FEDERAL RESERVE BANK OF ST.

LOUtS

JANUARY/FEBRUARY 1988

Protectionist Trade Policies: A Survey of Theory, Evidence and Rationale

Cletus C. Coughlin, K. Alec Chrystal and Geoffrey E. Wood

ROTECTIONIST pm-essures have been mounting won-idwide during the 1980s. These pn-essmmres are due to various economic problems incltnding the lam-ge and persistent balance of tnade deficits in the United States; the hard times expenienced by seven-al industries; and tire show growth of many foreign countn-ies. Pn-oponents of protectionist trade policies amgue that international trade has contributed substantially to these problems and that protectionist tmade policies will head to improved nesults. Professional economists in the United States, however-, generally agree that tmade restn-ictions such as tariffs and quotas substantially reduce a nations economic well~being.2 This article surveys the theory, evidence and nationale concen-ning protectionist trade policies. The first section illustrates the gains from fm-ce tn-ade using the concept of compan-ative advantage. Recent demelopments in international trade theorm that emphasize other reasons for- gains from trade are also reviewed. The theoretical discussion is followed hiy an examination of recent empirical studies that demonstm-ate the large costs of protectionist trade policies. Then, the rationale for restricting trade is presented. The concluding section summarizes the papers main arguments.

THE GAINS FRONt FREE TRADE

The most famous demonstration of the gains fn-om tnade appear-ed in 1817 in David Ricardos Principles of Political Economy and Ta,vation. We use his example inmolming tn-ade between England and Portugal to dem onstnate how both countries can gain fn-om tnade.. The two countnies pn-oduce the same two goods, wine and cloth, and the only pn-oduction costs ane labor- costs. The figures below list the amount of lahior leg.. won-ker-daysi requin-ed in each country to pn-oduce one bottle (if wnne on- one hiolt of cloth. Wine England Portugal Cloth

3
1

Since both goods ar-c tiione costly to produce in England than in Por-tugal, England is ahisolmrtely less efficient at producinig both goods than its prospective trading partner. Portugal has an absolute advantage in hioth wine and cloth. At fin-st glance, this appear-s to r-ule. (iut mutual gains fn-om trade; howeme,n-, as we demonstrate below, absolute advantage is irrelevant in discerniing whether tnade can hienefit both countn-ies. The natio of the pr-oductioni costs fon the two goods is diffenent in the two countries. hn England. a bottle (if wine will exchange for 3/7 of a bolt (if cloth becamnse the labor content of the wine is 3/7 of that for (:10th. In Portmtgal, a bottle of winie ivill exchanige for 1/5 of a bolt of cloth. Thus, wine is relatively cheaper in Portugal than in Englanid and, conmersels, cirith is relativehs cheaper in England than in Portmngal. the example indicates that Pom-tugal has a compam-atime admamitage in mvimie pr-oduction and England has a comiiparative advantage in cloth pnocluction. The different relative pm-ices provide the basis fom

Cletus C. Coughlin is a senior economist at the Federal Reserve Bank of St Louis. K. Alec Chnystal is the National Westminster Bank Professor ofPersonal Finance at City University, London. Geoffrey E Woodis a professor of economics at City University, London. This article was written while Chrystal was a professor ofeconomics 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 (Keart 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).

FEDERAL RESERVE SANK OF ST. LOUIS

both countries to gain fn-omn iriternational trade. The gains arise from both exchange and specialization. The gains from exchange can be highlighted in the following manner. If a Portn.mguese wine producer sells five bottles of wine at home, he receives one bolt of cloth. If he trades in England, he receives more than two bolts of cloth. Hence, he can gain by expon-ting his mvine to England. English cloth-pn-oducen-s are willing to trade in Portugal; for emeny 3/7 of a bolt of cloth the sell there, they get just (iven twri bottles of wine. The English gain from exporting cloth to land imponting wine fn-omi Por-tugal, and the Portuguese gain from expon-tingwine to land importing cloth fromi England. Each country gains by expon-tinig the good in which it has a compan-ative advantage and by impon-ting the good in which it has a compam-ative disadmantage. Gains fn-om specialization can be denionstrated in the follomming manner. tnitially, eachi country is ptoducing some of both goods. Suppose that, as a n-esutt of tr-ade, 21 units of labor are shifted fnom mvine to cloth pnoduction in England, while, in Portugal, 10 units of labor- are shifted from cloth to wine production. lhis reallocation of labor- does not altet- the total amount of labor used in the two countn-ies; howeven-, it causes the pn-oduction changes listed below. Bottles of Wine England Portugal Net 7 10 3 Bolts of Cloth 3
9

JANUARVIFEBRUARY lies
labor was the only resource used to produce the two goods in the example above; yet, labor is really only one of many resources used to prodmnce goods. The example also assumed that the costs of producing additional units of the goods ar-c constant. For example, in England, three units of labor an-c used to produce one bottle of wine regardless of the level of wine pn-oduction. In reality, unit production costs could either increase or- decrease as more is produced. A thim-d assumption was that the goods an-c produced in per-fectlv competitive mat-kets. tn other- won-ds, an individual firm has no effect on the price of the good that it pnoduces. Some industries, however-, are dominated by a small numhier of firms, each of which can affect the manket pr-ice of the good iw altering its production decision. Sonic of these extensions at-c discnnssed in the appendix. These theoretical developnients genem-atly have strengthened the case for an open rn-ading system. They suggest thn-ee soun-ces of gains fi-om trade. First, as the market potentially served by firms expands from a national to a world market, then-c are gains associated with declining per unit pnoduction costs. A second sour-ce of gains results fn-om the reduction in the monopoly power- of domestic firms. Domestic firms, facing more pressure from fon-eign competitors, are forced to produce the (iutput demanded by consumers at the lowest possible costlhird is the gain to consumers from increased product variety and lomver prices. Generally speaking, the gains fiomn tm-ade r-esult from the increase in competitive pressures as the domestic economy becomes less insulated from the world economy. The gains fmoni free trade can also he illustrated gt-aphicallv. The shaded insert on pages 14 and 15 examines the gains fiom ttade in pemfectly competitime markets using supply and demand analysis. The insert also analyzes the effects of tn-ade n-estn-ictions, a topic that we discuss below.

The shift of 21 rnnits of labor to the English cloth industr raises cloth production by three bolts, while reducing wine production by seven biotttes. In Pomtrngal, the shift (if 10 units of labor from cloth to wine raises winie pmoduction liv 10 bottles, while n-educing cloth pm-oduction by two liolts. This reallocation nif labor imicreases the total pnoduction of Iiothi goods wine by thn-ee bottles and cloth hiy one bolt. This increased output will be shared Iiy the two countries. Thus, the consumption of liothi goods and the wealth of both countries ane incn-eased by the specializa tion bm-ought about by trade based on compar-ative advantage.

COSTS OF TRADE PROTECTIONISM

Protectionist trade policies can take numenous forms, some of which are discussed in the shaded insert on pages 16 amid 17. All forriis of lirotectioni an-e

TRADE THEORY SINCE HICARDO

Since 1817, numerous analyses have gener-ated insights concerning the gains from tmade, lhiey chiefly examine the consequences rif relaxing the assumptions used in the pn-eceding example. For example, A 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 catted X-etficiency, has been stressed by Leibenstein (1980). The increase in competitive pressures due to international trade reduces the probability that costs are not minimized.

A Supply and Demand Analysis of the Gains from Free Trade and the Effects of a rliarjff

A starndaid flinstr-aliori tnt liii gains 1mm tree hat], arid tie ctin-ls ol il till ill is pn~crnturIhrIm~I lu ,rl,i\ sk issl_rnhis Irr-tr-I~m m-onrpctiti Lets tlniorrghoni.
-

nw-

(.ains trout tree 1 rude

tin tigtri-i I tit- rum-s S anti I) an tint L S. sirinpi\ unit] rienirunti rmr\ts Irir a lnptnliiehrral wind. lit-iultirrsprluin at B irsmrlls In the rritrrirlni-itrrir maltips tot rri~- arrd qmrarrlit\ ii I arid 0 \~strminuti1 etc-c] Stai,-s has a ronlpar-.itn ( dnsaLi\ uoNr4e

tiad, puikm liii nragrrilLrdr- mt tlrtse grin, and Io~s-s trsing thr- cm mt-ph ut roiIsmirniii- .uiti jil ii uiu,-i-i smnr- nltjs air inn seen iii tir4lri-i I - C mrrp,LrrrIt-is 1 gain i tno na s_ Iniliaiim ,tniistrnnitr-s i;irrclras-tI ft. al a purr 11cr trnil ii I - flitir Ire tiatlr. Ihm-~ inmrriliast V. at till- inner 11cc ill-n minnrt cr1 P I iii,, r~arrrrs r-epr-rsenteti mm tb, rcI-tanMie I lii -- I. - In atidrtioni tIn irn~,r 1niict irrnlirt-ts ,nrisirnltrs rn ii i-ri-_tsr Lli,ii- i)triilizisis h,rnn 0 tin iF -- liii, gain ~ n-pi r,rriirci Irm I ht Ir iarngic- tIC I -l he total g;tir I to 5 rininsLnrnrer-s is I IC i_ in nsirng the- itrnm-r case IcItu-i s to rrpi-t-srrit tiers a - Ii - r .\inalogoList\ piteltrici s lo,c un,- to the loner- ruin- tIr,-m ri-the br Uteri- nmitirmrt. . - anti kr,- to ihprrrnrrnti-artimnri nil ]~nrtitcIrun li-tin, to S - I he tori1 toss tir dnrrr-s is I - Bi I ora tic- inctimri as a nhrln ~zrinis inr,,nusp lint- curl snrnwr ialris mt a - In emcee,l tin innnmlmrcmisses otah~in -- c I inisanahsisranalsuIni-mrened irsirin4agunti ihat inc t nnIt,i5tttr,c\jflnri~ in ritin,-,nnri-cis tire t rlited States mmiii inam, a nunirpriilIi\n atlurnitag,- hr itt innrulrrrlnn cr1 a gtnrcl I or Liii export ~ocrd. the change tin lit, track- mmiii miLriurmrtirct-r gains liar i-u-c-mci rnrin,imnncr losses.

pitidirtioni cii this gnu,] lii, Pr icr- n ill lit ion crabroad than iii tie trnit,-tl stains i_it [Iris tinner ~~m,nici ri I- he P P rind asurne th,t t s prrtIisr-s drn n,,t altect this nonid Jilt- (.rapirinalk his , iiiiri-seintrmI iiv lint Iic,ir,,rrit,i t~mnrtcistr nptm tmrrmc, 1 S It onpalionslurti-tp trade this lunerncnrtl ~ has tnuu ittirts. I insI I s (trnstrriwrs mmiii inlrrvasp tin,-ir- c-runi.,Iirin nlicnul 1,1 I nrrci. t S. nnindnrm9s 1 mmiii contract their pi-(nultrm-tiorr In S~ iii, (mimss oh S muniistrnipliuri n.Imen-prciciLwnoin is I _S nnrrliasrs tiom tor-,-i~niprodtrr-er-s hat k. rnrportslint Ion cc ill_u_i- ~innmillanernu~Imhem-IrIs t roii srrrritrs ain,i hunts I _S iirintlircnr~ a fact tirat civil es tin,- nec-cot cturitiinu-rsral mli,c-tesruins in! I 5.

Ihe E/Jiets oju lari/J

Fiq.-ro
I

tiu maLi lie ariatmsisoi proteu-tiorli~~Irade rulirm t rs strrrghttmrrmmmrti is ncnsiinii tiii- jrrmji~t,-t nI is ;mri,ilm,.c-mt (urn, rain mn-mm aIim nonlerlrorrist li-aiimpotu:m inmnmm~-mi-r.rniirni.iralirgmnnmsirnairnncr- tmr cr11 mc-iiiirrce h, in,,- tr,itii- -esirits rn ligmrnc t wi uhrpIi rated ri lrgiri-i 2. Gim,ni ihc In-,-,- Lr-,icic- mmmuntd ~r~~- cr1 I. t S rcnnscrniniutroir rnn,lirctnnir arid irinpcni-ts aJ-r Ii s acid S. U~ I ruled a tar nacrsiirg he pi-it n irn lit \,-,trrne slates Iticrrs inirj_rmus,-d rin,-i-e;rw to t~ i lu p1-itt- in tire I rued Slates mrnmm r-u-i-c-ck prirmin tInt mmrutd irm the arriorunt mt tin,- unit P_U

I~
A
US

B
Pu

bHN.,c
F

s
w

I ~

mis

us
-

i he higher I ~ p ri 1 Jrtists t-miinsLrrnir it mai-s tin ,iucr-t-asi lr-mnrni i) Inn il. clcnnretir pm) dmncticnmn unnnnn-;rspli,rnrs InnS and rrlnirurlstmr denrnirp lninlnr 5 h) Iii S ii il. imni io,,ini~lint 1 tar itt nunistrnilers mist tIlt- ar-pa P_iC P. mm ci - p gamid prid,rier~gaun th i-area PH i and Urn-

FEDERAL RESERVE SANK OF ST. LOUIS

JANUARYIFEBRUARY ION

not-slit- intudtnm tnri priitertc-mi at tint

n\ nuiist~ru 1

titnnlnmslit ,-lniIsmlnie-s (Jilt- c-nnri iit-atirjmn sti-nlns minim tarili nemeuiint~ I am 1 ill rmenurc mmllicIl marl lit- m iemmemi as a gain Rn the gtu\ en runt-mt t-Cithiis the hrntt P_P tinlirs tlr, tjtnani

-nyu. p

tit~ nt ulm rumts 1 t 1(11 tnm I. tcr~st-scIi t

1Y

t In, il-men,rc is Ninal tin aica

C)m,-r-ahi lint- rnalnrnmn i,)sis ini-e;Imrsc lint ccumnsnrmiltrs

Al
F H

f
~

r~~1
W

:n.~ :y: m~

t:
-

i-

miemmed

-t--

-c itr,s nm-smiting mum

z
I

n_&emm-eni the prLruhnm-11-s g;rirls

::.m~t:~1Y;~:
imnehlicienit

I
u~ us
-

us
~-

nss

US

dm,mnestim inn ,InicI mum mmliiiian-agii-ailiti a mid mmti.ght imrss nisttitinng room nii-Iliiient lie mirmmcti c-unsnnnptnntn 11)54 amid mrrl ttnn hltic is ,, mm nmlslm imi~it ii In

mnili-rm,hith tim immmimllmm liii inrn_ilmnmr nt ninrnnn,stmr- nitntmmi tin IuinnIr;mn iiichmmii-ns I rn inn inn ,iinmmi Iinm-rnnngii imrim i-its Iinnt nir,-n n~sl hit,- lnm,nin, inrtii~i-limrn ml mini- trrn-,n&~nm r mnii,hnrm-l ,h,-mnr-tst- tIn, pm I nihnn,tnmnmn rmn~ts nI citimmirsi ml tim-iris mr srnmniiiiiimm mnsini,l Ilir- ,ic-n-I-~s ri

rlmrrnns inn inrniilit him i-~rmni nir it c nirrtn, iii 1 1 cr11 rnninrrits mimni, lmnncii_znm niiiihnn,im nir~nm niiiimc hnigiii-m nniris tmrrrln,-in imptrnts lii tim rninlictr,i mmniri,_,-I

rhmnri-n Iii liii- chrnmin,-stnr nmnriL,-t The specific goal of protectionist tn-ade policies is to expand domestic pn-oductioni in the protected indus ti-ies, benefiting the owners, workers and suppliers of r-esour-ces to the protected industry. The govennment imposing pr-otectioniist trade policies may also benefit, for example, in the form of taniff revenue.
-

Then-c have lieeni numerous studies of the costs of protectionism. We begin by examining three recent studies of pm-otectioriism in the United States, then proceed to studies examining developed and finally, developing countries.

Costs qf Protectionism in the Crated

States
Recent studies by Tan and Morkre 119841, t-lickok 1985) and Huliuauer et al. 11986) esttmated the costs of pi otcctnotusni in tin Unitc d St ntis I lit s studn s usc . - -. different estnmatnoni procedunes, examine diffem-ent - .. protectnonnst policies arid cover dnfferent tnriie periods. Norn thiclcss thcv ptovnde consnstenit mesults Tair and Mork,-e 11984) estimate annual costs to the U.S. economy of 812.7 billion 1983 dollars) from all tan ifs and from quotas on automobiles, textiles, steel and sugar-. Their cost estinuate is a net measur-e in which the losses of consumers are offset partially by the gains of domestic producers and the U.S. government. Estimates liv thickok 1985) indicate that tr-ade rt~ strictions on only three goods clothing, sugam-, and automobiles caused increased consumer expenidi tunes of $14 billion in 1984. I-Iickok also shows that low income families am-c affected mom-c than highincome fnnilies. The import restmaints on clothing, sugar and automobiles are calculated to be equivalent to a 23 percent income tax sun-charge that is, an additional

1 he expansiomi of doniestic pn-odnnctnon mni protected . - nndusti its is not costless it n quit is tddntnonal sources from other itidustrres. Consequently, output . , , . iii other domestic industm-ies is i-educed. these nndus - . . trres also might be made less competrtnve hecanise ot higher pt-ices for imported inputs. Since protectionist tr-ade policies frequently incr-ease the price of the protected good, domestic consumers are liar-med. lhey lose in two ways. I-inst. their consuniption of the pr-otected good is n-educed because of the, associated rise in its price. Second, they consume less of other goods, as their output declines and pm-ices r-ise. The preceding discussion highlights the domestic winniers and losers due to pnotectionist trade policies. Domestic produrcers of the liiotected good and the government if tariffs an-c imposed) gain; domestic consomers and otliei cloniestic pnoducem-s lose. Fon-eigni interests an-c also affected by trade resti-ictions. The pnotection of doniestic pn-oducers will harm some foreigni pmoducers; oddly emiourgbi, other foreign pmt

Fermi
N

Of Preticfio\nisrn
/ N N

~wa p fl*\

AU
lb a

In \iuapl nS fp 1

t nit

in~ns

in

mtbbmeima \p
eatesicrease I

t mi
;a

Thz ad

darn A on

o. I d

nIne
-

Thi tab
at,
tbc\

that

n4ie

at

~sncprdu
~,.

IS4t

~ dltYeier aqtwa1sh\~t t gn ~ nw the Aue$~n

\S~W~

d\
N

m$ht\

In

Tar4J~
T
\N ,/~/ /

N /N /

N
N

1:

which
b

iS

&
N

4nii
/ /

b/

twagN

ii U11~

U~e d
\\\\/NNN ~

N*~

/~. :,N~

t,p

N~>h
N, ~

t
N

~t
/ /N~\ N / N /

:s
-~ /(7

ta~
N
N N \N

v:
/ N

wartm$

/~

>N,

a
N N

It
~.

in
N

atn~
N

i$%ta
N

tt4zd~tcnySaniers
~ N \N~, N,~,, ~ ~,N,S, N N ON

I.
N,

0
N\~~

17110
N\~/\N\ N

ft t ntIS ide oov ill dollatamounts lb

kim 0ff
Np ,

twsay
,t hug tEes

ev Lw the
anfrate~
N N N

ci
I

bSct,w
fl
N N N~

,fhel S

Sat
tnt
nipt /

01K
_

1It,~l%i0ujdw

habeenhgoned
NO

tanf thG A
N

nt pin

LSES\

Phr,

ad
t

I
sib t
,

ifi

othbmn~vai~.

i~

frMnak?v
~ N

ci

r
N

NbJ,k,

aint the prop uwMrhNq!iN I ~e tIpniQ

N

tnitant
~ N

qo

I
N

s4
\N

n~

ci Io*y
N

~,

Quotas
/ NqN,

wbetb

lisa
thd~

cI.
LI

a
p
N

binned
lb

tnamnin nitntbei

ttiSf, tin 1 as 11
i(pac

ntiS pa St ani fr I) Tb land c&n en dni mM tb~ thatnmptnteNodrodtarept w~drngt. bIONitO ,da and atdN adb ~ t t o
N -

I tia

IIiqtbahm~

CII

ci

flKNOLkPflC

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY ION

sltim-imonis miram be lint imni nm,sitrtnni cit salt-tm on- iotltn 1 1 tmoni st(tmn,l:tn-cls that mu-n-c mini I~ im,mislm bt-imng met hm tomuizmn m;mms
.

i-:_1rIIum~tccj,nrols
_th ot 11mm alflimm- ri-late dir-entim tin tilt iimimm ol gomnci. \ hmmai class ml n-,-stm-ic-tiomis mmmnrks b~nt-strict imng ai-,-i-ss hr tb~ inr-emgri rmionnn n tqinin-tci to imnim hnimm4ui 4oocis. I tim tannnplt. a r~omem nimmitnI that mm stied mm l)itflcct its n\pmnmIimng mutt irnilirimi mtmumnpelimng irititnstm-ie, mat tim- t,m hold ils n~t-hang,ate artmht-nailm Iomm \s a mcstnil tcnnvigmt goods mmmnmmi~i ~ m-x nc-nisnmc- iii tic- boone mnnarkt-t mmliii,- inomi1 gomnds mmmnnimi lit- cheaji tim el-st-as bonn- puunmhnt-nt-s irnm ilim-itim au-n- smrbsrtin,tml anti horune ,omrstnnuc-r s liii 1 ,Iititlm all- tamed I mis jn,itic~ is nmm,nmnrailm man-ct tci 1 srmstainn lii, merlIn at barth mi holctimng tint- t-mu-hnng,i-ate domm ni has tin hunt- hmrvmgrn t-~c-liamngt with itorne~ tit- rum tint - I its nmmm It isscnmcl cionrmmstic cnnnmnem imnt-mt-aa-s the donirtstir nimomnt~m sunk amid evmmnttrailm m-amnsm-s imuiialimnnn tnmllnitiomnaum iioIntits am-p mini non unnalim r-egai-cied as ni s,rniInlt~mvam or h)m-l)tet-titnr4 dci nine_,tit immdmtstm I ln,t-n- is amn,ither asinec-t In) emr-hamnMc- nomnim-inis Iii, jrmstitimatitimr is tin(tt pm-~mc-mininng limimnun, m-esitipmnts tm-nnnim inimtstinis4 tr\ cm Sta, immtmietils ,Iinnnit,sti, gu-cmmm tin (us ml lead, to greater tttimmn,stn, nvtl imImestniitnlt tmi nm-alitm ml ,rtmld tim emanli_m tIne opptnsitu Restr imtinmg an-c css to tnmnt-ign ass,-ts nnrav raise tim, mamiznnee anti tnnmmem lint- nttnnm n It, dJmmilc9s iii ,tmnmiiestit mm tnltln. tnn tin, simmnmt 11111 ii alsni rmn(t\ nppnur erie lu dnmmnntsUr esnhanngm rzntt tund him-mtin - wiLt ticnrnmn-sth- pmmn~ cttntmns ins ,trmnm ),Iitimi 1
-

Suhsul:es
.\ri ;mtrenunarime to u-c-~ink-timm tint- tm-minus ninth-itmlntm In him n-igni-n-s man nomuhncte nun tine lmomime mar kc-I N to stibsmdm/e cloinuestir jnmumthmeen-s.sulnsuhes mmnam lie Iortmsntl unron ami imidItNin in M~~~ mr mtpon the eqnont aetnm itos ol tInt innntnnstnm ~mi -\armnple tnt tintm hia-insact In \I,inimi arid \lcgmin t,4 N time tm,mnmliimnati,run ol n-mu dii mnomzu_nnnns sint-m-ial tam nnm,-emntim c-s anrci diii-, t stubsidm pam mnnm-nts hat bemnttit tic I _s_ sinipimumhhng inndLmstmm - _~mn esanmn nle nil lint 1 latter us tine iimnamieial _msmstinnre inn lease e\jn~nn to Is nomi,l,-ci in lint t ~, I-.minnmmtimniptnmt ilanik thm-nmtngh nlnnucI loans imiamr ~irtarautues amid imnstnn amine. annit mtm._nnmmnmnt mn;mnis Iii titln,n case rim cn,itrttic,rn ill l\ln(nmmtt _\ni innllioi-iamni clill,n-,rin-t- lu-tmmet-tn strlisi,iies annni t;mniti, imnm mm i-s tint- u-im mnlmnt- imnriilit-ationms rim gt,mmmmn miemnt_ tin, lmmn-nmner mmnttnlmes be gomei-minimennt inn in~.\ imnn_r nitnt nim,nmn,m mmii, nm-nt, tam-ills gentnate nlm-otnn,- tom lint- grim em-nunri-unt -t me t-tt,-m-t mnnm ,ionmnestir rniidLnntiomi ;nniti mmeliart imnrmmnm tm tarn be tInt- saline unmien ,tmtn,i cImt-s as innnmier tantts amid tirmotas In all menses tin, mm m,tnmI,ti inicb_tEmm is hi-imig strinsidi-,c-iI hm iii- rent ni timtpt,nmimjrnnm

added to the normal income tax) for families with less than $10,000 in 1984 and a 3 percent income tax sin-charge fom families with incomes cxceeding $60,000.
tax incomes

consumet losses associated with protection in car-hon steel $6.8 billion), automobiles $5.8 billion) and dairy products $5.5 billion). The purpose of pr-otectionism is to IJmotect jobs in
specific industries. A useful approach penspective on consumer- losses is to

Hufbauer et al. 11986) examined 31 cases in which trade volumes exceeded $100 million arid the United 4 States imposed protectionist trade restrictions. They generated estimates of the welfare consequences for each major- group afected see table 1). The figures in the table indicate that annual consumer losses exceed $100 million in all hut six of the cases. The largest losses, $27 billion pen- year-, come fnom protecting the textiles and apparel industry. There also ane lan-ge

to gain some express these losses on a per-job-saved basis. In 18 of the 31 cases, the cost per-job-saved is $100,000 or more pen-year-; the consumer losses per-job-saved in benzenoid chemnicais, carbon steel )two separate periodsL specialty steel, and bolts, nuts and screws exceeded $500,000
per year.

~Whilethere 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, protectionisi policies were maintained indefinitely or removed because of favorable demand changes.

Table I also neveals that domestic producers were the primary beneficiaries of protectionist policies; however, there are some noteworthy cases where fom~ eign producers realized relatively lam-ge gains. For the US-Japanese voluntary export agn-eenient in automobiles, foreign producers gained 38 percent of what domestic consumers lost, while a similar computation

FECfRAL RESERVE BANK OF sr. LOWS

JANUARY/FEBRUARY lOSS

Table 1 Distribution of Costs and Benefits from Special Protection

-

consumer Losses Per job saved (dollars)

5 ~00 000
fl? 1 :n-tnt:.rl

Producer Gains Totals (million dollars)

S KTh

eltare costs of Restraints Gamn to foreigners (million dollars)

n--c

case
Murutaciun mg St-OK r-nanaoclmnra Benzpr-ond clrernca~

Totals (million dollars)

S 500 2 650

Tariff revenue (million dollars)

S 0

Efficiency loss (mmIlion dollars)

5 29

2 2h0 1-30 90
25

ceo meg rreg --cg

~

252 54 139 69 55 29
10

14 13 33 6 11 150 4 lOU 3 100 4850

50 120 33cr -cc 30 lb 7 3 r) 8 200
1

Rr,Lbe- tr,otwoa.
sc-am--; am:nrjen

200 230
9u

2Lno 000 30 000

47.500 240 oco 7~.00Q 22 njOO 37000 42000 240.000 62000(3 750 000 90000 1 000 Xc 55 000 420 000 93.000 550 000 11/000 105 000 50 000 270 000 60 000 690-acre 220 000 800 ~ow 000 dcrP

erarm;clc-s
Orarnge u-ce Gannenlnna laxnlr-s ond apoae Phase; lexlnles ond appasl Phase ml Texmnies and Hpnarci Phase UI Gareo-~ steel- Pt-aspi carbon steel rb-,~e tn Carbon steel Phase lit Bail 004nnnus Snou only sleeNr-nwbner trMnwea Color leiev-snom-s ~B aoo; Bolts -Us arge svews Petared imnsnrcorns Autrynob-les MolomnyL e~ St- rv.cos Mdr tnme nrouslnes Agr.cul~re ard fnsbemnes Sugar Davy pmoduris

116
525 9~ 9.400 20 000 oT 000 1.970 4 30 6 $00 43 520 700 420 55 ~10 3S 5.800 04 3000 930

135000

oP 39O
74 6700 8000 22000 1 330 2 /70 3 800 21 5

neq
ico 350 1 800 330 940 2.000 neq

1 56 243
2.535 290 a56 560 18 32 77 32 /5 /90
1

cc
250 190 n4 50 2.1-00 67 2 njOO 550 5 000 170

so
220 140 nt-a nan reg 2200 meg req

10

000 130

410

-,

5 500 1/0 800 560

6 900 67

S0 nag 135

3-4

I 370

Pcanuis Meat Fnsh

M-nnrg Pelroneum Leacljr-Lzrnc

9
44

14

150000 725 t-pao 21-300

160 000 30000

1 600 200
4 800 41-

170
2 000 4

1//
/0 I:

5
3 000

Neo mneg: gnoie Unless oherwnse sr-pc -fpc. t.ommro: ire oar wor-0r EsI marco eumncs colk.clecn on sn-n) mepanns on-fnrmeo -abroac i-i Ions case becaLse of Ihe way tnt: ouctas were alnoratoni the ganrs to iporir ms arc- Led ~Urjonnestnc metnnos ratit-r tar tnmengr exporters SOURcE Irude Protpn Inorl nt-ne Umnn:ed blaIcs 31 Case Stuc-es

FEDERAL RESERVE BANK OF ST. LOWS

for the latest phase of protection for carbon steel was 29 percent. Finally, table 1 indicates that the efficiency losses ane small in compan-ison to the total losses borne by consumers. These efficiency losses, which are defined precisely and illusu-ated in the fir-st shaded insert, result fr-on) the excess domestic production and the reduction in consumption caused lw pr-otectnonist trade policies. En laige cases such as textiles and apparel, petroleum, dairy pm-oducts and the maritime industries, these losses equal or exceed $1 billion. It is likely that these estimates understate the actual costs because they do not capture the secondary effects that occur as pn-oduction and consumption changes in one industry affect other industries.~ In addition, restrictive trade policies generate additional costs because of hun-eaucn-atic enforcement costs and efforts by the pi-ivate sector to influence these policies fon- their own gain as well as simply comply with administrative r-egulations.

JANUARY/FEBRUARY IS
value of exports. One r-ationale for this finding is that a reduction in the pun-chases of foreign goods reduces foreign incomes and, in turn, causes reduced foreign purchases of domestic goods. While the reduction in impot-ts increases employmnent in industries that produce pn-oducts similar to the previously imported goods, the reduction in cxpor-ts decreases employment in the expor-t industries. In other words, while some jobs are saved, other-s ar-c lost; however-, this economic n-ealitv may not be obvious to businessmen, labon- union leaders, politicians and others. Luttrell 1978) has stressed that the jobs saved by piotectionist legislation are more readily obsened than the jobs lost due to protectionist legislation. In other words, the jobs that are protected in, say, the textiles industry by U.S. import restrictions on foreign textiles are more readily apparent land publicized) than the jobs in agriculture and high technology industries that do riot materialize because of the impoi-t restrictions. These employment effects will net to appi-oximately zei-oY the OECt) study also stresses that developing counti-ies need exports to offset their debts. Thus, pr-otectionist tn-ade policies by developed countries affect not only the economic activity of the developing coun tries, but the stability of the intet-natiorial financial system as debtor nations find it increasingly difficult to service their debts. Not only does a free trade policy by developed countries benefit developing countries, hut a free trade policy by developing countries benefits developing countries. A recent World Bank study (1987) of 41 developing countr-ies compar-ed the perion-mance of countries following a fr-ce tr-ade policy with countries following a restricted trade policy. Table 2 lists the

Costs o/ Protectionism Throughout the

World
In 1982, the Organization for Economic Co opet-ation and Development OECD) began a pn-oject to analyze the costs and benefits of protectionist policies in manufacturing in OECD countnes. The DEC13 1983) highlighted a numbei- of ways that protectionist policies have generated costs far in excess of benefits. Since protectionist policies increase pr-ices, the i-epon-t concludes that the attainment of sustained noninflationary growth is hindered by such price increasing effects. Moreovei-, economic growth is po tentiallv reduced if the uncertainty created by varying tnace policies depnesses investment. Wood and Mudd 1978), and many others, have shown that impoi-ts do not cause higher unemployment. Conversely, the OECD study stresses the fact that a r-eduction in imports via trade restr-ictions does not cause gn-eater employment. A reduction in the value of imports i-esults in a similar reduction in the

Recent estimates of the costs of protectionist policies using general equilibrium models suggest thai 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.

eRecent evidence shows that protectionist legislation actually may reduce employment. Denzau (1987) estimated that 35,600 manufacturing 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 (14000) and producers 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 strategy, 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.

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARY/FEBRUARY 1988

Table 2

Annual Average Growth of Per Capita Real Gross National Product

Free Trade Period 1963-73 Strongly S;ngapore Soutn Korea Hong Kong 9 0-~ 71 60 Moderately Braznl lsraeTha-and l-ndonesna costa Rnca Malays.a Ivory coast co-orbna Guaremaa ~amerooi 55% 5.4 4 9 4 6 3 9 3 8 3 5 33 27 01 Moderately Ynngoslav-a Mexnco Nngena Turrs:a Kenya Phnlnppnnes Bolnvna Honduas E Sa-vador Madagascar N-caraoua Senegal 49% 43 42 40 39 22 20 1 9 1 4 1 1 1 I 06 Restricted Trade Strongly Turkey Dornnr-cun Pc-publnc Burur-dn Arger-t na Pakrstar Tanzanna Sm Lanka Erhnoona Chnle Peru Uruguay 7amb-a Inc-na Ghana Bang:aoesh Sudan Banglacesh mona Buruno Domnnncan Repub.rc Efhnoona
Sudan Peru

a 5--. 34 3 2 3 3 2 7 23 1 9 1 7
i

5 1 2
l

1.1

0 4 1.9 20 20 1 2 05 04
04 1

197355

Snngapore Hong Kong South Korea

65 63 54

Malaysn~ Tnanlano Tunrsna

Braz-Turkey

4 1 SO
29 1 ~ 1 4 04 0.4 0.1

Cameroon lndorc-sna
Sn Lanka

56 40
33

Pa~nsnan Yugoslavia
cola-tb-a

31 27
1 8 1 3 1 1

Isree. Uruguay
chnne

Mexnco
Phnlnpp-nes

Kcnya Honduras Senegan costa Rca Guatemala Ivory Coast En Savaoo Nncaragua SOURCE World Doveroomenn Reporl 1987 and The Economist l1987l

03 0 I 0 5 1 0 1 0 1 2 35 39

Tan?anna Argenrnna Zamh,a

Nnqerna

6 20 23 25

Bonnvna Ghana Madagascar

31 32 3 4

;rnirLnai nu-nar4n gn-uniii ni fl ii pin r.npil;r gnnnss Hr Innun_nI pniniinirl trr-emr-k tnt line II rniinritnir, tim 19mM 7% nd 97:; M_5 I hrrsr rrrtnmitniis hat did nut bias indi_ns tii;tijinimdnmnlionilo~~rndrhrdrnnni-stnnmnnarknl mumim\ mnslri-lnrnnms r~nn~\i t_n,lri natr-~tIn~nnn lutist- rlnnl nun r finn i-\mnnnpli- lInt n.m-n-ar_re amninnn.nl ginn~~tIm n-atm iii n-n-rI pt-n r-.r nil I inn nun,- Inn- lUh%-7% ~as U.U penlenni inn 1 1 nm-rnnnr)nunin--, sin-nmnnj.l~ inn triter Inn rye Marie rrnrl in lInt- et-n,nnrrnmies stnnninr_I~ nmnni-mnierl mm n_n sIr jeir-mI p adi- I inn 1)7% 55 lines. r nfl tin n nit-s flint- i 9 iir run 0 I ni-er- nil ni-s nei-tn~n-Is 1 Inn- stnnnI~ )nn)n(mni-, Inn nrIrnitii~ Ii.- Iin.nlrr)nlmrmniimnnirnr,,mnni mm liii LrruinrnI tinnninmnr_r \ Un\mnn nrnnrntnmnt ni nnn~\

innustninenil ~einrrured mmmi- aiidiimrnn 1 1 rnnnrnti ni-s luIInfl\nnlg a liii- 1mar11 pn i thur .1 ri stnin-ni-d path- pnnIni-~ Ike n1.nsunn is that a nilinlniiinnn-mit nlinn~snr~mnrnl 11r\\ in, is mmI hn~lnl~ Inn \~rlineniInss ~~lniin a rvstnirlnl traIt- n-nn~innnninn,ennltins linris n-r-nmnmnnnmnjr inremniiw

It j)nmntnrIirnnnn~.nnnis,mnnnnstl~.~~in~ i. mnnrli-n-Iir,niisnnn sin 1 nin\nsnn~ I !ni~snrtnnnln n-n~nn~~~ rnn.nlnnn nnr4nnnmninni. tint

FEDERAL RESERVE BANK OF Si. LOUIS

for restricting tiade and provides explanations for the existence of protectionist trade policies.

JAIIUARYIPEBRUARY INS
such a large producer- or consumer of a good that a change in its production or- consumption patterns influences world prices. For- example, by imposing a tarifL the countny can make tbr-eign goods cheaper. Since a tariff reduces the demand for foreign goods, if the tariff-imposing country has some mar-ket power-, the world price for the good will fall. The tariffimposing countr will gain because the pr-ice per unit of its imnports will have decreased. There ar-c a number of obstacles that preclude the widespread application of this ar-gnment. Few couritries possess the necessary market power and, when they do, only a small number of goods is cover-ed. Secondly, in a won-Id of shifting supply and demand, calculating the optimum tan-iff and adjusting the rate to changing situations is difficult. Finally, the possibility of foreign retaliation to an act of economic warfare is likely. Such metaliation could leave both countries won-se off than they would have been in a fm-ce trade environment -

National he/ease
The national defense argument says that impor-t barrier-s are necessary to ensure the capacity to produce cr-ucial goods in a national emergency. While this ar-gument is especially appealing for weapons dun-ing a war, then-c will likely be demands from other industn-ies that deem themselves essential. For- example, the footwear industry will demand protection because militam personnel need combat boots. The national defense ar-gument ignores the possibility of purchases from friendly countries dur-ing the emer-gency. The possibilities of storage and depletion r-aise additional doubts about the general applicability of the aigument. If crucial goods can be stored, for example, the least costly way to prepare for an ernergency might be to buy the goods from foreigners at the low world pr-ice befone an emer-gency and store them. If the crucial goods ar-c depletable miner-al resour-ces, such as oil, then the restriction of oil imports befon-e an erner-gency will cause a more rapid depletion of domestic r-eserves. Once again, stockpiling might be a far less costly alternative.

Balancing the Balance o/Trade

Many conntnies enact prDtectionist tnade policies in the hope of eliminating a balance of trade deficit on increasing a balance of tr-ade surplus. The desire to increase a balance of ti-ade surplus follows from the mer-cantilist view that larger trade surpluses are beneficial from a national perspective. This argument is suspect on a numben of grounds. First, there is nothing inherently undesir-able about a tr-ade deficit or- desirable about a surplus For example, faster economic growth in the United States than in the rest of the won-Id would tend to cause a trade deficit. In this case, the trade deficit is a sign of a healthy economy. Second, protectionist policies that r-educe irnports will cause exports to decrease by a comparable amount. Hence, an attempt to increase exports permanently relative to imports will fail. tt is doubtful that the trade deficit will be reduced even tempor-ar-ily because imnpont quantities do not decline quickly in r-esponse to the higher- impor-t pr-ices and the revenues of foreign producer-s might r-ise.

Income Redistribution
Since pr-otectionist tr-ade policies affect the distribution of income, a trade restriction might be defended on the grounds that it fayors some disadvantaged group. It is unlikely, however, that trade policy is the best tool for dealing with the perceived evils of income inequality, because of its bluntness and adverse effects on the efficient allocation of resources. Attempting to equalize incomnes directly by tax and transfer paymerits is likely less costly than rising trade policy. In addition, as Hickoks 19851 study indicates, trade restrictions on many items increase r-ather than decrease income inequality-

Optirn urn Tariff.zirgurnent

The optimum tar-iff argument applies to situations in which a country has the economic po\\el to alter won-Id prices. This power exists because the countr or a group of countries acting in consor-t like the Organization of Petroleum Exporting Countr-iesl is

See Pine (1984).

If a country such as the United States has no market power, the world price is fixed. Consequently, the price faced by U.S. consumers and producers rises by the full amount of the tariff. In the optimum tariff case, the price faced by U.S. consumers and producers 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. See Chrystal and Wood (1988) earlier in this issue.

FEDERAL RESERVE BANK OF Si. LOUIS

JANUARYIFEBRUARY INS
Even though the aggregate effect is lange, the harm to each consumen- may be small. this small cost, ofwhich an individual may not even be awar-e, and the costs of organizing consumers deter the formationi of a lobby against the legislation. On the other hand, worker-s arid other- nesoun-ce owner-s are vex-v concer-ned about protectionist legisla tiori for their industry. liieir benefits tend to be large individually and easy to identity. Their voting and campaign contributions assist politicians who support thein positions and penalize those who do riot. Thus, politiciarns are likely to respond to their demnands for- pr-otectionist Iegislationiz

Protection of Jobs

Public Choice

The protection of jobs ar-gumenit is closely related to the balance of trade argument. Sinceareduction in imports via trade n-estr-ictions will result in a similar reduction in exports, the over-all employment effects, as found in the OECD 1985) study and many others, are negligible. While the overall effects are negligible. won-kens land r-esour-ce owners) in specific industries ame affected differently. A domestic industry faced with increased impomts fr-ore its foreign competition is under pressure to n-educe production and lower- costs. Pn-oductive n-csources mnust mnove fi-om this industry to other domestic industmies. Wor-ker-s must change jobs and, in some cases, relocate to other cities. Since this change is forced upon these workers, these won-ken-s hear m-eal costs that they are likely to resist. A similar statement can be made about the owner-s of capital in the affected industry. Workers and other mesoinr-ce owner-s ~villllikely resist these changes by lobbying for trade r-estr-ictions. The pr-eviously cited studies on the costs of protectionism demonstt-ated that tr-ade restrictions entail substantial neal costs as well. lhese costs likely exceed the adjustmnent costs because the adjustment costs ar-c one-time costs, while the costs of protectionism continue as long as trade restrictions are maintained. An obvious question is why politicians supply the protectionist legislation demanded by workers arid otlier resource owner-s. A bm-anch of economics called public choice, which focuses on the interplay between individual pr-eferences and political outcomes, provides an answer. The public choice liter-atur-e views the politician as an individual who offer-s voters a bundle of governmentally supplied goods in order to win elections. Many argue that politicians gain by pr-ovid ing protectionist legislation. Even though the national economic costs exceed the benefits, the politician faces different costs and benefits. lhose hat-med by a protectionist trade policy for a domestic industry, especially household consumen-s, will incur a small individual cost that is difficult to identify. For example, a consumer is unlikely to ponder how much extra a shirt costs because of protectionist legislation for the textiles and apparel industry.

In/Cat .Industries
Ehe preceding ar-gunient is couched in terms of protecting a domestic industry. A slightly different argument, the so-called infant industry case, is couched in terms of promoting a domestic industry. Suppose an industry, already established in other countm-ies, is being established in a specific coinntrv. The country might riot he able to realize its compar-a tive advantage in this industry because of the existing cost and other advantages of foreign fir-nis. Initially, owner-s of the fledgling firm must be willing to suffer losses until the firm develops its mar-ket and lower-s its production costs to the level of its fon-eign rivals. In on-den- to assist this entr-ant, tan-ill pn-otection can be used to shield the fir-ni fnom some for-eign competition. After this temnporan~period of protection, free trade should be n-estor-ed; howeven-, the removal of tariff protection frequently is resisted. As the industry develops, its political power- to thwan-t opposing legislation also increases. Another- pr-oblenn with the infant industry arguniierit is that a tariff is not the best way to intervene. A production subsidy is super-ion- to a tariff if the goal is to expand pn-oduction. A suhsidv will do ibis directly, while a tariff has the undesin-able side effect of r-educ inig consumptioti. In many cases, intervention might not be appr-opri ate at all. If the infant industny is a good candidate for being competitive inter-nationudly, borr-ouing from the

The role of pressure groups, acting in their economic self-interest, has been stressed by Stigler (1g71) 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).

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

FEDERAL RESERVE BAtIK OF ST. I.OUiS

pr-ivate capital markets can finance the expansion. Investors are willing to ahsor-b losses temporarily if the prospects for- future profits ar-c sufficiently good.

JANUARYIFEBRUART INS
petition to favor- domestic over foreign firms and shift the excess returns in monopolistic mar-kets from foreign to domestic fir-ms. Knugman 1987) illustrates an example of the angurnent. Assume that there is only one firm in the United States, Boeing, and one multinational Iir-m in Europe Airbus, capable of producing a 150-seat passenger aircraft. Assume also that the aim-cr-aft is pn-oduced only for export, so that the returns to the firm can be identified with the national interest. This export market is pn-ofitable for either fin-nm if it is the only producer; however, it is unprofitable for both fir-ms to pnodtice the plane. Finally, assume the following payoffs are associated with the four combinations of production: 1) if both Boeing and Airbus pr-oduce the aircr-aft, each firm loses $5 million; 2) if neither Boeing non Ainbus pnoduces the aircn-aft, profits are zero; 3) if Boeing pn-oduces the aircraft and Air-bus does not, Boeing profits by $100 million and Air-bus has zero pr-oUts; and 4) if Air-bus produces the air-cr-aft and Boeing does not, Airbuns profits by $100 million and Boeing has zero profits Which finnn)s) will pr-oduce the air-craft? The example does not yield a unique outcome. A unique outcome can be genen-ated if one firm, say Boeing, has a head stant and begins production before Airbus. tn this case, Boeing will n-cap pn-ofits of $100 million and will have deter-n-ed Airbus from emitening the niarket because Air-bus will lose $5 million if it enters after Boeing. Stnategic trade policy, however-, suggests that judicious gover-nmental intervention can alter the outcome. If the Eunopean governments agree to subsidize Air-bus production with $10 million no mnatter what Boeing does, then Airbus will produce the plane. Production by Air-bus will yield more profits than not producing, no matter- what Boeing does. At the same time, Boeing will be cleterred fiom pnoducing because it would lose money. Thus, Airhus will captur-e the emitime rnar-ket and reap profits of $110 mnillion, $100 mnillion of which can be viewed as a tnansfer of pr-ofits from the United States. The criticismns of a strategic trade policy ar-c similar to the criticisms agaimist pr-otectimig a technologically pr-ogn-essive industry that gener-ates spilloverbenefils. Then-c are major- infonmational pn-obleniis in applying a

Spillover Effects
The justification for pn-otecting an industry, infant or otherwise, frequently entails a suggestion that the industry generates spilloven benefits for other in dustrres ot individuals fon which the industry is not compensated. Despite patent laws, one common suggestion is that cer-tain industn-ies are not fully compensated for their- r-esear-ch arid development expenditur-es. This argument is frequently clirected toward technologically progressive industries where some firms can capture the nesults of other- finns nesearch and development simply by dismantling a pr-oduct to see how it works. The application of this ar-gumnent, however-, enigetiden-s a niurnbem of pn-oblems. Spillovens of knowledge ane diffic,ult to measure. Since spilloven-s am-e not market tr-ansactions, they do not leave an obvious tr-ail to idemitifv their- heneficiar-ies. The lack of niar-ket tr-ansactions also complicates an assessment of the value of these spillovers. To determine the appr-optiate subsidy, one must be able to place a dollar value on the spillovers genen-ated by a given reseamch and developnient expenditure. Actually, the calculation requires munch mom-c than the alneady difficult task of reconstn-ucting the past. It r-equir-es complex estimates of the spillovens future worth as well. Since resources are moved fioni other industries to the targeted industry, the goven-nment must tinder-stand the functioning of the entire economy. Finally, then-c are political pr-ohlems. An aggressive application of this an-gumuent might lead to retaliation and a mutually destructive tn-ade war. In addition, as intenest groups compete for the gover-nmental assistance, there is no guar-antee that the tight gr-oups will be assisted or that they will use the assistance efficientl.

Strategic Trade Policy

Recent theor-etical developments have identified cases in which so-called stn-ategic tr-ade policy is superior- to fl-ce trade. As we discussed earlier, decreasing unit production costs and market structures that contain monopoly elements ar-e conirnon hi industr-ies involved in inter-miational tmade. Mar-ket imperfections immediately suggest the potential benefits of govern mental intervention. In the str-ategic trade policy argument, gover-nmnent policy can alter the terms of com-

3 A recent volume edited by Paul Krugman (1986) examines the olicy implications of the new trade literature, See Grossmans article in that volume for a discussion of the information requirements.

FEDERAL RESERVE BANK OF ST. LOUIS strategic trade policy. The government must estimate the potential payoff of each course of action. Economic knowledge about the behavior of industnies that have monopoly elements is limited. Firms mnay behave competitively on- cooper-ativel~and may compete by setting pm-ices on- outpr.rt. The behavior- of rival govem-nments also must he amiticipated. loreign retaliatiomi must be viewed as likely where substantial profits are at stake. In addition, many intem-est gr-oups will compete for the govem-nmental assistance. Though only a small number of sector-s can be conisidered poterrtiall~stmategic, many industries will make a case for assistance.

JANUARY/FEBRUARY 1988 ducers are relatively disadvantaged, the wisdom of a pn-otectionist response is doubtful. Again, the costs of pr-otectionismn exceed substantially the benefits froni a miatiorial perspective. In an attempt to reinfon-ce the ar-gument for fair tnade, pr-oponents also an-gue that n-etaliatory thn-eats, combined with changes in tariffs and miontamiff barriers, allow for the simultaneous pr-otectiomi of domestic industries against unequal comnpetition and indunce mon-c open fom-eign mnar-kets. This mom-c flexible appr-oach is viewed as super-ior to a onesided free trade policy. lhe suggestion that a fair tnade policy produces a tnading environment with fewer- trade nestrictions allows pmoponents to asser-t that such a policy serves to pn-omote both equity and efficiency. In other- words, not only will domestic and foreign pnoducem-s in the same industry be treated equally. but the gains associated with a fneer tr-ading envir-onment will he n-ealized. On the other- hand, critics of a fair trade policy argue that such a policy is simply disguised protectionisni it simply achieves the goals of specific intemest gn-otnps at the expense of the nation at lam-ge. In mnany cases, fair tr-ader-s focus ori a specific pr-actice that can be portr-ayed as pn-otectionist while ignon-imig the entine package of policies that are affecting a nations comnpetitive position. In these cases, the fon-eign coumitmv is muon-c likely either not to respond or retaliate by incr-easing r-athcr- than n-educing their tracle barriens. In the latter- case, the escalation of trade han-mien-s causes losses for- both nations, which is exactly opposite to the alleged efiects of an activist fair trade policy. Critics of fain trade proposals ar-c especially bother-ed by the use of bilateral trade deficits as evidence of unfair trade. In a world of many tn-ading countries, tle trade between two countnies need not be balanced for the tr-ade of each to be in global balance. Diffemimig demands and pr-oductive capabilities across countries will cause a specific coumitry to have trade deficits with some countries and sur-pluses with other- counitmies. These bilateral imbalances an-c a rionmual result of coumitmies tradimig on the basis of comnpar-ative advan4 tage. Thus, the focuns on the hilater-al tnade deficit can produce inappr-opriate conclusions about fain-ness and, niiore impor-tantly, policies attempting to eliminate bilatenal trade deficits ar-c likely to be very costly becaunse they eliminate the gains from a multilateral tn-ading system.

Reciproci~vand the Level Playing Field

Bhagwati and Irwin 1987) note that U.S. tnade policy discussions in recent year-s have frequently stressed the importance of fair tr-ade. The concept of fair trade, which is technically n-efen-red to as reciprocity, means diffen-ent things to different people. Under the Genen-al Agreement on Iariffs and made, negotiations to reduce trade ham-mien-s focus upon matching concessions. This form of reciprocity, known as first-difference reciprocity, attempts to n-educe trade barrier-s by r-equiring a country to provide a tariff rtduction of value compan-able to one pr-ovided by the other country. In this case, neciprocity is defined in terms of matching changes. Recent U.S. demands, exemplified by the Gephardt amendment to the cur-rent tn-ade legislation, r-eyeal an approach that is called fbll reciprocity. This approach seeks r-eciprocitv in terms of the level of pr-otection bilaterally and ovem a specific n-ange of goods. Becipn-ocitv n-equines equal access and this access can he deter-mined by bilatemal tnade balances. A tnade deficit with a trading partner is claimed to be prima fade evidence of unequal access. Examples ahounnd. For example, U.S. construction fin-mns have not had a major contract in Japan since 1965, while Japanese constm-uction fim-ms did $1.8 billion wonth ofbirsiness in the United States in 1985 alone. Recent legislation liar-s Japanese panticipation in U.S. public won-ks pnojects until the Japanese offer r-ecipn-ocal pr-ivmleges. As the name suggests, the fundamental argument for fain- trade is one of equity. Domestic producens irm a fl-ce tnade country argue that foneign trade harrier-s am-c unfair because it places them at a competitive disadvantage. In an extreme yer-sion, it is asserted that this unfain competition will vir-tually eliminate [1.5. mnanu factoring, leaving only jobs that consist primarily of flipping hamburgers at fast food restaun-ants on-, as Bhag%%-ati and Inwin have said, rolling mice cakes at Japanese-owned sushi bar-s. While domestic pm-o-

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

FEDERAL RESERVE BANK Of ST. LOUIS

JANUARY/FEBRUARY INS
Interest gr-oup pressunes from industries expeniencing difficulty and the gener-al appeal of a level playing field combine to make the reduction of trade hiar-riers especially difficult at the present timmic in the United States. Nonetheless, national inten-ests will be served best by such an admittedly difficult political comm-se. Imi light of the cur-rent tJmuguay Round negotiations under- the General Agn-eement on Taniffs and Tm-ade, as well as numerous bilateral discussions, this fact is especially timnely.

CONCLUSION
The pr-olifer-ation of pnotectm()nist tr-ade policies in recemit year-s pm-ovides an impetus to m-econsidem their wor-th. In the won-id of tr-aditionial trade theory, chan-ac terized by perfect competition, a definitive recoin mendatiori in favor- of fl-ce trade can he made. The gains fnonii initen-nationial trade result fm-om a m-ealloca tioni of pn-oducti~e m-esour-ces towam-d goods that cami he lino(lumced less costly at home than abroad amid the exchange of some of these goods for goods that can he pm-odunced at less cost aljnoad than at home. Recent developments in inter-national tm-ade theory have examitied the consequences of international tr-ade in mar-kets when-c tliene ar-c nianket imperfections, such as momiopoly and technological spillo\er-s. Do these imperfections justitv protectionist trade policies? The answer- continues to he no. While protectionist trade policies may offset monopoly power- over- seas or advantageously use domestic moniopol~ power, trade restrictions tend to reduce the competition faced by domestic producers, protecting domestic producer-s at the expense of domestic conisunIens. chic empirical evidence is clear-cut. The costs of pr-otectionist tr-ade policies far exceed the benefits. the losses suffer-ed by consumer-s exceeml the gains reaped by domestic pr-oducens and government. Lowinconiie consumer-s ar-c nelatively more adversely affected than high-income consumers. Not only are then-c inefficiencies associated with excessive domestic production and restricted consumption, but then-c ane costs associated with the enfor-cemnent of the pm-o tectionist legislationm and attempts to influence trade policy. The primam~reason for these costly protectionist policies relies on a public choice an-gumemit. The desire to influence trade policy arises froni the fact that trade policy changes benefit some gr-oups, while har-ming others. Consumer-s are ham-med by pr-otectionist legislation; however-, ignor-ance, small individual costs, and the high costs of or-ganizing consumers pr-event the consumers from beirig an eflective for-ce. On the other hand, wor-ken-s and other resour-ce owner-s in an industn-v ar-c mor-e likely to be effective politically because of their relative ease of om-ganizing and their- individually large and easy-to-identify benefits. Politicians inter-ested in reelection will most likely respond to the demands for- pmotectionist legislation of such an interest group. Ihe enmpimical evidence also suggests that the ad~er-seconsumer effects of pr-otectionist trade policies ame riot shor-t-lived. These policies genem-ate lower economic gn-owth n-ales than the rates associated with fr-ce trade policies. In turn, slow gr-owth contributes to additional pr-otectionist pressum-es.

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. 10930. Brander, James A. Intra-Industry Trade in Identical Commodities, Journal of International Economics (February 1981), pp. 114. Brander, James A.. and Paul R. Krugman. A Reciprocal Dumping Model of International Trade, Journal of International Economics (November 1983), pp. 31321. Chrystal, K. Alec, and Geoffrey E. Wood. Are Trade Deficits a Problem? this Review (January/February 1988), pp. 5i 3. Clarete, Ramon L., and John Whalley. interactions Between Trade Policies and Domestic Distortions, Center for the Study of International Economic Relations Working Paper 8522C (London, Ontario: University of Western Ontario, 1985). Coughlin, Cletus C. Domestic Content Legislation: House Voting and the Economic Theory of Regulation, Economic Inquiry (July 1985), pp. 43748. Denzau, Arthur T. How Import Restraints Reduce Employment, Washington University Centerfor 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 Economics, 1987). Dixit, Avinash K., and Victor D. Norman. Theory of International Trade (Cambridge University Press, 1980). Dixit. Avinash K., and Joseph E. Stiglit2. Monopolistic Competition and Optimum Product Diversity, American Economic Review (June 1977), pp. 297308. 74 The Economist. July 4,1987, p. . Grais, Wafik, Jaime de Melo, and Shujiro Urata. A General Equilibrium Estimation of the Effects of Reductions in Tariffs and Quantitative 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, Strategic Trade Policy and the New International Economics in Paul R. Krugman, ed. (MIT Press, 1986), pp. 4768. Helpman, Elhanan. International Trade in the Presence of Product Differentiation, Economies of Scale and Monopolistic Competition, Journal of International Economics (August 1981), pp. 305 40. Hickok, Susan. The Consumer Cost of U.S. Trade Restraints, Federal Reserve Bank of New York Quarterly Review (Summer 1985), pp. 112.

FEDERAL RESERVE BANK OF ST. LOWS Hufbauer, Gary Clyde, Diane T. Berliner, and Kimberly Ann Ellioti, 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. 2837. Kierzkowski, Henry K. Recent Advances in International Trade Theory: A Selective Survey, Oxford Review of Economic Policy (Spring 1987), pp. ii9. Krugman, Paul R. Is Free Trade Pass? Journal of Economic Perspectives (Fall 1987), pp. 131-44. Strategic Trade Policy and the New International Economics (MIT Press, 1986). Lancaster, Kelvin. Intra-Industry Trade Under Perfect Monopolistic Competition, Journal of International Economics (May 1980), pp. 15175. Variety, Equity, and Efficiency (Columbia University Press, 1979). Leibenstein, Harvey. Beyond Economic Man (Harvard University Press, 1980). Luttrell, Clifton B. Imports and Jobs The Observed and the Unobserved, this Review (June 1978), pp. 210. 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

JANUARY/FEBRUARY 1988 the Study of American Business, Working Paper #80 (July 1983). Organization for Economic Co-Operation and Development (OECD). Costs and Benefits ofProtection (1985). Page, Sheila. The Rise in Protection Since 1974, Oxford Review of Economic Policy (Spring 1987), pp. 37Si. Peltzman, Sam, Toward a More General Theory of Regulation, Journal ot 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 Journal 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. 58 73. Tarr, David G., and Morris E. Morkre. Aggregate Costs to the United States of Tariffs and Quotas on Imports: General Tariff Cuts and Removal of Quotas on Automobiles, Steel, Sugar, and Textiles, Bureau of Economics Staff Report to the Federal Trade Commission (December 1984). Venables, Anthony, and Alasdair Smith, Trade and Industrial Policy Under Imperfect Competition, Economic Policy (October 1986), pp.62172. Wood, Geoffrey E., and Douglas R. Mudd. The Recent U.S. Trade DeficitNo Cause for Panic, this Review (April 1978), pp. 27. World Bank. World Development Report 1987 (Oxford University Press, 1987).

FEDERAl. RESERVE BANK OF ST. LOUIS

JANUARYIFEBRUARY INS

Appendix Developments in International Trade Theory and the Gains from Trade
Since 1817, numerous developments have taken place in inter-national trade theory. The consequences of mone than one factor of pn-oduction, increasing amid decn-easing unit pn-oduction costs, amid imper-fectly competitive markets are examined mi this appemmdix. Special attention is focused on developments in intennatiomial tm-ade theory in the last decade. While resources may not be easily transfer-n-ed acr-oss industries imi the shont n-un, worker-s can change jobs and capital cart be moved as time passes. hf resources are mobile, then the longer-run consequences for- labom- and owmiens of capital an-c different from those described above. Even if lahor and capital are perfectly mnobihe, however, one set of mesource owner-s may benefit while another- gn-oup is har-mned by trade. The real world is more complicated than this discussion has allowed. Them-c are mon-c thami two factor-s of pr-oduction amid vamyimig degrees of mobility for these factor-s. For example, the U.S. labor- for-ce con taimis scientists and engirmeers as well as shorton-demcooks. Nonetheless, the underlying analysis does suggest somne generalizations. When trade occur-s, owners of the resoun-ces that ar-c mor-e specialized imi the pn-oduction of export goods will tend to become weal

Increasing the Number of Factors of Production

Assume that, in the United States, two resoun-ces, labor- arid capital (e.g., machinesi, ar-c used in the pr-oduction of two goods, automobiles amid air-planes. The pnices of these resources will be affected differ- ently by tr-ade. As trade develops, demnanid for- the exported good that is, the good in which the United States has a comparative advantage) will incnease amid demand for U.S. pr-oductiomi of the imported good will fall. hhis demand shift causes the price of the exported good to rise nelative to the pr-ice of the ire ported good. Similar-h, the shiff may also pn-oduce changes imm the pr-ices of r-esoun-ces: however-, these pm-ice changes an-c not always obvious. Initially, assumne that the resoum-ces cannot be tm-ans ferred across industnies. For examnple, the labor- and capital used to pr-oduce automobiles, the good im ported into the United States, cannot he used to produce aimplamies, the expon-ted good. Consequently, as the pr-ice of airphanes n-ises imi the United States, the compensation for- lahor and capital in the air-plane industry will rise; meanwhile, the decline in automobile pr-ices causes a decline in compensation for- labor and capital in the industry. It would riot be surpn-ising if labor and owners of capital in the imidustrv would mesist such changes by asking for- tn-ade pm-otection.

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 othercountries, 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 automobiles 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.

FEDERAL RESERVE BANK OF ST. LOUIS thier; those who own n-esoun-ces mon-c specialized in the production of import-competing goods will tend to lose wealth. People also gain or- lose, howeven, depending on what happens to the pr-ices of the goods they buy. Individuals who chiefly consume imported goods will benefit, while those who pr-efer- consuming the exported goods will lose. Thus, the net effect on any individual depends on both the gains or losses associated with the price changes on the goods that he consumes and the effect of trade on his wealth (orincome I.

JANUARY/FEBRUARY 1988 pontation costs. Each firm openates under what is termed a Cournot comijectun-e, meaning that each firm assumnes its pmoduction decision ivill not affect its rivals pmoduction decision. Before inter-national tnade, each firm has a monopoly position in its home man-ket. Aliowing for- fr-ce trade induces each fin-rn to emiter the othen br-ms niarket, because price exceeds mnamginal cost in each coumitry. Thus, the same good will flow to and fr-om each country. Kierzkowski l1987n has noted that the bulk of intn-aindustry trade involves differentiated rather than homogeneous goods. Two approaches, Lancasters (1979) chanacteristics approach and Dixit and Stighitzs (1977) hove of variety appr-oach, have pr-ovided the foundation for- trade models involving differentiated goods. In the chanactenistics approach, individuals have pn-eferences for- the characteristics of goods m-ather than for- collections of time goods themselves, A group of goods is defined as goods possessing the sanre characteristics but in diffem-ent pn-oportions. A diversity in consumner pr-eferences causes different consumner-s to pr-efen different products (i.e., vanieties) of a gr-oup of goods. Fleipman (1981) and Lancaster)lSSD) used the char-actem-istics approach to show how intra-industry tnade results fnom combining the demand for- variety with economies of scale, The change fr-om autan-chy to free trade enlarges the mar-ket and causes output of the existing varieties to increase and the production of new varieties to begin. Consumers gain from the productiomm of more varieties and lowem pr-ices as economies of scale are realized, The sources of gains from tr-ade are identical using the love-of-variety and character-istics appr-oaches. In the love-of-variety approach, which is used by Dixit and Norman )1980r, consumers have identical tastes and prefen to consume as many types of the differentiated product as possible. The introduction of imperfect competition and declining unit pm-oduction costs suggest three sources of gain from free trade. As the market potentially served by firms expands fi-om a national to a world market, there will be gains due to declining unit production costs. The second is the reduction in nmonopol~power of firnms faced with foreign competitors. The thin-d is the gain to consumer-s fm-onn lower prices and increased product variety. Gener-ally speaking. gains from trade result from the increase in competitive pr-essures as the domestic econonmy becomes less insulated from the world economy. Nonetheless, time numerous market structur-es and firm hehavions possible under imperfect competition preclude a definitive statement about the optimahity of fr-ce trade.

Increasing Unit Production Costs

A second assumption underlying the Ricandian example of the gains fr-ore trade is that unit production costs are constant. If unit production costs rise as more is produced, however, the general conclusions about the gains fr-nm tr-ade remain essentially unchanged. The major- difference is that rising unit production costs limit the extemit to which speciahizatiomi occurs.

Decreasing Unit Production Costs and Imperfect Competition

On the other hand, if unit production costs decrease as production increases, the extent to which actual trade patterns can be explained by comparative advantage becomes unclear. lt also fonces tr-ade theory to deal with nunmerous char-acteristics of international trade in the real wor-hd. The market structure ofindustries engaged in tnade is fr-equently highly concentnated. In other words, the individual firms in an industry, contrany to those in a perfectly competitive industry, can affect the manket price of their good by their productiomi and advertising decisions. In addition, trade statistics show that intra-industny trade (i.e., the simultaneous export and import of the output of the same industnz accounts for- incn-easinglv larger shames of won-Id trade. In Ihe last decade, trade theorists have developed numnen-ous models to deal with these facts. An exhaustive r-eview of this rapidly expanding hitenature is beyond the scope of this appendix; howeven, a few illustrative articles are discussed in order to establish some key points. Brander (1981) and Br-ander and Knugman (1983) developed models using a homogeneous good to highlight how imperfect competition can cause intra-industry trade and how intna-industry tn-ade can arise in the absence of cost differences. Assume two countries with one firm in each country. The firms an-c pmoducing a homogeneous good under identical cost situations and then-c am-c no trans-

FEDERAL RESERVE BANK OF ST. LOUIS

JANUARYIFEBRUARY ISa
The demand curve is D,, and the marginal cost curve is MC,,~. The mnarginah revenue curve associated with lJ,,~ ( is omitted.) The monopoly pr-ice and output are P,~ and Q~ The change from autarchy to free tnade transforms the national monopolies into a world duopoly. Assuninng the firnis follow a Cournot str-ate~, price declines from P~ to P,~, sales in the United States incnease from Q,,, to Ii,,,, and consumers gain an-ca P~ ES P,, or h + i. Profits ignoring fixed costs) in the United States decline fr-om area P,,, RWM or i + j + k to area P, SXM or j + k + 1. The domestic firm has onehalf of the domestic market, so its profits an-c j with k + I going to the foreign firm. The domestic firms expon-ts allow it to captune one-half of the foreign mar-ket. If the foreign mar-ket is identical to the domestic market, the firms profits on ibr-eign sales will equal k + I. Therefore, the net reduction in the domestic fin-rns profits is I and the overall welfare gain to the economy is h + I. If the assumption of identically sized domestic and foreign mar-kets is dropped, then a different conclusion is possible. If the foreign market is smaller- than the domestic, the profits of the domestic firm in the foreign market will be less than k + 1. Assuming zero exports, the domestic gains fr-nm trade are h k, and the domestic economy could lose from fnee trade. In this case, consumer- gains can be mon-c than offset by the shifting of profits fi-onm the domestic to the fbreign economy. lhis shifting reflects the contraction of an activity that is already too little to an even smaller- level.

Figure 3

Ttj~

TI

P4

w~x
D~5

ICUS

D~512~

Sometimes the benefits of expanded consumption resulting from fl-ce trade an-c less than the costs associated with distorted pr-oduction. Venables and Smith (1986) pn-ovide a graphical illustn-ation, duplicated in figure 3, of the preceding point using the Br-anderKr-ugman duopoly model. Assume the U.S. market and the market in the rest of the won-Id for- a specific good are monopolies, the good is produced at a constant marginal cost, and there are no tr-ansportation costs.