Natalia Lazzati

Mathematics for Economics

Note 3: The Implicit Function Theorem

Note 3 is based on Apostol (1975, Ch. 13), de la Fuente (2000, Ch.5) and Simon and Blume
(1994, Ch. 15).

This note discusses the Implicit Function Theorem (IFT). This result plays a key role in
economics, particularly in constrained optimization problems and the analysis of comparative
statics. The …rst section develops the IFT for the simplest model of one equation and one
exogenous variable. We then extend the analysis to multiple equations and exogenous variables.

Implicit Function Theorem: One Equation

In general, we are accustom to work with functions of the form x = f ( ) where the endogenous
variable x is an explicit function of the exogenous variable : This ideal situation does not always
occur in economic models. The IFT in its simplest form deals with an equation of the form

F (x; ) = 0 (1)

where we separate endogenous and exogenous variables by a semicolon.

The problem is to decide whether this equation determines x as a function of : If so, we have
x = x ( ) for some function x (:), and we say x (:) is de…ned "implicitly" by (1). Formally, we are
interested in two questions

(a) Under which conditions on (1) is x determined as a function of ?; and

(b) How do changes in a¤ect the corresponding value of x?

The IFT answers these two questions simultaneously!

The idea behind this fundamental theorem is quite simple. If F (x; ) is a linear function, then
the answer to the previous questions is trivial— we elaborate on this point below. If F (x; ) is
nonlinear, then the IFT states a set of conditions under which we can use derivatives to construct

1
a linear system that behaves closely to the nonlinear equation around some initial point. We
can then address the local behavior of the nonlinear system by studying the properties of the
associated linear one. So the results of Notes 1 and 2 turn out to be important here!
Before introducing the IFT let us develop a few examples that clarify the requirements of the
theorem and its main implications.

Example 1. Let us consider the function F (x; ) = ax b c. Here the values that satisfy
F (x; ) = 0 form a linear equation ax b c = 0: Moreover, assuming a 6= 0, the values of x
and that satisfy F (x; ) = 0 can be expressed as

b c
x( ) = + :
a a

If a 6= 0; then x ( ) is a continuous function of and dx ( ) =d = b=a.

Notice that the partial derivative of F (x; ) with respect to x is @F (x; ) =@x = a. So in this
simple case, x ( ) exists and is di¤erentiable if and only if @F (x; ) =@x 6= 0: N

In Example 1, assuming @F (x; ) =@x 6= 0, x ( ) is de…ned for every initial value of x and .
In the next example, x ( ) exists only around some speci…c values of these two variables.

Example 2. Let F (x; ) = x2 + 2 1, so that the values of x and that satisfy F (x; ) = 0
form a circle of radius 1 and center (0; 0) in R2 :
In this case, for each 2 ( 1; 1) we have two possible values of x that satisfy F (x; ) =
x2 + 2 1 = 0: [Therefore, x ( ) is not a function.] Note, however, that if we restrict x to
positive values, then we will have the upper half of the circle only, and that does constitute a
p
function, namely, x ( ) = + 1 2.

Similarly, if we restrict x to negative values, then we will have the lower half of the circle only,
p
and that does constitute a function as well, namely, x ( ) = 1 2.

Here for all x > 0 we have that @F (x; ) =@x = 2x > 0; and for all x < 0 we have that
@F (x; ) =@x = 2x < 0: Then the condition @F (x; ) =@x 6= 0 plays again an important role in
the existence and di¤erentiability of x ( ) : N

The last two examples suggest that @F (x; ) =@x 6= 0 is a key ingredient for x ( ) to exist, and
that in some cases x ( ) exists only around some initial values of x and/or : The next example

2
proceeds in a di¤erent way, it assumes x ( ) exists and studies the behavior of this function with
respects to :

Example 3. Consider the cubic implicit function

F (x; ) = x3 + 2
3 x 7=0 (2)

around the point x = 3 and = 4: Suppose that we could …nd a function x = x ( ) that solves
(2) around the previous point. Plugging this function in (2) we get

[x ( )]3 + 2
3 x( ) 7 = 0: (3)

Di¤erentiating this expression with respect to (by using the Chain Rule) we obtain

dx dx
3 [x ( )]2 ( )+2 3x ( ) 3 ( ) = 0: (4)
d d

Therefore
dx 1
( )= n o [3x ( ) 2 ]: (5)
d 3 [x ( )]2

At x = 3 and = 4 we …nd
dx 1
( )= : (6)
d 15
Notice that (5) exists if [x ( )]2 6= 0. Since @F (x; ) =@x = 3 x2 , the required condition
is again @F (x; ) =@x 6= 0 at the point of interest: N

Let us extend Example 3 to a general implicit function F (x; ) = 0 around an initial point
(x ; ) : To this end suppose there is a C 1 (continuously di¤erentiable) solution x = x ( ) to the
equation F (x; ) = 0, that is,
F [x ( ) ; ] = 0: (7)

We can use the Chain Rule to di¤erentiate (7) with respect to at to obtain

@F d @F dx
[x ( ); ] + [x ( ); ] ( ) = 0:
@ d @x d

Solving for dx=d we get

dx @F @F
( )= [x ( ); ]= [x ( ); ]: (8)
d @ @x

3
The last expression shows that if the solution x ( ) to F (x; ) = 0 exists and is continuously
di¤erentiable, then we need @F (x; ) =@x 6= 0 at [x ( ); ] to recover dx=d at : The IFT
states that this necessary condition is also a su¢ cient condition!

Theorem 1. (Implicit Function Theorem) Let F (x; ) be a C 1 function on a ball about

(x ; ) in R2 : Suppose that F (x ; ) = 0 and consider the expression

F (x; ) = 0:

If @F (x ; ) =@x 6= 0, then there exists a C 1 function x = x ( ) de…ned on an open interval I

about the point such that:

(a) F [x ( ) ; ] = 0 for all in I;

(b) x ( ) = x ; and

dx @F @F
(c) ( )= [x ( ); ]= [x ( ); ]:
d @ @x

Proof. See de la Fuente (2000), pp. 207-210.

The next example applies the IFT to the standard model of …rm behavior in microeconomics.
In ECON 501A we will study a general version of this problem. Although in the next example
the endogenous variable can in fact be explicitly solved in terms of the exogenous ones, we will
use the IFT to state how changes in the latter a¤ect the former. In this way we can corroborate
the predictions of the IFT hold.

Example 4. (Comparative Statics I) Let us consider a …rm that produces a good y by

using a single input x. The …rm sells the output and acquires the input in competitive markets:
The market price of y is p, and the cost of each unit of x is just w. Its technology is given by
f : R+ ! R+ ; where f (x) = xa and a 2 (0; 1). Its pro…ts are given by

(x; p; w) = pxa wx: (9)

The …rm selects the input level, x; in order to maximize pro…ts. We would like to know how its
choice of x is a¤ected by a change in w:

4
 Assuming an interior solution, the …rst-order condition of this optimization problem is

@
(x ; p; w) = pa (x )a 1
w=0 (10)
@x

for some x = x . (Check that the second order condition holds.)

Notice that here
F (x; p; w) = pa (x )a 1
w: (11)

Since @F (x ; p; w) =@x = pa (a 1) (x )a 2
< 0; we can use the IFT to get

dx @F @F 1
(p; w) = [x ; p; w] = [x ; p; w] = < 0: (12)
dw @w @x pa (a 1) (x )a 2

We conclude that if the price of the input increases, then the …rm will acquire less of it. That
is, the unconditional input demand is decreasing in the input price. [Con…rm this result by
…nding an explicit expression for x in (9) and di¤erentiating it with respect to w:] N

The next sub-section extends the previous ideas to a model that involves a systems of equations
and many parameters.

Implicit Function Theorem: System of Equations

The general problem involves a system of several equations and variables. Some of the variables
are endogenous, and the other ones are exogenous. The question of interest is under what
conditions we can solve these equations for the endogenous variables in terms of the remaining
variables (or parameters). The IFT describes these conditions and some conclusions about the
e¤ect of the exogenous variables on the endogenous ones. To motivate the requirements of the
theorem, we start again with a linear example.

Example 5. Consider a linear system of two equations

0 10 1 0 1
ax1 + bx2 1 =0 a b x
or @ A@ 1 A = @ 1
A: (13)
cx1 + dx2 2 =0 c d x2 2

In Note 1 we showed that the linear system (13) has a unique solution for (x1 ; x2 ) if the determi-
nant of the coe¢ cient matrix is di¤erent from 0, i.e. ad cb 6= 0. If this condition holds, we can

5
use Crammer’s rule to state
0 1 0 1
1 b a b 1d 2b
x1 = x1 ( 1; 2) with x1 ( 1; 2) = @ A = @ A =
d c d ad cb
2
0 1 0 1
a a b
x2 = x2 ( 1; 2) with x2 ( 1; 2) = @
1
A = @ A =a 2 c 1
:
c c d ad cb
2

Then if ad cb 6= 0 we can …nd a pair of functions x1 ( 1; 2) and x2 ( 1; 2) that simultaneously

solve (13) for all possible values of the exogenous variables 1 and 2. Moreover, these two
functions are C 1 and we can di¤erentiate, for instance, x1 ( 1; 2) with respect to 2 to obtain

@x1 ( 1 ; 2) b
= :
@ 2 ad cb

The Jacobian matrix of the left-hand side of system (13) with respect to (x1 ; x2 ) is
0 1
a b
@ A: (14)
c d

The determinant of (14) is ad cb: Therefore, we conclude that the system of equations (13) de…nes
x1 ( 1; 2) and x2 ( 1; 2) as continuous functions of the exogenous variables if the determinant
of its Jacobian matrix is di¤erent from zero. In the general IFT, the nonvanishing condition
of the determinant of the Jacobian matrix also plays a fundamental role. This comes about by
approximating a system of nonlinear equations by linear ones! N

To extend Example 5 to a nonlinear system, let us write the basic system of equations as

F1 (x1 ; :::; xn ; 1 ; :::; m) = 0 (15)

.. .
. = ..

Fn (x1 ; :::; xn ; 1 ; :::; m) = 0

where (x1 ; :::; xn ) is the vector of endogenous variables and ( 1 ; :::; m) are the exogenous ones.
Suppose there are n C 1 functions in a neighborhood of (x1 ; :::; xn ; 1 ; :::; m)

x1 = x1 ( 1 ; :::; m)
.. .
. = ..

xn = xn ( 1 ; :::; m)

6
that solve the system of equations (15). Then

F1 [x1 ( 1 ; :::; m ) ; :::; xn ( 1 ; :::; m) ; 1 ; :::; m] = 0 (16)

.. .
. = ..

Fn [x1 ( 1 ; :::; m ) ; :::; xn ( 1 ; :::; m) ; 1 ; :::; m] = 0:

We can use the Chain Rule to di¤erentiate (16) with respect to h at ( 1 ; :::; m) to get

@F1 @x1 @F1 @xn @F1

+ ::: + + = 0 (17)
@x1 @ h @xn @ h @ h
.. .
. = ..
@Fn @x1 @Fn @xn @Fn
+ ::: + + = 0
@x1 @ h @xn @ h @ h

where all the partial derivatives are evaluated at (x1 ; :::; xn ; 1 ; :::; m ). This system of equations
can be rewritten as 0 10 1 0 1
@F1 @F1 @x1 @F1
@x1 ::: @xn @ h @ h
B CB C B C
B .. .. .. CB .. C B .. C
B . . . CB . C= B . C (18)
@ A@ A @ A
@Fn @Fn @xn @Fn
@x1 ::: @xn @ h @ h

where the n n matrix of the left-hand side is the Jacobian of (16) with respect to (x1 ; :::; xn )
evaluated at (x1 ; :::; xn ; 1 ; :::; m) :

Solving for (@x1 =@ h ; :::; @xn =@ h ) we obtain

0 1 0 1 10 1
@x1 @F1 @F1 @F1
@ h @x1 ::: @xn @ h
B C B C B C
B .. C B .. .. .. C B .. C
B . C= B . . . C B . C: (19)
@ A @ A @ A
@xn @Fn @Fn @Fn
@ h @x1 ::: @xn @ h

We see from (19) that if the solution [x1 ( 1 ; :::; m ) ; :::; xn ( 1 ; :::; m )] to the system of equations
(15) exists and is di¤erentiable, then (@x1 =@ h ; :::; @xn =@ h ) exists if the determinant of the
Jacobian of (15) is di¤erent from zero at (x1 ; :::; xn ; 1 ; :::; m ). (Why?) The IFT shows again
that this necessary condition is also a su¢ cient condition!

Theorem 2. (Implicit Function Theorem) Let F1 ; :::; Fn : Rn+m ! R be C 1 functions:

7
Consider the system of equations

F1 (x1 ; :::; xn ; 1 ; :::; m) = 0 (20)

.. ..
. = .

Fn (x1 ; :::; xn ; 1 ; :::; m) = 0

as possibly de…ning x1 ; :::; xn as implicit functions of 1 ; :::; m. Suppose that (x1 ; :::; xn ; 1 ; :::; m)

is a solution of (20). If the determinant of the n n Jacobian matrix

0 1
@F1 @F1
@x1 ::: @x
B n C
B .. .. .. C
Dx F = B . . . C
@ A
@Fn
@x1 ::: @F
@xn
n

evaluated at (x1 ; :::; xn ; 1 ; :::; m) is non-zero, then there exist C 1 functions

x1 = x1 ( 1 ; :::; m)
.. .
. = ..

xn = xn ( 1 ; :::; m)

de…ned on an open ball B around ( 1 ; :::; m) such that

F1 [x1 ( 1 ; :::; m ) ; :::; xn ( 1 ; :::; m) ; 1 ; :::; m] = 0 (21)

.. ..
. = .

Fn [x1 ( 1 ; :::; m ) ; :::; xn ( 1 ; :::; m) ; 1 ; :::; m] = 0

for all ( 1 ; :::; m) in B and

x1 = x1 ( 1 ; :::; m)
.. .
. = ..

xn = xn ( 1 ; :::; m) :

@x1 @xn
Furthermore, we can calculate @ h ( 1 ; :::; m ) ; :::; @ h ( 1 ; :::; m) as in (19), or solve for one
particular element of this vector by using Crammer’s rule.

Proof. See de la Fuente (2000), pp. 211-212.

8
 SB, pp. 360-364, provide an interesting application. We o¤er a simpler one that extends
Example 4 to a …rm problem with two inputs.

Example 6. (Comparative Statics II) Let us consider again a …rm that produces a good y;
but now let us assume it uses two inputs x1 and x2 . The …rm sells the output and acquires the
inputs in competitive markets: The market price of y is p, and the cost of each unit of x1 and x2
is just w1 and w2 respectively. Its technology is given by f : R2+ ! R+ , where f (x1 ; x2 ) = xa1 xb2 ,
a + b < 1. Its pro…ts take the form

(x1 ; x2 ; p; w1 ; w2 ) = pxa1 xb2 w1 x 1 w2 x2 : (22)

The …rm selects x1 and x2 in order to maximize pro…ts. We aim to know how its choice of
x1 is a¤ected by a change in w1 . Notice that now w1 a¤ects the choice of x1 not only in a
direct way (as in Example 4) but also indirectly through its e¤ect on the other variable of choice,
x2 :
Assuming an interior solution, the …rst-order conditions of this optimization problem are
@
(x ; x ; p; w1 ; w2 ) = pa (x1 )a 1 (x2 )b w1 = 0 (23)
@x1 1 2
@
(x ; x ; p; w1 ; w2 ) = pb (x1 )a (x2 )b 1 w2 = 0
@x2 1 2
for some (x1 ; x2 ) = (x1 ; x2 ). (As we will study later, the second order conditions hold here by
the strict concavity of the production function.)
Let us de…ne

F1 (x1 ; x2 ; p; w1 ; w2 ) = pa (x1 )a 1
(x2 )b w1 (24)

F2 (x1 ; x2 ; p; w1 ; w2 ) = pb (x1 )a (x2 )b 1

w2 :

To achieve our goal we need to di¤erentiate the system of equations (24) with respect to w1
@x1 @x2
pa (a 1) (x1 )a 2
(x2 )b + pab (x1 )a 1
(x2 )b 1
1 = 0 (25)
@w1 @w1
@x1 @x2
pab (x1 )a 1
(x2 )b 1
+ pb (b 1) (x1 )a (x2 )b 2
0 = 0:
@w1 @w1
The Jacobian matrix of F = (F1 ; F2 )T with respect to x at x is
0 1
pa (a 1) (x1 )a 2 (x2 )b pab (x1 )a 1 (x2 )b 1
Dx F (x1 ; x2 ) = @ A: (26)
pab (x1 )a 1 (x2 )b 1 pb (b 1) (x1 )a (x2 )b 2

9
Notice that in this case Dx F (x1 ; x2 ; p; w1 ; w2 ) = pDx2 f (x1 ; x2 ). That is, the Jacobian of F with
respect to (x1 ; x2 ) is the market price times the Hessian of the production function. A similar
structure appears in many other models of microeconomics.
Combining (25) and (26), and rearranging terms, we get

0 10 1 0 1
@x1
pa (a 1) (x1 )a 2
(x2 )b pab (x1 )a 1
(x2 )b 1
1
@ A@ @w1 A=@ A: (27)
@x2
pab (x1 )a 1
(x2 )b 1
pb (b 1) (x1 )a (x2 )b 2
@w1 0
The requirement of the IFT is satis…ed if the determinant of (26) is non-zero: (Check that this
condition holds.) By the IFT and Crammer’s rule we get
0 1
1 pab (x1 )a 1 (x2 )b 1
@ A
a b 2
@x1 0 pb (b 1) (x 1 ) (x 2 )
(p; w1 ; w2 ) = 0 1 < 0:
@w1
pa (a 1) (x1 )a 2 (x2 )b pab (x1 )a 1 (x2 )b 1
@ A
pab (x1 )a 1 (x2 )b 1 pb (b 1) (x1 )a (x2 )b 2

We conclude that if the price of input 1 increases, then the …rm will acquire less of it. [Con…rm
this result by …nding an explicit expression for x1 and x2 in (23) and di¤erentiating
x1 with respect to w1 :] N

10