3

~!.
0

HA'PTER
-~---
jiJ' .
.··

Consymer Prefere·nces and Choice

Chapter Outline- 3-2 Gillette Introduces the Sensor and Mach3
3.1 Utility Analysis- Razors-Two Truly Global Products
3.2 Consumer's Tastes: Indifference Curves 3-3 Time as a Constraint
3.3 International Convergence of Tastes 3-4 Utility Maximization and Government Warnings
3.4 The Consumer's Income and Price Constraints: on Junk Food
'.fhe Budget Line 3-5 Water Rati_oning in the West
3.5 Consumer's Choice At the Frontier: The Theory of Revealed
Preference
List of Examples
3-1 Does Money Buy Happiness?

.r After Studying This Chapter, You Should Be Able to:

• Know how consumer tastes are measured or represented
f - I , ,

t •. Descri~ the relationship between money and happiness

:;~ 1{now how the consiimer's·constraints ate represented
J}. 1 Understand how the consumer r:iajcunizes satisfaction or reaches equilibrium
: • · Des~ribe how (:Pnsumer tas'tes or prefer,ences
1-:" --- - :!. ,.,.. '•' .,,,. ., • ,. ; , •.
c~ _b~ 1nf~rr~o·witho~t:.J,askir{g~tlt_e ~<;msumer
. . ... ..,.~· -~ 1( . 1.., '"- ...;;, -, -

n this chapter, we begin the formal study of microeconomics by examining the eco-

I nomic behavior of the c umer. A consumer 1s an individual or a household composed

· o one or more individuals. The consumer is the basic economic unit that determines
,!!rish,.commodities are purchased and_ in what guantities. Millions ~f such decisions are
made each day on the more than $13tn1Iion worth of goods and services produced by the
American economy each year. ,..
What guides these individual consumer decisions? .Why do consumers purchase
some commodities and not others? How do they decide how much to purchase of each
commodity? What is the aim of a rational consumer in spending income? These are some
of the important questions to which we seek answers in this chapter. The theory of con-
sumer behavior and choice is the first step in the derivation of the market demand curve,
the importance of which was clearly demonstrated in Chapter 2. ,,.
We begin the study of the economic behavior of the consumer by examining tastes.
Consumers' tastes can be related, to utility concepts gr indifference curves. These are

51
 of Consumer Behavior and Deman~
52 PART TWO Theory

. f the chapter. In Section 3.3, we ex~n~ the conver.

discussed in the first two sections oth . troduce the budget line, which gives the con.
· ti nally We en m · C tr ·
gence of tastes mtema O •, . basing goods and services. ons runts arise
strain.ts or limitations consumer s face m pure ants command a price in the marketplace
·t· ·· that the consumer w d Ii
because the commodi ies , h 1. •ted income. Thus, the bu get ne reflects
f ) · d the consumer as um ... • • -
(i.e., t:,bey are not ree · . f scarcity as it pertains to die md1v1cfual co~-
the familiar and pervasive economIC fact o

sumer. , ts are unlimited or, in any event, exceed his or her

Because_ the ~onsume_r ~an t that the consumer spend income so as to.max.
ability to ·satisfy them all, it is Idmpl~rtan 'd d to illustrate and predict how a rational
. . t' f tion Thus a mo e is prov1 e
1ID1ze sa 1s ac · • . . . r her tastes (indifference curves) and the
consumer maximizes sati~f~~tf1on, g(1tvhenbhi;g~i'1ine) The "At·the Frontie'r..-secf\6 n; pr~-
constraints that the consumer aces e. u . . - . . ' · ·ndiffi·e ,:
· · · consumer
sents a different way to examine · ·- tas tes and derive a consumer s 1 rence
/
curyes_. .. al al , . ld amples and important applications presented in;the'clfapter
Toe sever re -wor ex .• .d 11. •
•
demonstrate th.e re1evance an d use fulness of the theory of consumer. behavior
. an cuo1ce.

UTILITY ANALYSIS

In this section we discuss the meaning of utility, distinguish between total utility and
marginal utilicy, and examine the important difference between cardinal 'and ordinal ~til-
ity. Toe concept of utility is used here to introduce the consumer's tast~s-?'he an~ys1~ of
consumer tastes is a crucial step in determining how a consumer max1rmzes satisfaction
in spending income.

Total and Marginal Utility

I •..

to
Goods are desired because of their ~bility, satisfy hum.an ·\-\'.ants. Th_e .property of a g~
that enables it¼! sati~cy-human-\\lants is called utility. As individu~s consume more of a
gooC,. per_time period, .their .total utility (TU) or satisfaction increases, but their marginal
utility diminishes. Marginal utility (MU)"is the extra utility received from''d:msuming
one additional unit of the good per unit of time while holding constant the quantity con-
sumed of all other commodities.
For example, Table 3:1 indicates.tha,.t one hamburger per day (or, more generally, one
unit of goodX per period of time) gives the consum~r a-total ut,ility (TU) of 10 utils, where
a util is an.arbitrary unit of utility. Total utility £increases with each additional hamburger
consumed until the fifth one,,.which leaves-tptal ~tility unchanged. This is the saturation
point. Consuming the sixth hamburger- then. leads to a -decline in total utilitv because of
1
storage or disposal problems. The_third colull)Il 9f T&bl~ 3.1 gives the extr~ or marginal
utility resulting ~om th~ consumption of eaeh additional hamburger. Marginal utility is
positive buLdeclines until the,fifth ham,burger, for whi.c hj t is zero, and becomes negative
for the sixth hamburger.

1 That is, some effort (dis utility), no matter·how small,.is required to get rid of the sixth hamburger. Assuming
that the individual cannot sell the sixth hamburger, he or she would not want it even for free.
 CHAPTER 3 Consumer Preferences and Choice 53

Ox TUx MUx
0 0
1 10 10
2 16 6
3 20 4
4 22 2
5 22 0
6 20 2

TUx

22
~o

.: o 1 2 3 4 5 6 Qx
Quantity of X

MUx

:>< '
.. ..... J 2
_, •.;._
0
c 10
;.::::
·;:i 8
:::3
FIGURE 6
utir 31· ~otal and Marginal · Utility In the
. top panel,
·
total
shade ity (TU) increases by smaller and smaller acnounJ.S (th,e,
=·
"6b 4.
2
Pan ar~as) and so the marginal utility (MU) in the bottom
1 echnes. TU remains unchanged with the consumption of 0 Qx
the ~fth
-2 MUx
per da hambu~ger, and so MU is zero. After the fifth hamburger Quantity of X
y, TU declines and MU is negative.

Plotting the values given in Table 3;1, we obt3;111 Figure 3.1, with the top panel
showing total utility and the bottom panel showing marginal utility. The total and marginal
utility curves are obtained by joining the midpoints of the bars measuring TU and MU at
eaclf level of consumption. Note that the TU rises by smaller and smaller amounts (the
shaded areas) and so tlte MU declines. The consumer reaches saturation after consuming
 f Consumer Behavior and Demand
PART TWO Theory _o
. h ed with the consumption of the fifth
Th TU remams unc ang . MU. ..
.:0 urth hamburger. us, , ·,. b e-r TU declines and so is negative
the 1' • Af th fifth ham urg , ·
hamburger and MU is zero. ter eh - . ht inclination of the MU curve reflects the law
The negative slope or down';~d-to-t e ng
of diminishing marginal utility. f a · articular individual; that is, they are unique to
Utility schedules reflec~ tastes o - p articular subjective preferences and percep-
0
the indi~idual a~d r~~ect his : h~::ew~i;erent tastes and different utilit~ schedules.
tions. Different md1v1~uals m y d I the individual's tastes remam the same.
Utility schedules remam unchange so ong as

Cardinal or Ordinal Utility? _

The concept of utili;y di~cuss;d in the previous section was introd~ce~ at about the same
time, in the early 1870s, by William Stanley Jevons of Gre~t- Bntai?, ~~rl Menge~ of
Austria, and Leon Walras 'of France. They believed that the utility an mdividual re~eives
from consuming each quantity of a good or basket of goods could be measured cardmally
just like weight, height; or temperature. 2 . ·
Cardinal utility meaµs that an individual can attach specific values or numbers of
utils from consuming each· quantity of a good or basket of goods. In Table 3.1 ~e saw
that the individual received 10 utils from consuming one hamburger. He received 16
utils, or 6 -additional utils,_: from consuming two hamburgers. The consumption of the
third hamburger gave this, individual 4 extra utils, or two-thirds as many extra utils, as
the secoqd hamburger. Th~, Table 3.1 and Figure 3.1 reflect cardinal utility. They actu-
ally provjde an index of satisfaction for the individual.
In contrast, ordinal utility only ranks the utility received from consuming
various amounts of a good or baskets of goods. Ordinal utility specifies that con-
suming two hamburgers gives the individual mo re utility than when . consuming
one hamburg~r, but it does not specify~ctly .how muc_h additional utility the sec-
on4 _hamburger provides. Similarly, ordinal utility would say only that three ham-
burgers give this individual more utility than two hamburgers, but not how many
more utils. 3
Ordinal utility is· a' much weaker notion than cardinal utility because it only
requires that the consumer be able to rank baskets of goods in the order of his or her
preference. That is, when presented with a choice between any two baskets of goods,
ordi~al utility req~ires only that the individual indicate if he or she prefers the first bas-
~et: t~e second_~basket, o~_is indifferent between the two. It does not require that the
md1v1dual_sB~c1f~_how m~ny more utils he o_[ she r~ceive~ from the preferred basket. In
short, ordinal utility only ranks various consumption bundles; whereas cardina-z utility
pro_vides an actual index or measure° of ~atisfaction.

2
A market basket of goods can .be defined as containing specific quantiti'es of vanous
· · , goo ds and-'Services.
• for
example, one basket. may contam. one hamburger, one soft drink ' and a ticket to a ball h'l th
game, w 1 e ano er
basket may contam two soft dnnks and two movie tickets.
, . by the 1·nd·1v1·dual fromconsumm · g
3
To. ,b~ sure,
. numerical.values
. could
. be att!lched
. ,.,•· - to the
.;, utility received
~~mb~rger~ •. e~,e--:n w~t~ _ordmal ut1~1ty. -~~~ever, ..yith ordinif utility, higher utility values only
md1cate higher rankings ~f ut1hty, and no importance ,can be attached to actl.i'al numerical differences in
utility. For example, 20 utilS' ·can only be interpreted as giving more utility than 10 utils but not twice as
much. ~us·, to indicate rising utility rankings, numbers such as 5, 10, 20; 8, 15, 17; or j (lowest), II, and III
are equivalent. _
 CHAPTER 3 Consumer Preferences 9nd Choice 55

. - - · · ortant because a theory

The distinction between cardinal and ordmal uh 1ity is imp d" al •i·t
.
of consumer behavior can ·be developed on the we er assum ak ption of or m utl .1 Y
. . 1
without the need for a cardmal measure. And a theory th at reac hes the same .cone us1on
as another on weaker assumptions is ·a superior theory. 4 t1·1·1ty th""' "'ory provides·
. u
a. con-
.
venient ·introduction .to the analysis of consumer tastes and to the more ngorous rnd1 f-
.,...---:.--:- · ~ ··-:- - h ·
ference curve approach. It is also useful ror me analysis of consumer c mces 1 .
·n the face
.
of uncertainty, which is presented in Chapter 6. Example 3-1 examines the relatmnship
between money income and happiness .

.EXAMPLE 3-1
Ooes ·Mon~y Buy Happiness?
. "". __ _.-~ • -~ r--.:~: '. .,• • l __ ~\;f~ _; .w::.
Does money buy happiness? . Pliilosophers . have long ponaered this :question . .
-~onmpisis h~i:v~·nO\y,,gotten.tQvolY,e d_,Y.J.- try_ing,tq .~}1s~.q~ t}li~!age-old question. 1]1ey _.,
4
__ t.:;al~ulf}tedJhe ".µiean happine&s ratint'.-.(base~-,Qn.;a-j,§s<?f¼Y~of "very happy"_= 4,
· "p;-~tty:happy:' ~ -2, ~d ','qpt,:.too~hf!RPY~ -:- Q,)_,Jor-;iJ1.diY,iq9.aj_s..at different levels of per:-- .
. ""sonal income.at a: gi"Y;~n.:pQ_int .in.Ji!Ile,~d4 9r; Q!f:fe~m-n1~Ji.~,qs ov~r tin,ie. Whaf $ey__.. 0

. foµn~Lw~ that up JO,~ ·,inc9me.pei;1Gapjta- o(;ab9u,t l~Cl_.~ . ltiglie_F llll;_Om~ inJhe

. UniJe$1S~t~5:.~ei e ~o~itiv~ly _c9rr~!a~9~wir!i hY.RR!R~§~ J.eftpgpses, hut th~t after t:I;uJt, ,.
·.bfgh~r inconi~~-iJ!ld Jittl~, :if o<:lnY~-.et:f~~~~OI\: <1.b..se~ $W k~PPID~~~; Fa:tthermore, ~verag~ ·
individual happiJ;i.~_ss in~theUnited St~~e,§ r~m.l}if!~g,reµiar~~~~fflat ,&ince th~ 1950s in . ;
the face of a cons.iderable ip,crease·1~ average income. Similar results were found for~ .
other advanced
-
,riati<;>flS~_·;:,_--~!
'¾_~,_,,:~;
su,ch;;-~,;a~ . the .U,ni(ed. :--l!'"
t~(i~ --"' -
·K.tqgdpm, F1:<U1ce, .Germany,,
- ..-;,_9-,.~ · ;·..,~:---:--:-,;~~-i~: / - --·
and 'Jal)an.
- ·-
_ , .i
These results seem.··to'"'go' coutiter't<f lhe. o·asic-ecotiom.1c ·assumption that l:tj.gh.er .p e~'."
sonal.in~pme leaq.s toJ1Jgh~r;1I-ti_.liJY-: \; , .~ ;_,.3 : · _:-,., , .. ·._,::
.· · T~o .e~plan!ltions:are_giv~n: fo.r th½se_rel,!).arkabl~,,~4 ·I?!J.Zzling-results: (1) thaL
hapBine&~__,i_s _l,a~ed ,on_r~lativ,e ratq~r,:than,:,~bs.oJute m,.c.~m~ ;and {2.) that happiness
qui~ldy_ad~pts"t9, _c hMge_s;i~.,\ pe_:level ,of ipcome. ·Sp~.~ ~~j, higher incomes make
ind.jyidl,lals happj~r fof a whilp, buttQ~ir,~;ffec_tf~d~s. Y~rt ,quiqkly as indivi:du~s ad.just , -~
to th~ hj.gl;l~r incofI!e,an,9,sQO!J. tal<:~~ttJQ!':gfM,ly~, E9f ~~a.Il}ple, a generation ·~go, cen-:
tral heating :.was: -r~gjird~d:1~s ia lux4ry,.:, ·w-bil~_, today__ it is viewed as essentictl. ;
furthe@oxe,@ ~nPi_vjc:bJals _become richer.~ey ~ec~,n1e _happier, but when society as
a whole grows,.pcli~P,-"nobody seerosJ1-apJ!1~1-!}p !her ~ord_s1 people are ·o ften more
concerned-~bout their incomy ,relative Jo others.' tQaq ~bout their absolute income.
4
Pleasure ,at_your ow~, P?Y Jj~~ qm vanish ~he,_p ypu learn that a colleague has been 1
given ,ij similw, pay.increa,se. · ; - . il'.,1 :• 1 •
;i
The irnpl~cation of all, of tlµiS is that_people '~-effort. to work more in order to eam :-. i
aQd: spend more in advapced (tjch),soc_i~.ties,_does not.m~~,pepple any happier because .
others dq_t4e ,_salij,e. -(~ ,pog,ri,cpUJ1tliles, .. hi_~.9~r _incomes .. do ,make people happier).
Lower taxesh:in, 1:0eE -Umt~d
grow faster t an m urope, ,u , s;.
i~t hie~ dc
ed.QFo~age;p~9P,}~1tq:~ 9 _more aqd ;tJie nation ·to. . a
not,Il~Cl?S~¥Y,3/,l}l~~ Americans .happiet{t;ha)l o"i
. . ' ' ',. ~~i•J
,--~ ,._ ~. "--•'•l ~. - :~·,:1
 . . . · · iness researchers is that aft~r ~~ng enough to
Europeans. The consensus am~~ happ., ;,(;.'boUt $20,000), family, friends, and eoo,_
satisfy basic wants (a per c~plta 1ppo~ s-in life. . C • • •

munity tendto be ~e most tmp9rtant ·,· g . . . . ...

S . . . " . d Ha ~ness/?._Eco,wmic Jo~mal, ,July::2009, B. S. _Pre! and A.
oui't'es:R.A. Easterlin, Income• Learn an · PPHam· . IP.P
f. r o ' m nese""""'?
·' e.ss ,..., .....,. •.,.".Journal
· ·, ·oifE
. conomicLiterature, June

;
Stutzer, "What d, Can Econorms.~: . . · . . ;., S-- ;,,nc:.
u . ess "'Lessqnsfrom a.,.yew c...., "' .\~. . . . . '.n ~ndo~: Penguiµ, Jl. Dt :ella
. , . 5);.Wi
200 .,1\1\Land
2002·' ~-. .Lay.w nappin · Yot&· · ' '·c
, · . , · ."c - · '· · .,,,
:,,, . h. ''So·me' UsesofHap,nmessData,Jouma
-Perspectives, inter~. .
.,.-. P~ mz _____ f ' . • . dum··
R. MacC~oc • . . . -:.i I. . " : . . :,·, ; - .s ''R:elative Tnfum~ Happmess, an . . tty;
. pp. -25-46; and A. E. Cll!J'k, P.;FnpJtersd, and
Anfapl3D.!ltionfor the Eas~r~~-:ara ox an .
Journal
. ·~ ; .,
of°Economk Literature, March
· · . .

-~~-.2:::'.'1:1.~;;J~cc=~~--~~"'-~-.
CONSUMER'S TASTES: INDIFFERENCE CURVES

In this section, we define indifference curves and examine their characteristic,

Indifference curves were first introduced by the English economis.tf. Y. !½:fgeworth m the
1880s. The concept was refined and used extensively by the 'Italian economist Vlllredo
Pareto' in the early 1900s. Indifference curves ~ere popularized and greatly extended in
application in the 1930s byJ wo othe!:English economists: R. G. D. Allen and John R.
Hides. Indifference curves are a crucial tool of analysis because they are used to represent
·an ordinal :measure of the tastes and preferences of the consumer and to show how the
consumer maximizes utility in spending income,:~-- ---

Indifference Cti,:ves-What Do They Show?5

Consumers• tastes can be ei<amined with ordinal Utility. An ordinal measure of utility•
based on three ·assumptions. First,·we assume •that when faced with any two baskets of
goods, the consumer can determine whether he or she prefers basket A to basket B, B to A,
or whether he or-she is indifferent between the two. Second;we assume that the tastes of
the consumer are consistent or transitive. Thar is, if the consumer states that he or she
prefer, basket A to,tiasket B and also that he or she' prefers basket B to basket C, then tbaJ
conswher will prefer A to C. Third, we assume that more of a-commodity is preferred to
less; that is, we asswhe that the'_commodity is a good rather than a bad, and the consum~
is never satiated with the commodity.• The three assumPtions can be used to represent an
individual's tastes with indifference CUrves. In order to conduct the analysis by plane
geometry, we will assume throughout that there are only two goods, X and Y.
An indilferena, curve sh~w~ the various combinations of two goods that give the
consumer equal utility or satisfacnon. A higher indifferetice curve refers to a higher level
of satisfaction, and a lower indifference eurve refers to less satisfaction. However, we
have no indi_c ation as_ to ·h_ciw much additional satisfaction or utility a higher indifference
curve
ing of indicates. different tndifference curves simply provide an ordering or ran!<·
That ts,preference.
the individual's

s For a mathematical presentation of indifference curves and their characteristics using rudimentary calculus,
Section A. I of the Mathematical Appendix at the end of the book.
~x,mple, of bods ore poll"tioo, god,,ge, and di"''"• of whioh ,.., is pref"=<! to mon,_
 tHAPTER -3 Consumer Preferences and Choice 57

·' · For example, Table 3.2 gives an indifference schedule showing the various combina-
,1
. tions of hamburgers (good X) and soft drinks (good Y) that give the consumer equal sat-
isfaction. This information is plotted as indifference curve U1 in the left panel of Figure 3.2.
· The 'tigtit pan~l repeats indifference curve U 1 along with a higher indifference curve ( U2)
~d a lower one (Uo). - · - · ·'
, · · Indifference curve U 1 shows that one hamburger and ten soft drinks per unit of time
:.:~ · ''(combination A) give the consumer the same level of,satisfaction as two hamburgers and
six soft drinks (combination B), four hamburgers and three soft drinks (combination C), or
seven hamburgers_.and one soft drink (combination F). On the other hand, combination R
(four hamburgers and seven soft drinks) has both more h~burgers and more soft drinks
. than combination B (see the right panel of Figure 3.2), and so it refers to a higher level of
1 t.r ' ' • r ', - satisfaction. Thus, ·combination R and all the other combinations that give the same level of
satistacti_on as combination R define higher indifference curve U2 • Finally, all combinations
I '

. . - ! '

• .J

Hamburgers {X) Soft Drinks (Y) Combinations

A
B
' l jt; :r •• c ·.
p - .,

Qy '-.

Soft drinks ( Y)
per unit
of time 11
10
10 ;::....
4-c
0 8
..e-
...... 7
6
6 & 5
4
3'
..
3 2 U2
1
1
0 1 2 3 4 6 7 9 Qx
0 1 ·2 4 7
- •1 · Hamburgers (X) pei; unit of time Quantity of X
FIGURE 3.2 Indifference Curves The individual is indifferen! among combinations A, B, C, and F
·since they all lie on indifference·curve U, . U, refers·to a higher level of satisfaction than Uo, but to a
lower level than U2-
58 PART TWO Theory of Consumer Behavior and Demand

n u, give the same satisfaction as combination T, and combination T refers to both fe

o o . .-~ . ) b" . Wer
hamburgers and fewer soft drinks than (and there1ore 1s 11~e~or to com matton Bon {J
Although in figure 3.2 we have dr~wn: only three mdi~erence c~es, there is~
indifference curve going through each pomt m the XY plane (1.e., refemng to each P<>s .
ble combination of good X and good Y). That is, between any two indifference curves 81'
additional curve can aly,,ays be drawn. The entire set of jndifference curves is caneci :
indifference map and reflects the entire set ~f tastes and .preferences of the consUiner.

Characteristics ·of Indifference Curves-

Indifference cyrves are usually negatively slopec;l,_9annot int~rsect, and are convex to the
origin (see Figure 3.2). l!ldµferep.ce curves are negati~~ly sloped because if one basket of
,goods X and Y CQn~ns mor~ ,of X, it will have to contain less of Y than another basket in
order for the two baskets to give the same level of satisfaction and be on the same indiffer-
ence curve. For example, since basket B on indifference curve U1 in Figure 3.2 contains
more hamburgers (good X) than basket A, basket B must contain fewer soft drinks (gOOd
Y) for the consumer to be on indifference curve U 1• - , -.
A positiYely sloped curve would indicate tii.at 0ne basket containing more of both
commodities gives the same utility or satisfaction to the consumer as another basket con-
taining less of both commoditie~ (and I?-O other coinmodity). Because we are dealing with
goods rather than bads, such a cuJ."Ve could not possibly be an indifference curve. For
example, in the left panel of Figure 3.3, combination B' contains more of X and more ofY
than combination A', and so the positively sloped curve on which B' and A' lie cannot be
an indifference curve. That is, B' must be on a higher indifference curve than A' if X and Y
are both goods. 7

Qy
Qy
1

;:....
.....0 ;:....
.....0
c
-~
.:: -~C'i::
(<j
(<j B*
& & 2

. 0 Qx b
Qx
Quantity of X
Quantity of X
FIGURE 3.3 Indifference Curves Cannot Be Positively Sloped or Intersect
In the left panel, the positively sloped curve cannot be an indifference curve because it
shows that combination 8", which contains more of X and Y. than combination A', gives
equal satisfaction to the consumer as A'. In the right panel, since C is on curves 1 and 2,
it should give the same satisfaction as A* and 8*, but this is ,impossible because a· has
more of X and Ythan A*. Thus, indifference curves cannot intersect

7
Only if either X or y were b d . . . l f
F.1gure 3 .3 . a a wou 1d the md1fference curve be po&1tively
• sloped as in the left pane 0
 Indi~~rence ~tu;~es also cannot intersect. Intersecting curves are inconsistent with
the defirution of tndrfference curves. For example, if curve I and curve 2 in the right
~anel _of Figure 3.3 were indifference curves, they would indicate that basket A*
is equivalent to basket C* since both A* and C* are on curve 1 and also that basket
B* is equivalent to basket C* since both B* and C* are on curv.'2. By transitivity, B'
s~ould then be equivalent to A*. However, this is impossible because basket B* con-
~ams more of both good X and good Y than
intersect. · basket A*. Thus, indifference eurves cannot

lndiffere~ce curves ~e usually convex to the origin; that is, they lie above any tan-
gent to ~he curve. Conve_xity results from or is a reflection of a decreasing marginal rate
of substitution, w~ch is discussed next.

The Marginal Rate of Substitution

The marginal rate of substitution (MRS)_refers to the amount of one good that an iJtdj-
vidual is willing to give up for an additional unit ·of another g()od while- mai~tairiiii't tbe
same level of satisfc\ction or remaining on the same ind~fference Jurve. ·For example, .the
marginal rate of substitution of g·ood X.for good y (MRS~) refers to the amount of Y that
· the individual is willing.to exchange per unit of X-and maintairi the ;a.me leyel df satisf~9-
tipp. Note that MRSxr measures the downward ~ertical distance (the amount of Y thanhe
individual is willing to give up) per unit of horizontell distance (i.e., per·additional unit of X
required) to remain on the same indifference curve. That is, MRSxr = -.1Yl.1X. Because
of the reduction in Y, MRSxr is negative. However, we multiply by -1 and express MRSxr
as a positive value. ·
-For example; starting atpoint A on,U1 in Figure J .4, the individual is willing to give
up four unit,s of¥ for one ,additiona1 unit of X, and-'reach point B on U • Thus, MRSxr =
1
- .:_(-4/I}-= 4. This is the absolute (or positive value ·of the).·slope of the chord from point
· - ' -Alto pointB on·u 1• Between pointB and·point C on.:V1, MRSxr = 3/2 = 1.5 (the absolute
slope of chord-BC):- Betweeri points Cand 'F;·MRSxr = 2/3 == 0.67. At a particular point
· '· -on·lhe indifference curve; MRSxris given by the absolute slope of the tangent to the indif-
-ference curve at that point. Different individuals usually have different indifference
curves and different MRSxr (at points where their indifference curves have different
slopes).
We can relate indifference curves to the preceding utility analysis by pointing out that
all combinations of goods-X and ~ 011a;given·indift_eren.ee cu,1:.ve refer to the same level of
total utility for the individual. Thus, for a move~ent dpwn ~_given indifference curve, the
· ·gain'jn utility in ~oriSu~rtg more ~f g'.'°~
X iriust be.:e\i'Ua!_\o the loss in utility_ in con-
sumingl~ss·of good Y: Sp~c1~~ally,the 1~~reas~ m ~onsum~~1on of good~_( 6-X) tm_1es the
'marginal ~~litr that the ind1~1dua,1_r~~e~~es from ~o~su~~g ea~h ad~~onal umt of X
(MlJx) mus't.be equal fo the reduction 1_n (-6.Y) tl~es margmal utility of Y (MUy).
that is, . . - . . -. . "-

(liX)(MUx) = -(liY)(MUy) [3.1]

MUxlMUy = -liYl!iX = MRSxy [3.2]

' ' "' . . ' ' .
Thus, MRSxy is e-qual to the absolute slope of the indifference curve and to the ratio of the
marginal utilities . .
I
60 PART TWO Theory of Consumer Behavior and Demand

Qy

10 A

;:...
• ..,j< _
I
I
;:...
.....
0
c-
:.c 6
· i::
c<:I

&

FIGURE 3.4 Marginal ,R~te of Subs~itution (i1R~) S~rting at

toe
point A; the individual is willing to give u'p 4 u~its of Y qn7 .
1
additjonal unit of X and reach point Bon u,.Trus, MRSxr_ = 4 (the .
absolute slope of chord AB). Between points B and__c, MR~XY = .
0 1 ·2 · 4 6 7 Qx
3/2. Betw~en C and F, t:1RSXY = 2/3. MRSXY declines as the
inpividual moves down the indifferen~~ cuf"-'.e. Quantity of X .

Note that MRSxy (i.e.·, tile absolute slope pf -th~ i~diff~r~nce curve) declines as ~e
move down the indifference curve. ,This follows from_, or is a reflection of, the conveXIty
of the .indifference curve. That is, as the•indj:vidual woves dawn an indifference curve and
is ieft with less and less.Y(say, .soft drinks) and more ~nd more X (say, hamburgers), each
remaining unit of Ybes;ome_s.more valuapl.e to tpe indiyidual and each additional unit of
X becomes less. valuable. Thus, the indiviqu~ is willing ~o give up less and less of Y to
obtain each.additional unit of~- It is tbi-s -pNperty that m~es MRSxy diminish and indif·
feren~e. curves convex to the origin. We will see in Sec!i,on 3..5 the crucial role that con·
:vexity plays in con.sumer utility maJS:imization. 8 : •

.- ; l

I '

Some Special Types of -Indifference;-Curves-i• 11

, ! ,,.. . • ••"".• i I ...

_Although indiff~rence CUJ"Ves are usually_neg,ati;yt.\Y. sf~petj.__and convex to the origin,

they may somet~mes assume ott:ier s~,ape~, as sq~w,n in F!~ure 3.5. Horizontal indiffe~-
en,c~ curyes, as m the top left panel of F1gure"1.5, would indicate that commodity x,s
a ,neute_r~ ,that i~, tb.~ . c_onsumer i~ , ind*er~nt · befwiin J i~~ing more or less of the
~o~od~ty. Vertical md1fference curves, as m the top right panel of Figure 3.5, would
md1cate mstead that commodity Y is a neuter. · '
The• b?tto~ left panel of figure 3.5 shows indifference curves that are negativelY
sloped s~a1ght lines. Here, MRSxy or the absolute slope of the indifference curves is con·
stant. This means that an individual is always willing to give up the same amount of good
Y (say, two cups of_tea) for each additional unit of good X (one cup of coffee). Therefore,
good X and two umts of good Y are perfect substitutes for this individual
• \ I •

8
A movement along an indifference curv · th · •'
em e upward direction measures MRSrx, which also diminishes-
THE CONSUMER'S INCOME AND PRICE CONSTRAINTS:
THE BUDGET LINE

In this section; we introduce the constraints or limitations faced by a consumer in satis-

fying his-or her wants: In order to conduct the analysis by plane geometry, we assume
that the consumer spends all of his or-her income on only two goods, X and Y. We will
see that the constraints of the consumer can then be ~epresented by a line called the bud-
get line. The position of the budget line and changes in it can best be understood by
looking at its endpoints.
64
PART TWO Theory of Consumer Behavior and Demand

Definition of the Budget Line

In Section 3.2, we saw that we can represent a consumer's tastes with an indifference
map. We now introduce the constraints or limitations that a consumer faces in attempting ·
~o sati_sfy his or her wants. The amount of goods that a consumer can purchase over a
given period of time is limited by the consumer's income and by the prices of the goods
that he or she must pay. In wbat follows we assume (realistically) that the consumer can.
not affect the price of the goods he or she purchases. In economics jargon, we say that the
consumer faces a budget constraint due to his' or her limi~ed income and the given prices
of goods. .
By assuming that a c~nsumer spends all of his or her income on good X (ham-
burgers) and on good Y (soft drinks), we can express tµe budget constraint as

PxQx + Py(h= I _ (3.3]

where Px is the price of good X, Qx is the quan'tity of good X, Py is the price of good Y,
Qy is the quantity of good Y, ~nd ·1 is the consuwer's money income. Equation [3.3] pos-
tulates that the price of X times the quantity of X plus the price of Y times the quantity of
Y equals the consumer's moriey income. That is, the ·<!lllount of money spent on X plus
the amount spent on y equals the consumer's income. 10 •
Suppos~ that Px = $2, Py= $1, and/= $10 per unit of time. This could, for exam-
ple, be the situation of a student who has $10 per day to ·s pend on snacks of hamburgers
(good X) priced at $2 each and 01,1, soft drinks (good Y) priced at $1 each. By spending all
in~ome on Y, the consumer could_purchase lOY and OX. This defines endpoint Jon the
vertical axis of Figure 3.6. Alternatively, by spending all income on X, the consumer
could purchase 5X and OY. This defines endpoint Kon the horizontal axis. By joining end-
points J and K with a straight line we get the consumer's budget line. This line shows the
various combinations of X and Y that the consumer can purchase by spending all income
at the given prices of the two goods. For example: starting at endpoint J, the consumer
could give up two units of Y and use the $2 not spent on Y to purcha~e the first unit of X
- and· reach point L. By giving up another 2Y, he or she-could purchase the second unit of
X . The slope of - 2 of budget line J K shows that for each 2Y the consumer gives up, he or
she can purchase lX more.
By rearranging equation (3.3), we can express the consumer's budget constraint in a
different and more useful form, as follows. By subtracting the term PxQx from both sides
of equation [3.3) we get

PrQy=l-PxQx [3.3A]
• ' I ~• I 4 J ' ..- -

By then dividing both sides of equation [3.3A] by Py, we isolate Qy on the left-hand side
and define equation [3.4]:

Qy = /!Py - (Px/Py)Qx [3.41

10 '
_Equation
[3.3] could be generalized to deal with any number of goods. However, as pointed out, we deal
with only two goods for purposes of diagrammatic aQalysis.
 CHAPTER 3 Consumer Preferences and Choice
65

Qy
J
10

8
::,..
4-<
0 .c
e-
·a
6
c:::

·FIGURE _ 3.6 . The Budget Line With an income of/~

& 3
$10; and Pr= $1 and Px = $2, we get budget line JK This - 2
shows that the consumer can purchase lOY and OX fendpoint K
' J), 8Y and.lX (pointt): 6Y and 2X{pbint8), or .. . OYarid 5X
0- 1 2 4 5 Qx
(endpoint I(). I/Pr= $10/$1 = 10 is the vertical or Y-intercept
of the oudget lihe~m9 ._:_ pxfP/ = -$2/$1 =·.: . : 2 is ttfe slope. Quantity of X
.

The first term on the right-hand side of equation [3.4] is the vertical or f-intercept of the
budget line and -Px/Pyis the slope of the budget line. For example, continuing to use Px
= $2, Py= $1, and/= $10, we get/ /Py= 10 for the f-intercept (endpoint Jin Figure
=
3.6) and -Px IPy -2 for the slope of the budget line. The slope of the budget line refers
to the rate at which the two goods can be exchanged for one another in the market (i.e., 2Y
for lX).
The consumer can purchase any combination of X and Y on the budget line or in the
shaded area below the budget line (called budget space). For example, at point B the indi-
vidual would spend $4 to purchase 2X and the remaining $6 to purchjiSe 6Y. At point M,
he or she would spend $8 to purchase 4X and the remaining $2 Jo purch~e 2Y. On the
other hand, at a point such as Hin the shaded area below the budget line (i.e., in the bud-
get space), the individual would sper.d $4 to purchase 2X anq,-$3 to purchase 3Y and be
left with $3 of unspent income. In :what follows, we assume that the consumer does spend
all of his or her income and is on the budget line. Because of the inc~m~ and price con-
straints, the consumer cannot reach combinations of X and Y above the budget line. For
example, the individuai cannot purchase combination G (4X, 6Y) because it requires an
expenditure of $14 ($8 to'purchase 4X plus $6 to purcgase 6f).

Changes in Income and Prices and the Budget Line

A particular budger ijpe refers to a specific lev ei of the consumer's income and specific
prices of the two goods. If the consumer's income and/or the price of good X or good y
change, the budget line will also change. When only the consumer's income changes, the
butlget iiiie'will ..shift up if income (/•) ,rises and down if I falls, but the slope of the budget
line remains unchanged. For example, the left panel of Figure 3.7, shows budget line JK
(the -same as in -Figure 3.6 with I= $10), higher budget line J'K' with I= $15 , and still
higher budget line J"K" with I= $20 per day. Px and Pr do not change, so the three bud-
get lines' are parallel and their slopes are equal. If the consumer's income falls, the bud-
get line shifts down but remains parallel.
66 PART TWO Theory of Consumer Behavior and Demand

If only the price of good X changes, the vertical or Y-intercept remains unchan
and the budget line rotates upward or_counterclock~ise if Px falls and do-:vnwar!~
clockwise if Px rises. For example, the ng~t pane!, of_Figure 3. 7 shows budget_line JK. (the
same as in Figure 3.6 at Px = $2), budget~neJK w1~ Px = $1, and budgetlineJN'with
Px = $0.50. The vertical intercept (endpomt J) rel!lrun~ the same because/ and Py do not
.change. Toe slope of budget line JK" is -Px/Py = -~1=$1 = -1-. The_ slope of budget
line JN' is -1/2. With an increase in Px, the budget line rotates ~lockw1se and becomes
steeper.
On the other hand, if only the price_of ~-c"~apg~ , tge ,~?tizontal or X-intercept Will
_ be the same, but the budget line wiH totat~ upward if Py f~ls-and downward if Py rises.
For example,. with 1 = $10, Px = $2, and Py= $0.50 (rather than Py= $1), the new ver.
tical or Y~intercept is Qy = 20 anf the.s lope of,Ute._new_budget line is -Px /Py::::. -4.
With Py=:~~• .the new Y-intercept:is Qy·~ 5-atJ~,-;--PFPr = 1 (you should be able to
sketch these-line~).
.
:Fin.ally, with a proportionat~
. - in Px and Pr and constant 1,
reduction' .

there will be a -parallel upward shift in the ·budget line; with a proportionate increase in
Px and Py and -constant /, there will be a parallel downward shift in the budget line.
Example 3-3 .shows that time, instead of the consumer's income, can be a constraint.

Qy
]"
20

15

¥,
Qy '
J
IO IO

K" N'
0 . 7.5 IO ·Qx 0 Qx

jK
F~GURE 3.7 . Changes·in the B~~get Line The left pan.el shows budget line (the same as in Figure 3.6
with/= $10), higher budget hneJK with/= $15, and Stlll higher budget line J"K" with/= $20 per day. Px and
Py do not change, so the three budget lines are parallel and their .slopes are equal The right panel shows budget
line JK with Px = $2, budget line JK" with Px = $1, and budget line JN'! with Px = $050. The vertical or ·
Y-intercept (endpoint j) remains the same because income.and Pr do not change. The slope of budget line JK"
is -Px /Py= -$1 /$1 = - 1, while the slop·e of budget line JN' is - 1/2. .
[ 3.s l CONSUMER'S CHOICE

We will now bring together the tastes and preferences of the consumer (given by his or
her indifference map) and the income and price constraints faced by the consumer (given
by his or her budget line) to examine how the consumer determines which goods to pur-
chase and in what quantities to maximize utility or satisfaction. As we will see in the next
chapter, utility maximization is essential for the derivation of the consumer's demand
curve for a commodity (which is a major objective of this part of the text).

Utility Maximization
Given the tastes of the consumer (reflected in his or her indifference map), the rational
consumer seeks to maximize the utility or satisfaction received in spending his or her
income. A rational consumer maximizes utility by trying to attain the highest indifference
 68 PARTTWO Th eory of Consumer Behavior
. . and Demand

curve possible, given his or her budget line. This occurs where an indifference curve is tan.
gent to the budget line so that the slope of the indifference curve (the MRSxy) is equal 1
the slope of the budget line (Px/Py). Thus, the condition for constrained utility maJ(j~
mization, consumer optimization, or consumer equilibrium occurs where the con.
sumer spends all income (i.e., he or she is on the budget line) and

MRSXY=Px/Py [3.5]

Figure 3.8 brings together on the same set of axes the consumer indifference curves
of Figure 3.2 and the budget line of Figure 3.6 to determine the point ofutility maximiza-
tion. Figure 3.8 shows that the consumer maximizes utility at point B where indifference
curve U is tangent to budget line JK. At point B, the co_nsumer is on the budget line and
is
MRSxy = Px/Py = 2. Indifference curve U1 tbehighest that the consumer can reach with
1

his or her budget line. Thus, ·to m-~ z e utility the consumer should spend $4 to purchase
2X and the remaining $6 to purchase 6_Y. Any other c9mbination of goods X and Y that the
consumer could pur~hlise (those on or below lhe -budget line) provides less utility. For
example, the consumer couid spend all income to purchase combination L, but this would
be on lower indifference·curve Uo.
At point L the consumer is willing to give up more of Y than he or she has to in the
market to obtain one additional unit of X. That is, MRSxy (the.absolute slope of indiffer-
ence curve Uo at point L) exceeds the value of Px/Py (the absolute slope of budget line
JK). Thus, ;tarting from -point L, the con1amef can increase his or her satisfaction by
purchasing les_s of Y and more ofX until he o~she _reaches point Bon U1, where the slopes
of U and the budget line are equalJ i.e., ·MRSXY = PxfPy= 2). On the other ·hand, starting
1
from point M, where MRSxy < Px/Py, the consumer can increase his or her satisfaction
by purchasing less of X and more of Y until _he or she reaches point B on U 1, where
MRSx; = Px/Py. One tangency point such, ; Bis assured by the fact that there is an
indifference curve ~-oing through eadh point in the XY co~odity space. The consumer

Qy
.,
J
10
;,..
......
0

-~e--
....: 6
('ll

&
3
FIGURE 3.8 Constrained Utility Maximization The consumer U2
maximiZes utility at point B, where indifference curve U1 is tangent to
budget line JK At point 8, MRSxr = PJPr = 2. Indifference curve u1 is Qx
0 2 9
the highest that the consumer can reach with his or her budget line
Thus, the consumer should purchase 2X and 6 Y. Quantity of X
 ,,...__ "'t ·•,

EXAMPLfc,l ~.4
," U~ilit(~a~i~i~ation and ·Gove_
rnment'Warnings on J~~k Food
.~ .; J - . -· - • '!' ' :

·suppose that in F1gure 3.9; goodX,refers to milk ai:u;i good Yrefers to soda, Px = $1,
Pr = $1, and the consumer spends his o/ her entire weekly allowance of $10 on milk
and sodas. Supposeaiso that the constirrier maximizes utility .by spending $3 to pur-
chase three containers of milk $7 to purchase seven sod'as {point B on indifference and
curve U1) before any ·govertiment warninion the·danger.of dental cavities and obesity
from sodas. Aftyr the warning, the cqnsumer's tastes may change away from sodas and
_ ·toward milk. It inay be argued that go.vernil)en_t wru:niiigs change tbe.jnformation avail-
able to consumers rather than tastes; that is, the warning affects consumers~ perception

Qy
10

f·1w-.~~1H-J{i:;.t!; -"·~ :; ~· ·- H~
!rI~JiJ}·~_A~\·~-:,~f- '.Li'"!- • ,~,
h .:- r i .~.
\. \ . ,.Jg.,1-'!•~ . i .. • ~' ' .: • .. ;;
FIG!,JRE
'
3.9 ··Effect
-I'• ,l..{' '!J.~ l .,,.
of G'overnment Warning~ t . The ·
f", I, • •• r '. •

cons~rne( m~x_itnizes·utility by purchasing 3. containe~s ·of milk 4

and) sodai 'ti:)oifit B on'·indiffeirenc'.e ·euh,e the . :· ,, uJBetoYe
f gove'.~me~f;~, Prling 6,fr t~~ consun:iption_of.';~-q~~-1~~( ~h~ ;, •j'· .,
.warning, the ·consumer's tastes change and are sHown 'by ·dashed ·. · 0

. ind~erenc~ ~uives .O'o-and U' 1-. Trie'"consufner how, ma~imite!s ,\: ••'" : :

t Utility by purc"hasirfg'6 ccrntaihets\ J(niill< ana only 4 fodas' (poihf 1-,.:• Qx

3. · ' 6 10
· o.. ,
B', Where U.'1jJ:tangenuo the budget 'ffrie). ' ,~' ; : ,.. ,,}. Milk
''' i ' ' , i·~.,-,/
t l/
;, "',,.( .7.r
, ., ..,.
.1'4:·(,.J~ .,..j,,)'-• - , .. >,
- ' ,.r
\c• ·•~ "' I • '
r ,,;$.
• •,~ ~.,
., .;,
-«'..

11 For a mathematical presentation of utility maximization-using rudimentary calculus, see Section A.2 of
the Mathematical Appendix .
• 12See I. Ehrlich, ,"Partic_ipation i~ Ille~itimate Activities: A ~heoretic;al and Empirical Investigation,"
journal of Political Econom):, May/June I97~; y,I. T. Dickens, ' Crime and Punishment Again: The
1

Economic Approach with a Psychological Twist," National B~reau of Economic Research, Wurking Paper
No. J884, April 1986; and A. Gaviria, "Increasing Retums·and the' Evolution of Violent Crimes: The Case of
Colombia," Journal ·o f Development Economics, February 2000.
 CHAPTER 3 Consumer Preferences and Choice 73

Marginal Utility Approach to Utility Maximization

Until n;ow we have examined constrained utility maximization with ordinal utility (i.e.,
with indifference curves). If utility were cardinally measurable, the condition for con-
strained utility maximization would be for the consumer to spend all income on X and Y
in such a way that
· MVx MUy - [3.6] .
--=-
Px Py
Equation [3.6] reads, the margin~~ utility of good X divided by the pnce of good X equals
the marginal utility of good Y divided by the price of good Y. MUxiPx is the extra or mar-
ginal utility per dollar spent on X. Likewise, MUy/Py is the marginal utility per dollar
spent on Y,. Thus, for constrained utility maximization or optimization, the marginal util-
13
ity of the last dollar spent on X and Y should be the same.
' Fbr example, Table 3.3 shows a portion of the· declining ~arginal utility schedule for
good X and good Y (from Table 3-.1), on the assumption that MUx is independent of MUr
(i.e., that MUx is not affected by how much Y the individual consumes, and MUr is not
affected by the amount of X consumed). If the consumer's income is/= $10, Px = $2,
and Pr= $1: the consumer should spend $4 to purchase~2x-and tlie remaining $6 to pur-
chase 6Y so that equation [3.6fis satisfied. That is, ·
6 utils . 3 utils [3.6A]
$2=$1
If the consumer spent only $2 to purchase lX and the remaining $8 to purchase
-i•• 1 - , • •.

8Y, MUxlfx = 10/2 = 5 an~ MUy!Py = _111 .= 1. The last (second) dollar spent on X
_ _ , I • • •

thus gives the consurper ~,ye _times as ~~cl! .utility _as the fast (eighth) dollar spent on
Y and the consumer wo~io not be ri:laxiinizfog utility. To be at an optimum, the con-
sum~r shoulctpurchase-~oi;ej >(~ (~_Uxfalisj ana-°l_ess of y14(MUrrises) until he or she
purchases 2X and -6Y, ~h,ere equation [3_.6] is _satisfied. . This is the same result
obtained with .the indifference cur;e· appioach in Section 3.' 5. Note that even when the
consumer purchases lX and 4Y equa~ion ~3.6] is sa~sfied (MUxlPx = 10/2 = MUrlPr
= 5/1), but the consmner \Yould not be at an optimull! because he or she would be
spending only $6 of the $10 income. · ·

Qy MUr
Qx MUx
10 4 5
1 5 4
2 6
6 3
3 4
7 2
4 2
8 l
5 0

13We will see in footnote 14 that equation [3 .6] also holds for the indifference curve approach.
14By giving up the eighth and the seventh units of Y, the individual loses 3 utils. By using the $2 not spent
on Y to purchase the second unit of X, the individual receives 6 utils, for a net gain of 3 utils. Once the
individual consumes 6Y and 2X, equation [3.6] holds and he or she maximizes utility.
· TWO Theory of Consumer Behavior and Demand

The fact that the marginal utility approach gives the same result as the indifferen
curve approach (i.e., 2X and 6Y) should not be surprising. In fact, we can easily show w~e
this is so. By cross multiplication in _equation [3.6], we get y

MUx Px
M Uy - Py [3.7]
But we have shown in Section 3.2 that MRSxy = MUxlMUy (see equation [3.2]) and in
Section 3.5 that MRSxy = Px!Py when the consumer maximizes utility (see equation
[3.51). Therefore, combining equations [3.2], [3.5], and [3.7], we can express the condi-
tion for consumer utility maximization as • · ·

MUx . fx
MRSxy =- -=- [3.8]
MUy Py
Thus, the condition for consum<tr utility m~imization with the marginal utility approach
(i.e.,-equatioo [3.6]) is equivalen.t to that ~ith the 1n~ei-ence c_uzye approach (equation [3.5]),
except for comer solutions. With both approaches, the value,of equation [3.8] is 2.