Lecture 14 - General Equilibrium in a Pure

Exchange Economy

14.03 Spring 2003

1 General Equilibrium in a Pure Exchange Econ-

omy
1.1 Motivation
• To now, we’ve talked about 1 market at a time: labor, sugar, food, etc.
• But this tool is a convenient fiction. Not always a badly misleading fiction.
But still a fiction.
• Markets are always interrelated:

— Reduce sugar tariffs → reduce sugar prices

— → Drop in employment of sugar cane farmers
— → Cane workers apply for other farm jobs, depress wages for farm
workers generally
— → Arable land is freed for other uses
— → Gives rise to new crop production
— → Prices of other farm products fall
— → Real consumer incomes rise
— → Rising consumer income increases demand for sweets, a luxury
good
— → The dessert market grows and the café sector booms
— → Starbucks buys up all urban real estate in four major cities
— ...there is literally no end to this chain of events.

• At some level, there is no such thing as the market for a single good.
• All changes in quantities or prices feed back into the demand and/or supply
for other goods through:

 1. income effects
2. substitutability/complementarity of goods whose prices rise/fall
3. changes in the abundance/scarcity of other resources

• To understand this story, we need a model that can accommodate the

interactions of all markets simultaneously and determine the properties of
the grand equilibrium. What we need to develop is a general equilibrium
model, in contrast to the partial equilibrium models we’ve used thus far
this term.

1.2 The Edgeworth Box

• We need to reduce the dimensionality of the ‘all markets’ problem to
something analytically tractable. But we need to retain the essence of the
problem.
• The Edgeworth Box (after Jevons Edgeworth) allows us to do this. It turns
out that we only need 2 interacting markets to see the entire problem:

— 2 goods
— 2 people
— Pure exchange. (No production in this first model, though we will
add this.)
— Perfect expression of the economic concept of opportunity costs.

• Simple as this model is, it demonstrates two of the most fundamental

results in economics: the 1st and 2nd welfare theorems.

1.2.1 Figure 1: Edgeworth box, pure exchange

• Two goods: call them food F and shelter S.
• Two people: call them A and B.
• The initial endowment:
F S
EA = (EA , EA )
F S
EB = (EB , EB )

• Their consumption:
F S
XA = (XA , XA )
F S
XB = (XB , XB )

• Without trade:
XA = EA
XB = EB

2
• With trade, many things can happen, but the following must be true:
F F F F
XA + XB = EA + EB
S S S S
XA + XB = EA + EB

• Note the elements of this figure:

1. All resources in the economy are represented.

2. The preferences of both parties are represented.
3. The notion of opportunity costs is clearly visible.

• Starting from point E, the initial endowment, where will both parties end
up if they are allowed to trade?
• It is not fully clear because either or both could be made better off without
making either worse off. But it’s clear that they need to be somewhere in
the lens shaped region between UA0 and UB0 .
• How do we know this, because all of these points Pareto dominate E :
one or both parties could be made better off without making the other
worse off. There are potential gains from trade.
• A would prefer more food and less shelter, B would prefer less food and
more shelter.
• So hypothetically
S S
A gives up EA − XA
F F
A gains XA − EA
F F
B gives up XA − EA
S S
B gains EA − XA

• Are all points in the lens shaped region Pareto efficient?

• No. All of the points in the lens region are Pareto superior, but only a
subset are Pareto efficient.
• Q: What needs to be true at a Pareto efficient allocation?
• A: The indifference curves of A, B are tangent.
• Why? Otherwise, we could draw another lens.
• So trading should continue until a Pareto efficient allocation is reached.
• Pareto efficient allocation:

1. No way to make all people better off.

3
 2. Cannot make 1 person better off without making at least 1 other
person worse off.
3. All gains from trade exhausted.

• At a Pareto efficient allocation, the indifference curves of A, B will be

tangent.
• The set of points that satisfy this criterion is called the Contract Curve
(CC). All Pareto efficient allocations lie along this curve.
• We know that after trade has occurred, parties’ set of choices will lie some-
where on CC that passes through the lens defined by the points interior
to UA0 and UB0 .
14#1
ESB
B
Food

E
EFA EFB
UA0
U B0

A
ESA
Shelter

1.3 How do we get from E to a point on the contract

curve?
• Famous analogy: Auctioneer (Leon Walras → Walrasian auctioneer).

1. In the initial endowment, the market clears (that is, all goods con­
sumed) but the allocation is not Pareto efficient.
2. So, an auctioneer could announce some prices and then both parties
could trade what they have for what they preferred at these prices.

 3. Problem: Choices would then be Pareto efficient but would not nec­
essarily clear the market.
4. It’s possible there would be extra F and not enough S or vice versa.
5. So, must re-auction at new prices...

• See Figure 2.
14#2

Food

UB

UA

Proposed price

Shelter
At proposed prices:

— A wants to reduce consumption of shelter increase consumption of

good
— B wants to increase consumption of shelter decrease consumption of
food
— But, A wants to increase consumption of food more than B wants to
decrease
— A wants to decrease consumption of shelter more than B wants to
increase
— So:
F F F F
XA + XB > EA + EB ⇒ Excess demand
S S S S
XA + XB < EA + EB ⇒ Excess supply

• How do we know that it is inefficient for some of the shelter to go unused?

5
 • What should auctioneer do? Raise PF /PS .
• When the auctioneer gets the price ratio correct, the market clears. No
excess demand or supply for any good.
• This is a market equilibrium, competitive equilibrium, Walrasian equilib­
rium, etc:

1. Each consumer choosing his most preferred bundle given prices and
his initial endowment.
2. All choices are compatible so that demand equals supply.
3. Pareto efficient consumption (‘Allocative Efficiency’):
µ ¶ µ ¶
∂U/∂F ∂U/∂F
=
∂U/∂S A ∂U/∂S B

• Q: How do we know Allocative Efficiency will be satisfied?

1. Because both A, B face the same prices PF /PS .

2. Each person’s optimal choice will therefore be the highest indifference
curve that is tangent to her budget set given by the line with the slope
PF /PS that intersects E.
3. Because these choice sets (for A, B) are separated by the price ratio,
we know they will be tangent to one another but will not intersect.
(The set of prices forms a ‘separating hyperplane’ that divides the
indifference maps of the two consumers).

• This equilibrium price ratio will exist provided that:

1. Each consumer has convex preferences (diminishing marginal rate of

substitution) as we assumed during consumer theory.
2. Or, each consumer is small relative to the aggregate size of the market
so that aggregate demand is continuous even if individual preferences
are not. (This is obviously not relevant in the two person case rep­
resented by the Edgeworth box.)

1.4 Aside: How do we know that both (all) markets clear

simultaneously?
• How do we know that both (all) markets clear simultaneously?
• Consider two goods X, Y and two individuals A, B.
• Label A0 s demand and supply of each good as DxA , DyA , SxA , SyA and simi­
larly for consumer B.

 • Consumer A0 s budget constraint can be written as:

Px DxA + Py DyA = Px SxA + Py SyA ,

Px (DxA + SxA ) + Py (DyA + SyA ) = 0,
Px EDxA + Py EDyA = 0,

where EDxA is A0 s ‘excess demand’ for good X. Note that excess demand
can be positive or negative (so more precisely, excess demand or excess
supply).
• Note that the excess demand is the amount of good A would like to con­
sume relative to her current endowment (her ‘supply’).
• This equation states that given an initial supply (endowment) of goods
and a set of prices, an individual’s total excess demand for goods is zero.
Simply put, a consumer cannot buy more than the value of the goods she
holds, since the value of these goods is her budget constraint.
• A similar budget holds for consumer B:

Px EDxB + Py EDyB = 0.

• Putting these excess demand functions together,

Px (EDxA + EDxB ) + Py (EDyA + EDyB ) = Px EDx + Py EDx = 0.

and Px EDx = 0 ⇒ Py EDx = 0

Which is to say, that there cannot be either excess demand or excess supply
for all goods simultaneously.
• This observation — that total excess demand must equal zero — is called
Walras’ Law (after Leon Walras who first provided this proof).
• Hence, if there are n goods, and there is no excess demand for n − 1 of
these goods, then there is also no excess demands for the nth good.
• In our two-good exchange economy above, this proves that if the market
for food clears with no excess demand or excess supply, then the market
for shelter clears simultaneously.

1.5 How are equilibrium prices set? First Welfare Theo-

rem
• You do not need the auctioneer! Leon Walras proved that the mar­
ket can reach this equilibrium without assistance from a central planner
(auctioneer). “The self-organizing economy.” Process of Tattonment —
translation ‘groping.’ This is a fundamental result. [I will not prove this.
See your book.]

7
 • In partial equilibrium analysis, we have taken prices as exogenous. At the
individual level, this is true. For all practical purposes, my preferences do
not influence the price of Coca Cola.
• But in aggregate, prices depend on preferences. The market trade-off
among goods — that is, the price ratio — depends on the aggregation of the
psychic trade-offs among all potential consumers.
• What Walras showed, and what is clear from the Edgeworth box, is that
a competitive market will exhaust all of the gains from trade
• A free market, in equilibrium, is Pareto efficient.
• ...if the following conditions are true:

1. No externalities
2. Perfect competition
3. No transaction costs
4. Full information

• This was Adam Smith’s original brilliant insight:

Blind pursuit of self-interest by autonomous actors in a market setting
yields collectively welfare maximizing behavior.
• Q: Does the 1st Welfare Theorem guarantee that the market allocation
will be ‘fair’ or equitable?
• No. Giving everything to A in the initial endowment would be Pareto
efficient, as would giving everything to B.
• The 1st Welfare Theorem simply says that the market will enlarge the pie
as much as possible; it has nothing to say about who gets which slices.

1.6 Second Welfare Theorem

• So, is there a trade-off between enlarging and dividing the pie?
• Stated rigorously, given a Pareto efficient allocation of resources, will prices
exist such that this allocation is a market equilibrium?
• The 2nd Welfare Theorem proves that the answer is yes.
• 2nd Welfare Theorem: Providing that preferences are convex, any Pareto
efficient allocation can be a market equilibrium.
• The reasons are self-evident in the Edgeworth diagram (though this is a
far from a proof).

— Along the contract curve, every point represents the tangency point
of two indifference curves

8
 — This tangency point corresponds to a price ratio that separates the
two tangent indifference curves
— This price ratio clearly must exist if the indifference curves are tan-
gent and each is convex (so they don’t recross at some later point)
— This price ratio is therefore the market price vector that will support
that particular Pareto efficient allocation.

• Hence, it is immediate from the Edgeworth box that all Pareto efficient
distributions — that is, all points on the CC — are feasible as market equi­
libria.
• As long as the assumptions above are met, a competitive equilibrium will
exist merely because each person is self-interestedly maximizing her own
well-being.
• This equilibrium cannot be improved upon without making at least one
person worse off. In other words, it’s Pareto efficient.
• Moreover, any Pareto efficient allocation can be maintained as a compet­
itive equilibrium.
• This means that the problems of equity/distribution and efficiency can be
separated.
• Hence, another statement of 2nd welfare theorem is: There is no intrinsic
trade-off between equity and efficiency.
• When we discussed consumer versus producer surplus last class, I asserted
that it was justified to maximize the sum of the two rather than worry­
ing about their division. The 2nd welfare theorem is what justified that
assertion.

1.7 Conclusions
• The fundamental welfare theorems provide some very basic policy guid­
ance:

— The purpose of the price system is to ensure that all resources are
consumed in a Pareto efficient fashion — all gains from trade are
exhausted.
— This occurs automatically as prices adjust to clear the market.
— Distorting the price system to achieve equity is intrinsically a bad
idea (as we discussed in the partial equilibrium taxation example
prior to the first mid-term).
— Distorting the price system truly does create a trade-off between
efficiency and equity — which is exactly what the Welfare theorems
say we do not need to do.

9
 — However, this does not mean we should ignore equity. We can achieve
whatever ‘equitable’ allocations of resources is desirable through lump-
sum distributions.

• Is this dictum — don’t distort prices — always correct? No. Because the
strong assumptions underlying the Welfare Theorems are not always — or
perhaps ever — met.
• But it does build a prima facie case that free market outcomes may be
efficient — or at least hard to improve upon.
• Hence, one should not presume that one can improve upon the market.
Instead:

1. A careful diagnosis as to why the market allocation is not desirable.

2. A prescription that analyzes how a specific intervention will remedy
this fault.

• This insight—that the free market system generates a Pareto efficient equi­
librium — is non-obvious. Why would anyone assume that prices are other
than arbitrary social creations?
• And in fact in most of human history, prices and market operations have
been viewed with a great deal of suspicion.
• But economic theory suggests that market equilibria:

1. Have a fundamental logic

2. This logic is an emergent property of the rational, atomistic actions
of market participants.
3. This doesn’t mean that market outcomes cannot be improved upon.
But improving on them requires a careful analysis of why they are not
desirable; and preferably a prescribed fix that harnesses the efficiency
properties of markets rather than attempting to override them.

1.8 Aside: Why do we claim that this conclusion is non-

obvious?
• For most of history, market behavior has been viewed with great suspicion.
• In 1639 in Boston, the respected merchant Robert Keayne is charged with
a heinous crime: He has made over sixpence profit on the shilling, an
outrageous gain.
• The court is debating whether to excommunicate him for his sin.
• In view of his spotless past, in instead fines him 200 pounds.

10

• But Keayne is so distraught over his sin that he prostrates himself before
the church elders and “with tears acknowledges his covetous and corrupt
heart.”
• The minister off Boston cannot resist this golden opportunity to make an
example of Keayne.
• He uses the example of Keayne’s avarice to denounce during his Sunday
sermon on “some false principles of trade:”

1. That a man might sell as dear as he can, and buy as cheap as he can.
2. If a man loses by casualty of sea, etc., in some of his commodities,
he may raise the price of the rest.
3. That he may sell as he bought, though he paid too dear...

• From Helibroner (1953), The Worldly Philosophers (New York: Touch-

stone).

11