Chapier 4: Efficiency and tite ldealized Competiiioe Model 55
I
Efficiency
and the Idealized
Competitive Model
Toe Efficiency Benchmark: The Competitive
Economy
e v · t
ollective action enables society to produce distribute and
d b d ' , consume a great
arre Y an a un anee of goods. Most collective action arises from voluntary
agre�men�s among people-within families, prívate organizations and exchan e
�elatt�Shlp�. re. .J?Olicy analy_st, however, deals primarily with c;llective acti�n
nvo v1�g t e eg1hmate. c?erc1ve powers of government: public olic encour-
ages,. dis�u�ag�s,. prohíbíts, or prescribes private actions. Beg�iniwith the
prenuse t �t mdiv1d_uals g�nerally act in their own best interest, or at least what
the.Y perceive as their self-mterest, policy analysts bear the burden of providing
ratio:11-a es for any go�ernment.al _interference with prívate choice. The burden
apphe_st° 1 th
e eva�uatwn of exístíng policies as well as new initiatives It is ah
�!��:�iae �m�nhtt �ftnot the first �tep,_ in any analysis, and will often pro.vide the
insig in o comp ex situations,
Our approach to classifying rationales for publíc policy begins with the
��nc�p� of a perfectly competitive economy. One of the fundamental bodies of
. eory_m modern econorrucs deals with the properties of idealized economies
mvolvmg large numbers of profit-maximizing firms and utility-maximizing
consum�rs. Under certain assumptions, the self-motivated behaviors of these
cient in th actors_ lead to patterns of production and consumption that are effi-
c1e11 m t e specíal sense that it would not be possible to change the patterns in
suc a way so as to make sorne person better off without making sorne other
person worse off.
. Economists recognize severa! commonly occurring circumstances of rivate
choice, referred to as market failures, that violate the basic assumptionlof the
54
'deal:l.zed competitive economy and, therefore, interfere with efficiency in production
cmisumption. The traditional market failures, which we discuss in Chapter 5, pro-
vide widely accepted rationales for such public policies as the provision of goods and
1
� e regulation of markets by government agencies. Economists, until recently, have
aíd Isss attention, however, to the plausibility of sorne of the more fundamental
pssumptions about the behavior of consumers. For example, economic models usu-
� treat the preferences of consumers as unchanging. Is this reasonable? Do
m�wners always make the right calculations when faced with decisions involving
:uch 00mplexities as risk? Negative answers to these questions, which we consider in
Chapter 6, may also provide rationales for public policies.
Of course, efficiency is not the only social value. Human dignity, distributional
equlty, economic opportunity, and political participation are values that deserve con-
idei·ation along with efficiency. On occasion, public decision makers or ourselves as
.metnbers of society, may wish to give up sorne economic efficiency to protect human
life, make the final distribution -of goods more equitable, or promote fairness in the
ctiBtmblitio.n process. As analysts, we have a responsibility to address these multiple
values and the potential conflicts among them. We discuss distributional and other
valúes as rationales for public policies in Chapter 7.
Imagine a world where each person derives utilíty (one's perception of one's own
well-being) from personally consuming various quantities of all possible goods,
including things, services, and leisure. We can think of each person as having a utility
function that converts the list of quantities of the goods consumed into a utility index
such. that higher numbers imply greater well-being. We make severa! basic assump-
tíons, First, other things equal, the more of any good a person has, the greater that
person's utility. (We can incorporate unpleasant things such as pollution within this
framework by thinking of them as "goods" that decrease utility or result in dísutilíty.)
And, second, additional units of the same good give ever-smaller increases in utility;
in other words, they result in declíníng marginal utility.
Now make the following assumptions about the production of goods: Firms
attempt to maximize profits by buyíng factor inputs (such as labor, land, capital,
and materials) to produce goods for sale. The technology available to firms for con-
verting factor inputs to final goods is such that, at best, an additional unit of output
would require at least as many units of inputs to produce as the preceding unit; in
�ther words, it becomes more costly in terms of resources to produce each addi-
honal unit of the good. Firms behave competitively in the sense that they believe
that they cannot change the prices of factor inputs or products by their individual
actions. '
Each person has a budget derived from selling labor and his or her initial
endowments of the other factor inputs, such as capital and land. People maximize
t�eir well-being by using their incomes to purchase the combinations of goods that
grve them the greatest utility.
In this simple world, a set of prices arises that distributes factor inputs to firms
and goods to persons in such a way that it would not be possible for anyone to find a
reallocation that would make at least one person better off without making at least one
56 Pari JI: Conceptual Foundations for Problem Analysis
Chapier 4: Efficiency and the Idealized Competitive Model 57
person worse off.' Economists ref t h di . . .
a concept with great . t ití er o suc a istribution as berng Pareto etiicient lt . persons that sum exactly to $1,000. Any point on this line or inside the triangle it forms
m ui rve appeal· would 't d h Id ' b · :J�' • IS
an existing distribut ion if we could find n l, an � ou . n .t, "" e d 1ss ahs fied With with the axes would be a technically feasible allocation because it gives shares that
someone better off without makín an a ternatíve dístríbutíon that would make
other criteria for choosing betwee� :�0�1.e t��e
�orse�ff? Although we would neect
we should, unless we are malevolent alw is n u ions at were each Pareto efficient
sum to no more than $1,000.
The potential Pareto frontier indicates all the points that fully allocate the
$1,000. Any allocation that <loes not use up the entire $1,000 cannot be Pareto effi-
from inefficient distributions to effici�nt m�l:
want to make Pareto-improv.ing move� cient, because it would be possible to make one of the persons better off by giving
Figure 4.1 iIIustrates the concept of Pareto efficie . . her the remaining amount without making the other person worse off. The actual
$1,000 between two people. Assume that the two eony rn�f 1vrn&" the aUocation of Pareto frontier depends on the allocations that the two people receive if they reach
agreed-upon amounts of money that sum to no m�re 1:he wi 1 recerve any II:utuaIIy
represents the allocation to person 1 and the horizontal . an $ ,000. The vertical axis 110 agreement. If they each receive nothing in the event they reach no agreement,
pers�n 2. An allocation of all of the mone to erson 1 waxis represents the aU?cation to rhen the potential Pareto frontier is the actual Pareto frontier in that any point on it
vertical axis at $1,000· an allocation of allyof thpe tould appear as the pomt on the would make at least one of the persons better off without making the other person
worse off.
· . ' money o person 2 would a
po nt on the hor zon l axis at $1 000 The line í ppear as the Now imagine that, if these two people reach no agreement about an allocation,
r i ta conne t th .

ca 11 th e potenii·al Pareio frontier ' repr· esents all c rng ibl ese 1t1wo p.ornts ' whi c h we person 1 receives $100 and person 2 receives $200. This point ($100, $200) can be thought
th
' e possi e a ocations to the two of as the status qua point-it indicates how much each person gets in the absence of an
agreement. The introduction of the status quo point reduces the Pareto frontier to the
Iine segment between ($100, $900) and ($800, $200). Only moves to this segment of the
potential Pareto frontier actually guarantee that each person is no worse off than
under the status quo.
$1,000
Note that whether a particular point on the potential Pareto frontier is actually
Pareto efficient depends on the default allocations that comprise the status quo. More
generally, Pareto efficiency in an economy depends on the initial endowments of
($800, $200)
resources to individuals.
1 The idealized competitive economy is an example of a general equiiibrium model-
1
1 it finds the prices of factor inputs and goods that clear all markets in the sense that the
1
1
quantity demanded exactly equals the quantity supplied. Although general equilib-
1 rium models can sometimes be usefully applied to policy problems, limitations in data
1
1 and problems of tractability usually lead policy analysts to evaluate policies in one
1
1 market ata time.2 Fortunately, a well-developed body of theory enables us to assess
1
1
economic efficiency in the context of a single market. ·
1
1
1
1
1
1 Market Efficiency: The Meaning of Social Surplus
1
1
1

/
$100 -----�-------------------- We need a yardstick for measuring changes in efficiency. Social surplus, which mea-
- sures the net benefits consumers and producers receive from participation in markets,
1
$1,000 serves asan appropriate yardstick. In the context of the ideally competitive economy,
$200 the Pareto-efficient allocation of goods also maximizes social surplus. When we look
Allocation to Person 2 across markets, the set of prices and quantities that give the greatest social surplus is

Status quo: Point ($100, $200) usually Pareto efficient. Moreover, differences in social surplus between alternative
Potential Pai:eco F�ontier: Líne segment connecting ($1,000, $0) and ($0, $1,000 market allocations approximate the corresponding sum of differences in individual
Pareto Frontier: Line segrnent connecting ($800, $200) and ($1 OO. $900) )
":7elfares. As social surplus is the sum aj consumer surplus and producer surplus, we con-
Figure 4.1 Pareto and Potential Pareto Efficiency síder each in tum.
 1 2For a review of the use of general equilibrium models in education, see Thomas J. Nechyba,
For the history of general equilibrium theor E R ·
Competitive
of a Equilibrium· 1930-1 54 "Jo . 1 if E }'., Lee · ºY Wemtraub, "On the Existence "What
Can Be (and What Has Been) Learned from General Equilibrium Sirnulation Models of School
· 9 , urna o conomic iierature 21(1) 1983, 1-39. Finance?"
Nationa/ Tax [ournat 54(2) 2003, 387--414.
58 Part TI: Conceptual Foundations for Problem Analysis
Chapier 4: Efficiency and the Idealized Competitive Model 59

Consumer Surplus: Demand Schedules as Marginal
Valuations
Imag�,e that you have the last five tickets to the World Cup Soccer final match walk
into a room and ruu101:111-ce to thosepresei:it that you own all the remainin t'. You
t� the event �d that you will sell these tickets in the fallowing manner: Starting i<;kets
d et and other consumption opportunities. You now continue offering successively
bu gr stated prices until you sella second ticket, which a second person accepts at $180,
�weecond-highest valuation. Repeating this process, you sell the remaining three tick-
e
:t
t: $160, $140, and $120, respectively.
You were happy as a seller to get each person �o pay 0-e amount that he or s.h.e
ed the ticket. If, instead, you had announced pnces until you found one (specífí-

h1�h stated pn�e, say, �500, you "".ill keep lowering it until someone agrees to �u""'lth a
:ºu
a ticket. will continue lowermg the stated price until all five tickets a � �ase
(An auctíon such as this with declining prices is known as a Dutch auction: �
1
ª (ect.
an auctio� with ascen?ing prices is known asan English auction.) Each peTSo�:til t
roo� decides the maximum a�ount that he or she is willing to pay far a ticket. If �
maximum ª?1ount 1:1ust be_ paid, then each person will be indifferent between bu: thi
?-nd not buyíng the tick�t. Figure 4.2 displays the val1;1ations far the persons in desc�ln�
mg or_der from Ieft to nght. Altho:1gh the stated l?n�es begin at $500, the first acee d
tance is at $200. Toe pur�:haser obviously values this ticket at least at $200. Value in�
context means the máximum amount the person is willing to pay, given his or her
�ª:
1 $100) such that exactly five people wanted to buy tickets, then sorne of the pur-
"ª110
c� Y�rs would get tickets at a price substantially lower than the maximum amounts
they would have been willing to p�y. Far example, the person with the highest
Juation would have been willing to pay $200 but only has to pay your set price of
;100. Toe difference between the person's dallar valuation of the ticket and the price
that he or she actually pays ($200 - $100) is the surplus value that the person gains
from the transaction. In a similar way, the person with the second-highest surplus
� $-ffü ($180 - $100). The remaining three purchasers receive surpluses of $60
($160- $100), $40 ($140 - �100), and $20 ($120 - $100). Adding the �urpl:1ses realized
by these five purchasers yields a measure of the consumer surplus in this market far
World Cup tickets of $300.
Toe staircase in Figure 4.2 is sometimes called a marginal valuation schedule
bseause it indicates how much successive units of a good are valued by consumers in

$200
- we stated various prices and observed how many units would be purchased at each
príce, then we would obtain the same staircase but recognize itas a demand schedule. Of
course, we would also get a demand schedule by allowing individuals to purchase
$180 more than one unit at the stated prices. If we had been able to measure our good in
small enough divisible units, or if demanded quantities were very large, then the stair-
$160

$140
- i---
case would smooth out to become a curve.
How do we move from this conceptualization of consumer surplus to its mea-
surement in actual markets? We make use of demand schedules, which can be esti-
mated from observing market behavior.

.9
$120

$100 +$80 +$60 +$40

--
+$20 = $300 surplus
(Later we will interpret the curve as the demand schedule far all persons with access to
the market.) Note that this consumer values all units less than choke price, Pe, the price
that "chokes off" demand. Toe horizontal line drawn at PO indicates that she can pur-
t---- ---- r---- --------------- -----
g
.¡
$100 ---- ---- chase as many units of the good as she wishes at the constant price P0 At price P0 she
>
fínd that she could make herself better off by purchasing a little more because she would
valus additional consumption more than its cost to her. (Toe demand schedule lies
abovepxice far quantities less than Q0.) Suppose, on the other hand, she purchased more
than Q!), then she would find that she could make herself better off by purchasing a little
$50 le�s because she would value the savings more than the lost consumption. At given
pmee_ P0, the quantity Q0 is an eouilibrium because the consumer does not want to move
te arL�ternative quantity. The area of the triangle under the demand schedule but above
the pnce line, Pea P0, represents her consumer surplus from purchasing Q0 units of
the good at price p 0• ·
. Changes in consumer surplus are often the basis far measuring the relative effi-

o 2 3 4 5 6 igure 4.3 íf a government policy increases price from PO to P 1? Toe new consumer

World Cup Tickets s:tu:plus is the area of triangle PcbP1, which is smaller than the area of triangle P aP0 by
�e area of the trapezoid inscribed by P1baP0 (the shaded region). We interpret the area
0
Figure 4.2 Consumer Values and Surpluses O
the rectangle P1bcP as the additional amount the consumer must pay far the units
60 Part ll: Conceptual Foundations far Problem Analysis Chapter 4: Efficíency and the Idealized Competitive Model 61

tillty after the price change as befare. It thus serves as a dallar measure, or money
u etric for changes in welfare. If the demand schedule represented in Figure 4.3 were
� ri �d by observing how the consumer varied purchases as a function of price, h';>l_d-
1

. e utility constant at its initial level (it thus would be what we call a constant-utzltty
\ ;:;�,n,ud schedule), then the consumer surplus change would exactly equal the compen-
atíng variation. . . . .
F.i,gure 4.4 illustrates how compensatmg varíation can be mterpreted as a
oney jnetric, or proxy, for utility. Toe vertical axis measures expenditures by a per-
son on all goods other than good X; the horizontal axis measures how many units of
X she consumes. Suppose initially that she has a budget, B, but that she is not
allowed to purchase any units of X, say, because X is manufactured in another coun-
o 1 tr and imports of it are prohibited. She will, therefore, spend all of B on other
u 1
·e
e,
1 gJods. The índífference curve 10 indicates all the combinations of expenditures on
1
1 c.:,ther goods and units of X that would give her the same utility as spending B on
1
other goods and consuming no units of X. Now imagine that the import han is lifted
Po ---------------------�----------
c1
O
that she can purchase units of X at price P x· She can now choose any point on the
líne connecting B with the point on the horizontal axis indicating how many units of
1

1
D X she could purchase if she spent zero on other goods, B / P x· Once this new budget
1
1
llne is available, she will maximize her utility by selecting a point on the highest fea-
1 sible índífference I1 , by purchasing x1 units of X and spending her remaining budget
1
1 011 other goods. Once X is available, it would be possible to return her to her initial
1
1
leve! ef utility by reducing her initial budget by the distance between B and C on the
o Q¡
Quantity of X per Unit of Time c::z:i

'<i A
Loss in consumer surplus: P1baP0 eo : oi::
Revenue collected by government: P¡bcP0 eae: ·;:: füi
> t<:I
Deadweight loss: abe ·a·¡¡¡ 1il
"" O"> e:
o' ¡,¡¡
O

Figure 4.3 Changes in Consumer Surplus o :::l

o
...
o B
ll
�
of the good that she _continues to pu�chase and the area of the triangle abe as the sur·
o
:=l
O!)
e:
·.:i i::
c::z:i
e:
t<:I o o
plus the consumer gwes up by reducíng consumption from Q to Q . í@ ·g E
<e:
0
As � example �f a government policy that raises price, imagine the imposition
1
o g,·¡¡¡ �
of an excise (commodity) tax on each unit of the good in the amount of the difference eo: oE> �e:
e, u ·�

be_tween P 1 and P0• Then the area of rectangle P 1 bcP O corresponds to the revenue U)
>,
...l e
raised by the tax, which, conceivably, could be rebated to the consumer to offset that the equilibrium pnce off simply giving the tax collector a lump-sum
exactly that part of the consumer surplus lost. Toe consumer would still suffer the los and quanhty under the tax payment equal to the area of P1baP0 in return for
of the area of triangle abe. Because there are no revenues or benefits to offset this are not Pareto efficient-if removal of the excise ta:<
r�du�tion in cons:1mer surplus, economists define this loss in surplus due to a reduc· 1t were possible, the a
tion m cons1;11;11p�1on as _the deadweigh� loss of the tax. Toe deadweight loss indic�t� consumer would be better n
d
its associated deadweight loss. o
e::
The loss of consu�er surplus shown in Figure 4.3 approximates the m?st o
�
con:i�only used theoreh�al measure of changes in individual welfare: compensatttig
uariation. Toe compensatmg variation of a price change is the amount by which tJ,e
consumer's budget would have to be changed so that he or she would have the sa.lJle

Quantity of Good X
Figure 4.4 Money Metrics for Utility
 Chapter 4: Efficiency and the Idealized Competitive Model 63
62 Part Il: Conceptual Foundations far Problem Analysis
ve�t.ical axis. Th�s amou�t is her compensating variation associated with the avan, Producer Surplus: Background on Pricing
ability of X at pnce Px· It is a dollar value, or money metric, for how much value sh
places on being able to consume X. e In the ideal competitive. mode�, it �s standar� to assume that margi?.ªl �ost of pro-
Instead ?f asking .how �uch money we could take away to make the person as . for each firm rises with increases in output beyond eqmhbnum
well off after introduction of ímports of X as before, we could ask how much mon levels. dBuct. ion firms have sorne fixed costs that must be pa1id b e f ore any pro d uc
we would have to give her if X were not available so that she is as well off as shy tiion can
ecaustehe average cost of producing output first falls as the fixed costs are spread
· creasi·ng margi·na 1 �os t b �ru·:is t o
ocvceurra., larger number of units, then n·ses as t h e m
would be with imports of X. This amount, called equivalent variation, is shown as th e
distance between A and B on the vertical axis-if her budget were increased from B te0
A, then she could reach indifference curve 1 1 without consuming X. c�r:::�
d · ate because the use of sorne input, such as labor, becomes ess e icient.
�e�1tly, sorne output level minimizes the average cost of the firm. �e curve
rnarkeJ AC in Figure 4.5 represents a U-shaped aver�ge cost curve for the firm. The
�� �� n
In practice, we usu l y work with empiric ly estimated demand schedules that
�old constant cons1;1mers tncome (rather than utility) and all other prices. This constant plus, see Amold C. Harberger, "Three Basic Postulates for Applied Welfare
Economics," Journal of Economtc
tncome, or Marshallzan, demand schedule involves decreases in utility as price rises (and L
total consumption falls) and increases in utility as price falls (and total consumption i
rises). In comparison with a demand schedule holding utility constant at the initial t
level, the Marshallian demand schedule is lower for price increases and higher for e
r
price reductions. Fortunately, as long as either the price change is small or expendi- a
tures on the good make upa small part of the consumer's budget, the two schedules t
are close together and estimates of consumer surplus changes using the Marshallian u
demand schedule are close to the compensating varíations.I r
Now, moving from the individual to society, consider the relationship between e
9
price and the quantity demanded by all consumers. We derive this market demand (
schedule by summing the amounts demanded by each of the consumers at each price. 3
Graphically, this is equivalent to adding horizontally the demand schedules for all the )
individual consumers. The consumer surplus we measure using this market demand 1
9
schedule would just equal the sum of the consumer surpluses of all the individual 7
consumers. It would answer these questions: How much compensation would have te 1
be given in aggregate to restore all the consumers to their original utility levels after a ,
price increase? How much could be taken from consumers in the aggregate to restare 7
them all to their original utility levels after a price decrease? 8
5
. Thus, if we can identify a change in price or quantity in a market that would pro- -
duce a net positive increase in social surplus, then there is at least the potential for 9
making a Pareto improvement. After everyone is compensated, there is still something 7
left over to make someone better off. Of course, the change is not actually Pareto .
improving unless everyone is given at least his or her compensating variations from
the surplus gain.
Our primary use of consumer surplus in Chapter 5 is to illustrate the inefficien-
cies associated with the various market failures. For this purpose, an exclusive focus
on the potential for Pareto improvement is adequate. In the context of cost-benefit
analysis, the Kaldor-Hicks compensation principle advocates a similar focus on net
positive changes in social surplus asan indication of the potential for Pareto improve-
ments. When we consider cost-benefit analysis as a tool for evaluating policies in
Chapter 16, we discuss the implications of focusing on potential rather than actual
improvements in the welfare of individuals.

3Jn any event, the consumer surplus change measured under the Marshallian demand curve will líe
between compensating variation and equivalent variation. For a discussion of the use of consumer surpJus
changes as a measure of changes in individual welfare, see Robert D. Willig, "Consumer Surplus without
Apology," American Economic Revíew 66(4) 1976, 589-97. For a more intuitive justification of consumer su�-
 . ,o-inal cost curve is labeled MC. Note that the marginal cost curve crosses the aver- awe6c.ost· curve at the
latter's lowest poí· nt. W.h en marpi·n�1 cos t i·s 1 ower th an average
gst the latter must be falling. When marginal cost is hígher than average cost, the must be rising. Only
when marginal cost equals average cost <loes average cost
s is easily grasped by thinking about your average sc�re for a
f:t t;r
remain unchanged. Thi
series of examinations-only a new score above your current average can raise your
average. .
The total cost of producing sorne level of output, say, Q5, can be calculated in one of two ways. First,
because average cost is �imply t?tal cost �ivi�ed by the quan- tity of output, multiplying average cost at
this quantíty (AC5 in Figure 4.�) by Q5
yields total cost as given by the area of rec�angle AC5bQ50. Second, re��lling th�t
marginal cost tells us how much it costs to mcrease output by one additional umt,
we can approximate total cost by adding up the marginal costs of producing succes- sive units from the
first all the way up to Q5• The smaller our units of measure, the

�
a;
u
·ee,:
§
G
$ Ps -------- ----------------------------
tl o
a;
u
�n
,,.
<
fj ACs
s

o QL Qs
Output Leve! per Unit of Time

Figure 4.5 Average and Marginal Cost Curves

 Chapier 4: Efficiency and the ldealized Competitive Model 65
64 Part 11: Conceptual Poundations far Problem Analysis

closer will �et�� sum of their associated marginal costs to the total cost of produ ·
Q5, In the limiting case of infinitesimally small units the area under the marg�tng
�ost c_urve (MC in F_igure 4.5)_ from ze�o to Q5 exactly �quals total cost. (Those fa:�� e
iar with calculus wíll recogmze marginal cost as the derivative of total cost th
' t ti ' 1 . 50 at I

. 1 ki h , a ion is G
eq va ent to ta g t e area under the marginal cost curve be een zero th ::E S
....,

output level.) e ....,

...
11)
u
. Now imag�e. th�t the �arket price for the good produced by the firm is p . 5 Th "o'·�
e
firm would maxirruze its profits by producing Q , the quantity at which margina 1 u-o
e
5. � �
eq�a 1 s pn· ce_. B ecause average cost i· s 1 ess th an p nc e at output level Q , the firm wocuolsdt
5 '6'o,,..::.

en¡oy a profit equal to the area of rectangle P 5 abAC 5. Profit equals total revenue min
total cost. (Total revenue equals price, P 5, times quantity, Q5, or the area of rectan �s ��.�
...,
P 5aQ50; the t�t�l cost of prod�cing output �ev�l Q5 is the area of rectangle AC5bQ
I:1 the competitive model, pr�fit would be distributed to persons according to their iní-
i1 Gg
< .:
....,
... 11)

tial �ndowments of ow1�ership. But these shares of profits would sígnal to others t:hat � ¡i::
Uo3
profits could be made simply by copying the technology and inputs used by the fírm �.s
As more fir�s enter the industry, howe�er, total outp:it of the good would rise and: � �
<��
A Ce
�erefore, pnce woul� fall. At_ the same time the new firms would bid up the price ef

up. Eventually, pnce would fall t? PI.,, the level at which the new marginal cost equals A Cm
D

the mdustry.
With no constraint� �n the n_u��er o� iden!ical �irms that can arise to produce
each _good, the Pareto-�fficient equilibnum in the idealízed competitive model is char-
actenzed b;'. zero p�ofits for ali. frrm�. �Note that we are referring to economic profits, Quantityffime
not accou1:tmg profits, Economic profit is total revenue minus payments at competitive
market pnces to. all factors of. produ�ti?n, fncluding an implicit rental price for capital Competitive Pricing Monopoly Pricing
�wned by the fírrn. Account_i�g profit is simply revenue minus expenditures.) If the
Consumer Surplus: Pece Pmae
firm do�s not �ake an explicit payment to shareholders for the capital it uses, then Total Revenue: PcCQcO PmaQ,,¡0
acc�unhng p 7ofits may b� greater than zero even when economic profits are zero. To Total ACcdQcO ACmbQ1110
avoid confusión, economísts refer to economic profits as rents, which are defined as Cost: PccdACc PmabACm
any payments in_excess of the rr:imimum amounts needed to cover the cost of supply. Rent: O acf
Rents can occur m markets for mputs such as land and capital as well as in product
Deadweight
markets. Figure 4.6 Monopoly Pricing, Rents, and Deadweight Loss
Loss:
. In the real economic world, unlike our ideal competitive model, firms cannot be
mstantaneously replicated; at any time sorne firms may thus enjoy rents. These rents,
h?wever, attract i:iew firm� to the industry, so that over the long run we expect rents to
disappear. Only if sorne circumstance prevents the entry of new firms will the rents $99 /unit. Revenue increases by $90 (10 times $99 minus 9 times $100). Toe he�ght of
persíst. Therefore, we expect the dynamic process of profit seeking to move the econ-
the marginal revenue curve above 10 units is thus $90, which is less than the heíght of
omy toward the competitive ideal.
the demand schedule, $99. As long as marginal revenue exceeds marginal cost (MC),
To 1:1n�e!stand bet�er the concept of rent, it is useful to contrast pricing in a
profits can be increased by expanding output, The profít-maximizing level of output
?1onopohst�c md�stry w_ith one that is competitive. To begin, consider the case of an o�curs when marginal cost equals marginal re_venue (where fv!C intersects A:fR). In
m�ustry with a smgle fírrn that does not have to worry about future competition. Figure 4.6, this output level for the monopoly firm, Qm, results in a market pnce, Pm'
and profíts equal to the area of rectangle PmabACm: total revenue (P m times Q11) minus
T�is monopoly firm sees the entire demand schedule for the good, labeled D in enue i1:creases far each additional unit offered to the market. Toe marginal
Figure 4.6. It also sees a marginal revenue curve (MR), which indicates how much rev- revenue curve lies below the demand schedule because each additional unit
offered lowers
the �quilibr�um price not just far the last but far ali units sold. For example, imagine
total cost (AC times Q ).
that mcreasmg supply from 9 to 10 units decreases the market price from $100/unit to In contr��t to the ��se of monopoly, consider the production decisions of one of
the firms in a competitive industry. Because it provides a small part of the total indus-
try supply, it ignores the effects of its supply on market price and, therefare, equat.es
marginal cost with price (the intersection of MC and D), yielding price Pe and profits
 Chapter 4: Efficiency and ihe Ideolized Competítíve Model 67
66 Part JI: Conceptual Foundaiions far Problem Analysis

PccdACc Toe difference in profit between monopoly and competitive pricing is Graphically, this total cost equals the area under the
the supply curve from quantity zero
monopoly rent, a type of economic rent. t
Remembering that the profits of the firm go to persons, we should take account o
of these rents in our consideration of economic efficiency. A dallar of consumer t
sur- plus (compensating variation) is equivalent to a dallar of distributed economic h
profit. If we set the price and quantity to maximize the sum of consumer surplus e
and rent q
then we will generate the largest possible dallar value in the market, creating u
th� a
prerequisite for a Pareto-efficient n
allocation. t
Toe largest sum of consumer surplus and rent results when price equals i
mar- ginal cost. A comparison in Figure 4.6 of the sums of consumer surplus t
and rent between the competitive and monopoly pricing cases illustrates this y
general proposi- tion. Toe sum in the monopoly case, where marginal cost s
equals marginal revenue (MC = MR), is the area between the demand schedule u
and the marginal cost curve from quantity zero to Qm. Toe sum in the competitive p
case, where marginal cost equals price (MC = P), is the area between the demand p
schedule and the marginal cost curve from quantity zero to QC' Obviously, the l
sum of consumer surplus and rent under competitive pricing exceeds that under i
monopoly pricing by the shaded area between the demand schedule and the e
d
marginal cost curve from Qm to QC' This difference, the area of triangle acf is the
.
deadweight loss caused by monopoly pricing. That this area is the deadweight Suppose the market price is P3 so that the quantity
loss follows directly from the observation that the marginal benefit (D) exceeds the
marginal cost (MC) for each unit produced in expanding output from supplied is Q . Toen the total cost of producing Q3 is the
o, to QC' area P1 aQ30. Toe total revenue to the Arms, however,
equals price times quantity, given by the area of
rectangle P3aQ3 O. The difference between total revenue
and total cost equals the total rent accruing to the firms.
This difference, called producer surplus, is measured by the
Producer Surplus: Measurement with Supply Schedules total shaded area in Figure 4.7
i
We usually deal with markets in which many firms offer supply. Therefore,
we desire sorne way of conveniently summing the rents that accrue to all the firms n
sup- plying the market. Our approach parallels the one we used to estimate s
compensar- ing variations. First, we introduce the concept of a supply schedule. c
Second, we show how the supply schedule can be used to measure the sum of
rents to all firms supplying the market. r
Imagine constructing a schedule indicating the number of units of a good that i
firms offer at each of various prices. Figure 4.7 shows the common case of firms b
facing increasing marginal costs. Firms offer a total quantity Q2 at price P2• As price e
increases, d
firms offer successively greater quantities, yielding an upward-sloping supply
schedule. b
Toe schedule results from the horizontal summation of the marginal cost curves of
the y
firms. (For example, refer MC in Figure 4.5.) Each point on the supply curve tells P
us how much it would cost to produce another unit of the good. If we add up these
3
mar- ginal amounts one unit at a time, beginning with quantity equal to zero and
a
ending at
the quantity supplied, then we arrive at the total cost of producing that quantity. P
1.

.e,g
P2 --------------

P¡�-----

o
Quantity/Time
Loss in producer surplus resulting from a fall in price from P3 to P2: P3abP2

Figure 4.7 A Supply Schedule and Producer Surplus

Producer surplus need not be divided equally among firms. Sorne firms
.may have uní ue advantages that allow them to produce at lower cost than other
flr°:s,
even thoJ hall firms must sell at the same price. For instance, a for�ate f�rmer with
ver roductive land may be able to realize a rent at the market pnce, while anotl�er
far!lr on marginal land just covers total cost. Because the quantity of very
J.?roducttve land is limited, both farmers face rising margin�l c�sts that �ey
equate with mar et rice to determine output levels. Toe general point is that
umque resour�es-such as rspecially productive land, exceptional athletic talent,
easy-to-extr�ct mmerals-can earn rents even in competitive markets. Excess
payments to �uch uruque resources a_re
usually referred to as scarcíty rents. Unlike monopoly, scarciiu rents do not neceeeanlv
imply economic inefficiency. if
Changes in producer surplus represent changes in rents. For example, 1 we
want to know the reduction in rents that would result from a fall in p�ice from P 3 to
P 2, t�en we compute the change in producer surplus in the m�ket.. In Figure 4.7 the
reduction
in rents equals the dark shaded area P3abP2, the raduction in producer surplus.

Social Surplus
We now have the basic tools for analyzing efficiency in sp�cific m_arkets. A
necessary condition for Pareto efficiency is that it should not be pos.s1ble to
mcre�se the suf
ºi
111\
compensating variations and rents through any reallocation of factor inputs or
products. We have shown how changes in consumer surplus measure the
sum o
68 Part JI: Conceptual Foundaiions far Problem Chapier 4: Efficíency and the ldealized Conipetitiue Model
Analysis 69

caveats: Models and Reality

The general equilibrium model helps �s ��ers_tand a comple� v:7o_rld. As is the
ca�e
P¡ with all models, however� it .inv_olves simplifications that may limit its usefulness. It
is
worth noting three such Iimitations. .
Fírst, the general equilibrium model is static rather than_ dynamic,
Re�l
mies constantly evolve with the introduction of new goods, improvement
econº
in
techn@logies, and changes m· �ons�mer tas�es. A n g fea t 1:1re o f the pn·ce sys-
amazm ·

Po t rn is lts capacity for conveymg information among decentralized producers

o 1
o
·e and

�1
1

e, 1
e nsuro.ers about such changes, which Friedrich Hayek refers to as "the
----,----
1
1
particulars
1 tírne and place.:" Toe eguilibria of the competitive frame"".ork give us
1 snapshots rather than videos of the real world. Usually, the snafshot is
1

1
helpful �nd n?t t�o
1
1
roisleading. Nevertheless, policy analysts should realize that substantial gams
1
1
m social welfare result from innovations that were not, and perhaps coul_d not:
1 have been arrticipated. Policy analysts should be careful not to take too statíc a
1
1 view of ma,rkets. Large rents that seem well protected by barriers to entry, for
1
1
example, may
1
very well spur innovative substitutes. In discussing each of the mark�t failures
in
Q¡
ºº
Quantity/Time the next chapter, we consider sorne of the market responses that may anse to
reduce
social welfare
losses.
Second, the general equilibrium model can never be complete: modelers
simply
do not have enough information to incorporate all goods and services. If they
could, then it would be unlikely they could solve the model for its equilibrium. Our
switch-
ing from the general equilibrium II;odel to models of indi_vidual markets i� a
i:urpc:s�-
Figure 4.8 Inefficiencies Resulting from Deviations from the Competitive Figure 4.8 reviews the inefficiencies associated with deviations from the
Equilibrium
equi- librium price and quantity in a competitive market in terms of losses
of social
, surplu_s. Toe. efficien� competitive equilibrium occurs at price P0 and quantity
compensating variations and how changes in producer surplus measure Q0
changes in
rents. Toe sum of consumer and producer surpluses in all markets is defined as
social surplus. C�ai:1ges in social surplus, therefore, measure changes in the sums of
compen-
�atmg va_nabons and r�nts. Por evaluating the efficiency implications of small
changes
m the pnce and quantíty of any one good, it is usually reasonable to limit
analysis to changes in social surplus in its market alone.
ful restriction of the model so that it can be usefully apphed. In most applications that doing so may not fully capture all their implications in the wider
it is a reasonable approach, though sometimes goods are such strong economy. Nonetheless, we see this analysis as highly valuable in helping policy
complements or substitutes that it is not reasonable to look at them separately.5 analysts get started in understanding the complexity of the world in which they
Third, the assumptions of the general equilibrium modelare often violated in work.
the
real world. In the two chapters that follow, we consider the most important of
these violations of assumptions. We do so in the context of specific markets,
acknowledgin� Conclusion
the pomt of íntersection of the (5) and demand (D) schedules. A policy The idealized competitive economy provides a useful conceptual framework
supply that for
increases price to P1 involves a loss of social surplus given by the area of thinking about efficiency. Toe tools of applied welfare economics, consumer and
triangle pro- ducer surplus, give us a way of investigating efficiency within specific markets.
ab�-each forgone unit between Q1 and Q0 yields marginal value (as given by In the next chapter, we explícate four situations, the traditional market failures,
the in which equilibrium market behavior fails to maximize social surplus.
height of the demand schedule) in excess of marginal cost (as given by the
height of
the supply schedule). Consequently, social surplus can be increased by
lowering price so that the guantity supplied and demanded moves closer to Q •
A policy that
0
decreases price to involves a loss of social surplus given by ea of triangle
P2 the
ade-�ach additior:al unit supplied and demanded between and Q2 yields
Q aris in cess of 4F. A. Hayek, "The Use of Knowledge in Society," American Economic Reoieui 35(4) 1945, 519-30 at 522.
0
marginal cost (as g1ven by the height of the supply schedule)
th
at ex 5See Anthony E. Boardman, David H. Greenberg, Aidan R. Vining, and David L. Weimer, Cost-Benefit
marginal benefit (as given by the height of the demand schedule). Analysis: Concepts and Practice, 3rd ed. (Upper Saddle River, NJ: Pearson Prentice Hall, 2006), Chapter 5.
Conseguently, raising price to move the quantity supplied and demanded 6
R. G. Lipsey and Kelvin Lancaster, "General Theory of the Second Best," Review of Economic Studies
closer to Q íncreases 0
24(1) 1956-1957, 11-32.
social surplus.
70 Part JI: Conceptual Foundations far Proolem Analysis

5
For Discussion
1. Assume that the world market for crude oil is competitive, with an upward-slo .
supply schedu�� �d a do:"'nward-sloping demand schedule. Draw a diagrantig
shows.the equ1li�
num pnce and quantity. Now imagine that one of the ma·or �t
exportíng countnes undergoes a revolution that shuts clown its oil fields. drawºil
new supply schedule and show the loss in consumer surplus in the world 0·1 ª
resulting from the loss of supply. What assumptions are you making b market
demand for crude oil in Y?ur meas�ement of consumer surplus? a out the
2. Now assume that the Umted States is a net importer of crude oíl Show th · Rationales for Public
of �he price i1:crea se resulting from the loss of supply to the �orld m:rrpact
social surplus in the U.S. market. et on
Policy
Market Failures

T he idealized competitive model produces a Pareto-efficient allocation of

goods. That is, the utility-maximizing behavior of persons and the profit-
maximizing behavior of firms will, through the "invisible hand," distribute goods
in such a way that no one could be better off without making anyone else worse
off. Pareto efficiency thus arises through voluntary actions without any need for
public policy. Economic reality, however, rarely corresponds perfectly to the
assumptions of the idealized competitive model. In the following sections we dis-
cuss violations of the assumptions that underlie the competitive model. These
violations constitute market failures, that is, situations in which decentralized
behavior <loes not lead to Pareto efficiency. Traditional market failures are shown
as circumstances in which social surplus is larger under sorne alternative alloca-
tion to that resulting under the market equilibrium. Public goods, externalities,
natural monopolies, and information asymmetries are the four commonly recog-
nized market failures. They provide the traditional economic rationales for public
participation in prívate affairs.
71