Microeconomic Analysis, Lecture 6: An example of general equilibrium,

worked out to last consequences

Steen Hoernig, 24/09/2015, Nova SBE

How to calculate a General Equilibrium

1.1

Setup

We have two goods, and two consumers with utility functions

uA x1A ; x2A

x1A

x2A

uA x1B ; x2B

x1B

x2B

and endowments ! A = (! 1A ; ! 2A ) and ! B = (! 1B ; ! 2B ). Also let ! 1 = ! 1A + ! 1B

and ! 2 = ! 2A + ! 2B the total endowments of goods 1 and 2.

1.2

Demand functions

Let us rst look at consumer A. For some budget MA and prices p = (p1 ; p2 ),
he solves
max
uA x1A ; x2A s:t: p1 x1A + p2 x2A MA
1
2
xA ;xA

We have the Lagrangian L = (x1A ) (x2A )

rst-order conditions
@L
1
1
=
x1A
x2A
1
@xA
@L
= (1
) x1A
x2A
0 =
@x2A
MA = p1 x1A + p2 x2A

p1 =

0 =

(MA

x1A

p2 =

p1 x1A

x1A

x2A

x2A
n

x1A

p2 x2A ), with

x2A

p1
1

Divide the rst two conditions to get the "MRS = price ratio" condition
=x1A
=
) =x2A
1

u1
=
u2
(1
and solve for x2A :

p1
x2A
=
1
xA
p2

x2A =

p1 1
x , x1A (p; MA ) =
p2 A

p1 1
x
p2 A
Substitute into budget constraint to obtain
MA = p1 x1A + p2

MA
p1

p2

By symmetry we have
x2A (p; MA ) =

1.3

(1

) MA
p2

x1B (p; MB ) =

MB 2
(1
; xB (p; MB ) =
p1

) MB
p2

Gross Demands

Now we remember that consumersincome is given by their endowments at

given prices:
MB = p1 ! 1B + p2 ! 2B

MA = p1 ! 1A + p2 ! 2A ;
Thus gross demands are given by

p1 ! 1A + p2 ! 2A
=
p1
p1 ! 1A + p2 ! 2A
x2A (p; p ! A ) = (1
)
p2
2
1
p1 ! B + p 2 ! B
=
x1B (p; p ! B ) =
p1
p1 ! 1B + p2 ! 2B
2
xB (p; p ! B ) = (1
)
p2
x1A (p; p ! A ) =

1.4

! 1A +

p2 2
!
p1 A

= (1

p1 1
! + ! 2A
p2 A

! 1B +

p2 2
!
p1 B

= (1

p1 1
! B + ! 2B
p2

Excess demands, and Walraslaw

The aggregate excess demands are found by summing up over consumers:

z1 (p) = x1A + x1B

!1 =

z2 (p) = x2A + x2B

! 2 = (1

p2 2
!
+
p1 A
p1 1
)
! + ! 2A
p2 A

! 1A +

! 1B +

p2 2
!
p1 B

!1

+ (1

p1 1
! + ! 2B
p2 B

!2

Verify Walraslaw:
p1 ! 1A + p2 ! 2A + p1 ! 1B + p2 ! 2B
p1 ! 1
) p1 ! 1B + p2 ! 2B
+ (1
) p1 ! 1A + p2 ! 2A + (1
p1 ! 1A p1 ! 1B
= p1 ! 1A + p2 ! 2A + p1 ! 1B + p2 ! 2B
= 0

p1 z1 (p) + p2 z2 (p) =

Indeed, the value of the excess demands is zero.

p2 ! 2
p2 ! 2A p2 ! 2B

1.5

The equilibrium price

We know that we need to x one price, say p1 = 1, or solve directly for the
price ratio. We also know that then we can use the condition that one of
the excess demands is zero to nd the equilibrium. But which one? The one
that looks nicer to you! (The result must be equal!). Take the condition
z1 (p) = 0, or
! 1A +
or

p2 2
!
p1 B

!1 = 0

) ! 1A + (1

) ! 1B

! 1B +

p2 2
p2 2
!A +
! = (1
p1
p1 B

or

1.6

p2 2
!
p1 A

) ! 1B
) ! 1A + (1
! 2A + ! 2B

(1
p2
=
p1

Some comparative statics

What happens to the equilibrium price if preferences or endowments change:

Assume that ! 1A or ! 1B goes up, i.e. good 1 becomes relatively less
scarce ) pp12 increases, i.e. p1 decreases relative to p2 : more (less)
scarcity of a good leads to a relatively higher (lower) price!
Assume consumer As relative preference shifts towards good 1, i.e. his
increases ) pp21 decreases, i.e. p1 increases relatively to p2 : Relatively
more (less) desired goods have relatively higher (lower) prices

1.7

The equilibrium allocation

Let us nd the equilibrium gross and net demands. Gross demand:

x1A

=
=

) ! 1A + (1
) ! 1B 2
!A
! 2A + ! 2B
! 1A ! 2A + ! 1A ! 2B + (1
) ! 2A ! 1B
! 2A + ! 2B
! 1A +

(1

The other gross demands look similar and can be found from intelligent (and
lucky) cut-and-paste, for example for good 2:
x2A = (1

! 1A ! 2A + (1
) ! 2A ! 1B + ! 1A ! 2B
(1
) ! 1A + (1
) ! 1B
3

Net demands:
x1A

! 1A =

x2A

! 2A =

(1
(1

) ! 2A ! 1B (1
) ! 1A ! 2B
! 2A + ! 2B
) ! 1A ! 2B
(1
) ! 2A ! 1B
(1
) ! 1A + (1
) ! 1B

Depends on all endowments and on relative preferences for goods whether

net demands are positive or negative, i.e. whether a consumer wants to buy
or sell a good in equilibrium!
Check whether this is correct by verifying the budget constraint
0 = x1A

! 1A +

p2 2
x
p1 A

! 2A

or
0 =
+
=

) ! 1A ! 2B
) ! 2A ! 1B (1
! 2A + ! 2B
) ! 1A ! 2B
(1
) ! 1A + (1
) ! 1B (1
2
2
1
!A + !B
(1
) ! A + (1

(1
(1

! 2A

) ! 2A ! 1B
) ! 1B

0
=0
+ ! 2B

(u!)

1.8

Finding the equilibrium using oer curves

While the preceding was easy, this part is a bit more complicated mathematically (but not conceptually). We want to nd the point where the oer
curves cross in the Edgeworth box. That is, contrary to the above we rst
nd the equilibrium quantities, and then derive the equilibrium prices.
Remember that the oer curves were curves in the (quantity) Edgeworth
box, that is, we need to express them as "good 2 as a function of good 1",
without having any prices as arguments. So we go back to the rst-order
conditions of the consumers decision problem, that is MRS equal to price
ratio and the budget constraint,
1

x2A
p1
= ; p1 x1A + p2 x2A = p1 ! 1A + p2 ! 2A :
1
xA
p2

Divide the budget constraint by p2 and substitute out p1 =p2 using the rst
condition; then solve for x2A :
x2A =

(1
) ! 2A x1A
:
x1A
! 1A
4

The same for consumer B :

x2B =

(1
) ! 2B x1B
:
x1B
! 1B

Crossing these two curves in the Edgeworth box means taking into account
that demands must sum up to endowments, i.e. solving
(1
) ! 2A x1A
x1A
! 1A
(1
) ! 2B (! 1 x1A )
=
(! 1 x1A )
! 1B

x2A =
!2

x2A

We obtain the solutions x1A = ! 1A etc. (initial endowments) and

x1A =

! 1A ! 2A + (1

) ! 2A ! 1B + ! 1A ! 2B
= x1A
! 2A + ! 2B

etc. as before. Indeed, the oer curves cross at these points! The corresponding quantities for good 2 (just to check that everything is ne) are, from the
oer curve,
x2A ! 1A
x2A x1A

(1
) ! 2A ! 1A
= ! 2A
! 1A
! 1A
(1
) ! 2A x1A
= x2A
=
x1A
! 1A

The equilibrium price ratio is then found equating it to the marginal rate of
substitution:
p1
= M RSA =
p2
1

x2A
=
x1A
(1

Correct!

Short Break

breathe in breathe out ok!

! 2A + ! 2B
p
= 1
1
1
) ! A + (1
) !B
p2

3
3.1

E cient Allocations and the Welfare Theorems

The contract curve

The long way to nd the contract curve is to solve a full maximization problem:
max

x1A ;x2A ;x1B ;x2B

uA x1A ; x2A

uB x1B ; x2B
uB
1
1
1
xA + xB = !
x2A + x2B = ! 2

s:t:

The Lagrangian for this problem is

L = uA x1A ; x2A +
+ 1 ! 1 x1A

uB x1B ; x2B
x1B + 2 ! 2

uB
x2A

x2B

(beautiful, no?) Now the rst-order conditions with respect to the quantities
are
@L
@uA
=
1
@xA
@x1A
@L
@uB
=
1
@xB
@x1B

@L
@uA
=
2
@xA
@x2A
@L
@uB
= 0;
=
2
@xB
@x2B

= 0;
1

=0
2

=0

From this we obtain:

@uA =@x1A
=
M RSA =
@uA =@x2A

@uB =@x1B
= M RSB ;
@uB =@x2B

i.e. the indierence curves are tangent. The short way is to start from this
very point.
In the present case, we have
M RSA =

x2A
=
x1A
1

x2B
= M RSB
x1B

Together with the feasibility constraints, the contract curve is then given by

x2A
=
x1A
1

(! 2
(! 1
6

x2A )
x1A )

which we can solve for good 2:

x2A = C x1A

(1
(1

) ! 2 x1A
) !1 + (
) x1A

In the Cobb-Douglas economy the contract curve is a straight line if and

only if = , i.e. consumers have the same relative preferences for dierent
goods!

3.2

The rst welfare theorem

Theorem (the easy one): "Each market equilibrium is Pareto e cient"

Show that our equilibrium allocation lies on the contract curve! That is,
plug x1A into C (x1A ), close your eyes, prey, and hope that what comes out is
x2A :
C x1A

! 1A ! 2A + ! 1A ! 2B +(1
! 2A + ! 2B

) !2

(1
=

) !1 + (

(1
= (1

)! 2A ! 1B

! 1A ! 2A + ! 1A ! 2B +(1
! 2A + ! 2B

)! 2A ! 1B

) ! 2A ! 1B + ! 1A ! 2B
! 1A ! 2A + (1
(1
) ! 1A + (1
) ! 1B

= x2A
Yeah! We were right, our market equilibrium lies on the contract curve and
is therefore Pareto e cient!

3.3

The second welfare theorem

Theorem (the di cult one): "Each Pareto-e cient allocation can be market
equilibrium (with itself as initial endowment)"
(true here because preferences are indeed convex!)
Show that price ratio equal to slope of indierence curves at allocation
gives rise to demands which are equal to this allocation!
Fix some Pareto-e cient allocation x1A , x2A = C (x1A ), x1B = ! 1 x1A ,
x2B = ! 2 x2A , where ! 1 and ! 2 are total quantities of goods 1 and 2, and
let endowments be given by the same allocation: ! 1A = x1A , ! 2A = x2A .
Propose a price ratio equal to the (identical) slopes of the indierence
curves at this point:
p1
= M RSA =
p2
1

x2A
=
x1A
7

(1

!2
) !1 + (

) x1A

Now check whether gross demands for goods 1 and 2 at this price ratio
coincide with the original Pareto-e cient allocations:
x1A

p2 2
C (x1A )
1
+ !A =
xA +
p1
p1 =p2
2 1
(1
) ! xA
x1A +
!2
! 1A

(p; p ! A ) =
=
= x1A

We got our quantity of good 1 back! What about good 2?

p1 1
! + ! 2A = (1
p2 A
(1
) ! 2 x1A
=
(1
) !1 + (
) x1A
= C x1A = x2A

x2A (p; p ! A ) = (1

p1 1
x + C x1A
p2 A

Correct, we are done for consumer 1. I leave it to you to check consumer 2

:-))