4.

Introduction to General
Equilibrium
1 Introduction
General Equilibrium theory sees the economy as a closed and interrelated system
in which the equilibrium values of all variables of interest have to be determined
simultaneously. In other words, all markets clear at once and the entire economy
is at an equilibrium point, i.e. the vector of final consumption, the vector of
production and the vector of prices of all goods are determined at once. This
is the set of endogenous variables in the system, and whenever a change in
the exogenous variables is introduced, the equilibrium values of all endogenous
variables have to be recalculated.
The set of exogenous variables is reduced to the set of economic agents, the
available technologies, preferences and endowments.
The main assumptions of General Equilibrium theory are:

1. Agents optimize: consumers solve their utility maximization problems,

and producers their profit maximization problems.

2. Agents behave competitively, they take prices as given (and optimize in-
dependently).
3. There is a complete set of markets with perfect and complete information.

4. Markets clear, no agent would change his actions, and there is no excess
demand nor supply.

General equilibrium differs from partial equilibrium in the sense that in

partial equilibrium some endogenous variables in the economy are assumed to
remain constant when a perturbation occurs and only the change in some en-
dogenous variables is studied.
We will start by analyzing three examples of General Equilibrium settings:

• The Edgeworth box economy: 2 consumers, 2 goods and no production

(pure exchange).
• The one-consumer, one-producer economy.

• The 2x2 production model: 2 inputs, 2 factors of production.

2 Equilibrium and Pareto Optimality

We consider a private ownership economy with an arbitrary number of con-
sumers, commodities, and firms. Private ownership means that consumers own

1
the firms. Therefore, part of the consumer’s wealth will be claims to the firms’
profits.
The basic setting is

• I consumers, characterized by consumption sets Xi ⊂ RL

+ , and rational
preferences <i on Xi .
• J firms, characterized by nonempty Yj .

• Endowments ω = (ω 1 , ..., ω I ) ∈ RL .

An allocation (x, y) ∈ RL(I+J) is a specification of consumption vectors

xi ∈ Xi ∀i and production vectors yj ∀j. An allocation is feasible if
X X
xli = ω l + yli ∀l
i i

Let A be the set of feasible allocations. Then, (x, y) is Pareto Optimal if

@(x0 , y 0 ) ∈ A such that x0i <i xi ∀i and x0i i xi for some i. 
I J I
Given a private ownership economy {Xi , <i }i=1 , {Yj }j=1 , {(ωi , θi1 , ..., θiJ )}i=1 ,
(x, y) and p are a Walrasian equilibrium if:

1. ∀j, yj∗ solves the PMP in Yj , i.e. p · yj ≤ p · yj∗ ∀yj ∈ Yj .

n o
2. ∀i, x∗i is maximal for <i in xi ∈ Xi such that p · xi ≤ p · ωi + j θij p · yj∗ .
P

x∗i = ω + yj∗ .
P P
3. Market clearing condition: i j

3 The Edgeworth box economy

There is only exchange, but no production opportunities. The production set,
therefore, is Y = −R2+ , and the only possibility is free disposal. There are two
commodities. In order to characterize the economy, we need to specify:

• Consumer preferences <i over consumption vectors xi .

• Endowments ωi = (ω1i , ω2i ). The total endowment of good l is ω l =
ωl1 + ωl2 .

We need to define an allocation as x ∈ R4+ with x = (x1 , x2 ) = ((x11 , x21 ), (x12 , x22 )),
a distribution of the goods Pin the economy between the two consumers. P
Then, x is feasible if i xli ≤ ω l ∀l. Moreover, x is nonwasteful if i xli =
ω l ∀l. We will represent nonwasteful feasible allocations by means of the Edge-
worth box.
Given the initial allocation, the budget set is a function of prices:

Bi (p) = xi ∈ R2+ : p · xi ≤ p · ωi ,


2
which implies that wealth is endogenous and depends on prices.
The set of Pareto optimal allocations is the Pareto set. An allocation x
is Pareto optimal if @x0 s.t. x0i <i xi for i = 1, 2 and x0i i xi for some i.
An allocation x Pareto dominates x0 if xi <i x0i for i = 1, 2 and xi i x0i for
some i. The set of Pareto optimal allocations that Pareto dominates the initial
allocation is the contract curve. The Walrasian equilibrium allocation will be
on the contract curve.
For each price vector p the consumer solves a utility maximization problem,
leading to the demand function xi (p, p · ωi ). Consumer i’s set of optimal choices
depending on p, that is as p varies, is the offer curve.
A Walrasian equilibrium in the Edgeworth box is determined by the inter-
section of the offer curves of the two consumers. It is a price vector p∗ and an
allocation x∗ = (x∗1 , x∗2 ) in the Edgeworth box such that for i = 1, 2 we have
that x∗i <i x0i ∀x0i ∈ Bi (p). Since there is no production it holds that y ∗ = 0. At
a Walrasian equilibrium, no good is in excess demand or supply.
In general, the Walrasian equilibrium implies tangency between the indiffer-
ence curves of the two individuals, except in corner solutions.

Example 1 In a two-person, two-good economy, consumer A has an endow-

ment (1, 0) and preferences uA (x1 , x2 ) = x1 x2 , whereas consumer B has endow-
ment (1, 2) and preferences uB (x1 , x2 ) = min {x1 , x2 } . In order to determine the
equilibrium, we need to analyze both consumers’ utility maximization problems.
First, the Pareto set is the diagonal, and the contract curve is the portion
of the diagonal in between the two indifference curves that contain the initial
allocation. The Walrasian equilibrium will be somewhere on the diagonal.
Consumer A solves (after normalizing the price of the second good p2 = 1
and setting p1 ≡ p) :

maxx1 x2
x≥0
s.t. px1 + x2 = p,

and in this case the solution is interior (you should be able to verify this) as
long as 0 < p < ∞. The optimal bundle is derived from
x1 p1
= and from x1 p + x2 = p,
x2 p2

and therefore, x1 = 12 and x2 = p2 . If p = 0, then the solution is x2 = 0, and if

p goes to infinity, then there is no solution.
On the other hand, consumer 2 solves

max min {x1 , x2 }

x≥0
s.t. px1 + x2 = p +2

Given that this utility function is not differentiable, the optimal solution is
derived by analyzing his utility function carefully. The optimal bundle is always

3
such that x1 = x2 , which implies that
p+2
x1 = x2 =
p+1
Then, the Walrasian equilibria are:

• p = 0 and xA = (0, 0), xB = (2, 2)

• p = 1 and xA = ( 12 , 12 ), xB = ( 23 , 32 )

4 The one-consumer, one-producer economy

In this case, L = 2, I = 1, and J = 1. The two goods are labor and the
consumption good. Labor can be transformed into the consumption good by
means of th production function f (z), where z is the amount of labor used in
production.
Moreover, w denotes the salary, and thus, the firm solves

max pf (z) − wz,

z≥0

which yields z(p, w), q(p, w), and π(p, w).

Since the firm is owned by the unique consumer, he solves

max u( x1 , x2 )
x1 ,x2 ≥0

s.t. px2 ≤ w(L − x1 ) + π(p, w).

Then, a Walrasian equilibrium in this economy is (p∗ , w∗ ) and x∗1 , x∗2 , q ∗ , z ∗

such that the markets for consumption and labor clear, i.e.

x2 (p∗ , w∗ ) = q(p∗ , w∗ )
∗ ∗
z(p , w ) = L − x1 (p∗ , w∗ ).

Example 2 Assume that the consumer’s preferences on leisure and the con-
sumption good are such that u(x, R) = a ln x + (1 − a) ln R, where R is leisure.
Moreover, his total endowment of time is 1.
The firm’s technology is given by x = aL, where L is labor input employed
by the firm. Hence, in equilibrium, R + L = 1.
The firm’s problem, given p, w is
p 1
max paL − wL ⇒ pa = w ⇒ = , and profits are zero
L≥0 w a
The consumer, therefore, maximizes

max a ln x + (1 − a) ln R
x,R≥0
s.t. px ≤ w(1 − R)

4
 Since the solution is interior, we know that
a R p 1
= = ,
1−a x w a
a2
and thus, we get x = 1−a R. Making use of the market-clearing condition in the
market for labor,

a2
R = a(1 − R) ⇒ R = 1 − a, x = a2 .
1−a
√
Example 3 Now, if the production function was x = L instead, the firm’s
problem would be
√ p  p 2
max p L − wL ⇒ √ = w ⇒ L =
L≥0 2 L 2w
p2 p 2 p2


Thus, π(p, w) = 2w − w 2w = 4w 6= 0. This result arises because of
decreasing returns to scale. Then, the consumer’s problem is

max a ln x + (1 − a) ln R
x,R≥0

p2
s.t. px ≤ w(1 − R) + ,
4w
and therefore,
a R p a
= ⇒ xp = Rw.
1−a x w 1−a
Then, making use of the budget constraint, the optimal consumption of
leisure will come from

p2 p2
 
a
wR = w(1 − R) + ⇒R= 1+ (1 − a),
1−a 4w 4w2

which implies that, using the market-clearing condition:

 p 2  p2

+ 1+ (1 − a) = 1.
2w 4w2
q q
Then, wp = 2 2−a a
, implying that R = 2 1−a
2−a and x = a
2−a . Profits are
thus positive in equilibrium.

5 The 2x2 production model

Now we have two firms, each firm producing one output, using two inputs.
The production functions are fj (zj ), which have CRS. The economy has total
endowments (z 1 , z 2 ), owned by consumers, who do not want to consume them.

5
 Let output prices be fixed at p = (p1 , p2 ). Our goal is to determine the
equilibrium factor prices (w1 , w2 ). The procedure will be to set up and solve
the two firms’ cost minimization problems, so as to derive their factor demands.
The equilibrium factor prices will be those that clear the markets for the two
inputs.
Assume that f1 (z1 , z2 ) and f2 (z1 , z2 ) are homogeneous of degree one. This
implies constant returns to scale. Let cj (w) be the minimum unit cost, and
aj (w) = (a1j (w), a2j (w)) be the input combination needed to achieve that cost
cj (w). Given the firm’s technology, a unit isoquant for firm j is {(z1j , z2j ) ∈
R2+ : fj (z1j , z2j ) = 1}. Similarly, we have the level curve of the unit cost function
{(w1 , w2 ) : cj (w1 , w2 ) = c}. The gradient of these curves are, respectively, w
and aj (w).
We can thus construct an Edgeworth box of sizes z 1 , z 2 to characterize non-
wasteful allocations of resources. The Pareto set is the set of efficient allocations
of resources. The Pareto set lies entirely above or below the diagonal or is the
diagonal itself.
The production possibility set is (q1 , q2 ), that is, these are all output vectors
that are feasible given the inputs of the economy. The combinations q1 , q2
correspond to the points on the Pareto set. In order to determine equilibrium
factor prices, we have to use
c1 (w1 , w2 ) = p1 and c2 (w1 , w2 ) = p2 ,
and both conditions must hold simultaneously in equilibrium. The equilibrium
is determined by the intersection of c1 (w) = p1 and c2 (w) = p2 .
In order to make sure that we have a unique solution, it is convenient to
make the following assumption: The production of good 1 is relatively more
intensive in factor 1 than the production of good 2 if
a11 (w) a12 (w)
> ∀w.
a21 (w) a22 (w)
In words, the unit cost curve in (w1 , w2 ) is always steeper for good 1.
Knowing w∗ , we can immediately determine
∗
z11 a11 (w) z∗ a12 (w)
∗ = and 12
∗ = a (w)
z21 a21 (w) z22 22

Graphically, z ∗ is the intersection of the two rays with slopes equal to the
ratio of the unit factor proportions.
The factor price equalization theorem says that the equilibrium factor prices
depend only on the technologies of the two firms and on output prices.
A change in p1 only, will result in an increase of the equilibrium price of the
factor relatively more used in good 1, and will decrease the price of the other
factor. This is the Stolper-Samuelson theorem.
The final comparative statics result is Rybcszynski’s theorem, which states
that if total endowment of factor 1 increases, the production of the good that
uses it relatively more intensively increases, while factor prices remain unaltered
since technology has not changed.