Updated: 6 July 2020

ECO 602: MICRO ECONOMICS

Lecture 3

Topics to be covered:
a. Definition of Economic Efficiency
b. Efficiency in Production
c. Efficient Choice of Inputs for a Single Firm
d. Efficient Allocation of Resources among Firms
e. Efficient Choice of Output by Firms
f. Theory of Comparative Advantage
g. Efficiency in Product Mix
h. Competitive Prices and Efficiency
i. Efficiency in Production
j. Imperfect Competition
k. Externalities
l. Distribution of Economic Welfare
m. Exchange with Initial Endowments
 ECO 602: ADVANCED MICRO ECONOMICS
The Efficiency of Perfect Competition
Nicholson Chapter 13 Cont.

Adam Smith put forth the idea that a competitive market would ensure that resources would find
their way to where they were most valued (the invisible hand). He provided the framework that
has given rise to modern welfare economics.

The fundamental Theorem of Welfare Economics states that there is a close correspondence
between the efficient allocation of resources and the competitive pricing of these resources.

Definition of Economic Efficiency – Pareto Efficient Allocation

An allocation of resources is Pareto efficient if it is not possible (through further reallocations) to

make one person better off without making someone else worse off.

Efficiency in Production

Production Efficiency: An allocation of resources is efficient in production (or technically

efficient) if no further reallocation would permit more of one good to be produced without
necessarily reducing the output of some other good.

Y Production possibilities frontier.

B

A C A = technically inefficient
B & C = technically efficient

Technically efficiency is a precondition for overall Pareto efficiency. However, technically

efficiency does not guarantee Pareto efficiency. The firm may be producing the wrong goods
efficiently.

1
Production efficiency: There are three aspects:
1. Resource allocation within a single firm
2. Allocation of production resources among firms
3. Coordination of firms’ output decisions.

1. Resource Allocation within a single firm

Efficient choice of inputs for a single firm that produce two goods X and Y. We showed this
previously with the Edgeworth Box diagram by assuming the following:
i. firm produces two outputs X, Y.
ii. fixed amounts of inputs K, L .
iii. production function for X = f (KX, LX).
iv. assume full employment of both factors, hence, KY = K – KX and LY = L – LX
v. production function for Y = g (KY, LY) = g ( K – KX, L – LX)

Technical efficiency requires that the output of X be maximized for any given level of Y eg. Y .
Setting up the Lagrangian equation for this problem yields

L=f (Kx, Lx) +  ( Y – g ( K – KX, L – LX))

and the first order conditions for a maximum are


(1) = fK + gk = 0
K X

(2) = fL + gL = 0
L X

(3) = Y – g ( K – KX, L – LX) = 0


fK g
From (1) and (2), we get = K
fL gL

RTSX (K for L) = RTSY (K for L)

2
2. Efficient Allocation of Resources Among Firms that produce same good X

It is necessary for overall efficiency. Resources should be allocated to those firms that can use
them most efficiently.

Assume that
i. two firms produce X
ii. production functions are f1 (K1, L1) and f2 (K2, L2)
iii. total supplies of capital and labor are given by K and L

Allocation problem is to maximize

X = f1 (K1, L1) + f2 (K2, L2)

subject to:
K1 + K2 = K
L1 + L2 = L

Substituting the constraints into objective function, maximization problem becomes

X = f1 (K1, L1) + f2 ( K – K1, L – L1)

First order conditions for a maximum are

,
X f f f f
= 1 + 2 = 1 - 2 =0
K 1 K 1 K 1 K 1 K 2

because K1 + K2 = K 
f 2 f
=− 2
K 1 K 2
X f f f f
= 1 + 2 = 1 - 2 =0
L1 L1 L1 L1 L2
or
f 1 f 2 f 1 f
= and = 2
K 1 K 2 L1 L2

As a result, the marginal physical product of any resource in the production of a particular good
must be the same no matter which firm produces the good.

3
Example: Gains from Efficiently Allocating Labor
To examine the quantitative gains in output from allocating resources efficiently, suppose two
rice farms have production functions of the simple form
q = k 1 / 4l 3 / 4 ,

but one rice farm is more mechanized than the other. If capital for the first farm is given by k1=16
and for the second farm by k2 =625, then

q1 = (16)1 / 4 l13 / 4 = 2l13 / 4

q2 = (625)1 / 4 l 23 / 4 = 5l 23 / 4

If the total labor supply is 100, an equal allocation of labor to these two farms will provide total
rice output of
Q = q1 + q 2 = 2(50 ) 3 / 4 + 5(50 ) 3 / 4 = 131 .6

The efficient allocation is found by setting the marginal productivities equal:

q1  3  −1 / 4 q  15 
MPL1 = =  l1 = MPL2 = 2 =  l2−1 / 4
l1  2  l2  4 

Hence, for efficiency labor should be allocated so that

 3  −1 / 4  15  −1 / 4
 l1 =  l2
2  4

Given the greater capitalization of farm 2, practically all of the available labor should be devoted
to it. With 100 units of labor, 97.4 units should be allocated to farm 2 with only 2.6 units to farm
1. In this case total output will be:
Q = q1 + q2 = 2( 2.6)3 / 4 + 5(97 .4)3 / 4 = 159 .1

This represents a gain of more than 20 percent over the rice output obtained under the equal
allocation.

4
If capital were not fixed in this problem, marginal productivities of both capital and labor should
be equal for firm 1 and firm 2.

Efficient Choice of Output by Firms

Even though resources may be allocated efficiently both within the firm and between firms, it
must be the case that firms also must produce efficient combinations of outputs.

Output
of RPT1A=1.5/1 Output RPT B0 = 1/1
Cars of
Cars

RPT0A=2/1 RPT B1 =1.5/1

Output of Trucks
Firm A Output of Trucks Firm B

2 1
Initially assume that RPTA0 =  RPTB0 = .Hence firm A should produce more cars and firm B
1 1
should produce more trucks. In this way RPTA0 will fall and RPTB0 will increase so that they will
1 .5 1 .5
be equal i.e. RPTA1 = = RPTB1 =
1 1

Theory of Comparative Advantage

First proposed by David Ricardo: Countries should specialize in producing goods of which they
are relatively more efficient producers.

Suppose we now have two countries A and B,

5
Y(Output of Cars)
RPT1A=? Y(Output of Cars)
RPTB0=1/1
Y 1A

Y0B

Y 0A
RPT0A=2/1 Y1B RPTB1=?

X1A X0A X(Output of Trucks) X0B X1B X(Output of Trucks)

Country A Country B

Both countries would gain if they produced more of the good for which they have a comparative
advantage. Country A should produce more of Y and country B should produce more of X.
Country A would be able to sell some of the Y produced to country A in exchange for purchase
of X. Both countries can be made better off.

Initially country A can give up one unit of X and produce 2 units of Y. For country B at its initial
position it can be produced one more X for one unit of Y. Hence, if A were to produce one unit
less of X and hence produce two more units of Y, it could trade with Country B one of those units
of Y for one unit of X and it would be still better off by one unit of Y. At the same time, country
B would be left no worse off. World welfare would be improved.

Note that if a linear production possibility function existed in both countries, to maximize world
welfare each country should completely specialize in the production of one good. Not the case if
production functions are concave.

Efficiency in Product Mix

Technical efficiency is a necessary but not a sufficient condition for Pareto efficiency. The right
mix of goods must be produced. There is a need to bring individual’s preferences and production
possibilities together.

Assume that a society with 2 goods and all individuals have identical preferences.

6
 Y

P
F

E
G
U2
U1
U3
H

P X

Along PP technically efficient maximum utility is at E where MRSXY = RPTXY that is the
required condition for overall efficiency.

Note that G which is an “inefficient technical” point is preferred to the “efficient technical” point
H.

A mathematical proof
To demonstrate this result mathematically, assume again
i. there are two goods X and Y
ii. society where all individuals have identical preferences has a utility function U(X, Y)
iii. production possibility frontier written implicitly as T(X, Y) = 0

Setting up the Lagrangian expression for this problem yields

L = U (X, Y) +  [T(X, Y)]

and the first order conditions for an interior maximum are
L U T
(1) = + =0
X X X
L U T
(2) = + =0
Y Y Y
(3) ∂L/∂λ = T(X, Y) = 0

U
Combining (1) and (2), we get
X
/ UY =
T
X
/ TY
or
dY
MRS (X for Y) = – (along T) = RPT (X for Y)
dX

7
Example: Utility Maximization Product Mix

Suppose that production possibilities frontier for X and Y is given as X2 + 4Y2 = 100 and that
utility function is given as U (X, Y) = XY . The problem is to maximize utility subject to PPF
constraint.

Setting up the Lagrangian expression yields,

L= XY +  (100 – X2 – 4Y2)  L=X0.5 Y0.5 + (100-X2-4Y2)

L
(1) = 0.5 X-0.5 Y0.5– 2X = 0
X
L
(2) = 0.5 X0.5 Y-0.5– 8Y = 0
Y
L
(3) = 100 – X2 – 4Y2 = 0

Dividing (1) by (2), we have

Y 1 X
 =  X2 = 4Y2
X 4Y
Now, substituting X2 = 4Y2 into production possibilities frontier X2 + 4Y2 = 100, we get
8Y2 = 100
Y2 = 12.5
Y = 3.54 and correspondingly
X2 + X2 =100,
2X2 =100, X2 = 50 and
X= 50
X = 7.07
Therefore, maximum utility is= 12.5 50 = (625)1/4=5

Competitive Prices and Efficiency : The First Theorem of Welfare Economics.

Pareto efficient allocation of resources requires that the rate of trade off between any two goods
X and Y should be the same for all agents (both producers and consumer’s). In a perfectly
competitive economy, PX/PY provides this common rate of trade off to which all agents wil
adjust. Because prices are treated as fixed parameters both in individuals’ utility maximizing
decisions and in firms’ profit maximizing decisions, all trade off rates between X and Y will be
equalized to the rate at which X and Y can be traded in the market (PX/PY). Also, because all

8
agents face the same prices, all trade off rates will be equalized and an efficient allocation will be
achieved.

Hence, MRSXforY = RPTXforY = PX/PY.

Efficiency in Production

Competitive pricing of inputs K and L, with wage rate “w” and cost of capital “v”. The firm to
w
minimize costs will need to the rate of technical substitution of inputs set RTSKL =
v

w
Also firm produces two goods X and Y, RTSXK,L = RTSYK,L =
v
We learned in ECON 501.

Profit maximizing firm will set the value of the marginal product of labor equal to wage rate:
PXfL = w

This holds for all firms , so we have PXf1L = PXf2L = w. What is more, Px is same for all firms,

Hence, f1L = f2L

Consequently, for a given wage rate each firm has same marginal product in the production of x
for a given factor of production.

MC X
Also, RPTXforY = and PX = MCX and PY = MCY
MCY
MC X PX
Hence, = Market prices give signals for efficient production.
MCY PY
MC X
Efficiency in Product Mix requires RPTX,Y = = MRSXY
MCY

Competitive prices are set so that demand = supply

9
 Y

P
F

Y*
E

U0
− P* X
Slope =
P *Y

X* P X

Only combination (X*, Y*) is a Pareto optimal. A competitive price ratio of P*X/P*Y will lead
the economy to this Pareto efficient solution.

Only at this price ratio will supply be equal to demand for both commodities. Equilibrium will
occur at an efficient product mix, and with a Pareto efficient output mix. This is the fundamental
theorem of welfare economics:
Adam Smith argued that laissez fare policies would lead to this result. However, market
distortions will lead us to different results.

Example: Efficiency and Inefficiency

The efficiency of competitive pricing can be shown with the simple general equilibrium model
discussed in the previous lecture.

The base case allocation ( x * = y * = 50 ) is technically feasible in both of the case when technical

progress took place or tastes changed, but it is not the best use of resources except in the base
case. For the situation where there is a technical progress in the production of good x (shift in
supply), the base case allocation lies inside the PPF. Another way to see this is, if we hold y*
constant at 50 , it is possible to produce x*= 2 50 once the technical in good x is taken into
account. Opting for the base case allocation would forgo a substantial amount of x.

10
 Y

U= 8.5 C

3 C

B
A
U=7.07 U=10 C

2 X

0.5 0.5
U = X 0.5Y 0.5 = 50 50 = 50 = 7.07

The base case allocation would also be inefficient when preferences shift toward good y (change
in demand). With the new utility function, the base case would yield:

0.5
U = X 0.5Y 0.5 = (2 50 ) 0.5 50 = 2 50 = 100 = 10

Alternatively, if tastes change as U= X0.1Y0.9, the optimal allocation ( x * = 10 , y = 3 10 ) yields

utility of:

0.1
U ( x, y) = x 0.1 y 0.9 = 10 (3 10) 0.9 = (10) 0.05 30.9 (10) 0.45 = 30.9100.5 = 8.50

Clearly, efficiency requires that preferences and technology be tied together properly. Utility is
increased follow 7.07 it was before the technical change.

11
Note: Very Important Example

The excess burden of a tax: Suppose government places a 200% tax on good x, but maintains
purchasing power by rebating the tax proceeds to consumers in a lump sum. To model this tax,
we let Px/Py be the price ratio without the tax – this is the ratio firms see. Consumers, on the other
hand, see a price ratio of 3 Px/Py - that is, they must pay the firm Px and the government 2Px
whenever they buy a unit of good x. Now equilibrium is described by:
RPT = Px/Py = x / y (supply)
and
3 Px/Py = y / x =MRS (demand)

3 Px y
MRS= =
Py x

Px y
=
Py 3x
MRS = RPT (PPF: x2 + y2 = 100)
y x
=
3x y
Hence, x / y = y / 3x or y2=3x2. Substituting this into original PPF (x2 + y2=100) yields the
following after- tax equilibrium:

x2 + 3x2=100
x2 = 25, therefore x* = 5 , and
25 + y2 =100

y2 = 75, therefore y* = 5 3 , Also

PX 5 1
PY = =
5 3 3

12
 Y

5√3

√50 Px/Py
E
3Px/Py
U0
U1

√50 X
5

Post tax utility in this situation is:

U ( x, y ) = x 0.5 y 0.5 = 50.5 (5 3) 0.5 = 5(3) 0.25 = 6.58

The reduction in utility from 7.07 to 6.58 is a measure of the excess burden of this tax. Here,
because tax proceeds are rebated to consumers, there is no other burden of this tax. The welfare
loss arises solely because the tax discourages x consumption by creating a wedge between what
consumers pay for the good and what producers receive for it.

Types of Market Distortions:

1. imperfect competition
2. externalities
3. public goods

1. Imperfect Competition:
Imperfect competition is the situation where economic agents exert some market power in
determining the price.

Suppose X is produced under imperfect competition where MRX<PX* but Y is produced under
competition where MRY = PY*
MC X MR P* X
RPTXforY = = * X < * = MRSXforY
MCY PY P Y

13
 Y
P
Y1 MRX
B Slope = –
MRY
Y0 E

U1 U2
− P* X
Slope =
P *Y

P
X 1
X0 X

As a consequence, X1 < X0 and Y1 > Y0.

Monopoly has the same effect as a tax
2. Externalities

Pollution is not included in the market prices that agents are reacting to. We need to internalize
the extra social costs and equate the Social Rate of Transformation in production with the Social
Marginal Rate of Substitution in Consumption.

3. Public Goods
Non-rival or non-exclusive in consumption.
We will consider this later in more detail.

Distribution of Economic Welfare

Pareto efficiency can occur with many different distributions of welfare among individuals. Many
of these distributions of income may be very inferior to others if we were to measure it according
to some “social” welfare functions.

To study distribution in its simplest way, assume that

there are only two individual Smith and Jones, and

two goods X and Y to be distributed among these people are in fixed supply.

14
Analyze with Edgeworth Box Diagram

OJ

UJ1

UJ2
M4

UJ3 US4
Total Y

M3

UJ4 M2 US3

M1 US2

US 1
A
OS Total X

All points along the contract curve OSOJ are Pareto Efficient. MRSS = MRSJ. Any point within
the Edgeworth box in which the MRS for Smith is unequal to that of Jones offers an opportunity
for a Pareto improvement adjustment.

Point A is Pareto inefficient with US1 and UJ3 as Smith could be made better off by moving to M2,
(US2) without making Jones worse off i.e. he still would be on UJ3. Or alternatively Jones could
be made better of by moving to M1 where he would have utility of UJ4 while Smith would be left
with US1.

When the marginal rates of substitution for the two individuals are equal, such mutually
beneficial trades are not available.

Other points such as M3 and M4 would make at least one of the individuals worse off.

Along the contract curve, OSOJ, the individual’s preferences are rival in the sense that one
individual’s situation may be improved only if someone else is made worse off.

Exchange with Initial Endowments

Suppose the individuals in the exchange possess specific quantities of the goods.

15
Suppose initial endowment is the quantities associated with point A. Smith has Sx and Sy and
Jones has Jx and Jy.

OJ

UJA
Total Y

Jy
M2
M1

USA
A
Sy

OS
Sx Jx
Total X

Neither Smith nor Jones will accept trading outcomes that give utility less than USA or UJA. It
would be preferable for an individual to refrain from trading rather than accept an inferior
outcome.

Thus we should focus only on those combinations between M1 and M2 on the contract curve in
our study of free exchange not the entire contract curve OSOJ.

The range of efficient outcomes for voluntary exchange has been greatly narrowed by the
distribution of the initial endowments.

If the initial distribution favors one party, the final allocation will favor the same party.

Voluntary transactions can not overcome large differences in initial endowments. Hence, the
individual endowments will affect the total welfare in society as measured by the aggregation of
individual utilities.

It is a very slow and difficult process to change the distribution of income within a society.

Example: A Two Person Exchange Economy

16
Consider an exchange economy in which there are exactly 1,000 soft drinks (x) and 1,000
hamburgers (y). If we let Smith’s utility be represented by

U s = ( x s , y s ) = x s2 / 3 y1s / 3
and Jones’s utility by
U j = ( x j , y j ) = x1j / 3 y 2j / 3

We can compute the efficient ways of allocating soft drinks and hamburgers. We might expect
from given utility functions that efficient allocations would give relatively more soft drinks to
Smith and relatively more hamburgers to Jones.

Suppose we let Smith start at any preassigned utility level, Ūs. Our problem now is to choose xs,
ys, xj and yj to make Jones’s utility as large as possible given Smith’s utility constraint. Setting up
the Lagrangian for this problem yields
 − 
L = U j ( x j , y j ) +  U s ( x s , y s ) −  s 
 
 − 
= x1j / 3 y 2j / 3 +   x s2 / 3 y1s / 3 −  s .
 
Remember that Jones simply gets what Smith doesn’t, and vice versa. Hence
x j = 1000 − xs

and
y j = 1000 − y s .

Our Lagrangian is therefore a function of only the two variables xs and ys ,

−
L = (1000 − xs )1/3 (1000 − ys ) 2/3 +  ( xs2/3 y1/3
s − s ).

The first order condition for a maximum are

2/3 1/ 3
(1) L 1  1000 − y s  2  y s 
= −   +   =0
xs 3  1000 − xs  3  xs 

17
 1/ 3 2/3
(2) L 2  1000 − xs    xs 
= −   +   =0
y s 3  1000 − y s  3  y s 

Dividing (1) by (2) gives us the following equation for the contract curve.

1  1000 − y s  y 
  = 2 s 
2  1000 − x s   xs 
or

xs  4 y s  →This is the equation of the contract curve

=  
(1000 − xs )  1000 − y s 

This is our required condition for efficiency. We can use above equation to calculate any number
of Pareto efficient allocations. For example, if we arbitrarily choose Xs=Xj =500, by using
contract curve equation, we find that Pareto efficient outcome requires Ys=200 and Yj=800.

To illustrate why points off this contact curve are inefficient, consider an initial allocation in
which Smith and Jones share X and Y equally (this allocation is not on the contract curve). With
500 units of each item, both Smith and Jones receive a utility of 500, Us= (500)2/3(500)1/3 and
Uj=(500)1/3(500)2/3. But, we can find using the equation for the contract curve that there are many
allocations on the contact curve that offer more than utility of 500 to both people. Consider Xs
=660, Ys =327, Xj =340, and Yj =673. For this allocation Smith’s utility is
Us=(660)2/3(327)1/3=522 and Jones’s is Uj= (340)1/3(673)2/3= 536. Both are clearly better off than
at the initial allocation, and one might expect some sort of trading to take place that moves them
toward the contact curve.

Effects of Initial Endowments: To see how initial endowments may restrict the range of Pareto
efficient solutions in this economy, suppose Smith starts in a very favorable position with Xs
=800, Ys = 800. Then Jones gets Xj =200, Yj =200, and the initial utility levels are Us=800 and
Uj=200. There are Pareto improvements that might be made from these initial endowments, but
none of them will improve Jones’s situation very much. For example, if we hold Smith’s utility at
800, the efficient allocation Xs =884, Ys =657, Xj =116, Yj =343 (calculations provided by text
book) will increase Jones’s utility from 200 to 239. But that is the best that Jones can do given the

18
constraint that Smith cannot be made worse off. The efficiency gains to Jones, while significant,
do very little to move the overall allocation toward more equal outcomes.
The End

19