General Equilibrium in Pure

Exchange Economies
 Barter/ Exchange Economies
• We will start with equilibrium allocations in
economies without production (so-called "barter
equilibrium” for pure exchange economies)
• Let consumer 1 be initially endowed with
units of good 1 and 2, respectively.
– Similarly, consumer 2’s endowment.
• We depict the initial endowment and any
other allocation thru the Edgeworth box.
Explain orientation of axis for each consumer, Indifference curves
and endowment e
 General Equilibrium in an exchange
economy
• Movement from the initial endowment e to
allocation A cannot be a barter equilibrium, since
consumer 1 is worse off at A, thus being able to
block allocation A.
• Only points in the lens-shaped area
can be accepted as equilibria by both consumers
• This lens shaped area (later to be called the core
of the economy) is:

– ff condition holds:
Edgeworth Box
Other points in the lens/ core
• Q: Is point B, which belongs to
this lens-shaped area, a barter
equilibrium?

• A: No! other points, such as D,

would still make both consumers
better off (i.e. higher IC) than at B.

• Any point on the cc curve would

be an equilibrium (no Pareto
improvements are possible).

• The cc curve is the "contract

curve"
 Some Definitions
• Feasible allocation – an allocation x=(x1,
x2,…xI) for I consumers is said to be feasible if

• Pareto-Efficient Allocation – A feasible

allocation in which there is no other feasible
allocation yi which is w.p. by all consumers iεI
over xi (i.e. yi w.p. xi), and at least strictly
preferred by one consumer (i.e. yi s.p. xi)
 Some definitions
• Blocking coalitions – A subset S of I
consumers will block a feasible allocation x if
there is an allocation y that is:
1. Feasible:

2. yi w.p. xi for all i a member of S with at least one

individual also a member of S yi s.p. xi
 Equilibrium and the Core
• An equilibrium in an exchange economy is
therefore a feasible allocation x, given
endowment e, such that no coalition of consumer
blocks x

• Core [C(e)] of the economy is the set of all

unblocked allocations given e, i.e. set of equilibria
points defined earlier
– This is the lens shaped area around e (see previous
slide)
 Some notes on the Equilibrium in a
Pure Exchange Economy
• Equilibrium is represented by the tangency of
the indifference curves of the two consumers
• All points in the contract curve are pareto
optimal (they are also non-comparable)
• If the local non-satiation condition is not
satisfied, the equilibrium allocation
represented by point D may not be Pareto
optimal, say, for the case of thick indifference
curves
 Some notes on the Equilibrium in a
Pure Exchange Economy
• Because point D is a tangency between two
indifference curves then it satisfies two
conditions:
 Some notes on the Equilibrium in a
Pure Exchange Economy
• Condition (a) is a necessary condition for
equilibrium
• Condition (a) is also a sufficient condition with
our assumption of strict quasiconcavity (i.e.
unique solution)
General Equilibrium in Competitive
Exchange Economies
 Assumptions about Consumers
• Consumer utility function is continuous,
stritctly increasing, and strictly quasiconcave
• Max Ui s.t. pxi l.e. pei for al p>>0 (i.e. strictly
positive prices)
• Has a unique solution x(p, pei) which is
continuous in p
– Walrasian demand correspondence
 Excess Demand Function
• Given x(p, pei) we can define the excess
demand function for good k zk(p) as:
– (write on board)
• zk(p)>0, implies that aggregate demand for
good k exceeds the aggregated endowment for
good k
• zk(p)<0, implies that aggregate demand for
good k is smaller the aggregated endowment
for good k, i.e. excess supply of good k
 Properties of zk(p)
• z(p) is continuous in p
• Homogeneity z(ap)=z(p) for all a>0
– Follows from homogeneity in Walrasian demand
in p
• Walra’s Law: pz(p)=0
– (Write proof on board)
• Implications of Walra’s law in a two good
economy:
– p1z1(p)+p2z2(p)=0
– p1z1(p)=-p2z2(p)
• if there is excess demand in market 1, z1(p) > 0,
there must be excess supply in market 2, z2(p) < 0
• Similarly, if market 1 is in equilibrium, z1(p) =0,
then market 2 is also in equilibrium, z2(p)=0
• For n markets, if markets of n-1 goods are in
equilibrium, then so is the nth market, i.e. you
only need to look at n-1 markets.
 Walrasian equilibrium and conditions
for it’s existence
• Is a vector of prices p* in Rn++ is a walrasian
equilibrium if z(p*)=0, i.e. the aggregate excess
demand is zero at p*
• The walrasian equilibrium exists if
1. z(p) is continuous for all p>>0 in Rn++
2. pz(p)=0; i.e. Walra’s law holds for all p>>0 in Rn++
3. If {pm} is a sequence of strictly positive price vectors,
pm in Rn++, converging to pbar not equal to 0, and pbark = 0
for some good k, then for some good k’ with price pbark’ =
0, the associated sequence of excess demands in the
market for good k’, {zk’(pm)}, is unbounded above
 What the conditions say?
• First two conditions follow from the
assumptions we imposed on consumer’s utility
functions
• Third condition intuitively says that, if the
prices of some but not all goods are arbitrarily
close to zero, then the excess demand for at
least one of those goods is arbitrarily high
– Use to prove that p*>>0
 Proof of Existence of WEA
• Use Brouwer’s fixed point theorem
– For a compact, convex and non-empty set S, Let f:
S->S (i.e. f maps S unto itself). Then there exist at
least one x* in S such that f(x*)=x* and x* is
called the fixed point of f
• Proof .. Write on board
WEA in the Edgeworth box
Alternative Graphical Interpretation
• Another way to see CE is through the
concept of the offer curve

• Offer Curve - locus of optimal

consumption for varying prices (red
for consumer 1; blue for consumer 2)

• CE is where the offer curves of each

consumer intersect

• Offer curves always pass through the

initial endowment point e

• Intersection aside from the initial

endowment point is the CE point
 Some Unusual Cases and Non-
Existence of Equilibria
• Linear Indifference
curves ( perfect
substitutes)
• There would still be
a competitive
equilibria but
marginal rates of
substitutions will not
be equal
 Some Unusual Cases and Non-
Existence of Equilibria
• Consumer 1’s preferences
have nonconvex portions

• As a consequence there is a
break in his offer curve
(see blue line)

• Therefore Offer Curves

will not meet aside from
that at the initial
endowment therefore no
equilibria
WEA and the Core
 Relationship between WEA and the
Core
• If each consumer’s utility function is strictly
increasing, then every WEA is in the Core C(e)
• Proof: (by contradiction)
 Some implications
• The Core C(e) will always be non-empty, it
should contain at least the WEA
• Since all core allocations are Pareto optimal
(i.e. no blocking coalition will form), then all
WEA that are part of C(e) is pareto optimal
(First Welfare Theorem)
• Previous proof is also a proof of the First
Welfare Theorem
Second Welfare Theorem
Suppose you have the following
situation:

From initial endowment e, the

associated WEA is x’. Note x’ is
pareto optimal because it is in the
core C(e)

However, society prefers xbar which

is outside the C(e)

What to do? Society can alter

endowments to e”; such that
p*e”=p*xbar
After “transfers” let the market
work and would lead to new WEA
xbar