The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
January/February 1988 Vol. 70, No. 1 3 Are IWide Deficits a P rob lem ? 12 Protec tionist Trade Policies: A Survey o f T h e o ry » Evidence and Rationale 30 T h e B orrow ed -R eserves O perating P ro ced u re: T h eo ry and Evidence 55 F a rm Policy and M an d atory Supply C ontrols — The Case o f Tobacco THE FEDERAL A RESERVE Jtk RANK of A r ST. LOUIS F e d e ra l R e se rv e Bank of St. L ouis Review January/February 1988 In This Issue . . . T h ere is co n sid era b le c o n c e rn th at th e su b stan tial U.S. trad e d eficits in cu rred in re cen t y ears are a sym ptom or, even w orse, a ca u se o f gen eral w eak n ess in the U.S. eco n om y. In th e first article in th is R ev iew , K. Alec C hrystal an d Geoffrey E. W ood an sw er th e qu estion , "Are T rad e D eficits A P ro b lem ?” T h ey first d em o n strate th e relatio n sh ip b etw een trade d eficits an d cap ital su rp lu ses u sin g th e U.S. b alan ce o f p ay m en ts a cco u n t. T h ey th e n d istin guish betw een circ u m sta n ce s in w h ich trade d eficits are p ro b lem a tic — for exam p le, th o se p ro d u ced by inflation aiy d o m estic m o n etary p o licy — an d circ u m sta n ce s in w h ich they are not — for exam ple, th o se p ro d u ced bv high private investm ent relative to the d o m estic savings available to fund th em . T h ey co n clu d e by n otin g that the p resen t trad e d eficits ap p ear to have resu lted from a co m b in a tio n o f large federal d eficits and investm en t spen din g. In sofar as th e federal d eficit an d private savings behavior are taken as given, th e ch o ice s that th e U nited States faces are co n tin u ed high lev els o f private in v estm en t d em an d an d trad e d eficits o r b a la n ced trad e and slow real grow th. In th e cu rren t case, red u cin g th e trade d eficit m ight well p ro d u ce th e g reater p roblem . * * * P ro tectio n ist p ressu res have b ee n m o u n tin g w orldw ide d uring th e 1980s. In the se co n d article in th is R eview , C letus C. C oughlin, K. Alec C hrystal an d Geoffrey E. W ood survey th e theory ev id en ce an d ratio n ale co n cern in g p ro tectio n ist trade p o licies. T h e au th o rs illu strate th e gains from free trade u sin g th e co n c e p t of com parative advantage an d review re ce n t d evelop m en ts co n cern in g th e c o n s e q u en ces o f in tern a tio n a l trade in im p erfectly com p etitive m arkets. T h ey argue that, w hile p ro tectio n ist trade p o licies o cca sio n a lly m ay offset foreign m on op oly p ow er or advantageously u se d o m estic m o n o p o ly pow er, trad e restrictio n s g enerally re d u ce b o th th e co m p etitio n faced by d o m estic p ro d u cers and the co n su m p tio n p o ssib ilities o f d o m estic co n su m ers. T h e em p irical evid en ce is clea r-cu t. T h e co sts o f p ro tectio n ist trade p o licies b o rn e by co n su m ers far e x ceed th e gains o f d o m estic p ro d u cers an d governm ent. M oreover, th e adverse c o n su m e r effects are n o t short-lived. P ro tectio n ist trade policies g en erate low er ec o n o m ic grow th rates th a n free trade p o licies. C o n se quently, n ation al in terests will b e served by th e red u ctio n o f trad e b arriers. * * * In the third article, en titled ' T h e B orrow ed R eserves O perating P roced u re: T h eory and E v id en ce,” D aniel L. T h o rn to n analyzes th e Fed eral R eserve’s o p era t ing p ro ced u re as a m eth o d for co n tro llin g th e m o n ey stock o r th e fed eral funds rate. T h o rn to n d em o n stra tes th at th e borrow ed reserves o p eratin g p ro ced u re is n ot an effective m eth o d for co n tro llin g th e m o n ey stock. Ind eed , h e show s th at an in terest rate targeting p ro ced u re w ould be m o re effective in co n tro llin g m oney. Fu rtherm ore, h e show s th at th e borrow ed reserves op eratin g p ro ced u re is an effective m eth o d o f co n tro llin g in terest rates in th e sh o rt ru n only w h en th e variation in borrow ing is due solely to shifts in th e d em an d for total reserves; in th e long run, it is an effective m eth o d o f co n tro llin g in tere st rates only w h en the borrow ings fu n ctio n is stable. T h o rn to n also investigates th e u se o f th e b orrow ing fu n ctio n s sin ce O ctob er 1982. His evid ence suggests th at th e borrow ings o p eratin g p ro ced u re h as b een u sed to offset th e effect o f p erm an en t shifts in borrow ings on th e federal funds rate. 1 In This Issue . . . 2 T h e farm se c to r o f the U nited States ex p e rie n ce d a severe d o w n tu rn in the 1980s d esp ite over $95 b illion in governm ent su p p ort. M any an alysts believe th at su ch govern m en t su p p o rts an d o th e r tradition al U.S. farm p o licies have d on e little to alleviate ag ricu ltu re’s problem and m ay have actu ally co n trib u te d to the farm crisis. As a result, so m e have p ro p o sed m an d atory co n tro ls o n crop p ro d u c tion as an alternative to th e cu rren t policy o f voluntary acreag e red u ctio n and d irect governm en t paym ents. Su ch m an d atory co n tro ls have b een u sed in th e U nited States to su p p o rt the p rice o f to b a c co sin ce the 1930s. In th e final article in th is R eview , “Farm Policy and M andatory Supply C on trols — T h e C ase o f T o b a c c o ,” K en n eth C. C arraro exam in es U.S. to b a c co policy to analyze th e likely co n se q u e n ce s o f exten d in g m an d ato iy su p p ly co n tro ls to o th er m a jo r U.S. crop s. C arraro show s that the United States w as able to u se m an d atory su p p ly co n tro ls su ccessfu lly for m any y ears only b ec a u se b o th w orld su p p ly and w orld d em an d w ere in elastic. As th e w orld elasticity o f b o th su pply an d d em an d in crea sed over tim e, how ever, the effectiveness o f U.S. to b a c co policy erod ed . T h e a u th o r c o n c lu d e s that m a n d a to iy su pply co n tro ls for o th e r crop s w ould n ot b e su cce ssfu l b ec a u se th e co n d itio n s of elasticitv th at w orked well previously for to b a cco are n ot p resen t for o th e r m ajo r crop s. FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Are Trade Deficits a Problem ? K. Alec Chrvstal and Geoffrey E. W ood I n 1986, the U.S. trad e deficit ex ceed e d $140 billion. Su ch su bstan tial trade d eficits often are co n sid ere d a sign o f w eak n ess in th e eco n om y. W hile this situ ation is so m eth in g o f a novelty for th e U nited States, m any o th e r co u n tries have h ad trad e d eficits off an d on th rou gh o u t th e p ostw ar period.' T h e p u rp o se o f th is article is to explain w hat is m eant by trade deficits w ithin th e co n tex t o f th e b a l a n ce o f paym ents, to ou tlin e th e circ u m sta n ce s u n d e r w h ich the state o f the b a la n ce o f pay m ents m ay be sy m p to m atic o f a problem , and to co n sid e r w hat this analysis im plies cu rren tly for th e United States. W ith regard to th e last, w e will suggest that co n c e rn about th e U.S. trad e d eficit h as b ee n overstated. Ind eed , a trade d eficit can be indicative o f a h ealth y and strongly grow ing eco n om y. THE BALANCE OF PAYMENTS ACCOUNTS T h e b alan ce o f pay m ents a cc o u n ts are a reco rd o f tra n sactio n s b etw een d o m estic resid en ts an d th e rest o f the w orld over a sp ecific perio d o f tim e. Like any K. Alec Chrystal is the National Westminster Bank Professor of Per sonal Finance at City University, London. Geoffrey E. Wood is a professor of economics, also at City University, London. This article was written while Chrystal was a professor of economics at the University of Sheffield, Sheffield, England. Vincent T. Waletzki provided research assistance. 'The United States did run deficits in the 19th century, but not quite as big relative to GNP as are current U.S. trade deficits. See Mudd and Wood (1978). d ou ble entry bookkeeping system , th e b a la n ce o f pay m en ts a cc o u n ts m u st b alance.- T h e re is n o th in g m ys teriou s about this, n o r d o es it involve any statem en t about h ow th e w orld w orks. T h e sim p lest form in w h ich th e b a la n ce o f pay m en ts a cc o u n ts ca n b e ex p ressed is as follow s: (1) CA + K + F = 0, w h ere (.'A is th e cu rren t a cc o u n t b alan ce, K is net n o n official cap ital flows an d F is official reserve financing. T h e se item s are defined in su ch a w ay th at th ey m u st sum to zero. Let u s co n sid e r ea ch o f th em in tu rn. Current Account T h e cu rren t a cc o u n t h as two m ajo r co m p o n e n ts. T h e se are th e trad e b a la n ce an d th e seiv ices or “invisi b le s ” b alan ce. T h e form er, w h ich generally gets the m ost atten tio n , is th e d ifference b etw een th e value of goods exp o rted an d th e value o f good s im ported. T h e se exp orts an d im p o rts are o f ph y sical o b jects w h ich , in prin cip le, co u ld b e observed cro ssin g the bord er. In co n tra st, “in v isib les” are services for w h ich in tern atio n al p ay m en ts are m ad e b u t that do not 2Because of measurement errors, the actual accounts add in a “statistical discrepancy” which when included in (1) ensures bal ance. The reason we say that they must balance, however, is not a statement about the accuracy of the statistics. The current and capital account (including official balance) are defined to be equal and opposite. Think of the current account as the excess of income over spending. The capital account is then merely net saving, which is equal to income minus spending. If you measure saving as negative and the excess of income over spending as positive, they will obviously add up to zero. 3 JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS involve th e d irect tran sfer o f a p h y sical p ro d u ct. For exam ple, if a New York sh ip p in g co m p an y w ere to in su re a cargo w ith Lloyds o f London, th e p u rch a se o f th at in su ran ce co n tra ct w ould rep rese n t an invisible im port for th e United States and an invisible exp ort for the United Kingdom . Invisibles take m any different form s. Tw o exam p les are w o rth m entio n in g in ad d ition to su ch financial send ees as in su ran ce an d banking. First, if a n ation has eith e r a ssets or liabilities overseas, th e n et paym ent o f in terest o r dividends is m easu red as an invisible im p ort or export. A positive n et retu rn on foreign assets is co u n te d as an invisible export, b eca u se it g en erates an inflow o f pay m en ts into the eco n om y ju st as an export o f good s d o es. Seco n d , in tern atio n al to u rism is co u n te d as part o f th e invisible co m p o n e n t o f th e cu rren t a cco u n t. If U.S. citizen s spend m ore on over seas trips th an foreigners sp en d on U.S. vacations, it is m easu red as an invisible n et im port in th e U.S. b a la n ce of p aym ents. Non-Official Capital Account T h e cap ital a cc o u n t o f th e b alan ce o f paym en ts m easu res th e ch an g e in n et in d eb ted n ess b etw een th e d o m estic eco n o m y and th e rest o f th e w orld. It is im p o rtan t to get this clear, as th ere is so m etim es co n fu sio n abou t w hat th e cap ital a cco u n t co n ta in s. It d oes n ot involve im p o rts and exp orts o f cap ital goods, su ch as m a ch in e tools an d co m p u ters. T h e se are all p h y sical goods, and th e ir im port an d exp ort are th e re fore co u n te d in th e trad e a cco u n t. T h e cap ital a cco u n t involves th e tran sfer o f fin an cial claim s o f various kinds. T h e se claim s are referred to as “ca p ita l” b e ca u se th ey re p resen t claim s to in terest o r dividend p aym en ts and, in th e ca se o f co m p any shares, do involve ow nership o f u nd erlying real assets. T h e term in ology co m m o n ly u sed to d escrib e the cap ital a cc o u n t is rath er co n fu sin g w h en it is related to th e way in w h ich cap ital a cc o u n t item s are m e a su red. In th e cu rren t a cc o u n t o f th e b a la n ce o f pay m ents, goods leaving th e co u n try is m easu red as a plus item . In th e cap ital acco u n t, how ever, w hat is generally called a cap ital “outflow ” is m easu red as negative. Only th e term in ology h ere is confusing, h o w ever; a cc o u n ts are qu ite logical. W hat w e m ean by a cap ital outflow is th at d o m estic resid en ts are buying foreign assets. In o th e r w ords, th ey are “im p o rtin g ” foreign shares, titles or secu rities. T h u s, all p u rch a ses of foreign goods, secu rities (stocks, bon d s, bills) or any o th e r asset are m easu red as negative (im ports), an d all sales to foreigners are m easu red as positive (exports) irrespective o f w h eth er they are goods sales or asset sales. http://fraser.stlouisfed.org/ 4 Federal Reserve Bank of St. Louis In prin cip le, th e cap ital a cc o u n t o f th e b a la n ce of pay m en ts m easu res th e ch an g e in th e n et asset/ liability p o sitio n b etw een the h o m e eco n o m y an d the rest o f th e w orld. W e say “in p rin cip le ” b ec a u se th ere is o n e re sp ect in w h ich th is is n ot co rrect. T h e capital a cco u n t m easu res th e value o f th e n et flow o f financial in stru m en ts (stocks, bon d s, bills, etc.) th at p asses b e tw een d o m estic and overseas resid en ts. But the ex ter nal in d e b te d n e ss o f an eco n o m y ch a n g es not ju st as a ssets ch an g e h an d s. It also ch an g es as a result of ch an g es in values o f a sse ts th at have n ot ch an g ed h an d s. F o r exam ple, U.S. resid en ts m ay ow n sh ares in Rolls Rovce w h ich rise in value. T h is cap ital gain (or loss) elem en t o f th e extern al asset/liability p o sitio n is n ot m easu red as part o f th e b a la n ce o f paym en ts a cc o u n ts u ntil it is realized by an asset sale. Only th e flow o f fin an cial claim s is in clu d ed .3 Official Balance T h e final item in th e b a la n ce o f p ay m en ts a cc o u n ts is the b a la n ce for official fin an cin g. T h is co m p rises ch a n g es in the official foreign ex ch a n g e reserves o f th e d o m estic eco n om y. T h e se reserves are m ainly claim s against foreign governm ents lor cen tra l banks), for exam ple, Fed h old ings o f D eu tsch e m arks. F o r m ost co u n tries, reserves are h eld as a m e a n s o f in tervening in foreign ex ch a n g e m arkets to su p p o rt th e value o f the d o m estic cu rren cy .4 T h is item is a sp ecial official se c to r co m p o n e n t of th e cap ital a cco u n t. It is treated sep arately for h isto rical rea so n s a sso cia te d w ith the fixed ex ch an g e rate system w h ich o p erated alm ost w orldw ide from W orld W ar II u ntil 1973.3 U nd er a freely floating ex ch a n g e rate regim e, th e official fin a n c ing b a la n ce is alw ays zero. If F in eq u atio n 1 is zero, clearly, CA an d K m u st b e equal an d o f o p p o site sign. U.S. Balance o f Payments Table 1 show s the U.S. b a la n ce o f pay m en ts for 1986. It show s a cu rren t a cco u n t d eficit o f a little over $141 billion. T h e cu rre n t a cc o u n t is m ad e up o f item s 1 and 2. T h e cap ital a cc o u n t su rp lu s o f $117 billio n is show n in lin es 3 an d 4. C hanges in U.S. official reserves are show n in lin e 5. T h ere w as a very sm all fall o f $0,312 billion in 1986 (a p lu s sign in d ica tes a d eclin e in h old ings o f foreign assets). T h is in d ica tes th at th e U.S. au th o rities in terv en ed little d uring 1986 as a w hole. 3Some have claimed that the United States has become a net debtor vis-a-vis the rest of the world. This claim ignores the capital gains on U.S.-owned foreign assets; in reality, the United States is likely still to have positive net external assets. 4See Balbach (1978). 5See Batten and Ott (1983). JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS Table 1 U.S. Balance of Payments: 1986 (millions of dollars)___________________ 1) Merchandise trade 2) Invisibles net Balance on current account (1 plus 2) 3) Change in U.S. assets abroad (increase - ) 4) Change in foreign assets in U.S. (increase + ) Balance on capital account (3 plus 4) 5) Change in U.S. official reserves (increase - ) Statistical discrepancy (1+ 2 + 3 + 4 + 5) $-144,339 +2,987 $-141,352 -96,294 +213,387 +117,093 +312 $23,947 NOTE: The merchandise trade balance is exports minus imports. The invisibles balance is the sum of: net military transactions; net investment income; other service transactions; net remittances, pensions and other transfers; and U.S. government grants (non-military). SOURCE: U.S. Department of Commerce. sp en d on its exports. T h is will p ro d u ce an im b alan ce in its foreign ex ch a n g e m arket.6 A ttem pted sales o f d o m estic cu rre n cy (for foreign cu rren cy) will exceed a ttem p ted p u rch a se s. T h e m arket value o f th e cu r ren cy will fall u ntil th e qu an tity o f th e cur ren cy d e m an d ed is equal to th at su p p lied . At th is point, eith e r th e cu rren t a cc o u n t h as a d ju sted so th at it is no lon ger in deficit; o r th e n et exp ort o f a sse ts (in d uced as a ssets in the co u n try b eca m e ch ea p er, th rou gh d o m estic cu rren cy devaluation, an d th u s m ore attractive to for eigners; an d p rices o f foreign assets b eca m e h ig h er and h e n c e less attractive to U.S. citizen s! is just equal to th e cu rren t a cco u n t deficit. T h u s, th e ex ch a n g e rate will a d ju st to en su re th at th e cu rren t an d cap ital a cc o u n ts are exactly offsetting. T h ere is n o th in g m agical abou t this o u tco m e. T h e en d resu lt is th e sam e for any individual. If y ou sp en d m ore th an y o u r in co m e, y ou m u st b orrow o r sell th e equivalent value o f y o u r assets to cover th e d ifference; if y o u sp en d less th a n y o u r in co m e, y ou m u st inevita bly acq u ire in crea sed claim s on so m eo n e else. Sim i larly, a n atio n that ru n s a cu rre n t a cc o u n t d eficit m u st eith e r borrow from abroad o r sell off som e o f its assets, w h eth er th ese assets are d o m estic o r foreign. Like w ise, a cu rren t a cc o u n t su rp lu s m u st be a sso cia ted w ith eith e r an in crea se o f claim s on foreigners o r a re d u ctio n o f previous borrow ings. A nother im p licatio n o f the defin ition of b a la n ce of pay m en ts is th e follow ing identity: (2) CA = GNP —GDE. T h u s, th e d o m in an t p ictu re is on e o f U.S. resid en ts buying m ore goods and services overseas th a n foreign resid en ts are buying from th e U nited States an d of foreigners in creasin g th e ir net hold ing o f claim s against th e U nited States. N otice, how ever, th at th ere is a fairly large statistical d iscrep an cy . T h e p re se n c e o f th is d iscrep an cy in d i ca tes that th e data do not in clu d e som e trade and/or cap ital flow s. W hile it is im p o ssible to say w h ere the in a ccu ra cies arise, it is often p resu m ed that th e g reat est errors are likely to be in th e cap ital a cco u n t, p ri m arily b ec a u se asset tran sfers are m ore difficult to keep reco rd s on. If th e d ata h ad n o o m issio n s, th en the cu rren t and cap ital a cc o u n ts (includ ing official flows) w ould add to zero. It is n o t obvious at first g lan ce w hy th e cu rre n t and cap ital a cc o u n ts m u st offset ea ch o th e r exactly. W hat w ould h ap p en if they did no t? Su p p o se for exam ple, th at at cu rren t ex ch a n g e rates a co u n try is ru n n in g a cu rren t a cc o u n t d eficit bu t its p la n n e d n et cap ital flows are zero. T h is m eans that th e co u n try is trying to sp en d m ore on im p orts th an foreigners are w illing to T h e cu rren t a cc o u n t su rp lu s (or m inu s th e cu rren t a cc o u n t deficit) is equal to gross n atio n al p ro d u ct m inu s gross d o m estic ex p en d itu re. T h is identity show s th at th e cu rren t a cc o u n t o f th e b a la n ce of p aym en ts is the difference b etw een th e value o f w hat th e n ation p ro d u ces an d w h at it sp en d s. T h e form er (GNP) ca n also be th ou gh t o f as th e value o f the n a tio n ’s gross in co m e. Id entity (2) is useful b eca u se it m akes cle a r th at any n a tio n th at sp en d s m ore th an it p ro d u ces will have a trad e deficit. T h e in terestin g qu estion , o f co u rse, is w h eth er su ch an im b alan ce is good or bad. WHAT MAKES THE CURRENT ACCOUNT BALANCE A PRO BLEM ? T h e n atu re o f w hat is u sually term ed a b a la n ce of paym en ts problem varies co n sid erab ly, d ep en d in g 6See Chrystal (1984). 5 FEDERAL RESERVE BANK OF ST. LOUIS upon w h eth er th e co u n try in q u estio n has a fixed o r a floating exch an ge rate regim e. T h e p roblem p ro d u ced by a deficit on the cu rren t a cc o u n t can be m ost a cu te if th e n ation is m ain tain in g a fixed ex ch an g e rate re gim e.7 In th is case, "th e p ro b lem " is felt d irectly by the cen tral bank. M aintaining a fixed ex ch an g e rate vis-a-vis on e or m ore co u n tries requ ires th e pegging nation s cen tral ban k to h o ld foreign ex ch an g e reserves w ith w h ich to intervene in the foreign ex ch an g e m arket. T h is in ter v en tion can be n ecessary to stop th e ex ch an g e rate from m oving in eith e r d irectio n . Su ppose, for exam ple, that th e co u n tiy has a cu rren t a cc o u n t d eficit an d no d esired n et private cap ital flow s. In o rd er to m ain tain th e existing ex ch an g e rate, th e cen tral bank m u st sell foreign ex ch an g e for its d o m estic cu rren cy . W h eth er the origin or so u rce o f th e n et supply o f d o m estic cu rre n cy in foreign ex ch an g e m arkets is from the cu rren t or cap ital a cc o u n t side o f the b alan ce o f pay m ents is irrelevant. T h e d o m estic cu rre n cy value of reserves sold in a p articu lar period is th e official fin an cin g b alan ce, F, in eq u atio n 1. B eca u se it involves the sale to foreigners o f a d o m estically h eld asset, a net loss of reserves is m easu red as positive in th e b a la n ce o f p aym ents a cco u n ts. U nd er a fixed exch an ge rate regim e, ex ch an g e rate p ressu re p o ses a p roblem if th e cen tral bank in q u e s tion starts to run ou t o f foreign ex ch an g e reserves. T h is possibility m akes th e p roblem w orse b eca u se hold ers o f th e d o m estic cu rren cy , fearing a devalua tion, will tiy to bu y foreign cu rren cy . Speculative sales of th e d o m estic cu rre n cy in foreign ex ch an g e m arkets force th e cen tral bank to sell even m o re foreign ex ch an g e reserves. Inevitably, th e n atio n m u st eith e r devalue its cu rre n cy o r in tro d u ce m easu res to cu t d o m estic sp en d in g (includ ing sp en d in g on foreign goods). T h is actio n is u navoidable; o th erw ise, th e c e n tral bank will run out o f foreign ex ch an g e reserves. T h is d escrib es th e n atu re o f m ost b a la n ce o f pay m ents crise s ex p erien ced bv co u n tries attem p ting to m aintain fixed ex ch an g e rates in th e 1950s an d 1960s. It is w orth noting, how ever, that th e U nited States u n d e r th e p ostw ar "B retto n W o o d s” regim e w as n ot th e sam e as o th e r co u n tries.8All o th e r co u n tries in the system pegged th eir cu rre n cies to the d ollar an d h eld The exception to this is when a currency is depreciating at a fast rate. This is a symptom of acute internal problems normally associ ated with hyperinflation. 8The system was named after the place in New Hampshire where the final negotiations setting it up were held in July 1944. http://fraser.stlouisfed.org/ 6 Federal Reserve Bank of St. Louis JANUARY/FEBRUARY 1988 d ollar reserves for th is p u rp o se. T h e lin ited States, th erefore, did n o t n eed to su p p o rt its ow n ex ch an g e rate and, in fact, did not h old significant reserves of foreign cu rre n cy d uring th is period .0 S in ce th e spring of 1973, w h en all the m a jo r in d u s trial co u n tries m oved to a floating ex ch a n g e rate re gim e (the United K ingdom h ad floated in Ju n e 1972), th e n atu re o f b a la n ce o f p aym en ts pro b lem s has ch a n g ed .10 U nder a floating ex ch a n g e rate system , a cen tral ban k d o es not have to u se its foreign ex ch an g e reserves to fin a n ce a d eficit in th e n on -official part o f th e b a la n ce o f p ay m en ts; in fact, th ere w ill b e n o n e ." In eq u ation 1 above, the term F b e co m e s zero. In stead o f cen tra l ban k in tervention, th e ex ch an g e rate m oves to assu re th at th e cu rren t a cc o u n t an d th e capital a cc o u n t su m to zero on th e ir own. WHY WORRY ABOUT THE TRADE BALANCE? C o n cern about the state o f the trade b a la n ce lias a long h isto iy . It is useful to put this c o n c e rn in h isto ri ca l co n text, as it lead s n atu rally to th e analysis o f w h en su ch co n c e rn is justified. In th e follow ing d iscu ssio n , w e take it as given that trade itself is ben eficial, a p o in t not clearly esta b lish ed until R ica rd o ’s fam ous d em o n stratio n p u b lish ed in 1817. T h ere w as, how ever, so m e co n n e ctio n h isto ri cally betw een the ca se against trade d e fic it s an d th e u n d erstan d in g o f w hy trad e in gen eral w as a good thing. Only w h en the gains from trad e w ere properly u n d ersto o d cou ld p eop le begin to m ake sen sib le a s sessm en ts o f th e ca u se an d effect o f trad e d eficits. T h e co n tex t in w h ich th e early d eb ates took p lace w as an in tern atio n al eco n o m y in w h ich p ay m en ts for extern al trad e w ere largely m ad e in p recio u s m etals, esp ecially gold. T h e effect o f ru n n in g a trad e su rplu s w as th at a n atio n w ould a ccu m u la te gold. In m any 9The U.S. authorities agreed to convert dollars into gold at $35 per ounce. This commitment was abandoned for all but official holders in March 1968 and for official holders in August 1971. See Batten and Ott (1983) for evidence on exchange market intervention. ,0Note that even today the majority of small countries peg their exchange rates to either a major currency or a weighted basket of currencies. Reserve shortages still may cause acute problems for them. 11In fact, none of the major currencies are floating freely. All the major central banks have intervened from time to time to influence ex change rates. Intervention to support the dollar has been especially heavy since the “ Plaza Accord” of September 1985. FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 p e o p le ’s m ind s, the accu m u latio n o f gold itself b e cam e th e o b je ct o f trad e: trad e su rp lu ses w ere ‘g o o d ” and trad e d eficits “b a d .” T radin g in o rd er to bu ild up gold h old ings b eca m e know n as m ercan tilism . flation .13 In d eed , it w as a perio d o f rapid an d p ro longed e c o n o m ic grow th. T h ere is th u s at least one co u n te rex a m p le — an d a m a jo r on e — to R ica rd o ’s gen eralization . How can th is be exp lain ed ? M ercan tilism w as criticized by several em in en t w rit ers, in clu d in g David Hum e (1752), w ho sh ow ed th at a co n tin u in g trade su rp lu s w as u n attain able. An e x ist ing trad e su rplu s, he n oted , p ro d u ces an inflow of gold. B ecau se gold is a form of m oney, th e qu antity of m o n ey in th e co u n try rises. This, in turn, p ro d u ces a rise in p rices, w h ich co n tin u es as long as m o re gold flows in. As th e co u n try ’s goods b eco m e m ore e x p e n sive relative to th o se p ro d u ced overseas, how ever, few er will be bought, eventually elim in atin g the trade su rp lu s.12 At th e tim e R icardo w as w riting, h is claim w as d isp uted , m ost notably, bv H eniy T h o rn to n . T h o rn to n argued that, alth ou g h p rio r excessive m o n ey ex p a n sion w as in d eed su fficien t to p ro d u ce a trad e deficit, it w as n o t a n ecessa ry co n d itio n for a trad e deficit. Som e y ears later, David R icardo (1817) u sed th is d em o n stratio n to sh o w w hy trade d eficits o ccu rred . His an sw er to this q u estio n brings us d irectly to ou r cen tra l point: trade d eficits can resu lt from a variety of so u rces, not all o f w h ich are “b a d .” RICARDO, THORNTON AND TRADE DEFICITS'* R icardo argued that a trade d eficit w as the inevita ble co n se q u e n ce o f p rices in th e d eficit co u n try bein g "too h ig h .” T h e se p rices, in tu rn, w ere p ro d u ced by excessive p rio r m o n etary ex p an sio n from d o m estic so u rces th at w ere u nrelated to p rio r trad e su rp lu ses.14 He argued, in o th e r w ords, th at excessive m o n etary ex p an sio n w as n o t only a su fficient co n d itio n for a trade d eficit to o ccu r, it w as also a n e c e ssa iy co n d itio n an d vice versa for trade su rp lu ses. T his d escrib es w hat h ap p e n ed in m any co u n tries d uring th e B retto n W oods regim e. W hile th is se q u en ce o f events portrays a co m m o n ca u se o f trade deficits, how ever, it is not th e only cau se. In th e 65y ea r perio d b etw een 1830 and 1895, th e U nited States h ad a cu rren t a cc o u n t d eficit in alm ost eveiy year; th ere w ere only 13 y ears in w h ich a su rp lu s was reco rd ed . Yet this w as not a period o f su stain ed in ,2Hume, although dealing explicitly with the mercantilist argument, dealt implicitly with the notion that an export surplus is necessary for growth. Since a perpetual export surplus is impossible, if an export surplus were essential for growth, growth would have stopped. It did not, however, and to date has not. ,3An extensive discussion of the ground covered in this section can be found in Perlman (1986). ,4The issue of domestic bank notes partially backed by gold was a topic of controversy between the “ currency” and “banking” schools through the 19th century in Britain. T h o rn to n d istin g u ish ed b etw een trade d eficits a ris ing from real ca u ses an d th o se arising from excessive m o n ey creatio n . T h e form er ca n o c c u r b ec a u se in d i viduals in a co u n try w an t to sp en d m ore th a n th e ir cu rre n t in co m e, that is, th ey w ish to re d u ce th e ir net fin an cial w ealth or in cre a s e th e ir n et in d e b te d n e ss.16 In term s o f eq u atio n 2 above, an yth in g th at ca u ses d o m estic sp en d in g to ex ceed o u tp u t w ill p ro d u ce a trade deficit. Of co u rse, th e b a la n ce o f p ay m en ts d eficit from this ca u se ca n n o t p ersist forever. It will d isap p ear w h en individuals have re a ch ed th e ir n ew lo w er d esired w ealth level; in th e sam e m an n er, a trade d eficit p ro d u ced by ex cess m o n ey crea tio n w ill en d w h en the e x c e ss m o n ey h a s b e e n d isp e rse d o v erseas (or deflated by h ig h er p rice s).17 In sum m ary, a trad e d eficit ca n b e p ro d u ced n ot ju st by ex cess m o n etary exp an sio n , bu t by dissaving.'8 Both o f th ese will p ro d u ce d eficits that are tem porary; how ever, th e se d eficits will b e elim in ated eventually by different m e ch a n ism s. D issaving and th e a sso ci ated d eclin e in fin an cial w ealth can be p ro d u ced by several factors; exam in in g so m e m ajo r o n es h elp s to u n d ersta n d the cu rre n t U.S. situ ation . ,5For more details on this, see Mudd and Wood (1978) and Friedman and Schwartz (1963). "This highlights the fact that a trade deficit can be a symptom of a problem, but is not itself a problem. Alternatively, it may be a symptom of something that is not a problem at all. l7Note that, when we talk about a “ lower desired wealth level," we are referring only to financial wealth. If financial assets are being con verted into physical capital, the composition rather than the level of wealth is changing. If the physical capital offers a greater rate of return than financial assets, this change actually will increase peo ple’s wealth. This distinction is central to the argument that a trade deficit associated with high levels of domestic real investment could lead to faster real growth, increased wealth and higher output in the future. 18Monetary expansion need not always lead to a trade deficit. In a classic paper, Robert Mundell (1963) showed that, with perfect capital mobility, floating exchange rates and sticky goods prices, monetary expansion causes capital outflows (purchases of foreign assets). This causes the currency to depreciate and results in a current account surplus. Similar results are found in the modern “overshooting” literature. 7 FEDERAL RESERVE BANK OF ST. LOUIS WHY SHOULD TH ERE BE DISSAVING? In o rd er to d iscu ss th e p o ssib le so u rces o f dissaving in th e d o m estic econ om y, it is co n v en ien t to set out a n o th e r id en tity :1'1 13! CA = ( S - I ) + ( T - G l . T his show s th at th e cu rren t a cc o u n t su rp lu s m u st be equal to th e ex cess o f private saving over private in vestm en t (S —I), plus th e govern m ent bu dget su rplu s (T —G). In o th e r w ords, the su rp lu s for th e eco n o m y as a w h ole can be broken dow n in to th e private se c to r su rplu s plus th e p u blic se c to r su rp lu s. T h is cla ssifica tio n suggests p o ssib le d irectio n s in w h ich to look for ca u ses o f th e trade deficit: a fall in private saving, a rise in private investm ent o r an in crea se in th e governm ent budget deficit. A fall in private saving m u st be a sso ciate d w ith an in crease in co n su m p tio n relative to in co m e. T his cou ld h ap p en if th ere w ere a tem porary fall in in co m e due, for exam ple, to a crop failure o r a natu ral d isaster. It is w ell estab lish ed that, at tim es w h en in co m e is abn orm ally low, p eop le attem p t to m aintain th eir c o n su m p tio n p attern s by dissaving. If th e n atio n as a w h ole d o es this, it will n ecessarily involve a trade deficit. It sh o u ld b e em p h asized that, w hile crop fail ures or o th e r n atu ral d isasters are u nfo rtu n ate, the ability to a d ju st to th e se events by dissaving an d th u s im portin g good s from abroad is preferable to red u cin g d o m estic co n su m p tio n . In extrem e cases, th e ch o ice m ay be b etw een ru n n in g a trad e d eficit an d starvation. W hile natu ral d isasters can explain som e trade d e ficits, it is unlikely to explain the U.S. d eficits in the 1980s. After all, this h as b een a period o f fairly steady in co m e grow th. T h e seco n d alternative su ggested by id entity (3) is a rise in private investm ent, cau sed by an ex p ected rise in th e productivity o f d o m estic cap ital (relative to th at overseas). T his alternative is an extrem ely h ealth y sign for th e d o m estic eco n om y. It in d icates th at th e ex p e cted profitability o f in vestm en t w as su ch th at firm s w ere p rep ared to borrow in o rd er to fin a n ce the h ig h er investm ent. If private investm en t ex ceed s p ri vate saving (for a b a la n ced governm ent budget), the private se c to r m u st borrow from overseas. W e have seen alread y th at n et borrow ing from overseas im plies a cu rren t a cc o u n t d eficit in th e b alan ce o f paym ents. ,9This can be derived as follows: GNP = C + I + G + CA from the expenditure accounts. It is also true that GNP = C + S + T from the income accounts. Sol + G + CA = S + T and CA = (S - l) + (T -G ). http://fraser.stlouisfed.org/ 8 Federal Reserve Bank of St. Louis JANUARY/FEBRUARY 1988 If ov erseas-fin an ced grow th in private in v estm en t lies b eh in d the trade deficit, w e have to be carefu l in in terp retin g th e sta tem e n t th at th e trad e d eficit is a sso cia ted w ith dissaving o r a red u ctio n in w ealth . It is true that the private se c to r will be in creasin g its net fin an cial liabilities (or red u cin g n et fin an cial assets). At th e sam e tim e, how ever, it is converting th o se liabili ties into real cap ital. T h e retu rn on that real cap ital is ex p ected to b e g reater th a n th e co st o f th e borrow ing. H ence, th is provides th e b asis for in co m e an d w ealth grow th in th e future and, presum ably, exp lain s w hy th e U n ited S ta te s h ad s u s ta in e d tra d e d eficits th rou gh o u t th e seco n d h alf o f th e 19th cen tu ry. R ap idly grow ing co u n tries th at attract cap ital from over seas typically will have trade deficits. T h e final p ossibility is th at th e cu rren t a cco u n t deficit reflects th e governm ent budget deficit. Obvi ously, if private saving an d in vestm en t w ere equal, th e bu dget d eficit an d th e cu rre n t a cc o u n t d eficit w ould be equal. We shall not p u rsu e th e q u estio n o f w h eth er th e bu dget d eficit is “g o o d ” o r “b a d ” for th e eco n om y. A ssum ing th at th e bu dget d eficit re p rese n ts th e d elib erate ch o ice o f policym akers, how ever, it follow s th at th e a sso cia te d trad e d eficit m u st be preferred to th e alternatives. T h u s, w e have seen th at a rising cu rren t a cc o u n t deficit m u st b e a sso cia te d w ith eith e r a rise in invest m en t relative to saving (or fall in saving relative to investm ent) or a rise in th e budget d eficit o f th e gov ern m en t. W e alread y h ad see n th at cu rren t a cc o u n t d eficits co u ld resu lt from excessive m o n etary ex p a n sion, a ca se th at is co n siste n t w ith id en tity (3): th e attem p t to sp en d th e ex cess m o n ey will resu lt in eith e r a fall in S —I (higher co n su m p tio n , low er saving or h ig h er investm ent) or a fall in T —G (m ore govern m ent sp en d in g relative to taxes). The Evidence f o r the United States We n ow look at th e p o ssib le ca u ses o f th e U.S. cu rren t a cc o u n t deficit. First, w e co n sid e r th e argu m ent, favored by Ricardo, o f fast m o n etary grow th a sso cia te d w ith high d o m estic inflation. At first sight, th is ap p ears a likely possibility. M on etary grow th a c celera ted after 1982 (chart 1) at th e sam e tim e as th e cu rren t a cc o u n t p lu n g ed into d eficit (chart 2). How ever, U.S. inflation fell (chart 3) an d rem a in ed c o n s is t ently below th e OECD average d uring th is p eriod . Also, b o th the real and effective ex ch a n g e rates a p p reciated strongly u ntil 1985. T h e in flation an d ex ch a n g e rate behavior are signs o f m o n etary tigh tn ess, n o t m o n e- JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS C hart 1 U.S. M o n e y G r o w th , Effective a n d Real Exchange Rates Percent Index N O T E : The e f f e c t i v e e x c h a n g e r a t e is th e F e d e r a l R e s e r v e B o a r d o f G o v e r n o r s ’ t r a d e - w e i g h t e d e x c h a n g e r a t e , a w e i g h t e d i n d e x (1973=100] o f t h e d o l l a r ' s v a l u e in t e r m s o f 10 i n d u s t r i a l c o u n t r y c u r r e n c i e s . The r e a l e f f e c t i v e e x c h a n g e r a t e is o b ta in e d b y d i v i d i n g th e n o m i n a l e f f e c t i v e e x c h a n g e r a t e b y the r a t i o o f c o n s u m e r p r i c e i n d e x e s (CPI] o f th e 10 i n d u s t r i a l c o u n t r i e s , ( t r a d e - w e i g h t e d , t h e s a m e a s t h e e x c h a n g e r a te s ] to t h e CPI o f th e U n i t e d S t a t e s ; a l l CPIs a r e i n d e x e d t o 1 9 7 3 , 1973 =1 00. taiy ease.2" Only the high U.S. m on ey grow th in 1986 looks co n siste n t w ith R icard o ’s ex p lan ation : b o th the real an d n o m in al ex ch an g e rates fell during 1986. T h e in crea se in th e trad e deficit in 1986, how ever, w as sm all. H ence, little w eight can be a tta ch e d to the m o n etary exp lan ation o f th e trade deficit. In d eed , w hy th e rapid m oney grow th o f 1 9 8 2 -8 6 did not crea te 20lt is possible that the Mundell model referred to above is relevant here. This predicts that monetary tightness causes capital inflows, a currency appreciation and a current account deficit. We think this unlikely to be relevant here. There is no clear evidence of sufficient monetary tightening over the entire 1981-86 period to explain what happened. More importantly the same outcome is predicted from the Mundell analysis as resulting from fiscal expansion. Hence mone tary neutrality combined with fiscal expansion would be sufficient. It is the latter which seems to us to dominate in this case. inflation is still so m eth in g o f a m ystery. T h e re was, over th is period, a sign ifican t d eclin e in th e velocity of circu latio n , w h ich m ean s th at th e extra m o n ey b a l a n ce s w ere w illingly h eld ra th er th a n sp en t d o m esti cally.-' A m u ch m ore plau sib le sto iy em erg es from a plot o f th e private an d p u b lic se c to r su rp lu ses (chart 2). No tice th at w e sh o w h ere I —S ra th er th a n S —I, b ec a u se it is ea sier to see its co rre sp o n d e n c e w ith T —G. Before 1982, th e relatio n sh ip b etw een th e p u blic se c to r d e ficit and th e private se c to r su rp lu s w as rem arkably clo se. As a result, cu rren t a cc o u n t d eficits an d su r p lu ses generally w ere sm all. After 1982, how ever, the 21See Stone and Thornton (1987). 9 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 C h a rt 2 Relationships B e tw ee n Public a n d Private Sector Deficits a n d the Current Account Billions of dollars p u b lic se c to r d eficit stayed high w hile private invest m ent rose relative to private saving. Bv 1986, th e p ri vate se c to r invested in ex cess o f its saving. H ence, the co n tin u ed p u blic se c to r d eficit is n ecessarily m atch ed by a cu rren t a cc o u n t d eficit o f equivalent size. Insofar as th e governm ent bu dget d eficit is taken as given, the ch o ice s th at the U.S. faces are high levels o f private in vestm ent and a trade deficit or b alan ced trade and slow real grow th. Billions of dollars general, a trade d eficit m u st b e a sso cia ted w ith som e co m b in atio n o f private an d p u b lic se c to r d eficits. Until 1982, the budget d eficit w as approxim ately fin an ced by private se c to r su rp lu ses. T h e p resen t situation, how ever, is th e inevitable resu lt o f th e co m b in a tio n of a budget d eficit an d high in v estm en t relative to private saving. REFERENCES CONCLUSION Balbach, Anatol B. “The Mechanics of Intervention in Exchange Markets,” this Review (February 1978), pp. 2-7. A trade deficit arises w h en a co u n try buys m ore from overseas th an foreigners buy from it. T h e co u n terpart o f a trad e d eficit in the b alan ce o f paym en ts a cc o u n ts is an in crease in borrow ings (or red u ctio n in net lending) from th e rest o f th e w orld. T rad e deficits cou ld resu lt from inflationary d o m estic m o n etary pol icies; th ere is no evidence, how ever, th at su ch policies are th e cau se o f the cu rren t U.S. trade deficit. In Batten, Dallas S., and Mack Ott. “What Can Central Banks Do About the Value of the Dollar?” this Review (May 1984), pp. 5-15. http://fraser.stlouisfed.org/ 10 Federal Reserve Bank of St. Louis Chrystal, Alec K. “A Guide to Foreign Exchange Markets,” this Review (March 1984), pp. 5-18. Friedman, Milton, and Anna Jacobson Schwartz. A Monetary His tory of the United States 1867-1960 (Princeton University Press, 1963). Hume, David. Political Discourses (Edinburgh, 1752). JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS C hart 3 U.S. a n d OECD Inflation 1968 70 72 74 76 78 80 82 84 1986 Mudd, Douglas R., and Geoffrey E. Wood. “The Recent U.S. Trade Deficit — No Cause for Panic,” this Review (April 1978), pp. 2-7. Perlman, Morris. “The Bullionist Controversy Revisited,” Journal of Political Economy (August 1986), pp. 745-62. Mundell, Robert. "Capital Mobility and Stabilization Policy Under Fixed and Flexible Exchange Rates,” Canadian Journal of Eco nomics and Political Science (November 1963), pp. 475-85. Stone, Courtenay C., and Daniel L. Thornton. “Solving the 1980's Velocity Puzzle: A Progress Report,” this Review (August/Septem ber 1987), pp. 5-23. Ricardo, David. The Principles of Political Economy and Taxation (London: J. M. Dent & Sons, Ltd., reprinted 1948). 11 JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS Protectionist Trade Policies: A Survey of Theory; Evidence and Rationale Cletus C. Coughlin, K. Alec Chrystal and Geoffrey E. W ood JL ROTECTIONIST p ressu res have b een m ou n tin g w orldw ide d uring th e 1980s. T h e se p ressu res are due to various ec o n o m ic p ro blem s in clu d in g th e large and p ersisten t b alan ce o f trade d eficits in th e United States; th e h ard tim es ex p e rie n ce d by several in d u s tries; an d th e slow grow th o f m any foreign co u n tries.' P ro p o n en ts o f p ro tectio n ist trade p o licies argue that in tern atio n al trade h as co n trib u ted su bstan tially to th e se p ro blem s an d th at p ro tectio n ist trad e p o licies will lead to im proved resu lts. P rofessional eco n o m ists in th e U nited States, how ever, gen erally agree that trad e re strictio n s su ch as tariffs and qu o tas su b sta n tially re d u ce a n a tio n ’s e co n o m ic w ell-being.2 T h is article surveys th e theory, evid en ce an d ra tio n ale co n cern in g p ro tectio n ist trad e p o licies. T h e first sectio n illu strates th e gains from free trade using th e co n c e p t o f com p arative advantage. R ecen t devel o p m en ts in in tern atio n al trad e th eory th at em p h asize o th e r reason s for gains from trade are also reviewed. T h e th eo retical d iscu ssio n is follow ed bv an ex am i n ation o f re cen t em p irical stud ies that d em o n strate th e large co sts o f p ro tectio n ist trade p o licies. T h en, th e ration ale for restrictin g trade is p resen ted . I’he co n clu d in g sectio n su m m arizes th e p a p e r’s m ain argum ents. Cletus C. Coughlin is a senior economist at the Federal Resen/e Bank of St. Louis. K. Alec Chrystal is the National Westminster Bank Profes sor of Personal Finance at City University, London. Geoffrey E. Wood is a professor of economics at City University, London. This article ivas written while Chrystal was a professor of economics at the University of Sheffield, Sheffield, England. Thomas A. Pollmann provided research assistance. 'See Page (1987) for a detailed examination of trade protectionism since 1974. 2 This consensus was found in a survey published in the late 1970s (Kearl et at., 1979). Recent developments in international trade theory, which can be used to justify governmental intervention in trade policy, have not altered the consensus (Krugman, 1987). http://fraser.stlouisfed.org/ 12 Federal Reserve Bank of St. Louis THE GAINS FROM FR E E TRADE T h e m o st fam ou s d em o n stra tio n o f th e gains from trade a p p eared in 1817 in David R icard o ’s P rin c ip le s o f P o litic a l E c o n o m y a n d T ax a tion . W e u se h is exam ple involving trade b etw een E n glan d an d Portugal to d em o n strate how b o th co u n tries ca n gain from trade. T h e two co u n tries p ro d u ce the sam e tw o goods, w ine and clo th , an d th e only p ro d u ctio n co sts are la b o r co sts. T h e figures below list th e am o u n t o f la b o r (e.g., w orker-days) requ ired in e a c h co u n tiy to p ro d u ce one bo ttle o f w in e o r o n e b o lt o f clo th . E ngland Portugal W ine Cloth 3 1 7 5 Sin ce b o th good s are m ore co stly to p ro d u ce in E n gland th a n in Portugal, E n glan d is ab so lu tely less efficient at p ro d u cin g b o th good s th an its p rospective tradin g p artn er. Portugal h a s an ab so lu te advantage in b o th w in e an d clo th . At first g lan ce, th is ap p ears to rule out m u tu al gains from trade; how ever, as w e d em o n strate below , ab so lu te advantage is irrelevant in d iscern in g w h eth er trad e ca n ben efit b o th co u n tries. T h e ratio o f th e p ro d u ctio n co sts for th e two goods is different in th e tw o co u n tries. In England, a b o ttle of w ine will ex ch an g e for 3/7 o f a b o lt o f clo th b e c a u se th e lab or co n ten t o f th e w in e is 3/7 o f that for clo th . In Portugal, a b o ttle o f w in e will ex ch a n g e for 1/5 o f a bolt o f clo th . T h u s, w ine is relatively c h e a p e r in Portugal th an in England and, conversely, clo th is relatively c h e a p e r in England th an in Portugal. T h e exam ple in d icates that Portugal h as a com p arative advantage in w ine p ro d u ctio n an d E ngland h as a co m p arative ad vantage in clo th p ro d u ctio n . T h e different relative p rices provide th e b asis for JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS both co u n tries to gain from in tern atio n al trade. T he gains arise from b o th ex ch an g e an d sp ecializatio n . T h e gains from e x c h a n g e can be high lighted in the follow ing m an n er. If a P ortu gu ese w in e p ro d u c e r sells five b o ttles o f w ine at h om e, h e receives on e b o lt of clo th . If he trades in England, he receives m ore than two bolts of clo th . H ence, h e can gain by exp o rtin g his w ine to England. E nglish clo th -p ro d u cers are w illing to trade in Portugal; for eveiy 3/7 o f a bolt o f clo th they sell th ere, they get ju st over two b o ttles o f w ine. T h e English gain from exportin g clo th to (and im porting w ine from) Portugal, an d th e P ortu gu ese gain from exportin g w ine to (and im p ortin g clo th from) England. E ach co u n try gains bv exp orting th e good in w h ich it has a com p arative advantage an d by im portin g th e good in w h ich it h as a com p arative disadvantage. (rains from s p e c ia liz a tio n can be d em o n strated in th e follow ing m an n er. Initially, e a c h co u n try is p ro d u cin g so m e o f b o th good s. Su p p o se that, as a resu lt of trade, 21 u n its o f lab or are sh ifted from w ine to clo th p ro d u ctio n in England, w hile, in Portugal, 10 u n its of lab or are shifted from clo th to w ine p ro d u ctio n . T his reallo catio n o f lab or d oes n o t alter th e to tal am o u n t o f lab or u sed in th e tw o co u n tries; how ever, it cau ses the p ro d u ctio n ch an g es listed below . B ottles o f W ine England Portugal Net - 7 Bolts o f Cloth + 3 + 10 - + 3 + 1 2 T h e shift o f 21 u n its o f lab or to th e English clo th in d u stry raises clo th p ro d u ctio n by th ree bolts, w hile red u cin g w in e p ro d u ctio n by seven bo ttles. In P ortu gal, th e shift o f 10 u nits o f lab or from clo th to w ine raises w ine p ro d u ctio n by 10 bo ttles, w hile red u cin g clo th p ro d u ctio n by tw o bo lts. T h is reallo catio n of lab or in cre a se s th e total p ro d u ctio n o f b o th goods: w ine by th ree bo ttles and clo th by on e bolt. This in crea sed o u tp u t will be sh ared by th e tw o co u n tries. T h u s, th e co n su m p tio n o f b o th good s and th e w ealth o f both co u n tries are in crea sed by the sp e cia liz a tion brou ght abou t by trade b ased on com p arative advantage. lab or w as th e only re so u rce u sed to p ro d u ce th e two good s in th e exam p le above; y et, la b o r is really only o n e o f m any re so u rces u sed to p ro d u ce good s. T h e exam p le also a ssu m ed th at th e co sts o f pro d u cin g ad d ition al u n its o f th e good s are co n sta n t. F or ex a m ple, in England, th ree u n its o f lab or are u sed to p ro d u ce on e bo ttle o f w ine regard less o f th e level o f w ine p ro d u ctio n . In reality, u nit p ro d u ctio n co sts cou ld eith e r in crea se o r d ecrea se as m o re is p ro d u ced . A third a ssu m p tio n w as th at th e goods are p ro d u ced in p erfectly com p etitiv e m arkets. In o th e r w ords, an in d i vidual firm has no effect on th e p rice o f th e good that it p ro d u ces. Som e in d u stries, how ever, are d om in ated by a sm all n u m b er o f firm s, e a c h o f w h ich ca n affect the m arket p rice o f the good by altering its p ro d u ctio n d ecisio n . Som e o f th e se ex ten sio n s are d iscu sse d in th e append ix. T h e se th e o retica l d ev elo p m en ts g en erally have stren g th en ed th e ca se for an o p en tradin g system . T h ey suggest th ree so u rce s o f gains from trade. First, as th e m arket poten tially served by firm s exp an d s from a n ation al to a w orld m arket, th ere are gains a sso cia te d w ith d eclin in g p e r u nit p ro d u ctio n co sts. A seco n d so u rce o f gains resu lts from th e re d u ctio n in th e m on o p o ly p o w er o f d o m estic firm s. D om estic firm s, facing m ore p re ssu re from foreign co m p etito rs, are forced to p ro d u ce th e o u tp u t d em an d ed by c o n su m ers at th e low est p o ssib le co st.3 T h ird is th e gain to co n su m ers from in crea sed p ro d u ct variety an d low er p rices. G enerally speaking, th e gains from trade result from the in cre a se in com p etitiv e p ressu res as the d o m estic eco n o m y b e co m e s less in su lated from the w orld econ om y. T h e gains from free trad e ca n also b e illu strated graphically. T h e sh ad ed in sert on pages 14 an d 15 exam in es th e gains from trade in perfectly com p etitive m arkets u sin g su p p ly an d d em an d analysis. T h e in sert also analyzes th e effects o f trad e restrictio n s, a to p ic th at w e d iscu ss below . COSTS OF TRADE PROTECTIONISM P ro tectio n ist trade p o licies can take n u m ero u s form s, so m e o f w h ich are d iscu sse d in th e sh ad ed in sert on pages 16 an d 17. All form s o f p ro tectio n are TRADE THEORY SINCE RICARDO Sin ce 1817, n u m ero u s analyses have g en erated in sights co n cern in g th e gains from trade. They chiefly exam in e th e co n se q u e n ce s o f relaxing th e a ssu m p tio n s u sed in th e p reced in g exam p le. F or exam ple, 3A profit-maximizing firm produces its output at minimum cost. When firms are insulated from competition, costs are not necessarily being minimized. This situation, which is called X-efficiency, has been stressed by Leibenstein (1980). The increase in competitive pres sures due to international trade reduces the probability that costs are not minimized. 13 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 A Supply and Demand Analysis of the Gains from Free Trade and the Effects of a Tariff A stand ard illu stration o f th e gains from free trade an d th e eifects o f a tariff is p resen ted below . T h e analysis assu m es perfectly com p etitive m ar kets th rou gh ou t. Gains From Free Trade In figure 1 th e lines Sus and Dus are th e U.S. supply a n d d em an d curves for a h y p oth etical good. T h e ir in tersectio n at B resu lts in th e equilibriu m values for p rice and qu antity o f Pu, and Q us. A ssum ing th e United States h as a com p arative disadvantage in th e p ro d u ctio n o f th is good, th e p rice will be low er abroad th an in th e U nited States. Let this low er w orld p rice be Pw, and assu m e that U.S. p u rch a ses do n o t affect th is w orld price. G raphically, this is re p rese n te d bv th e h o rizo n tal w orld su pply curve S„ . If o n e allow s for free trade, th is lo w er w orld p rice h as tw o effects. First, U.S. co n su m ers will in crea se th e ir co n su m p tio n to D 'us. Second , U.S. p ro d u cers will co n tra ct th eir p ro d u ctio n to S 'us. T h e e x cess o f U.S. c o n su m p tio n o v erp ro d u ctio n is U.S. p u rch a ses from foreign p ro d u cers (that is, im ports). T h e low er p rice sim u ltan eo u sly ben efits U.S. c o n su m ers an d h arm s U.S. p ro d u cers, a fact that u n d erlies th e re ce n t controversial d iscu ssio n s o f U.S. F ig u re 1 http://fraser.stlouisfed.org/ 14 Federal Reserve Bank of St. Louis trade policy. T h e m agn itu d e o f th e se gains and lo sses u sin g the co n c e p ts o f c o n su m e r an d p ro d u ce r su rp lu s ca n b e se e n in figure 1. C on su m ers gain in two w ays. Initially, co n su m e rs p u rch a sed Q uk at a p rice p er unit o f Pus. W ith free trade, they p u rch a se Q us at th e low er p rice p e r unit o f P„. T his gain is rep rese n te d by th e rectan g le Pus BE P„. In ad d ition, th e lo w er p rice in d u ce s co n su m e rs to in crea se th e ir p u rch a se s from Q„s to D 'us. T h is gain is re p resen ted by th e triangle BCE. T h e total gain to co n su m ers is Pus BC P„ or, u sin g th e low er case letters to rep resen t areas, a + b + c. Analogously, p ro d u cers lo se due to the low er p rice th ey receive for th e ir o utput, S 'us, an d d ue to th e ir c o n tra ctio n o f p ro d u ctio n from Q„s to S 'us. T h e total loss to p ro d u cers is Pus BF P,v o r a. T h e n atio n as a w h o le gains b e c a u se th e c o n su m er gains o f a + b + c ex ceed th e p ro d u cer lo sses o f a by b + c. T h is analysis ca n also b e viewed u sing a good th at th e U nited States exp orts. In o th e r w ords, th e U nited States w ill have a com p arative advantage in th e p ro d u ctio n o f a good. F o r the exp ort good, th e ch a n g e to free trade w ill ca u se p ro d u cer gains th at ex ceed c o n su m e r losses. The Effects o f a Tariff T o m ake th e analysis o f p ro tectio n ist trade policy as straightfoi’w ard as p o ssib le, th e im p act o f a tariff is analyzed . (One ca n view any p ro tectio n ist trade policy, how ever, in an an alogous m anner.) For c o n ven ien ce, th e free trad e resu lts in figure 1 are d u p li ca ted in figure 2. Given the free trad e w orld p rice o f Plv, U.S. co n su m p tio n , p ro d u ctio n an d im p orts are D 'us, S ' _ and S 'us D 'lls. A ssum e a tariff is im posed , cau sin g the p rice in th e United States to in crea se to P ,. T h e p rice in th e U nited States n ow e x ceed s p rice in th e w orld by th e am ou n t o f th e tariff, PW PT. T h e h ig h er U.S. p rice ca u se s c o n su m e r p u r ch a se s to d e cre a se from D'„, to D"us, d o m estic p ro d u ctio n to in crea se from S 'us to S"us, an d im p o rts to d ecrea se from S 'usD 'us to S"„sD"us. By im p osin g the tariff, co n su m ers lo se th e area PTJCP„. o r d + e + f + g an d p ro d u cers gain th e area PTIFP„ o r d. Do- FEDERAL RESERVE BANK OF ST. LOUIS F ig u re 2 JANUARY/FEBRUARY 1988 m e stic p ro d u cers are p ro tected at th e ex p e n se of d o m estic co n su m ers. O ne co m p lica tio n stem s from tariff revenue. T a r iff revenue, w h ich ca n b e view ed as a gain for th e governm ent, eq u als th e tariff, PW PT, tim es th e q u a n tity o f im ports, S"usD"us. T h is reven u e is equal to area IJGH o r f. Overall, th e n a tio n lo ses b e c a u se th e c o n s u m e rs’ lo sses o fd + e + f + g e x ceed th e p ro d u c e rs’ gains o f d an d th e g overnm ent gains o f f by e + g. Area e is called a “d ead w eight p ro d u ctio n lo s s ” an d ca n b e viewed as a lo ss resu lting from in efficien t (excess) d o m estic p ro d u ctio n , w hile area g is called a “d ea d w eight co n su m p tio n lo s s ” an d ca n be view ed as a loss resu ltin g from in efficien t (too little) co n su m p tio n . in ten d ed to im prove th e p o sitio n o f d o m estic relative to foreign p ro d u cers. T h is can be d on e th rou gh po li cies that in crea se th e ho m e m arket p rice o f the foreign prod u ct, d ecrea se th e p ro d u ctio n co sts o f d o m estic firm s, o r so m eh o w restrict th e a c c e ss o f foreign p ro d u cers to th e d o m estic m arket. 1'he sp ecific goal o f p ro tectio n ist trad e p o licies is to expand d o m estic p ro d u ctio n in th e p ro tected in d u s tries, ben efitin g th e ow ners, w orkers and su p p liers of re so u rces to the p ro tected in d u stiy. The governm ent im posing p ro tectio n ist trade p o licies m ay also benefit, for exam ple, in th e form o f tariff revenue. T h e ex p an sio n o f d o m estic p ro d u ctio n in p ro tected in d u stries is not co stle ss; it requ ires ad d ition al re so u rces from o th e r in d u stries. C onsequently, ou tp u t in o th e r d o m estic in d u stries is red u ced . T h e se in d u s tries also m ight be m ad e less com p etitive b ec a u se of h ig h er p rices for im p o rted inp u ts. Sin ce p ro tectio n ist trad e p o licies frequ en tly in crea se th e p rice o f th e p ro tected good, d o m estic co n su m ers are h arm ed. T h ey lose in two w ays. First, th e ir co n su m p tio n o f th e p ro tected good is red u ced b eca u se o f th e asso cia ted rise in its p rice. Seco n d , th ey co n su m e less o f o th e r goods, as th e ir o u tp u t d eclin es and p rices rise. T h e p reced in g d iscu ssio n highlights th e d o m estic w in n ers and lo sers d ue to p ro tectio n ist trad e p o licies. D om estic p ro d u cers o f th e p ro tected good and the governm ent (if tariffs are im posed) gain; d o m estic c o n su m ers and o th e r d o m estic p ro d u cers lose. Foreign in terests are also affected by trad e restrictio n s. T h e p ro tectio n o f d o m estic p ro d u cers will harm som e foreign p ro d u cers; oddly enou gh, o th e r foreign p ro d u cers m ay b enefit. For exam ple, if qu otas are p laced o n im ports, som e foreign p ro d u cers m ay receive h ig h er p rices for th eir exp o rts to th e p ro tected m arket. T h ere have b ee n n u m ero u s stu d ies o f the c o sts of p ro tectio n ism . W e b egin by exam in in g th ree re cen t stu d ies o f p ro tectio n ism in th e U nited States, th e n p ro ce e d to stu d ies exam in in g developed and, finally, d eveloping co u n tries. Costs o f Protectionism in the United States R ecen t stu d ies by T a rr an d M orkre (1984), Hickok (1985) an d H ufbauer et al. (1986! estim ated the co sts of p ro tectio n ism in th e U nited States. T h e se stu d ies use d ifferent estim atio n p ro ced u res, exam in e different p ro tectio n ist p o licies an d cover d ifferent tim e period s. N on eth eless, th ey provide co n siste n t results. T arr an d M orkre (1984) estim ate an n u al co sts to the U.S. eco n o m y o f $12.7 billion (1983 dollars) from all tar iffs an d from qu otas on au tom ob iles, textiles, steel and sugar. T h e ir co st estim ate is a n et m easu re in w h ich the lo sses o f co n su m ers are offset partially by th e gains o f d o m estic p ro d u cers an d th e U.S. governm ent. E stim ates by Hickok (19851 in d ica te that trad e re strictio n s on only th ree good s — cloth in g, sugar, and a u to m o b iles — ca u sed in crea sed c o n su m e r e x p e n d i tu res o f $14 billion in 1984. Hickok also show s th at lowin co m e fam ilies are affected m o re th an h ig h -in co m e fam ilies. T h e im port restrain ts on cloth in g, su gar and a u to m o b iles are ca lcu la ted to b e equivalent to a 23 p e rce n t in co m e tax su rch arg e (that is, an ad d itional 15 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Form s Of Protectionism P ro tectio n m ay b e im p lem en ted in n u m ero u s w ays. All form s o f p ro tectio n are in ten d ed to im prove th e p o sitio n o f a d o m estic relative to foreign p ro d u cer. T h is ca n be d o n e by p o licies th at in cre a se th e h o m e m arket p rice o f th e foreign p ro d uct, d ecrea se th e co sts o f d o m estic p ro d u cers or re strict th e a c c e ss o f foreign p ro d u cers to th e hom e m arket in so m e o th e r way. in crea se in th e h o m e m arket. T h is in d u ce s d o m es tic p ro d u cers to in crea se p ro d u ctio n an d c o n su m ers to re d u ce co n su m p tio n . O ne d ifference b e tw een a tariff an d a qu ota is th at th e tariff g en erates revenue for th e governm ent, w h ile th e qu o ta g en er ates a revenue gain to th e o w n er o f im port licen ses. C on sequen tly, foreign p ro d u cers m ight ca p tu re som e o f th is revenue. Tariffs In re ce n t y ears, a slightly different version of quotas, called eith e r ord erly m arketing ag reem en ts o r voluntary exp ort restrain ts, h as b een used. In an ord erly m arketin g ag reem en t, th e d o m estic govern m en t asks th e foreign governm ent to re strict the qu an tity o f exp orts o f a good to th e d o m estic c o u n try. T h e req u est ca n b e view ed as a d em an d , like the U .S.-Japan au tom ob ile a g reem en t in th e 1980s, b e ca u se th e d o m estic co u n try m akes it cle a r that m ore restr ictive a ctio n s are likely u n less th e foreign govern m en t "volu ntarily" co m p lies. In effect, th e orderly m arketing a g reem en t is a m u tu ally agreedupon quota. Tariffs, w h ich are sim ply taxes im p o sed on goods en terin g a co u n tiy from abroad, resu lt in h ig h er p rice s and have b ee n th e m o st co m m o n form of p ro tectio n for d o m estic p ro d u cers. Tariffs have b ee n p o p u lar w ith govern m ents b ec a u se it ap p ears th at th e tax is bein g p aid by th e foreig n er w ho w ish es to sell his good s in th e h o m e eco n o m y and b ec a u se th e tariff revenue ca n be u sed to fin an ce governm ent sen d ees o r re d u ce o th e r taxes. In th e 20th cen tu ry, U.S. tariff rates peaked as a resu lt o f th e Sm oot-H aw ley T ariff o f 1930. F o r ex am ple, in 1932, tariff revenue as a p ercen tag e o f total im p orts w as 19.6 p e rce n t. An id en tical calcu la tio n for 1985 yield s a figure o f 3.8 p ercen t. T h e d eclin e w as due prim arily to tw o reaso n s. First, sin ce m any o f th e tariffs u n d e r Sm oot-H aw ley w ere set as sp e cific d ollar am o u n ts, th e rising p rice level in th e U nited States erod ed th e effective tariff rate. S e c ond, sin ce W orld W ar II, n u m ero u s tariff red u ctio n s have b een n eg o tiated u n d e r th e G eneral A greem ent on Tariffs an d Trade. On th e o th e r hand, various o th e r form s o f p ro te c tion, frequ en tly term ed n o n -tariff barriers, have b e co m e in creasin gly im p ortan t. A few o f th e m ore frequ ently u sed devices are d iscu sse d below . Quotas A qu ota seem s like a sen sib le alternative to a tariff w h en th e in ten tio n is to restrict foreign p ro d u c e rs’ a c c e ss to th e d o m estic m arket. Im p o rters typically are lim ited to a m axim u m n u m b er o f p ro d u cts that th ey can sell in th e h o m e m arket over sp ecific period s. A qu ota, sim ilar to a tariff, ca u se p rices to http://fraser.stlouisfed.org/ 16 Federal Reserve Bank of St. Louis Regulatory Barriers T h ere are m any o th e r w ays o f restrictin g for eign ers' a c c e ss to d o m estic m arkets. M un ger (1983) h as n o ted th at th e tariff co d e itself te n d s to limit trade. T h e 1983 T a r i f f S c h e d u le s o f t h e U n ited S ta te s A n n o ta te d co n sists o f 792 pages, plus a 78-page appen d ix. Over 200 tariff rates pertain to w a tch es an d clo ck s. Sim ply a scerta in in g th e ap p rop riate tariff classificatio n , w h ich requ ires legal a ssista n ce an d ca n b e su b je ct to d ifferen ces o f o p in ion , is a d eterren t. P rod u ct sta n d a rd s are a n o th e r co m m o n regu la tory b arrier. T h e se stan d ard s ap p ear in various form s an d are u sed for m an y p u rp o ses. T h e sta n d ards ca n b e u sed to serve th e p u b lic in terest by en su rin g that im p orted food p ro d u cts are p ro ce ssed acco rd in g to a ccep ta b le san itary stan d ard s an d th at drugs have b ee n scre e n e d b efo re th e ir in tro d u ctio n in th e U nited States. In o th e r ca ses, the stan d ard s, so m etim es in ten tion ally, p ro tect d o m estic p ro d u cers. An exam p le o f u n in te n d ed re- JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS strictio n s m ay be th e im p o sitio n o f safety o r p o llu tion stan d ard s th at w ere not previously bein g m et by foreign cars. Exchange Controls An alternative to restrictin g th e term s u n d er w h ich foreigners can co m p ete in th e h o m e m arket is to su bsid ize d o m estic p ro d u cers. Su bsid ies m ay be focu sed u p o n an in d u stry in g en eral o r u p o n th e exp ort activities o f th e in d u stiy . An exam p le o f th e form er, d iscu sse d by M orici an d M egna (1983), is th e co m b in atio n o f cred it program s, sp ecial tax in cen tiv es and d irect su bsid y p ay m en ts th at benefit th e U.S. ship bu ild ing ind u stry. An exam p le o f th e la tter is th e fin ancial a ssista n ce to in cre a se exp orts provided by th e U.S. E xp o rt-Im p o rt Bank throu gh d irect loans, lo an g u aran tees an d in su ran ce, an d d isco u n t lo an s. In eith e r case, p ro d u ctio n will expand . All o f th e above relate d irectly to th e flow of g ood s. A final cla ss o f re strictio n s w orks by re strict ing a c c e ss to th e foreign m on ey requ ired to buy foreign good s. F o r exam ple, a govern m en t that w ish e d to p ro te c t its ex p o rtin g an d im p o rtco m p etin g in d u stries m ay try to h o ld its ex ch an g e rate artificially low . As a result, foreign goods w ould a p p ea r expensive in th e h o m e m arket w hile h o m e good s w ould b e ch e a p overseas. H om e p ro d u cers im plicitly are su b sid ized an d h o m e co n su m ers im p licitly are taxed. T h is p o licy is norm ally h ard to su stain . T h e ce n tra l bank, in h old ing th e ex ch a n g e rate dow n h as to bu y foreign ex ch a n g e w ith d o m es tic cu rren cy . T h is new ly issu ed d o m estic cu rre n cy in crea ses th e d o m estic m o n ey stock an d eventually ca u ses inflation. In flation ary p o licies are n ot n o r m ally regard ed as a sen sib le w ay o f p ro tectin g d o m e stic ind ustry. An im p o rtan t d ifferen ce b etw een su bsid ies and tariffs involves th e revenu e im p licatio n s for govern m ent. T h e form er involves th e govern m ent in p ay ing out m oney, w h ereas tariffs g en erate in co m e for th e governm ent. T h e effect on d o m estic p ro d u ctio n an d w elfare, how ever, ca n be th e sam e u n d e r su b si dies as u n d e r tariffs an d qu o tas. In all cases, th e p ro tected in d u stiy is bein g su bsid ized by th e rest o f th e eco n om y. T h e re is a n o th e r a sp e ct to ex ch a n g e co n tro ls. T h e ju stifica tio n is th at preven tin g h o m e resid en ts from investing overseas b en efits d o m estic grow th as it lead s to g reater d o m estic real in vestm en t. In reality, it co u ld do exactly th e o p p o site. R estrictin g a c c e ss to foreign a sse ts m ay raise th e v arian ce and lo w er th e retu rn to o w n ers o f d o m estic w ealth . In th e sh o rt run, it also m ay a p p recia te th e d o m estic ex ch a n g e rate and, thereby, m ake d o m estic p ro d u cers less com p etitive. Subsidies tax ad d ed to th e n orm al in co m e tax) for fam ilies w ith in co m es less th an $10,000 in 1984 and a 3 p ercen t in co m e tax su rch arg e for fam ilies w ith in co m es ex ceed in g $60,000. H ufbauer et al. (1986) ex am in ed 31 ca se s in w h ich trade volum es ex ceed e d $100 m illion an d the United States im p o sed p ro tectio n ist trade re strictio n s.4 They g en erated estim ates o f th e w elfare co n se q u e n ce s for ea ch m a jo r group affected (see table 1). T h e figures in th e table in d icate th at an n u al c o n su m e r lo sses ex ceed $100 m illion in all bu t six o f th e cases. T h e largest losses, $27 billion p e r year, co m e from p ro tectin g the textiles and ap p arel in d u stiy . T h e re also are large “While there were cases in which the industry adjusted to its new competitive position and the protection was terminated, these cases were more the exception than the rule. In far more cases, protection ist policies were maintained indefinitely or removed because of favorable demand changes. co n su m e r lo sses a sso cia te d w ith p ro tectio n in carb o n steel ($6.8 billion), a u to m o b iles ($5.8 billion) an d d aily p ro d u cts ($5.5 billion). T h e p u rp o se o f p ro tectio n ism is to p ro tect jo b s in sp ecific in d u stries. A u seful a p p ro a ch to gain som e pei"spective on c o n su m e r lo sses is to exp ress th ese lo sses on a per-job-saved basis. In 18 o f th e 31 cases, th e co st p er-job-sav ed is $100,000 or m ore p e r y ear; the co n su m e r lo sses p er-jo b -sav ed in b en z e n o id c h e m i cals, ca rb o n steel (two sep a ra te periods), sp ecialty steel, an d bolts, n u ts an d screw s ex c e e d e d $500,000 p e r year. Table 1 also reveals th at d o m estic p ro d u cers w ere th e p rim a iy b en eficiaries o f p ro tectio n ist p o licies; how ever, th e re are so m e n o tew o rth y ca se s w h ere for eign p ro d u cers realized relatively large gains. F o r the U .S.-Japan ese voluntary exp o rt ag reem en t in a u to m o biles, foreign p ro d u cers gain ed 38 p e rce n t o f w hat d o m estic co n su m ers lost, w hile a sim ilar co m p u tatio n 17 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Table 1 Distribution of Costs and Benefits from Special Protection Consumer Losses Case Manufacturing Book manufacturing Benzenoid chemicals Glassware Rubber footwear Ceramic articles Ceramic tiles Orange juice Canned tuna Textiles and apparel: Phase I Textiles and apparel: Phase II Textiles and apparel: Phase III Carbon steel: Phase I Carbon steel: Phase II Carbon steel: Phase III Ball bearings Specialty steel Nonrubber footwear Color televisions CB radios Bolts, nuts, large screws Prepared mushrooms Automobiles Motorcycles Services Maritime industries Agriculture and fisheries Sugar Totals (million dollars) $ 500 2,650 270,000 2,000 neg. 102 1,000 930 550 410 5 130 5,000 250 34 1,370 170 1,600 neg. 135 9 44 14 145 560 60,000 690/acre 220,000 1,800/cow 1,000/acre 160,000 225/head 21,000 200 170 177 15 6,900 67 160,000 30,000 2,0003 4 70 11 3,000 5 170 1,800 305 2,250 neg. neg. 130 90 25 62 390 74 8,700 18,000 22,000 1,330 2,770 3,800 21 420 250 190 14 60 13 2,600 67 neg. neg. neg. neg. neg. $ 7 neg. 350 1,800 330 930 2,000 neg. 50 220 140 neg. neg. neg. 2,200 neg. 54 139 69 55 128 10 1,158 2,143 2,535 290 556 560 18 32 262 77 32 16 25 790 21 13 33 6 11 130 4 1,100 3,100 4,850 50 120 330 neg. 30 16 7 5 1 0.8 200 17 4,800 46 $ 0 252 Efficiency loss (million dollars) 3,000 Peanuts Meat $ Tariff revenue (million dollars) 200 230 95 116 525 91 9,400 20,000 27,000 1,970 4,350 6,800 45 520 700 420 55 110 35 5,800 104 5,500 Mining Petroleum Lead and zinc Totals (million dollars) Welfare Costs of Restraints Gain to foreigners (million dollars) $ 100,000 over 1 million 200,000 30,000 47,500 135,000 240,000 76,000 22,000 37,000 42,000 240,000 620,000 750,000 90,000 1,000,000 55,000 420,000 93,000 550,000 117,000 105,000 150,000 Dairy products Fish Per job saved1 (dollars) Producer Gains $ 29 14 Neg. = negligible. 'Unless otherwise specified, figures are per worker. Estimated duties collected on ship repairs performed abroad. 3ln this case, because of the way the quotas were allocated, the gains to importers accrued to domestic refiners rather than foreign exporters. SOURCE: Trade Protection in the United States: 31 Case Studies. http://fraser.stlouisfed.org/ 18 Federal Reserve Bank of St. Louis FEDERAL RESERVE BANK OF ST. LOUIS for the latest p h ase o f p ro tectio n for carb o n steel w as 29 p ercen t. Finally, table 1 in d icates that the efficien cy losses are sm all in co m p ariso n to the total lo sses b o rn e bv co n su m ers. T h e se efficien cy lo sses, w h ich are defined p recisely and illu strated in th e first sh ad ed insert, result from the e x cess d o m estic p ro d u ctio n and the red u ctio n in co n su m p tio n cau sed by p ro tectio n ist trad e p o licies. In large ca se s su ch as textiles and apparel, p etroleum , d aiiy p ro d u cts an d the m aritim e industr ies, th ese lo sses equal o r exceed $1 billion. It is likely th at th e se estim ates u n d erstate th e actu al co sts b ec a u se th ey do n o t cap tu re th e seco n d ary effects th at o c c u r as p ro d u ctio n and co n su m p tio n ch an g es in one ind u stry affect o th e r in d u stries.’ In ad d ition, re s tric tive trad e p o licies g en erate ad d ition al co sts b e c a u se of b u rea u cratic en fo rce m en t co sts and efforts by th e private se c to r to in flu en ce th e se p o licies for th e ir ow n gain as well as sim ply co m p ly w ith adm inistrative regu lations. Costs o f Protectionism Throughout the World In 1982, th e O rg a n iz a tio n for E c o n o m ic C o o p eratio n and D evelopm ent (OECD) beg an a p ro je c t to analyze th e co sts an d b en efits o f p ro tectio n ist p o licies in m an u factu rin g in OECD co u n tries. T h e OECD (1985) highlighted a n u m b er o f w ays that p ro tectio n ist p o licies have g en erated co sts far in e x cess o f benefits. Sin ce p ro tectio n ist p o licies in crea se p rices, th e report co n clu d e s th at th e attain m en t o f su stain ed n o n in flatio n ary grow th is h in d e re d by su ch p ricein creasin g effects. M oreover, e co n o m ic grow th is p o ten tially re d u ce d if th e u n certain ty create d by varying trade p o licies d ep resse s investm ent. W ood an d M udd (1978), and m any o th ers, have sh ow n th at im p orts do n ot ca u se h ig h er u n em p lo y m ent. Conversely, th e OECD stud y stresse s th e fact that a red u ctio n in im p orts via trade re strictio n s d oes not ca u se g reater em p loy m en t. A red u ctio n in th e value o f im p o rts resu lts in a sim ilar re d u ctio n in th e 5Recent estimates of the costs of protectionist policies using general equilibrium models suggest that the secondary effects, to the limited extent they are measurable, are substantial. For example, Grais, de Melo and Urata (1986) estimate that the elimination of quotas in Turkey in 1978 would have caused a 5.4 percent rise in gross domestic product, while Clarete and Whalley (1985) estimate that the elimination of tariffs, quotas and export taxes in the Philippines in 1978 would have caused a 5.2 percent rise in gross national product. JANUARY/FEBRUARY 1988 value o f exports. O ne ratio n ale for th is finding is th at a re d u ctio n in th e p u rch a se s o f foreign good s red u ces foreign in co m es and, in turn, ca u se s red u ced foreign p u rch a se s o f d o m estic goods. W hile th e re d u ctio n in im p o rts in crea ses em p lo y m ent in in d u stries th at p ro d u ce p ro d u cts sim ilar to the previously im p o rted goods, th e red u ctio n in ex p orts d ecrea se s em p lo y m en t in th e export in d u stries. In o th e r w ords, w h ile so m e jo b s are saved, o th ers are lost; how ever, th is e c o n o m ic reality m ay not be obvi ous to b u sin essm en , lab or u nion lead ers, p o litician s an d o th ers. Luttrell (1978) has stressed that th e jo b s saved by p ro tectio n ist legislation are m o re readily observed th an th e jo b s lo st d ue to p ro tectio n ist legis lation. In o th e r w ords, th e jo b s that are p ro tected in, say, th e textiles in d u stry by U.S. im p o rt re strictio n s on foreign textiles are m o re readily a p p aren t (and p u b li cized) th an the jo b s in agricu ltu re an d high te c h n o l ogy in d u stries th at do not m aterialize b ec a u se o f the im port restrictio n s. T h e se em p lo y m en t effects w ill net to approxim ately zero.6 T h e OECD stud y also stresse s that d eveloping co u n tries n ee d exp orts to offset th e ir d ebts. T h u s, p ro te c tio n ist trad e p o licies by d eveloped co u n tries affect n ot only th e e c o n o m ic activity o f th e developing c o u n tries, but th e stability o f th e in tern atio n al financial system as d eb to r n a tio n s find it in creasin gly difficult to service th e ir d ebts. Not only d o es a free trade policy by developed co u n tries ben efit developing co u n tries, bu t a free trade policy by d eveloping co u n tries b en efits d evelop ing co u n tries. A re ce n t W orld Bank stud y (1987) o f 41 developing co u n tries co m p a red th e p erfo rm an ce of co u n tries follow ing a free trad e p o licy w ith co u n tries follow ing a re stricted trad e p olicy.7 Table 2 lists the 6Recent evidence shows that protectionist legislation actually may reduce employment. Denzau (1987) estimated that 35,600 manu facturing jobs were lost as a result of the September 1984 voluntary export restraints that limited the level of U.S. steel imports. Despite an increased employment for producers of steel (14,000) and pro ducers of inputs for steel producers (2,800), these increases were more than offset by the 52,400 job losses by steel-using firms. These losses are due to the higher steel prices that cause steelusing firms to be less competitive in export markets and subject them to more foreign competition in the U.S. market. The World Bank study divides trade strategies into two groups: outward oriented and inward oriented. An outward-oriented strat egy, which we call a free trade policy, is one in which trade and industrial policies do not discriminate between production for the domestic market and exports, nor between purchases of domestic and foreign goods. An inward-oriented strategy, which we call a restricted trade policy, is one in which trade and industrial policies are biased toward production for the domestic market relative to the export market. 19 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Table 2 Annual Average Growth of Per Capita Real Gross National Product Restricted Trade Free Trade Strongly Period Moderately 1963-73 Singapore South Korea Hong Kong 9.0% 7.1 6.0 Brazil Israel Thailand Indonesia Costa Rica Malaysia Ivory Coast Colombia Guatemala Cameroon 1973-85 Singapore Hong Kong South Korea 6.5 6.3 5.4 Malaysia Thailand Tunisia Brazil T urkey Israel Uruguay Chile Moderately Strongly 5.5%. 5.4 4.9 4.6 3.9 3.8 3.5 3.3 2.7 -0.1 Yugoslavia Mexico Nigeria Tunisia Kenya Philippines Bolivia Honduras El Salvador Madagascar Nicaragua Senegal 4.9% 4.3 4.2 4.0 3.9 2.2 2.0 1.9 1.4 1.1 1.1 0.6 Turkey Dominican Republic Burundi Argentina Pakistan Tanzania Sri Lanka Ethiopia Chile Peru Uruguay Zambia India Ghana Bangladesh Sudan 3.5% 3.4 3.2 3.1 3.1 2.7 2.3 1.9 1.7 1.5 1.5 1.2 1.1 0.4 -1 .4 -1 .9 4.1 3.8 2.9 1.5 1.4 0.4 0.4 0.1 Cameroon Indonesia Sri Lanka Pakistan Yugoslavia Colombia Mexico Philippines Kenya Honduras Senegal Costa Rica Guatemala Ivory Coast El Salvador Nicaragua 5.6 4.0 3.3 3.1 2.7 1.8 1.3 1.1 0.3 0.1 0.8 1.0 1.0 1.2 3.5 3.9 Bangladesh India Burundi Dominican Republic Ethiopia Sudan Peru Tanzania Argentina Zambia Nigeria Bolivia Ghana Madagascar 2.0 2.0 1.2 0.5 -0 .4 -0 .4 -1.1 -1 .6 -2 .0 -2 .3 -2 .5 -3.1 -3 .2 -3 .4 SOURCE: World Development Report 1987 and The Economist (1987). an n u al average grow th in real p e r cap ita gross n a tional p ro d u ct for e a ch o f th e 41 co u n tries for 1 9 6 3 -7 3 an d 1 9 7 3 -8 5 . T h o se co u n tries th at did n o t bias in d u s trial p ro d u ctio n tow ard th e d o m estic m arket by trade restrictio n s grew at faster rates th a n th o se th at did. F or exam ple, th e average an n u al grow th rate in real p e r cap ita in co m e for 1 9 6 3 -7 3 w as 6.9 p e rce n t in the eco n o m ies strongly o rien ted to free trade an d 1.6 p e rce n t in th e eco n o m ies strongly o rie n te d to re stricted trade. F o r 1 9 7 3 -8 5 th e se grow th rates w ere 5.9 p e rce n t and —0.1 p ercen t, respectively. ARGUMENTS FOR RESTRICTING TRADE T h e study p ro ceed s to identify th e m a cro e co n o m ic reason for th e gen eral finding. A given am o u n t o f new If p ro tectio n ism is so costly, w hy is p ro tectio n ism so pervasive? T h is sectio n reviews th e m a jo r argu m en ts http://fraser.stlouisfed.org/ 20 Federal Reserve Bank of St. Louis in vestm en t g en erated m o re ad d ition al o u tp u t in co u n tries follow ing a free trad e p o licy th a n a re stricted trade p olicy. T h e rea so n is th at a free trade en v iron m en t allow s cap ital to flow to its m o st highly valued u ses, w hile a restricted trad e en v iron m en t d is torts ec o n o m ic in cen tives. FEDERAL RESERVE BANK OF ST. LOUIS for restrictin g trad e and provides ex p lan atio n s for the ex iste n ce o f p ro tectio n ist trad e p o licies. National Defense T h e n atio n al d efen se arg u m ent savs that im port barriers are n ecessary to en su re th e cap acity to p ro d u ce cru cial good s in a n atio n al em erg en cy . W hile this argum ent is esp ecially ap p ealin g for w eap o n s d uring a war, th ere will likely be d em an d s from o th e r in d u s tries that d eem th em selves essen tial. F or exam ple, the footw ear in d u stry will d em an d p ro tectio n b eca u se m ilitary p e rso n n el n ee d co m b at boots." T h e natio n al d efen se argum en t ignores th e p o ssib il ity o f p u rch a se s from friendly co u n tries d uring the em erg en cy . T h e p o ssib ilities o f storage and d ep letio n raise ad d itional d oubts abou t th e gen eral ap p licability o f th e argum ent. If cru cial good s can b e stored , for exam ple, the least co stly w ay to p rep are for an em er g en cy m ight be to buy th e goods from foreigners at the low w orld p rice before an em erg en cy and store them . If th e cru cia l good s are d ep letab le m ineral resou rces, su ch as oil, th e n the restrictio n o f oil im p orts before an em erg en cy will ca u se a m o re rapid d ep letio n o f d o m e stic reserves. O n ce again, stockp iling m ight be a far less co stly alternative. In com e Redistribution Sin ce p ro tectio n ist trade p o licies affect th e d istrib u tion o f in co m e, a trade restrictio n m ight be d efen d ed on th e g roun d s that it favors som e disadvantaged group. It is unlikely, how ever, th at trade policy is the best tool for dealing w ith th e perceived evils o f in co m e inequality, b e c a u se o f its b lu n tn e ss an d adverse effects on th e efficient allo catio n o f re so u rces. A ttem pting to eq u alize in co m es d irectly by tax an d tran sfer pay m en ts is likely less co stly th an u sin g trad e policy. In ad d ition, as H ickok’s (1985) stud y in d icates, trad e re strictio n s on m an y item s in crease rath er th an d e crea se in co m e inequality. Optimum Tariff Argument T h e op tim u m tariff arg u m en t ap p lies to situ atio n s in w h ich a co u n try h as th e ec o n o m ic pow er to alter w orld p rices. T h is p ow er exists b ec a u se th e co u n try (or a group o f co u n tries actin g in co n so rt like th e O rganization o f P etroleu m Exp ortin g C ountries) is 8See Pine (1984). JANUARY/FEBRUARY 1988 su ch a large p ro d u cer or c o n su m e r o f a good that a ch an g e in its p ro d u ctio n o r co n su m p tio n p attern s in flu en ces w orld p rices. F or exam ple, by im p osin g a tariff, th e co u n try ca n m ake foreign good s ch ea p er. Sin ce a tariff re d u ce s th e d em an d for foreign goods, if the tariff-im p osing co u n try has som e m arket pow er, the w orld p rice for th e good will fall." T h e tariffim p osin g co u n try w ill gain b e c a u se th e p rice p e r unit o f its im p orts will have d ecreased . T h e re are a n u m b er o f o b sta cle s that p reclu d e the w id esp read ap p lica tio n o f th is argum en t. Few c o u n tries p o sse ss th e n ecessa ry m arket p o w er and, w h en th ey do, only a sm all n u m b er o f goods is covered. Second ly, in a w orld o f shifting supply an d dem and , calcu latin g th e op tim u m tariff an d a d ju stin g th e rate to ch an g in g situ atio n s is difficult. Finally, th e p o ssib il ity o f foreign retaliatio n to an a ct o f ec o n o m ic w arfare is likely. Su ch retaliatio n cou ld leave b o th co u n tries w orse off th an th ey w ould have b een in a free trade environm ent. Balancing the Balance o f Trade M any co u n tries en a ct p ro tectio n ist trad e p o licies in th e h o p e o f elim in atin g a b a la n ce o f trad e d eficit or in creasin g a b a la n ce o f trad e su rplu s. T h e d esire to in crea se a b a la n ce o f trad e su rp lu s follow s from the m ercan tilist view th at larger trade su rp lu ses are b e n e ficial from a n ation al p erspective. T h is arg u m en t is su sp ect on a n u m b er o f g round s. First, th ere is n o th in g in h eren tly u n d esirab le about a trade d eficit o r d esirab le about a surplus.'" For ex a m ple, faster ec o n o m ic grow th in th e United States th an in th e rest o f th e w orld w ou ld ten d to ca u se a trade deficit. In th is ca se, th e trade d eficit is a sign o f a h ealth y eco n om y. Seco n d , p ro tectio n ist p o licies that red u ce im ports will ca u se exp orts to d ecrea se by a co m p arab le am o u n t. H ence, an attem p t to in crea se exp o rts p erm a n en tly relative to im p o rts will fail. It is doubtful th at th e trad e d eficit will b e red u ced even tem porarily b ec a u se im port q u an tities do n ot d eclin e quickly in re sp o n se to th e h ig h er im port p rices and th e reven u es o f foreign p ro d u cers m ight rise. 9lf a country such as the United States has no market power, the world price is fixed. Consequently, the price faced by U.S. con sumers and producers rises by the full amount of the tariff. In the optimum tariff case, the price faced by U.S. consumers and pro ducers rises, but not by the full amount of the tariff. This must be the case because the world price falls and the amount of the tariff is the difference between the world price and the U.S. price. ,0See Chrystal and Wood (1988) earlier in this issue. 21 JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS Protection o f J o b s — Public Choice T h e p ro tectio n o f jo b s arg u m ent is clo sely related to th e b alan ce o f trade argum ent. Sin ce a red u ctio n in im p orts via trade re strictio n s will resu lt in a sim ilar re d u ctio n in exports, th e overall em p lo y m en t effects, as fou nd in the OECD (1985) study and m any oth ers, are negligible. W hile th e o v e r a ll effects are negligible, w orkers (and re so u rce ow ners) in sp ecific in d u stries are affected differently. A d o m estic in d u stry faced w ith in crea sed im ports from its foreign co m p etitio n is u n d e r p ressu re to re d u ce p ro d u ctio n and low er co sts. Productive re so u rces m u st m ove from th is in d u stry to o th e r d o m es tic in d u stries. W orkers m u st ch an g e jo b s and, in som e cases, relo cate to o th e r cities. Sin ce th is ch an g e is forced u p o n th e se w orkers, th e se w orkers b e a r real co sts th at they are likely to r esist. A sim ilar sta tem en t ca n be m ad e abou t th e ow ners o f cap ital in th e af fected industry. W orkers an d o th e r re so u rce ow ners w ill likely resist th e se ch an g es by lobbying for trade restrictio n s. T h e pr eviously cited stu d ies on th e co sts o f p ro tectio n ism d em o n strated th at trad e restrictio n s entail su bstan tial real co sts as w ell. T h e se co sts likely e x ceed the a d ju st m ent co sts b eca u se th e a d ju stm en t co sts are o n e-tim e co sts, w hile th e co sts o f p ro tectio n ism co n tin u e as long as trad e re strictio n s are m aintain ed . An obvious q u estio n is w hy p o litician s su pply the p ro tectio n ist legislatio n d em an d ed by w orkers and o th e r re so u rce ow ners. A b ra n ch o f ec o n o m ics called p u blic ch o ice , w h ich focu ses on th e in terp lay betw een individual p re fe ren ces an d p o litical o u tco m es, p ro vides an answ er. T h e p u blic c h o ice literatu re views the p o litician as an individual w ho offers voters a b u n dle o f governm entally su p p lied good s in o rd er to w in e le c tio n s ." M any argue th at p o litician s gain by provid ing p ro tectio n ist legislation. Even thou gh th e nation al eco n o m ic co sts ex ceed th e ben efits, th e p o litician faces different co sts and benefits. T h o se h arm ed by a p ro tectio n ist trad e p o licy for a d o m estic industry, esp ecially h o u se h o ld co n su m ers, will incur- a sm all individual co st that is difficult to identify. F o r exam ple, a co n su m e r is unlikely to p o n d er h ow m u ch extra a sh irt co sts b e c a u se o f p ro te c tio n ist legislation for th e textiles an d ap p arel industry. "The role of pressure groups, acting in their economic self-interest, has been stressed by Stigler (1971) and Peltzman (1976). For references, as well as an example of an international trade study focused on the interaction of politicians and interest groups, see Coughlin (1985). 22 Even thou gh th e aggregate effect is large, th e harm to ea ch co n su m e r m a y b e sm all. T h is sm all co st, o f w h ich an individual m ay n o t even b e aw are, an d th e co sts of organizing co n su m ers d eter th e form ation o f a lobby against th e legislation. On th e o th e r h an d , w orkers an d o th e r re so u rce ow ners are very co n c e rn e d abou t p ro tectio n ist legisla tio n for th e ir ind ustry. T h e ir ben efits ten d to be large individually an d easy to identify. T h e ir voting and cam paign co n trib u tio n s assist p o litician s w ho su p port th e ir p o sitio n s an d p en alize th o se w ho do not. T h u s, p o litician s are likely to resp o n d to th e ir d e m an d s for p ro tectio n ist legislation. Infant Industries T h e p reced in g arg u m en t is c o u ch e d in term s of p ro tectin g a d o m estic ind ustry. A slightly different argu m en t, th e so -ca lled in fan t in d u stry ca se, is co u ch e d in term s o f p r o m o t in g a d o m estic ind ustry. Su p p ose an industry, alread y esta b lish ed in o th er co u n tries, is b ein g esta b lish ed in a sp ecific cou n try. T h e co u n try m ight n o t b e able to realize its co m p a ra tive advantage in th is in d u stry b ec a u se o f th e existing co st an d o th e r advantages o f foreign firm s. Initially, ow ners o f th e fledgling firm m u st b e w illing to suffer lo sses until the firm develops its m arket and low ers its p ro d u ctio n co sts to th e level o f its for eign rivals. In order' to assist th is en tran t, tariff p ro tectio n can be used to sh ield th e firm from so m e for eign co m p etitio n . After th is tem p o rary p eriod o f p ro tectio n , free trade sh ould be resto red ; how ever, the rem oval o f tariff p ro tectio n frequ ently is resisted . As th e in d u stry d e velops, its p o litical p o w er to th w art o p p o sin g legisla tion also in creases. A n o th er p ro b lem w ith th e in fan t in d u stry argum en t is th at a tariff is n ot th e b est w ay to in terven e. A p ro d u ctio n su bsid y is su p erio r to a tariff if the goal is to exp an d p ro d u ctio n . A su bsid y will do th is directly, w hile a tariff h as th e u n d esirab le side effect o f re d u c ing co n su m p tio n . In m any cases, in terven tio n m ight not be ap p ro p ri ate at all. If th e infant in d u stry is a good can d id a te for bein g com p etitive in tern ation ally, b orrow ing from the ,2Special interests benefiting from trade will likely resist the forces for protectionist legislation. Destler and Odell (1987) identify exporters, industrial import users, retailers of imported products, businesses providing trade-related services, foreign exporters, and foreign gov ernments as interest groups capable of exerting some anti protection pressure. Decisions about protectionist legislation result from the interaction of both pro-protection and anti-protection forces. FEDERAL RESERVE BANK OF ST. LOUIS private cap ital m arkets can fin an ce th e exp an sio n . Investors are w illing to abso rb lo sses t e m p o r a r ily if th e p ro sp ects for future profits are su fficiently good. Spillover Effects T h e ju stificatio n for p ro tectin g an ind ustry, infant or otherw ise, frequ en tly en tails a su ggestion th at the in d u stiy g en erates spillover ben efits for o th e r in d u stries o r individuals for w h ich th e in d u stiy is not co m p en sated . D esp ite p aten t law s, on e co m m o n su g gestio n is th at certain in d u stries are n o t fully co m p e n sated for th e ir re sea rch and developm ent ex p en d i tures. T h is argum en t is frequ en tly d irected tow ard tech n o log ically progressive in d u stries w here som e firm s can cap tu re th e resu lts o f o th e r firm s’ re sea rch an d d evelopm ent sim ply bv d ism antling a p ro d u ct to see how it w orks. T h e ap p licatio n o f this argum ent, how ever, en g en ders a n u m b er o f problem s. Spillovers of know ledge are difficult to m easu re. Sin ce spillovers are not m ar ket tra n sactio n s, th ey do n o t leave an obvious trail to identify th e ir b en eficiaries. T h e lack o f m arket tra n sa c tion s also co m p lica te s an a sse ssm en t o f th e value of th ese spillovers. To d eterm in e th e ap p rop riate su b sidy, o n e m ust be able to p lace a d ollar value on the spillovers g en erated by a given re se a rc h an d develop m en t exp en d itu re. Actually, th e calcu latio n requires m u ch m ore th an th e alread y difficult task o f re c o n stru ctin g th e past, ft requ ires co m p lex estim ates o f the spillovers' future w orth as w ell. Sin ce re so u rces are moved from o th e r in d u stries to th e targeted ind u stiy, th e governm ent m u st u n d e rstan d th e fu n ctio n in g of the en tire econ om y. Finally, th ere are political p ro blem s. An aggressive ap p licatio n of th is argum en t m ight lead to retaliation and a m utually d estructive trade w ar. In ad dition, as in terest groups co m p ete for th e g overnm en tal assis tan ce, th ere is no gu aran tee th at th e right groups will b e assisted o r th at they will u se th e assista n ce efficiently. Strategic Trade Policy Recent th eo retical d evelop m ents have id entified ca ses in w h ich so -called strategic trade policy is su p e rior to free trade. As w e d iscu sse d earlier, d ecreasin g unit p ro d u ctio n co sts and m arket stru ctu res th at c o n tain m on o p o ly elem e n ts are co m m o n in in d u stries involved in in tern atio n al trade. M arket im p erfectio n s im m ed iately suggest the p o ten tial ben efits o f govern m ental intervention. In th e strategic trad e p o licy argu m ent, governm ent p o licy can alter th e term s o f c o m JANUARY/FEBRUARY 1988 p etitio n to favor d o m estic over foreign firm s an d shift the ex cess retu rn s in m o n o p o listic m arkets from for eign to d o m estic firm s. Krugm an (1987) illu strates an exam p le o f th e argu m en t. A ssu m e th a t th ere is only o n e firm in th e United States, Boeing, an d o n e m u ltin atio n al firm in Europe, Airbus, cap ab le o f p ro d u cin g a 150-seat p a sse n g e r aircraft. A ssum e also th at the aircraft is p ro d u ced only for export, so th at th e retu rn s to th e firm ca n be id entified w ith th e n atio n al in terest. T his exp o rt m ar ket is p rofitable for e ith e r firm if it is th e only p ro d u cer; how ever, it is u n p ro fitab le for bo th firm s to p ro d u ce th e plan e. Finally, a ssu m e th e follow ing payoffs are a sso cia ted w ith the four co m b in a tio n s o f p ro d u ctio n : 1) if b o th B oeing an d Airbus p ro d u ce th e aircraft, ea ch firm lo ses $5 m illion; 2) if n eith e r B oeing n o r Airbus p ro d u ces th e aircraft, profits are zero; 3) if B oeing p ro d u ces th e aircraft an d A irbus d oes not, B oeing profits by $100 m illion an d Airbus h as zero profits; and 4) if Airbus p ro d u ces th e aircraft an d B oeing d oes not, Airbus profits by $100 m illion an d B oeing has zero profits. W h ich firm(s) w ill p ro d u ce th e aircraft? T h e ex am ple d oes n ot yield a u n iq u e o u tco m e. A u niqu e o u t co m e can be g en erated if on e firm, say Boeing, has a head start an d beg in s p ro d u ctio n before Airbus. In th is case, B oeing will reap profits o f $100 m illion and will have d eterred Airbus from en terin g th e m arket b eca u se Airbus will lose $5 m illion if it en ters after Boeing. Strategic trade policy, how ever, suggests th at ju d i cio u s g overn m en tal in terven tion ca n alter th e o u t co m e. If th e E u ro p ea n g overn m en ts agree to su bsid ize A irbus’ p ro d u ctio n w ith $10 m illion no m a tter w hat Boeing does, th e n Airbus w ill p ro d u ce th e p lan e. Pro d u ctio n by Airbus will yield m o re profits th an not prod u cin g , no m a tter w h at B oeing does. At th e sam e tim e, B oeing w ill b e d eterred from p ro d u cin g b eca u se it w ould lo se m oney. T h u s, Airbus w ill ca p tu re th e en tire m arket an d reap profits o f $110 m illion, $100 m illion o f w h ich ca n b e view ed as a tran sfer o f profits from the U nited States. T h e criticism s o f a strategic trad e policy are sim ilar to th e criticism s against p ro tectin g a tech n o log ically progressive in d u stiy th at g en era tes spillover b en e fits.13 T h ere are m a jo r in form ation al p ro b lem s in applying a 13A recent volume edited by Paul Krugman (1986) examines the jDolicy implications of the new trade literature. See Grossman's article in that volume for a discussion of the information require ments. 23 FEDERAL RESERVE BANK OF ST. LOUIS strategic trade policy. T h e govern m ent m u st estim ate th e poten tial payoff o f ea ch co u rse o f actio n . E c o n o m ic know ledge abou t th e behavior o f in d u stries th at have m o n o p o ly elem en ts is lim ited . Firm s m ay behave com p etitively or cooperatively and m ay co m p ete by setting p rices or output. T h e behavior o f rival governm ents also m u st be an ticip ated . Foreign retalia tio n m u st be view ed as likely w h ere su b stan tial profits are at stake. In ad dition, m any in tere st grou p s will co m p ete for th e governm ental assistan ce . Though only a sm all n u m b er o f secto rs can be co n sid ered poten tially strategic, m any in d u stries w ill m ake a case for assistan ce. Reciprocity and the “Level Playing Field” Bhagw ati and Irwin (1987) n o te that U.S. trade policy d iscu ssio n s in re ce n t y ears have frequ ently stressed th e im p o rtan ce o f “fair trad e .” T h e co n c e p t o f fair trade, w h ich is te ch n ica lly referred to as recip rocity , m eans different things to different people. U nder th e G eneral A greem ent on Tariffs an d Trade, n eg o tiation s to re d u ce trad e b arriers fo cu s u p on m a tch in g c o n c e s s io n s . T h is form o f recip ro city , k now n as first-d ifference recip rocity , attem p ts to re d u ce trad e b arriers by requiring a co u n try to provide a tariff re d u ctio n o f value co m p arable to on e provided by th e o th e r cou ntry. In this case, recip ro city is defined in term s o f m atch in g ch an g es. R ecen t U.S. d em an d s, exem plified by th e G ephardt am en d m en t to th e cu rren t trade legislation, reveal an a p p ro ach that is called full recip rocity . T h is ap p roach seeks recip ro city in term s o f th e level o f p ro tectio n b ilaterally an d over a sp ecific range o f good s. R eci p ro city requires equal a c c e ss and this a c c e ss can be d eterm in ed bv bilateral trad e b alan ces. A trade d eficit w ith a trading p a rtn er is claim ed to b e p r im a f a c i e ev id ence o f u neq u al a cc e ss. E xam ples abou n d . For exam ple, U.S. co n stru ctio n firm s have not had a m ajo r co n tra ct in Ja p a n sin ce 1965, w hile Ja p a n e se c o n stru ctio n firm s did $1.8 b illion w o rth o f b u sin ess in th e United States in 1985 alo n e. R ecen t legislation bars Ja p a n e se p articip atio n in U.S. p u blic w orks p ro jects until the Ja p a n e se offer recip ro cal privileges. As th e n am e suggests, th e fu n d am en tal argum en t for fair trade is o n e o f equity. D om estic p ro d u cers in a free trade co u n try argue that foreign trade b arriers are u nfair b e c a u se it p laces th em at a com p etitive d isad vantage. In an extrem e version, it is asse rte d th at th is unfair co m p etitio n will virtually elim in ate U.S. m a n u facturing, leaving only jo b s th at co n sist prim arily o f flipping ham bu rgers at fast food restau ran ts or, as Bhagw ati an d Irw in have said, rolling rice cak es at Ja p a n ese -o w n ed su sh i bars. W hile d o m estic pro http://fraser.stlouisfed.org/ 24 Federal Reserve Bank of St. Louis JANUARY/FEBRUARY 1988 d u cers a r e relatively disadvantaged, th e w isd om o f a p ro tectio n ist re sp o n se is d oubtful. Again, th e co sts of p ro tectio n ism ex ceed su b stan tially the b en efits from a n atio n al p erspective. In an attem p t to rein force th e arg u m en t for fair trade, p ro p o n en ts also argue th at retaliatory threats, co m b in ed w ith ch a n g es in tariffs an d n o n -tariff b a rri ers, allow for th e sim u ltan eo u s p ro tectio n o f d o m estic in d u stries against u n eq u al co m p etitio n an d in d u ce m ore o p en foreign m arkets. T h is m ore flexible a p p ro a ch is view ed as su p erio r to a “o n e -sid e d ” free trad e policy. T h e su ggestion th a t a fair trad e policy p ro d u ces a trading en viron m en t w ith few er trad e re strictio n s allow s p ro p o n e n ts to a ssert th at su ch a policy serves to p ro m o te b o th equity an d efficiency. In o th e r w ords, n o t only w ill d o m estic an d foreign p ro d u cers in th e sam e in d u stry b e treated equally, b u t the gains a sso cia ted w ith a freer trading en v iron m en t will b e realized. On th e o th e r hand , critics o f a fair trade policy argue that su ch a p o licy is sim ply d isguised pr o te ctio n ism — it sim ply achieves th e goals o f sp ecific in terest groups at th e ex p en se o f th e n a tio n at large. In m any cases, fair trad ers focu s on a sp ecific p ra ctice th at can be portrayed as p ro tectio n ist w hile ignoring th e en tire package o f p o licies th at are affecting a n a tio n ’s c o m petitive p o sitio n . In th e se ca ses, th e foreign co u n try is m ore likely eith e r n o t to resp o n d o r retaliate by in creasin g ra th er th an red u cin g th eir trade barriers. In th e la tte r case, th e esca la tio n o f trad e b arriers ca u ses lo sses for b o th n atio n s, w h ich is exactly o p p o site to th e alleged effects o f an activist fair trade policy. C ritics o f fair trad e p ro p o sals are esp ecially b o th ered by th e u se o f bilateral trad e d eficits as evid en ce of u nfair trade. In a w orld o f m any tradin g co u n tries, the tr ade b etw een two co u n tries n eed n ot b e b a la n ced for th e trad e o f e a c h to b e in global b a la n ce. Differing d em an d s an d prod uctive cap ab ilities a cro ss co u n tries will ca u se a sp ecific co u n try to have trad e d eficits w ith som e co u n tries an d su rp lu ses w ith o th e r co u n tries. T h e se b ilateral im b a la n ce s are a n orm al resu lt of co u n tries trading on th e b a sis o f com p arative advan tage.'4 T h u s, th e focu s on th e b ilateral trad e d eficit can pro d u ce in ap p ro p riate co n clu sio n s abou t fairness and, m o re im portantly, p o licies a ttem p tin g to elim i n ate bilateral trad e d eficits are likely to b e very co stly b eca u se they elim in ate th e gains from a m u ltilateral trading system . ,4Bergsten and Cline (1985) estimate an equilibrium U.S.-Japanese bilateral trade deficit of $20-$25 billion annually. JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS CONCLUSION T h e p roliferation o f p ro tectio n ist trade p o licies in re cen t y ears provides an im p etu s to re co n sid e r th eir w orth. In th e w orld of trad itional trade theory, c h a r a c terized bv p erfect co m p etitio n , a definitive re co m m en d atio n in favor o f free trad e ca n be m ade. T h e gains from in tern atio n al trad e resu lt from a re a llo ca tion o f productive re so u rces tow ard goods th at ca n be p ro d u ced less co stly at ho m e th an abroad and the ex ch an g e o f som e o f th e se good s for good s th at can be p ro d u ced at less co st abroad th a n at h om e. R ecen t d evelop m ents in in tern atio n al trad e theory have exam in ed th e co n se q u e n ce s o f in tern atio n al trade in m arkets w here th ere are m arket im p erfec tions, su ch as m on o p o ly an d tech n o lo g ical spillovers. Do th e se im p erfectio n s ju stify p ro tectio n ist trad e p o li cies? T h e an sw er co n tin u es to be no. W hile p ro te c tio n ist trad e p o licies m ay offset m on o p o ly p o w er over seas o r ad vantageou sly u se d o m estic m o n o p o ly pow er, trad e re strictio n s ten d to re d u ce th e co m p e ti tion faced by d o m estic p rod u cers, p ro tectin g d o m es tic p ro d u cers at th e ex p e n se o f d o m estic co n su m ers. T h e em p irical evid ence is clea r-cu t. T h e co sts of p ro tectio n ist trade p o licies far ex ceed th e benefits. T h e lo sses suffered by co n su m ers ex ceed th e gains reaped by d o m estic p ro d u cers an d governm ent. Lowin co m e co n su m ers are relatively m o re adversely af fected th an h ig h -in co m e co n su m ers. Not only are th ere in efficien cies a sso ciate d w ith excessive d o m es tic p ro d u ctio n and re stricted co n su m p tio n , but th ere are co sts a sso ciate d w ith th e en fo rce m en t o f th e p ro tectio n ist legislation and attem p ts to in flu en ce trade policy. T h e prim ary reaso n for th e se co stly p ro tectio n ist p o licies relies on a p u b lic c h o ice argu m ent. T h e desire to in flu en ce trad e policy arises from th e fact that trade policy ch an g es ben efit som e groups, w hile harm in g o th ers. C on su m ers are h arm ed by p ro tectio n ist legis latio n ; how ever, ig norance, sm all individual co sts, an d th e high co sts o f organizing co n su m ers prevent the co n su m ers from bein g an effective force. On th e o th e r hand , w orkers and o th e r re so u rce ow n ers in an in d u s try are m o re likely to be effective p o litically b e c a u se of th e ir relative ease o f organizing an d th e ir individually large an d easy -to-id entify ben efits. P olitician s in ter ested in re -e le ctio n will m ost likely resp o n d to the d em an d s for p ro tectio n ist legislation o f su ch an in ter est group. T h e em p irical evid en ce also su ggests that th e a d verse c o n su m e r effects o f p ro tectio n ist trad e p o licies are n o t short-lived. T h e se p o licies g en erate lo w er e c o n o m ic grow th rates th an th e rates a sso ciate d w ith free trade p o licies. In turn, slow grow th co n trib u te s to ad d ition al p ro tectio n ist p ressu res. In terest g roup p re ssu res from in d u stries e x p e rie n c ing difficulty an d th e g en eral ap p eal o f a "level playing field " co m b in e to m ake th e re d u ctio n o f trade barriers esp ecially difficult at th e p resen t tim e in th e United States. N o n eth eless, n atio n al in tere sts will be served b est by su ch an ad m itted ly difficult p o litical co u rse. In light o f th e cu rre n t Uruguay R ou nd n eg o tiation s u n d er th e G eneral A greem ent on Tariffs an d Trad e, as well as n u m ero u s b ilateral d iscu ssio n s, this fact is esp ecially tim ely. REFERENCES Bergsten, C. Fred, and William R. Cline. The United States-Japan Economic Problem (Institute for International Economics, October 1985). Bhagwati, Jagdish N., and Douglas A. Irwin. “The Return of the Reciprocitarians — U.S. Trade Policy Today,” World Economy (June 1987), pp. 109-30. Brander, James A. “ Intra-Industry Trade in Identical Commodities,” Journal of International Economics (February 1981), pp. 1-14. Brander, James A., and Paul R. Krugman. “A Reciprocal Dumping' Model of International Trade,” Journal of International Economics (November 1983), pp. 313-21. Chrystal, K. Alec, and Geoffrey E. Wood. “Are Trade Deficits a Problem?" this Review (January/February 1988), pp. 5-13. Clarete, Ramon L., and John Whalley. “ Interactions Between Trade Policies and Domestic Distortions,” Center for the Study of Inter national Economic Relations Working Paper 8522C (London, On tario: University of Western Ontario, 1985). Coughlin, Cletus C. “ Domestic Content Legislation: House Voting and the Economic Theory of Regulation,” Economic Inquiry (July 1985), pp. 437-48. Denzau, Arthur T. “ How Import Restraints Reduce Employment,” Washington University Center for the Study of American Business, Formal Publication #80 (June 1987). Destler, I. M., and John S. Odell. Anti-Protection: Changing Forces in United States Trade Politics (Institute for International Eco nomics, 1987). Dixit, Avinash K., and Victor D. Norman. Theory of International Trade (Cambridge University Press, 1980). Dixit, Avinash K., and Joseph E. Stiglitz. “ Monopolistic Competition and Optimum Product Diversity," American Economic Review (June 1977), pp. 297-308. The Economist. July 4, 1987, p.74. Grais, Wafik, Jaime de Melo, and Shujiro Urata. “A General Equilib rium Estimation of the Effects of Reductions in Tariffs and Quanti tative Restrictions in Turkey in 1978,” in T. N. Srinivasan and John Whalley, eds. General Equilibrium Trade Policy Modeling (MIT Press, 1986). Grossman, Gene M. “ Strategic Export Promotion: A Critique," Stra tegic Trade Policy and the New International Economics in Paul R. Krugman, ed. (MIT Press, 1986), pp. 47-68. Helpman, Elhanan. “ International Trade in the Presence of Product Differentiation, Economies of Scale and Monopolistic Competi tion,” Journal of International Economics (August 1981), pp. 30540. Hickok, Susan. “The Consumer Cost of U.S. Trade Restraints,” Federal Reserve Bank of New York Quarterly Review (Summer 1985), pp. 1-12. 25 FEDERAL RESERVE BANK OF ST. LOUIS Hufbauer, Gary Clyde, Diane T. Berliner, and Kimberly Ann Elliott. Trade Protection in the United States: 31 Case Studies (Institute for International Economics, 1986). Kearl, James R., Clayne L. Pope, Gordon C. Whiting, and Larry T. Wimmer. “A Confusion of Economists?" American Economic Review, Papers and Proceedings (May 1979), pp. 28-37. Kierzkowski, Henry K. “ Recent Advances in International Trade Theory: A Selective Survey,” Oxford Review of Economic Policy (Spring 1987), pp. 1-19. Krugman, Paul R. “ Is Free Trade Passe?” Journal of Economic Perspectives (Fall 1987), pp. 131-44. _________Strategic Trade Policy and the New International Eco nomics (MIT Press, 1986). Lancaster, Kelvin. “ Intra-Industry Trade Under Perfect Monopolistic Competition,” Journal of International Economics (May 1980), pp. 151-75. ________ Variety, Equity, and Efficiency (Columbia University Press, 1979). Leibenstein, Harvey. Press, 1980). Beyond Economic Man (Harvard University Luttrell, Clifton B. “ Imports and Jobs — The Observed and the Unobserved,” this Review (June 1978), pp. 2-10. Mill, John Stuart. Principles of Political Economy, W. J. Ashley, ed. (Longman, 1909). Morici, Peter, and Laura L. Megna. U.S. Economic Policies Affecting Industrial Trade: A Quantitative Assessment (National Planning Association, 1983). Munger, Michael C. “The Costs of Protectionism: Estimates of the Hidden Tax of Trade Restraint,” Washington University Center for 26 JANUARY/FEBRUARY 1988 the Study of American Business, Working Paper #80 (July 1983). Organization for Economic Co-Operation and Development (OECD). Costs and Benefits of Protection (1985). Page, Sheila. "The Rise in Protection Since 1974,” Oxford Review of Economic Policy (Spring 1987), pp. 37-51. Peltzman, Sam. “Toward a More General Theory of Regulation,” Journal of Law and Economics (August 1976), pp. 211—40. Pine, Art. "Footwear Industry Tells Congress Shoe Gap’ Threatens U.S. Defense," Wall Street Journal, August 24, 1984. Ricardo, David. The Principles of Political Economy and Taxation (Penguin, 1971). Stigler, George J. "The Theory of Economic Regulation," Bell Jour nal of Economics and Management Science (Spring 1971), pp. 3 21 . Stolper, Wolfgang, and Paul A. Samuelson. “ Protection and Real Wages,” Review of Economic Studies (November 1941), pp. 5873. Tarr, David G., and Morris E. Morkre. Aggregate Costs to the United States of Tariffs and Quotas on Imports: General Tariff Cuts and Removal of Quotas on Automobiles, Steel, Sugar, and Textiles, Bureau of Economics Staff Report to the Federal Trade Commis sion (December 1984). Venables, Anthony, and Alasdair Smith. “Trade and Industrial Pol icy Under Imperfect Competition,” Economic Policy (October 1986), pp. 621-72. Wood, Geoffrey E., and Douglas R. Mudd. “The Recent U.S. Trade Deficit — No Cause for Panic,” this Review (April 1978), pp. 2-7. World Bank. World Development Report 1987 (Oxford University Press, 1987). FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Appendix Developments in International Trade Theory and the Gains from Trade Sin ce 1817, n u m ero u s developm ents have taken p la ce in in tern atio n al trad e theory. T h e c o n se q u e n ce s o f m o re th an on e facto r o f pro d u ctio n , in creasin g and d ecreasin g u n it p ro d u ctio n co sts, an d im perfectly com p etitive m arkets are exam in ed in th is ap p en d ix. Sp ecial atten tio n is focu sed on d evelop m ents in in ter n atio n al trade th e o iy in th e last d ecad e. Increasing the N u m b er o f Factors o f Production A ssum e that, in th e U nited States, two reso u rces, la b o r an d cap ital (e.g., m achin es), are u sed in the p ro d u ctio n o f two goods, au to m o biles and airplan es. T h e p rices o f th e se re so u rces will be affected differ en tly by trade. As trad e develops, d em an d for the ex p o rted good (that is, th e good in w h ich th e United States has a com p arative advantage) will in cre a se and d em an d for U.S. p ro d u ctio n o f th e im p o rted good will fall. T h is d em an d shift ca u ses th e p rice o f th e ex p o rted good to rise relative to th e p rice o f th e im p orted good. Sim ilarly, th e shift m ay also p ro d u ce ch a n g es in th e p rices o f re so u rces; how ever, th ese p rice ch an g es are n o t alw ays obvious. Initially, assu m e that the re so u rces ca n n o t b e tra n s ferred a cro ss in d u stries. F o r exam ple, th e lab o r and cap ital u sed to p ro d u ce au to m o biles, th e good im p o rted in to th e U nited States, ca n n o t be u sed to p ro d u ce airp lanes, the ex p o rted good. C on sequently, as th e p rice o f airp lan es rises in th e U nited States, th e co m p en satio n for lab o r an d cap ital in th e airplane in d u stry w ill rise; m eanw hile, th e d eclin e in au to m o bile p rices ca u ses a d eclin e in co m p en sa tio n for lab or an d cap ital in th e in d u stiy. It w ould n ot be su rprisin g if lab or an d ow n ers o f cap ital in th e in d u stry w ould resist su ch ch an g es by asking for trad e p ro tectio n . W hile re so u rces m ay n o t be easily tran sferred a cro ss in d u stries in th e sh o rt run, w orkers ca n ch an g e jo b s an d cap ital ca n b e m oved as tim e p asses. If re so u rces are m obile, th e n th e lo n g er-ru n c o n s e q u en ces for la b o r an d o w n ers o f cap ital are different from th o se d escrib ed above. Even if lab or an d cap ital are p erfectly m obile, how ever, on e set o f re so u rce ow ners m ay ben efit w hile a n o th e r group is h arm ed by tra d e .1 T h e real w orld is m ore co m p lica ted th an this d is cu ssio n h as allow ed. T h ere are m o re th an tw o factors o f p ro d u ctio n an d varying d egrees o f m obility for th e se factors. For exam ple, th e U.S. lab or force c o n tain s scien tists an d en g in eers as w ell as sh o rt-o rd er cooks. N on eth eless, th e u nd erlyin g analysis d o es sug g est so m e g en eralizatio n s. W h en trad e o ccu rs, ow ners o f th e re so u rces th at are m ore sp ecialized in the p ro d u ctio n o f export goods will ten d to b eco m e w ea l 'Who wins and who loses? It depends on the U.S. endowment of capital to labor relative to other countries. If the United States has relatively larger amounts of capital to labor relative to other countries, then owners of capital would benefit, while labor would be harmed. This result follows from the Stolper-Samuelson Theorem (Stolper and Samuelson, 1941). In the example, the United States is defined to be capital-abundant. The example also implicitly assumes that airplanes are produced by capital-intensive methods and automo biles by labor-intensive methods. Thus, the production of airplanes requires the use of more capital relative to labor than automobiles. Since the United States is relatively well-endowed with capital and the production of airplanes is capital intensive, the United States will have a comparative advantage in the production of airplanes. With the elimination of trade barriers, the relative price of airplanes to automobiles will increase. The Stolper-Samuelson Theorem shows that an increase in the relative price of the capital intensive good will increase the return to capital relative to the prices of both goods and reduce the return to labor relative to the prices of both goods. 27 FEDERAL RESERVE BANK OF ST. LOUIS th ier; th o se w ho ow n re so u rces m ore sp ecialized in th e p ro d u ctio n o f im p o rt-co m p etin g goods will tend to lose w ealth . People also gain or lose, how ever, d ep en d in g o n w hat h ap p en s to the p rices o f th e goods they buy. Individuals w ho chiefly co n su m e im ported goods w ill ben efit, w hile th o se w ho p refer co n su m in g th e exp o rted goods will lose. T hu s, th e net effect on any individual d ep en d s on b o th th e gains or lo sses a sso ciate d w ith th e p rice ch an g es on th e good s th at h e co n su m es and th e effect o f trade on his w ealth (or incom e). Increasing Unit Production Costs A se c o n d assu m p tio n u nd erlying th e R icardian e x am ple o f th e gains from trad e is th at u nit p ro d u ctio n co sts are co n stan t. If u nit p ro d u ctio n co sts rise as m ore is p ro d u ced , how ever, th e gen eral co n clu sio n s abou t th e gains from trad e rem ain essen tially u n ch an g ed . T h e m a jo r difference is th at rising u nit p ro d u ctio n co sts lim it th e ex ten t to w h ich sp ecializatio n o ccu rs. Decreasing Unit Production Costs and Imperfect Competition On th e o th e r hand , if u n it p ro d u ctio n co sts d e crease as p ro d u ctio n in creases, th e ex ten t to w h ich a ctu al trad e p attern s ca n b e ex p lain ed by com parative advantage b e c o m e s u n clear. It also fo rces trad e th eo ry to deal w ith n u m ero u s ch a ra cte ristics o f in tern atio n al trad e in th e real w orld. T h e m arket stru ctu re o f in d u s tries engaged in trad e is frequ en tly highly c o n c e n trated. In o th e r w ords, th e individual firm s in an industry, co n trary to th o se in a p erfectly com p etitive industry, ca n affect th e m arket p rice o f th e ir good by th e ir p ro d u ctio n and advertising d ecisio n s. In ad d i tion, trad e statistics sh ow th a t in tra-in d u stry trade (i.e., th e sim u ltan eo u s exp ort an d im p o rt o f th e o u tp u t o f th e sam e industry) a cc o u n ts for in creasin gly larger sh ares o f w orld trade. In th e last d ecad e, trad e th eo rists have developed n u m ero u s m o d els to deal w ith th e se facts. An ex h a u s tive review o f th is rapidly exp an d in g literatu re is b e y ond th e sco p e of th is append ix; how ever, a few illu s trative articles are d iscu sse d in o rd er to estab lish som e key p o in ts. B ran d er (1981) and B ran d er an d K rugm an (1983) developed m od els u sing a h o m o g en e o us good to highlight ho w im p erfect co m p etitio n ca n ca u se in tra-in d u stry trad e an d h ow in tra-in d u stry trade can arise in th e a b se n ce o f co st d ifferences. A ssum e tw o co u n tries w ith o n e firm in e a c h co u n try. T h e firm s are p ro d u cin g a h o m o g en eo u s good u n d e r id en tical co st situ atio n s an d th ere are n o tra n s http://fraser.stlouisfed.org/ 28 Federal Reserve Bank of St. Louis JANUARY/FEBRUARY 1988 portation co sts. E a ch firm o p erates u n d e r w h at is term ed a “C ou rn ot c o n je c tu re ,” m ean in g th at ea ch firm a ssu m es its p ro d u ctio n d ecisio n w ill n o t affect its rival's p ro d u ctio n d ecisio n . B efore in tern atio n al trade, ea ch firm h as a m o n o p o ly p o sitio n in its h o m e m arket. Allowing for free trad e in d u ce s e a c h firm to e n te r th e o th e r firm ’s m arket, b eca u se p rice ex ceed s m arginal co st in e a c h co u n try . T h u s, th e sam e good w ill flow to an d from e a c h cou n try. Kierzkow ski (1987) h as n o ted th at th e bulk o f in tra in d u stry trad e involves d ifferentiated ra th er th a n h o m o g e n eo u s g oo d s. Tw o a p p ro a c h e s , L a n c a s te r’s (1979) ch a ra cte ristics a p p ro a ch an d Dixit an d Stiglitz’s (1977) “love o f variety” ap p roach , have provided the fou n d ation for trad e m o d els involving d ifferentiated goods. In th e ch a ra cte ristic s ap p roach , individuals have p re fe ren ces for th e ch a ra cte ristic s o f good s rath er th a n for co lle ctio n s o f th e good s th em selv es. A group of goods is defin ed as good s p o ssessin g th e sam e ch a ra cte ristics b u t in different p ro p o rtio n s. A diversity in c o n su m e r p re fe ren ces ca u ses different co n su m ers to p refer d ifferent p ro d u cts (i.e., varieties) o f a g roup of goods. H elpm an (1981) an d L a n ca ster (1980) u sed th e c h a r a cteristics a p p ro a ch to sh o w h ow in tra-in d u stry trade resu lts from co m b in in g th e d em a n d for variety w ith eco n o m ies o f scale. T h e ch a n g e from a u ta rch y to free trad e enlarges th e m arket an d ca u ses o u tp u t o f the existing varieties to in cre a s e an d th e p ro d u ctio n of new varieties to begin. C on su m ers gain from th e p ro d u ctio n o f m ore varieties an d low er p rices as e c o n o m ies o f sca le are realized . T h e so u rces o f g ains from trad e are id en tical using th e love-of-variety an d ch a ra cte ristic s ap p ro a ch es. In th e love-of-variety ap p ro a ch , w h ich is u sed by Dixit an d N orm an (1980), co n su m ers have id en tical tastes an d p refer to co n su m e as m any types o f th e d ifferenti ated p ro d u ct as p ossible. T h e in tro d u ctio n o f im p erfect co m p etitio n an d d e clin in g u n it p ro d u ctio n co sts suggest th ree so u rces o f gain from free trade. As th e m arket poten tially served by firm s ex p a n d s from a n atio n al to a w orld m arket, th ere will b e gains d ue to d eclin in g u nit p ro d u ctio n co sts. T h e se co n d is th e red u ctio n in m on o p o ly p ow er o f firm s faced w ith foreign co m p etito rs. T h e th ird is th e gain to co n su m ers from low er p rices an d in creased p ro d u ct variety. G enerally speaking, gains from trad e resu lt from th e in cre a se in com p etitive p ressu res as th e d o m estic eco n o m y b e c o m e s less in su lated from th e w orld eco n om y. N on eth eless, the n u m ero u s m arket stru ctu res an d firm behaviors p o s sible u n d e r im p erfect co m p etitio n p reclu d e a d efin i tive sta tem e n t abou t th e optim ality o f free trade. FEDERAL RESERVE BANK OF ST. LOUIS F ig u re 3 JANUARY/FEBRUARY 1988 T h e d em a n d curve is Dlls an d th e m arginal co st curve is MCUS. (The m arginal revenu e curve a sso cia te d w ith Dus is om itted.) T h e m o n o p o ly p rice an d o u tp u t are Pus an d Q lls. T h e ch a n g e from au tarch y to free trad e tran sform s th e n atio n al m o n o p o lies into a w orld duopoly. As sum ing th e firm s follow a C ou rn ot strategy, p rice d eclin es from Pus to Plv, sales in th e U nited States in crea se from Q us to D 'us, an d co n su m ers gain area Pus RS P„ or h + i. Profits (ignoring fixed costs) in th e U nited States d eclin e from area Pus RWM o r i + j + k to area P„ SXM o r j + k + 1. T h e d o m estic firm h as o n eh alf o f th e d o m estic m arket, so its profits are j w ith k + 1 going to th e foreign firm. T h e d o m estic firm 's exp orts allow it to ca p tu re o n e-h a lf o f th e foreign m arket. If th e foreign m arket is id en tical to th e d o m estic m arket, th e firm ’s profits on foreign sales w ill eq u al k + 1. T h e re fore, th e n et re d u ctio n in th e d o m estic firm 's profits is i — 1 an d th e overall w elfare gain to th e eco n o m y is h + 1. So m etim es th e ben efits o f exp an d ed co n su m p tio n resu lting from free trad e are less th an th e co sts a s so c i ated w ith d istorted p ro d u ctio n . V enables and Sm ith (1986) provide a g rap h ical illu stration, d u p licated in figure 3, o f the p reced in g point using th e B rand erKrugm an d uopoly m odel. A ssum e the U.S. m arket and th e m arket in the rest o f th e w orld for a sp ecific good are m on op olies, the good is p ro d u ced at a co n sta n t m arginal cost, an d th ere are n o tran sp o rtatio n co sts. If th e a ssu m p tio n o f id en tically sized d o m estic and foreign m arkets is d ropped, th e n a different c o n c lu sion is possib le. If th e foreign m arket is sm aller than th e d o m estic, th e profits o f th e d o m estic firm in the foreign m arket will b e le ss th a n k + 1. A ssum ing zero exp orts, th e d o m estic gains from trad e are h — k, an d th e d o m estic eco n o m y co u ld lose from free trade. In th is case, c o n su m e r gains ca n b e m ore th a n offset by th e shifting o f profits from th e d o m estic to th e foreign eco n om y. T h is shifting reflects th e co n tra ctio n o f an activity th at is alread y to o little to an even sm aller level. 29 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 The Borrow ed-Reserves Operating P roced u re: Theory and Evidence Daniel L. Thornton I n LATE 1982, th e Federal Reserve sw itch ed from a n on borrow ed -reserves to a borrow ed -reserves o p era t ing p ro ced u re.' Analysts generally believe that the a d op tion o f this p ro ced u re, w h ich involves th e use o f a “borrow ings a ssu m p tio n ” sp ecified by th e Fed eral O pen M arket C om m ittee (FOMC), rep rese n ts a p o licy reversal tow ard th e settin g o f th e federal fu n d s rate an d aw ay from d irect m o n ey stock con trol. T h is p ap er d iscu sse s th e m erits o f th e borrow ed reserves op eratin g p ro ced u re as a m eth o d for m on ey stock or in terest rate con trol, analyzes th e rela tio n sh ip s b etw een th e borrow ings assu m p tio n , th e federal funds rate and th e d isco u n t rate, an d provides som e evid ence on h ow th e new p ro ced u re h as b een u sed sin ce late 1982. THE NEW OPERATING PROCEDURE T h e co rn e rsto n e o f th e borrow ed -reserves o p era t ing p ro ced u re is th e borrow ings fu n ction , w h ich reflects th e b asic e co n o m ic factors that in d u ce d ep o s itory in stitu tio n s to borrow from th e Fed eral Reserve. It is u su ally argued th at th e level o f borrow ings (Borr) from th e F ed eral Reserve is in flu en ced prim arily by th e sp read betw een th e federal fund s rate (FFR) and th e Fed eral R eserve's d isco u n t rate (DR). A ccordingly, th e borrow ings fu n ctio n is (1) Borr, = b„ + b,(FFR, — DR,) + v„ w h ere b0 an d b, are co n sta n ts (b, > 0). T h e rand om erro r term , v„ cap tu res th e effect o f all o th e r factors that d eterm in e d ep ository in stitu tio n s’ borrow ings. It ca n be thou gh t to rep resen t “tran sitory " sh o ck s to the borrow ings fu n ction , w hile ch an g es in b0 rep resen t Daniel L Thornton is a research officer at the Federal Reserve Bank of St. Louis. Rosemarie V. Mueller provided research assistance. 1For a discussion of this change, see Roley (1986), Wallich (1984) Federal Reserve Bank of New York (1986) and Gilbert (1985). http://fraser.stlouisfed.org/ 30 Federal Reserve Bank of St. Louis “p e rm a n e n t” shifts in th e fu n ctio n . T h is d istin ctio n will b e u seful later. T h e equilibriu m FFR, given DR, is d eterm in ed by the d em an d for an d th e su pply o f total reserves. T h e su pply o f total reserves is co m p o sed o f n o n b o rro w ed reserves (NBR) an d reserves su p p lied th rou gh th e d is co u n t w indow . T h e d em an d for total reserves is co m p osed o f th e d em an d for requ ired plus e x cess re serves. T h eory su ggests th at th e d em an d for reserves is inversely related to th e federal fund s rate. Equating th e d em an d for an d th e su pply o f to tal reserves resu lts in an equilibriu m eq u atio n for th e fed eral fu n d s rate o f the gen eral form (2) FFR, = |x„ - jj.,NBR„ |x,„ |x, > 0. T his eq u ation show s all p o ssib le co m b in a tio n s o f n o n b orrow ed reserves an d fed eral fund s rates for w h ich th e su pply o f an d th e d em an d for to tal reserves are equal. E qu ation 1, p re sen ted in figure lb , an d eq u ation 2, p resen ted in figure la , can be u sed to illu strate how the borrow ings p ro ced u re o p era tes an d to sh o w sim i larities and d ifferen ces b etw een a borrow ings o p era t ing p ro ced u re an d a fed eral fu n d s rate targeting p ro ced u re. Su p p o se th e borrow ings assu m p tio n is set at Borr", sh ow n in figure lb . T h e borrow ings a ssu m p tio n rep rese n ts th e am ou n t o f reserves th at th e Fed w ish es to in d u ce d ep ository in stitu tio n s to b orrow from the d isco u n t w indow . T h is im plies FFR m u st eq u al FFR*. Given th e d em an d an d su p p ly fu n ctio n s, th e F ed ca n h it its borrow ings target by su pplying n o n b o rro w ed reserves equal to NBR*. T h is esta b lish es an eq u ilib rium FFR at FFR* co n siste n t w ith th e borrow ings objective. W hile figure 1 clarifies th e relatio n sh ip b etw een the b orrow ings objective, th e fu n d s rate an d n o n b o r row ed reserves, it m ay b e u nfam iliar to m an y read ers. C on sequen tly, from th is point on, th e analysis will be illu strated in term s o f th e m ore fam iliar figure 2. (See ap p en d ix A for a d iscu ssio n in term s o f figure 1 an d th e d etails co n cern in g th e v arian ces o f th e fed eral fu n d s rate an d borrow ings stated in th e text.) In figure FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 F ig u re 1 Equi l i bri um in the Reserv e Ma r k e t u nder a Bo r r owi n gs Pr oc edu r e FFR FFR 2, th e su pply o f total reserves is o btain ed by adding borrow ed reserves (from eq u atio n 1) to th e d esired level o f n on borrow ed reserves, NBR*, on th e a ssu m p tion th at b„ > 0 (see th e sh ad ed insert). T h e eq u ilib rium fed eral fund s rate, FFR*, is d eterm in ed by th e in tersectio n o f th e su pply an d d em an d curves in figure 2. As before, th e target level o f borrow ings is ach ieved by providing th e ap p rop riate am o u n t o f n o n borrow ed reserves. F ig u re 2 E q u i l i b r i u m in the Re s e r v e M a r k e t u nd e r a Borrowings Procedure H o w the Fed Should R esp on d to Shocks under a Borrowings Operating Procedure To u n d e rstan d p rop erly th e efficacy o f th e borrow ings p ro ced u re as a m eth o d o f m o n ey stock o r in terest rate co n tro l, it is im p o rtan t to see how th e F ed reacts to sh o ck s w h en u sin g it. First, co n sid e r its re sp o n se to an in crea se in th e d em an d for total reserves, illu s trated in figure 3a. O th er things co n stan t, an in crea se in th e d em an d for to tal reserves ca u se s th e eq u ilib rium funds rate to rise from FFR* to FFR *' and borrow ings to rise from th e d esired level (TR* —NBR*) to (TR*' —NBR*). T o bring borrow ings back to its d esired level, th e supply of n o n b o r r o w e d reserves m u st be in creased via an o p en m arket p u rch a se o f governm ent secu rities. T h is re d u ce s th e federal fu nd s rate and 31 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 The Borrowings Function: A Supply o r Demand Curve? T h e F ed d oes n o t set quantitative re strictio n s on th e level o f aggregate borrow ings. C onsequen tly, it is u n cle a r w h eth er th e qu antity o f reserves in je cted throu gh th e d isco u n t w indow p rop erly reflects “su p p ly " o r “d em a n d ” co n sid eratio n s. In o th er w ords, d o es th e borrow ings p o rtio n o f the TRScurve in figure 2 rep rese n t a su pply curve, as is co m m o n ly assu m ed , o r a d em an d curve? T h ou g h th e an sw er to th is q u estio n h as no bearin g on th e analysis p resen ted in th e text, th e issu e is in terestin g in its ow n right. It is im p o rtan t to n o te th at d isco u n t w ind ow borrow ing is a privilege for d ep ository in stitu tio n s, not a right. N evertheless, u ntil th e m id-1960s, the d isco u n t w ind ow w as co n sid ere d “o p e n " in the sen se th at th o se w ho w an ted to borrow w ere m ore or less au to m atically gran ted a loan . Only if a bank h ad b ee n d eem ed to have m ad e too frequ en t or in ap p ro p riate u se o f th e privilege w ould it b e “d is co u rag ed " from fu rth er borrow ing. F u rth erm o re, for various reason s, so m e banks m ay be “re lu cta n t to b o rro w ” from th e Fed. M any e co n o m ists believe th at th e re lu cta n ce to borrow is in d u ced by th e F ed 's ad m in istratio n o f th e d is co u n t w ind ow u n d e r w h ich th e am o u n t o f b o r row ed reserves available is d eterm in ed by a co m p li cated system o f p rice an d n o n -p rice rationing. A ccording to th is view, in th e a b se n c e o f ad m in is trative restrictio n s, d ep ository in stitu tio n s w ould borrow from th e Fed w h en ever the d isco u n t rate is belo w th e fed eral fu nd s rate, an d obtain fu n d s in th e federal fund s m arket w h en ev er th e federal fund s rate is below th e d isco u n t ra te.1 T h u s, in the a b se n ce o f any ad m in istratio n o f th e d isco u n t w in dow, th e d isco u n t rate sh o u ld establish an effective ceiling for th e federal fu nd s rate. If true, th e su p p ly o f reserves sh o u ld b e perfectly in tere st-elastic at o r below th e cu rre n t d isco u n t rate. Sin ce d isco u n t borrow ings are sm all (usually b etw een $1 billion an d $2 billion) an d th e federal fund s rate is generally a b o v e th e d isco u n t rate, it is often assu m ed th at th ere is n o n -p rice ration in g at th e d isco u n t w indow . Presum ably, th e m arginal, n o n -p ecu n iary rationing co st is th e am o u n t by w h ich th e federal fund s rate ex ceed s th e d isco u n t rate. T h is sp read is th e m arginal p rem iu m d ep o si 'See Goodfriend (1983). http://fraser.stlouisfed.org/ 32 Federal Reserve Bank of St. Louis tory in stitu tio n s w illingly pay to o b tain fund s from so u rces o th e r th an th e Fed. C on sequen tly, a c c o rd ing to this view, th e borrow ings fu n ctio n is e s s e n tially a su p p ly curve th at ch a n g es as th e Fed ch a n g es its n o n -p ecu n ia ry p rice o f fund s o btain ed at th e d isco u n t w indow . T h is view h as certa in d eficien cies. First, it im plies th at borrow ings sh ould b e zero w h en ev er th e d is co u n t rate is at o r above th e fed eral fund s rate — th at is, th e in tercep t term in eq u atio n 1, b„, sh ould b e zero. E stim ates o f eq u atio n 1, how ever, typically y ield a sign ifican t positive in tercep t term . A ppar ently, d ep ository in stitu tio n s borrow at th e d is co u n t w in d ow even w h en it is th e m ore costly so u rce o f m arginal funds. F o r som e d ep o sito ry in stitu tio n s at least, borrow in gs a p p ea r to b e d e te r m in ed by factors o th e r th an the relative c o s t o f funds, at least as m e a su red by th e fed eral fund s rate/discount rate spread . Riefler (1930) argued th at th e positive in tercep t is d ue to a "n e e d ” th at ca n n o t b e satisfied in the m arket. At on e level, R iefler’s n ee d s th eory cou ld result from a lack o f fin an cial so p h istica tio n . For exam ple, so m e banks m ay seld om , if ever, b o rro w or len d in th e fed eral fu n d s m arket. W h en th e se in sti tu tio n s ex p e rie n ce a reserve d eficien cy , for w h a t ever reason , th ey m ay go to th e F ed ra th er th a n th e u nfam iliar fed eral fun d s m arket, even if th e d is co u n t rate is above the fed eral fun d s rate. Viewed from th is p erspective, th e b orrow ing fu n ctio n is a d em an d curve, at least to th e ex ten t that ch a n g es in d em an d -related facto rs ca u se th e curve to shift. Today, th is exp lan atio n lacks th e cred ib ility it m ight have h ad at R iefler's tim e. T h e in crea sed so p h is tic a tion o f d ep o sito ry in stitu tio n s an d th e w id esp read use o f co rresp o n d en t banking have w eaken ed the validity o f th is argum ent. A different exp lan atio n for th e positive co n sta n t term in th e borrow in gs fu n ctio n is th at th e average federal fund s rate d o es n ot rep rese n t th e relevant op p o rtu n ity co st for all d ep o sito ry in stitu tio n s. Som e in stitu tio n s, for exam p le, m ay b e able to b o r row in th e federal fund s m arket on ly at a very high federal fu n d s rate, if at all. F or th e se in stitu tio n s, d isco u n t w in d o w borrow ings m ay b e attractive, even w h en th e average level o f th e fed eral fu n d s rate is w ell below th e d isco u n t rate. In th is ca s e too, how ever, shifts in th e borrow in gs fu n ctio n reflect ch a n g es in d em an d ra th er th a n supply. FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 F ig u re 3 The Effect of an Increase in the D em and for Total Reserves u n d er a B o rrow ing s P ro ce d u re FFR FFR (o) brings borrow ings back to Borr*; th at is, (TR* —NBR*) eq u als (T R * '—NBR*’), as show n in figure 3b. C o n se quently, n eith e r borrow ings n o r th e fed eral fun d s rate is ch an g ed ; in stead , th e d em an d for total reserves is satisfied by an in crea se in NBR. Alternatively, th e borrow ings fu n ctio n co u ld shift to th e right, illu strated by th e rightw ard shift in th e TRS fu n ctio n in figure 4a. O th er things th e sam e, borrow ings will in crea se and th e equilibriu m federal fund s rate will d eclin e from FFR* to F FR *'. If th e borrow ings assu m p tio n is m ain tain ed at Borr*, th e su p p ly o f n o n borrow ed reserves m u st b e in creased . T h is will put fu rth er dow nw ard p re ssu re on th e fed eral fund s rate until it re a ch es FFR*" (see figure 4b), in d u cin g borrow ings back to Borr*. In th is in sta n ce , borrow ings are u n ch a n g ed an d th e fund s rate falls. The Borrow ings Procedure: An Ineffective Tool f o r M oney Stock Control T h e above analysis su ggests th at strict ad h e re n ce to th e borrow ings p ro ced u re w ill n ot provide effective m o n ey stock co n tro l. If borrow ings are kept at the a ssu m ed level, ch an g es in th e d em an d for m on ey and, h en ce, reserves will be a cco m m o d ated bv c o m p e n sa tory ch an g es in th e su p p ly o f reserves. T h is is illu s (b) trated by th e u sual m o n ey supply/m oney d em an d paradigm sh ow n in figure 5a. Here, th e m o n ey supply is positively related to th e FFR an d is draw n h old ing th e d isco u n t rate an d th e level o f n o n b o rro w ed r e serves u n ch a n g e d .2As usual, th e d em an d for m o n ey is negatively related to th e in terest rate. If th e borrow ings p ro ced u re is u sed to co n tro l th e m o n ey stock, a m o n ey target, M*, m u st b e estab lish ed . Given th e d e m an d for m o n ey an d th e d isco u n t rate, th is requires achieving a sp ecific in terest rate, FFR*. Given th e b o r row ings fu n ctio n , th is im plies a target level o f b o rro w ings (Borr*) an d target settin g for n o n bo rro w ed re serves (NBR*). 2The supply of money is positively sloped on the assumption that borrowings are positively related to the funds rate and that the demand for other reservable components of the monetary base is negatively related to the funds rate. See Thornton (1982b) for a model that incorporates these assumptions. The federal funds rate is not used commonly as the representa tive interest rate for money demand. But it is used commonly in money supply models as well as those that incorporate both money supply and money demand functions. This may not be desirable, especially if the relationship between the funds rate and the true representative rate in the money demand equation is either highly variable or affected by changes in policy or policy-related variables. It is adequate, however, if there is a fixed proportionate relationship between these rates. 33 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 F igu re 4 The Effect of an Increase in the Borrowings Function under a Borrowings Procedure FFR NBR FFR* FFR* T R * TR*' (a) Su p p o se th e d em an d for m oney in crea ses from Md to M',,.3 O th er th in gs th e sam e, th e resu ltin g rise in the fu nd s rate w ould in cre a se borrow ings. In o rd er to red u ce borrow ings back to Borr*, th e su pply o f n o n borrow ed reserves m u st in crease from NBR’ to NBR*', shifting the m on ey supply sch e d u le to the right as sh ow n in figure 5a. B eca u se borrow ings d ep en d only on th e level o f th e federal fund s rate (given th e d is co u n t rate and assu m in g the fu n ctio n is oth erw ise stable), th e d esired level o f b o rro w in g s ca n he achieved by supplying th e requ isite qu an tity o f n o n b orrow ed reserves. H ence, all shifts in th e d em an d for m on ey are a cco m m o d a ted bv co rresp o n d in g shifts in th e m oney su pply if Borr* rem ain s u n ch an g ed . In this case, no differ e n c e b etw een m o n ey stock co n tro l u n d er a borrow ings p ro ced u re and a fed eral fund s rate targeting p ro ced u re exists. 3Total reserves demand is composed of the demand for required and excess reserves. The demand for required reserves can be thought of as a derived demand, derived from the demand for money via the relationship between checkable deposits and required reserves. Because the demand for money generally is estimated to be interest-inelastic, the demand for required reserves should also be interest-inelastic. Of course, during the lagged reserve accounting (LRR) period prior to February 1984, the demand for required reserves should be perfectly interest-inelastic. (See Thornton (1983) for a discussion of the differences between lagged and contemporaneous reserve accounting.) The demand for excess reserves usually is found to be relatively interest-insensitive as well. http://fraser.stlouisfed.org/ 34 Federal Reserve Bank of St. Louis (b) Now, su p p o se th e borrow ings fu n ctio n tem porarily in crea ses, th at is, v, > 0. T h is p ro d u ces a tem p orary rightw ard shift in th e m o n ey su p p ly from M s to M 's, as illu strated in figure 5b.4 O th er things th e sam e, the feder al fund s rate falls an d borrow ings in crea se. To bring borrow ings b a ck to Borr*, n o n b o rro w ed reserves m u st b e in crea sed , shifting th e m o n ey su p p ly s c h e d ule still fu rth er to th e right from M 's to M"„. As a result, the m o n ey stock is fu rth er aw ay from its targeted level, M*. Str ictly en fo rced , th e borrow ings o p eratin g p ro c e dure yield s less sh o rt-term co n tro l over th e m o n ey stock th an a straight forw ard fed eral fund s rate target ing p ro ced u re as long as th e borrow ings fu n ctio n is su b je ct to som e variability an d th e Fed m akes no allow ance for the shift.5 The Borrowings Procedure as a Federal Funds Rate Target T h e borrow ings p ro ced u re p ro d u ces resu lts that are id en tical to th o se u sin g a fed eral fu n d s rate target ing p ro ced u re if all sh o ck s em a n a te from th e d em an d s for m o n ey o r reserves. T h e two p ro ced u res y ield dif- 4The assumption here is that the discount window is assumed “open,” given a set of unchanged administrative constraints. 5This point has also been made by VanHoose (1988). FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Figu re 5 The Effect of an Increase in the Demand for M oney and the Borrowings Function on the M oney Supply under a Borrowings Procedure feren t resu lts, how ever, w h en th ere are shifts in th e borrow ings fu n ctio n . T h e borrow ings p ro ced u re e x a c erb ates the effect o f sh o rt-ru n flu ctu atio n s in the b o r row ings fu n ctio n on th e funds rate.1’ If an y part o f su ch shifts is offset bv ch an g es in NBR, th e borrow ings p ro ced u re will p ro d u ce g reater variability in th e fed eral fund s rate and less variability in borrow ings th an a d irect fund s rate targeting p ro ced u re. Ind eed , u n d e r a borrow ings p ro ced u re, th e variability o f borrow ings will b e less th an th e variability o f th e fund s rate u n d er fairly general con ditions/ N evertheless, if th e bo rro w ings fu n ctio n is stable (in th e sen se th at all flu ctu a tio n s are transitory), flu ctu atio n s in th e fed eral fund s rate will n et out over tim e [E(vt) = 0]; th erefore, the borrow ings o p eratin g p ro ced u re ca n b e u sed to achieve a fed eral funds rate target over a som ew h at lon g er-term h o rizo n .8 O f co u rse, th e borrow ings fu n ctio n also co u ld ex h ibit p e rm a n en t shifts a sso cia te d w ith ch a n g es in b0 in eq u atio n 1. In th is in sta n ce, th e a ssu m ed level of borrow ings w ould be ach ieved only w ith a su bstan tial ch a n g e in th e fed eral fund s rate.‘J F or exam ple, if borrow ings are m ain tain ed at a p red eterm in ed level d esp ite a p e rm a n en t d ecre a se in th e borrow ings fu n c tion, n o n b o rro w ed reserves m u st b e re d u ce d u ntil th e fed eral fu n d s rate rises en o u gh to retu rn borrow ings to th eir form er level. On th e o th e r hand , if th e fed eral fu n d s rate is kept at its fo rm er level, th e borrow ings assu m p tio n m u st b e low ered . 6Roley (1986) argues that, while the Fed moves quickly to offset changes in the demand for total reserves, it does not do so for changes in the borrowings function. Thus, he argues that the federal funds rate will vary more in the short run under a borrowed-reserves procedure than under a federal-funds-rate targeting procedure. Roley’s assertion implies the Fed can distinguish between shifts in these two functions. 8The two procedures are equivalent if the shocks to both the demand for total reserves and borrowings exhibit no persistence and if no attempt is made to offset such temporary shocks. See appendix A for details. 7The relative variability of borrowings will be less if the slope of the total reserves supply function in figure 2 is flatter than that of the demand for total reserves — a condition that is likely to hold — or if the Fed is reasonably successful in offsetting the effect of shifts in the borrowings function. See appendix A for details. 9Wallich (1984), p. 26, notes that the borrowings function is unstable. Therefore, he contends the borrowings procedure cannot be re garded as a form of "rate-pegging,” because the " . . . chosen level of borrowing is consistent with any range of values of the funds rate.” 35 FEDERAL RESERVE BANK OF ST. LOUIS In sum m ary, th e borrow ings o p eratin g p ro ced u re is a useful su rrogate for a sh o rt-ru n federal fu nd s rate targeting p ro ced u re o n ly if all ch an g es in aggregate borrow ings are p ro d u ced by shifts in th e d em an d for total reseives. It is a useful su rrogate for a longer-ru n fed eral fu nd s rate targeting p ro ced u re only if th e b o r row ings fu n ctio n is stable, th a t is, su b je ct only to tem porary, ran d o m shifts. It is u n su ited for a federal fund s rate target w henever there are p erm an en t shifts in th e borrow ings fu n ctio n , u n less th e borrow ings assu m p tio n is ch a n g ed sufficiently. The Policy Implications o f a Change in the Borrowings Assumption Usually, m onetary aggregates or in tere st rates are ch o se n as in term ed iate policy targets. W hy is the borrow ings p ro ced u re u sed w h en th e m o n ey sto ck or th e fund s rate co u ld be m o re d irectly co n tro lled by o th e r p ro ced u res? W hat are th e policy im p licatio n s o f a ch an g e in th e borrow ings assu m p tio n ? W ithout p re cise in form ation abou t th e in term ed iate p o licy target, it is difficult to an sw er th e se q u estio n s definitively; n ev ertheless, so m e g en eralizatio n s ca n be m ad e. If th e borrow ings fu n ctio n is stable, an in crea se in th e borrow ings assu m p tio n can be in terp reted as a m ove tow ard restrain t in th at it red u ces th e su p p ly of reserves relative to d em an d . Conversely, a d e cre a se is a m ovem ent tow ard ease. If th e borrow ings fu n ctio n is u nstable, in th e sen se th at p erm an en t shifts o ccu r, how ever, ch an g es in th e borrow ings a ssu m p tio n m ay reflect th e F e d ’s aw aren ess o f th ese shifts an d its d esire to m itigate th e ir effect o n th e funds rate. A failure to ch an g e th e borrow ings assu m p tio n , on th e o th e r hand , co u fd be in terp reted as a m ovem en t to w ard ease o r restraint, d ep en d in g on th e d irectio n of th e shift o f th e borrow ings fu n ctio n . The Relationship Between Changes in the Borrowings Assumption and Changes in the Discount Rate C hanges in th e borrow ings a ssu m p tio n an d th e d isco u n t rate can be view ed as su b stitu tes. B eca u se it d ep en d s on th e d isco u n t rate, th e T B Scurve shifts w ith a ch an g e in th e d isco u n t rate. F o r exam ple, a d isco u n t rate in crea se shifts th e slop ed p o rtio n o f th e TRS curve to th e left at all levels o f th e fu nd s rate. If th e b o rro w ings a ssu m p tio n is u n ch an g ed , th e qu antity o f n o n b orrow ed reserves m u st be re d u ce d u ntil th e funds rate rises en o u gh to resto re borrow ings to th e ir d e sired level. On average, th e fed eral fu n d s rate will http://fraser.stlouisfed.org/ 36 Federal Reserve Bank of St. Louis JANUARY/FEBRUARY 1988 ch an g e p o in t-fo r-p o in t w ith a ch a n g e in th e d isco u n t rate u n d e r a strictly en fo rced borrow ings p ro ced u re. T h e sam e ch a n g e in th e eq u ilibriu m fed eral fu n d s rate co u ld b e o b tain ed by ch an g in g th e borrow ings assu m p tio n in stead . C on sequen tly, ch a n g es in th e borrow ings assu m p tio n an d ch a n g es in th e d isco u n t rate are su b stitu tes in th e ir effect on th e fed eral fund s rate u n d e r a strictly en fo rced borrow ings p ro ced u re."’ Table 1 rep o rts ch a n g es in th e d isco u n t rate an d the borrow ings a ssu m p tio n from O cto b er 1982 th rou gh D ecem b er 1986. T e ch n ic a l d isco u n t rate ch an g es, re p orted ly m ad e s o le ly to keep th e d isco u n t rate in lin e w ith m arket in terest rates, are d en o te d by a T; th o se m ad e for o th er, policy -related , rea so n s are d en o ted by a P ." As th e table show s, ch a n g es in th e borrow ings assu m p tio n an d th e d isco u n t rate gen erally o ccu rred aro u n d th e sam e tim e: five of th e 11 ch a n g es in the d isco u n t rate cam e w ith in abou t o n e w eek o f a ch an g e in th e borrow ings assu m p tio n , w hile tw o w ere w ith in tw o w eeks. M oreover, all ch a n g es th at o cc u rre d clo se to g eth er w ere in th e sam e d irectio n , in d icatin g c o n sisten t m ovem en ts in b o th th e borrow ings a ssu m p tio n an d th e d isco u n t rate. T h e table also show s altern atin g p eriod s o f ea se and restrain t. From O cto b er 1982 th rou gh th e en d o f the year, th e borrow ings a ssu m p tio n an d th e d isco u n t rate w ere red u ced . W hile ch a n g es in th e borrow ings a ssu m p tio n w ere m o d est (even cum ulatively) an d on e d isco u n t rate ch a n g e w as te ch n ica l, p o licy eased m od erately d uring th is period. From spring 1983 to sprin g 1984, p o licy m oved to w ard restrain t. T h e borrow in gs a ssu m p tio n w as raised bv $600 m illion from M arch th rou gh August 1983, low ered in O cto b er 1983 by $150, th e n in crea sed by $350 m illion in M arch 1984. T h e last in cre a se w as follow ed clo sely by a 50 b a sis-p o in t te ch n ica l in crea se in th e d isco u n t rate. Policy w as ea sier d uring th e fall o f 1984. T h e b o rro w ings assu m p tio n w as red u ced by $700 m illion from early O cto b er to late D ece m b er an d two policy-related cu ts in th e d isco u n t rate re d u ce d it by a full p e rc e n t age point. T h e re w ere no large, co n siste n t m ovem ents in th e borrow ings a ssu m p tio n d uring 1985 an d n o n e after early Febru ary 1986, d esp ite fou r cu ts in th e d isco u n t rate (three o f w h ich w ere p olicy-related ). '“Because a one percentage-point change in the discount rate is associated with about a $420 million change in borrowings over this period, a $420 million change in the assumed level of borrowings should have an effect on the funds rate equal to a one percentagepoint change in the discount rate. See Thornton (1986). "See Thornton (1986,1982a) for a discussion of the classification of discount rate changes into technical and non-technical changes. JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS Table 1 Changes in the Discount Rate and the Borrowings Assumption Change in the Change in the discount rate borrowings assumption (in basis points) (in millions of dollars) October 6,1982 October 12,1982 November 17, 1982 November 22, 1982 December 14, 1982 December 22,1982 $ -5 0 -5 0 T -5 0 -5 0 P - 50 P -5 0 50 100 100 250 100 -1 5 0 March 30,1983 May 25, 1983 June 24, 1983 July 14, 1983 August 24, 1983 October 5, 1983 March 28, 1984 April 9, 1984 October 3,1984 November 8,1984 November 23,1984 December 19, 1984 December 24, 1984 February 14, 1985 March 27, 1985 May 20, 1985 May 22,1985 August 21, 1985 October 2, 1985 November 6, 1985 December 18,1985 February 13, 1986 March 7, 1986 April 21, 1986 July 11, 1986 August 21, 1986 350 50 T -2 5 0 -1 7 5 -5 0 P -2 7 5 - 50 P 50 50 - 50 P -5 0 75 75 -5 0 -1 0 0 -5 0 -5 0 -5 0 -5 0 -5 0 P T P P tion in borrow in gs abou t its m ean level.'2 F or th e b orrow ings p ro ced u re to b e u sed effectively as a fed eral fund s rate target over th e lo n g er run, th e borrow ings fu n ctio n m u st b e stable. It is im portan t, therefore, to d eterm in e w h eth er th ere have b ee n p erm an en t shifts in th e borrow ings fu n ctio n . If so, th e key issu e is h ow th e borrow in gs a ssu m p tio n w as ch a n g ed in re sp o n se to th e se shifts. To ex am in e this issu e, eq u atio n 1 w as estim ated using rand om co efficien t regression , w h ere b o th the co n sta n t term , b„, an d th e slop e coefficien t, b„ are allow ed to vaiy th rou gh tim e .13 C hart 1 p resen ts ran d o m -co efficien t-reg ressio n estim a tes o f th e c o n stant term an d a b an d re p resen tin g plus or m inu s one stan d ard error. T h e vertical lin es sh ow the d ates on w h ich th e borrow ings a ssu m p tio n w as raised o r low ered, as in d icated . T h e in tercep t show s co n sid era b le variability. W ith b u t th ree sign ifican t ex cep tio n s, th e borrow in gs a s su m p tio n w as ch a n g ed in th e d irectio n co n siste n t w ith m itigating th e effect o f shifts in th e borrow ings fu n ctio n on th e fed eral fu n d s ra te.14 T h e th ree e x c e p tio n s o cc u rre d in O cto b er 1984, August 1985 an d O cto b e r 1985. In O cto b er 1984, w h en th e borrow in gs fu n c tion sh ifted upw ard, th e borrow ings a ssu m p tio n w as red u ced from $1 b illio n to $750 m illion. In b o th August a n d O cto b er 1985, th e borrow ings a ssu m p tio n w as raised, d esp ite th e d ow nw ard shift in th e borrow ings fu n ctio n . B oth in crea s es w ere relatively sm all ($75 m illion each ), how ever, an d b o th w ere co m p letely offset by th e m id -D ece m b er d ecrea se. T h e se resu lts are co n siste n t w ith m ovem en ts in borrow ings, th e borrow ings assu m p tio n an d th e fed eral fund s rate p re sen ted in ch art 2. T h e O cto b er 1984 ch a n g e in th e borrow ings a ssu m p tio n p reced ed m ovem en ts in b orrow ings; how ever, th is a ctio n fol low ed a 100 b a sis-p o in t drop in th e fund s rate from its T Indicates changes made solely to keep the discount rate in line with market interest rates P Indicates changes made for other, policy-related, reasons. 12This is for the period from October 1982 through June 11,1986. See Thornton (1986). This same function estimated for the 222 weeks prior to October 6,1979, has an R2of about .70. EMPIRICAL EVIDENCE ON THE USE OF THE BORROWINGS PROCEDURE An im p o rtan t fa cto r in d eterm in in g th e im p a ct of th e borrow ings p ro ced u re on th e fu nd s rate and the m o n ey sto ck is th e stability o f the borrow ings fu n c tion . H istorically, th e borrow ings fu n ctio n has b een s u b je ct to co n sid erab le ran d o m variation: by itself, the sp read betw een th e fed eral funds rate an d the d is co u n t rate exp lains less th a n 50 p e rce n t o f th e varia 13The procedure used here is suggested by Garbade (1977). The equation was first estimated allowing only the constant term to vary. It was then reestimated allowing both the constant and slope coeffi cients to vary; this was done to determine whether variation in the slope coefficient was being inappropriately attributed to the constant term. The results presented in chart 1 are from the latter estimation. The qualitative interpretation of the relationship between changes in the borrowings assumption and shifts in the borrowings function was not affected by the different estimation procedures. ,4There were two other exceptions: they occurred on October 6,1982, and May 22, 1985. In both instances, however, these changes predate the shift by only one week. Including these in subsequent statistical tests does not affect the results. 37 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 C h a rt 1 V a r y i n g P a r a m e t e r Intercept .11________________ Jan .'82 The b lu e v e r tic a l lin e s The g re y v e rtic a l line s re fe r to p o s itiv e c h a n g e s in th e b o rro w in g s assu m p tio n . F u rth er evid ence on th e effects o f using th e borrow ings p ro ced u re ca n be o btain ed by analyzing sca tte r p lots o f borrow ings an d th e federal fund s rate during p erio d s in w h ich b o th th e borrow ings assu m p tio n an d th e d isco u n t rate w ere u n c h a n g e d .“ T h e previous resu lts suggest th at th ere are co n sid era b le tem p orary shifts in th e d em an d for bo rro w ed reserves. If m o st o f th e effect o f sh o rt-ru n variation in th e borrow ings fu n ctio n on th e federal fu n d s rate w ere offset quickly, th ere w ould be little variation in th e fed eral fund s rate bu t co n sid era b le variation in borrow ed reserves. In th e extrem e, if th e effect o f all su ch shifts on th e fund s rate w ere quickly an d co m p letely offset, all observa tio n s w ould lie along a v ertical lin e rep resen tin g th e average o f th e fed eral fund s rate in a sc a tte r plot of borrow ings an d th e fed eral funds rate. On th e o th er hand , if borrow ings w ere kept clo se to th e a ssu m ed 15The FOMC meeting was held on October 2,1984. The federal funds rate had fallen to 9.84 percent on September 26, though it averaged 10.73 percent for the week ending Wednesday, September 26. The weekly peak was 11.77 percent for the week ending Wednesday, August 22; the daily peak occurred on August 1, when the federal funds rate was 12.04 percent. >6This procedure was suggested to me by R. Alton Gilbert. It is interesting to note that the variability of borrowings could be reduced by simply “tying” the discount rate to the federal funds rate. This point was made by Thornton (1982b) and more recently by VanHoose (1987). cy clical peak for th e w eek end ing August 2 2 .15 Nearly all o th e r ch an g es in th e borrow ings assu m p tio n w ere p re ce d e d by m ovem ents in borrow ings an d th e fed eral fu n d s rate in th e sam e d irectio n . A Com parison o f the Variability o f Borrowings and the Federal Funds Rate http://fraser.stlouisfed.org/ 38 Federal Reserve Bank of St. Louis FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 C h a rt 2 B o r r o w i n g s , B o r r o w i n g s A s s u m p t i o n , a n d F e d e r a l F unds R a t e Perce nt M i l l i o n s of d o l la r s 5000 4000 rate CALE^ 3000 2000 Borrow ings ^S C A LE 1 1000 Borrow ings assumption ISCALE -1000 -2000 level, the funds rate sh o u ld vary relatively m ore th an borrow ings. In this case, th e o bserv ations sh o u ld be clu stered abou t a h o rizo n tal lin e at the assu m ed level for borrow in g s.17 D uring th e p o st-O cto b e r 1982 period, th ere w ere six p eriod s w ith 10 or m ore w eeks in w h ich b o th the borrow ings a ssu m p tio n and th e d isco u n t rate w ere u n c h a n g e d .S c a tt e r plots of borrow ings and the fed eral fund s rate for th ese p erio d s are p re sen ted in l7The variability of borrowings and the funds rate depend on the slopes of the TRSand TRd curves and the extent to which random shocks are offset. If more than 50 percent of such shocks are offset during the period, however, there will be more variability in the funds rate than in borrowings regardless of the slopes of these curves. ,aPlots for the omitted periods show no pattern. They consist, how ever, of very few observations. ch a rts 3a th rou gh 3f.19 T h e data u sed in th e se ch arts have b een n o rm alized . T h e actu al level o f borrow ings w as n o rm alized by dividing it by the level o f the borrow ings a ssu m p tio n for th e resp ectiv e per iod. T h e fed eral fund s rate w as n o rm alized by dividing it by its average rate for th e p erio d .20 All ch a rts have id en tical sca les for b o th variables to m ake it easy to co m p a re th e relative variability. T h e solid h o rizo n tal an d vertical lin es d en o te w h ere th e n o rm alized variables are equal 19These data exclude outliers such as the “window-dressing” borrow ings during the final reserve period of the year and the unusually large borrowings associated with Continental Bank of Illinois. See Thornton (1986) for a discussion of the latter episode. “ Because the mean of the normalized rate spread equals one, the rate spreads will be scattered symmetrically about the vertical line. In contrast, the data points will be scattered asymmetrically above (below) the vertical line depending on whether the borrowings assumption is below (above) the average level of borrowings for the period. 39 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 C hart 3 Selected Scatter Plots of Normalized Borrowings (11=12) Decem ber 29. 1982 - March 23, 1983 October 12, 1983 - M arch 21, 1984 (N=23) M illio n s of d o lla rs M i l l i o n s of d o l l a r s P e rc e n t P e rc e n t (a) (b) M ay 29, 1985 (N= 12) Ap ril 18, 1984 - September 26, 1984 (N=20) - August 14, 1985 M i l l i o n s of d o l l a r s J IOU 2930 • 2680 • 24 30 21 80 19 40 — • • 1690 1440 1190 - • •• 940 700 — • 45 0 0.85 0.8 8 0.91 0.94 0.97 0 .0 0 P e rc e n t 1.03 1.0 6 1 .0 9 "1 200 0.85 1 1 1 1 1 1 1 1 1 1 1 1 1 0.88 0.91 0.94 1 1 1 1 I I 0.97 0 .0 0 P e rc e n t 1.03 1.0 6 1 I 1 1 .0 9 (d) The vertical re fe re nc e lines refer to the n o r m a l i z e d me an of the f e d e r a l funds rate; the h o r i z o n t a l reference lines refer to the n o r m a l i z e d m e an of borrowi ngs. http://fraser.stlouisfed.org/ 40 Federal Reserve Bank of St. Louis JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS and the Normalized Federal Funds Rate August 27, 1986 - December 24, 1986 (N=18) April 30, 1986 - July 9, 1986 (N=11) 0.85 0.88 0.91 0.94 0.97 0 .0 0 P e rc e n t 1.03 1.06 1 .0 9 0.85 0.88 0.91 0.94 0.9 7 0 .0 0 P e rc e n t 1.03 1.06 1. 0 9 (f) to on e. Som e d escrip tive sta tistics for th e raw d ata are p re sen ted in table 2. December 27, 1978 - July 18, 1979 (N=30) Finally, ch art 3g is a sc a tte r plot from late D ecem b er 1978 to late Ju ly 1979, w h en it is gen erally ack n o w l edged th at th e Fed w as targeting th e fed eral funds rate. D uring this period , a 75 b asis-p o in t target range for the fed eral fu n d s rate w as sp ecified .21 W ith th e ex cep tio n p erh a p s o f ch art 3c, no period su ggests a rapid a d ju stm en t to m ain tain borrow ings at th e a ssu m ed level. In co n trast, tw o p eriod s (charts 3a an d 3e) sh ow relatively little variability in th e fed eral fund s rate. In d eed , a co m p a riso n o f th e se ch arts w ith ch art 3g show s th at th e fu n d s rate flu ctu ated less aro u n d its m ean d uring th e se p eriod s th a n it did aro u n d th e m id p oin t o f th e Fed ’s narrow range for the federal fund s rate in early 1979. T h ere sh o u ld b e less variability in borrow ings an d m ore variability in th e fed eral fund s rate u n d e r a 0.85 0.88 0.91 0.94 0.97 0 .0 0 P e rc e n t (g) 1.03 1.06 1 .0 9 21T h is w a s th e on ly e x te n d e d pe riod in w h ic h th e federal fu n d s rate b a nd w a s bo th n a rro w an d u n chang ed. 41 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Table 2 Descriptive Statistics for Borrowings and the Federal Funds Rate Borrowings Period Dec. 27,1978July 25, 1979 Federal funds rate Mean1 SD2 CV Mean SD CV $1,190.47 $367.0 .3083 10.16% .1549 .0153 Dec. 29, 1982March 2 3 ,19832 326.3 165.9 .5084 8.56 .1176 .0137 Oct. 12,1983March 21, 1984 688.0 214.9 .3124 9.51 .1863 .0196 April 18,1984Sept. 26, 19843 951.3 301.8 .3173 11.15 .4935 .0443 May 29, 1985Aug. 14,1985 559.0 213.6 .3821 7.74 .2618 .0338 April 30, 1986July 9,1986 284.9 79.6 .2794 6.89 .0525 .0076 Aug. 27, 1986Dec. 24, 1986 343.8 184.7 .5372 5.98 .1742 .0291 NOTE: SD denotes the standard deviation. CV denotes the coefficient of variation, i.e., SD/Mean. 'In millions of dollars. 2Excludes the “window-dressing” borrowings for the week ending January 5, 1983. 3Excludes the four weeks of unusually large borrowings (May 16-June 6) associated with the problems of Continental Bank of Illinois. borrow ings target th an u n d e r an in terest rate target. T able 3 p re sen ts the m ean, stan d ard deviation ISD) and a m easu re o f relative variability, the co efficien t o f variation (CV), for w eekly d ata d uring th e p eriod o f the borrow ings o p eratin g p ro ced u re and d uring an equal n u m b er o f w eeks u n d e r an in terest-rate-targ etin g re gim e. T h e resu lts are gen erally co n siste n t w ith th o se d iscu ssed above. T h e variability o f borrow ings differs little in eith e r abso lu te o r relative term s b etw een th e tw o p eriod s. T h e variability o f th e fed eral fu nd s rate, how ever, fell consid erably; its SD d eclin ed nearly 30 p ercen t, w hile its CV d eclin ed n early 50 p e rce n t.22 The Im pact o f Changes in the Borrowings Assumption on the Federal Funds Rate If ch an g es in th e borrow in gs assu m p tio n w ere m ad e prim arily to offset shifts in th e borrow ings fu n c tion, th ere sh o u ld be no sign ifican t relatio n sh ip b e 22This result is only marginally affected by the switch from a one-week to a two-week reserve accounting period. If only reserve period data are used for the CRR period, the standard deviation of the funds rate is 1.60 percent and the coefficient of variation is .19. http://fraser.stlouisfed.org/ 42 Federal Reserve Bank of St. Louis tw een ch a n g es in th e borrow in gs a ssu m p tio n and m ovem en ts in th e federal fund s rate. If ch a n g es in the borrow ings a ssu m p tio n are m ad e for o th e r reason s, th ey sh o u ld p ro d u ce a sign ifican t effect on th e federal funds rate.2:1 23This is true only if a discount rate change shows significant direct effect on the federal funds rate; Thornton (1986) argues such an effect should be small and insignificant. Indeed, this may provide an expectations-effect-free method of assessing the direct effect of a discount rate change on market interest rates. See Thornton (1986) for discussion of three potential effects of a change in the discount rate on the federal funds rate. Because the Federal Reserve makes a public statement when it changes the discount rate, it is difficult to separate the direct and announcement effects. In contrast, the levels of the borrowings assumption for the previous calendar year are made public in the Spring issue of the Federal Reserve Bank of New York Quarterly Review. Because appropriately scaled changes in the discount rate and the borrowings assumption have equivalent announcementfree effects on the supply of credit in the market, the direct effect of discount rate changes can be gauged by investigating the effect of changes in the borrowings assumption if the Federal Reserve moves quickly to stabilize the level of borrowings at the new as sumed level. If changes in the borrowings assumption are made to offset the effect of shifts in the borrowings function on the federal funds rate, they will not produce a significant effect on the federal funds rate and will not provide an announcement-effect-free test of the direct effect of a discount rate change. JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS To investigate this, th e follow ing eq u atio n w as estim ated : L (3) AFFR, = ot„ + S c^ A F F R ,., + p„ADRT, i= l K + P.ADRN T, + S iXiABA,., + e,. i= 0 Here, AFFR d en o tes th e ch an g e in th e fed eral funds rate, ADRT and ADRNT d en o te "te c h n ic a l” and “n o n te ch n ica l'' ch an g es in the d isco u n t rate, ABA d en o tes th e ch an g es in th e borrow ings assu m p tio n an d e d en o tes a rand om erro r term . A ch an g e in th e b o rro w ings assu m p tio n w as assu m ed to be effective th e day after th e d ecisio n w as m ad e.24 T h is eq u atio n w as e sti m ated u sing ordinary least squ ares (OLS) for th e p e riod from O cto b er 1 ,1 9 8 2 , throu gh D ece m b er 3 1 ,1 9 8 6 ; how ever, th e eq u atio n w as estim ated sep arately for the period o f lagged reserve requ irem en ts, LRR (up to Febru ary 1, 1984) an d co n tem p o ra n e o u s reserve re qu irem en ts, CRR.23 T h e eq u atio n w as estim ated using daily, w eekly an d reserve-p eriod data (one w eek b e fore Febru ary 1, 1984, and two w eeks thereafter). B eca u se it is n o t know n h ow quickly ch an g es in th e borrow ings assu m p tio n are im p lem en ted or how rap idly th e fed eral fun d s rate m ight resp o n d , lags o f ABA w ere in clu d ed ; how ever, th e F -statistic for in clu d in g lagged values o f th e ABA, Faoalacs, show s th at th e rela tio n sh ip b etw een th e AFFR and ABA is co n te m p o ra n e ous regard less o f w h eth er daily, w eekly or reserveperiod data are u sed .28 T h e resu lts are rep o rted in table 4. T h ey in d ica te a statistically sign ifican t p o s itiv e relatio n sh ip betw een ch a n g es in th e funds rate an d ch an g es in th e b o rro w ings assu m p tio n only for w eekly data during th e CRR period. A fu rth er investigation o f this relatio n sh ip , how ever, show s it to be qu ite fragile (see ap p en d ix B 24AII changes in the borrowings assumption but one were made at regularly scheduled meetings of the FOMC. 25The equation was estimated for separate periods for several rea sons. First, it would be inappropriate to estimate the equation using reserve-period data for the entire sample period with OLS because the error terms of one-week and two-week average data will be different and OLS would not reflect the heteroskedasticity induced by the change in the reserve accounting period. Second, the coeffi cients do not appear to be stable overthe entire period as the results of table 4 suggest. Third, there is a pronounced quarterly seasonal spike during the LRR period (as is readily evident in chart 2) that is not statistically identifiable during the CRR period. Finally, there is low-order autocorrelation in the error term during the CRR period which is not evident during the LRR period. 26The exception was for daily data during the CRR period; however, the sum of the coefficients was not significantly different from zero. Table 3 Descriptive Statistics on Borrowings and the Funds Rate under Interest Rate and Borrowings Operating Procedures_____ Statistic Borrowings Federal funds rate July 9 ,1 9 7 5 - October 3,1979 Mean SD CV $567 533 .94 6.88% 2.09 .30 October 6,1982 - December 31,1986 Mean SD CV $660 571 .87 8.60% 1.42 .17 NOTE: SD denotes the standard deviation. CV denotes the coefficient of variation, i.e., SD/Mean. for details). H ence, th ere is n o strong, statistically sig n ifican t relatio n sh ip b etw een ch a n g es in th e borrow ings a ssu m p tio n an d ch a n g es in th e fu n d s rate. T h ese resu lts are c o n siste n t w ith th e previous o n es th at ch a n g es in th e borrow ings a ssu m p tio n w ere m ad e prim arily to a cco m m o d a te shifts in th e borrow ings fu n ctio n .27 SUMMARY AND CONCLUSIONS T h is p ap er a sse sses th e u sefu ln ess o f th e b o rro w ings op eratin g p ro ced u re in co n tro llin g th e m oney stock o r th e in terest rates. T h e borrow ings p ro ced u re is a p o o r m eth o d for achieving m o n ey sto ck co n tro l. In fact, a fed eral fund s rate targeting p ro ced u re is su p e rior for b o th m o n ey stock an d in tere st rate con trol. I’h e borrow ings p ro ced u re is an effective m ean s of targeting th e fed eral fu n d s rate in th e sh o rt run only w h en the variation in borrow in gs is due solely to shifts in th e d em an d for total reserves. It is an effective m ean s o f targeting th e fed eral fu n d s rate over lo n g er p erio d s only w h en th e borrow in gs fu n ctio n is stab le. If th ere are p e rm a n en t shifts in th e borrow ings fu n ction , ^Alternatively, these results could be interpreted as evidence that the announcement-free, “direct effect” of a discount rate change on the federal funds rate is nil. 43 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Table 4 Estimates of Equation 3 October 1,1982February 1,1984 February 2,1984December 31,1986 Reserve period’ Daily Weekly -0.01 (0.46) 0.04 (1.02) 0.01 (0.44) 0.01 (0.26) 0.03 (0.63) ADRT, 1.32* (2.08) 0.49 (0.55) 0.13 (0.19) 0.23 (0.30) -0 .0 5 (0.08) ADRNT, 0.85* (2.09) 1.59* (2.65) 0.72 (1-82) 0.69 (1.78) 0.80* (1.96) ABA, 0.0004 (0.50) 0.0005 (0.50) 0.0019* (2.38) 0.0004 (0.58) Daily Constant p 'AFFRLAG S 2 8.44* 15.36* -0.0018* (2.12) 9.82* 7.01* Reserve period 0.34 F 3 r ABALAGS 1.27 1.80 2.56* 1.53 1.91 R2 0.1854 0.5302 0.1188 0.1640 0.0397 SEE 0.2860 0.2942 0.4709 0.3815 0.3247 Indicates statistical significance at the 5 percent level; two-tailed test. 'During the LRR period, the reserve period was one week. 2F-statistic for the lagged values of AFFR. A quarterly seasonal was included for weekly data for the LRR period. 3F-statistic for lagged values of ABA. The reported results are for an equation that did not include lagged values of ABA if they were not significant. th e fed eral fu n d s rate w ill vary w ith shifts in th e borrow ings fu n ction , an d th e borrow ings p ro ced u re ca n be u sed to target th e fed eral funds rate only if co m p en sato ry ch an g es in th e borrow ings assu m p tio n are m ade. Evidence in d icates that the borrow ings fu n ctio n is u n stab le. Also, it suggests that generally th e b o rro w ings assu m p tio n h as b e e n ch an g ed in th e d irectio n th at offsets th e effect o f p e rm a n en t shifts in th e b o r row ings fu n ctio n on th e federal funds rate. "The Federal Dickey, David A., and Wayne A. Fuller. “ Distribution of the Estima tors for Autoregressive Time Series With a Unit Root,” Journal of the American Statistical Association (June 1979), pp. 427-31. Engle, Robert F., and C. W. J. Granger. “Co-Integration and Error Correction: Representation, Estimation, and Testing,” Econometrica (March 1987), pp. 251-76. http://fraser.stlouisfed.org/ 44 Federal Reserve Bank of St. Louis Garbade, Kenneth. “Two Methods for Examining the Stability of Regression Coefficients,” Journal of the American Statistical Asso ciation (March 1977), pp. 54-63. Gilbert, R. Alton. “Operating Procedures for Conducting Monetary Policy,” this Review (February 1985), pp. 13-21. Goodfriend, Marvin. “ Discount Window Borrowings, Monetary Pol icy, and the Post-October 6, 1979 Federal Reserve Operating Procedure,” Journal of Monetary Economics (September 1983), pp. 343-56. Plosser, Charles I., G. William Schwert and Halbert White. “ Dif ferencing as a Test of Specification,” International Economic Re view (October 1982), pp. 535-52. REFERENCES Board of Governors of the Federal Reserve System. Reserve Discount Window,” 1980. Federal Reserve Bank of New York. “ Monetary Policy and Open Market Operations in 1985,” Quarterly Review (Spring 1986), pp. 34-53. Polakoff, Murray E. “ Reluctance Elasticity, Least Cost, and Member-Bank Borrowing: A Suggested Integration,” Journal of Finance (March 1960), pp. 1-18. Riefler, Winfield. Money Rates and Money Markets in the United States (Harper and Brothers, 1930). Roley, V. Vance. “ Market Perceptions of U.S. Monetary Policy since 1982,” Federal Reserve Bank of Kansas City Economic Review (May 1986), pp. 27^t0. JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS Thornton, Daniel L. "The Discount Rate and Market Interest Rates: Theory and Evidence," this Review (August/September 1986), pp. 5-21. VanHoose, David D. “A Note on Discount Rate Policy and the Variability of Discount Window Borrowing," Journal of Banking & Finance (December 1987), pp. 563-70. _________ “ Lagged and Contemporaneous Reserve Accounting: An Alternative View,” this Review (November 1983), pp. 26-33. _________ “ Discount Rate Policy and Alternative Federal Reserve Operating Procedures in a Rational Expectations Setting,” Board of Governors of the Federal Resen/e System, Finance and Eco nomics Discussion Series, 12 (February 1988). ________. “The Discount Rate and Market Interest Rates: What’s the Connection?” this Review (June/July 1982a), pp. 3-14. _________ “Simple Analytics of the Money Supply Process and Monetary Control,” this Review (October 1982b), pp. 22-39. Wallich, Henry C. “ Recent Techniques of Monetary Policy,” Fed eral Reserve Bank of Kansas City Economic Review (May 1984), pp. 21-30. Appendix A Complete Results for a Simple Model of the Reserves Market T h is ap p en d ix develops the resu lts stated in th e text in term s o f a sim ple m odel o f the m oney stock. T he m od el co n sists o f the follow ing eq u atio n s: (All TKtl = a„ - a , I I P, + u (A2I B o rr = b„ + b,(FFR —DR) + v (A3! TR„ = NBR + Borr, and (A4) TR„ = TR„, w h ere TR d en o tes total reserves an d th e su b scrip ts “d " and “s" d en o te "d e m a n d ” an d “su p p ly .”' Borr d en o tes th e am o u n t o f borrow ings an d NBR th e su p ply o f n onborrow ed reserves, w h ich is assu m ed to be co n tro lled by the Fed. FFR an d DR d en o te th e fed eral fund s and d isco u n t rates, respectively, an d u an d v are ran d o m errors su ch that E(u) = E(v) = E(uv) = 0. E qu atio n s A1 - A4 can be co m b in ed to yield the ex p ressio n for th e eq u ilibriu m fed eral funds rate (A5) FFR = - X - ' I N B R + (b0- a „ ) - b,DR + (v - u ) ], w h ere X = (a, + b ,). Figure A l-a show s th e ex p ected value o f th is eq u ilibriu m eq u atio n .2 Given th e d isco u n t rate an d th e stru ctu ral p aram eters, it show s all p o ssi ble co m b in atio n s o f FFR an d NBR su ch that the re serve m arket is in equilibriu m . Figure A l-b reflects the ex p ected value o f th e borrow ings fu n ctio n , equation A2. 'The “time” subscript, t, is dropped for convenience. 2The curve slopes downward on the assumption that the interest rate intercept is positive. A sufficient condition for this is that a0 > b0. If th e F ed esta b lish es a borrow ings objective, Borr*, th e fed eral fu n d s rate m u st equal FFR*, given the d isco u n t rate. T h e eq u ilibriu m trad e-o ff curve in d i ca tes th at th e target level o f borrow ings ca n b e hit by providing n o n b o rro w ed reserves equal to NBR*. T his illu strates th e rela tio n sh ip b etw een a borrow ings o p erating p ro ced u re an d a fed eral fund s rate targeting p ro ced u re. If th e Fed d oes n ot resp o n d to sto c h a stic sh ocks, th e varian ce o f borrow ings will be id en tical u n d e r eith e r p ro ced u re, as will th e varian ce o f the federal fu n d s rate. D ifferen ces betw een th e two p ro ced u res em erge w h en th e F ed a cts to offset d istu rb a n ces in b o rro w ings, v. T h e resu lts d ep en d on th e tim e perio d over w h ich th e d istu rb a n ces are operative an d th e a ssu m p tion m ade abou t th e d istrib u tio n s of u an d v. For exam ple, if sh o ck s o c c u r e a c h day an d if v an d u are w h ite noise, su ch shifts essen tially will b e im p ossible n n to offset. F u rth erm o re, b eca u se 2 u/n an d S v/n i= 1 i= 1 a p p ro a ch zero as n g ets large, th ere is no n eed to offset th ese shifts if th e p lan n in g h o riz o n is fairly long. Over sh o rte r p erio d s su ch as a reserve p eriod (one w eek before Febru ary 1984 an d tw o w eeks thereafter), th ese errors will seld om “average o u t;” th erefore, it m ay be d esirab le to offset part o f th e se sh o ck s. Also, th ese sh o ck s m ay exh ibit p e rsisten ce, e.g., u, = cp,u,_, + e, and v, = tp,v,_, + T|t, w h en e, an d r|t are w h ite n oise. In th is case, th e Fed m ay also find it advantageous to offset som e shifts d uring th e reserve period (or, for that m atter, over a so m ew h at sh o rte r or lo n g er period) d ep en d in g on th e m ag n itu d e o f tp, an d ip,. 45 FEDERAL RESERVE BANK OF ST. LOUIS The M odel with Complete Adjustment to Shocks T h e borrow ings o p eratin g p ro ced u re can be differ en tiated from an in terest rate targeting p ro ced u re by co m p arin g the ap p rop riate resp o n se to sh o ck s in eith e r borrow ings o r th e fed eral funds rate u n d e r ea ch p ro ced u re. Initially, th is is d on e u n d er th e a ssu m p tion th at the Fed co m p letely offsets all sh o ck s. U nder the borrow ings p ro ced u re, th e ap p rop riate re sp o n se to sh o ck s is to ch an g e n o n bo rro w ed re serves in a cc o rd a n ce w ith th e rule: (A6) dNBR = u + (a,/b,)v.3 T hu s, no n bo rro w ed reserves sh o u ld ch an g e d o llar for d ollar w ith a sh o ck to th e d em an d for to tal reserves an d by a larger o r sm aller am o u n t (d epend ing on th e relative m ag n itu d es o f a, and b,) for a sh o ck to borrow JANUARY/FEBRUARY 1988 ings. T h e se ca s e s are illu strated in figures A2 an d A3. In figure A2, a fully a n ticip a te d in crea se in th e d em an d for total reserves shifts th e m arket eq uilibriu m curve by u; th e borrow ings fu n ctio n rem ain s u n affected by th is sh ock. C on sequen tly, th e target level o f th e fed eral fund s rate is u n ch an g ed , b u t NBR is in crea sed bv u. In figure A3, a positive value o f v shifts the bo rro w ings fu n ctio n to th e right by v an d th e m arket eq u ilib rium curve to th e left by v. As a result, th e level o f the fund s rate th at is co n siste n t w ith th e borrow ings o b jective is lower- an d n on borrow ed reserves m u st be ex p an d ed by (a,/b,) to bring th e fund s rate dow n en ou gh to m ain tain borrow ings at th e target level. If th e Fed fully offsets shifts in th e d em an d for total reserves, n e ith e r borrow ings n o r th e fu n d s rate will ch an g e. If th e borrow ings fu n ctio n shifts, how ever, borrow ings w ould rem ain at th e ir target level b u t th e fed eral fu n d s rate w ill ch an ge. U nder a fed eral fu n d s rate targeting p ro ced u re, the ap p rop riate ru le for ad ju stin g n o n b o rro w ed reserves w ould b e 3This rule is obtained by substituting A5 into A2, totally differentiating the result and setting it equal to zero. Technically, the result is dNBR = du + (a,/b,)dv; however, since the results are presented about the expected value, du and dv have been replaced with u and v. http://fraser.stlouisfed.org/ 46 Bank of St. Louis Federal Reserve (A7I dNBR = u - v. Note th at th e resp o n se to a sh o ck in to tal reserves d em an d is th e sam e as u n d e r th e borrow ings op erat- FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 F ig u re A 2 The Effect of a Rise in v u nde r a B o r r o w i n g s O p e r at i n g P r oc edu r e F ig u re A 3 The Effect of a Rise in v u nder a Bo rr o w i n g s Pr oc e d u r e 47 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Figu re A 4 The Ef fect of a Rise in v u nder a F ed er a l Rat e T ar get i ng P r o c e d u r e ing p ro ced u re, eq u atio n A6. T h e d ifference in th e two p ro ced u res co m es in th e re sp o n se to shifts in the borrow ings fu n ctio n , in th e ca se of an in terest rate target, th e Fed offsets th e effect o f an in crea se in the borrow ings fu n ctio n by re d u c in g n o n borrow ed re serves by v (illustrated in figure A4), w hile u n d e r a borrow ings op eratin g p ro ced u re, th e Fed in crea ses n o n bo rro w ed reserves by (a,/b,) o f th e shock. U nder an in terest rate target, if th e Fed offsets all shifts in th e d em an d for total reserves, n eith e r b o rro w ings n o r th e fund s rate will deviate from th e ir target levels (as u n d e r th e borrow ings p ro ced u re). If th e Fed offsets shifts in th e borrow ings fu n ction , th e fund s rate will not vary; how ever, th ere will be variability in borrow ings. The M odel with Incom plete Adjustment to Shocks T h e above analysis is b ased on the assu m p tio n that th e Fed h as p erfect foresight an d co m p letely offsets sh o ck s to total reserves or borrow ings. Now assu m e th at th e Fed only offsets part o f th e sh o ck s. T h at is, eq u atio n A6 can be rew ritten as (A8I dNBR, = 8u + 8(a,/b,)v, http://fraser.stlouisfed.org/ 48 Federal Reserve Bank of St. Louis w h ere 8 rep rese n ts th e p ro p o rtio n o f sh o ck s w h ich th e Fed offsets over a given p lan n in g h orizo n , 0 8 *£ 1. 8 = 1 is th e co m p lete a d ju stm en t m odel, 8 = 0 rep rese n ts a m od el in w h ich th e F ed m akes n o a t tem p t to offset sh o ck s. 8 w ould likely in crea se w ith the len gth o f th e plan n in g h orizon . T h e varian ce o f borrow ings an d th e fun d s rate u n d er a borrow ings op eratin g p ro ced u re can b e ex p ressed as (A9I Var( B o rr |Borr*) = b 2(l - S)2 \~2 a 2 + [1 —b, X~'(l + 8a,/b,)l2 ct; and (A10) VarlFFR |Borr*) = (1 - 8)2 \ “2 of, + X“-(l + 8 a./b^cr;, respectively.4 Note th at VarlBorr |Borr*) eq u als zero if 8 = 1, an d X.“2(b;CT;i + a;cr;) if 8 = 0. Also, VarlFFR |Borr*) equals (of/b;) if 8 = 1, an d + &;) if 8 = 0. T h e v ariance o f borrow ings an d th e fu n d s rate u n d er a fund s rate target ca n b e ex p ressed as These expressions are obtained by applying the definition of the variance, e.g., E[Borr - E (Borr)]2, and replacing NBR - E(NBR) with equation A8. JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS (A ll) VariB orr |FFR*) = b 2(l - 8)2 \ - 2ct2 o f th e fu n d s rate u n d e r an in terest rate targeting p ro ced u re. + (1 - I ) , \ - ' l l - 8 ) P o f and (A12I Var(FFR|FFR*) = \ “211 - 5 i2Ict2 +orl, respectively. T h e VariBorr |FFR*) eq u als cr; if 8 = 1 and \ "2(b2(r2 + a,of) if 8 = 0 , w hile th e VarlFFR |FFR*) equals zero if 8 = 1 and \_~lerf, + erf) if 8 = 0. A co m p ariso n o f eq u atio n s A9 and A l l show s that th e variance o f borrow ings will be s m a lle r u n d e r a borrow ings p ro ced u re th a n u n d e r an in terest rate targeting p ro ced u re for 8 > 0 and equal for 8 = 0. T h e v ariance o f th e funds rate will be la r g e r u n d e r a b o r row ings p ro ced u re th an u n d e r an in terest rate target ing p ro ced u re for 8 > 0 and equal for 8 = 0. Also, it is p o ssible to estab lish co n d itio n s u n d e r w h ich th e v ariance o f borrow ings wall be sm all relative to th e varian ce o f th e fed eral fund s rate u n d e r a borrow ings target. Solving eq u atio n A10 for II —8)2 A."’erf, and su b stitu tin g th e resu lt into A9, y ield s W hile it m ay seem od d that the ex p ressio n s for the relative v ariance do n ot d ep en d on ct2, th is result is qu ite intuitive. V ariation in th e d em an d for total re serves affects th e varian ce o f borrow ings only throu gh its effect on th e variation o f th e federal fund s rate, not d irectly th rou gh th e borrow ings fu n ctio n . C o n se quently, variability in th e d em an d for total reserves only p ro d u ces variability in th e m arket in tere st rate; given th e borrow ings fu n ctio n , th is tra n sla tes in to an equal am o u n t o f ap p rop riately sca led variability in borrow ings. T h is resu lt is illu strated in figure A5 u n d er the assu m p tio n th at 8 = 0. T h is also ex p lain s w hy co n tro l errors, i.e., NBR = NBR* + a), w h ere w re p rese n ts a ran d om co n tro l error, in crea se the variability o f b o th borrow ings an d the fed eral fu n d s rate, b u t do n o t affect th e variability of borrow ings relative to th e fu n d s rate. T h is is illu s trated in figure A5 alternatively as NBR above (NBR') or below INBR") th e target level INBR*). What I f u and v Are Correlated? (A13) VariBorr | Borr*) = b; Varl FFR | Borr* l + [1 —b, \"M1 + 8a,/b,)]2 a —b2 11 + 8a,/b,)2 of. Sin ce the term b2Varl FFR |Borr*) is m erely th e varian ce of th e in terest rate exp ressed in u n its co m p arab le to VariBorr |Borr*), after so m e sim plification, th e vari a n ce o f borrow ings relative to th e fed eral fund s rate u n d e r a borrow ings op erating p ro ced u re ca n be w rit ten as (A14) VariB orr |Borr*) — b 2 Varl FFR [ B o rr’ ) = II IA16) FFR = —\_l INBR + lb0- a „ ) - b.DR - II —g) u]. 0 12 erf — 0 2ct2, w h ere 0 = b, \ “'ll + 8a/b,l. 0 is a m o n o to n ic in c re a s ing fu n ctio n o f 8. T h e righ t-h an d side o f A13 is negative if 0 > 1/2. T h is co n d itio n will hold if b, > a, o r if 8 2= 1/2. H ence, u n d e r so m e fairly gen eral co n d itio n s, th e vari a n ce o f borrow ings w ill b e less th a n th e v arian ce o f th e federal fund s rate u n d e r a borrow ings op eratin g p ro ced u re. Likew ise, eq u atio n A12 ca n b e solved for \_2I1 —Sp a2 a n d th e resu lt su b stitu ted into eq u atio n A l l . T his yield s IA15) VariBorr |FFR*) - b 2 Varl FFR |FFR*) = 11 —4*)- ct; — i|r a 2, w h ere i|i = b, \ “'(1 —8). i|i is a m o n o to n ic d ecreasin g fu n ctio n o f 8. T h e righ t-h and side o f eq u atio n A15 will b e negative if i|/> 1/2. T h is w ill b e satisfied if b, > a, o r if 8 > 1/2. C on sequently, if th e Fed is able to offset m ore th a n h a lf o f th e sh o ck s over its p lan n in g h orizon , th e variance o f borrow ings will b e larger th a n th e variance O ne p ossibility th at d eserves co n sid era tio n is the ca se w h ere u an d v are co rrelated , th at is, sh o ck s to the d em an d for total reserves, u, p ro d u ce a ch an g e in the d em an d for b o rro w ed reseiv es, v. To see h o w this affects th e resu lts, co n sid e r first th e sp ecial ca se in w h ich th e sh o ck s are p erfectly co rrelated , e.g., v = £u. A ssum e th at £ is positive, alth o u g h th is assu m p tio n is n o t critica l to th e resu lts. Given th e se assu m p tio n s, eq u atio n A5 ca n b e rew ritten as Note th at (1 —£) is positive if 0 =£ £ < 1, zero if £ = 1 an d negative if £ > 1. Given th is assu m p tio n , no shifts in th e borrow ings fu n ctio n are in d ep en d en t o f shifts in th e d em an d for to tal reseiv es. H ence, th e d ifference that th e co rrela tio n b etw een th e erro r term s m akes ca n be see n by co m p arin g th e effect o f a ch an g e in u u n d e r b o th a ssu m p tio n s. In th e m o d el th a t a ssu m ed in d e p e n d e n ce , th e equ ilibriu m in tere st rate curve shifted to th e right by u w h ile th e borrow ings fu n ctio n did n ot shift, as in figure A2. U nd er p erfect positive co rrelatio n , th e m arket eq u ilibriu m curve shifts by (1 —£)u, w hile th e borrow ings fu n ctio n shifts by £u. T h e se shifts d eterm in e th e ex ten t to w h ich o p en m arket o p era tion s m u st b e u n d ertak en to stabilize borrow ings at the target level.5 It also ca n be sh o w n th at th e assu m p - 5lf i < 0 and equal to - b, a,, nonborrowed reserves will not have to change to stabilize borrowings at the target level. In this case, the leftward shift in the borrowings function just cancels the effect of the rightward shift in the equilibrium curve on nonborrowed reserves. 49 JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS F ig u re A 5 A n Il l ustrati on of W h y the Var i a n ce of Bo r r o w i n g s Rel at i ve to the Fed er al F un ds Rat e is U na f f e ct e d by 6 FFR N B R "N B R * NBR' NBR (a) tion o f p erfect co rrelatio n has n o effect on co n clu sio n s abou t th e variability o f borrow ings and/or th e federal fund s rate u n d e r th e alternative o p eratin g p ro ced u re. W hat if th e sto c h a stic d istu rb an ces are not p erfectly co rrelated ? F or exam ple, assu m e th at v = £11 + t ) , w h ere T| is id en tically and in d ep en d en tly d istribu ted w ith a m ean zero an d a variance cr2. Given th is a s su m p tio n , th e erro r term o f eq u atio n A5 is sim ply — + (1 —£;) u]; th e sam e as th at o f A5 ex cep t th at u is rep laced by (1 —£)u and T| re p laces v. C onsequently, 50 all o f th e previously stated resu lts h old .6 6The intuition for this is straightforward. The variability of borrowings under a borrowings operating procedure relative to that under a federal funds rate operating procedure depends only on the variabil ity of the borrowings function. Since variability of the borrowings function is the same under any of these assumptions, i.e., v = f;U or even v = £u + r), for both the borrowings and federal funds rate targets, the assumption made does not affect the general conclusion about the variability under these procedures. This is also the reason the general conclusions about the variance of borrowings relative to the federal funds rate under the borrowings operating procedure are unaffected by this assumption. JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS Appendix B A More Detailed Analysis of the Effect of a Change in the Borrowings Assumption on the Federal Funds Rate T h e p u rp o se o f th is ap p en d ix is to p re sen t d etailed resu lts on th e effect o f a ch an g e in th e borrow ings assu m p tio n on th e federal fund s rate. O ne w ay to calib rate su ch effects is to estim ate a red u ced form eq u atio n for th e le v e l o f th e fed eral fu n d s rate (FFR): ( B l) FFR, = a + PDR, + jjiB A , + e„ w h ere DR and BA d en o te the level o f th e d isco u n t rate and borrow ings assu m p tio n . U nder a strict borrow ed reserves op eratin g p ro ced u re, (3 sh ould be positive an d equal one. OLS estim ates o f eq u ation B l are re ported in th e top h a lf o f table B l for th e LRR and CRR period s. T h ree significant a sp e cts o f th e se resu lts d e serve p articu lar atten tio n . First, th e h yp o th esis th at (3 = 1 is re je cte d at th e 5 p e rce n t level d uring b o th period s. Seco n d , th e Q statistic d o es not in d icate loword er serial co rrelatio n during th e LRR period, but d oes in d icate it during th e CRR period. N evertheless, the resid u als sh ow a p ro n o u n ced qu arterly season al spike d uring th e LRR period (clearly evident from ch art 2 o f the text). Third , th e stand ard erro r o f th e eq u ation in crea ses d ram atically during the CRR p e riod, in d icatin g in creased variability o f th e FFR u n d e r CRR. (This is tru e w h eth er w eekly o r reserve period data are used.) B eca u se o f th e sea so n a l spike d uring th e LRR period an d serial co rrelatio n o f th e resid u als d uring th e CRR period, th e eq u a tio n s w ere reestim a ted in clu d in g lagged d ep e n d en t variables. T h e resu lts are rep o rted on th e b o tto m h alf o f table B l. (Four lags o f FFR are in clu d ed d uring the CRR period ; in ad d ition, FFR,_13 is in clu d ed d uring th e LRR period.) D uring the LRR period, th e co efficien t on BA in crea sed som ew hat, alth ough its t-ratio d eclin ed . Also, th e estim a te o f (3 d eclin ed su b stan tially an d th e h yp o th esis that p = 1 is re je cte d at very low sig n ifican ce levels. F or th e CRR period, th e estim ated co efficien t on BA d e clin e d by n early tw o-th ird s an d th e t-ratio d eclin ed dram atically. T h ere are several rea so n s for q u estio n in g th e e sti m ates from the level eq u atio n s. T h e first rea so n relates to the tim e-series p ro p erties o f th e individual series th em selves. BA is highly au to co rrelated , as table B2 in d icates. T h e fact th at th e levels o f BA an d FFR are highly a u to co rrela ted affects th e relatio n sh ip b etw een th em . T h is is evident in th e sim p le co rrelatio n co ef ficien ts given in table B3. T h e sim p le co rrela tio n o f FFR an d BA is h ig h er th a n th at o f FFR an d a ctu a l a d ju st m en t plus seaso n al borrow ing, Borr, d uring th e LRR period ; how ever, th e co rrela tio n co efficien t o f first Table B1 Estimates of Equation B1 Constant DR BA October 1,1982February 1, 1984 2.23* (2.37) .72* (6.78) February 2, 1984December 31, 1986 0.81* (2.22) October 1, 1982February 1,1984 February 2,1984December31,1986 Period DL Q3 SEE .0017* (9.60) — 4.69 .3041 .80* (13.4) .0029* (10.95) — 38.05* .3763 5.39* (5.23) .42* (2.02) .0022* (6.64) 5.71*' 5.95 .2616 0.53 (1.76) .28’ (2.29) .0011* (2.90) 10.13*2 0.78 .3072 ’ Indicates statistical significance at the 5 percent level. ’Test that FFR,_, - FFR,_„ and FFR,_13are jointly zero. 2Test that FFR,., - FFR,_4are jointly zero. 3Test for white noise residuals, distributed x2(6). 51 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Table B2 Autocorrelations of Time-Series Variables for Reserve-Period Data Lag 1 2 3 4 5 FFR .55 BA Borr .98 .37 .53 .95 .37 .48 .92 .39 .46 .88 .25 .37 .85 .27 Variable 6 7 8 9 10 11 12 13 14 15 .17 .60 .17 .17 .55 .12 .49 .34 .44 .16 .14 .42 .00 .39 .15 .01 .34 .16 .48 .40 .12 .43 .33 .10 .38 .26 .09 .33 .20 .04 .28 .14 .07 .23 .09 .00 October 1,1982 - February 1,1984 .36 .81 .31 .32 .76 .36 .23 .71 .27 .23 .66 .09 February 2,1984 - December 31,1986 FFR .97 .94 .89 .84 BA Borr .97 .50 .93 .32 .89 .25 .83 .27 .78 .76 .19 .72 .69 .17 .66 .61 .09 Table B3 Simple Correlations of Time-Series Variables for Reserve-Period Data Variables BA Borr ABA ABorr October 1,1982 - February 1,1984 FFR AFFR .605* — .540* — — .166 — .533* February 2,1984 - December 31,1986 FFR AFFR .903* — .492* — — .175 — -.0 2 4 * indicates statistical significance at the 5 percent level. d ifferences o f FFR an d BA is d ram atically different from that o f th eir levels. T h is is not true, how ever, of th e co rrelatio n b etw een B o rr an d FFR an d A Borr and AFFR. F o r th e CRR period, w h en th e ir au to co rre la tio n s m a tch closely, th e co rrelatio n b etw een FFR an d BA is high. Yet in first-d ifference form , th e co rrelatio n is essen tially the sam e as d uring th e LRR period an d is n ot statistically significant. A se c o n d reaso n to be cau tio u s o f th e level eq u ation resu lts h as to do w ith th e lon g -ru n stability o f th e b orrow ings fu n ctio n itself. T h e borrow ings a ssu m p tio n d o es n o t rep rese n t an exog en ou s su p p ly o f b o r row ings; m o re precisely, it is an exog enou s target level th at th e Fed attem p ts to in d u ce d ep ository in stitu tio n s to hold by altering th e su pply o f n on borrow ed http://fraser.stlouisfed.org/ 52 Federal Reserve Bank of St. Louis .60 .54 .07 .54 .47 .10 reserves. C on sequen tly, actu al borrow ings can , and do, deviate from th e d esired level. N evertheless, over a lo n g er tim e period, th e average level o f borrow in gs ca n b e clo se to th e d esired level. T h is is esp ecially likely if a d ju stm en ts are m ad e to n o n b o rro w ed reserves or if th e borrow ings a ssu m p tio n itself is ch a n g ed to keep it in lin e w ith a ctu al borrow ings levels. T h erefore, w h en th e le v e l o f th e fu n d s rate is re g ressed on th e le v e l o f th e borrow in gs assu m p tio n , th ere is a te n d en cy to retrieve this lon g -ru n rela tio n ship to a g reater or sm aller degree, d ep en d in g on h ow clo sely th e borrow ings a ssu m p tio n m im ics a c tual borrow in g s.1 In o rd er to m ore clo sely ca p tu re the effect of an exo g en o u s c h a n g e in th e borrow ings assu m p tio n on th e fu n d s rate, first d ifferen ces o f th e fu n d s rate are reg ressed on first d ifferences o f th e borrow in gs a s su m p tio n . T h is sh o u ld yield c o n siste n t estim ates of th e im m ed iate re sp o n se o f th e fed eral fu n d s rate to an exo g en o u s ch a n g e in th e borrow ings assu m p tio n , even if th e level sp ecifica tio n is co rrect.- M oreover, it 'Augmented Dickey-Fuller tests for stationarity applied to borrowings and the funds rate indicate that both series are integrated of order one, i.e., 1(1) for the LRR period. When the test is applied to the residuals from OLS estimates of equation 1, however, the results indicate that borrowings and the funds rate are cointegrated in the Engle-Granger (1987) sense. The augmented Dickey-Fuller test indicates that BA and FFR are l(2) over the CRR period. Yet the test indicates that the residuals from equation 1 estimated over this period are stationary. The OLS estimate of b,, of equation 1 from the text for the LRR period is 471. This yields an implied coefficient estimate of (3 of equation B1 equal to .0021 (1/471). The implied estimate of 0 for the CRR period using resen/e-period data is .0038 (1/260). 2See Plosser, Schwert and White (1982). FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Table B4 Table B5 t-ratios for ABAj from equation B1 Estimation of the Funds Rate Equations with ABA Partitioned as in Table B1 October 1,1982February 1,1984 February 2,1984December 31,1986 Reserve-Period' Weekly Reserve-Period — 2.78 AN 1.23 A -1 .9 8 AN -1 .6 6 AN -2 .1 4 AN -0 .4 8 0.42 -0.56 0.56 0.34 -0.81 0.52 0.96 -1 .1 7 AN -0.3 5 - 1.76 AN 0.89 0.03 1.00 -0 .3 7 2.69 A 1.13 2.12 A 0.22 0.31 -0 .6 4 0.27 -0 .3 4 0.83 0.26 -0 .3 8 0.37 -0 .5 4 -1 .0 0 0.65 0.53 -0 .3 5 0.13 0.64 0.66 1.80 A 2.51 A 0.75 1.57 A -0 .8 7 -0 .1 5 -0 .2 0 -0 .2 0 0.43 -0.21 -0 .1 7 -0 .3 9 0.06 0.19 0.06 0.33 0.35 -2 .0 9 AN 0.35 (A, M-2 M-3 M-4 fJ-5 M"6 M"7 M-8 M-9 M-10 M-12 M-13 M-14 ms M"16 M-17 M-18 M-19 M-20 M*21 NOTE: A denotes those that are positive and significant at a 25 percent level. AN denotes those that are negative and significant at a 25 percent level. 'During the LRR period, the reserve period was one week. sh o u ld avoid sp u rio u s co rrelatio n often ex p erien ced w h en th e levels o f n o n statio n ary series are used. Finally, b eca u se th e borrow ings assu m p tio n is ch an g ed infrequently, ch an g es in th e borrow ings a s su m p tio n can be p artitio n ed into th o se th at do have a significant effect on th e fed eral funds rate and th o se th at do not. T his is d on e bv estim atin g th e eq u ation : K 1B2) AFFR, = a„ + 2 a,AFFK,_, + p„ ADRT, + (3, ADRNT, i= 1 October 1,1982February 1,1984 Reserve-Period1 February 2,1984December 31,1986 Weekly Reserve-Period Constant 0.04 (0.98) 0.015 (0.47) 0.033 (0.87) ADRT 0.20 (0.22) 0.247 (0.33) 0.082 (0.15) ADRNT 2.280* (3.60) 0.755* (1.99) 0.891* (2.36) ABA-A 0.0087 (1.16) 0.0055* (3.10) 0.0058’ (3.21) ABA-B 0.0007 (0.77) 0.0011 (1.25) 0.0002 (0.26) -0.0160* (2.53) — -0.0076’ (2.37) R2 0.5737 0.1875 0.2046 SEE 0.2803 0.3761 0.2951 6.57* 0.65 ABA-AN F AFFRLAGS 15.81* 'indicates statistical significance at the 5 percent level. 'During the LRR period, the reserve period was one week. borrow ings a ssu m p tio n over th e sam p le p eriod .3 Th e estim ated stan d ard erro r from th is eq uation, cr;, m ea su res th e co n d itio n a l varian ce o f AFFR for p eriod s w h en ABA = 0. H ence, th e t-ratio for th e jth ABA in d ica tes h ow m u ch th e fed eral fund s rate m oved d uring th is period relative to period s w h en th e b o r row ings assu m p tio n w as u n ch an g ed . T h e t-ratios for ea ch ABA are rep o rted in table B4. T h e resu lts in d icate that, o f th e 17 ch a n g es in the borrow ings a ssu m p tio n d uring th e LRR period, 10 w ere in v erse ly related to ch a n g es in the fed eral funds rate. D uring th e CRR period , eith e r seven or eight of th e 21 ABAs w ere inversely related to th e fu n d s rate, d ep en d in g on w h e th e r w eekly or reserv e-p eriod data are u sed . T h e resu lts in table B4 ca n be u sed to partition ABA in to th o se th at have a positive signifi- L + X (JLiABA;, + e ,, j= l w h ere ABA,: takes on th e value o f th e jth ch an g e in the borrow ings assu m p tio n d uring th e perio d an d is zero o th erw ise. L d en o tes th e n u m b er o f ch an g es in the 3lf daily data were used, L is equal to the number of changes on the borrowings assumption over the sample period. When weekly or reserve-period data are used, the data are averaged on a pro-rata basis. Consequently, L denotes the number of weeks or reserve periods that are affected. This is usually larger than the number of changes in the borrowings assumption itself. 53 FEDERAL RESERVE BANK OF ST. LOUIS can t effect on th e fund s rate, ABA-A, th o se th at have a significant negative effect, ABA-AN, and all others, ABA-B. T h is w as d one by in clu d in g in th e A o r AN groups all ABAs th at are significant at th e 0.25 p e rce n t level u sing stand ard analysis. T h o se ch an g es in BA th at are in th e A or AN groups are d esignated co rre spond ingly in table B4. E stim ates o f th e sam e b asic eq u atio n w ith th e p arti tioned data are p resen ted in table B5. In all cases, http://fraser.stlouisfed.org/ 54 Bank of St. Louis Federal Reserve JANUARY/FEBRUARY 1988 excep t th e single observ ation in th e A group d uring th e LRR period, th e co efficien ts on th e A an d AN partitio n s are sign ifican t at th e 5 p e rce n t level. M ore im portantly, th e co efficien ts on th e ch a n g es in th e borrow ings assu m p tio n in th e B p artition , w h ich a c co u n t for th e vast m ajo rity o f ch a n g es in th e borrow ings assu m p tio n , w ere uniform ly in sign ifican t at th e 5 p e rce n t level. T h is evid en ce in d ica tes th at th e link b etw een ch a n g es in th e borrow ings assu m p tio n and th e fed eral fund s rate is, at best, w eak. FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 Farm Policy and Mandatory Supply Controls — The Case of T obacco Kenneth C. Carraro F ROM 1980 throu gh 1986, th e U nited States sp en t $43.9 billion in d irect p ay m en ts to farm ers and $52.3 billion on o th er p rice su p p o rt program s.1D esp ite su ch exp en d itu res, th e U.S. farm se c to r has ex p erien ced a severe d ow nturn. Falling exports, d eclin in g farm land values, high rates of farm lo an d elin q u en cie s an d in creasin g d e p e n d en ce on governm ent su p p o rt p ay m en ts w ere visible sym ptom s o f th e farm s e c to r’s difficulties. B eca u se o f th e great ex p en se an d th e ap p aren t failure o f farm program s, som e policym akers have called for the use o f m an d ato iy su pply co n tro ls to lim it crop p ro d u ctio n an d raise prices.- Advocates assert th at su ch co n tro ls co u ld gu aran tee farm ers a “fair” p rice an d im prove th eir in co m es, w hile d rasti cally cu ttin g th e co st o f farm program s and elim in a t ing farm co m m o d ity su rp lu ses. Kenneth C. Carraro is an economist at the Federal Reserve Bank of St. Louis. Dawn M. Peterson provided research assistance. 'U.S. Department of Agriculture, History of Budgetary Expenditures of the Commodity Credit Corporation, Book 2, and Agricultural Outlook (December 1987), p. 53, table 32. 2The Harkin-Gephardt “ Save the Family Farm Act” is the most prominent domestic example of mandatory supply control legisla tion currently being debated in Congress. In 1986, Congress allo cated $10 million for the study of mandatory controls and the polling of farmers. Mandatory supply controls have recently been proposed in the European Economic Community to limit milk production. T h is article ex am in es th e effects o f m an d atory su p ply co n tro ls. T h e an alysis beg in s w ith a th eo retica l d iscu ssio n o f th e effects o f m an d ato ry su pply co n tro ls o n eco n o m ies th at are clo sed to in tern atio n al trade an d th o se th at engage in in tern atio n al trade. Next, the ex p e rie n ce o f th e U.S. to b a c co in d u stry an d its m a n d a tory supply co n tro ls is exam in ed .3 Finally, th e key p o in ts from th e th e o retica l d iscu ssio n an d th e U.S. to b a cco in d u stry 's e x p e rie n ce are co m b in ed w ith sp e cific facts abou t U.S. cro p s to suggest th e likely c o n s e q u en ces o f th e su p p ly legislatio n cu rren tly u n d e r c o n sid eration . THE ECONOMICS OF SUPPLY CONTROLS IN A CLOSED ECONOMY Supply co n tro l p rogram s are d esign ed to in crea se th e p rice o f a good above its free m arket p rice by restrictin g th e qu an tity o f th e good th at re a ch es the m arket. T h e su pply re strictio n s typically are e sta b lish ed by a governm ent ag en cy or a co n so rtiu m of p ro d u cers. T h e O rganization o f P etroleu m Exporting C ou n tries (OPEC) is o n e exam p le o f a group o f p ro d u cers w ho agree (usually) to re strict p ro d u ctio n as a m ean s o f secu rin g a h ig h er p rice for cru d e oil. 3U.S. tobacco policy has used mandatory supply controls since the 1930s. 55 JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS p rice. If th e qu an tity o f a p ro d u ct d em an d ed ch a n g es p ro p o rtio n ately less (in a b so lu te value) th a n th e ch an g e in p rice, th e d em an d is referred to as in elastic. Figure 1 Supply Controls in a Closed Ec o n o m y Sin ce a 1 p e rce n t in crea se in p rice ca u ses le ss th a n a 1 p e rce n t d ecrea se in quantity d em an d ed w h en d e m and is in elastic, th e p rice in cre a s e ca u se s total reve n u e to in crea se. Conversely, a p rice d ecrea se ca u ses total revenue to d ecrea se w h en d em an d is in elastic. T h e effects on total revenue o f p rice ch a n g es are reversed w h en d em an d is elastic. E la stic d em an d ex ists w h en th e qu an tity o f a p ro d u ct d em an d ed ch a n g es p ro p o rtio n ately m o re (in a b so lu te value) th an th e ch an g e in p rice. Sin ce a 1 p e rce n t in crea se in p rice ca u ses a m o re th a n 1 p e rce n t d ecrea se in qu an tity d em an d ed w h en d em an d is elastic, th e p rice in crea se ca u ses total revenue to d ecrea se. Conversely, a p rice d ecrea se ca u ses total revenue to in crea se w h en d e m an d is elastic. A final possibility, know n as u n ita iy elasticity, is that a 1 p e rce n t ch a n g e in p rice lead s to a 1 p e rce n t ch a n g e in q u an tity d em an d ed , w h ich h as no effect on total revenue. Supply Limit Quantity Figure 1 d em o n strates how p rices are d eterm in ed in an eco n o m y that is clo sed to in tern atio n al trade an d ho w su p p ly co n tro ls ca n in crea se th e p rice o f a good above its free m arket level. T h e su p p ly curve, lab eled S, rises upw ard an d to th e right, ind icatin g th at p ro d u cers w ill su pply larger qu an tities o f a good as its p rice is in creased . T h e sh o rt-ru n d em an d curve, labeled Ds, slop es d ow nw ard to show th at co n su m ers will buy sm aller q u an tities o f a good as its p rice rises. In a free m arket, the in tersectio n o f the supply and d em an d curves at p o in t A d eterm in es that th e p rice w ould be P, w hile th e qu an tity su p p lied w ould equal th e qu antity d em an d ed , at Q,. Sin ce th e qu an tity o f the good su p p lied to th e m arket at th at p rice exactly satisfies co n su m e r d em and , n eith e r p ro d u cers n or co n su m ers have an incen tive to ch an g e th e ir p ro d u c tio n o r co n su m p tio n p attern s. By im p osing a su p p ly lim it at Q ,, th e p rice ca n b e in crea sed from P, to P,. T h is w ou ld b en efit p ro d u cers, how ever, only if it in crea sed th e ir profits. Sin ce p ro d u ctio n d eclin es, th e total co sts in cu rred by p ro d u cers w ill d eclin e also. As long as total revenue is n o t re d u ce d by an am o u n t larger th a n th e re d u ctio n of total co sts, profits will rise. T h e ch an g e in total revenue resu lting from a p rice ch an g e d ep en d s u p on th e p rice elasticity o f dem and . T h e p rice elasticity o f d em an d m e a su res th e re sp o n siveness o f th e qu antity d em an d ed to a ch a n g e in http://fraser.stlouisfed.org/ 56 Federal Reserve Bank of St. Louis In figure 1, th e su p p ly con trol, w h ich re d u ce d th e qu an tity su p p lied from Q, to Q,, ap p ears to have ca u sed th e p rice ap p roxim ately to d ou ble from P, to P,. T h e qu an tity d em an d ed , how ever, a p p ears to have d ecrea sed m u ch less. In o th e r w ords, th e d em an d is co n sid ered to b e in elastic in th at p rice range. W hen th e d em an d for a p ro d u ct is in elastic, a su pply co n tro l program in crea ses th e total revenue o f p ro d u cers. Sin ce total co sts will have fallen also, profits m u st in crease. W h en th e d em an d for a p ro d u ct is elastic, a su pply co n tro l program w ould re d u ce th e qu an tity d e m a n d ed p ro p o rtio n ately m o re th a n th e p rice in crea se. T h e re d u ctio n in total revenue m akes it p o ssi ble th at th e su p p ly co n tro l program co u ld lead to re d u ce d profits. In general, a su p p ly co n tro l program is b en eficial to p ro d u cers facing an in ela stic d em an d . A variety o f factors in flu en ce th e elasticity o f d e m an d for a p ro d u ct. O ne o f th e m ost im p o rtan t of th e se is th e availability o f su b stitu tes for th e p ro d u ct. A p ro d u c t’s d em a n d is m o re likely to b e elastic if a c c e p t able su b stitu tes fo r th at p ro d u ct exist. F o r exam ple, th e p rice elasticity o f b e e f likely ex ceed s th a t o f g aso lin e b e c a u se th ere are n u m ero u s su b stitu tes for b e e f w hile th e re are few su b stitu tes for gasoline. A n oth er extrem ely im p o rta n t in flu en ce on d em an d elasticity is tim e. In th e sh o rt run, a p ro d u c t’s d em an d is generally less ela stic th a n over th e lon g ru n b eca u se co n su m ers find su b stitu tes or learn to co n serv e on the co n su m p tio n o f th e p ro d u ct over tim e. D em an d b e co m e s m o re ela stic th e lo n g er th e tim e p erio d as FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 F igu re 2 Supply Controls in a W orld Economy W ith Trade Panel A Panel C Panel B Domestic Eco nom y Rest of the World Entire World Supply limit Price Price 'DA UDA«C SA SA Quantity co n su m ers read ju st th e ir co n su m p tio n p atte rn s.4 Figure 1 portrays th e effect o f ch an g es in d em an d elasticity over tim e. T h e curve D Lportrays th e lon g-run d em an d curve for th e p ro d u ct an d is m u ch flatter th an th e sh o rt-ru n d em an d curve Ds. T h is reflects the g reater elasticity th a t is co m m o n over th e lon g run. T h e su p p ly co n tro l th at resu lted in th e d oublin g of p rices from P, to P, in th e sh o rt ru n is m arkedly less ben eficial to p ro d u cers over th e long ru n. In th is case, th e im p o sitio n o f th e su p p ly restrain t h as a relatively sm all effect on th e p rice, raising it only to P.,. F u rth er m ore, it ap p ears that th e total revenue h as d eclin ed th rou gh th e u se o f th e co n tro ls. T h e sh o rt-ru n strategy th at ap p eared to in crea se profits m ay lead to low er future profits if th e lon g-run d em an d b e co m e s elastic. THE ECONOMICS OF SUPPLY CONTROLS IN AN OPEN ECONOMY Quantity Quantity d u ced in co u n tries o u tsid e o f th e co u n try (or group of cou n tries) a ttem p tin g to in crea se retu rn s th rou g h a su pply co n tro l policy, an d 2) th at th e co n tro lled good ca n b e trad ed b etw ee n co u n tries. In a clo sed e c o n omy, a p ro d u ct ’s p rice is d eterm in ed solely by d o m es tic su p p ly an d d em an d . W ith th e ad d ition o f trade, p rice d eterm in atio n o cc u rs in th e w orld ra th er th an d o m estic m arket. Figure 2 p ortrays p rice d eterm in atio n in th e w orld m arket. Panel A re p rese n ts th e d o m estic m arket for a good. Panel B re p rese n ts th e su pply an d d em an d of th e p ro d u ct for all o th e r co u n tries in th e w orld. F i nally, p an el C is th e w orld econ om y, w h ich is derived by horizon tally co m b in in g th e su pply an d d em an d curves o f th e d o m estic an d rest-of-th e-w orld e c o n o m ies. So far, w e have fo cu sed o n a sim p le eco n o m y w ith out in tern atio n al trad e to illu strate fu n d am en tal p o in ts abou t su pply co n tro l program s. T h is sectio n exp an d s th at analysis to in clu d e su p p ly co n tro ls in a w orld eco n o m y w ith trade. T h e ad d ition o f trad e to th e analysis im p lies: 1) th a t a p ro d u ct m ay b e p ro Ignoring tra n sp o rta tio n co sts, th e equilibriu m p rice for b o th th e d o m estic eco n o m y an d th e rest o f the w orld is Pw. In th is ca se, th e eq u ilibriu m p rice is above w hat the d o m estic p rice w ould have b een in a clo sed eco n om y. A ccord in g to p a n el A, at th e w orld p rice, d o m estic p ro d u cers su p p ly a larger qu an tity (QSA) th a n d o m estic co n su m ers are w illing to p u rch a se ( Q da ). T h e d ifference b etw ee n th e se tw o is ex p o rted to th e re st o f th e w orld w h ere, at Pvv, co n su m ers d em an d a larger qu an tity (QDB) th a n p ro d u cers in th e rest o f the w orld are w illing to su p p ly (Q SB) as sh o w n in p a n el B. 4For example, Houthakker and Taylor (1966) estimated the long-run price elasticity for gasoline at - .7, while the short-run elasticity was estimated to be much more inelastic at - .2. T h e d o m estic eco n o m y in figure 2 is p ortrayed as th e d o m in an t w orld su p p lier o f a p ro d u ct for w h ich th e d em an d is in elastic. T h e im p o sitio n o f a su pply 57 FEDERAL RESERVE BANK OF ST. LOUIS co n tro l in th e d o m estic eco n o m y at th e qu an tity Q (: ch an g es th e w orld su pply from Su to Sc. T h is shift, in turn, ca u ses th e w orld p rice to ju m p from Pu to P ,. B eca u se o f su pply co n tro ls in th e d o m estic econom y, th e qu antity su p p lied falls from Q SAto Q ,. At th e h ig h er p rice o f P( , foreign p ro d u ctio n in crea ses from Q sli to Q 's„; w hile foreign co n su m p tio n falls from Q„„ to Q 'UB. As a resu lt o f th e se ch ang es, th e level o f exp o rts from th e d o m estic eco n o m y to th e rest o f th e w orld d e clin es. T h e sh ares o f w orld trad e and w orld p ro d u c tio n h eld by th e d o m estic eco n o m y also d eclin e. T h e lo ss o f sh ares o f w orld p ro d u ctio n an d trad e is a p red ictab le o u tco m e o f a su pply co n tro l m easu re. W hile an exporting co u n try m ight p refer n ot to lose its sh ares o f w orld p ro d u ctio n and trade, it is m ore likely to a cc e p t th e se lo sses if th e su pply co n tro ls result in hig h er retu rn s to p ro d u cers. In figure 2, it appeal's that retu rn s w ould b e in crea sed in th e sh o rt ru n b ec a u se th e in e la stic w orld d em an d curve an d th e in elastic foreign su pply curve resu lt in h ig h er total revenue for d o m estic p rod u cers.'1 T h e se sh o rt-ru n retu rn s will erode, how ever, b e ca u se th e p rice elasticities o f bo th d em and an d supply in crea se over tim e. A given d o m estic su pply co n tro l resu lts in a sm aller p rice in crea se in th e long ru n th an in th e sh o rt ru n. T h is effect is even m o re p ro n o u n ced w ith in tern atio n al trad e b e c a u se th e elasticity o f for eign, as w ell as d o m estic, su pply generally in crea ses over tim e. In th e sh o rt run, p ro d u cers are u n ab le to resp o n d fully to a p rice in crease b ec a u se the capital b ase u sed for p ro d u ctio n is fixed. Over a lo n g er period, p ro d u cers can in cre a se o u tp u t by add ing p ro d u ctio n cap acity, im proving tech n o lo g y and ad op tin g n ew technology. T h is long-run foreign su p p ly resp o n se co n trib u te s to th e d eclin e in th e sh are o f w orld p ro d u ctio n and trad e o f th e d o m estic co u n try by in c re a s ing foreign p ro d u ctio n and, in th e p ro cess, red u cin g th e d em an d for th e d o m estic c o u n tiy ’s exp orts. T h e foreign su pply re sp o n se b e c o m e s in creasin g ly m ore im p ortant b ec a u se o f th e grow ing foreign sh are of w orld p ro d u ctio n . 5The example of OPEC is instructive at this point. When OPEC reduced production as a means of increasing the price of crude oil, it was logical to expect that its share of both oil exports and production would fall. While its share fell, it was able to greatly increase its returns because of the elasticities of world demand and supply. With a lack of acceptable energy sources as substitutes, the world demand for crude oil was extremely inelastic. The world supply of oil also was extremely inelastic because of the small share of world production held by non-OPEC countries and the difficulty, expense and time required to find and tap new oil reserves. If non-OPEC countries had been able to expand production easily and quickly in response to higher prices, the price increases would not have been as great. http://fraser.stlouisfed.org/ 58 Federal Reserve Bank of St. Louis JANUARY/FEBRUARY 1988 In sum m ary, th e in tro d u ctio n o f in tern atio n al trade m akes th e d ecisio n to u se su pply co n tro l m easu res d ep en d en t on th e elasticity o f w orld d em a n d an d w orld supply. It is im p o rtan t to n o te that, w hile the elasticity o f foreign (rest-of-the-w orld) su p p ly is im portant, it is th e en tire w o rld ’s elasticity o f su pply that d eterm in es if a d o m estic su pply co n tro l program will be effective. F or exam ple, foreign su pply m ay be very elastic over a sm all range; b u t if foreign p ro d u ctio n re p rese n ts only a sm all sh are o f to tal w orld p ro d u c tion, th e d o m estic su pply co n tro l program m ay still be very profitable. T h is is tru e b e c a u se th e foreign supply resp o n se, w h ile very elastic, m ay have only a sm all effect on th e total qu an tity su p p lied in th e w orld if d o m estic p ro d u ctio n dwarf’s foreign p ro d u ctio n . A natu ral c o n se q u e n ce of d o m estic su pply co n tro ls and foreign su pply elasticity, how ever, is an in crea se in the foreign sh are o f w orld p ro d u ctio n an d a resu lting in crea se in th e w orld su pply elasticity. THE ORIGINAL TOBACCO PROGRAM T h e cu rren t to b a cco program has its roots in the farm legislation o f th e 1930s know n as th e Agricultural A d ju stm en t Act (AAA). T h is legislatio n u sed p ro d u c tion co n tro ls on m o st agricultu ral p ro d u cts as a m ean s o f in crea sin g p rices. Of th e n u m ero u s supply co n tro l program s p ro p o sed in th e original AAA legisla tion, on ly th e to b a cco an d p ea n u t program s have m a in tain ed d irect p ro d u ctio n co n tro ls. T h e to b a c co program fu n ctio n ed , an d co n tin u es to fu n ction , by first estab lish in g a su p p o rt p rice .6 In i tially, farm ers w ere assig n ed allo tm en ts th a t in d ica ted th e n u m b er o f a cres o f to b a c co e a c h farm er cou ld c u ltiv a te . In th e 1 9 6 0 s an d 1 9 7 0 s , th e a c r e a g e a llo t m en ts w ere su p p le m en ted w ith m arketin g q u o tas that lim ited th e n u m b er o f p o u n d s o f to b a cco ea ch farm er co u ld sell. T h e se q u o tas w ere b a se d on estim a tes of th e qu an tity th at co u ld be sold at th e su p p o rt price. T h e p rice su p p o rt m e ch a n ism h a s ch a n g ed only slightly over tim e. Initially, if a farm er did n o t receive an offer g rea ter th a n th e su p p o rt p rice, th e govern m en t p u rch a se d th e farm er’s to b a c co an d h eld it u ntil it co u ld b e sold at th e su p p o rt p rice. In th e 1940s, a system o f g ro w ers’ coop erativ es w as organized to p u r ch a se an d h o ld th e su rp lu s to b a cco . T h e coop erativ es received, a n d co n tin u e to receive, govern m en t fin a n c ing. 6From its inception in the 1930s until 1985, the tobacco support price was based on a “parity index" which measures the ratio of prices received by farmers to prices paid by farmers. The parity ratio is typically criticized for having no relationship to market prices. FEDERAL RESERVE BANK OF ST. LOUIS F or a long period, th e to b a cco program w as c o n sid ered extrem ely su ccessfu l. T h e p rice o f U.S. to b a cco co n tin u ed to rise, an d th e program w as ru n at little co st to th e g overnm ent. In ad d ition, th e qu o ta rights to grow an d sell to b a c co w ere m arketable; in fact, they g en erated as m u ch as $800 m illion p e r y e a r in in co m e for qu ota o w n ers.7 It is, in part, b ec a u se o f th e ap p a r en t s u c c e ss o f th e to b a c co program th at in tere st in supply co n tro ls h as resu rfaced for o th e r crop s. T h e to b a c co p ro g ram ’s ability to en d u re w hile g en erating su bstan tial w ealth th rou gh th e sale an d le a s ing o f qu otas w as attribu table to th e in e la stic n atu re of b o th w orld d em an d an d su pply o f to b acco . T h e m ajo r rea so n for th e in e lastic su pply re sp o n se w as th at th e United States h eld a large sh are o f th e w o rld ’s p ro d u c tion an d sales o f p articu lar varieties o f to b a c co .8 As recen tly as th e 1950s, th e United States p ro d u ced m ore th an 80 p e rce n t o f th e w o rld ’s bu rley to b acco . It is im p o rtan t to n o te th at th e U.S. d o m in an ce in to b a c co p ro d u ctio n an d th e in elasticity o f w orld su p ply w ere even g reater w h en o ne co n sid ers th e im p o r tan t d istin ctio n o f to b a cco quality. Owing to sp ecial soil an d clim atic co n d itio n s and grow ing ex p erien ce, U.S. to b a c co generally w as regarded to be o f u n m a tch ed quality.3 T h is fu rth er d ifferentiated it from to b a cco grow n in o th e r co u n tries. If o th e r co u n tries w ere u nab le to grow su p erio r quality to b a c co even as its p rice in creased , th e su pply o f th at to b a c co w ould be co n sid ere d p erfectly in elastic. Perfectly in elastic supply m ean s th at th e qu antity su p p lied w ould not ch an g e w h en th e p rice changed . T h e d em an d for to b acco , in general, w as also in e la s tic. O ne so u rce estim ated th e in term ed iate-ru n d e m an d elasticity o f to b a c co at — .1 and th e long-run elasticity at —.5.'" T h e m ajo r reaso n for th e in elastic n atu re o f to b a c co d em an d is th e lack o f su bstitu tes. T h e addictive n atu re o f to b a c co fu rth er re d u ce s se n s i tivity to p r i c e ch an g es. F u r t h e r m o r e , to b a c co p u r 7Sumner and Alston (1985), p. 13. The U.S. General Accounting Office study found that, although farmers were the intended benefi ciaries of the tobacco program, 68 percent of quota owners were not active farmers. U.S. General Accounting Office (1982), p. 18. 8There are numerous varieties of tobacco. Two varieties, flue-cured and burley, account for more than 90 percent of the tobacco grown in the United States. There are other varieties used in the blending of cigarettes that are not grown in this country, such as Oriental tobacco. JANUARY/FEBRUARY 1988 ch a se s gen erally re p rese n t only a sm all sh are o f a c o n su m e r’s budget, a fact th at u sually re d u ces the elasticity o f d em an d . W hile to b a c co u sers ca n sw itch from U.S. to foreign to b a c co (or cigarettes), th ere are few su b stitu tes for to b a c co in g eneral. By using su pply co n tro ls, U.S. to b a cco p ro d u cers initially earn ed h ig h er in co m es. W hile th e qu an tity of to b a cco m arketed fell, th e resu ltin g p rice in crea se w as large en ou gh to ca u se th e total revenue received by q u o ta ow ners an d to b a c co grow ers to in crea se. B e ca u se o f th e h ig h er p rice, U.S. exp orts fell as foreign co n su m ers red u ced th e am o u n t o f to b a cco p u rch a sed at th e h ig h er p rice. Foreign su p p liers resp o n d ed to th e h ig h er p rice by p ro d u cin g larger q u an tities of to b a cco . SOME LONG-TERM TRENDS T h e su pply an d d em a n d analysis su ggested th at the ad o p tio n o f a su p p ly co n tro l p o licy w ould lead to both a re d u ctio n in U.S. p ro d u ctio n an d a sm aller U.S. share o f w orld trad e an d w orld p ro d u ctio n . An exam in ation of to b a cco p ro d u ctio n an d qu ota tren d s d o cu m en ts th e lon g -term p ro c e ss o f red u cin g th e d o m estic to b a c co in d u stiy as a m e a n s o f m ain tain in g th e p rice su p p o rt m ech a n ism . C hart 1 tracks th e p ro d u ctio n of to b a cco in th e U nited States against th e p ro d u ctio n of to b a cco in th e rest o f th e w orld over th e p ast 30 years. It show s th at d o m estic p ro d u ctio n , th o u gh variable, h as b een tren d in g d ow nw ard w h ile foreign to b a cco p ro d u ctio n h as grow n steadily. Sin ce 1966, d o m estic to b a c co p ro d u ctio n h a s fallen by 38.8 p e rce n t, w hile foreign p ro d u ctio n h as grow n by 56.5 p ercen t. A lo n g er-term p ersp ectiv e on th e im p act o f th e to b a cco program re strictio n s ca n b e gain ed by ex a m ining acreag e data. T h e to b a cco program initially a t tem p ted to co n tro l p ro d u ctio n solely by restrictin g th e n u m b er o f a cres th at farm ers cou ld grow. C hart 2 show s th e lon g-term tren d o f falling acreag e a llo t m e n ts." As y ield s in creased , acreag e lim itatio n s b eca m e less effective in co n tro llin g p ro d u ctio n an d w ere aug m e n ted by m arketing qu otas th at lim ited th e n u m b er o f p o u n d s o f to b a c co farm ers co u ld m arket. C hart 3 show s th e tren d o f falling m arketing qu o tas for fluecu red and bu rley to b a cco s, th e two varieties th at a c c o u n t for 90 p e rce n t o f all d o m estic to b a c co p ro d u c- 9Starkey (1985), p. 50 and U.S. General Accounting Office (1982), p. 18. ,0Tweeten (1970), p. 201. These measures of demand elasticity are interpreted to mean that a 1 percent increase in price would lead to only a .1 percent decrease in quantity demanded in the intermediate run and to a .5 percent decrease in the long run. 11Although not shown in the graph, tobacco acreage in 1986 was at its lowest point since 1874 as a result of the supply control program. 59 JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS C hart 1 U.S. a n d Foreign Tobacco Prod uction Index Index 1970=100 60 1959 1970=100 62 65 68 S o u r c e : U .S . D e p a r t m e n t o f A g r i c u l t u r e tion. T h e ch art sh ow s that, after initially rising, po u nd ag e qu otas for th e se tw o to b a c co s generally have b e e n d ecreasin g in th e 1980s. As in d icated earlier, a red u ctio n in th e U.S. sh ares o f w orld to b a cco p ro d u ctio n and total ex p o rts is an ex p ected resu lt o f th e su pply restrictio n . T able 1 d o cu m ents th ese sh are lo sses. For exam ple, in th e 1 9 5 5 -5 9 period, th e United States a cc o u n te d for m ore th an 80 p e rce n t o f th e w o rld ’s p ro d u ctio n o f burlev to b a cco . By 1985, th e U.S. share o f bu rley p ro d u ctio n h ad fallen to 38 p e rce n t. Sim ilar tren d s are evident for flu e-cu red to b a c co and for th e category lab eled “all to b a c c o .” Not only have th e U.S. sh ares o f w orld p ro d u ctio n an d trad e fallen, but th e u se o f im p o rted to b a cco has risen su bstantially (see table 2). Until th e 1970s, th e u se o f im p o rted bu rley and flu e-cu red to b a cco w as n eg li gible. In 1969, less th an 1 p e rce n t o f all burley to b a cco u sed in th e United States w as im p orted. Bv 1985, im p orts a cco u n ted for m ore th an 24 p e rce n t o f all bu rley u se. O th er varieties n o t p ro d u ced in th is c o u n http://fraser.stlouisfed.org/ 60 Federal Reserve Bank of St. Louis try, su ch as O riental to b a cco , co n tin u ally have b een im p o rted for b len d in g p u rp o ses. A n oth er im p o rtan t tren d is th e re d u ctio n o f th e quality advantage th at U.S. to b a c co h old s over foreign to b a cco . N um erous so u rces a ssert th at th e quality gap b etw een foreign an d d o m estic to b a c co is n arrow in g.12 T his reflects th e fact that attem p ts to in crea se the p rice o f high-qualitv to b a cco have provided foreign p ro d u cers w ith an in cen tiv e to im prove th e quality of th e ir to b a cco . T h e resu lt o f a sm aller quality advantage an d risin g p rice s h a s led, p red ictably, to th e lo ss of b o th d o m estic an d foreign m arkets for U.S. to b a cco . Over tim e, th e d em an d for U .S.-p ro d u cea to b a cco h as b e c o m e m o re ela stic as o th e r so u rces o f su pply from th e rest o f th e w orld have ap p eared . T h e elastic- 12Starkey (1985), p. 50 and U.S. General Accounting Office (1982), p. 18. FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 C ha rt 2 H a r v e s t e d A c r e a g e o f B u rle y a n d F lu e -C u r e d T o b a c c o M illions of acres M illions 1.7 of acres 1.7 .5 1920 25 30 35 40 45 S o u r c e : U .S . D e p a r t m e n t o f A g r i c u l t u r e ity o f su pply also h as in creased . In th e sh o rt run, foreign to b a c co p ro d u cers w ere lim ited in th e ir re sp o n se to h ig h er p rices by th e ir lan d b ase an d o th e r facto rs su ch as th e know ledge an d tech n o lo g y n eed ed to p ro d u ce h igher-qu ality to b a cco . W ith tim e, h o w ever, foreign p ro d u cers have acq u ired th ese a d d i tional inp u ts. T h e result h as b een a d ram atic in crea se in th e qu antity o f to b a c co su p p lied by th e rest o f the w orld. As a co n se q u e n ce , th e im p act o f U.S. to b a cco policy on w orld to b a cco m arkets has d eclin ed . A lthough th e lon g -ru n ben efits o f su pply co n tro l p o licies m ay b e in q u estio n for U.S. to b a cco farm ers, b en efits for foreign p ro d u cers are obvious. T h e se b en efits are co n ferred in tw o w ays. First, by restrictin g th e su pply o f U.S. to b a c co initially th rou gh qu otas and la ter th rou gh th e m a in ten a n ce o f th e lo an stocks by th e grow ers' cooperatives, a h ig h er w orld p rice is m aintained . Second , the program crea te s a strong in cen tive for foreign p ro d u cers to im prove the quality of th e ir to b a cco by m aintain in g a h ig h er p rice in the m arket for high-quality to b a c co th an w ould oth erw ise result. N one o f th e se lo n g -term tren d s o f d ecreasin g p ro d u ctio n , falling q u o tas o r falling U.S. sh ares, how ever, w ere ca u se for co n c e rn . T h e p u rp o se o f su p p ly c o n trols w as to raise th e co m m o d ity ’s p rice and, m ore im portantly, to raise th e net revenue o f farm ers. For m any y ears, th e to b a cco program w as su cce ssfu l in th is resp ect. Over a re ce n t period, how ever, the program led to low er revenu es for to b a c co grow ers. From 1982 to 1985, th e p ou nd age a llo tm en ts for bu rley to b a c co fell by 30.4 p e rce n t. Over th is sam e period , how ever, the average p rice p aid to grow ers for bu rley fell by 11.9 p ercen t. T h e co m b in a tio n o f lo w er o u tp u t an d low er p rice tran slated into a 38.7 p e rce n t d eclin e in to b a cco re ceip ts for bu rley farm ers. RECENT PROGRAM DEVELOPMENTS AND CHANGES In the 1980s, the to b a c co p rice su p p o rt m ech an ism led to m a jo r p ro b lem s. T h e to b a c co p rice su p p o rt w as, 61 FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 C hart 3 P o u n d a g e A llo tm e n ts for B u rle y a n d F lu e -C u re d T o bacco Millions of M illions pounds 2 .7 of pounds ------------------------12 . 7 2 .5 2 .3 2.1 1.9 1.7 1.5 1.3 1.1 1971 73 S o u r c e : U. S. D e p a r t m e n t o f A g r i c u l t u r e Table 1 U.S. Percentage of World Tobacco Production and Exports Burley Year 1955-59 1960-64 1970 1975 1980 1981 1982 1983 1984 1985 Flue-Cured All tobacco Production Exports Production Exports Production Exports 82% 80 62 52 44 51 49 33 42 38 60% 57 33 27 27 21 30 22 18 26 41% 40 30 28 20 18 13 13 12 10 60% 52 46 33 29 26 23 22 24 22 23% 25 19 18 16 16 13 11 12 10 35% 30 28 20 20 18 18 17 17 18 SOURCE: U.S. Department of Agriculture, Tobacco Outlook and Situation Report (December 1986), p. 42. http://fraser.stlouisfed.org/ 62 Federal Reserve Bank of St. Louis FEDERAL RESERVE BANK OF ST. LOUIS JANUARY/FEBRUARY 1988 th e stored to b a c co in crea sed . T h e a sse ssm en ts for 1986 w ere estim a ted at 30 ce n ts p e r p o u n d o r $600 p e r acre. Table 2 Percentage of Imported Flue-Cured and Burley Tobacco Used Domestically Year beginning July 1 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 Flue-Cured Burley 0.9% 1.6 1.7 1.9 2.8 3.4 3.5 4.6 8.3 9.3 13.1 11.7 11.5 17.7 17.6 20.9 24.1 0.6% 0.6 0.9 1.6 5.4 8.4 8.4 7.2 14.7 15.1 18.6 22.3 19.1 24.1 25.8 28.9 24.5 SOURCE: U.S. Department of Agriculture, Tobacco Outlook and Situation Report (April 1987), p. 14. an d still is, ad m in istered by g ro w ers’ cooperatives, w h ich p u rch ase d su rp lu s to b a cco an d h eld it u ntil it co u ld be sold at a p rice above th e su p p o rt level. Any lo sses on th e su rp lu s stocks w ere ab so rb ed by the C om m od ity C redit C orp oration (CCC), w hile gains w ere red istribu ted to th e cooperative's m em b ers. As th e stocks h eld by coop eratives co n tin u ed to grow bu t p ro sp e cts for selling th e se stocks at a gain seem ed rem ote, th e p o ten tial co st to th e governm ent in crea se d greatly. Legislation in 1985, how ever, relieved grow ers o f the p o ten tial liability for lo sses on th e stored to b a cco . T h e CCC took title to th e su rp lu s stocks an d sold th e m at d isco u n ts o f u p to 90 p ercen t, resu ltin g in a n et loss o f ap p roxim ately $373 m illion . T h is loss w ill n o t b e re covered th rou gh th e No Net C ost Act. In ex ch a n g e for th e g ov ern m en t’s rescu e, to b a cco farm ers a cc e p te d lo w er su p p o rt p rice s. B e ca u se o f th e low er p rices, th e a sse ssm en ts fell to only 2 ce n ts p e r p ou nd . T h e U nited States D ep artm en t o f A griculture (USDA) also w as given in crea sed freedom to red u ce to b a c co p rices fu rth er if n eed ed an d w as p erm itted to u se a m ore m ark et-o rien ted m eth o d o f calcu latin g su p p o rt p rices an d settin g q u o ta s.13 T h e n ew to b a c co program has resu lted in su b sta n tially low er p rices. T h e average to b a c co p rice paid to grow ers fell from $1.80 p e r p o u n d in 1985 to $1.45 p e r po u n d in 1986. As a result, to b a cco exp o rts ro se in 1987. Im p orts also fell an d n o w re p rese n t a sm aller sh are o f th e to b a c co u sed in th e U nited States. M arket ing qu otas also have b e e n in crea sed in an ticip a tio n of grow ing sales. CAN SUPPLY CONTROLS BE USED EFFECTIVELY ON OTHER CROPS? T h e initial s u c c e s s o f th e to b a c co p ro g ram ’s u se of su p p ly co n tro ls c a n b e a ttrib u ted to su pply an d d e m an d ch a ra cte ristic s th at are n o t p re sen t for o th e r m a jo r crop s. T h e to b a c co program ben efited from the fact th at th e d em an d for U.S. to b a c co w as in elastic b ec a u se o f a lack o f a good su b stitu te. A dditionally, th e w orld supply w as in e la stic b ec a u se th e U nited States held a d om in an t sh a re o f th e w o rld ’s p ro d u ctio n . In resp o n se, the No Net C ost T o b a cco Program Act o f 1982 w as p assed . T h is a ct stip u lated that the to b a cco program be run at no n et co st to th e govern m en t o th e r th an adm inistrative co sts. U nder th is law, a sse ssm en ts w ere levied on grow ers an d bu yers to su p p o rt lo sses in cu rred by th e program . In 1985, b o th buyers an d p ro d u cers o f flu e-cu red to b a c co w ere re qu ired to pay a sse ssm en ts o f 7 ce n ts p e r p o u n d to cover p rogram co sts. T h is am o u n t w as roughly equiv alen t to $140 p e r acre for farm ers. ,3The support price formerly had been determined by a combination of the parity index and limits set by the Secretary of Agriculture. Tobacco support prices currently are determined by a formula using five-year moving averages of tobacco prices and year-to-year changes in costs of production. This approach is substantially more “market-oriented” than the previous method, which was driven by costs of numerous products unrelated to the open market for to bacco. U.S. to b a c co su rp lu ses grew as th e gap b etw een th e su p p o rt p rice an d th e w orld p rice w id ened and im p o rts gain ed a larger sh are o f U.S. to b a c co m arkets. W ith less d o m estic to b a c co bein g sold on th e m arket, th e cooperatives p u rch ase d m ore su rp lu s to b acco . As a result, th e grow ers' p o ten tial liability for lo sse s on The USDA determines tobacco quotas based on three factors. The first factor is the intended purchases of tobacco by cigarette manufacturers based on the support price. Cigarette manufacturers must provide these estimates and purchase a minimum of 90 percent of their stated intentions or face a penalty. The remaining two factors are the average tobacco exports of the past three years and an estimate of the quantity of tobacco needed to maintain tobacco stocks at desired levels. 63 JANUARY/FEBRUARY 1988 FEDERAL RESERVE BANK OF ST. LOUIS M ost, if not all, o th e r m ajo r cro p s do not en jo y th ese ch a ra cte ristics. F or exam ple, if th e U nited States w ere su cce ssfu l in restrictin g th e p ro d u ctio n o f co rn an d raising its m ar ket price, co n su m ers w ould m o st likely sw itch to any o f th e n u m ero u s co a rse grains su ch as barley, so r ghum , m illet o r oats, w h ich are a ccep ta b le su b stitu tes for m an y o f th e feed u ses o f co rn . On an in tern atio n al level, th e U.S. sh are o f th e w o rld ’s co arse grains is sm all. If it w ere to im p o se su pply co n tro ls on corn , it w ould be n ecessa ry to re strict greatly th e im p ortation of foreign grain th at w ould o c c u r in re sp o n se to h ig h er U.S. p rice s. Su ch trad e restrictio n s m ight negatively affect th e ability to exp ort o th e r U.S. co m m o d ities. In so m e crop s, th e U nited States d o e s have a large sh are o f th e w o rld ’s p ro d u ctio n . B eca u se of th e avail ability o f su bstitu tes, how ever, su pply restrictio n w ould be ineffective. T h e U nited States, for exam ple, p ro d u ces m ore th an h alf o f th e w o rld ’s soybean s. U nfortu nately for ad vocates o f su pply co n tro ls, o th e r cro p s like co rn , co c o n u t an d co tto n seed can b e su b stitu ted for so ybean s as in p u ts for ed ible oil p ro d u c tion. An ad d ition al facto r restrictin g th e p o ten tial u se of su pply co n tro ls for o th e r crop s is th e w orld elasticity o f su pply o f th e se crop s. M ost cro p s for w h ich supply co n tro ls have b ee n co n sid ere d in th e U nited States ca n b e p ro d u ced th rou gh o u t th e w orld. W heat, for exam ple, is p ro d u ced in m ore th an 100 co u n tries. If th e U nited States w ere su cce ssfu l in raising w h eat p r ic e s b y r e d u c in g p r o d u c tio n , o th e r w h e a t- elasticities an d th e foreign sh are o f w orld p ro d u ctio n have in crea sed . T o a large exten t, th e d eclin e in m ar ket p ow er ca n be a ttrib u ted to U.S. p o licy a ctio n s. In re sp o n se to this d eclin e, th e su pply co n tro l program h as b ee n altered to b e m ore m ark et-o rien ted in settin g su p p o rt p rices. T h e o th e r m a jo r cro p s for w h ich su p ply co n tro l legislation h as b ee n p ro p o sed do n o t have th e n ecessa ry su pply an d d em an d ch a ra cte ristics n eed ed to su ccessfu lly im p o se a su pply co n tro l p ro gram , even in the sh o rt term . REFERENCES Belongia, Michael T. “The Dairy Price Support Program: A Study of Misdirected Economic Incentives,” this Review (February 1984), pp. 5-14. Galston, William A. A Tough Row to Hoe: The 1985 Farm Bill and Beyond (Hamilton Press, 1985). Grise, Verner N. “The Tobacco Program: Recent Changes and Issues in the Eighties,” Tobacco Outlook and Situation Report, U.S. Department of Agriculture, Economic Research Service (U.S. Government Printing Office, March 1984). Houthakker, H. S., and Lester D. Taylor. Consumer Demand in the United States-1929-1970 (Harvard University Press, 1966). Johnson, Paul R. The Economics of the Tobacco Industry (Praeger Publishers, 1984). Pugh, Charles R. “County Lease Rates for Flue Cured Tobacco: Revisited,” Economics Special Report No. 66 (North Carolina State University, May 1981). Starkey, James H. “An Industry Viewpoint of the Tobacco Program and its Effects,” in Removal of Price Supports and Supply Controls for U.S. Tobacco (National Planning Association, Washington, D.C., 1985). Sumner, Daniel A., and Julian M. Alston. Removal of Price Supports and Supply Controls for U.S. Tobacco: An Economic Analysis of the Impact (National Planning Association, Washington, D.C., 1985). p ro d u cin g co u n tries w ould b e able to resp o n d quickly by in c re a sin g p ro d u c tio n w h ile th e n o n -w h ea tp ro d u cin g co u n tries w ould have incen tives to b egin to p ro d u ce w heat. Tomek, William C., and Kenneth L. Robinson. Prices (Cornell University Press, 1981). SUMMARY U.S. Department of Agriculture, Agricultural Marketing Service. Annual Report on Tobacco Statistics, various issues. C ontrolling th e su pply o f ag ricultu ral p ro d u cts has received atten tio n recen tly as a p o ssib le so lu tio n to th e p ro blem o f falling farm p rice s and grow ing co m m odity su rp lu ses. T h e original to b a c co program p ro vides an insigh t into th e likely effects o f su ch farm p o licy ch an g es. T h e to b a cco program en jo y ed initial su cce ss b eca u se o f u niq u e ch a ra cte ristic s o f th e su p ply o f an d d em an d for to b acco . T h e m arket p ow er of th e U nited States in th e w orld to b a cco m arket, h o w ever, h as d ecrea se d over tim e as su p p ly an d d em an d http://fraser.stlouisfed.org/ 64 Federal Reserve Bank of St. Louis Agricultural Product Tweeten, Luther. Foundations of Farm Policy (University of Ne braska Press, 1970). _________ Agricultural Stabilization and Conservation Service. History of Budgetary Expenditures of the Commodity Credit Corpo ration, Fiscal Year 1980-1986 Actual, December 29, 1986. _________Economic Research Service. Tobacco Outlook and Sit uation Report, formerly the Tobacco Situation Report, various is sues. _________ Tobacco, Background for 1985 Farm Legislation, Agri culture Information Bulletin Number 468 (National Economics Divi sion, September 1984). U.S. General Accounting Office. “Tobacco Program's Production Rights and Effects on Competition,” Report by the Comptroller General of the United States, CED 82-70 (April 23, 1982). F e d e r a l R e s e r v e B a n k o f S t. L o u is Post O ffice Box 442 St. L ouis, M isso u ri 63166 T h e R e v ie w is p u b lis h e d six tim es p e r y e a r by th e R e s e a rc h a n d P u b lic In fo rm a tio n D e p a rtm en t o f th e F ed era l R es erv e R a n k o f St. L ou is. S in g le-co p v s u b s c rip tio n s a re available to th e p u b lic f r e e o f c h a r g e . Mail r e q u e s t s f o r s u b s c rip tio n s , b a c k is s u e s , o r a d d r e s s c h a n g e s to: R e s e a rc h a n d P u blic In fo rm a tio n D ep a rtm en t, F ed e ra l R es erv e R an k o f St. L ouis, P.O. Rox 4 4 2 , St. L ou is, M isso u ri 6 3 1 6 6 . T h e views e x p r e s s e d a r e th o s e o f th e in d h id u a l a u th o rs a n d d o n o t n ece ss a rily re fle c t official p o s itio n s o f th e F e d e ra l R es erv e R a n k o f St. L o u is o r th e F e d e ra l R es erv e S ystem . A rticles h ere in m ay b e r e p rin t e d p ro v id e d the s o u r c e is c re d ite d . P lea se p ro v id e th e R a n k ’s R e s e a rc h a n d P ublic In fo rm a tio n D e p a rtm en t with a c o p y o f r e p rin t e d m aterial.