes ro se m ore slow ly th a n plan n ed , chiefly b e ­
ca u se e co n o m ic grow th w as overestim ated, bu t
th e ir grow th w as dow n sharply from th e 1 9 7 6 -8 0
tren d . Real co rp o ra te in co m e taxes slow ed, but
n o t to th e ex ten t ou tlin ed in th e 1981 plan. Real
social in su ra n ce co n trib u tio n s grew at rates very
clo se to w hat w as fo reca st in 1981. B o th th e excise
an d “all o th e r" co m p o n e n ts o f to tal receip ts w ere
estim ated in co rrectly, b u t this h ad little c o n s e ­
q u en ce sin ce th ey are su ch sm all p ro p o rtio n s of
total receip ts.

CONCLUSION
In 1981, th e new ly inau gu rated Reagan ad m in is­
tration form ulated a bu d g et p lan d esign ed to slow
th e grow th o f govern m en t an d b o o st in cen tives
(via taxes) to save, invest an d work. In clu d ed in the
p ro jectio n s w as a m ovem en t tow ard a b alan ced
fed eral bu dget by 1986. T h e a ctu a l rise in th e fed ­
eral d eficit sin ce 1981, cu lm in atin g w ith a $221

JANUARY/FEBRUARY 1989

28

billio n d eficit in 1986, suggests th at th e Reagan
budget program failed. E xam ination o f th e factors
co n tribu tin g to th e deficit as w ell as th e co m p o si­
tion o f b o th outlays an d receip ts, how ever, in d i­
ca tes broad ly w hy th is resu lt o ccu rred an d p o in ts
out th at th ere w ere a n u m b er o f su c c e sse s as well
as failures w h en individual co m p o n e n ts o f the
fed eral bu d g et are co n sid ered .
T o tal re ceip ts in 1986 w ere overestim ated by
about $170 billion, m ainly b ec a u se th e 1981-82
re cessio n w as n o t an ticip ated . T h e m ajo r tax p ro ­
posals w ere ad opted, althou g h n o t in th e ir exact
form n o r acco rd in g to th e p ro p o sed tim etable.
B ecau se o f d ifferences in tim ing and su b seq u en t
legislation, th e actu al co m p o sitio n o f total receip ts
varied som ew hat from th e 1981 p ro jectio n s. T h e
d irectio n o f m ovem ent, how ever, w as generally as
p ro jecte d for individual in co m e taxes, co rp orate
in co m e taxes an d so cial in su ran ce co n tribu tio n s.
T h e largest erro r in th e p ro je c te d co m p o sitio n of
total receip ts w as for ex cise taxes, chiefly b eca u se
the forecast o f oil p rices w as in erro r w ith th e
resu lt th at th e w indfall profits tax did n o t p ro d u ce
revenues as exp ected .
T o tal outlays w ere u nd erestim ated by about $30
billion in 1986 (or m ore th an $70 billion if th e 1981
estim ate in clu d es “u n allo cated savings”). F u rth er
exam in ation o f outlays revealed a $73 billio n error
in th e p ro jectio n o f n et in terest. T h is erro r w as
largely offset, how ever, b eca u se th e actu al d efen se
bu ild -up fell abou t $69 billion belo w p ro jectio n s
by 1986. T h e n o n in terest p ortion o f n o n d efen se
sp en d ing w as u n d erestim ated by $15 billion, o r 5
p ercen t.
In general, if o n e looks at bu dg et outlays, th e
Reagan program en jo yed som e su cce ss: th e d e ­
clin e in th e relative role o f d efen se outlays w as
reversed; p aym en ts for individuals relative to total
outlays co n tin u ed roughly as plan n ed ; all o th er
g rants an d th e resid ual category of “all o th e r o u t­

19See appendix A and Carlson (1984).

http://fraser.stlouisfed.org/
RESERVE
BANK OF ST. LOUIS
Federal ReserveFEDERAL
Bank of St.
Louis

la y s” co n tin u ed to d eclin e from th e ir p eaks in th e
late 1970s o r 1980. T h e m a jo r ex cep tio n to th e 1981
plan w as th e rise in n et in terest outlays p ro d u ced
by failures to fo reca st th e 1 9 8 1 -8 2 re ces sio n
(w hich slow ed th e grow th o f receip ts), th e level of
in terest rates, an d th e cum ulative effect on outlays
o f co m p ou n d in g in terest on a grow ing n atio n al
d eb t.19

REFERENCES
America's New Beginning: A Program for Economic Recovery
(The White House, Office of the Press Secretary, February
18,1981).
Belongia, Michael T. “Are Economic Forecasts by Government
Agencies Biased? Accurate?” this Review (November/Decem­
ber 1988), pp. 15-23.
Boskin, Michael J.
1987).

Reagan and the Economy (ICS Press,

Carlson, Keith M. “Trends in Federal Revenues: 1955-86,”
this Review (May 1981), pp. 31-39.
-------------- “Trends in Federal Spending: 1955-86,” this
Review (November 1981), pp. 15-24.
-------------- “The Critical Role of Economic Assumptions in the
Evaluation of Federal Budget Programs,” this Review (Octo­
ber 1983), pp. 5-14.
-------------- “ Money Growth and the Size of the Federal Debt,”
this Review (November 1984), pp. 5-16.
Congressional Budget Office. An Analysis of President
Reagan’s Budget Revisions for Fiscal Year 1982, Staff Work­
ing Paper (March 1981).
Mills, Gregory B., and John L. Palmer, eds. Federal Budget
Policy in the 1980s (The Urban Institute Press, 1984).
Modigliani, Franco. “ Reagan’s Economic Policies: A Critique,"
Oxford Economic Papers (September 1988), pp. 397-426.
Niskanen, William A.
1988).

Reaganomics (Oxford University Press,

Office of Management and Budget.
Revisions (March 1981a).

Fiscal Year 1982 Budget

________ Federal Government Finances (March 1981 b).
________ Historical Tables, Budget of the United States
Government Fiscal Year 1990 (U.S. Government Printing
Office, 1989).

29

Appendix A
The Im pact of E co n o m ic Assumptions on the 1981
Reagan Budget Plan
An alternative m eth o d o f evaluating th e 1981
budget plan is to sim u late th e effect on the budget
o f eco n o m ic co n d itio n s different from th o se a s­
su m ed in p lan n in g th e budget. An u p d ated ver­
sion o f a bu dget m od el previously p re sen ted in
th is R ev iew w as u sed to do th is.1 T h e m o d el c o n ­
sists o f th ree parts: an estim ate o f th e im p act o f
inflation and real grow th on b o th (1) p rim aiy re­
ceip ts and (2) p rim aiy outlays, an d (3) an e sti­
m ated eq u atio n for n et in terest outlays. T h e latter
reflects th e in d irect effects o f inflation an d real

grow th on receip ts an d outlays as well as th e ef­
fect o f in terest rate ch an g es.
First, th e 1981 R eagan bu d g et p lan w as sep a ­
rated into prim ary receip ts, p rim ary outlays an d
n et in terest (see table A l). T h e effects o f actu al
inflation, real grow th an d in terest rates th e n w ere
calcu lated , yield in g sim u lated values o f prim ary
receip ts an d outlays an d in terest co st for the
1 9 8 1 -8 6 period . T h e se sim u lation resu lts are
sh ow n in table A2.

Table A1
The 1981 Reagan Budget Plan (billions of dollars)

Year

Primary
receipts

Primary
outlays

1980
1981
1982
1983
1984
1985
1986

$508.3
587.7
637.0
694.0
752.2
831.2
921.3

$541.4
614.7
643.8
703.5
756.1
830.2
898.4

Primary
surplus/deficit
$ -3 3 .1
-2 7 .0
- 6 .8
- 9 .5
- 3 .9
1.0
22.9

Federal
Reserve
deposits
$11.8
12.6
13.3
15.1
18.5
18.7
18.9

Total
receipts
$520.1
600.3
650.3
709.1
770.7
849.9
940.2

Net
interest
$52.5
64.1
68.2
68.9
67.8
64.9
62.8

Total
outlays

Surplus/
deficit

$593.9
678.8
712.0
772.4
823.9
895.1
961.2

$ -7 3 .8
-7 8 .5
-6 1 .7
-6 3 .3
-5 3 .2
-4 5 .2
-2 1 .0

Total
outlays

Surplus/
deficit

$593.9
684.3
736.7
808.3
858.9
934.7
999.2

$ -7 3 .8
-8 0 .3
-1 0 5 .5
-1 8 2 .7
-2 0 4 .6
-2 2 8 .6
-2 5 5 .3

Table A2
The 1981 Reagan Budget Plan; Simulated1(billions of dollars)
Year

Primary
receipts

Primary
outlays

1980
1981
1982
1983
1984
1985
1986

$508.3
591.2
616.0
611.1
638.6
689.0
725.5

$541.4
613.4
650.2
711.3
747.6
810.4
864.7

Primary
surplus/deficit
$ -3 3 .1
-2 2 .2
-3 4 .2
-1 0 0 .2
-1 0 9 .0
-1 2 1 .4
-1 3 9 .2

Federal
Reserve
deposits2

Total
receipts

Net
interest

$11.8
12.8
15.2
14.5
15.7
17.1
18.4

$520.1
604.0
631.2
625.6
654.3
706.1
743.9

$52.5
70.9
86.5
97.0
111.3
124.3
134.5

'Using actual values of economic variables (real growth, inflation and interest rates)
2Actual

'See Carlson (1984), appendix B, for a summary of the model.
The only difference from that model is that the interest cost
equation was reestimated with data through 1986. For an ex­
tended discussion of the role of economic assumptions in
budget estimates, see Carlson (1983).



JANUARY/FFRRIIARY 1QAQ

30

Table A3
Actual Budget Results (billions of dollars)
Year

Primary
receipts

Primary
outlays

1980
1981
1982
1983
1984
1985
1986
1987

$505.3
586.5
602.6
586.1
650.8
717.0
750.7
837.3

$538.4
609.5
660.7
718.5
740.7
816.9
854.3
865.2

Primary
surplus/deficit

$ -33.1
-2 3 .0
-58.1
-132.4
-8 9 .9
-9 9 .9
-103.6
-2 7 .9

T h e resu lts in d icate th at had th e 1981 bu dget
plan b ee n fully im p lem en ted , it w ould have
y ield ed a deficit o f about $255 billion in 1986.
T h e se resu lts are b ased on th e actu al co u rse o f
inflation, real grow th an d in terest rates from 1981
to 1986. Sin ce th e actu al 1986 d eficit w as $221 b il­
lion (see table A3), ap p aren tly th e 1981 program
w as n o t im p lem en ted as p lan n ed . Specifically,
from 1981 to 1986, n eith e r total re ceip ts n o r pri-


http://fraser.stlouisfed.org/
Federal Reserve Bank
of St. RESERVE
Louis
FEDERAL
BANK OF ST. LOUIS

Federal
Reserve
deposits

Total
receipts

Net
interest

Total
outlays

Surplus/
deficit

$11.8
12.8
15.2
14.5
15.7
17.1
18.4
16.8

$517.1
599.3
617.8
600.6
666.5
734.1
769.1
854.1

$ 52.5
68.7
85.0
89.8
111.1
129.4
136.0
138.6

$ 590.9
678.2
745.7
808.3
851.8
946.3
990.3
1003.8

$ -7 3 .8
-7 8 .9
-127.9
-207.8
-185.3
-212.3
-221.2
-149.7

m a iy outlays in crea sed to th e ex ten t originally
p lan n ed ; to tal outlays in c re a se d m o re th an
p la n n ed b e c a u se o f large erro rs in estim atin g n et
in terest. T h u s, th e a ctu a l beh av io r o f key eco n o m ic
variables "o v erexp lain s” th e deficit. T h at is, if p ri­
m ary receip ts an d outlays h ad perfo rm ed a cc o rd ­
ing to th e 1981 plan, th e 1986 d eficit w ould have
b een m u ch larger th an it tu rn ed ou t to be.

31

Appendix B
Composition of Fed eral Outlays
F ed eral outlays ca n b e classified in term s o f two
analytical stru ctu res: bu dget fu n ctio n and m ajo r
program category. T h e fu n ctio n al classificatio n
p re sen ts outlays acco rd in g to th e p u rp o se th at a
fed eral program is in ten d e d to serve. T h e se fu n c­
tio n s in clu d e, for exam ple, n atio n al d efense, in ter­
nation al affairs, energy program s, tran sp o rtatio n ,
health, in co m e security, etc. Tw o ad d itional c a te ­
gories — n et in terest and u nd istrib u ted offsetting
receip ts — do n o t ad d ress sp ecific fu n ctio n s, but
are in clu d ed to cover th e en tire budget.
T h e classificatio n o f federal outlays by m ajo r
program category fo cu ses on th e m eth o d o f carry­
ing ou t an activity. T h e m a jo r program categ ories
are n atio n al d efense, ben efit p ay m en ts for individ­
uals, gran ts to state an d lo cal govern m ents (oth er
th a n for b en efit paym ents), n et in tere st an d all
o th e r outlays. N ational d efen se an d n e t in terest
co rresp o n d to th e fu n ctio n al categ o ries o f th e
sam e nam e, b u t th e rem ain ing m a jo r program
categ o ries do n o t co rresp o n d to a sim p le su m ­
m ing o f fu n ctio n al categ o ries. N o n eth eless, a p ­
p ro xim atio n s ca n be m ad e. T h e acco m p an y in g
table groups 1988 outlays by fu n ctio n to provide
a d d ed inform ation abo u t th e m a jo r program ca te ­
gories d iscu ssed in th e text.




Table B1
1988 Federal Outlays (billions of
dollars)
Category

Amount

Total

$1,064.0

National defense

290.4

Benefit payments for individuals1
Health and medicare
Social security
Income security
Veterans benefits and services

501.4
123.4
219.3
129.3
29.4

Other grants to state and local governments'
Natural resources and environment
Agriculture
Transportation
Community and regional development
Education, training, employment and social
services
General government
Net interest

50.0
3.7
2.1
18.1
4.3
19.9
1.9
151.7

All other'
International affairs
General science, space and technology
Energy
Natural resources and environment
Agriculture
Commerce and housing credit
Transportation
Community and regional development
Education, training, employment and social
services
Administration of justice
General government
Undistributed offsetting receipts

70.4
10.5
10.8
2.3
10.9
15.1
18.8
9.2
1.0
12.0
9.2
7.6
-37.0

'Amounts shown are the sums for the functions listed under
them and differ slightly from the categories discussed in the
text.

JANUARY/FEBRUARY 1989

32

Cletus C. Coughlin
and G eoffrey E. Wood
Cletus C. Coughlin is a senior economist at the Federal Reserve
Bank of St. Louis and Geoffrey E. Wood is a professor of
economics at City University, London. Thomas A. Pollmann pro­
vided research assistance.

An Introduction to Non-Tariff
B arriers to Trade

STRICTIO N S on in tern atio n al trade,
p rim arily in th e fo rm o f n o n -ta riff b a rrie rs ,
have m ultiplied rapidly in th e 1 9 8 0 s .1 T h e
Ja p a n ese , fo r exam ple, b eg an re strictin g
au tom obile ex p o rts to th e U nited States in 1 981.
O ne y e a r later, th e U .S. govern m en t, as p a rt o f
its ongoing in terv e n tio n in th e su gar m ark et,
im posed qu otas on su gar im ports.
T h e in creasin g u se o f p ro tectio n ist trad e
policies raises natio n al as w ell as in tern atio n a l
issues. As m any o b serv ers have n oted , in te rn a ­
tio n al tra d e re strictio n s g en erally have costly
n atio n al co n se q u e n ce s.2 T h e n e t b e n e fits r e ­
ceived b y p ro tected d o m estic p ro d u cers (that is,
b e n e fits red u ced b y lobbying costs) ten d to b e
outw eighed by th e losses asso ciated w ith e x ­
cessive p ro d u ctio n and re stricte d con su m p tion
o f th e p ro tected goods. P ro te ctio n ist tra d e
policies also cau se fo reig n ad ju stm en ts in p ro ­

1See Page (1987) for a general discussion indicating that
the proliferation of trade restrictions in recent years has
taken the form of non-tariff, as opposed to tariff, barriers.
A recent Congressional Budget Office study (1987) notes
that the average tariff rate for most developed countries is
less than 5 percent. There is no evidence of rising tariff
rales or coverage. For example, U.S. tariff revenue as a
percentage of total imports has changed very little be­
tween 1975 (3.9%) and 1986 (3.6%). See the Statistical

http://fraser.stlouisfed.org/
FEDERAL
RESERVE BANK OF ST. LOUIS
Federal Reserve Bank
of St. Louis

d u ction and co n su m p tio n th a t risk s retaliatio n
b y th e a ffe cte d co u n try .
As a type o f p ro tectio n ist policy, n o n -ta riff
b a r r ie r s p ro d u ce th e g en era l co n se q u e n ce s id en ­
tified above; h o w ev er, th e r e a re n u m ero u s
reaso n s, besid es th e ir p ro lifera tio n , to fo cu s a t­
te n tio n solely on n o n -ta riff b a r r ie r s .3 N on-tariff
b a r rie rs en co m p ass a w id e ra n g e o f specific
m ea su res, m any o f w h o se e ffe c ts a re n o t easily
m easu red . F o r exam ple, th e e ffe c ts o f a g o v e rn ­
m en t p ro c u re m e n t p ro cess th a t is b iased to w ard
d om estic p ro d u cers a re d ifficu lt to qu antify. In
addition, m an y n o n -ta riff b a rrie r s d iscrim in ate
am ong a co u n try 's trad in g p a rtn ers.
T h is d iscrim in atio n violates th e m ost-favoredn atio n prin cip le, a co rn e rsto n e o f th e G en eral
A g reem en t on T a riffs an d T ra d e (GATT), th e
m u ltin ation al a g ree m en t g overn in g in tern a tio n a l
trad e. Not only does th e m ost-favored -n ation

Abstract of the United States (various editions) for the
figures for other years.
2For example, see Coughlin et al. (1988).
3See chapter 1 in Laird and Yeats (forthcoming) for a
discussion of the policy issues raised by non-tariff barriers.

33

prin cip le re q u ire th a t a co u n try tr e a t its tradin g
p a rtn e rs identically, b u t it also re q u ires th a t
tra d e b a r r ie r red u ctio n s negotiated on a
b ila te ra l b asis b e exten d ed to all G A TT m em ­
b e rs. By su bstitu ting b ilateral, d iscrim in ato ry
ag reem en ts fo r m u ltilateral ap p roach es to trad e
negotiations and d ispute settlem en t, co u n tries
raise doubts ab o u t th e long-run viability o f
GATT.
T h is p ap er provides an in tro d u ctio n to nonta r iff b a rrie rs. W e beg in by id entifying n u m e r­
ous n o n -ta riff b a rrie rs and d ocu m en t th e ir p ro ­
liferatio n . W e th e n u se supply and dem and
analysis to id en tify th e g en eral e ffe c ts o f tw o
freq u en tly u sed n o n -ta riff b a rrie rs: qu otas and
v olu ntary ex p o rt re stra in ts. Next, w e co n sid er
w h y n o n -ta riff b a rrie rs a re used instead o f
ta riffs. A b r ie f h isto ry o f G A TT’s attem p ts to
co u n te ra ct th e exp an sio n o f n o n -ta riff b a rrie rs
com p letes th e body o f th e paper.

NON-TARIFF BA R R IER S:
TY PE S AND USE
A ta r iff is a tax im posed on fo reig n goods as
th e y e n te r a co u n try ; n o n -ta riff b a rrie rs, on th e
o th e r hand, a re non -tax m easu res im posed by
g ov ern m en ts to fav o r d om estic o ver foreig n
suppliers. N on-tariff b a rrie rs en com p ass a w ide
ran g e o f m easu res. Som e have relatively u nim ­
p o rta n t trad e e ffects. F or exam ple, packaging
and labeling re q u irem e n ts can im pede trad e,
b u t usually only m arginally. O th er n o n -tariff
m e a su res su ch as quotas, v o lu n tary ex p o rt
re stra in ts, tra d e re stra in ts u n d e r th e M u ltifiber
A rrang em en t, non-au tom atic im p o rt a u th o riza­
tio n s and v ariable im p o rt levies h ave m u ch
m o re significan t e ffe c ts .4 T h e se "h a rd -co re" nonta r iff m easu res a re designed to re d u ce im ports
and, th e reb y , b e n e fit d om estic p ro d u cers. T h e
discussion belo w fo cu ses on th e se h ard -co re
b a rrie rs.

Quotas

co u n tries g en erally . T w o exam ples illu strate
th e se d iffe re n t ch a ra cte ristic s. T h e U nited States
im poses a g en era l qu ota on dried m ilk im ports;
licen ses a re g ra n ted to c e rta in U.S. trad in g co m ­
panies, w h o a re allow ed to im p o rt a m axim um
qu an tity o f dried m ilk b ased on th e ir previous
im ports. In a d iffe re n t situation U.S. su gar im ­
p o rts a re lim ited b y a quota th a t sp ecifies the
sh a re s o f individual co u n tries; th e righ t to sell
su gar to th e U nited States is given d irectly to
th e g ov ern m en ts o f th e se co u n tries.

Voluntary E xp o rt R estraints an d
the M ultifiber A rra n gem en t
V olu n tary ex p o rt re stra in ts, w h ich a re n early
id en tical to quotas, a re a g reem en ts b etw ee n an
ex p o rtin g and an im portin g co u n try lim iting th e
m axim um am ou n t o f ex p o rts in eith e r value o r
qu an tity te rm s to b e sold w ith in a given period.
C h aracterizin g th e se re stra in ts as "v o lu n ta ry ” is
so m ew h at m isleading b eca u se th e y a re f r e ­
qu en tly designed to p re v en t o fficial p ro tectiv e
m easu res by th e im p ortin g co u n try . In th e
1 9 8 0 s, fo r exam ple, ex p o rts b y th e Ja p a n ese
au tom obile in d u stry to th e U nited States and
th e U nited Kingdom h ave b e e n lim ited “volun­
ta rily ” to p rev en t th e g ov ern m en ts o f th ese
co u n tries fro m d irectly lim iting im p orts of
Ja p a n ese autos.
An exam ple o f a v olu n tary ex p o rt re s tra in t on
a m u ch b ro a d e r scale is th e M u ltifib er A rra n g e­
m ent. O riginally signed in 1 9 7 4 as a tem p o ra ry
ex cep tio n to G A TT and re n ew ed th re e tim es
since, th e M u ltifib er A rra n g em en t allow s fo r
special ru les to g o v ern tra d e in tex tiles and ap­
parel. U n d er th is ag reem en t, qu otas a re set on
m ost im p orts o f tex tiles and ap p arel by
developed co u n tries fro m developing co u n tries,
w hile im p orts o f tex tiles and ap p arel fro m o th e r
developed co u n tries ex cep t Ja p a n a re n o t su b ­
je c t to any re strictio n s. M u ltilateral vo lu n tary
ex p o rt re stra in t a g ree m en ts a re freq u en tly
called "o rd erly m ark etin g a g re e m e n ts.”

A quota is sim ply a m axim u m lim itation,
sp ecified in e ith e r value o r p hysical units, on
im p orts o f a p ro d u ct fo r a given period. It is e n ­
fo rce d th ro u g h licen ses issued to e ith e r im ­
p o rte rs o r e x p o rters and m ay b e applied to im ­
p o rts fro m sp ecific co u n tries o r fro m all foreig n

Non-Automatic Im p ort
Authorizations

4This subset of non-tariff barriers is taken from Laird and
Yeats (forthcoming). This subset excludes a number of
non-tariff barriers that can also have sizeable effects.
Among these are government procurement policies, delays

at customs, health and sanitary regulations, technical stan­
dards, minimum import price regulations, tariff quotas and
monitoring measures. See appendix 4 in Laird and Yeats
for a glossary of terms associated with non-tariff barriers.




N on-autom atic im p o rt au th o rizatio n s a re n o n ­
ta r iff b a rrie rs in w h ich th e approval to im port
is n o t g ran ted free ly o r au tom atically. T h e re

JANUARY/FEBRUARY 1989

34

a re tw o g en eral categ o ries o f non-au tom atic
licensing.
D iscretio n ary licensing, o ften called lib eral
licensing, o ccu rs w h en an im p o rte r’s g o v e rn ­
m en t m u st approve a sp ecific im port; how ever,
p re cise cond itions to e n su re approval a re n o t
specified. F req u en tly , this fo rm o f licensin g is
u sed to ad m in ister qu antitative lim its. U nd er th e
c u rre n t re stra in ts on U.S. im p orts o f steel, a
d om estic u se r can re q u e st au th o rizatio n to e x ­
ceed th e m axim um im p o rt lim itation if th e
sp ecific p ro d u ct is unavailable d om estically a t a
re a so n a b le cost. Exactly how availability and
co st co n sid eratio n s a ffe c t th e p ro bability o f an
approval a re le ft to th e d iscretio n o f th e
au th o rities.
T h e seco n d categ o ry o f n on-au tom atic im p o rt
licensin g re q u ire s th e im p o rte r to m eet sp ecific
cond itions, su ch as m inim um e x p o rt p e r fo r ­
m an ce, th e u se o f th e im p orted good fo r a
sp ecific p u rp o se o r re q u ired p u rch ases o f
d om estic p rod u cts. In an ex p o rt-im p o rt linkage
sch em e, a firm ’s value o f im p orted co m p on en ts
is lim ited to a m axim um p e rce n ta g e o f th e value
o f its ex p o rts. T h is m easu re is in ten d ed to im ­
p ro ve a co u n try ’s tra d e b a la n ce and p ro tect
d om estic p ro d u cers o f co m p o n e n ts.5 Exportim p o rt linkage re q u irem e n ts a re n u m ero u s. F o r
exam ple, in Yugoslavia d uring th e early 1980s,
au th o rized im p o rters o f au tom obiles w e re r e ­
qu ired to ex p o rt goods totalin g at least 30 p e r­
c e n t o f th e value o f ea ch im p orted au to m o b ile.6

V ariable Im p o rt Levies
V ariable im p o rt levies a re special ch arg e s set
to equalize th e im p o rt p rice o f a p ro d u ct w ith a
5See Herander and Thomas (1986) for a theoretical
demonstration that an export-import linkage scheme might
not improve a country’s trade balance.
6For details on the policies of Yugoslavia as well as
numerous other countries, see “ Survey of Automotive
Trade Restrictions Maintained by Selected Nations”
(1982).
A/ariable import levies, which are actually variable tariffs,
are considered non-tariff barriers in this study for two
reasons. First, the international trade literature generally
characterizes variable import levies as non-tariff barriers.
See Nogues et al. (1986) for another list of non-tariff bar­
riers that includes variable import levies. Second, Laird
and Yeats (forthcoming) provide the most up-to-date data
on non-tariff barriers and we have no way to remove
variable import levies from their data.
8The numerical example is from Coughlin and Carraro
(1988).
9One weakness of the coverage ratio as a measure of pro­
tectionism is that more-restrictive non-tariff barriers tend to

http://fraser.stlouisfed.org/
Federal Reserve FEDERAL
Bank of St.RESERVE
Louis
BANK OF ST. LOUIS

d om estic ta rg e t p rice. T h e levies a re v ariab le so
th a t as th e w o rld p rice o f a p ro d u ct falls (rises),
th e levy rises (falls).7 T h e re su lt is th a t p rice
ch an g es in th e w o rld m a rk et will n o t a ffe c t
d irectly th e d om estic p rice. T h e se m e a su res are
an in teg ral asp e ct o f th e E u ro p ean C om m unity’s
C om m on A g ricu ltu ral Policy. F o r exam ple, in
M a rch 1 9 8 7 , th e E u rop ean C om m unity’s p rice
fo r w h ea t w as $ 8 .5 3 p e r bu sh el, w hile th e
w o rld p rice w as $ 1 .9 5 p e r b u sh el. P ro sp ectiv e
im p o rters w ere fa ced w ith a levy o f $ 6 .5 8 p e r
b u sh el.8

T h e Use a n d E xpansion o f NonT a riff B a rriers
In a c u rre n t study, L aird and Y eats (fo rth co m ­
ing) m e a su re th e sh a re o f a c o u n try ’s im p orts
su b je ct to h a rd -co re n o n -ta riff b a rrie rs . B ecau se
co u n tries freq u en tly im pose n o n -ta riff b a rrie rs
on th e im p orts o f a sp ecific good fro m a
sp ecific co u n try , b u t n o t on im p o rts o f th e
sam e good fro m a n o th e r co u n try , th e y disag­
g reg ated ea ch co u n try ’s im p o rts b y b o th p ro ­
d u ct and co u n try o f origin to p erm it calcu latio n
o f th e to tal value o f a co u n try ’s im p orts su b je ct
to n o n -ta riff b a rrie rs. E ach c o u n try ’s "co v era g e
ra tio ” is sim ply th e value o f im p o rts s u b je ct to
n o n -ta riff b a r rie rs divided b y th e to tal value o f
im p o rts.9
Table 1 shows th e trade coverage ratio fo r 10
European Com munity and six oth er industrial
countries fo r 1981 and 1986. In com puting this
ratio, th e 1981 and 1986 non-tariff m easures are
applied to a constant 1981 trade base. Thus, the
figures identify changes in the use, b u t n ot the
intensity, o f specific non-tariff m easures, while
holding constant the effects o f trad e changes.
receive a lower weight in the construction of the coverage
ratio than less-restrictive ones. For example, a non-tariff
barrier that eliminated all imports of a good from a country
would have a smaller impact on the coverage ratio than a
less-restrictive measure. Assume that one country’s im­
ports are valued at $100, $15 of which comes from coun­
try A, and there are no non-tariff barriers. In this case, the
coverage ratio is zero. Suppose that a non-tariff barrier is
now imposed on imports of goods from country A. In the
first case, assume that imports from country A decline
from $15 to $10; alternatively, suppose that imports
decline from $15 to zero. The non-tariff barrier in the sec­
ond case is more restrictive; however, the change in the
coverage ratio does not reflect this fact. The coverage
ratio becomes 10.5 percent ($10/$95) in the first case and
zero percent ($0/$85) in the second. Thus, the “ intensity”
of the protection provided by non-tariff barriers is not
measured accurately by this coverage ratio. An alternative
measure focusing on the share of trade “ affected” by non­
tariff barriers, which also highlights the proliferation of
non-tariff barriers, can be found in Laird and Yeats (1989).

35

Table 1
Non-tariff Trade Coverage Ratios for
OECD Countries
Trade Coverage Ratio2
Importer1

1981

1986

Belgium-Luxembourg
Denmark
Germany, Fed. Rep.
France
Greece
Great Britain
Ireland
Italy
Netherlands
EC (10)3

12.6%
6.7
11.8
15.7
16.2
11.2
8.2
17.2
19.9
13.4

14.3%
7.9
15.4
18.6
20.1
12.8
9.7
18.2
21.4
15.8

Switzerland
Finland
Japan
Norway
New Zealand
United States

19.5
7.9
24.4
15.2
46.4
11.4

19.6
8.0
24.3
14.2
32.4
17.3

0.1
0.1
-0.1
-1 .0
-1 4 .0
5.9

All above

15.1

17.7

2.6

Difference

1.7%
1.2
3.6
2.9
3.9
1.6
1.5
1.0
1.5
2.4

NOTE: Non-tariff measures include variable import
levies, quotas, non-automatic import authorizations in­
cluding restrictive import licensing requirements, quan­
titative “ voluntary” export restraints and trade restraints
under the Multifiber Arrangement.
’ The following Organization for Economic Cooperation
and Development (OECD) countries — Australia,
Canada and Sweden — were excluded from the com­
putations because of problems in compiling their non­
tariff measures.
2The share of total imports (by value) subject to hard­
core non-tariff measures. In computing this index, 1981
and 1986 non-tariff measures are applied to a constant
1981 trade base. Petroleum products have been exclud­
ed from the calculations.
3European Community intra-trade is excluded.
SOURCE: Laird and Yeats (forthcoming).

1 9 8 6 . T h ird , th e U nited States had th e larg est
p ercen tag e-p o in t in crea se, as its co v erag e ratio
in crea s ed fro m 1 1 .4 p e rce n t in 1 9 8 1 to 17 .3
p e rc e n t in 1 9 8 6 . T h e 5 .9 p ercen tag e-p o in t in ­
c re a se w as m o re th a n double th e in c re a se fo r
all co u n tries.
Laird and Y eats provide evid en ce th a t exp o rts
fro m developing co u n tries to in d u strial co u n ­
trie s a re a ffe cte d to a la rg e r e x te n t th a n tra d e
am ong in d u strial co u n tries. F o r exam ple, th e
1981 tra d e co v erage ratio w as 1 8.8 p e rce n t fo r
developing co u n try ex p o rts to in d u strial co u n ­
trie s and 14.3 p e rce n t fo r in tra-in d u strial co u n ­
tr y tra d e. A sim ilar p a tte rn p revailed in 1 9 8 6
w ith a co v erag e ra tio o f 2 0 .6 p e rc e n t fo r
developing co u n try ex p o rts to in d u strial co u n ­
trie s and 17.5 p e rce n t fo r in tra-in d u strial co u n ­
try tra d e .10
T a b le 2 co n tain s co v erage ratio data on a p ro ­
d u ct basis. As a re su lt o f th e M u ltifib er A r­
ra n g em en t, tra d e in tex tiles and clo th in g is su b ­
je c t to n o n -ta riff b a rrie rs. F or exam ple, slightly
m o re th an on e-th ird o f E u rop ean C om m unity
and U.S. im p orts o f tex tiles a re a ffe cted , w hile
ap p roxim ately tw o-th ird s o f E u rop ean Com ­
m u n ity and th re e -q u a rte rs o f U .S. im p o rts o f
cloth in g a re a ffe cted . Sin ce th e se goods are
am ong th e m ost im p o rtan t m a n u fa ctu red e x ­
p o rts fro m developing co u n tries, co v erage ratio s
fo r im p orts fro m developing co u n tries relativ e
to in d u strial co u n tries ten d to b e h ig h er.
T a b le 2 also id en tifies som e o th e r m a n u fa c­
tu re d goods a ffe cte d su bstan tially b y n o n -ta riff
b a rrie rs , especially iro n and steel and tra n sp o rt
eq uipm ent. M ore th a n th re e -q u a rte rs o f U.S. im ­
p o rts o f iro n and steel and m o re th a n 4 0 p e r­
c e n t o f tra n sp o rt eq u ip m en t a re a ffe cted . T h e
co rresp o n d in g fig u res fo r th e E u ro p ea n Com ­
m u n ity a re 4 6 .2 p e rce n t and 2 3 .6 p e rce n t.

A n u m b er o f fa cts em erg e. F irst, th e coverage
ra tio v aries su bstantially a cro ss co u n tries. In
1 9 8 1 , th e co v erag e ra tio ran g ed fro m 6 .7 p e r­
c e n t in D en m ark to 4 6 .4 p e rc e n t in New Z ea­
land and, in 1 986, fro m 7.9 p e rce n t in D en m ark
to 3 2 .4 p e rc e n t in New Zealand. Second, fo r
m o st co u n tries, th e co v erag e ratio h as in creased .
T h is cau sed th e co v erage ratio using th e w orld
tra d e fig u res o f all 16 co u n tries to in crea se
fro m 15.1 p e rce n t in 1981 to 1 7 .7 p e rce n t in

W h ile tra d e in m a n u fa ctu red goods is a ffected
su bstan tially by n o n -ta riff b a rrie rs, tra d e in
a g ricu ltu ra l goods is a ffe cte d to an ev en g re a te r
ex ten t. T h e co v erag e ra tio s fo r a g ricu ltu ral
goods sh ow n in tab le 3 a re su bstan tially above
th o se fo r m a n u fa ctu red goods sh ow n in ta b le 2.
T h e ag ricu ltu ral co v erag e ratio s freq u en tly e x ­
ce ed 70 p e rce n t; see, fo r exam ple, th e U.S.
ra tio s fo r su gar an d h o n ey (91.9 p ercen t), d airy
p ro d u cts (87.8 p ercen t) and oil seeds and n uts
(74 p ercen t). E ven h ig h er ag ricu ltu ra l co v erage

10While this differential may reflect discrimination directed at
developing countries, another interpretation is that the differential is product-based. Chow and Kellman (1988), for

example, show that the relatively higher tariff rates faced
by developing countries can be explained by product
characteristics.




JANUARY/FEBRUARY 1989

36

Table 2
Coverage Ratios of Selected Non-tariff Measures on Selected Manufactured
Goods: 1986
SITC
61
62
63
64
65
66
67
68
69
71
72
73
81
82
83
84
85
86

Description

EC (10)'

Leather products
Rubber products
Wood and cork
Paper and articles
Textiles
Cement, clay and glass
Iron and steel
Non-ferrous metals
Metal manufactures, n.e.s.
Non-electric machinery
Electric machinery
Transport equipment
Plumbing & lighting fixtures
Furniture
Travel goods
Clothing
Footwear
Instruments

7.7%
9.1
1.0
5.9
34.7
2.9
46.2
0.8
2.1
3.1
11.1
23.6
0.0
0.3
0.9
65.7
11.3
3.8

Switzerland
30.8 %
0.0
1.9
0.0
0.0
0.0
1.0
1.9
5.6
4.7
0.0
84.7
0.0
0.0
53.0
18.6
74.6
0.0

Finland
0.0%
0.0
0.0
0.0
1.6
0.0
0.0
3.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
12.1
0.0
0.0

Japan
47.0 %
13.6
0.0
0.0
55.5
24.1
0.0
0.4
1.0
4.4
0.3
17.3
0.0
0.0
0.0
11.3
6.9
14.1

Norway

New Zealand

United
States

0.0%
0.7
0.0
0.0
6.1
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.0
86.5
0.3
0.0

59.9%
53.9
53.0
48.6
27.4
54.5
64.1
8.7
35.3
35.9
64.0
22.1
68.2
0.0
100.0
52.2
82.9
5.3

0.0%
0.0
0.0
0.0
34.5
0.1
76.3
0.0
11.0
0.0
1.4
41.1
0.0
1.1
18.9
76.4
0.1
0.0

NOTE: See table 1 for the list of hard-core non-tariff measures. The coverage ratio is, for each given product and country,
the imports subject to a hard-core non-tariff measure divided by total imports.
'European Community intra-trade is excluded.
SOURCE: Laird and Yeats (forthcoming).

Table 3
Coverage Ratios of Non-tariff Measures on Selected Agricultural Goods: 1986
SITC
00

01
02
03
04
05
06
07
08
09
11
12
21
22
23
24
25
26
29

Description

Live animals
Meat
Dairy products
Fish and seafood
Cereals and preparations
Fruits and vegetables
Sugar and honey
Coffee and cocoa
Animal feeds
Food preparations
Beverages
Tobacco
Hides and skins
Oil seeds and nuts
Rubber
Wood and cork
Pulp and paper
Silk, wool, cotton, etc.
Crude animal & vegetable matter

EC (10)'
60.2%
77.8
99.7
4.6
96.9
36.0
85.8
17.5
11.9
10.2
24.9
0.0
0.0
24.8
0.0
0.6
0.0
9.0

19.0

Switzerland
100.0%
97.8
45.5
58.3
87.8
44.8
0.0
0.0
30.9
13.4
76.4
0.0
99.1
56.0
0.0
39.6
0.0
24.8
78.0

Finland
95.3%
89.3
100,0
9.7
83.4
51.6
89.1
0.0
5.3
0.0
88.0
0.0
0.0
100.0
0.0
0.0
0.0
0.0
5.3

Japan
1.2%
65.7
73.2
100.0
32.5
18.3
84.6
0.0
13.7
17.3
70.7
84.3
18.1
4.3
0.0
0.0
0.0
1.2
51.8

Norway
98.0%
99.7
82.1
80.4
100.0
100.0
100.0
100.0
92.7
100.0
100.0
0.0
0.0
100.0
0.0
0.0
0.0
4.6
69.1

New Zealand
0.0%
14.4
12.7
3.6
5.1
39.2
0.9
0.9
16.9
73.7
5.6
5.1
0.0
0.0
0.0
2.4
0.0
16.4
11.2

United
States
0.0%
0.0
87.8
0.0
0.0
0.9
91.9
2.3
0.3
0.4
0.0
0.0
3.2
74.0
0.0
0.0
0.0
2.1
11.0

NOTE: See table 1 for the list of hard-core non-tariff measures. The coverage ratio is, for each given product and country, the imports subject to a hard-core non-tariff measure divided by total imports.
'European Community intra-trade is excluded.
SOURCE: Laird and Yeats (forthcoming).

http://fraser.stlouisfed.org/
Federal Reserve Bank
of St. RESERVE
Louis
FEDERAL
BANK OF ST. LOUIS

37

Table 4
The Use of Selected Non-tariff Measures________________________________
Change in the Share of Imports
Facing NTMs, 1981-86=

Share of Imports Facing NTMs, 19811
Importer

Belgium-Luxembourg
Denmark
Germany, Fed. Rep.
France
Greece
Great Britain
Ireland
Italy
Netherlands
EC (10)3
Switzerland
Finland
Japan
Norway
New Zealand
United States
All above

QUOT

0.3%
0.3
0.5
5.8
8.2
2.2
0.1
7.5
0.4
2.6

VER

MFA

NAIA

VIL

QUOT

VER

MFA

0.0%
-0.1
-0 .6
0.0
0.0
0.0
0.0
-0.1
-0 .2
-0 .2

NAIA

VIL

5.1%
2.6
3.0
1.2
4.8
2.0
4.6
0.8
2.0
2.3

1.2%
2.3
4.9
1.8
1.2
2.9
1.3
1.8
3.0
3.0

5.7%
1.1
3.0
7.1
3.9
5.1
2.2
7.0
14.0
5.6

5.2%
1.4
2.0
2.2
3.8
4.4
2.2
6.6
6.3
3.7

1.1%
0.1
0.4
1.6
0.4
-0 .9
0.1
0.6
2.5
0.5

2.2%
1.2
2.0
1.8
4.4
2.3
1.5
1.2
3.6
2.1

2.5
0.9
14.2
5.2
25.3
0.5

0.0
0.0
0.0
0.0
0.0
6.9

0.4
0.2
0.0
0.0
0.0
3.2

2.8
6.7
7.7
2.2
25.6
0.0

0.5
1.8
1.8
0.0
0.0
0.0

0.0
0.0
0.1
-0 .5
1.6
1.5

0.0
0.0
0.0
0.0
0.0
4.4

0.0
0.1
0.0
0.0
0.0
0.0

0.0
0.0
0.0
1.1
-8 .8
0.0

0.0
0.0
0.0
0.0
0.0
1.4

4.0

3.1

2.3

4.2

2.0

0.7

2.2

-0.1

-0.1

0.4

0.0%
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0

0.0%
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0

'Petroleum products have been excluded from the calculations. The abbreviations for the non-tariff measures are as follows:
QUOT—quotas; VER—voluntary export restraints; MFA—restrictions under the Multifiber Arrangement; NAIA—non-automatic
import authorizations; and VIL—variable import levies.
2The change is the 1986 share less the 1981 share.
3European Community intra-trade is excluded.
SOURCE: Laird and Yeats (forthcoming).

ratio s a re fou nd fo r th e E u rop ean C om m unity
and Jap an .
A n o th er d im ension o f th e u se o f n o n -tariff
b a rrie rs c o n c e rn s d iffe re n c e s in th e u se o f
sp ecific b a rrie rs a cro ss co u n tries. T a b le 4 show s
th e sh are o f im p orts (by cou n try ) th a t fa ced dif­
fe re n t n o n -ta riff m e asu res in 1981 and h ow this
sh a re chang ed by 1986. A n u m b er o f fa cts
em erg e. In 1981, n on-au tom atic im p o rt au th o ri­
zations and qu otas a ffe cte d th e larg est sh are o f
im p orts w h en all 16 co u n tries a re consid ered;
b y 1 9 86, th is w as no lo n g er th e case. V olu n tary
ex p o rt re stra in ts, w h o se u se in th e U nited
States, G reece, th e N etherland s and G reat B ri­
tain ro se su bstantially, a ffe cted th e larg est sh are
o f im ports (5.3 p ercen t) b y 1 986. M eanw hile,
th e sh are o f im p orts a ffe cte d b y qu otas ro se
fro m 4 p e rce n t in 1981 to 4 .7 p e rce n t b y 1986.
C om parisons o f th e sp ecific m easu res a cro ss
co u n tries in d icate th a t volu ntary ex p o rt r e ­
stra in ts w e re u sed m o re exten sively b y th e
U nited States th a n b y o th e r co u n tries. By 1 986,



11.3 p e rc e n t o f U.S. im p orts w e re a ffe cte d by
v olu n tary ex p o rt re stra in ts; G reece, w ith 9.2
p e rce n t, had th e n ext-h igh est sh a re o f its im ­
p o rts a ffe cte d b y th e se re stra in ts.

SU PPLY AND DEMAND ANALYSIS
USING QUOTAS AND VOLUNTARY
E X P O R T RESTRAIN TS
A lthough th e qu an titativ e e ffe c ts o f n o n -ta riff
b a r rie rs are n o t alw ays easily id en tified and
m easu red , a th e o re tic a l id en tificatio n o f th e ir
m ajo r e ffe c ts ca n b e d erived using supply and
d em and analysis. W e b eg in by exam in in g th e e f­
fe cts o f a quota, th e n d iscuss h o w a volu n tary
ex p o rt re stra in t ca n b e an alyzed sim ilarly.
In fig u re 1, DD re p re se n ts th e U.S. im p o rt
dem and cu rv e fo r som e good p ro d u ced b y U.S.
and fo reig n p ro d u cers. T h e fo reig n supply
cu rv e (that is, th e supply cu rv e fo r im p o rts into
th e U nited States) fo r th e good is SS. W ith fre e
trad e, th e U nited States w ill im p o rt Q f u nits o f
th e good and pay a p rice p e r u nit o f Pp.

JANUARY/FEBRUARY 1989

38

Figure 1
The Price and Quantity Effects of a Quota
and a Voluntary Export Restraint
Price

Q

q

Qf

Q uantity of
Imports

Now, suppose th a t an im p o rt quota o f Q q is
im posed b y th e U nited States. T h is re strictio n
cau ses th e im p o rt supply cu rv e to b eco m e v e r­
tica l at th e re stricte d qu antity. T hu s, th e im port
supply cu rv e is th e k inked cu rv e SCS'. T h e
re strictio n red u ces th e qu antity o f im p orts fro m
Op to Q q , th e d om estic p rice to rise fro m Pp to
P q and th e fo reig n p rice to d ecline fro m Pp to
P g .11 T h e h ig h er d om estic p rice re d u ce s total
U.S. con su m p tio n o f th e good, b u t in c r e a s e s
U.S. pro d u ctio n ; th u s, U.S. p ro d u cers o f th e
good b e n e fit at th e exp en se o f U.S. co n su m ers
in g en eral. T h e d iffe re n c e b e tw e e n w hat
d om estic and fo reig n co n su m ers pay, P b P q , is a
p rem iu m p e r u nit o f im p orts th a t can b e ap ­
p ro p riated b y ex p o rters, im p o rters o r g o v e rn ­
m ent. T h e m ethod u sed to allocate im port
licen ses d eterm in es th e d istribu tion o f th ese
prem iu m s am ong th e potential claim ants.

ta ry ex p o rt re s tra in t re d u ce s th e qu an tity o f im ­
p o rts, w h ich , in tu rn , cau ses th e d om estic p rice
to rise and th e fo reig n p rice to fall as sh ow n in
fig u re 1. Again, th e h ig h er d om estic p rice
b en e fits U.S. p ro d u cers o f th is good a t th e e x ­
p en se o f U.S. co n su m ers. Finally, th e d iffe re n c e
b e tw e e n w h a t d om estic and fo reig n co n su m ers
pay, P b PQ; is a p rem iu m p e r u n it o f im p orts
th a t can b e ca p tu red by e x p o rters, im p o rte rs or
gov ern m en t.
W h ile th e supply and d em and analysis isolates
th e m a jo r e ffe c ts o f tw o freq u en tly used nonta r iff b a rrie rs , it con veys v irtu ally no in fo rm a ­
tio n ab o u t e ith e r th e m agnitu d e o f th e co sts and
b e n e fits o f n o n -ta riff b a rrie rs o r th e ir dynam ic
c o n se q u e n ce s.12 V ariou s case studies, h o w ev er,
h ave provided estim ates o f th e se co sts and
b en e fits. A rev iew o f th is lite ra tu re ca n b e
fou n d in L aird and Y eats. T w o case studies are
provided in th e shaded in se rts on pages
and
as exam ples o f su ch analyses. T h e firs t ex a m ­
ple exam in es th e im pact o f th e U .S. qu ota on
su gar im ports; th e secon d exam in es th e e ffe c t
o f th e U .S.-Jap an ese a g ree m en t to lim it Ja p a n ese
au tom obile ex p o rts to th e U nited States.
As a p ro tectio n ist policy, n o n -ta riff b a rrie rs
a re a m eth o d fo r red istrib u tin g w ealth fro m
co n su m ers in g en era l to selected firm s and
w o rk e rs. T h is red istrib u tio n is a b ette d b y c o n ­
su m er ig n o ra n ce and th e costs o f m obilizing an
e ffe c tiv e fo r c e to co u n te ra ct p ro tectio n ist
dem ands. As Coughlin e t al. (1988) have d em on ­
stra ted re cen tly , th e b en e fits receiv ed b y se­
lected groups o f firm s and w o rk e rs a re fa r o u t­
w eighed by th e co sts b o rn e b y th e re st o f th e
population.

W HY USE NON-TARIFF B A R R IER S
INSTEAD OF T A R IFFS ?

A vo lu n tary ex p o rt re stra in t has th e sam e
g en eral e ffe c ts as an equivalent quota. A v olu n­

Sin ce n o n -ta riff b a rrie rs have b e e n used in ­
creasin g ly in re c e n t y ea rs, an obvious qu estion
is w h y n o n -ta riff b a rrie rs ra th e r th a n ta r iff bar-

11Figure 1 can also be used to illustrate a variable import
levy. While a quota limits the quantity of imports, a
variable import levy is used to fix the price. Assuming a
target (domestic) price of P q , when world prices fall below
this price, the levy will be altered automatically to maintain
the price of P q . Thus, no matter how far world prices
decline, the quantity of imports will not rise above Q q .
Consequently, a variable import levy and a quota have the
same effect, even though they are implemented differently.

text, two of which are mentioned below. Since many
markets for internationally traded goods are imperfectly
competitive, a standard topic in introductory international
trade texts is to identify the effect of an import quota in
the presence of monopoly. See Krugman and Obstfeld
(1988) for an elementary discussion. Since voluntary
export restraints discriminate among trading partners, the
effects of this differential treatment have been explored.
See Jones (1984) for such an analysis.

1theoretical research on the impact of non-tariff barriers
has explored various issues that we do not mention in the

http://fraser.stlouisfed.org/
FEDERAL
RESERVE BANK OF ST. LOUIS
Federal Reserve Bank
of St. Louis

39

A V oluntary E xp ort R estraint in P ractice: The U.S.
Jap an ese Autom obile A greem ent
O ne w ell-know n exam p le o f a vo lu n tary e x ­
p o rt re stra in t is th e Ja p a n e se re stra in t on
au tom obile ex p o rts to th e U nited States. In
early 1 981, th e Ja p a n e se im posed re stra in ts
to p reem p t m o re re strictiv e m e a su res ad ­
v o cated by m any, esp ecially la b o r groups,
w ithin th e U nited S tate s.1 T h e se p ro tectio n ist
p re ssu re s in creased d uring th e late 1 9 7 0 s and
early 1980s as au tom obile sales b y U .S. p ro ­
d u cers d eclined and fo reig n p ro d u cers ca p ­
tu re d la rg e r sh a re s o f th e U.S. m ark et.
Collyns and D unaw ay (1987), as w ell as
m any o th ers, estim ated th e e ffe c ts o f th e
re stra in ts. T h e se au th o rs exam in ed th e
re stra in ts fro m 1981 to 1 9 8 4 . T h e ex am in a­
tion revealed th a t th e ex p ected re su lts did
m aterialize.
W ith th e re stra in ts, th e p rices paid b y U.S.
co n su m ers fo r Ja p a n e se au tom obiles rose.
T h is red u ced th e com p etitive p re ssu re s on
U.S. p ro d u cers and n o n -Jap an ese ex p o rte rs to
th e U nited States w ith th e e ffe c t o f in c re a s­
ing p rices fo r th e se au tom obiles, b u t n o t as
m u ch as th e rise in Ja p a n e se p rices. T h e
h ig h er au tom obile p rices re d u ce d U.S. p u r­
ch ases, b u t th e e ffe c ts on U .S. and nonJa p a n e se p ro d u cers w e re m itigated by th e
relatively la rg e r rise in th e p rices o f Ja p a n ese
au tom obiles and th e resu lting sh ift aw ay
fro m Ja p a n ese au tom obiles.
T h e re stra in ts also ind u ced quality ch an g es
as Ja p a n ese p ro d u cers shifted th e ir m ix o f
ex p o rts to w ard la rg e r and m o re lu xurious
m odels th a t g en era ted m o re p ro fits p e r unit.
In addition, m o re "o p tio n al” eq u ip m en t w as
installed in ea ch unit. C onsequently, th e
av erage tra n sa ctio n p rice o f Jap an ese
au tom obiles in crea sed b ec a u se o f th e p u re
p rice e ffe c t as w ell as th e quality e ffe c ts
asso ciated w ith th e re stra in ts.

sold in th e U nited States an d co m p licate th e
estim ation . F o r all n ew ca rs sold in 1 9 8 4 , Col­
lyns an d D un aw ay (1987) estim ated an
av erag e in cre a se o f $ 1 ,6 4 9 (17 p ercen t),
w h ich co n sisted o f a p u re p rice e ffe c t o f
$ 6 1 7 p e r c a r and a quality e ffe c t o f $ 1 ,0 3 2
p e r ca r. T h e h ig h er p rice led to a red u ctio n
in 1 9 8 4 p u rch a se s o f ap p roxim ately 1.5
m illion.
As suggested above, th e ex p o rt re stra in ts
h ad d iffere n tia l e ffects. F o r exam ple, th e
p rice in cre a se fo r dom estically pro d u ced
au tom obiles o f $ 1 ,1 8 5 (12 p erce n t) w as less
th a n th e in crea se fo r im p o rts fro m Ja p a n o f
$ 1 ,7 0 0 (22.5 p ercen t). T h is relativ e p rice
ch a n g e allow ed th e U.S. p ro d u cers to in ­
cre a s e th e ir m a rk et sh a re b y 6 .7 5 p e rce n ta g e
points, en o u gh to leave d om estically p ro d u c­
ed u n it sales u n ch an g ed d esp ite a d ecline o f
u n it sales in th e U nited States. T h u s, th e U.S.
red u ctio n in 1 9 8 4 p u rch a se s o f 1.5 m illion
w as b o rn e by fo reig n p ro d u cers. T h e se p ro ­
d u ction ch an g es w e re estim ated to g en era te
in crea sed U.S. au tom otive em p lo y m en t in a
ra n g e fro m 4 0 ,0 0 0 to 7 5 ,0 0 0 jobs.
T h e h ig h er au tom obile p rice s re p re se n t one
fa c e t o f th e losses fo r co n su m ers. T h e p u re
p rice e ffe c t cau sed U.S. co n su m ers to su ffe r
a loss o f co n su m ers’ su rplu s o f $ 6 .6 billion in
1 9 8 4 . In addition, U.S. co n su m ers w ere w o rse
o ff to th e e x te n t th a t qu otas lim ited th e ir
ra n g e o f au tom otive ch o ices. P u rch a ses o f in ­
cre a se d quality resu ltin g fro m th e qu ota to ta l­
ed $ 1 0 .7 5 billion in 1 9 8 4 . T h e w e lfa re loss
asso ciated w ith th e se quality ex p en d itu res
w as n o t estim ated , b u t it is cle a r th a t this
loss is possibly g re a te r th a n th e loss asso­
ciated w ith th e p u re p rice e ffect.

In fact, th e fa cto rs u nd erly ing th e p rice
ch a n g e a ffe c t th e p rice s o f all au tom obiles

T h e losses o f U.S. co n su m ers a re prim arily
tr a n s fe rs fro m co n su m ers to d om estic and
fo reig n p ro d u cers. E stim ates o f th e b en e fits
fo r d o m estic and fo reig n p ro d u cers h in ge on

1Feenstra (1985) provides numerous details concerning
legislation designed to restrict imports. In early 1981,
Sens. Danforth and Bentsen introduced a bill to restrict
automobile imports from Japan to 1.6 million units an­
nually during 1981-83, which is very close to the volun-

tary export restraint of 1.68 million. Other proposed
legislation was more restrictive in providing for smaller
import quotas and in specifying the minimum content of
American parts and labor for automobiles sold in the
United States.




IAKII I A D V / C C D D I I A D V 1QOQ

40

th e assu m ption ab o u t th e d istribu tio n o f th e
p u re p rice e ffe c ts. If th e ex p o rt re stra in ts led
to equivalen t p u re p rice e ffe c ts on dom estic
and im p orted cars, th e n U.S. p ro d u cers gain ­
ed $5 billion in 1 9 8 4 and fo reig n p ro d u cers
gained $1 .5 billion. O f th e fo reig n p ro d u c e rs ’
gain, Ja p a n e se p ro d u cers re ceiv ed $1 billion.
O n th e o th e r hand, if th e ex p o rt re stra in ts
led to equivalen t quality e ffe c ts, th e n U.S.

p ro d u cers gain ed $ 1 .2 5 billion in 1 9 8 4 and
fo reig n p ro d u cers gained $ 5 .5 billion. O f th e
fo reig n p ro d u cers' gain, Ja p a n e se p ro d u cers
re ceiv ed $ 5 .2 5 billion. If a cc u ra te , th is fig u re
p rovides an obvious re a so n w h y th e Ja p a n e se
g o v ern m en t co n tin u ed th e re stra in ts b ey o n d
ea rly 1 9 8 5 w h en th e R eagan ad m in istratio n
d ecid ed n o t to re q u e st an ex ten sio n o f th e
a g ree m en t.2

2ln early 1985, the Reagan administration decided that
the domestic automobile industry had adjusted to
foreign competition and announced they would not ask
for an extension. Nevertheless, in early 1985, the
Japanese government extended the restraints through
early 1987 at a level 24 percent above the previous

level and in 1987 extended the restraints for another
year without a further increase in the ceiling. The
unilateral decision to extend the restraints is a clear in­
dication that the Japanese, especially automobile pro­
ducers, were benefiting from the restraints.

rie rs have b eco m e so p o p u lar.13 A rev iew by
D e a rd o rff (1987) co n clu d es th a t th e re cu rre n tly
is n o d efinitive an sw er to th is qu estion; h o w ­
ev er, n u m ero u s reaso n s have b e e n suggested.

T h e Im pact o f GATT: An Institu­
tional Constraint o n the Use o f
T a riffs
G A TT is an in stitu tio n w hose original m ission
w as to r e s tric t th e use o f ta riffs. Given th is co n ­
strain t, p o licy m ak ers w illing to resp o n d to p ro ­
te ctio n ist d em ands w e re fo rce d to u se n o n -ta riff
devices. T h u s, in this case, n o n -tariff b a rrie rs
a re simply a su b stitu te fo r ta riffs. In fact, r e ­
se a rch by Ray (1981) in d icates th a t n o n -ta riff
b a rrie rs have b e e n u sed to re v e rse th e e ffe c ts
o f m u ltilateral ta r iff red u ctio n s negotiated
u n d e r G A TT.14
13Dating from Bhagwati’s seminal discussion in 1965, com­
parisons of the theoretical effects of tariffs and non-tariff
barriers have been a frequent topic in the international
trade literature. Under various circumstances, a tariff and
a specific non-tariff barrier, say, a quota, can cause dif­
ferent final prices and production despite reducing trade
by equal amounts. These circumstances produce what is
termed nonequivalence. Tariffs and quotas are equivalent
when markets are perfectly competitive. In this case, there
is no reason to prefer one to the other.
Bhagwati (1965, 1968) has demonstrated that the
equivalence of tariffs and quotas breaks down in imper­
fectly competitive markets. Numerous situations can be
characterized as imperfectly competitive. To date, how­
ever, the literature has provided no compelling reasons for
preferring non-tariff over tariff barriers. For a recent exam­
ple from this literature, see Krishna (1985).
14A question remains, however, as to why the framers of
GATT chose to focus primarily on tariffs rather than non­
tariff barriers.

http://fraser.stlouisfed.org/
Federal Reserve Bank
of St. Louis
FEDERAL
RESERVE BANK OF ST. LOUIS

Certainty o f D om estic B en efits
D e a rd o rff (1987) su ggests th a t n o n -ta riff b a r ­
rie rs a re p re fe rre d to ta riffs b ec a u se policy­
m a k ers and d em an d ers o f p ro tectio n believe
th a t th e e ffe c ts o f ta riffs a re less ce rta in . T h is
p e rce p tio n could b e due to variou s reason s,
som e re a l and som e illusory. F o r exam ple, it
m ay b e m u ch ea sier to see th a t a qu ota o f 1
m illion lim its au tom obile im p o rts to 1 m illion
th a n to d em o n stra te co n clu sively th a t a ta r iff
of, say, $ 3 0 0 p e r c a r w ould re su lt in im p o rts o f
only 1 m illion au tom obiles.
In p art, d oubts th a t ta riffs w ill h ave th e
d esired e ffe c t is b a sed on th e possibility o f a c­
tion s th a t could b e ta k en to o ffse t th e e ffe c ts o f
h ig h er ta riffs. F o r exam ple, th e im position o f a
ta r iff m ay in d u ce th e ex p o rtin g co u n try to su b ­
sidize th e ex p o rtin g firm s in an a ttem p t to
re d u ce th e ta r iff’s effectiv en ess. T h e e ffe c ts o f
quotas, on th e o th e r hand , a re n o t a ltered by
su ch su b sid ies.15
15Deardorff’s (1987) review provides another perspective on
the role of uncertainty. The optimality of trade policy tools
has been explored extensively using trade models with
uncertainty. These models, which rely on risk aversion
(that is, an individual requires a higher expected return as
compensation for an increase in risk) and uncertainty
originating outside a country, conclude that quotas are
preferred to tariffs. The country is insulated from the
uncertainty stemming from randomness in world prices or
import supply curves by a quota that stabilizes the price
and quantity of imports. One problem with this explana­
tion, however, is that the quota is instituted before the
uncertain state of the world is known, while in the real
world protection is generally provided after a change in the
world market.

41

A Non-Tariff B a rrie r in P ractice:
Im port Quota
Sin ce 1 982, th e U nited States has im posed
qu otas on su gar im p o rts to su p p ort a
d om estic p rice g u aran tee b y th e fed eral
g ov ern m en t th a t ex ceed s w orld m ark et
lev els.1 T h e high p rice has stim ulated U.S.
su gar p ro d u ctio n and shifts in dem and
to w ard o th e r sw ee te n ers, w h ich h as n e c e s­
sitated large red u ctio n s in su gar im port
qu otas in re c e n t y ears.

The U.S. Sugar

Figure 2
The Effects of Trade Restrictions on the U.S.
Sugar Market
Price
(cents per pound)

T a r r and M o rk re (1984) estim ated th e costs
o f th e su gar im p o rt quota fo r fiscal y e a r
1 9 83 (O cto ber 1 982-S ep tem b e r 1983). A ctu al­
ly, th e qu ota is co m bin ed w ith a ta riff, so
ta r iff rev en u es as w ell as qu ota rev en u es
arise. T h e qu ota re v en u es a re cap tu red b y 24
fo reig n co u n tries w ho have th e righ t to sell
su gar in th e U nited States.
F ig u re 2 illu strates som e o f th e e ffe c ts o f
th e U.S. tra d e re strictio n s in 1983. T h e lines
SS and DD a re th e U.S. supply and dem and
cu rv e s fo r sugar. T h e w o rld p rice w as 15
ce n ts p e r pound, and U.S. p u rch ase s w ere
assum ed to have n o e ffe c t on th is p rice. W ith
fre e trad e, U.S. pro d u ctio n , co nsu m p tion and
im p orts w ould h ave b e e n 6 .1 4 billion pounds,
1 9 .1 8 billion pounds and 1 3 .0 4 billion pounds.
T o raise th e in tern al (U.S.) p rice to 2 1 .8 ce n ts
p e r pound, a ta r iff o f 2 .8 ce n ts p e r pound
and a quota o f 5 .9 6 billion pounds w e re used.
T h e value o f th e quota is 4 .0 ce n ts p e r
pound, b eca u se 2 .8 ce n ts p e r pound o f th e
6 .8 ce n ts p e r pou nd d iffere n tial b etw ee n th e
U .S. p rice and th e w orld p rice is due to th e
ta riff.

n u e, w h ich is re p re se n te d b y a re a i. C on se­
qu en tly, th e n et e ffe c t fo r th e U nited States
is a loss o f $ 4 8 3 m illion, w h ich is th e sum o f
a re a s g, j and h. A rea g is th e loss due to in ­
e ffic ie n t p ro d u ctio n and a re a j is th e loss due
to in e fficie n t con su m ption . A rea h, w h ich is
equal to $ 2 3 8 m illion, is th e value o f th e im ­
p o rt licen ses receiv ed b y fo reig n suppliers. In
o th e r w ord s, th e qu o ta en tails a tr a n s fe r
fro m U.S. co n su m ers to foreig n p ro d u cers o f
$ 2 3 8 million.

T h e w elfa re e ffe c ts o f th e trad e re strictio n s
a re ind icated b y th e are as f, g, h, i and j. T h e
p rice-in creasin g e ffe c ts o f th e tra d e r e s tric ­
tions cau se co n su m ers to su ffe r a loss o f co n ­
su m er surplus eq u al to $ 1 ,2 6 6 billion, th e
sum o f are as f, g, h, i and j, 2 P ro d u cers gain,
in th e fo rm o f p ro d u cer surplus, are a f
w h o se value is $ 6 1 6 m illion. T h e U.S. g o v ern ­
m en t also gains $ 1 6 7 m illion in ta r iff reve-

T h e p reced in g analysis, w hile effectiv ely
highlighting th e w in n e rs and lo sers fro m th e
U.S. su gar p ro g ram , is n ot th e e n tire story.
T h e se estim ates p erta in to on e y e a r only.
Sin ce th e U.S. su gar policy is ongoing, th e
losses a re ongoing as w ell. In addition, im p o r­
ta n t dynam ic in terrela tio n sh ip s b etw ee n
policy ch a n g es and p ro d u ctio n and tra d e
ch an g es exist.

1Maskus (1987) concluded that U.S. sugar production
and trade have been directed by government policies
almost continuously for 200 years.

studies. Maskus (1987) surveyed studies of the costs
borne by U.S. consumers and found estimates ranging
from $1 billion to $2.7 billion.

Source: Krugman and Obstfeid (1988).

2Tarr and Morkre’s (1984) estimate of the consumer cost
of the U.S. sugar program is consistent with other



IA Ml IA D V / C C D D I IA D V

1QOQ

42

M askus (1987) has id en tified a n u m b er o f
th e dynam ic co n se q u e n ce s o f th e U.S. su gar
p rog ram , m any stem m ing fro m th e fa c t th a t
su g ar has sev eral clo se su bstitu tes. C orn
sw e e te n e rs; n o n -calo ric sw ee te n ers, h o n ey
and specialty su gars a re all clo se su bstitu tes.
H igher su gar p rices have ind u ced th e p ro d u c­
tio n o f alternativ e sw ee te n ers th a t co m p ete
w ith and, co n seq u en tly, th re a te n U.S. su gar
p ro d u cers.
T h e fa c t th a t su gar is used in d iffe re n t
goods has set in m otion a n u m b er o f ad­
ju stm en ts. Exam ples abou n d o f th e d istortion s
in d u ced b y th e artificially high U.S. su gar
p rice. F o r exam ple, th e large p rice d iffe re n ­
tial b e tw e e n U.S. and fo reig n su gar provides
a co st advantage to fo reig n , esp ecially C ana­
dian, food -p rocessing firm s. T h e su gar policy
can b e view ed as a tax on U.S. re fin e rs and
p ro c e sso rs th a t w as n o t levied on fo reig n
firm s.
T ra d e flow s resp o n d ed to th e se p rice
ch an g es as a rapid exp ansion in im p o rts o f
su gar-containing goods ensu ed . In fact, th e
d iffere n tia l b e tw e e n U.S. and w orld su gar
p rice s b eca m e so large at on e tim e th a t
su gar-containing goods w e re im p orted solely
fo r th e ir su gar co n ten t. F o r exam ple, d uring
1 9 8 5 , w o rld su gar p rices d eclin ed so sharply
th at, in Ju n e 1 9 8 5 , th e U.S. su gar p rice w as
77 6 p e rc e n t o f th e w orld p rice. T h is d if­
fe re n c e in d u ced som e firm s in th e U nited
States to im p o rt C anadian p an cak e mix,
w h ich w as n o t su b je ct to th e quota, and p ro ­
cess it to e x tra c t th e sugar.
T h e ind u ced ch an g es in p ro d u ctio n and
tra d e have fo rce d a n u m b er o f additional
U.S. actio n s to m aintain th e su gar p rices. F o r
fiscal y e a r 1985, th e U .S. su gar im p o rt quota
w as re d u ce d 1 7 p e rce n t. T h is w as follow ed
by red u ctio n s o f 2 7 .6 p e rce n t in 1 9 8 6 and
4 5 .7 p e rc e n t in 1 987. T ra d e re strictio n s on
su gar su b stitu tes also have resu lted . T w o o f
th e se are: 1) an em erg en cy b a n on im p orts o f
ce rta in syrups and blend ed su gars in bu lk in
Ju n e 1 983; and 2) em erg en cy qu otas on a
b ro ad ran g e o f su gar-containing articles in
b o th b u lk and retail fo rm s in Ja n u a ry 1 9 8 5 .




T h e in creasin gly re strictiv e im p o rt b a rrie r s
have p ro d u ced ten sio n s w ith n u m ero u s e x ­
p o rte rs o f sugar, m o st o f w h om a re develop­
ing co u n tries. T o co n fo rm w ith th e G en eral
A g reem en t on T a riffs and T ra d e , th e im p o rt
qu otas m u st b e applied in a nond iscrim in ato ry fash ion . T h e U nited States ap ­
plied th is provision by basin g its qu o ta a llo ca­
tio n on im p o rts d uring th e relatively freem a rk et perio d o f 1 9 7 5 -8 1 . A ttem pts to m ain ­
tain co n sta n t sh a res fo r m ost co u n tries,
h o w ev er, ra n into p ra ctica l p ro b lem s. C oun­
trie s ex p erien cin g rap id g ro w th in su gar e x ­
p o rts to th e U nited States b e tw e e n 1 9 7 5 and
1 9 8 1 w e re su b je cted to su b stan tial cu ts b e t­
w een th e end o f th e fre e -m a rk e t perio d and
th e beg in n in g o f th e qu otas. F or exam ple,
su gar ex p o rts fro m H onduras w e re red u ced
fro m 9 3 ,5 0 0 to n s in 1 9 8 1 to 2 8 ,0 0 0 to n s in
1 983.
T h e e ffe c t o f th is c u t w as m itigated som e­
w h a t in 1 9 8 3 w h en th e U nited States tr a n s ­
fe rre d 52 p e rc e n t o f N icaragua’s qu ota to
H onduras, an a ctio n th a t sim ultaneou sly
pu n ish ed th e Sand inista reg im e an d re w a rd e d
a n eig h b o rin g sta te th o u g h t to b e in d an ger
fro m th e N icaraguan -su pported reb ellio n .
T h is a ctio n violated G A TT ru les and g en e­
ra ted m u ch criticism o f th e U nited States.
Su ch a quota system in crea s es th e likelihood
th at tra d e policy is used fo r n on econ om ic
reason s.
T h e lesson s fro m th e U.S. su gar p ro g ram
a re stra ig h tfo rw a rd . F irst, sign ifican t costs
have b e e n im posed on U.S. co n su m ers. Se­
cond , th e resu ltin g d isto rtion s in eco n o m ic in ­
cen tiv es h ave h a rm ed U.S. p ro d u cers d ep en ­
d en t on sugar. T h ird , eco n o m ic re sp o n ses to
th e legislation h ave rev ealed a n u m b e r o f
loopholes th a t h ave n ecessita ted additional
re strictio n s and d isto rtion s so th a t U .S. su gar
p ro d u cers cou ld co n tin u e to b en e fit. F ou rth ,
U.S. attem p ts to en su re fa irn e ss have n e c e ssi­
ta ted su b stan tial re so u rc e s to a sce rta in p ro ­
d uctio n and tra d e b eh av io r. Finally, th e p ro ­
g ram h as b e e n used fo r political p u rp o ses to
re w a rd and pu nish fo reig n co u n tries.

43

B en efits to O ther Parties
T h e supply an d d em and analysis o f quotas
and v o lu n tary ex p o rt re stra in ts highlights th e
d iffe re n c e p e r u nit o f im p o rt b etw ee n w h at
d om estic and fo reig n co n su m ers pay. T h is p rice
d ifferen tial re fle c ts th e e x te n t o f th e gains th at
a re available fo r som e grou p to app rop riate.
W ith ta riffs, th e p rice d ifferen tial is cap tu red by
th e d om estic g o v ern m en t in th e fo rm o f ta r iff
rev en u e. W ith n o n -ta riff b a rrie rs, th e dom estic
g o v ern m en t is n o t a d ire ct b e n e ficia ry u n less it
sells th e righ ts to im p o rt to th e h ig h est bidd ers.
O th erw ise, d om estic im p o rters, fo reig n e x ­
p o rte rs and fo reig n g ov ern m en ts cap tu re th ese
gains. T h e poten tial d istribu tion o f th e se b e n e ­
fits can in flu en ce th e d om estic g overn m en t's
ch o ice b etw ee n ta r iff and n o n -tariff b a rrie rs.
W ith volu n tary ex p o rt re stra in ts, th e p rice dif­
fe re n tia l id en tified above is typically cap tu red
b y th e exp o rtin g firm s fro m th e fo reig n co u n ­
try . T h is re su lt m ay re d u ce th e likelihood th a t
th e fo reig n co u n try w ill re ta lia te against su ch
re strictio n s. Given c e rta in d em and con ditions in
b o th th e U.S. and fo reig n m ark ets, vo lu n tary e x ­
p o rt re stra in ts can en tail a su bstan tial re d istrib ­
u tion fro m co n su m ers in th e im p orting co u n try
to selected p ro d u cers in th e exp o rtin g co u n try .
F or exam ple, Collyns and D unaw ay (1987)
estim ate th a t th e U .S.-Jap an ese v o lu n tary ex p o rt
re stra in t on au tom obiles yielded in creased
b e n e fits to selected Ja p a n ese auto p ro d u cers
ran ging fro m $1 billion to $ 5 .2 5 billion in 1984.
Hillman and U rsp ru n g (1988) ex ten d th e
p reced in g idea using a sim ple m odel o f trad e
policy form u lation in w h ich a d em o cratic
g o v ern m en t is choosing b etw ee n a ta r iff and a

16Husted (1986) also connects foreign lobbying to the
domestic economy. He finds that the dollar value of
foreign lobbying in the United States is small relative to
other traded service flows and that the returns to foreign
lobbying generate large returns. For example, Husted
calculated that the expenditure in the United States of
$1.4 million on foreign lobbying by the world automobile
industry came primarily from Japan. Given the estimates
by Collyns and Dunaway (1987) and others indicating
Japanese automobile rents exceeded $1 billion in 1984,
U.S. politicians do not appear to be capturing much of
these rents.

v o lu n tary ex p o rt re s tra in t.16 A sim plification in
th is m odel, w h o se im p o rta n ce is discussed
b elow , is th a t rival political can d id ates p lace no
value on ta r iff rev en u e. A ssum e a v o lu n tary e x ­
p o rt re stra in t and a ta r iff g en era te id entical
d om estic p ro d u cer b en e fits. P oliticians will sup­
p o rt th e v olu n tary ex p o rt re stra in t o ver th e
ta r iff b eca u se th e v o lu n tary ex p o rt re stra in t
g en era tes b en e fits fo r foreig n p ro d u cers th a t, in
tu rn , ca n b e ap p rop riated p artially b y th e politi­
cian s in th e fo rm o f cam paign co n trib u tio n s. On
th e o th e r hand, th e ta r iff re v en u e is assu m ed to
have n o value fo r politicians. C andidates fo r
electiv e o ffice a re view ed as an n o u n cin g tra d e
p olicy positions to m axim ize cam paign co n trib u ­
tion s fro m d om estic and fo reig n p ro d u cer
in terests.
In addition to in crea sin g th e p ro b ab ility th at
p ro tectio n ism w ill ta k e th e fo rm o f v o lu n tary
ex p o rt re stra in ts ra th e r th a n ta riffs, th e arg u ­
m e n t rev eals a w ay th a t political can d id ates can
perso n ally ca p tu re re v en u es th at, w ith ta riffs,
w ould have a cc ru e d to th e d om estic g o v ern ­
m ent. N onetheless, th e assu m ption ab o u t th e
p erceiv ed value o f ta r iff re v en u e to p oliticians
an d th e fa c t th a t co n su m er in tere sts a re ig­
n o red in th e analysis suggests one should b e
cau tio u s in g en eralizin g th is resu lt.
T h e possible b e n e fits to d om estic politicians of
u sing n o n -ta riff ra th e r th a n ta r iff b a rrie rs a re
n o t re stricte d to cam paign co n trib u tio n s. F or e x ­
am ple, a ta r iff is a n exp licit ta x on co n su m ers
w h ile a quota is an im plicit ta x on th em . P olicy­
m a k ers m igh t find it e a sie r to su p p o rt quotas
and o th e r n o n -ta riff b a rrie rs b eca u se th e y will
n o t b e d irectly a sso ciated w ith a tax in cre a s e
th a t co n su m ers, as v o ters, m ight r e sist.17

he also found some major differences. Tariffs are biased
toward low-skill rather than capital-intensive industries and
are unrelated to product heterogeneity and the
geographical dispersion of domestic production facilities.
On the other hand, non-tariff barriers are biased toward
capital-intensive industries producing fairly homogeneous
products. Production in these industries tends to be
distributed across regions consistent with the distribution
of population.

17A neglected issue in the preceding comparison of non­
tariff barriers with tariffs is the distribution of these restric­
tions across industries. While Ray (1981) found that non­
tariff barriers and tariffs are biased toward industries in
which the United States has a comparative disadvantage,



JANUARY/FEBRUARY 1989

44

GATT AND NON-TARIFF
B A R R IER S

u n less th e subsidized goods a re sh ow n to b e
cau sin g (or th reaten in g ) “m a teria l” in ju ry to a
d om estic p ro d u cer. T h is co d e also allow s a

T h e h isto ry o f m u ltilateral tra d e neg otiations
dealing w ith n o n -ta riff b a rrie rs is b r ie f.18
M ultilateral tra d e n egotiation s a re co n d u cted
u n d e r th e au spices o f th e G en eral A greem en t
on T a riffs and T rad e, w h ich w as cre a te d sh o rt­
ly a fte r W o rld W a r II. GATT, a te rm th a t e n ­
co m p asses th e m u ltilateral a g ree m en t govern in g
in tern atio n al trad e, th e bod ies ad m inisterin g th e
ag reem en t, and all associated trad e-related a c ­
tivities, has focu sed on th e red u ctio n o f ta r iff
ra th e r th a n n o n -ta riff b a rrie rs. T o date, seven
ro u n d s o f G A TT negotiations have b ee n co m ­
pleted, w ith th e first six co n c e rn e d alm ost e x ­
clusively w ith ta r iffs .19

co u n try to seek re d re ss fo r ca ses in w h ich
a n o th e r c o u n try ’s subsidized ex p o rts d isplace its
ex p o rts in th ird -co u n try m ark ets.

The Tokyo R ound
T h e T okyo Round, th e m o st re ce n tly co m ­
pleted ro u n d lasting fro m 1973 to 1 979, w as a
co m p reh en siv e e ffo r t to re d u ce tra d e o bstacles
stem m ing fro m ta riffs and n o n -ta riff m easu res.
New o r re in fo rc e d ag reem en ts, called "c o d e s,”
w e re re a ch e d on th e follow ing n o n -tariff m ea ­
su res: 1) subsidies and co u n tervailin g duties; 2)
g o v ern m en t p ro cu re m en t; 3) te ch n ica l stan ­
dard s; 4) im p o rt licen sing p ro ced u res; 5) cu s­
to m s valuation; and 6) anti-dum ping.20
T h e co d e on subsidies and cou ntervailing
duties p ro h ib its d ire ct ex p o rt subsidies, ex cep t
u n d e r c e rta in situations in ag ricu ltu re. T his
co d e is n o te w o rth y in exten d ing GATT's p ro h ib i­
tio n o f ex p o rt subsidies to tra d e in ra w m a te r­
ials. B ecau se n early all g ov ern m en ts subsidize
d om estic p ro d u cers to som e ex ten t, th e cod e e s­
tablished c rite ria to distinguish b e tw e e n a do­
m estic and an ex p o rt subsidy. D om estic su bsi­
dies th a t tr e a t d om estic and ex p o rt activities
id entically a re g en erally allow ed. C ountervailing
duties, w h ich a re ta riffs to o ffse t a subsidy
receiv ed by a fo reig n ex p o rter, a re p ro h ib ited

18For a brief history of multilateral trade negotiations, as well
as details on the current negotiations, see The GATT
Negotiations and U.S. Trade Policy, a 1987 study by the
Congressional Budget Office. For additional details on the
current multilateral negotiations, see Anjaria (1986) and
the 1987 report by the United States International Trade
Commission, Operation of the Trade Agreements Program.
19The sixth round, known as the Kennedy Round, marked
the first time for a GATT agreement on non-tariff barriers.
Agreements were reached on an anti-dumping code and
the elimination the U.S. system of American Selling

http://fraser.stlouisfed.org/
Federal Reserve Bank
of St. Louis
FEDERAL
RESERVE BANK OF ST. LOUIS

T h e cod e on g o v ern m en t p ro c u re m e n t states
th at, fo r qualifying n on m ilitary p u rch ases,
g ov ern m en ts (including g ov ern m en t-co n tro lled
entities) m u st tr e a t fo reig n an d d om estic p ro ­
d u cers alike. In addition to resolvin g disputes,
th e cod e estab lish es p ro ced u res fo r opening and
aw ard in g bids.
T h e cod e on te ch n ica l stand ard s attem p ts to
e n su re th a t te ch n ica l regu latio n s and p ro d u ct
stand ard s su ch as labeling, safety, pollution and
quality re q u ire m e n ts do n o t c r e a te u n n e ce ssa ry
o b stacles to trad e. T h e cod e does n o t sp ecify
stand ard s; h o w ev er, it estab lish es ru les fo r set­
ting stand ard s and resolvin g disputes.
T h e cod e on im p o rt licen sin g p ro ced u res,
sim ilar to th e code on te ch n ica l stand ard s, is
n o t spelled ou t in detail. G enerally speaking,
g ov ern m en ts stated th e ir co m m itm en t to sim ­
plify th e p ro ced u res th a t im p o rte rs m u st follow
to o b tain licen ses. R ed ucing delays in licen sin g
and p ap erw o rk a re tw o are a s o f special
in terest.
T h e code on cu sto m s valuation estab lish ed a
u n ifo rm system o f ru les to d eterm in e th e cu s­
tom s value fo r im p orted goods. T h is cod e u ses
tra n sa ctio n p rices to d eterm in e value and is
designed to p reclu d e th e u se o f a rb itra ry values
th a t in cre a s e th e p ro tectiv e e ffe c t o f a ta r iff
rate.
Finally, th e anti-dum ping co d e p re scrib e s ru les
fo r anti-dum ping investigations, th e im position
o f anti-dum ping duties and settling disputes.
T h e stand ard s fo r d eterm in in g in ju ry a re c la ri­
fied. T h is cod e obligates d eveloped co u n tries to
tr e a t developing co u n tries p referen tially .

Prices, which applied a tariff rate for certain imports to an
artificially high dutiable value. The dutiable value was set
equal to the price of a competing good produced
domestically instead of to the import’s actual invoice price.
This system was applied to a small portion of total imports,
primarily benzenoid chemicals and rubber footwear. Both
agreements were blocked by Congress, but were accepted
in the next round of negotiations.
20Non-tariff barriers were also reduced in civil aircraft and
selected agricultural goods, primarily meat and cheese.

45

T h e U ruguay R o u n d
T h e T okyo R ound cod es have relied on goodfaith com p liance, w h ich has ten d ed to u n d e r­
m ine th e ir effectiv en ess. Stream lin in g and
resolving disputes is a p rio rity d uring th e c u r ­
re n t ro u n d o f m u ltilateral n egotiations, th e
U ruguay Round. T h e T o k yo R ound cod es will
b e rev iew ed and possibly m odified d uring th e
U ru gu ay Round. In p articu lar, b ro ad en in g th e
g o v ern m en t p ro c u re m e n t cod e to inclu d e s e r ­
v ice co n tra cts w ill b e d iscussed. C o n cern in g th e
te ch n ica l stand ard s code, ag reem en ts dealing
w ith th e m u tu al a ccep ta n ce o f te st data g e n e r­
ated b y o th e r p arties and th e op en n ess o f th e
activities o f stand ard s bod ies w ill b e sought. A
m a jo r issue in th e anti-dum ping cod e is ho w to
han d le inp ut dum ping (that is, e x p o rt sales o f
p ro d u cts th a t co n tain inputs p u rch ased at
dum ped prices).
T h e U ruguay Round, b eg u n in Sep tem b er
1 9 8 6 , h as and w ill d iscuss a n u m b er o f nonta r iff b a r r ie r issues, m any o f w h ich exten d
bey o n d th e cod es o f th e T okyo Round. T ra d e
issues involving ag ricu ltu re and serv ices (bank­
ing, co n stru ctio n , in su ra n ce and tran sp o rtatio n )
a re o f p aram o u n t im p o rtan ce. T h e U nited States
h as p rop osed th e elim in ation o f all trad e- and
prod u ction-d istortin g ag ricu ltu ral policies. W hile
th e m ajo r ag ricu ltu ral nation s have ag reed to
th e p rin cip le o f liberalizing ag ricu ltu re, th e
sw eeping n a tu re o f th e U.S. prop osal h as b e e n
re sisted b y som e n ations, esp ecially th e E u ro ­
pean Com m unity. W ith re sp e c t to services, th e
p rim a ry goal is to establish p rin cip les fo r e x te n ­
ding G A TT co v erag e to th is trad e.
A re c e n t study b y th e C ongressional Budget
O ffice (1987) p red icts th a t th e p e rfo rm a n c e o f
th e U ruguay Round will b e judged largely on its
handling o f n o n -ta riff b a r r ie r issues. GATT has
n o t effectiv ely co m b atted risin g n o n -ta riff b a r ­
rie rs fo r m any reaso n s. T w o reaso n s a re th at
th e e ffe c ts o f n o n -ta riff b a rrie rs a re less tra n s­
p a re n t th a n th e e ffe c ts o f ta riffs and, in m any
cases, n o n -ta riff b a rrie rs a re designed to satisfy
a d om estic ra th e r th a n an in tern atio n al o b je c ­
tive. A m ajo r o b stacle is d eterm in in g at w h at
po in t a nation al eco n o m ic policy, w h o se in te rn a ­
tio n al e ffe c ts a re so m ew h at u n certa in , b eco m es
an in tern atio n ally u n acce p tab le n o n -ta riff b a r ­
rie r. T h e se national eco n o m ic policies have f r e ­
qu en tly resu lted fro m th e lobbying e ffo rts o f
stro n g d om estic co n stitu en cies su ch as
ag ricu ltu ral in tere sts. T hu s, m ajo r tra d e policy



re fo rm will b e m et w ith m u ch re sista n ce fro m
th e se groups.

CONCLUSION
N on -tariff b a rrie rs have e ffe c ts sim ilar to
th o se o f ta riffs: th ey in cre a s e d om estic p rices
and im pede tra d e to p ro te c t selected p ro d u cers
at th e ex p en se o f d om estic co n su m ers. As
sh ow n in th e ca se studies o f su gar and au tom o­
b iles, th ey also h ave o th e r e ffe c ts, g en erally
ad verse.
D espite th e ad verse n atio n al co n seq u en ces,
th e use o f n o n -ta riff b a rrie r s h as in crea sed
sh arp ly in re c e n t y e a rs. T h e ch a n c e s fo r a r e ­
v ersal o f this tre n d ap p ear to b e small. T h e
v a riety o f n o n -ta riff m easu res, th e d ifficu lties o f
id en tify in g and m easu rin g th e ir e ffe c ts and th e
b en e fits receiv ed b y sp ecific grou p s co m b in e to
m ake a sign ifican t re d u ctio n o f n o n -ta riff b a r ­
rie rs in th e ongoing U ru gu ay R ound n eg o tia­
tio n s unlikely.
T h e original m ission o f GATT, w h ich has b ee n
largely ach ieved , w as to re d u ce ta riffs. T h e
qu estion , h o w ev er, o f w h y p o licy m ak ers have
p re fe rre d to u se n o n -ta riff b a r rie rs ra th e r th a n
ta riffs in re c e n t y e a rs rem ain s. T h e m o re c e r ­
ta in p ro tectiv e e ffe c ts o f n o n -ta riff b a rrie rs is
on e plausible explan ation . A seco n d explanation,
w h ich fo cu se s on th e d istribu tio n o f th e b e n e ­
fits, is th a t th e b e n e fits o f n o n -ta riff b a rr ie rs
ca n b e ca p tu red by fo reig n p ro d u cers and
d om estic politicians. Su ch an allocation o f b e n e ­
fits in cre a s e s th e p ro b ab ility th a t th e political
p ro cess g en era tes la rg e r am oun ts o f n o n -ta riff
b a r rie rs relativ e to ta riffs. A fin al exp lan ation is
th a t th e ir ad v erse e ffe c ts a re g en erally less o b ­
vious to co n su m ers th a n th e e ffe c ts o f ta riffs.

REFERENCES
Anjaria, S.J. “ A New Round of Global Trade Negotiations,”
Finance and Development (June 1986), pp. 2-6.
Bhagwati, Jagdish N. “ On the Equivalence of Tariffs and
Quotas,” in R.E. Caves et al., eds. Trade, Growth, and the
Balance of Payments: Essays in Honor of Gottfried Haberler
(Rand McNally, 1965), pp. 53-67.
________“ More on the Equivalence of Tariffs and
Quotas,” American Economic Review (March 1968),
pp. 142-46.
Chow, Peter C. Y., and Mitchell Keliman. “ Anti-LDC Bias in
the U.S. Tariff Structure: A Test of Source Versus Product
Characteristics,” Review of Economics and Statistics
(November 1988), pp. 648-53.

JANUARY/FEBRUARY 1989

46

Collyns, Charles, and Steven Dunaway. “ The Cost of Trade
Restraints: The Case of Japanese Automobile Exports to
the United States,” International Monetary Fund Staff
Papers (March 1987), pp. 150-75.
Coughlin, Cletus C., and Kenneth C. Carraro. “ The Dubious
Success of Export Subsidies for Wheat,” this Review
(November/December 1988), pp. 38-47.
Coughlin, Cletus C., K. Alec Chrystal, and Geoffrey E. Wood.
“ Protectionist Trade Policies: A Survey of Theory,
Evidence and Rationale,” this Review (January/February
1988), pp. 12-29.
Deardorff, Alan V. “ Why do Governments Prefer Nontariff
Barriers?” in Karl Brunner and Allan H. Meltzer, eds.
Bubbles and Other Essays, Carnegie-Rochester Con­
ference Series on Public Policy (North-Holland, 1987),
pp. 191-216.
Feenstra, Robert C. “ Automobile Prices and Protection: The
U.S.-Japan Trade Restraint,” Journal of Policy Modeling
(Spring 1985), pp. 49-68.
Herander, Mark G., and Christopher R. Thomas. “ Export
Performance and Export-lmport Linkage Requirements,”
Quarterly Journal of Economics (August 1986), 591-607.
Hillman, Arye L., and Heinrich W. Ursprung. “ Domestic
Politics, Foreign Interests, and International Trade Policy,”
American Economic Review (September 1988), pp. 729-45.

Laird, Sam, and Alexander Yeats. “ Nontariff Barriers of
Developed Countries, 1966-86,” Finance & Development
(March 1989), pp. 12-13.
Laird, Sam, and Alexander Yeats. Quantitative Methods for
Trade Barrier Analysis (Macmillan, forthcoming).
Maskus, Keith E. “ The International Political Economy of
U.S. Sugar Policy in the 1980’s,” United States Depart­
ment of State, Bureau of Economic and Business Affairs,
Planning and Economic Analysis Staff, Working Paper #1
(September 1987).
Nogues, Julio J., Andrzej Olechowski, and L. Alan Winters.
“ The Extent of Nontariff Barriers to Industrial Countries'
Imports,” The World Bank Economic Review (1986),
pp. 181-99.
Page, Sheila. “ The Rise in Protection Since 1974,” Oxford
Review of Economic Policy (Spring 1987), pp. 37-51.
Ray, Edward John. “ The Determinants of Tariff and Nontariff
Trade Restrictions in the United States,” Journal of
Political Economy (February 1981), pp. 105-21.
“ Survey of Automotive Trade Restrictions Maintained by
Selected Nations.” Office of International Sectoral Policy,
U.S. Department of Commerce, in hearings on Fair
Practices in Automotive Products Act before the Subcom­
mittee on Commerce, Transportation and Tourism, March
2, 1982, pp. 113-23.

Husted, Steven. “ Foreign Lobbying and the Formation of
Domestic Trade Policy,” paper presented at Western
Economic Association Meeting, San Francisco, July 1986.

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 Commission (December 1984).

Jones, Kent. “ The Political Economy of Voluntary Export
Restraint Agreements,” Kyklos (1984), pp. 82-101.

U.S. Congress, Congressional Budget Office. The GATT
Negotiations and U.S. Trade Policy (GPO, June 1987).

Krishna, K. “ Trade Restrictions as Facilitating Practices,”
National Bureau of Economic Research, Working Paper
#1546 (1985).
Krugman, Paul R., and Maurice Obstfeld. International
Economics (Scott, Foresman, 1988).


http://fraser.stlouisfed.org/
Federal Reserve Bank
of St. Louis
FEDERAL
RESERVE BANK OF ST. LOUIS

U.S. Department of Commerce, Bureau of the Cen­
sus.Statistical Abstract of the United States: 1988 (GPO,
1987).
U.S. International Trade Commission. Operation of the Trade
Agreements Program—39th Report, 1987 (USITC, July
1988).

47

Michael T. Belongia
and Werner Hermann
Michael T. Belongia is a research officer at the Federal
Reserve Bank of St. Louis. Werner Hermann, an economist at
the Swiss National Bank in Zurich, is a visiting scholar at the
Federal Reserve Bank of St. Louis. Dawn Peterson and Laura
Prives provided research assistance.

Gan a Central Bank Influence
Its Currency's Real Value?
The Swiss Case
T

J L HE SW ISS National B ank (SNB) is on e o f
th e few ce n tra l b an k s th a t co n d u cts m o n etary
policy b y an n o u n cin g and g en erally achieving a
ta rg e ted g ro w th ra te fo r th e m o n ey stock.
Policy is co n d u cted in th is m a n n er b ec a u se SNB
officials, believing th a t excessiv e g ro w th in th e
m on ey stock is th e cau se o f inflation , have
estab lish ed long-run p rice stability as th e ce n tra l
b an k 's p rim ary objective. M o reo v er, b eca u se
large, u n ex p ected ch an g es in m oney g ro w th are
th o u g h t to c r e a te u n certa in ty th a t ca n raise re a l
in te re st ra tes and re d u ce output, SNB o fficials
b eliev e th e average ra te o f m oney g ro w th n ot
only should b e low (to ach iev e low inflation
rates) b u t stable as w ell.1 T hu s, in Sw itzerland ,
b o th rapid m oney g ro w th and large, u n ex p ected
ch an g es in th e m o n ey stock have b ee n ra re .
T h e h isto rical evid en ce clearly in d icates th a t
SNB actio n s have m et th e ir o b jectiv es: g ro w th in
th e m o n etary b ase sin ce 1 9 8 2 , fo r exam ple, has
b ee n targ eted at ra tes b etw ee n 2 p e rce n t and 3
p e rc e n t and h as b ee n , on average, 2 .4 p e rce n t
o v er th e se sev en y ears. T h e average ra tes o f in ­
flatio n and re al G ross D om estic P ro d u ct (GDP)

1See, for example, the arguments put forth by Mascaro and
Meltzer (1983).
2Whether profits rise or fall will depend on the elasticity of
demand for exports. See Belongia and Hermann (1989) for



g ro w th o v er th e sam e sev en y e a rs h ave b ee n
2 .0 p e rc e n t and 2 .3 p e rce n t, resp ectiv ely.
D esp ite its co m m itm en t to m o n ey g ro w th
ta rg e ts, th e SNB rea liz es Sw itzerlan d is a sm all
o p en eco n om y th a t ex p o rts ab o u t 4 0 p e rc e n t o f
G ross National P ro d u ct (GNP). T h u s, dom estic
re a l activity ca n b e a ffe cte d ad v ersely b y ap­
p recia tio n s o f th e Sw iss fra n c th a t raise th e re a l
p rice o f Sw iss goods to fo reig n b u y e rs relativ e
to p rices ch a rg ed by com p etin g foreig n sup­
pliers. If, fo r exam ple, Sw iss e x p o rte rs respon d
to an ex ch a n g e ra te a p p reciatio n by red u cin g
Sw iss fra n c p rices (w hich will m ain tain th e
fo reig n cu rre n c y p rice o f th e ir goods), th e
q u an tity o f ex p o rts sold w ill n ot ch an g e b u t
th e ir p ro fit m argin s will sh rin k. On th e o th e r
hand, if ex p o rte rs m aintain c u rre n t Sw iss fra n c
p rices, th e p rices paid by fo reig n b u y ers will
rise (because o f th e ex ch an g e ra te ap preciation)
and th e qu an tity o f exp o rts sold w ill d eclin e.2
T h e se sp ecific e ffe c ts on Sw iss e x p o rters p ose a
policy p ro b lem b ecau se, in th e aggregate, th ey
a re likely to cau se som e red u ctio n in re a l GNP
g ro w th .

some estimates of the responsiveness of Swiss exports to
exchange rate changes.

48

W h a t m akes th e Sw iss case in tere stin g in th is
co n tex t is th a t a ce n tra l b a n k co m m itted to
m oney g ro w th ta rg e ts and a low, stable in fla ­
tio n ra te m ade an abru p t, b u t te m p o rary , policy
sh ift b eca u se o f ex ch an g e ra te p re ssu res. Sp ecif­
ically, th e SNB aban d o n ed its m oney g ro w th
ta rg e ts in 1978-79 in an attem p t to re d u ce th e
re al value o f th e Sw iss fra n c in fo reig n e x ­
ch an g e m ark ets and av ert a re ce ssio n .3 As
fig u re 1 show s, th e fra n c had ap p reciated
sharply b o th against th e D eu tsch e m ark (DM)
and th e dollar, w h ich re p re se n t th e tw o m ost
im p o rtan t cu rre n cie s fo r Sw iss trad e . In
resp o n se to this cu rre n c y ap p reciatio n , th e
Sw iss m o n etary b ase w as expand ed at an a n ­
nual ra te o f 95 p e rc e n t b e tw e e n Ju ly 1 9 7 8 and
Ja n u a ry 1979.
T h e rapid m oney g ro w th and ex ch an g e ra te
m ovem ents show n in th e fig u re provide a case
study to help an sw er th e type o f qu estio n faced
b y m any co u n tries, including th e U nited States,
in re c e n t y ears: If th e re al value o f a nation 's
c u rre n c y has risen “too h ig h ,” o v er w h at tim e
h o riz o n and b y w h at am ou n t can actio n s by th e
ce n tra l b a n k re d u ce th e re al ex ch an g e ra te ?4 In
th is article, w e use th e Sw iss e x p e rie n ce o ver
th e p eriod o f flexib le ex ch an g e ra te s and som e
o rth o d o x re su lts fro m eco n om ic th e o ry to sug­
g est g en eral co n clu sio n s ab o u t th e e ffe c ts o f
m o n etary actio n s on th e real ex ch an g e rate.

NOMINAL AND REAL EXCHANGE
RATES
T h e re al ex ch an g e ra te is defined as th e
n om inal spot ra te ad justed fo r p rice level
d iffe re n c e s a cro ss co u n tries. If p u rch asin g
p o w er parity (PPP) cond itions w e re m et co n ­
tinuously, th e re al ex ch an g e ra te w ould b e
co n stan t. B ecau se eco n om ic d evelopm ents o ften
3See Rich and Beguelin (1985, p. 85) for a discussion of
this episode and, in particular, SNB reference to a
DM/Swiss franc exchange rate lower bound of 0.80 (their
footnote 9).
4Note that, in many respects, the Swiss debate parallels
that of the United States in the early 1980s. During that
period, analysts discussed the relationships between ex­
change rates and exports [see, for example, Batten and
Belongia (1986)] and the appropriate responses of both the
Federal Reserve and foreign central banks to a rising
dollar [see Batten and Kamphoefner (1982) and Batten
and Ott (1984)]. This attempt to change the level of the ex­
change rate is to be contrasted with efforts to reduce ex­
change rate volatility. Gartner (1987) offers some evidence
on the latter case for the SNB.
5Unless we are dealing in very special cases—such as a
world with indexed contracts or no unexpected price

http://fraser.stlouisfed.org/
Federal Reserve Bank
F F n Fof
R ASt.
I Louis
RFSFRVF

RANK OF ST

I O IIIS

a ffe c t th e n om in al spot ra te and p rice levels
a cro ss co u n tries w ith d iffe re n t lags and in d if­
fe r e n t w ays, h o w ev er, th e re a l ex ch a n g e ra te
g en erally v aries th ro u g h tim e. M ovem ents in
th e re a l ex ch a n g e ra te, th e re fo re , re p re s e n t
th o se ch a n g es in th e n om in al ra te th a t ca n n o t
b e a ttrib u ted to in flatio n d ifferen tials. Sp ecifical­
ly, ch an g es in th e re a l ex ch a n g e ra te re fle c t
stru c tu ra l ch an g es in re a l eco n o m ic p e r fo r ­
m a n ce a cro ss co u n tries.
D istinguishing b e tw e e n th e re a l and nom inal
ex ch a n g e ra te is cru cia l to an y analysis o f th e
e ffe c ts o f ex ch a n g e ra te m ov em en ts b ec a u se o n ­
ly ch an g es in a cu rre n cy 's re a l value a ffe c t
tra d e flow s. A ch an g e in th e n om in al ex ch a n g e
ra te alon e w ill n o t a ffe c t tra d e flow s; th e p o te n ­
tial b e n e fits fro m im p ortin g Sw iss goods due to
th e d ecline in th e fr a n c ’s n om in al value w ill b e
o ffse t exactly b y th e h ig h er p rice s fo r Sw iss
goods th a t cau sed th e nom inal d ep recia tio n .5
T h u s, if a ce n tra l b a n k ’s actio n s a re in ten d ed to
in flu en ce tra d e flow s — n o t sim ply to ch an g e
th e relatio n sh ip b e tw e e n fo reig n and d om estic
p rice levels — th e analysis m u st fo cu s on w h a t
h ap p en s to th e re a l ex ch a n g e r a te .6 M oreov er,
b ec a u se w e a re in tere ste d in how m o n eta ry a c ­
tion s m ight a ffe c t th e re a l ex ch a n g e ra te, w e
m u st exam in e eco n om ic m odels th a t p erm it c e n ­
tra l b a n k actio n s to do so.

TH EO RETICAL MODELS OF
EXCHANGE RA TE RESPONSES TO
MONETARY CHANGES
E stablish in g a relatio n sh ip b e tw e e n ch a n g es in
th e m o n ey stock and re a l eco n om ic m agnitu d es
p ro d u ces som e co n flict am ong com p eting
eco n om ic th eo ries. O ne class o f m odels posits
d iffe re n t speeds o f ad ju stm en t a cro ss m a rk ets
changes—nominal exchange rate changes also will be
changes in the real exchange rate. For some simple ex­
positions of these relationships and the distinction between
real and nominal exchange rates, see Batten and Luttrell
(1982) and Batten and Belongia (1986).
6At least two important issues are ignored from this
perspective. First, the rapid money growth associated with
the 1978-79 intervention was followed by a rapid increase
in the Swiss inflation rate. Second, Swiss importers and
consumers, who benefit from a higher exchange rate, will
be made worse off if the exchange rate declines. Thus, a
discussion of the net benefit of a lower exchange rate is
considerably more complicated than a narrow focus on the
welfare of Swiss exporters alone.

49

Figure 1

Real Exchange Rate of the Swiss Franc/$, Swiss Franc/DM
and the Swiss Monetary Base

in resp o n se to a m o n etary change. In th is case,
re a l m agnitud es ca n b e a ffe cte d by fullyan ticip ated m o n etary ch an g es b ecau se , say,
p rices o f fin an cial assets re a c t m o re quickly
th a n p rices o f d u rable goods and, as a co n se ­
q u en ce, relative p rices, ou tp u t and o th e r real
m agnitudes m ay b e a ffe cte d in th e sh o rt ru n.
A n o th er class o f m odels hyp oth esizes th a t fully
an ticip ated even ts w ill n o t a ffe c t re al variables
b ec a u se th e y alread y w ill in co rp o ra te th e se e x ­
p ectatio n s into c u rre n t values. T hu s, only
"sh o ck s” o r "su rp rise s” a re allow ed to a ffe c t real
m agnitudes. D espite th e ir p a rticu la r d iffere n ces,
h o w ev er, m odels fro m b o th classes p re d ict th a t
m o n etary ch an g es w ill a ffe c t re al variables only
te m p o ra rily .7 W e d iscuss sp ecific m odels o f each
type in th e sectio n s below .

Fully A nticipated M onetary
C hanges: T h e D o rn b u sch M odel
O ne m odel th a t re la te s ex ch a n g e ra te m ove­
m en ts to a ctu al ch an g es in th e m on ey stock is
th e D o rn b u sch (1976) m odel o f ov ersh ootin g
re a l ex ch a n g e ra tes. T h e m odel is derived fo r a
sm all co u n try w h o se actio n s ca n n o t a ffe c t th e
w o rld eco n om y; m o reo v er, it is assu m ed th a t
p e rfe c t cap ital m obility exists (that is, in te re st
ra te p arity holds continuously) and peop le fo rm
ex p ectatio n s rationally. T h e m odel in clu d es a
m on ey m a rk et w ith a stan d ard m oney dem and
fu n ctio n an d a m a rk e t fo r d om estic goods.
In th e long ru n , th is m odel assu m es th a t e x ­
ch an g e ra te s w ill b e co n siste n t w ith p u rch asin g

7Note that this also implies trade will be affected, if at all,
only temporarily and after some lag.



IAKI I I A D V / C C D D I I A D V

1QOQ

50

Figure 2
Exchange Rate Adjustments in the
Dornbusch Model

p o w er p arity. In th e sh o rt ru n , h o w ev er, th e
possibility th a t ex ch an g e ra te s m ay o v ersh o o t
th e ir long-run PPP values is in tro d u ced th rou g h
d iffe re n t speed s o f ad ju stm en t in fin an cial and
goods m ark ets.
F ig u re 2 illu strates th e e ffe c ts o f a m o n etary
ch an g e on th e re a l ex ch an g e ra te, e/P, (the p rice
o f th e fo re ig n c u rre n c y divided b y th e d om estic
p rice level; th e fo reig n p rice level is assu m ed to
b e con stan t) and th e d om estic p rice level, P. At
all points on C^, th e m o n ey m a rk et is in eq u ilib ­
rium , th a t is, in te re st ra te parity holds. T h e
goods m a rk et equ ilibriu m is re p rese n te d b y th e
v ertical line lab eled PPP b ecau se , on th is line,
p u rch asin g p o w er p arity holds. M o reo v er, it is
assu m ed th a t th e goods m a rk et eq u ilibriu m o c ­
cu rs fo r a given re al ex ch an g e rate. A m o n etary
exp an sio n im plies th a t Q0 sh ifts u pw ard to Q r
T h e n ew long-run equilibriu m , at point C, is
given by th e in tersectio n o f PPP and Q, at th e
sam e re a l ex ch an g e ra te b u t a h ig h er d om estic
p rice level. T h is im plies th a t an ex p an sio n ary
m o n etary policy ch an g e does n o t a ffe c t re a l
v ariables in th e long ru n .
T h e tran sitio n fro m th e old to th e n ew eq u i­
librium , h o w ever, does n o t tak e place th ro u g h a
d ep reciatio n o f th e nom inal ex ch an g e ra te th a t

http://fraser.stlouisfed.org/
Federal Reserve Bank
of St. RESERVE
Louis
FEDERAL
BANK OF ST. LOUIS

is exactly in line w ith in crea ses b o th in th e
d om estic p rice level and th e d om estic nom inal
in te re st ra te. In th e sh o rt ru n th e m o n ey
m a rk e t w ill d om inate th e goods m a rk e t b ec a u se
it re a c ts in stan tan eou sly to a m o n eta ry ch an g e
and finds its n ew eq u ilibriu m im m ediately. T h e
goods m ark et, h o w ever, is o u t o f eq u ilib riu m in
th e sh o rt ru n b eca u se p rice ad ju stm en ts lag and
th e qu an tity o f goods dem anded ex ceed s q u an ti­
ty supplied a t th e existin g p rice level.
T h is ex cess liquidity w ill ca u se sh o rt-te rm in ­
te r e s t ra te s to fall and, th u s, w ill m ake th e
d om estic c u r re n c y less a ttra ctiv e to hold. F u r­
th e rm o re , in v esto rs know th a t th e cu rre n c y
m u st d ep recia te to re s to re an equ ilibriu m b e ­
tw e en th e goods m a rk et and m on ey m ark et.
T h e re fo r e , th ey will m ove to sh ift p o rtfo lio s im ­
m ediately fro m d om estic assets in to fo reig n
assets. T h is po rtfo lio shifting, in d u ced b y th e e x ­
cess supply o f d om estic c u rre n cy , w ill cau se
b o th nom inal and re a l ex ch a n g e ra te s to d ep re­
ciate u ntil in te re st ra te p arity is reestab lish ed .
At this in term ed iate stage o f th e ad ju stm en t
p ro cess, in d icated b y p oin t B in fig u re 2,
d om estic in te re st ra te s a re b elo w fo reig n in ­
te re s t ra tes, an d th e d om estic p rice level has
n o t y e t adjusted. W h e n th e p rice level ev en tu a l­
ly does adjust, th e re is a m o v em en t along Q, in
th e fig u re to w a rd th e n ew long-term
equilibriu m , C. At C, th e n om in al ex ch a n g e ra te
has d ep recia ted b u t th e re a l ex ch a n g e ra te has
re tu rn e d to its initial value.
T h e m ech a n ics o f th e D o rn b u sch m od el—
specifically, th e initial ad ju stm en t fro m p o in t A
to p oin t B follow ed b y th e p erm a n en t, lon g-run
ad ju stm en t to p o in t C—im ply th a t th e SNB ca n
in flu en ce re a l ex ch a n g e ra te s in th e sh o rt ru n
at th e p rice o f a te m p o ra ry in flation. T h e m odel
also in d icates th at, in th e long ru n , m o n eta ry
policy h as n o e ffe c t on re a l ex ch a n g e ra tes.
B oth co n clu sio n s a re valid, h o w ev er, only if
o th e r ce n tra l b a n k s do n o t o ffse t th e m e a su res
ta k en b y th e SNB. O verall, given c e rta in sim pli­
fying assu m ption s, th e D o rn b u sch m odel o ffe rs
tw o testa b le p roposition s: Do m o n eta ry actio n s
cau se ch an g es in th e re a l ex ch a n g e ra te and, if
so, o v er w h a t p erio d o f tim e?

An Alternative M odel o f the R eal
E x ch a n g e R ate: T h e In flu e n c e o f
U nexpected M onetary Changes
A n o th er strateg y in m odeling th e re a l e x ­
ch a n g e ra te, co n siste n t w ith th e ea rlie r discus-

51

sion and follow ing th e asset ap p roach to e x ­
ch a n g e ra te d eterm in atio n, has b ee n to fo cu s on
u n e x p e c te d ch an g es in a sso rted m acro e co n o m ic
v ariables. In co n tra st to th e D o rn b u sch m odel,
w h ich allow s actu al d iffe re n c e s b e tw e e n dom es­
tic and fo reig n variables to have sh o rt-ru n real
e ffe c ts th ro u g h ad ju stm en t lags, o th e r m odels
h ave focu sed on d iffere n ces b e tw e e n actu al and
e x p ected values o f ex p lan ato ry v ariab les in f o r ­
eign and d om estic econ om ies. In this case, al­
th o u gh d iffe re n c e s in th e com m od ity m a rk et
and cu rre n c y m a rk et ad ju stm en t p ro cesse s still
m ay b e im p ortant, th e em ph asis is d irected
m o re to th e in flu en ce o f "su rp rise s” on th e e x ­
ch an g e rate. V ariables included, am ong o th ers,
h ave b e e n u n e x p ecte d ch an g es in th e m on ey
stock, th e g o v ern m en t bu d g et surplus (or
deficit) and re al GNP.
In each case, th e variables fo r th e se m odels
w e re th o u g h t to d eterm in e o r m easu re dif­
fe re n c e s in re a l activity a cro ss co u n tries so th a t
an u n ex p ected ch an g e in th em w ould signal a
re a ssessm e n t o f real p e rfo rm a n c e a cro ss co u n ­
trie s and, h en ce, a reassessm en t o f relative c u r ­
re n c y values. D espite th e ir th e o re tic a l appeal,
h o w ever, m odels o f th is type h ave had lim ited
su ccess in explaining su bstan tial am ou nts o f th e
v ariation in re al ex ch an g e rates. Su rveys by
B o m h o ff and K o rtew eg (1983) and B o m h off
(1987) have provided eco n om ic and eco n o m etric
re a so n s to explain th e decided em p irical failu res
o f th e o re tic a l ex ch an g e ra te equations.
F o r o u r in tere st in th e n a rro w issue o f SNB
m o n etary policy and th e re al ex ch an g e rate, a
m odel adopted by H ooper (1983) and S h a fer and
L oop esko (1983) o ffe rs a stra ig h tfo rw a rd ap­
p ro a ch . Starting w ith cond itions o f u n co v ered
in te re st p arity an d a long-run equ ilibriu m c u r­
re n t a cco u n t b alan ce o f zero , an ex p ressio n fo r
th e log o f th e re al ex ch an g e ra te (RER) can be
w ritte n as:
(1)

RER = (r -

ti‘)

- (r* - n '*) + aXCAB,

w h e re r an d n' d en o te th e n om inal in te re st ra te
and ex p ected in flation, resp ectively, * d en o tes a
fo reig n variable and a lC A B is th e cu m u lative
c u rre n t a cco u n t b alan ce. T hu s, th e log o f th e
re a l ex ch an g e ra te is stated as a positive fu n c ­
8Dropping this variable is justified on two grounds. First,
the Swiss current account balance has been nearly cons­
tant over time such that variations in it are unlikely to be
an important source of exchange rate fluctuations. The se­
cond reason is the theoretical result that it is a persistent
change in CAB — not its level — which will affect the real
exchange rate. For a discussion of this independence



tion o f b o th th e d om estic-fo reign real in tere st
d iffere n tia l and th e cu m u lative c u rre n t a cco u n t
b alan ce.
E qu ation 1 in its c u r r e n t fo rm , h o w ev er, is
n ot d irectly u sefu l fo r o u r p u rp o ses b ec a u se th e
policy qu estio n applies to ch a n g e s in th e re a l e x ­
ch a n g e ra te, n o t its level. M oreov er, w e a re in ­
te re s te d only in th e sim ple b iv a ria te relation sh ip
b e tw e e n m o n eta ry actio n s an d ex ch a n g e ra te
ch an g es. Finally, and p erh ap s m ost im p ortan t,
eq u ation 1, as w ritten , has no sp ecific re fe re n c e
to m o n eta ry policy actions.
T o apply eq u atio n 1 to th e c u r re n t in vestiga­
tio n o f m o n eta ry policy’s in flu en ce on th e re a l
ex ch a n g e ra te , th e CAB te rm w as dropped and
th e rem ain in g te rm s w e re d iffere n ced so th at
ch an g es in th e re a l ex ch a n g e ra te w ere related
to ch an g es in th e re a l in tere st d iffere n tia l.8
T h a t is:
(2)

ARER = A(r - n') - A(r* -

ti'*).

A fter th e se sim plifying assu m ption s and m an ip­
u lations, w e have, in eq u atio n 2, red u ced th e
p rob lem to o n e in w h ich ch an g es in e ith e r th e
d om estic o r fo reig n real in te re st ra tes, o r both ,
cau se a ch a n g e in th e re a l ex ch a n g e ra te .9
T o allow a ro le fo r m o n eta ry policy actio n s in
eq u ation 2, a long h isto ry o f eco n om ic lite ra tu re
suggests th a t u n ex p ected ch an g es in m oney
g ro w th ca n a ffe c t th e re a l in te re st ra te, a t least
tem p o rarily , b y alterin g in fla tio n a ry ex p e cta ­
tion s o r th ro u g h a liquidity e ffe c t. M oreover, if
eq u atio n 2 is a g en erally c o rr e c t exp ressio n fo r
th e re a l ex ch a n g e ra te, an an n ou n ced , cred ib le
policy b y th e SNB to in c re a s e m on ey g ro w th to
re d u ce th e fr a n c ’s re a l value should h ave no e f­
fe c t b eca u se ra tio n a l ag en ts w ill in c o rp o ra te th e
in fo rm a tio n in to rev ised ex p ectatio n s fo r h ig h er
fu tu re in flatio n and, as a co n seq u en ce, h ig h er
nom inal in te re st ra tes and a lo w er nom inal e x ­
ch an g e ra te. But, w ith all n om in al m agnitud es
ad justing b y th e e x a c t am ou n ts and w ith ou t
lags, th e r e is n o latitud e fo r a te m p o ra ry
ch a n g e in th e re a l ex ch a n g e rate.
B ecau se this seco n d m odel suggests th a t fully
an ticip ated m o n eta ry policy actio n s w ill leave
of a stable CAB level and the exchange rate, see Mussa
(1985).

9This abstracts from the special case in which the domestic
and foreign rates change in a way that leaves the differen­
tial unchanged.

. I A N I I A R V / F F R P I I A R Y 1 QHQ

52

th e real in te re st d ifferen tial and, h en ce, th e real
ex ch an g e ra te u n a ffected , eq u ation 2 suggests
th a t a su cce ssfu l attem p t b y th e SNB to re d u ce
th e fra n c's re a l value m u st b e b o th u n a n tic­
ip ated and n o t o ffse t b y th e actio n s o f o th e r
ce n tra l ban ks. W h e th e r th e p red ictio n s o f eith e r
m odel a re su p p orted b y th e data is investigated
in th e n e x t section.

EM PIRICAL ADAPTATION
B ecau se th e sole qu estion o f in te re st is w h e th ­
e r m o n etary actions, h o w ev er m easu red , a ffe ct
th e re al ex ch an g e rate, w e did n o t attem p t to
estim ate stru c tu ra l eq u atio n s d erived fro m
e ith e r o f th e th e o re tic a l m odels d iscussed
e a rlier. Instead, w e ch o se to exam in e statistical
te sts th a t in d icate w h e th e r m o n etary action s
"c a u se ” a ch an g e in th e re a l ex ch an g e r a te .10
M o reo v er, b ec a u se Sw iss m o n etary actio n s th at,
c e te r is p a r ib u s , w ould a ffe c t th e re a l ex ch a n g e
ra te m ay b e o ffse t b y actio n s o f o th e r ce n tra l
b an k s, o u r cau sality tests w e re estim ated b ased
on d iff e r e n c e s b e tw e e n ch an g es in th e g ro w th
ra te s o f th e m o n eta ry b a se a cro ss co u n tries.
T h e g en eral fo rm o f th e eq u atio n s estim ated
fo r th e se tests is d epicted as:

(3) ARER, = a + £ b iBt_i + I c, ARER.^ + £(;
i=o
j=l
w h e re RER is th e re a l Sw iss franc/DM o r Sw iss
franc/dollar ex ch an g e ra te , B is a m e asu re o f
relativ e m o n etary actions, a, b ( and c ; a re co e ffi­
cien ts to be estim ated and £, is a ran d o m e r r o r
term . Lag length s fo r th e exp lan ato ry variables,
p and q, w e re ch o sen by a final p red ictio n e r ­
ro r (FPE) c r ite rio n .11 T h e re al ex ch an g e ra tes
used a re m onthly averages o f th e Sw iss
franc/dollar and Sw iss franc/DM nom inal ra tes
ad justed b y ratio s o f th e resp ectiv e co u n trie s’
co n su m er p rice in d exes. T h e m o n etary b ase

10See Jacobs, et al. (1979) for some of the more common
critiques of causality testing. Also see Zellner (1979) for a
more general discussion of causality tests and their ap­
plication. Although Fratianni, et al. (1987) apply this testing
procedure to the money-exchange rate relationship, their
study does not include the Swiss franc and uses the
nominal, rather than real, exchange rate.
11See, for example, Batten and Thornton (1985).
12These measures of unanticipated monetary changes
should be “ white noise,” series whose movements cannot
be explained by their own past behavior or the behavior of
other variables. Tests for white noise indicated that the

http://fraser.stlouisfed.org/
Federal Reserve FEDERAL
Bank of St.RESERVE
Louis
BANK OF ST. LOUIS

w as ch o sen as th e b a sic m e a su re o f m o n eta ry
actio n s in all th r e e co u n tries. By allow ing co n ­
tem p o ran eo u s values o f m o n eta ry a ctio n s to
e n te r th e se reg ressio n s, w e explicitly assu m e
th a t m o n eta ry policy d ecisions a re exogenous.
T o te st th e D o rn b u sch m odel, th e B v ariab le
in eq u atio n 3 w as m easu red as th e d iffe re n c e
b e tw e e n Sw iss and G erm an o r Sw iss and U.S.
m o n eta ry b a se g ro w th ra tes. F o r th e secon d
m odel, m easu res o f u n ex p ected ch a n g es in th e
b a se g ro w th r a te s w e re n eed ed . In fa ct, th e
relatio n sh ip s d iscussed e a rlie r in d icated th a t B
should b e re p re se n te d as th e d iffe re n c e b e ­
tw e en Sw iss and G erm an o r Sw iss and U.S.
m o n etary su rp rises. T o c o n stru c t th e se m ea s­
u res, seco n d d iffe re n c e s o f logarith m s w ere
u sed to re p re s e n t u n an ticip ated ch an g es in ea ch
individual m o n eta ry b a se series. T h e n th e se in ­
dividual series w e re u sed to c o n s tru c t th e d if­
fe re n c e s b e tw e e n Sw iss an d G erm an o r Sw iss
and U.S. m o n eta ry policy su rp rise s.12
T h e relatio n sh ip s d escrib ed by eq u atio n 3
w e re estim ated w ith m on th ly data o v er a
1 9 7 3 -8 6 sam ple period ; th e re su lts a re sh ow n in
ta b le 1. Sectio n A o f th e tab le, b a sed on d if­
fe re n c e s b e tw e e n a ctu al g ro w th ra te s o f th e
m o n eta ry b a se a cro ss co u n tries, in d icates a
m arginally sig n ifican t e ffe c t o f m o n eta ry actio n s
o n th e re a l Sw iss fra n c ex ch a n g e ra te. T h e FPE
criterio n , p icked only co n tem p o ra n eo u s
m easu res o f m o n eta ry a ctio n s as th e b e s t
re p rese n ta tio n o f th e m odel in th e U.S. case; fo r
th e G erm an case, co n tem p o ra n eo u s an d tw o
lagged values o f rela tiv e m o n eta ry a ctio n s w ere
ch o se n .13 A lthough th is v aria b le w as m arginally
sign ifican t in th e U.S. case, th e sign o f th e
estim ated co e ffic ie n t fo r relativ e m o n eta ry a c ­
tio n s is in c o rre c t. Sin ce th e o ry suggests a
positive resp o n se fo r th is sp ecificatio n o f th e
data, (relatively fa ste r Sw iss m on ey g ro w th will
in cre a s e th e n u m b er o f Sw iss fra n cs p e r dollar)
th e n egative sign in th e U.S. ca se is puzzling.
F o r th e G erm an case, th e estim ated co e fficie n ts
second differences of logarithms of each country’s
monetary base series had this characteristic.
13An alternative would be to estimate models for a variety of
lag lengths and look for patterns in the results. This also
was done for all pairs of possible lags, up to 12 months,
for both the monetary variable and the exchange rate (144
regressions for each of the four equations reported). The
general finding was that significant effects of monetary
actions—whether actual or unanticipated—were found for
all lag pairs up to six months.

53

Table 1
Estimated Relationships Between Monetary Actions and Changes in the
Real Swiss Franc Exchange Rate_____________________________________
A. The Effects of Actual Monetary Actions (B is specified as Ain SWMB - Ain USMB or Ain SWMB - Ain GEMB)

Swiss franc/$:

Ain RER, = - 0.002 - 0.172 (B ,) + 0.310 Ain RER,_,
(1.02)
(1.85)
(4.21)

_2
R = 0.11

Swiss franc/DM:

Ain RER, = - 0.001 + 0.043 (B ,) - 0.036 (B,_,) + 0.149 (B,_2) + 0.449 Ain RER,„ - 0.197 Ain RER,.2
(1.04)
(0.99)
(0.84)
(3.48)
(5.98)
(2.60)

_2

R = 0.20

B. The Effects of Relative Monetary Surprises (B is specified as AAln SWMB - AAln USMB or AAln SWMB - AAln GEMB)

Swiss franc/$:

Ain RER, = - 0.002 - 0.148 (B ,) + 0.330 Ain RER,_,
(0.75)
(2.17)
(4.51)

_2

R = 0.12

Swiss franc/DM:

Ain RER, = - 0.001 + 0.015 (B ,) - 0.032 (B,_, ) + 0.095 (B,_2) + 0.080 (B,_,) + 0.460 Ain RER,.,
(1.39)
(0.39)
(0.70)
(2.06)
(2.07)
(5.98)
- 0.178 Ain RER,.2
(2.29)

_2

R = 0.20
NOTE: Absolute values of t-statistics are in parentheses.

fo r th e co n tem p o ran eo u s and last lag o f th e B
v ariable tak e th e ex p ected positive sign b u t only
th e la tte r te rm is significantly d iffe re n t from
z ero . In g en eral, th e co n clu sio n seem s to b e
th a t d iffe re n c e s b e tw e e n actu al m o n etary
ch an g es a cro ss co u n tries have w eak, short-lived
and u n p red ictab le e ffe c ts on th e re al ex ch an g e
rate.
In terp retin g th e se in co n sisten t re su lts is m ade
so m ew h at easier, h o w ever, b y re tu rn in g to
fig u re 1 and som e points m ade e a rlie r in th e
p ap er. T h e fig u re show s th a t th e large in crease
in th e Sw iss m o n etary b ase d uring 1 978-79 w as
a sso ciated w ith te m p o ra ry ap p reciatio n s o f bo th
th e dollar and DM against th e Sw iss fra n c. T his



relatively sm all and short-lived e ffe c t w as
follow ed, h o w ev er, b y tw o distinctly d iffe re n t
path s fo r th e d ollar and DM against th e Sw iss
fra n c , w ith th e dollar risin g sh arp ly u ntil early
1 9 8 5 an d th e DM re tu rn in g to a p ath o f small,
g rad ual d ep reciatio n s. Also re ca ll th a t th e SNB
d iscussed its policy sta n ce o v er this in terval in
te rm s o f a 0 .8 lo w er b o u n d fo r th e Sw iss
franc/DM ex ch a n g e ra te. T h e DM, o f co u rse,
dom inated o th e r cu rre n c ie s in SNB decisions
b ec a u se G erm an y is S w itzerlan d ’s larg est
trad in g p a rtn e r. O verall, th e se b its and p ieces o f
ev id en ce—th e path fo r th e Sw iss franc/DM e x ­
ch a n g e ra te in th e fig u re, SNB policy statem en ts
reg ard in g a Sw iss franc/DM ob jective, and th e
" c o r r e c t” sign fo r m o n eta ry actio n s in th e Sw iss

IA K IIIA D V /C C D D IIA D V

1 QQQ

54

franc/DM eq u atio n —suggest th a t SNB action s
designed to re d u ce th e fr a n c ’s re a l value did
h ave som e w eak e ffe c t relativ e to its ta rg e t c u r­
re n c y .14 T h a t e ffe c t, h o w ev er, w as dissipated
quickly.
Sectio n B o f th e table, w h ich re d e fin es th e
m o n etary v ariable as th e d iffe re n c e b e tw e e n
u n an ticip ated m o n etary ch an g es a cro ss co u n ­
trie s, show s m o n etary e ffe c ts th a t a re m o re
stron g ly sign ifican t b u t still o f puzzling signs.
T h e ch o se n lag lengths a re relatively sh o rt, sug­
g esting th e tra n sito ry n a tu re o f m o n etary a c­
tio n s on th e re a l ex ch an g e rate.
Nonetheless, th e gen eral results again are p ro b ­
lem atic. T h e o ry suggests th a t an u n ex p ected in ­
c re a se in Sw iss b a se g ro w th th a t is la rg e r th a n
an u n e x p ecte d ch an g e in fo reig n b a se g ro w th in
th e sam e d irectio n should b e re la ted positively
to a ch an g e in th e re al ex ch an g e ra te specified
as franc/foreign cu rre n cy . T h u s, th e positive
signs in th e G erm an case a re co n siste n t w ith
th is re su lt w hile th e negative sign fo r lagged
relativ e m o n etary su rp rises in th e U.S. case is
not. T h e in co n sisten t U .S. resu lt, as in th e
previo u s case, m ight b e explained b y th e in ­
stru m e n ts and o b jectiv es arg u m en t d iscussed in
fo o tn o te 14. Even th e G erm an re su lt is tro u b le ­
som e, h o w ev er, in th a t la g g ed values o f
m o n etary su rp rises have a sign ifican t im p act on
th e re a l ex ch an g e ra te. P resu m ably, a su rp rise
should have its e ffe c t fe lt only in th e p eriod it
o cc u rs bu t, in th is case, its e ffe c ts o cc u r only
w ith lags o f tw o and th r e e m on th s. As w ith th e
re su lts in Sectio n A o f th e table, th e se resu lts
in d icate sign ifican t b u t u n p red ictab le in flu en ces
o f m o n etary actio n s on th e re al ex ch an g e ra te.

CONCLUSIONS
T e n y e a rs ago, th e SNB tem p o rarily a b a n ­
d oned its m o n eta ry ta rg e ts and p u rsu ed w h at
h as b e e n in te rp re te d as a su ccessfu l in te rv e n ­
tio n to re d u ce th e Sw iss fra n c's re al value and
re sto re ex p o rt se c to r com p etitiven ess. A lthough
eco n o m ic th e o ry g en erally suggests th a t su ch an
in terv en tio n —even w h en o th e r c e n tra l b an k s

14A basic rule for policy actions is that policymakers must
have at least as many instruments as the number of objec­
tives they hope to achieve. The SNB has only one
instrument—the monetary base—and could use it to
achieve one exchange rate objective, such as the Swiss
franc/DM rate. Without more instruments, however, it

http://fraser.stlouisfed.org/
Federal Reserve Bank
of St. RESERVE
Louis
FEDERAL
BANK OF ST. LOUIS

co o p e ra te—m ay h ave little e ffe c t o n th e c u r r e n ­
cy ’s re a l value, eco n o m ic policy sum m its o v er
r e c e n t y e a rs o ften h ave d iscu ssed th e possibility
o f co o rd in ated m o n eta ry actio n s to a ffe c t e x ­
ch an g e ra tes. In th is co n tex t, th e Sw iss e x ­
p e rie n ce p re sen ts an in tere stin g case study o f
m o n eta ry e ffe c ts on ex ch a n g e ra tes.
O u r em p irical ev id en ce suggests th a t m o n e­
ta ry actio n s m ight b e related sign ifican tly to
re a l ex ch a n g e ra te m ovem en ts in th e sh o rt ru n .
T h e tro u b le w ith th is re su lt is th a t th e se e ffe c ts
ap p ear to b e u n p red icta b le an d n o t en tirely
co n siste n t w ith stan d ard m od els o f ex ch a n g e
ra te d eterm in atio n . C ausality te sts in d icated th a t
ch an g es in relativ e m on ey g ro w th ra te s b e ­
tw e en co u n tries—w h e th e r actu al o r
u n an ticip ated —in flu en ce th e re a l ex ch a n g e ra te
fo r up to six m on th s. B u t w h ile sh o rt-ru n re la ­
tionships a re co n siste n t w ith th e co n clu sio n th a t
m o n eta ry actio n s a re n o t likely to have e x ­
ch an g e ra te e ffe c ts o f a size o r d u ratio n th a t
w ill b rin g ab o u t su b stan tial in c re a s e s in exp o rts,
th e co n clu sio n s fo r p o licy m ak ers a re less clear.
T h e lack o f co n sisten cy in th e sign o f th e re la ­
tio n sh ip b etw ee n m o n eta ry actio n s and e x ­
ch an g e ra tes a cro ss co u n tries d oes n o t give a
cle a r signal as to w h ich m o n eta ry actio n should
b e ta k en to p ro d u ce a given ex ch a n g e ra te
resp o n se.

REFEREN CES
A k a ik e , H iro tu g u . “ In fo rm a tio n T h e o ry a n d a n E x te n s io n of

the Maximum Likelihood Principle,” in Boris N. Petrov and
Frigyes Csaki, eds., Second International Symposium on In­
formation Theory. Budapest, 1973.
Batten, Dallas S., and Michael T. Belongia. “ Monetary
Policy, Real Exchange Rates, and U.S. Agricultural Ex­
ports,” American Journal of Agricultural Economics (May
1986), pp. 422-27.
Batten, Dallas S., and James E. Kamphoefner. “ The Strong
U.S. Dollar: A Dilemma for Foreign Monetary Authorities,”
this Review (August/September 1982), pp. 3-12.
Batten, Dallas S., and Clifton B. Luttrell. “ Does ‘Tight’
Monetary Policy Hurt U.S. Exports?” , this Review
(August/September 1982), pp. 24-27.
Batten, Dallas S., and Mack Ott. “ What Can Central Banks
Do About the Value of the Dollar?” , this Review (May
1984), pp. 16-26.

could not simultaneously move to hit a Swiss franc/dollar
objective. This lack of instruments to hit multiple objectives
could explain the “ correct” results for the Swiss franc/DM
causality tests and the anomolous results for the Swiss
franc/dollar case.

55

Belongia, Michael T., and Werner Hermann. “ The Demand
for Swiss Exports: An Empirical Analysis,” Geld, Wahrung
und Konjunktur, forthcoming.

Kohli, Ulrich. “A Simple Structural Model of the Swiss
Franc—U.S. Dollar Exchange Rate,” Applied Economics (19:
1987), pp. 381-92.

Bomhoff, Eduard J. “ The Dollar-Yen Exchange Rate,”
Bank of Japan Monetary and Economic Studies (December
1987), pp. 1-31.

Mascara, Angelo, and Allan H. Meltzer. “ Long- and ShortTerm Interest Rates in a Risky World,” Journal of Monetary
Economics (November 1983), pp. 485-518.

Bomhoff, Eduard J., and Pieter Korteweg. “ Exchange Rate
Variability and Monetary Policy under Rational Expecta­
tions: Some Euro-American Experience 1973-1979”, Journal
of Monetary Economics (March 1983), pp. 169-206.
Dornbusch, Rudiger. “ Expectations and Exchange Rate
Dynamics,” Journal of Political Economy (December 1976),
pp. 1161-76.
Frankel, Jeffrey A. “ On the Mark: A Theory of Floating Ex­
change Rates Based on Real Interest Differentials,”
American Economic Review (September 1979), pp. 610-22.
Fratianni, Michele, Hyung-Doh Hur, and Heejoon Kang.
“ Random Walk and Monetary Causality in Five Exchange
Markets,” Journal of International Money and Finance
(December 1987), pp. 505-14.
Gartner, Manfred. “ Intervention Policy Under Floating Ex­
change Rates: An Analysis of the Swiss Case,” Economica
(November 1987), pp. 439-53.
Hooper, Peter. “ Movements in the Dollar’s Real Exchange
Rate Over Ten Years of Floating," mimeo, Board of Gover­
nors of the Federal Reserve System (June 1983).
Jacobs, Rodney, Edward Learner and Michael Ward. “ Dif­
ficulties with Testing for Causation,” Economic Inquiry
(July 1979), pp. 401-13.




Mussa, Michael L. "Commentary on ‘Is the Strong Dollar
Sustainable?’ ” in The U.S. Dollar - Recent Developments,
Outlook, and Policy Options, A Symposium Sponsored by
the Federal Reserve Bank of Kansas City, 1985, pp. 133-55.
Rich, Georg, and Jean-Pierre Beguelin. “ Swiss Monetary
Policy in the 1970s and 1980s,” in Karl Brunner, et al.,
Monetary Policy and Monetary Regimes (Graduate School
of Management, University of Rochester 1985).
Shafer, Jeffrey R., and Bonnie E. Loopesko. “ Floating Ex­
change Rates after Ten Years,” Brookings Papers on
Economic Activity (1: 1983), pp. 1-86.
Thornton, Daniel L., and Dallas S. Batten. “ Lag Length
Selection and Tests of Granger Causality between Money
and Income,” Journal of Money, Credit and Banking (May
1985), pp. 164-78
Zellner, Arnold. “ Causality and Econometrics,” in Karl Brun­
ner and Allan H. Meltzer, eds., Three Aspects of Policy and
Policymaking: Knowledge, Data and Institutions, CarnegieRochester Conference Series on Public Policy, No. 10,
(1979).

IA M l I A B V / C E R D I l& D V 1 QDQ

F e d e ra l R e s e rv e B an k o f St. L o u is
Post O ffice Box 442
St. Louis, Missouri 63166

The R e v ie w is published six
times per year by the Research
and Public Information
Department o f the Federal
Reserve Rank o f St. Louis.
Single-copy subscriptions are
available to the public fr e e o f
charge. Mail requests f o r
subscriptions, back issues, or
address changes to: Research
and Public Information
Department, Federal Reserve
Rank o f St. Louis, P.O. Rox 442,
St. Louis, Missouri 63166.
The views expressed are those
o f the individual authors and do
not necessarily reflect official
positions o f the Federal Reserve
Rank o f St. Louis or the Federal
Reserve System. Articles herein
may be reprinted provided the
source is credited. Please provide
the Bank’s Research and Public
Information Department with a
copy o f reprinted material.