Professional Documents
Culture Documents
Distortions
GeneralEquilibriumModel
of Production
RonaldW.Jones
Universityof Rochester
I
The basic equilibriumrelationshipsin this two-commodity,two-factor
model of generalequilibriumare provided by the pair of full-employ-
1 In his unpublished paper, "Factor Market Distortions and the Pure
Theory of
International Trade," Magee also considers this aspect of the problem.
2 This aspect of the problem is discussed in Herberg and Kemp (1971), Magee (see
need no longerbe the case if distortions(in the labor market) are suf-
severe.
ficiently
Simple calculationreveals that IXI is given by:
X1X2 i aLl aL2
LK
-aKlaK2_
aK1 aK2
and
OL1W + OKlr = IN - OLlal, (7')
OL2W + OK2r = P2 - OL2 2* (8')
O -b ) =- (ol -A2) +
0 (&2- 1))
0| O ( 2) (9)
0K2 O
01a-2).(0
A
W P2ThTP1P2) Al _
where (11)
B = [XLl0Kl + XKl0L1]0L20f1 + [XL2OK2 + XK2OL2]0L10o2 and B > 0.
The conclusionsto be derived from(11) are straightforward. First,
note again that at constantcommoditypricesand distortions,the link
betweenendowmentchangesand outputchangesdependsas in standard
theoryonly upon the physical rankingof factorproportions.Second,
note that the effecteitherof price changes or of distortionchanges at
constantendowmentsdepends cruciallyon the sign of XI 01. In tradi-
tional theory,an increasein a commodity'srelativepricemust increase
the relativeoutput of that commodity.However,if the industrythat is
capital-intensive in termsof physicalproportionsnonethelesspays labor
a highershare of its proceeds than does the otherindustry,thistradi-
tional price-outputpattern is reversed.The reasoningis clear: an in-
crease in the relativepriceof the commoditythat is labor-intensivein a
value sense must raise the wage-rentratio. This serves to encourage
II
In this sectionI concentrateon the incomeredistribution effectsof an
increasein the premiumpaid to labor in one industryover that paid
in the other.Let the "unionized" sectorbe industry1. In the preceding
section,it was shown that if commodityprices are held constant,an
increasein a, mustraise the unionwage and depressthe returnto capital
if the unionizedsectoris capital-intensivein a "value" sense. If, on the
contrary,labor earns a higherdistributiveshare in the unionizedsector,
even if in physical termsit is capital-intensive,an increase in its pre-
miumearned over nonunionizedlabor must, at constantprices,reduce
union wages, raise rentals, and, as well, reduce nonunion wages. It
remainsto be shown whetherthis gain or loss in union wages is dis-
sipated, or perhaps reversed,by the subsequent change in commodity
pricesas marketsare cleared. I focusprincipallyon distinguishing those
cases in whichit would pay a labor unionto demand a higherpremium.'
Clearly,the linesof analysiswilldepend in part on whetherthe physi-
7 The criterionfor gain to the union is the change in the union real wage, w1/p1
and/or W1/P2-
P2 SD s s
/
a~~~~~~~
\b/ /
// /
0 XI
X2
FIG.1
where
Q1 XLlKl + XK1OL1,
Q2 XL26K2 + XK2OL2,
QD -X| 101,
and 2QQi1.
But B, definedin (11), is also expressiblein termsof the positivefrac-
tionsQ, and Q2, so that the inducedpricechange can be given the alter-
native form(13'):
)
+ - (
Al_
A L2Q1a-l OLlQ2a-2
shownby (13') into equations (9) and (10) forthe change in union real
wages to obtain (14) and (15):
OKi
( -1 i5) = {Q202 -
OL21 X I D}(1(a2) (14)
S
P1 D
P2
a?--
b~~~~~~
0
O ~~~~~~~~~X?
X2
FIG. 2
III
The principalquestionraised by Herbergand Kemp (1971) concernsthe
shape of the (distorted)transformation schedule.This turnsout to be a
difficultquestionto deal withcategorically,even in the highlyrestricted
case in which each production functionexhibits the same constant
elasticityofsubstitution.To open the discussion,I shall treatthegeneral
case and discuss,first,the relationshipbetweenthe price ratio and the
slope of the distortedtransformationschedule. A differentdistorted
transformation scheduleexistsforeach value of al,/a2. This impliesthat,
formovementsalong any given transformation schedule,da = &2 = 0.
Regardlessof the degreeof distortionin factormarkets,competitive
pressuresensure that the value of output, (p1X1 + p2X2), is equal to
the national income,(wjLl + w2L2+ rK). Differentiation of this equal-
ity leads to
(pjdXl + P2dX2)+ (Xidp1+ X2dp2)= (wldL, + w2dL2+ rdK)
+ (Lidw, + L2dw2+ Kdr).
X2
Xl
(W > W2)
FIG. 3
L?+ K (X 1- X2)
12
Since LI is aLlXl, dLl/dX, is just aLl { I?+ (dLl/Xl) Withdistortion
levelsand }.
factor endowments constant, both AL, and XI can be expressed in terms of ( -r).
This yields LK1[XjIa
a 1 -
K |
l
XK26L + XL26K
(XK2
OKiN A_
aK1-aL1J- w -r)
) (
- - + (aK2 aL2)
4L2 OL1
- ('1 + 02 - 1)( r-r)
substitutingfor(
Furthermore, r-f) yields:
(XK2Oi1) -6L + ) +
(Ji?J1 0- 1)(X1 - X2)
XL2 0L1 (L+ aK)
_dX2
;= IXl 1?l + -IX 1 -(ai/al)](al + oi-2 )t (Xi *) (20)
dXI I (1/oi)(a2/ai)? (1/CO2)1~ (aL ? 6K) ' (0
where
OK1XK2
OL1XL2 + OK1XK2
0101
K ~~~~~CapitalI
Labor
FIG. 4
FIG. 4
tract curve lies below the diagonal (regions[2] and [3]). For a relative
premiumpaid to labor in industry1 of an amount representedby the
ratio of the slopes of the two isoquants along the diagonal, IXI becomes
zero and the distortedtransformation schedule must be equivalent to
the chordjoiningits end points.For lowervalues of a2/al, IXI becomes
negativeand, since ar1+ 'r2 > 1 in the Cobb-Douglas case, the transfor-
mation schedule must be uniformlybowed in toward the origin.It is
forvalues of a2/ai greaterthan unity (region[2], in which,because a2/
to statewhat
al > 1, the unionizedsectoris industry2) that it is difficult
happensto theshape ofthetransformation schedulein the Cobb-Douglas
case. For sufficientlyhigh a2/al, the transformationcurve may be
bowed in locally and may change its curvaturealong the curve. More
preciseconditionsare given by Herbergand Kemp.
The crucialfeatureof the Cobb-Douglas case is the constancyof dis-
tributiveshares, so that X1101 can become negative only by having
XI change signs. In the general CES case, 01 may change signs as
well. Consider,3,as definedin (21):
OL1 OL2
(21)
OK1 OK2
( K1) OK2)
Now
(OLi)A
OKL) (7b-r + ('Li-aKi)=(-j(-) + (1- Oi)a
Therefore,
= (02 - ) )2)))2(
01 ~~~~Capital
Labor _
(I)~~~~~~~4
02
=2<
FIG. 5
Capitol
Labor
(2)
-
?N 2 > I 02
FIG. 6
IV
Althougha number of differentfacets of the problem presentedby
factormarketdistortionshave been discussedin the precedingsections,
the entireanalysis has been based on the standardtwo-factormodel in
which changes in the degree of distortionhave taken the formof an
exogenouslydeterminedwideningofthe gap betweenwhat a factorlabor
earns in the two sectorsof the economy.This approach may be appro-
priateforcertaintypesof distortions,forexample,whena factorin one
sectoris subject to a tax that is different
fromthat imposedon the same
factorin the othersector(Harberger1962). However,in thosesituations
in whicha factorin one sectorattemptsto improveits positionin real
terms,the primaryweapon used may involve restrictionon entryin-
stead of an attempt to increase the premiumearned over that of the
same factorin otherareas.
Ideally, a model to examine the distortionsproblemin this fashion
would be a three-factor model. In particular,it would be a simplified
model in whichonly two factorsenterinto the productionof any com-
modity.For example,union labor and capital are employedin the first
sectorand nonunionlabor and capital in the second. An increaseddis-
tortionon the part of unions would thus take the formof restricting
labor in the union sector (literallythrowingout labor into the open
marketin a nongrowingeconomy),assumingthis released labor finds
employmentin the nonunionizedsector.In a recentpaper (Jones1971),
I used such a model to explore issues involved in economic history,
capital theory,and trade theory.The virtueof such a model forthe dis-
Appendix
The expressionfor(-dX2ldX1) is givenbyequation(20) in thetext.Thetrans-
formationschedulemustbe bowedout iftheexpression in bracketsis positive.
iftheunionized
thatitwillbe positive
I wishtoestablish sectoris labor-intensive.
In such a case, IXI I0 l is positiveas, by assumption,is IXI (1 - a2/al). If the
unionis locatedin the firstsector,both XI and (1 - a2/al) are positive.
Therefore,theexpressionin bracesin (20) couldbe negativeonlyifal + r2is
smallenough.In moredetail,thisrequires:
1 + (I6I/ci) al
sidemustbe negativeandtheinequality
inwhichcase,ofcourse,theright-hand
To see this,notethat
notsatisfied.
OL2/c2 - I XI (6K1/XL2)
OK26/1 + I X I (6L1/XK2)
References
Bhagwati, J. N., and Srinivasan,T. N. "The Theory of Wage Differentials:
Production Response and Factor Price Equalization." J. Internat.Econ.
(1971): in press.
Harberger,A. C. "The Incidence of the CorporationIncome Tax." J.P.E. 70
(June 1962): 215-40.
Herberg,H., and Kemp, M. C. "Factor Market Distortions,the Shape of the
Locus of CompetitiveOutputs,and the Relation betweenProduct Prices and
Equilibrium Outputs." In Trade, Paymentsand Welfare,papers in interna-
tional economicsin honorof Charles P. Kindleberger,edited by J. N. Bhag-
wati,R. W. Jones,R. A. Mundell,and J. Vanek. Amsterdam:North-Holland,
1971, in press.
Johnson,H. G. "Factor Market Distortionsand the Shape of the Transforma-
tion Curve." Econometrica34 (July1966): 686-98.
Johnson,H. G., and Mieszkowski,P. M. "The Effectsof Unionizationon the