LAWRENCE B. KRAUSE
Brookings Institution
assisted by JOHN A. MATHIESON
Brookings Institution
How
Much
of
Current
Unemployment
Did
We
Import?
1970, unemploymentin the United States
rose to 4 percentof the laborforce,the presumedfull employmenttarget.
By the first quarterof 1971,unemploymentreachedthe undesirablelevel
of nearly6 percent.Overthissameone-yearperiod,importsinto the United
States increasedby $4 billion (1971:1 over 1970:1 at annualrates) and
exportsincreasedby only $3 billion,reducingthe net positivetradebalance
by $1 billion.An increaseof competitiveimports-consideredin isolation
-has the directconsequenceof reducingjob opportunities.A reductionof
exportshas a similarjob-destructioneffect.By the sametoken,a reduction
of importsand an increaseof exportshave the oppositeconsequenceon
job opportunities.In view of the trade development,a seeminglynatural
questionis how much of currentunemploymentdid we import?'
Actuallythe questionis not as naturalas it appearsbecauseinternational trade cannot be consideredin isolation from other economicdevelopments.In an efficientlyevolvingdynamiceconomy,job opportunities
are constantlybeinglost becauseof manyfactors,but the losses are more
than offset by job-creatingdevelopments.Shifts in consumerdemand,
changesin productivetechnology,restructuringof governmentexpendiIN THE FIRST QUARTER OF
1. Some observershave alreadygiventheiranswersto it. See, for example,AFL-CIO,
IndustrialUnion Department,Crisis:Importsvs.Jobs(Viewpoint,Vol. 1, Summer1971).
417
418
BrookingsPaperson EconomicActivity,2:1971
tures,as wellas increasesin imports,causeunemploymentamongdomestic
workerswhoseproductis no longerneeded.Generalunemploymentneed
not occur,however,as long as aggregatedemandis maintainedin the economy, if need be throughgovernmentalpolicy. In the particulareventsof
1970-71,full employmentwas not maintainedand,therefore,the contribution of the declineof net exportsto the increasein unemploymentbecomes
meaningful.To be sure,the employmenteffectsof the declineof net exports
could have been offset by appropriatefiscal stimulation,but it can be
arguedthat trade movementsare relativelyunpredictableand, therefore,
some unintendedunemploymentmay have resulted.
It is not legitimate,however,to count the entiredeclinein net exports
since 1964 as a cause of currentunemployment.Up to 1970, displaced
workersin the aggregatewereabsorbedelsewherein the economyand did
not createan unemploymentproblemas such.Indeed,whilethe tradebalancewas decliningfrom $6.8 billionin 1964to $0.7 billionin 1969,unemploymentalso declinedfrom 5.2 percentto 3.5 percent.
In additionto the issueof total employment,the economicwelfareof the
societydependson whatkind of job opportunitiesare createdin exchange
forthosebeingdestroyed.Whenspecificjobs arelost throughtechnological
change,for instance,thereis a presumptionthat economicwelfareis improvedeven though the workersactuallydisplacedmay not find equally
remunerativework. A similarpresumption,based on the theory of comparative advantage,exists concerningdisturbancesarising from internationaltrade. Society'swelfareshouldimproveeven thoughthe individuals displacedmay suffera loss. This presumption,however,is dependent
on the country'shaving an equilibriumexchangerate. An overvalued
exchangerate would not directlypreventa country from reachingfull
employment;but at full employment,resourceutilizationwould be distorted away from tradablegoods towardnontradablegoods and services
with resultingwelfareloss for the world as a whole.
Methodology
The purposeof this exerciseis to obtainan estimateof the employment
consequencesof the declinein U.S. net exportsbetweenthe firstquarterof
1970andthe firstquarterof 1971,the periodin whichunemploymentrose.
Thefirststepin the estimationprocesswasto recordthe changesin exports
and importsat annualrates. Becausedifferentproductshave widelydif-
LawrenceB. KrauseandJohnA. Mathieson
419
ferentlaborinputs,some disaggregationwas required.Sincethe intention
is to relatetradechangesto employmentlevels, the classificationscheme
utilizedin employmentreportswas adoptedas shownin Tables1, 2, and3.2
Thetradestatisticsweretakenfromthe Censusreportby end-usecategories
andaggregatedto reachthe properclassification.3
Thetradedataare given
in Table1, withexportsandimportsshownseparately.Sincethe datacome
from Censusreports,they differfrom the merchandisetotals reportedin
the balanceof paymentsbecausethey areunadjustedfor timingandcoverage. Furthermore,the rise in militarysales($307 millionannualrate)and
unclassifiedexports($131 million annualrate) have been excludedfrom
exports.The growthof unclassifiedimportswas similarlyexcluded($267
million), as was the decline of noncompetitiveimportslike coffee ($127
million)sincethese declinesdid not lead to increasesin Americanproduction. The resultingincreaseswere $4.1 billion in importsand $3.1 billion
in exports.
The secondstep in the estimationprocesswas to evaluatethe numberof
jobs lost andjobs gainedas a resultof the tradechanges.Thereis both a
directjob effectwithinthe trade-impactedindustryand an indirecteffect
in industriessupplyinginputsto it.4The technicalcoefficientsfor both the
direct and indirectemploymenteffectswere taken from a study by the
Bureauof LaborStatistics.5This studyis based on the 1958input-output
tables, but projectedto 1970based on trendsin productivity.The prices
are also on a 1958base, which causessome difficulty.One would like to
deflatethe trade valuesby indexesof pricesof importsand exports.But
sinceindexesof tradedgoods of individualindustriesarenot available,no
adjustmentwas made.This will tend to bias upwardthe estimatesof both
jobs lost andjobs gained,but probablywill not be very significanton the
net balance.6
2. Employmentand Earnings,Vol. 18 (July 1971).
3. U.S. Bureau of the Census, Highlightsof U.S. Export and ImportTrade, FT990
(March 1970), and Highlightsof Exportsand Imports,FT990 (March 1971).
4. WalterS. Salant and BeatriceN. Vaccara,ImportLiberalizationand Employment
(BrookingsInstitution,1961).
5. U.S. Bureau of Labor Statistics, Projections1970: InterindustryRelationships,
PotentialDemand,Employment,Bulletin 1536 (1966), Table V-2.
6. As a crude check on the direct employmentconsequences,net value added per
employeewas calculated,based on 1969pricesfrom the AnnualSurveyof Manufactures,
and used to estimatethe directemploymentconsequencesof trade changes.The results
were not significantlydifferenton balance, but they show larger absolute numbersof
jobs lost and jobs gained. See Bureau of the Census, AnnualSurveyof Manufactures,
1969: GeneralStatisticsfor IndustryGroupsand Industries,M69(AS)-1(1971).
u W,
ffi
>
^>ol
00
e
o
0 00
t
t_ on
tn
t
Lt~~~~~~~~~~0
00
~~~~~~~~~~~t-
2
en
en
tn
I'l
N
oo \co oo 's
CS*{oo
~~~~~~~~~~~~~~oo
en
oF
..q..q..
4 {moIn
.
~
-4
C1 C1^ n
00
t tW>
(ON
C
M W
10
N:
C
en
W>^
0
.. n
.
ge
.0 I'X
.............
.
C1 "Nt en V-- V- 0
t- "o
en
o en oo tn
m
>
o
oOo~~~~~~~~~~st. Nc
bo,
N
-o
cn
[E~~~~~~~~~~~~~~b"
tWb
oN
Oo t-NNb
00 0%
oo
%
N
kn
o~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~b
w
e
F
O
rv
t~~~~~~~~~~~~~~~~~~~~~~~f
~~~~~~~~~~~~~~~~~~~~N
X 0
tn
-~~~~~~~~~~~~
(Z
S
WcX
t-
oo
o WoONlc
N eon o
N O
o
--
a
eq
b
N
T
44, C
rN
oFo
N
X t*FY
WImNNn
C)
O 4bN ( N
S
n
00o
t abt*
r-
CN
t
C
*o
m en
eno
tzz
o~~~~~~~~~~~~~~7
Pa~~~~~~~~~~~~0
c
0
t-
C4o
-
"Z
C
h
.0b
1
0
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I4
6~.
Z > W oo ? oo as b ooe ot
N
cn
6too
N
tA
~~~~~~~~~~~~~
;
tC
cy
'
?
et)S
bo
>1
835,
Cd
A YI
10
A t aS
> g 2 ? ?
r;
X t?e5Y.Xg~~~~~~~~~~~U
4 g Q d X 00
S~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1g
@
-
lo
I
BC'?s
mb
v~~~~~~~~~~
e
LawrenceB. KrauseandJohnA. Mathieson
421
Findings
The consequencesof the trade changesfor employmentare shown in
Table 2. In directemployment,an estimated47,800jobs werelost due to
decreasesin some exportsand 134,400jobs were lost due to increasesof
imports,makinga total loss of 182,200jobs. However,181,400jobs were
gainedthroughincreasesin exportsand a further1,300jobs werecreated
throughsome import declines,making a total of 182,700job increases.
Thustherewas on balancea slight edge towardjob creationwhen direct
employmenteffectsare takenby themselves.
,Theinclusionof indirecteffectsshifts the balanceslightlytowardjob
destruction.Therewas a net loss of about17,100jobs dueto indirecteffects
(227,600losses, offsetin partby 210,600job gains).It is reasonableto expect the indirecteffectsto be more unfavorableto employmentthan the
directeffects.Muchof the growthof exportswasin agricultureandmining.
In both of these industries,the ratio of valueaddedwithinthe industryto
total sales is much higherthan it is in manufacturing,wheremost of the
import increaseswere recorded.Since the indirect employmenteffects
come fromthat portionof salesvaluenot addedby the industryitself,the
observeddifferentialwas to be expected.
r,,:Whenthe directand indirecteffectsare takentogether,a loss of 16,600
jobs between1970:1and 1971:1can be attributedto changesin exports
and imports.If the displacedworkersstayedin the labor force and were
not absorbedinto employment,then approximately16,600 people were
addedto the unemploymentrolls fromthis cause.Duringthis span,total
unemploymentrose by 1.7 millionpersons.Thus internationaltradecontributedless than 1 percentof the increase.To put the resultssomewhat
differently,if unemploymenthad increasedonly becauseof tradedislocations, the unemploymentrate wouldhave risenfrom 4.16 percentto 4.18
percent,7ratherthanto the actual1971:1 averageof 5.93 percent.The net
job loss may appearsurprisinglysmall, given the declinein net exports.
relative
But it shouldbe rememberedthat U.S. exportsare labor-intensive
to U.S. imports, as Leontief establishedand others subsequentlyconfirmed.8TheUnitedStatesneednot havebalancedincreasesin exportsand
7. Assumingno change in the labor force.
8. See commentsin WilliamH. Bransonand Helen B. Junz, "Trendsin U.S. Trade
and ComparativeAdvantage,"this volume.
i
t
R
oW-4
F
o
C)~~~~~~~S
r
Q
2
S
? 0
on
C)
V~~~~~~~~~W _^
V
t
tn ens
o
t ~~~~tntn an
2
t
t
tn
oo oo
tn
EX
oB
W ff
j
'W fW't
e^-
ef0
r
Oe.I
t
I~~~
I
~
o
M
0
'O
W1
00
0%
00S
ei 4
m
N
S*
~~~~~~~~~~0
T-4
*
*
..NW
:,1
*
CO
'r i o
ei
*
0%
14 1~6
06
r.=..
*
0
0%
C; C
.
.
.
a
.
0
e
C
4
*
4.4~~~
00
*
.
00O O
~~~0g~~~~o
.0O
~~~~~
0
*
..
0
~0
0
0 t2
T-4
'O
tT-4
.
*
0..
e
0
0
ce*
c;~~~2v-
*
-40%
O
0'O
O
0
.C's
E
E
lo ffl,
E
8,
U
j
i
00
0) la0
'0)
cd
z60w
.
.
cl~
0
C
l .
t
3
424
Brookings Papers on Economic Activity, 2:1971
importsto maintainemploymentin trade-relatedindustries.If imports
increasemorethan exports,as occurredin the periodunderinvestigation,
job losses may not occur.
Whileonly an insignificantportionof total unemploymentcan be attributed to internationaltrade, the employmentsituation in particularindustriescouldbe greatlyaffected.Table3 relatesthe net directjob changes
due to tradeto the employmentsituationin specificindustries.Onlydirect
job effectswereusedsinceindirecteffectsarespreadwidelythroughoutthe
economy.As is seen, employmentwas down4.6 percentin primarymetals
(mainly steel); 2.5 percentin motor vehicles;and about 1.5 percentin
textiles.On the otherhand,coal mininggained6.6 percentin employment;
other transportationequipment(mainly aircraft),4.4 percent; and machinery,2.0 percent.Thesedata suggestthat a good case can be madefor
adjustmentassistanceto industriesthat are seriouslyaffectedby imports.
However,the imposition of import barriersis very dangeroussince it
usuallyleadsto retaliation,whichcan haveseriousconsequenceson export
industries.The difficultiesimportscausetextilefirmsare well known,but
little attentionis paid to the effecton the hard-pressedaircraftindustryif
it could not export.
QUALIFICATIONS
The interpretationof thesefindingsrequirescertainqualifications.First,
withanexpandinglaborforce,theavoidanceofjob destructionis not enough
to ensurefull employment:Jobs must be createdin sufficientnumbersto
accommodatenew workers.Thus,eventhoughinternationaltradedid not
destroymanyjobs on balance,neitherdid it carryits weightin creatingjobs
for the new entrantsinto the laborforce.Second,evenwhenthe total number of jobs createdis equalto the total destroyed,if the absolutenumbers
arelarge,frictionalunemploymentcould increase.To establishthis point,
the varianceof employmentdue to changesin internationaltrademustbe
comparedwiththe variancein employmentfrom all othercauses,and this
calculationwas not made.Finally,net incomeflows in the economywere
reducedby $1 billion due to the declinein net exports of goods, and the
multipliereffectson employmentwerenot estimated.However,thesereductions weremorethanoffsetby the $2.7billionrisein net exportsof services
duringthe sameperiod.In any event,the multipliereffectswouldbe small
in an economy as large as that of the United States and impossibleto
calculatebecauseexpenditurepatternscannotbe traced.
425
LawrenceB. KrauseandJohnA. Mathieson
withandwithoutEffectsof Changesin Exportand
Table3. Employment,
ImportTrade,by IndustryGroup,First Quarter1971
Thousandsof persons,seasonallyadjusted
Industrygroup
Net
primary
job loss
Potential Job loss as
due to
changesin employmentpercentage
Average exportand without of potential
employtrade
employment, import
ment
trade
1971: la
effect
-0.5
-0.5
74,026.7
22,535.9
Total
Traderelated
74,027.2
22,536.4
Manufacturing
18,749.2
59.2
18,808.4
10,100.8
Durable goods
1,255.3
Primarymetals
1,327.7
Fabricatedmetals
1,806.0
Machinery
1,791.0
Electricalequipment
Motor vehiclesand equipmento
879.5
914.3
Othertransportationequipment"
2,127.0
Other durables
33.0
60.3
0.0
-34.9
9.9
22.1
-38.4
14.0
10,133.8
1,315.6
1,327.7
1,771.1
1,800.9
901.6
875.9
2,141.0
0.3
4.6
...
-2.0
0.5
2.5
-4.4
0.7
Nondurablegoods
Food and kindredgoods
Textile mill goods
Apparel,finishedtextiles
Othernondurables
8,648.4
1,781.3
946.7
1,378.7
4,541.7
26.3
-23.3
3.6
15.3
30.7
8,674.7
1,758.0
950.3
1,394.0
4,572.4
0.3
-1.3
0.4
1.1
0.7
Agriculture
3,380.0
-49.6
3,330.4
-1.5
Mining,total
624.0
-10.1
613.9
-1.6
Mining, traderelated
Coal and relatedfuelso
Crudepetroleumand naturalgas,
407.2
152.9
254.3
-10.1
-9.4
-0.7
397.1
143.5
253.6
-2.5
-6.6
-0.3
Services
Contract construction
48,043.0
...
48,043.0
3,231.0
...
3,231.0
b
b
0.3%
...
Source: Employmentand Earnings, Vol. 18 (July 1971). Figures are rounded and may not add to totals.
a. Quarterly average.
b. Less than 0.05 percent.
c. Not seasonally adjusted.
426
Brookings Papers on Economic Activity, 2:1971
JobsandDollarDevaluation
One of the pillarsof the new economicpolicy announcedby President
Nixon on August15, 1971,was a groupof measuresthat shouldleadto the
devaluationof the dollar.As noted above,this shouldimproveeconomic
welfarefrom internationaltrade, since the dollar was overvalued.The
devaluationwill also stimulateU.S. exports.Domestic employmentwill
rise as a directconsequenceof the devaluation,for U.S. exportswill be
stimulatedby it, and as importsrise in price, Americansshould switch
consumptiontoward domesticallyproduced goods. It is importantto
recognize,however,that the devaluationis desirablenot to alleviateunemployment,but becausethe dollarwas overvalued,as evidencedby the
disequilibriumin the overallU.S. balance of payments.At full employment,therewill not be morejobs becauseof the devaluation,but different
jobs. Employmentin trade-relatedindustrieswill be greaterand employment in purelydomesticindustrieswill be lower becauseof it. The most
fundamentalprincipleof the InternationalMonetaryFund is that depreciationof a currencyshouldnot be used to correctan unemployment
problem,for that merelyexportsthe unemploymentto a nondepreciating
country,a practicethat can lead to competitivedevaluationsand a host of
otherbeggar-my-neighbor
policies.
an
Thereis importanttie, however,betweenthe timingof a devaluation
and the existenceof unemployment.Devaluationby itself is stimulative
and inflationary.Countriesusually combinedevaluationwith monetary
andfiscalpoliciesdesignedto offsetthe stimulationso as not to losethrough
domesticinflationthe competitivegainthey achievevia devaluation.Given
the currentslackin the U.S. economy,the stimulativeconsequencesof devaluationcanbe accommodatedthroughgreaterproduction,but the direct
inflationaryeffectswill still be of someconcernas the higherU.S. pricesof
importedgoodsraisethe pricelevel.
As part of the new economicpolicy,PresidentNixon imposeda temporary 10 percentimport surcharge,presumablyfor the purposeof urging
other countriesto appreciatetheir currenciesto effectuatethe dollar devaluationthe United States seeks. It shouldbe noted, however,that the
surchargeinterfereswiththe formulationof a newpatternof exchangerates
via marketaction since it artificiallystrengthensthe dollar, as does the
maintenanceof capital controls.Permanentrestrictionson trade merely
LawrenceB. KrauseandJohnA. Mathieson
427
maintainan overvalueddollar, thus sacrificingexportswhile restraining
importsand on balancereducingjob opportunitiesfor Americans.
Conclusion
How muchof currentunemploymentdid we import,then?None, apparsmallportion.Domesticeconomicpolicy
ently,or at most an insignificantly
musttakethe blamefor the overallunemploymentsituation.Nevertheless,
individualindustriesare greatlyaffectedby shifts in trade, some losing
employmentandsomegainingit. Publicpolicyshouldhelpeasethe burden
of adjustmentto shifts in trade as it should for other causes of unemployment.
The properpolicy instrumentsfor dealingwith generalunemployment
are monetaryand fiscal policies. Since these work best when used in
tandem,monetarypolicycannotbe, as is sometimessuggested,separately
purposes-at least not without signifitargetedfor balance-of-payments
cantcosts. Monetaryand fiscalpolicymustbe free of externalconstraints,
and this can be accomplishedby keepingexchangerates at equilibrium
values. Maintainingequilibriumwill requiresome flexibilityin spot exchangerates.With equilibriumexchangerates, one need not worryabout
the employmenteffects of a decliningtrade balance.Clearly, politicians
mustforgothe simplisticbut appealingtie betweenimportsand unemploymentif a viableinternationalsystemis to be maintained.
Discussion
WALTER SALANT FOUND IT difficult to interpret the question Krause
was posing.He felt that the issue on the importside was betterphrasedin
termsof the changein the propensityto importratherthan the changein
the volumeof imports.Sincerealimportsrosedespitethe recessionandthe
greatcut in use of capacitywhentheywouldhavebeenexpectedto fall, the
actual rise in imports understatesthe rise in the propensityto import.
Analytically,the Krauseapproachis similarto the failureto distinguish
betweeninducedandautonomouschangesin fiscalpolicy,usingthe change
in the actualbudgetdeficitor surplusinsteadof the changein the full em-
428
Brookings Papers on Economic Activity, 2:1971
ploymentbudgetas an indicator.Krauseagreedthat a measurementof the
shift in the importfunctionwould be an interestingexercise,but a much
more ambitious-and perhapsimpractical-one.
RobertHall expressedanotherconcernaboutthe questionposed in the
title of the paper.In general,a rise in unemploymentis not matchedby a
complementaryreductionof employment.In point of fact, employment
has been essentiallyon a plateauin 1970-71,but unemploymentrose becausejobs did not increaseas the labor force expanded.Secondly,even
in a recession,much unemploymentis caused by people changingjobs,
and, as Krause showed, alteredpatternsof exports and importsforced
peopleto changejobs. Krauseacknowledgedthese points,notingthat his
exercisewas intendedto refute some invalidclaimsthat were framedin
termsof unemploymentand wereequallysubjectto Hall'scriticisms.
WarrenSmithand CharlesBischoffemphasizedthat an adverseshift of
the trade balancecould have negativesecondaryeffects on employment
throughthe multiplier,evenif the primaryeffectwerezero. The multiplier
attachesto the primarychangein income,which is negative,ratherthan
to the primarychangein employment.
Salantwishedto qualifyKrause'spropositionthat devaluationis not a
propertool for curingunemployment.Whilehe agreedin general,he felt
that, to the extentthat unemploymentis due to overvaluationof the currency, restorationof high employmentby devaluationis preferableto a
cure by monetaryand fiscal expansion.The assumptionunderlyingthat
propositionis that resourceallocationwith a propervaluationof the currencyis in somesenseoptimal;if thatassumptionis valid,devaluationtends
to restorethat allocation.
© Copyright 2026 Paperzz