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.