Free

82 lnternational Economics - Vol. 1

Part One TAc Tlreorr of IntenvttintLlTi,ult

sector {the one using the country's scarce facior intensively;. The changes in prodr:c- tion have one set of effects on incomes in the short run. but another in the long run.

Short-Run Effects of Opening Trade In the *$:*rt't ru* laborers. plots of land. and other inputs are tied to their curreirt lines of prodiiction. The denand for these t-actors, and therefore the incornes or returns they earn, depend on the sector in which they are empioyed. Sorre peopie rvill enjoy higirer demand tbr the factors thev have to offer, because their iactors are employed in the sector that is atteinpting to expand its production. With the open- nig to tree trade, the expanding industries in oLrr exarnple are u,heat in the tJnited States anil cloth in the re st of the world. in the Uniled States landlords in rvheat- grorving ateas can clrarge higher rents because thelr land is in strong derna*d. U.S.

iurn1**k.rs in wheat-glowrng areas are likelv to get (temporarily) irigher wages. Foreign clothrvorkers cin also dernand and get higher wage rates. Foreign landlords

in thJ areas raising cotton atld u,ool ibr cloth-making can also get higher re*ts' rlrLd'rntrii!. rlru rwllrrc vl fcrvrvr., t- th,c clcclir,ir,6 iutlu;tri'+t g.g. a!'31.11'.t'ot!.et'S'

U.S. landlords in areas supplS'ing the cloth-nraking industry, fbreign rvheat-area laitd- lords and farnrhands- lose income through reduced clemand and therelbre reduced prices fbr fheir services.

Fol the shorl run" tiierl. gains and losses divide by output secfor: All groups tied to rising sectors gain, and all groups tied to declining sectoi's lose. One lvould expect empiovers. landlolds. and workers in the declining sectors to uttite in protest.

The Long-Run Factor-Price Response ln the i<,r*q r*-*n factors can move between sectors in respr'rnse to ditferences in returns. Sellers of the same factor r.viil eventually respond to the inconle gaps that have

been openecl up in the short run, Sorne U.S. clothrl,orkers will fincl better-payingiobs iu the wlieat sector. r\s the supply of labol into the nheat sector increases. lvages ilt the rvheat sectol'decline. As the remaining supply af labor to the clolh sector shritrks, wages in the cioth sector increase. Sorne U.S. cotion- and wool-raisirig land will also get better rents by converting to lvheat-related producfion, bringing rents in rjilfbrent

areas back in 1ine. Similarlv, foreign farmhands and landlords r,vill flnd the pay better in the cloth-relate<l sector. As the sr:pplies of the f'actors to the two sectors change in the iong run. wages and rents iri the cioth sector decrease, and wages aud rcnts in tlte

lr,heat sector increase. The full process of the effects olopening trade on l'aclor prices in the long run is summarized in Figure 5.1.

when the facto|s respond by rnoving to the better-paf ing scctors' will all n'ages arid rents be bicl back to their pretrade levels'l No. tliey will not. In the long run. wage rates encl r:p lorver tbr att U.S. workers and higher for all frrreign workers (each relative to its leve1 with no trade). A1l land lents end up liigher in the United Stales and lower in

-fhe.re-strf',the ruorid.{each relative to its level r'vith tto trade)' What drives thrs crucial resuit is the imbalance in the changes in factor supply and

demand. Wlieat grolving is more land-intensive and less labor-intensive than cloth making. Therefbre, the amounts o1'each factor beitrg hired in the erpandillg sector

tvi1l not match the amounts being released in the other sector. The imbaiances create pressures f'or f-actol prices to adjust.

ii:i.:,ii;ii:--', 1:

l1or.,, Ft'ee l-rade

Ailects Ittcotr:e Distribution rn the L-ong Rr:n: The l1'hole Chain of inl-luence

lnternational Economics, 1 5th Edition 83

Chapter 5 \Iilrr.r (l,rrtts rtncl \Illo Lo.re.r f-on -Ir.r.Jt''

iAt this point^ fhe shor.t-run effects come into play. bi-rt the ecr:noll)'colltinues to ll1ove torvard the longerlr"tn etfects shorvn in the rest ofthis fisure.

i1 the ljniteci States. for example, exp{tnding wheat production creates demand for

a lot of land and very f-ew workers, whereas cutting cloth productiott releases a lot of workefs and not so fiuch lancl.1 SrNnethirrg has to give.'fhe onl'; \\/ay the etnploynlent of labor nild land can acljusi to the available national supplies is for tactor prices to

change. The prodgction shill toward land-intensive. labor-sparing \l4leat raises rents and crrrs \L/ages thrtntglrtni the Liniteci States in the long run. The rise in rents alld lhe fa11 in ruagei Lroth continr-te until producers cor:le up rvitll ttrore land-saving and labor- gsiltg ways of rnaking wheat aud cioth. Once they do, rents aurl rvages stabilize-{:rrt Li.S. renls stili eld up higher ancl ra,ages lower than betbre trade opened up.

-[he sane

kind olreasoning makes ttte opposite results hold for the iest of tlie rvorld. 'L"ade. then. makes some people absoluteiy better oll and others absolutely worse

ofl in each ol tire trading countries. The gainers ancl losers in fhe sholt ruli are

This passace uses conr.renieni shorthand that is qrrantiiatiirely riague: "'r lot cf " lartcl' "i'rety fet"'" vvotkets'

and sc on.:Ihese sliouicl give the riqhi rmpressions ir';iIh a n-rinimurr o1'verbiaqe. Foi rrrote precision aboui

tlre irrplred inequallttes, see the nurneticai exam;rle in the i:ox "A i:acicr-Ratio Pata<jox."

ln the Rest o{ the Worldln the United States

Wheat expensive, cioth cheap

Wheat cheap, cloth expensive

lnitial pricesr

Trade opens:

- wheat--*->.

< al61[-

Pwheat uP, P.1o,5 down

Produce more wheat. Produce less cloth.

Pr"1"u1 down, Pcto& uP

Produce less wheat. Produce more cloth,

Prices respond to trade.

Prociuction responds to prices.*

For each unit o{ wheat sacri{iced, much land and few workers laid off; extr"a cloth demands many workers and little land,

For each unit o{ cloth sacrificed. many workers and a small amount of land laid of{; extra wheat demands few workers and much land.

Crucial step- National factor markets change.

National factor prices respond.

Wage rates fall and rents i Wage rates rise and rents rise {in both sectors). j iall (in both sectors}.

Product prices equalized between countries. Net gains {or both countries but different ef{ects on different groups. Wilners: U'S' lando,,n;ners, foreign workers. Losers: U'5' workers, {orelgn landowners,

Lonq-run results:

,: l"iTW y*r S#?7. Part O:re

F1&*&€,s"* Winnersaild Losers; Sliart Run versus

Long Rutl'

:-,tj'1.-.l,1..],i,,.].:,,'',.i :':.'::l:t";tlt"lr''""..lt;".'1".-:;;::*::,*::;J; lea'ing zero net gai:r' lnthe long run

n.*ino,,,rh" s1i* T1l:t',1'.n'"lX::x..il",1ff#ij::T'flJ::f#lt;;:f:J'##;;;;r*v botli coutrties get net

g.arns'^ r, .i. , - r G^+^r r , l i;'*;# *;i;tt*niex of displaced bctors'

-- ^.1:.'dlfla

, somel'hat different frou those in the lonq'ryr1.'?'"*o mlie adjus"yent can oicur

m

the long run. Figprg 52.,d;;;;- th* li*'eossiffiilj$Fi *o io.*'' in the'short

and long runs'

-e$*Hs *F THr H..*'T*4r*RY ' THRtrE ls/lPtlcAl tu*s

:: ' , . nplications for factor incomes. rhese,

:1;,*nn-,1ril#f;*11?,1tytr^ffif'.H;o,[*p,*ious secdon :

The'Stol Per-Samuelson Theorem

".s*il+s**rc*;: t*gg' This'tireorem state; tal" grv'en cenain conditions

and

,Four important conditions and assumptions ""1'1'1^t:iifi":t','.-ii[;:ll,$-:1T}'Jl1J;1",.

tT'#i;j":#*:i1i;*uly''-:fi ili-3'iff ;;;1ffi :lq*1ii;ltil*ffi Ix'"'""

ilH!ti*rl;:f"'m:::l;l':?'xl"lill'"'ii[1.':':'.'ffi:*;:]Ltlrln*l:H?1"J;"q"'. are fixed. though rhis

can be.relaxed some|h1t €f":I!:#r[iin pr"a*ing a product double' ...i,fr'";' '.']1u.,

.onuun'. return5 to scale (e.9., it all TaclUI5 u>cu ll l H' "*-

then output o{ the producr qoubles)' 'i

.,a

i ':l

90 ,nternational Economics - Voi. i

Patt One Thi, Ihe,rrr .4 litfrinlrtiorr.rl li.i{L

What we ncw know ahlout the mixtures of productive factors that make up the exports and imports of leading nationi has been learned largely because Wassily Leontief was puzzled in the 1950s. Leontief. a Nobei prize winner in economics. set off a generation of fruitful debate by foliowing the soundest of scientific instincts: testing whether the predictions cf a theory really fit the facts.

Leontief decided to test the Heckscher-Ohlin theory that coun'tries will export products whsse production requires more of the country's abun- dant factors and import products whose produc- tion relies more on the cauntry's scarce factors. He assumed that the U.5. economy at that time u,'as c;pital-abundant (and labor-scarce) relative to the rest o{ the world.

LEdNTIEF.S 'OI

TEST Leontief computed the ratios of capital stocks to numbers of wcrkers in the U.5. export and import-competing industries in 1947. This com- putation required figuring out not only hcw much capital and labor were used diretly in each of these several dozen industries but also how much capital and labor were used in producing the maierials purchased from other industries. As the main pioneer in input*output analysis, he had the acivantage of knowing just how to multi- ply the input-output matr;x of the U.5. economy by vectors o{ capital and labor inputs, export vaiues, and import values to derive the desired estimaies of capitalllabor ratios in exports and import-competing production. 5o the test of

the H*0 thecry was set. lf the United States was relatively capitai-abundant, then the U.S. export bundle should enrbody a higlier capitailiabor ratio {K,/1,), tfran the capital/lat}or ratio embod- ied in the U.5. production that competed nrith imports (K^lL*).

Leontief's results posed a paradox that puz- zled hirn and others: ln 1947, the Uniteci States was exporting relatively labor-intensive goods to the rest of the world in exchange for rela- tively capital-intensive importsl The key ratio (K,/L,J||K_/L^) was only 0.77 when !-l-0 said it shoutd be well above unity. Otiler studie3 con- firmed the bothersome Leontief paradox for the United States between World War ll and 'ig70.

BFOADER AND BETTTR TTSTS The most fruitful response to the paradox was to introduce other factors of production besides just capital and labor. Perhaps, reasoned many economists iincluding Leontief himself), we should make ure of the fact that there are .dif- ferent kinds of labor, different kinds of natural resources, different kinds of capital, anci so forth. Eroader.'calculations of factor content have paicl off in extra insights into the basis for U.S. trade. True, the United States was somewhat capital- abundant, yet it failed to export more capital services thah it imported. 8ut the post-Leontief studies showed that the United States was aiso abundant in farmland and highly skilled labor. And the United States was indeed a net exporter of products that use these factors intensively, as H-O predicts.

lbr physical caprtal. highly skilled labor, and unskilled labor. For niediurr-skillecl labor. the international contrasts are not neariv so sliarp. Countries have them in shares that tencl to be in the nriddle of the abundance -scarcitv spc.cffum. with C,-hina being an interesting exception.

Figure 5.3 conlinns rvhat rve knolv about the distribrition of the tvorld's arable (farrnable) land an<i forestland. These types of land are relatively concentrated

Gi.' &l' si,&r.'' F

E *-l Hi F3, Hi R g.

# H tc

&.r F' B &

pi

$1,

$i

$,' *,,i 4ti!.

ffi f+r $il

. ,, Fr. $i,

Part Or-re Ti:. fii..,,r., ll hfenlrirr:r:.rj Jir.:.,

in Figurr' 6' I low averilge cost is achieved at iu output size of 100.000 units. Anri horvlarge is the unit-cosr penalty' fbr sr'alrer ourpur r.*lt, ir"li;;';.ffi;'L''.*ro** .or,penaltl'lor a scale of 50.000 urtits. conrpaied witli lorvest aver;rse cost at 100.000units" is about 15 percent 1: 12-3 * 2allzal.The ansivers to these questions are inrpor-tant ftrr the type olstructure that arises on the seller sirle of the market.If scale econonies are moclest or moderaie. then there is room in the indirstry fbr alaige nuntber of flrnts. If, in acldition, pr"o,r.ti n.* cilfferertiated. then ne ha'e a 'rildibnn of- i mperf'ect colt.lpetition cal I ecl,...,,,,,,.,,,, _,,, 1,,. ; ;,

strucrure iri which , r&; ;;;;,;#t;;;;,;':;,;;*;.'"iil*n'';,1,t:"::n":fi3,.Tj producing and seJling varieties.of the basic ptorti*r fJecause each firm,s proclirct issomeu,har criffere'nt. each firm rras some co,riroi;;..;;;;;;;.;;;;;,':i,;r*., rbr irspt'oduct' -l-his cttrltrasts with the perfectly cornpetitive narket structure Lrseci in stan-darcl trade fheorlr wrth perlbct conrplrtitiol, each of the inciListry s rnany snrall firnsproduces an identical cotrluodity-lile proouct, ,or*r the market price 65 given, ancibelie'es rhaf it has no direct contr.or or intir,.n.. o"-ri,i, market prce.If scale econotnies ale subsrantial,or.er u rrrt.-rung. of output. then it is likely tliat afblv tlrnts wili grow to be large in.ordei'to reap'ttie-icir*e

"conornies. I j.a ferv lan:e tirn-rsdominate the glohal inrltrtlr.l rrer*pno-be..r"'..,.g.r.riohrtrdr buarc econ0nltis, then wehave an :;i:' . ' .

-fhe large iirms'i,, un origoplrl;'l;; that the1, can conrror or inflr-r-ence pdces';\ ke,r'issLte in on ortgopoty i, irJ*ju.ilu-lf drgrg large rir.rns cor,pere

'virh eaclt other

il^'i:ij: lot cqmpete,-too aggressiveiy, rtren rr rs possibte for rhe ftrns roealr econofitlc (or p,re) profit' prolit -qreater than tire noi'nrut retur-n to inuerte,l capital.In a iater section of this cltapter. we present in-rlepth anal;,sis of hos, r\€ cau usemonopolistic competition to better unclerstand ttre clririers and irnplicatio's of interna-rionai trade' we then take a lock at the rore oi";;il;;;in inrernarionar rracre.

External Scale Economies 'fhe

second rype of scale economlbs is external to any individual firm. I;xt*r,:l*i:**le ecg:3}*rxi*s are based on the ri.* oiun'*nru=lncuro, wifhin a specific geo-graphic area' The aver"age cost of the rypicat iirrr'pt"ir"i"g tlre procru6 in rhis areadeclines as rhe output ol ti.,* indrrytn {all the local fitrvithin rrre area is targer. rxt*r,,ui..o,,onries exprain ,n-t};rL;tll,:ltJiilj'fr,il:]||:?ol'some products in specific geographic.r.ur. wrruiu.",rr. sources of externar scaleeconomi es l) T here are several-pois i tri I i ti es. External ecotlonttles can arise if concenlration of an industryls flnls in a ge0graphicarea arracts greater local supplies of specializ.o ,*roi.*u ior the inclustry or iarger pooisof specialized kinds of laboirequired b1.the il;r.),';;;" r*o*on lhal filmrnakins trmsclusrer in Hotvrootl lancl irr Niunibai" fornierrl.. Bo;;; il;. il;;.ilffi,;wood.)is that a deep pool of r'vorkers sr<irreotirii,e-";;l{;",its orfirrnrnakrng.nd a mrlgeol compatries provirling special services to facilitate fihnmaking are available in iheselocati'ns' (You can also see tire self-reinfgrcing ;$;;';;;.rernal scale economies-*ifyou lvant to break into fjims and fihlmaking."or io offer-seruices to rii'r companies,you knorv where you need to go. ) External ecinonrjes .nn rrrn result as n.ru t norlt*ag.about prodtlct ancl production-tecluology (or other useful business intor.rnatjon) cliifusesquickly alnong flrrns in the area, through direct coatacts unrong the f irms or as skilleilr'vorkers fia'sfbl from firm to firm. K_nowledge ;rft;;; is rapid and perso'ner arecontinuallv shifting among the high_tec}rnot..,g1r-**flteirottrr"re.

semiconcluctor. and

lnternational Economics, .l 5th Edition 1A7

Chapter 6 Sr:rle Econorrtiel, ln'$cfiect Oornperirilrr. rn,l lrtrtlc ::i:'

related fir-rns in Silicon Va1le.v. Ottrer examples of clustering driven by exteriral scale economies include banking and flnance in London and in New York City; stylish clotlt-

ing. shoes, and accessories in ltaly; and watches in Srvitzerland. The final seclion of this chapter examines how extertral scale econotriies add to our

ability to analyze infernational trade'

i i:.i I it.iii' ; r!"i i3 i-j'$ T' iey gffi ee i"ll fl

According to our basic theories of comparative advantage. if a cotntry trades in a product, ii shsuicl either urostlv export the product (based on its relatively loiv prOdttc- tiot-r co.ts) or mostly import the product {as a result of its relatively high production

costs). \ve wr:uld see only int€r-t*l*{.*$try gr*d*" in rvhich a coLnltry eKports some products in tracle lbr irnports of other. qLrite diflerent products'

While much trade is inter-industr-v, especially in agricultural products and other pri-

rnary procficts. there is also substantial l*i:"i ;r't:f iilif i !.1;jl.lli iiiii--tivo-way trade in rvhictr a corntry both exports and inrports the same or very simiiar products (p:oduct

varieties that are such close substitutes that lhe-v are classified rvithin the sanre indus-

trv). Figure 6.2 provides some exarnples of product categories in rvhich the United Siates engages iri intra-industfy tlacle. For exanple, tbr each of pelfitntes and costnet-

ics. the amounr that the United States exports is close to the amount that it imporls. We cau quantib/ the relative importance of IIT in a cortntry's trade in a productby

splitting the total irade {the sum of exports plus impolts) into two conlpolients. 'Ihe

first cornponelt is i3*t cr,}{**. the difference belween expofts and ilnports of that prochict. Net trade is not intra-industry trade. Rather, net trade shows the product'.s importance in the country's inter-industry trarle. in which son"ie products are (net) exported. and other products are (net) irnportecl. In measurentent rve often use the con-

veirtion that net trarle is a positive value if the country is net exporter of the product. and it is a negative value if the country is net importer of the product. {Tlie "positive" and "negative" are nct vaiue judgments. just ways clf nreasuring.)

Intla-industrv trade is then the otirer corrponent of the country's total trade irr tlte proclgct, the an-iount of trade in rvhicli the country is both exporting and importing in the

iarne proriuct category. l'here are tr.vo eqLtivalent \vays to treastlrc tlte atnount of a coutt-

triri intra-industry trade in a procluct. Fils:, it is equal to hvice the value olthe smaller of expotts or imports icapfrriilg the amount of exporting of the prodrict fhat is nratched b1' the sarr'le arlrount of importing ol the proclLict). Second IIT is the part of total trade in the product (expolts plus irlports) that is not net trade. Using this latler approach,

IIT: (.Y + ktt - lx * 1t4i rvbere X is the value ol export$ ol the. product and .44 is the vah.ie of imports o1'the prorluct. Fnrrherntore. rve often \yant to compare the irnportance of IIT across different prr:ducts or ciiltbrent countries. Because the total amotlni of trade diftbrs naftrraliy tbr

dilferent products or dillerent countries" \re can meastlre the relttlire inrportaltce 0f intra-industrv trade. as a share ol total trade:

(I +.1,1) - d- ,1/lftT. ll Illlsnare= ---, . lotal tradr

.\-- r11l ll- I - rx+ hktx + h{)

lnternational Economica . Vol, l

*S Part One Tl:r- Thearl of lntettuttronalT;rulc

rf*{JH€

For each product, intra'ildustry trade is the difference beiween total.trade atld {the absolute value ofl net trade. The {IT Share is intia-

industfytIad€asapelcent4geoftotaltrade,.].... The,numbers in parentheses xe the Sdigit eode, Stanalard lntemational,Trade Classification, Revision 3.

t Source: Data & exports and imports fmm UnitedNationS,,Cammodit!: Dilision, IlilCantrade DstabLte.

,The values of these vaiious measures are shown in Figue 6,.2. For six of the seren products (all butlargeaircraft),roverhalf of U.S.'total trade,inthe product is IIf.

How large lslintra-indlstry tradein geneJal? It is more important for trade. in manufac- twed.products that it is for tade in agricultrnal products and other primary products, so let's focus on nonfcod, manufactured productsi Figure 6.3 shows the average share

of IIT.in the noafood.manufactures trade of'six indestriatized countries. The estimates

are based on dividing the sector,inlo over 1,3'00'diftrent prOducts"'To develap:the estimatei..the uT shareis calc$lated far'eachofthese,products, and then the weighted average is.catculated,aoross all these'products (using the counhy's total trade in the produ;t as weights, so:that prodtrcts,with m-.ore total trade receive more weight'in lhe

overall average).|The fine'level,ofproduct detail is necessary so'that we have mean- ingful, narrowly;defiued indus111-es W-e,t{y to-aVoid brasing the measurement of IIT

tWe can show the formula for this weighted average of llT- shares. To be clegr, let us add the subscript

I to the value for each products exports,and:'import5, to emphasize that we are ar.'eraging over a

number of different PlQduct categories.

How lm portant ls'lntra.lndustry ifaAef

tn this formula, tho first katio) term is thle propcrtionate weight lor each produci category,, and the

serond tetm is the calculated llT.share for each product category' :

IntraJndustry Trade for the United States, Selected Products, 2009

] t',r,n,,.-r,]

lnternational Economics, 15th Edition I 1'09

Chapter 6 Sc,rL' Ecrr)l0r.rri(i' ltnlri'iiccr ( 'r']nJ)' iiiiiltl ({llrl J jiitl'

Avelnge

Perceutage

Shates o{ intra- lnilusfry Trade irr the Countr:yi Total Ti'ade in Nonfood Manutactured Frodricts

Country

United 5tates L dildud t^^-^)dPdt I

Germany France United Kingdom

1989

55.3 54,3 l/.6 62.6 71 .3 69.0

2S05

58.3 63.2 41.2

67,' 13.9 11 -1

tr. $rl F. il.

F' sir..

g:;

*!'i t".:

ir ! ::l

i,,

i, i:i

.1.

:.' a;

..

Iol e thflll hal l of overll l For $ade il iloulbarl malufucturerl goocls (slTC 5 throtrghSi, il1tla"in<lirstry tlade

is n'

trade for rtost intiustliaiized ,ount'it'' The e'riilrates ore based on ot"er l 'J00 tlillerent indn'idta]

ploduct

.*ijeo'i*'(fhe5-ttigittereloftlreSl.1'C.iter,isiorr2).

!t!ttilu!i,'tu! Itt,h hl'l trntnoli!r ll'tralwv '

$pwafrl rvhicir codd happen if we instead tsed broad p{O€luct categofies i1 *arich

\ve wollld be {rr:istakeniyi tind;ng {l'f rvhen a country expofiS "ap1}les" and irlpCIlts

':orrr,g.r;' tir ihe too-broad prodtlct calcgory *'as "fruit")lt ,,-

As r,,,e can see ir'r tigur.L.3, tb.fir.e'of ihe six countrjes. IIT is rrore tllftn h1l1.Ltt-

total trade in nonfood ,ianut'acrures. ln additioo, the ilnportance of II'l has tieet rising

in the past several decades f-or:lll these cortnffies. f he iuttto** is especiallS'iarge tbr

Japan. h} this aspecl of its itade, Japan it still rlifl'elent li'E:in most ather indttstrialized

Cou11tl'ies"btititisn0w1es5c]il1-ererrtthanitLl5edt0be. trve know fr.orll otirer studies that II't is mcre pfevaletli-where tratie balriers

and

,transporr cosrs aie J;;;: ;;;;lti,, f,=iaontial-trade areas like the Europeatr Unio'' i;t-i!t.6.3. the aveiage IIT:shar"es are particularly high lbr the iitr*e E*ropean cou*rries. Furrhermcre,ltr. i, nrore chalacieristic ol rhe {higlr-lncome) ilrduslri- aiir*J ."""tti*s. aod average tlT sk*tes tencl to be lorver tbr the (lor'-itlcolne] #;Jiopl;'.-i,.itr*. ror ittsTance, for, Clhtnai rade in-I992'-the ave.rage ttl shlt1 l.,lr its trarle in nonfoOd nratlufactures'was 20,9 pefcent' Even:this ditference

is nar:row-

ing. tjror.igh. lbr the Oeoelopirrg cogntries fik* bhit'o that are iniegratirrg into the glottai

econony, Sy 2004. tfi* ot*ug; share ol ilT in llhina s tlade in nolltbod irtanufactured

prodilcts had riscn to 4t.i percent.

What Explains lnfa'lndustry Trade? why elo \ve'see so much intra-inclnstry trade'i'rhele are several reasoils'

some mea-

,ur*,] ltt probably reflects trade tlrir,'err by comparative advantage' For in-stance,' a ,producr caregory *ui'Siili ne ,io" liroad- so ttrai it.include.'i tlifferent products

that

ate Fr$duceci using dil.ferent productign tuetilodS' Rrithin a cal*gol'y: ior ittstanee'

the united States nay exBol't ipecific products that are produced using skilled labor

;ni*,iuiuety nnd impoir oifi*t:tp*citic piodgcts that are prnducecl rtsing unskilled iab*t:

'ln ihe analysis of intra-industry tra(ie, a'.'ieil-defined product category.con:ists of prclduct varie{ier

rrierveri as close sui:stitutesiy.ontrrn.ro a,rci prodL:ci-cl using verylimiiar f*ctor proporlions' ihe la,tter..is

;;;;;;;;or.n to tt.tn deiinitiorr oi an indi.istry stressed in lhe l-leck*ciret-Chirn thecry of coirrpalallve

acivantage, Altlrough we.;;;;i'p;;ti .t...n Jt these over 1,30t) product caiegori€s n-ie'eis lltis

deiinitisi r, il seems unlikelv ihat lhe c&iegorles generaliy ate too btoac'

110 lnternational Economics - Vol. 1

Part One T'lre 'fleln: of Inrrtirrtionrrl Ttrrl.

intensively. As another example. for some agricultural products. III-'measurcd over a Year may reflect seasonai comparative advantages. The Uniled States expor"ts cherries in July but irrports cherries in January.

Much intra-industry trade is driven by something other than comparative aclvan- tage. and the leading explanation focuses on the role of glr*{**{t *jff*r**€*;l?i*r,-- consLlmers view the varieties of a product otTered by difierenl lilns irr an industrv as close but not perfect substitutes for each other. We can see the roie of diiTerentiation for the products sliown in Figure 6.2. Product difl'erentiation is rampant in perfumes and cosn':etics, r.vith strong brand na;nes. exotic packaging. and rvide-ranging claims for eff-ectiveness ancl high quality. For clothes wa-shing machines. brancls tiom out- side the tJnited States. including lVliele. Bosch, LG, Samsung, and Flaier. contpete ivith Whirl;:ool. Maytag, GE, and other U.S. brands. in integrated circuits there is ciifl'erentiation by specific circuit design as nell as by functional specificarion, rvitir U.S. firms like hitel. Texas Instrunenls. and Micron battling with tbreign iirms like 'foshiba" Samsung" and Infineon. f-he range of automobile models on oi'fur is obvi- ouslv very large. {n iarge civilian aircraft the two firms are Boeing_ and Airbgs. rvith each offering models tliat compete rvith the other\ models on design, lechnolerg1,. apg otlter specifications. For cameras. consruners can choose fiom a range of branc{ names (including Canon. Sonv" Samsung, Kodak. and General Electric) and ntodels that clif- fer by fbatures and qnalit.v.

Procluct differentiation easil-v can be a basis for ffade. Sonre buyers in a country prefer varieties cf the product produced in a foreign country, so thev $'ant to import thc- product. The foreign varieties are iiot necessarill' cheaper than the varieties pro- duced by local ftrms. Rather, each loreign rrariet)'has dilleretrt characteristics. and these buvers find the fbreign rrarieties' charactedstics to be desirable. Ar rhe sanre time, sonte fbreign buyers prefer varieties produced b-v firms in this coLtntr,v, and these firms are able to expolt to the foreign country. Even if there are no relative cost ciifl-er- ences of the tvpe ernphasized by theories of comparative advantage, there is interna- tional trade. And, mucli of it can be intra-industry trade. rvith a counry exporting and importing diflbrent variants of the same basic product.

Yet. it is not only producf differentiation at r.r'ork in this explanation. Taken to ihe extleme, each of-ns rvould u,anl to bu-v the riniclue product variant that exactly matches our personal preference. This would be dift'erent fiom anyone else's exact variant. so the product variant would be completely customized to orr own individrul taste. But we know that tiris is not the case for the products shown in Figure 6.2. Each has a limiteci nutnber of variants. The customizatiort is lirnited. What limits the number of variants, and the ailtount of customization'J The answer is sonie forr:r of internal scaie econom;/. so that there is a cost advarrtage to prodLrcing larger anounts ola specific variani.

:.,i i: i.j ii Fli.l: t $ i? i { {_ i:} re ::} il'Ti ?' + 4:.} ix,e. ili i}'r & & t} r To understand rvhy \ve see so rluch intra-industry trade" we need a model of trade driven by product dillbrentiation and scale econonies. ProdLlct ilifibrentiation is a deviation from the assurnption of a homo-eeneous product, otle of the standard assump- tions used in the analvsis of perfect competition" so we are entering the realm of imper- fuct contpetition. Our analysis is based on a rnild fbrrn of imper'1bct competitian called