115

CHAPTER 5 Market Demand and Elasticities

h tO demand estimation,
examines some new revolutionary marketing researc~ approac es d b regression
while the optional appendix sQ.ows how demand is estimated and forecaSle Y
analysis.

aDL--~T~H=---E~M~AR~K~ET~D~E~M~AN~D~F~O~R~A~(~O~MM~O~D~ITY~--------
· d"t · derived
In this section, we examine how the market demand curve for a commo 1 Y is
from the individuals' demand curves. The market demand curve for a commodity is
simp y the horizontal summation of thedemand curves of all the consumers in the mar-
, ~T us, the market guantity demanded at each price is the sum of the ind1v1dual quan-
tities demanded at that price. For example, in the top of Figure 5.1, the market demand

Individual 2 Market
Individual 1
Px($) Px($) Px($)

2.00 --- - ----------- 2.00 ---------- 2.00

;
i 1------- -----------
..c:
'o 1.00 1.00 f---,-'-+-11~- - - C.. - - - - . 1.00
Q)
u
·c::
p.. 0.50 ------- 0.50 ------ 0.50
, ·•d1 <½

0 2,; 4 . 6 . Qx 0 4 10 16 Qx
0 , 2 -6 10 Qx
Hamburgers. (X) per unit of time

Individual ~.., ._ ~farket

Px($)
Px($)
2.00
"'Q) 2.00
I-,

bO
I-,
::l
s
..0
csj
..c: ' 1.00 .
4-<
0 1.00
Q)
u -
·c::
p.. 0.50
0.50
Dx

2 6 10 Qx 0 2 6 10 Qx
0
Number of hamburgers Millions of hamburgers

FIGURE 5.1 From Individual to Market Demand· The top part of the figure shows that the market
demand curve for hamburgers, 0, is ·obtained from the horizontal summation of the demand curve for
l 1 (d1) an_d individual 2 (d2). The bottom part of the figure shows an individual's
hamburgers of individua_
demand curve, dx, and the market demand curve, Dx, on the assumption that there are 1 million individuals
in the market with demand curves for hamburgers identical to dx. ·
 ,f

-5.2 PRICE ELASTICITY OF MARKET_ DEMAND

In this section, we show how to measure the price elasticity of demand, both algebraically
and graphically. We also examine the impon:ant relationship between the price elasti~
pf dem~<.L~nd.the total...expenditures_of_cQnsumers o~-1:E.~ commodity. That is, when the
price ofJ_ commodity changes, will consumers' expenditures on the commodity i~e,
dec{ptSe, or remain unchang€d? Finally, we examine the determinants or the factors that
affect the value of the price··elasticity of demand. ·

Measuring the Price Elasticity of Demand 3

The price elasticity of demand measures the responsiveness in the quantity dem~ded_
of a commodit)'. to a change in its price. This could b~ measured by the inverse of the
slope of the demand c·urve(fe., by /1 QTl1P). 4 The disadvantage is that the inverseol

3
For a discussion of the price elasticity of demand using calculus, see Section A.5 of the Mathematical
Appendix at the end of the book. (hr
4 • h . IOt price on
Smee t e turn of the century, the convention in economics (started by Alfred Marshall) 1s to P • Id I'<
· I axis
vert1ca · an d quantity · pnce cou
· on the horizontal axis. Therefore, the quantity response to a change 10
measured by ~QI L\ P, which is the inverse of the slope of the demand curve.
 CHAPTER 5 Market Elemand and Elasticities 119

the
. slope is expressed in terms of the units . of measurement. A change of 100,000 umts .
~n the ~uantity demanded of a commodity is very large if the commp,cjity is new hous-
ing
h umts ' but it is not very 1arge 1·f the commodity · 1s· hamburgers.~,.: ._:_.l ar Iy, a pnce
e!,!!!l ·
c ange. of one dollar is insignificant for houses, but very large for hamburg~hus,
measunng
. the
. responsive ness m· t he quantity
. demanded of a commodity . to a change m .
pnce b~ the inverse of the slope of the demand curve is not very useful. Furthermore,
companson of changes in quantity to changes in price across commodities is meaning-
~ess. Th~se problems can be resolved by using percenta£_e rather than absolute changes
m quantity and prices. --~ - ,,,.,,.__
!n order to have a measure of the responsiveness in the quantity demanded of a com-
modity to a change in its price that is independent of the units of measurement, Alfred
Marshal_l, the great English economist of the beginning of the twentieth century, refined and
populanzed the conc;~pt of the price elasticity. of demand. This measure is defined in
term~ ?f relative or percentage changes in quantity demanded and price. As such, price
elasticity of demand i!.; pure number {i.e., it has no units attached to it), and its value is
not affected by changes m the units of measurement. This also allows meaningful com-
parisons in the price' elasticity of demand of different commodities. ,,,-
The price_-el?Sticity !Jf demand is given by the percentage change in the quantity
demanded of a commodity divided by the percentage change in its price. Letting T/ (the
Greek letter eta/ stand for the coefficient of pri·ce elasticity of demand, tl Q for the change
/ ~ in quantity demanded, and t. P for the change in price, we have the formula for the price
~ -?, . j elasticity of demand:
1/~/4;· T/ =
tlQ/Q
tlP IP
· D.Q P
= .tlP . Q ,.,
[5.1]

/ Q
' t::. Since quantity and pricf; move in opposite' _girections 1 the value of T/ is negative. To com-
pare price elasticities, howe.ver, we use the_ir absolute value.(i.e., their value without the
negative sign) . .Thus, we say that a. demand ,curve with a price elasticity of -2 is more
elastic than a demand curve with a price elasticity of -1 (e.ven \hough -2 is algebraically
smaller than -1). Note that the inverse of the slope of the demand curve (i.e., tl QI tlP)
is a component, but only a component, of the price elasticity formula.
Formula [5 .1] measures point elasticity of demand or the elastici~_at_a eartic,Elar _
the demand curve. More frequently, we are interested in the price el_a~ticity
between two points on the demand curve. We then calculate the arc elasbc1ty of
demand. If we. used formula [5.1] to measure arc elasticity, however, we would get dif-
5
ferent results depending on whether the price rises or falls. To avoid this, we use the
average of the two prices and the average of the two quantities in the calculations.
Letting p refer to ,the lu_gher of the two prices (with Q1 the quantity at P1) and P 2 refer
1
to the lower of the two prices (with Q2 the corresponding quantity), we have the for-
6
mula for arc elasticity of demand :

tlQ (P1+P2)/2 llQ (P1+P2) [5.2]

17 = tlP . (Qi+ Q2)/2 = tlP . (Q, + Q 2)

5 As we will see below, this results because a different base is used in calculating percentage changes for a
price increase than for a price decrease. _ _ •
6For the second ratio in the formula , we could use P / Q, where the bar on P and Q refers to their average

value.
 120-
PART TWO Theory of E:onsumer Behavior and Demand

· ·tYfor a price decline from point

Using formula [5. l] to measure theelasttct . B (Q ,,._
0 50) on the market demand curve in p·1 " 2
P = $2.00) to point G (Q = 10, P = $ · &Ure s. ·
1
I

, ,<) ,.I' () 8 2 __ 16 = -5.33

-'- '? 17 = (- 1.50) (2) - 3
-l / 7 o' .
3//'/1 On the other hand, measuring elasticity for a price increase from pomt G to point Bon the
~.. <&1 i' 7<_O, same demand curve, we get
v - I 0, 7\. s a.so _ -~ = _0 _27
'y 17 = - (1.50) (10) - 15

Using formula [5.2] for arc elasticity, we get

. 8 2.50 __ 20 _ -1 11 /
17 = - (1.50) (12) - 18 - .

The price elasticity of dem~nd is usually different at and between di_ffer~nt points on
the demand curve, and it can range anywhere.from zero to very large or mfimte. Demand
is said to b elastic if the absolute value of 1J is larger thanJ, f:..':Jtq,_,y ejas~ic if the absolute
value of 17 equ- s 1, and inelas_tk;ji the a6'solute value of 1J is smaller than ,

~ c e Elasticity Graphically -
We can also measure graphically the price elasticity at any point on a linear or nonlinear
demand curve. To measure the price elasticity at point£ on Dxin the left panel ofFigure5.2
(the same as in the right bottom panel 'in Figure 5. It we proceed as follows. We draw
tangentAEH to point Eon Dx and drop perpendicular EJ to the quantity axis. The slope of
tangent line AEH is negative and constant throughout and can be measured by
t:iP JE
!:iQ = - JH
The first component of the price elasticity fonnula is the inverse of the slope of the
demand curve or .

!:iQ JH
-=--
t:iP JE
The second component of the price elasticity fonnula is

p JE
Q = OJ
Reassembling the two components of the el,asticity fonnula, we have

TJ == !:iQ . P = _JH . JE _ JH · 6
t:iP Q JE OJ - - OJ = -6 = -1
That is, the price elasticity of D . .
Since EJH, AKE and AOH x_at _pomt Em the left panel of Figure 5.2 is equal to 1.
' are surular triangles (see the left panel of Figure 5.2), tbe
 \', \~ 6 . 121
CHAPTER 5 Market Demand and.Elasticities
L\ 7 \
\
Px($) Px ($)
A
"'... 2.00
II) 2.00
V
...;:lbD e,o
;:l
.a 1.50
E
.c
c....,
0
1.00 'o 1.00
(.)
u V
u
·c 0.50 ·c:
Cl,.
c... 0.50

H
0 2 6 10 12 Qx 0 3 6 9 12 Qx
Millions of hamburgers Millions of hamburgers
FIGURE 5.2 Measurement of Price Elasticity of Demand Graphically In the left panel, the price elasticity at
point Eon Dx is measured by drawing tangent AEH to point Eon Dx and dropping perpendicular EJ to the horizontal
axis. At point E. TJ = -JH/OJ = -6/6 = - lln the right panel, the absolute value of T/ = 1 at point E (the midpoint
of D'x), TJ > 1 above the midpoint and T/ < 1 below the midpoint

price elasticity of Dx at point E can be measured by any of the following ratios of

distances: ·
JH KO EH
1} = -- = -- = -- [5.3]
OJ AK AE
The price elasticity of demand at any other point on Dx can be found in a similar way by
drawing a tangent to Dx at that point and then proceeding as indicated above (see Problem
2). This provides a convenient and easy way to measure the price elasticity of demand at ·
any point on a nonlinear demand curve.
The same procedure can be used to measure the price elasticity at any point on a
straight-line demanci,curve. For example, by inspecting the right panel of Figure 5.2, we
can find that 1J = -9/3 = -3 at point Con Dx, 1J = -3/9 = -1/3 at point F, and
17 = -6/6 = -1 at point E (the midpoint of Dx). Furthermore, 1J oo at point A and
17 = 0 at point H (see Problem 3). Thus, while the slope of a straight-line demand curve is
constant throughout, its price elasticity varies between each point on (and its absolute
value declines as we move down) the demand curve. As a general rule, a straight-line
demand curve is unitary elastic at its geometric midpoint, price elastic above its midpoint,
and inelastic below its midpoint (see the right panel of Figure 5.2).
Two other simple rules are useful in considering the price elasticity of demand. Toe first
is that of two parallel demand curves (linear or nonlinear). the one further to the right has a
smaller price elasticity at eac~ pric~ (s~ blem 4(a), with answer at the end of the text) .
. Second, when two demand curves intersect, the flatter of the two is more price elastic at the
point of intersection (see Problem 4(b), with the answer also provided at the end of the text).

~ sticity and Total Expenditures

An important relationship exists between the pric: ~ asticity of deman~ a!1d the total
expenditures of consumers .0n the commodity. This relationship is often used in econom-
- ics.1t p~ that a decline in the commodity price results in an increase in total
expenditures if demand is elastic, leaves total expenditure~ unchanged if demand is uni-

- --- -
tary elastic, and results in a decline, in total expenditures if demand .is inelastic.
122 PARTTWO Theory of Consumer Behavior and Demand

5t
TABLE 5.1 Total xpenditures and Price Ela icity of D~
Total Expenditures Absolute Val
Px Ox f lie
(million S) 0 1/
(S) (million units)
Point
2.00 0
0
------
00
3

G
4.5
3 /
1.50 6.0 l
1.00 6
4.5 1/3
0.50 9
F 0 0
0 12 ..,,
H

Specifically, ':hen the ~rice o~ a commodity falls, total_ expendi~res (p~ce times
quantity) increase _if demand 1s elasuc be~ause the percentage 1q~rease m gu~tity (Which
by itself tends to increase total expenditures). exceeds
) 'T' the . percentage
d' decline
. in pnee
.
(which by itself tends to reduce total expen 1t~res . 1.ota expen 1tures .are maxun
d 1
when Ir, I = 1 and decline thereafter. That is, when IT/ I < 1, 'a reduction 'in the commoct~lll
1
price leads to a percentage increase in the quantity demanded of the commodity that ~
smaller than the percentage reduction in price, and so total ~xp~nditures on the comma~~
ity decline. This is shown in Table 5.1, which refers to Dim Figure 5.2.
From Table 5.1 we see that between points A and E, IT/I > 1 and total expenditur
. · d 1· Th ' es
on the .commodity increase as the commo d1ty pnce ec mes. e opposite is tru
between_points E and F over which IT/ I < 1. Total expenditures are maximum at poin~
E (th~ geo~etric midpoint of Dx in Figure 5.2_). T~e general rule summarizing the rela-
tionship among total expenditures: I?rice,.and the price elasticity of demand is thafiotal
expenditures and price move in opposite directions if demand is elastic and in the ~e
direction if demand is inelastl~lsee Table 5.1). · ·
Figure 5.3 shows a dertuind curve that is unitary elastl.c throughout. Thus, ry ==
-JH /JO= -6/6 == -1 at point Eon D*, 71 = -LJ/OL = -3/3 = -1 at point B', and
T/ = -HN/OH = -12/12 = -1 at point G'. Note that total expenditures (price times
quantity) are constant ($6 million) at every point on D*. This type of demand curve is a
rectangular hyperbola. Its general equation is
Q=cp [5.4] '

where Q is the quantity demanded, Pis its price, and C is a constant (total expenditures). ·
Thus, P · Q = C. For example, at point B', (P)(Q) = ($2)(3) = $6. At point£, ($1)(6)==
/ d at point G', ($0.50)(12) = $6 also. / . .

J what Determines Price Elasticity?

Because the price elasticity of demand is so useful (i.e., it tells us, among otl\er things,
happens to the level of total expenditures on the commodity when its price changes),
~tis important to identify the forces that determine it4alue. The size of the price elaSric·
1ty ,of demand
. depends pnman · ~ lYon two factors
· ';l'/ and
. Hifst · · o'1
y" foremost, the price elasticity
demand . for a commod'ty · z ..,,.
1 is arger the closer and the greater are the number of avai
·[able
substitutes · For examP1e, the demand for coffee is more elastic than the demand ,ior salt
because
ercentagcoffee
. has b_e tter and more substitutes (tea and cocoa) than salt. Thus, the saJlle.0
P e mcrease m the · f d uon 1
the quantity dem d d pnce O coffee and salt elicits a larger percentage re uc
an e of coffee than of salt. .
 CHAPTER 5) ? nd . 123

4 \11""-~
\e," ,,-.
1'i~f
n>f
't/\J-1'
/' 1~ / '
/) ' } , ~ ,d' \
2 . r
,,.,,,, ~ ·
7
/ ' " ~er-
1 /I v,r".\? ;r-
A y-;P~-(V\ \-.
,,., ,. ~' v1' ·• :,.
0.5 - ~ - - - -D* , ;P /
/'
.,.
71
A
'6 .
y
L H N ; 1 0..J/
0 3 6 12 24 Q /J ~ ~
Quantity (million units) 7 \_? v ·~
FIGURE 5.3 Unitary Elastic Demand C~rve · Demand curve o• has unitary elasticity throughout. ~/l ,,.~-r-
Thus, 11 = -JH/QJ= -6/6=:;:;. -1 atpointf,1J= -LJ/OL= --3/3= -1 atpointB',and ,,,-J,._<~t_
11 = -HN /OH= - 12 / 12 = - 1 at point G'. Total expenditures (P . Q) are the same ($6 million) -
at every point on D*. This demand curve is a rectangular hyperbola.

In general, a commodity has closer substitutes and thus a higher price eJasticity of
demand the more narrowly the commodity is defined. For example, the price elasticity for
Marlboro cigarettes is much larger than for cigarettes in general, and still larger than for
all tobacco products. If a commodity is defined so that it has perfect substitutes, its price
elasticity of demand is infinite. For example, if a wheat farmer attempted to increase his
or her price above the market price, the farmer would lose all sales as ·buyers would
switch ~ heir wheat purchases to other farmers (who produce identical wheat).
Jl(ond, price elasticity is larger, the longer -is the period of time allowed for con-
sumers to adjust to a change in the commodity price. The reason for this is that it usually
takes time for consumers to lt:~~o~~ E-~e f h~p_g~.and to fl!.!!Y.J_c:!.S.pond.or adjust their Q~-;
chases. For example, consumers may not be able to reduce much the quantity demanded of
electricity soon after learning of an increase in the price of electricity. Over a period of sev-
eral years, however, households can replace electric heaters with gas heaters, purchase
appliances that consume less electricity, and so on. Thus, for a given price change, the
quantity response per unit of time is usually much greater in the long run than in the short
run, anJi so the absolute value of T/ is larger in the former than in the latter time period. This
7
7's'clearly shown in Example 5-2.

7 Sometimes it is 3tated that the price elasticity of demand is larger the greater is the number of uses of the
commodity. However, no satisractory reason has been advanced as to why this should be so. It is also
sometimes said that price elasticity is lower the smaller is the importance of tl)e commodity in consumers'
budgets (i.e., the smaller is the proportion of the consumers' incom~s spent on the commodity). However,
empirical estimates often contradict this.
 CHAPTER 5 Market Demand and Elasticities 125

-~ Y ;;r'?'T':::r-'--... ~ ~ ~ . ,- ~ " T " " ¥~

...,.
.. h"1"r-"' •
- • j 'IJ. t -
· ·' ,>i:,·Mall.y economic , li ~- ( '' ' ·
, ).. .. . . , P? cies such as redubing American dependence on imported
1
petr? feubm re$ly _cru~uilly on price elasticities. ·For examr1le· with.the U.S. price of gaso-
line) o a$Iout 3·50pet gail··on .·m 2007 and a short-run price P '
elasticity of 0.25• (see Table
$4
5 2 a ·00 al;ax
per gallon would increase the price of gasoline from about $3.50 to
:z~:. "-
_ 50%
3 [r-~;(;;;)•
7: . • rJ ~.
~r by about}9%, and reduce the_ quantity demanded ~f gasoline. b_y
-0. 25 )(0.~9) =-7 .3% =%~Q] m the short run. With the price elas:
-
· ... t:1c1ty equa
. tft . 'l. . 1,to 0.92 m ·the Iong .run, the red.uctton
. -m. .the quantity
· , ' . •.demanded of gasoline : ,..
IL· __ ,m. e on~ 1:1n .would be abop! 27%. In fact, there is evidence (Hughes 2006) that'"tii~ " ·
. . pnce . elasticity
. of
_ demand fior gasoline m • the Umted
. ·States· · has sharply 'fallen smce
. the ',
,'.J;_.,_. , begtnntng of the decade Th1·s " 'ould
_ .-~·~,:~"" •• .:.:s::"o~··•: _ • • . • ,w_ 1 · h 11 ·
resu t m a µme sma er response m the quantity_ ·
. -~ - · __ _ ;~::__'°.~gasoline demanded m the Uruted States as a result of a $1.00 tax on gasoline. '

2 E E~TICITY OF DEMANDS 4 .
In Section 4.1 we defined the Engel curve as showing the amount of a commodity that a
consumer would purchase per unit of time at various income levels, while holding prices
and tastes constaw:. We can measure the responsiveness or sensitivity in the quanllty
demanded of a commodity at any point on the Engel curve by the income elasticity of
demand. This is defined as .

..p . D,.Q/Q D,.QI 55

•. / = = M . Q [ - ]

~p~e
M /I

\ .· 11Q is the change in the quantity purchased, 1:),,/ is the change in income, Q is the
1 \ original quanti~y. and / is the original money income of the consumer.
l, .JI' A commodity is normal if rJ 1 is positive and inferior if rJ 1 is negative. A normal good can
//t be further classified as a necessity if rJ 1 is less than 1 and as a luxury if rJ 1 is greater than 1. In
the real world, most broadly defined commodities such as food, clothing, housing, health care,
education, and recreation are normal goods. Inferior goods are usually narrowly defined inex-
pensive goods, such as bologna, for which good substitutes are available. Among normal
goods, food and clothing are necessities while education and recreation are luxuries.
This classification of goods into inferior and normal, and necessity and luxury, cannot
be taken too seriously, however, because the same commodity can be regarded as a luxury
by some individuals or at some income levels, and as a necessity or even as an inferior
good by other individuals or at other income simple geometric method can
ermine if a commodity is a luxury, a necessity, or an inferior good at each income level.
the tangent to the Engel curve is positively sloped and crosses the income axis, r, 1
eeds 1 and the good is a luxury at that income level. If the tangent crosses the origin,

711
=
1. If the tangent crosses tfieliorizontal axis, 7/'i is less than 1 and the commodity is a
necessity at that income level., Finally, -4.,the tangent to the Engel curve is negatively
~ oped, thec ommodity is an inf~~O
_!"j.00~

8 For a discussion of the income elasticity of demand using calculus, see Section A.5 of the Mathematical
Appendix at the end of the book. .
9 Indeed, some economists feel that the necessity-luxury classification of goods is entirely spurious and

meaningless.
126
PART TWO Theory of Consumer Behavior and Demand

For example, Table 5.3 and Figure 5.4 show that the student in Chapters an
3 dd 4
would regard hamburgers as a luxury at income levels (allowances) of up to $15 Per
Hamburgers would become a necessity for daily allowances of between $15 and $)Q, ay_
would be regarded as an inferior good at higher incomes (where the student couict aff~nd
steaks and lobsters). td

TABLE 5.3 ncarr,i~ ~Efa 'ia~1ficatio·~ of Hamburgers (X)

at Vaneu$,,: . :AUowances _
I Ox %L1Qx %L1/ T/1 Classification
10
15
20
2
4
5
100
25
50
33
2.00
0.76
-------
Luxury
Necessity
30 6 20 50 0.40 Necessity
40 4 - 33 33 - .1.00 Inferior

/ ($)

40 Engel
curve

,....._ 30 T' ,....._ 30

~ -
"C "C ,,
..."'
'-..
..."'
'-..
J;l J;l
0 0
"C
.._., "C
.._.,
20
s 20
II)

s
II)

0 0

- -
u u
r:: 15 r:: 15

IO 10

0 2 4 5 6 Qx 0 I 2 4 5 6 Qx
Number of hamburgers (X) Number of hamburgers (X)
FIGURE 5.4 Engel Curve and Income Elasticity Because the tangent to the
Engel curve is positively sloped and crosses the income axis up to the daily income
allowance of $15, hamburgers are a luxury for this individual. The tangent goes
through the origin at/= $15 and T/t = 1 at that income level. Since the tangent is
positively sloped and crosses the quantity axis from/= $15 to $30, hamburgers are
a necessity between these income levels. For/ higher than $30, the Engel curve is
negatively sloped and hamburgers become an inferior good for this individual.
 CHAPTER 5 Market Demand and Elasticities 127

The concept and measurement of the income elasticity of demand and Engel curve
can refer to a single customer or to the entire market. When referring to the entire market,
Q and Q are the total or tqe market quantity purchased and its change, while / and IJ,,./
~e the total or aggregate money income of all consumers in the market and its change. 10
As pointed out in Section 4.1, the proportion of total expenditures on food declines
as family incomes rise. This is referred to as Engel's law. Indeed, the higher the propor-
tion of income spent on food, the poorer a family or nation is taken to be. For ·example,
in the United States less than 20% of total family incomes is spent on food as compared
with over 50% for India (a much poorer nation). As Example 5-3 shows, the income elas-
ticity of demand can be very different for different products. 4'
 CROSS ELAS-TICITY OF DEMAND 11

We have seen in Section 2.2 that one of the things held constant in drawing the
demand curve for a COilll?J.odity is the price of s~bstit11te and c~ plementary co ll1ark~t
ties. Commodities X and Y are substitutes if m"oi-e otx' is purchase~ he· Pn~odi-
goes up. For example, consumers usually purchase more ceffee when the priceCeofOf y
- .
rises. Thus, coffee and te_a ?re _sub~titutes. (?ther. examples of substitutes include bu~:
and m~r1e, hamburgers and hot dogs, Coca-Cola and
so on. /
P~f
si, electri~ity and ga~ and

On the other hand, comm?dities X and Y are complements if less of Xis purchasect
when the_price of Y goes up. For example, consumers usually purchase fewer lemons
when the price of tea goes up. Thus, le~ons and tea are complements. Other examples of
commodities that are complements are coffee and cream, han:iburgers and b~ns, hot dogs
and mustard, cars and gasoline, and so on. __/
An increase in the price of a commodity le&ds to a reduction in the quantity
demanded of the commodity (a movement along the dema~cj curve for the commodity)
but causes the demand curve for a substitute to shift to the right and the demand curve for
a complement to shift to the ieft. For ·example, a~ incre;~e 1n the price of tea will cause
the demand for coffee (a substitute of tea) to shift. to the right (so that more coffee is
demanded at each coffee price) and the demand for lemons (a complement of tea) to shift
to the left (so that fewer lemons are demanded at each lemon price). ,,--
We can measure the .!",~§R9JlSiv.eness.. or_~~~s~ in the quantity purchased of com-
modity X as a result of a change in the price of commodity Y by the cross elasticity of
demand ('Y/xy). This is given by:
. :;;;a-~
----F-
!:::,. QXI QX ! ti QX Py l\ /
'i

rJxy = ti Py/ Py = \ t::.Py · Q::..t ' [5.6]

.
where ti Q x is the change in the quantity purchased of X, ti Py is the change in the pnce
of Y, Py is the original price of Y, and Qx is the original quantity of X. Note that in mea-
suring 1JXY, we hold constant Px, consumers' incomes, their tastes, and the number of
consumers in the market. ·
'1!. is greater than zero: X and y are substitutes because an increase in p~ leads ~o
1JXY
an increase in Qx as Xis substituted for Yin consumption. On the other hand, 1f 1JXY !5
less than zero, X and Y are complements because an increase in Py leads to a reduction 111
(Qy and)'Qx. The absolute value (i.e., the value without the sign) of the cross elasticity of
/

11
For a discussion of the cross elasticity of demand using calculus, see Section A.5 of the Mathematical
Appendix at the end of the book. · '
 CHAPTER 5 Market Demand and Elasticities 129

demand measures the degree of substitution or complementarity. For example, if T/XY

between coffee and tea (s found to be.larger than that betw~t;m coffee and hot chocolate,
this means that coffee ·and tea are better substitutes than coffee and hot chocolate. If T/xr
is close to zero, X and Y are independent commodities. This may be the case with cars and
pencils, telephones and chewing gum, pocket calculators and beer, and so on.
Several additional things must be kept in mind with respect to the cross elasticity of
demand. First, the value of /Jxr need not equal the value of T/YX because the responsive-
ness of Qx to a change in Pr need not equal the responsiveness of Qr to a change in Px.
For example, a change in the price of coffee is likely to have a greater effect on the quan-
tity of sugar (a complement of coffee) demanded than the other way around, since coffee
is the more important of the two in terms of total expenditures. .
Second, a high positive cross elasticity of demand is often used to define an industry
since it indicates that the various commodities are very similar. For example, the cross
elasti~ity of de~anq oetween Chevrolets and Oldsmobiles is very high, and so they
belong to the same (auto) ind~-stry. This can lead to some difficulty, however. For exam-
ple, how high must the positiv:e cross elastic_ity between two commodities be for them to
be in the same industry? Also, if the cross elasticity between cars and station wagons and
between station wagons and trucks is "high," but the cross elasticity of demand between
cars and trucks is "low," are cars and trucks in the same industry? In these cases the def-
inition of the industry usually depends on the problem to be studied . .,. .,. -
Third, the above defi~ition ef substitutes and_complements is sometimes referred to
as a "gross" definition; as such, it refers to.the entire market-response and reflects both the
incom_J\.and the substitution effects. For an individual consumer, there is a more rigorous
12
definition (in tenns-~r'th;substitution effect only) discussed in more advanced treaties.
Example 5-4 gives the estimated gross elasticity of demand between a number of prod-
ucts and shows its usefulness in the analysis of important economic issues. _.,,,.--

. ·'EXAMPLE 5-4
Margarine·and Butter Are Substitutes, Entertainment and Food Are Complements
. ¥1' ' -

. The first row of Table 55 sho,w s ~ha;..tl1e,. c~pss .ela~ticity of,c,lepiand of margari,ne with
respect t~ the ~rice· ofb~tter'k r'.?3o/f This rti~~~s-~at _~ 1~ }ncrease in the ~ri~_e ~f
butter leads tg a 1.5~% mcr~ase m the demand for margarine. Thus, rnarganne and
0

.,.. · butter are substitutes in the United-States. On the other hand, row 6 of Table 5.5· sh~ws
tha~ the:cross•efasticity o{demanctof entertainment with respect to food is -0.72. This
means that •,i' .J% . ,ii;icrease· in· the :price ·of -entertainment leads to a reduction in the
, demand for'ifood by' 0.72%.<I1hus, en~ertainhterit and food are complements in the
United States. The :table ·also,shows. the cross elasticity. of demand of other selected
;commodities: ., i'
. Cross-price elasticities"of demand have imponant ,economic applications-even
• t inf the' c:ourtroom;1:as:,th~ 1eeleotated Cellophane Case shows (see U.S. Reports, Vol.
., · '35l/ Waihittgt6,d;50@:,.~1~. :9-ov·e mment Printing Office,• 1956, p. 400). In that case,
the court decided that puPoht had not _monopolized: the market for cellophane even
1

"ltAJ ~,n t.t1'k•tttiti :~JJ.i:! 1uJ!f•h~ ,Jt!- 'i~d':i.l£;:·'f!liftitiii,V~tl_;:';J?ri:~·,/tl-!iJ.f.?27~! nl'fo)'i~':-. ,··;"(\\~}'· "1
1
~··.

12 See J. R. Hicks, Value and Capital (New York: Oxford University Press, 1946), p. 44.