Simon Fraser University Prof.

Karaivanov
Department of Economics Econ 301

THE FIRM’S PROFIT MAXIMIZATION PROBLEM

These notes are intended to help you understand the firm’s problem of maximizing profits
given the available technology. Both a general algebraic derivation of the problem and the
optimality conditions and specific numerical examples are presented. This is done separately
for the short and long run.

Profit Maximization - Overview

We assume that firms are in business to make as much money as possible, i.e. they strive
to maximize their profits. This assumption has its rationale in the idea of ”natural selection”
or ”survival of the fittest” - if a firm is not maximizing profits its competitors who do would
eventually drive it out of business by employing more efficient (and more profitable) methods
of production. Even if the firm is monopoly, however it would most probably want to maximize
profits - after all firms are owned by people form whom we assumed more is always better.
Thus more firm’s profit would mean more income (or wealth) for its owners which makes them
happier. Thus profit maximization seems a reasonable assumption about firms’ behavior. The
firm maximizes profits (revenues minus costs) by choosing the most efficient way to produce,
i.e. by choosing the optimal amounts of the factors of production to employ. The firm chooses
these optimal amounts taking into account the available technology embodied in the production
function which gives the relationship between the amounts of inputs put into production and
the maximum possible amount of output that can be produced.

Short vs Long Run

The firm’s problem of maximizing profits differs between the short and the long run. Re-
member that we call ”the short run” a time period in which at least some factors of production
are fixed. Thus in the short run the firm is unable to vary all factors - some of them are fixed at
predetermined levels and cannot be changed. Thus the only thing that the firm can choose in
this case is the quantities of variable inputs (not fixed) to hire. For example we can imagine
that we produce output using labor and a plant and the size of the plant is fixed in the long
run. The only choice variable for the owner of such firm in the short run (the only thing that
can be varied to maximize profits) is then the number of people hired.
In contrast, in the long run all factors are variable so the firm can choose how much to hire
of all inputs, which makes its problem a bit more complicated. In our example above, the
owner of the firm can choose both the number of people to hire and the size (or number) of
plants.

Notation
To simplify things, in the discussion below we will use an example with 2 inputs and one
output. We will use the following notation:
- x1 , x2 - the quantities of the two inputs hired by the firm

1
 - y = f(x1 , x2 ) − the amount of output produced from x1 and x2 where f (x1 , x2 ) is the
production function
- w1 , w2 - the per unit market prices (wages) for the two inputs
- p - the market price of output

We will assume that the firm takes input and output prices w1 , w2 and p as given - i.e. it
cannot influence them. Thus we are looking at the case where both input and output markets
are competitive (each firm is too small to affect the prices).

Profit Maximization in the Short Run

Since we are in the short run (SR) assume that factor 2 for example is fixed, i.e. x2 = x̄2
(we just have our single factory). The firm’s problem then is to maximize profits by choice of
x1 - the amount of input 1 to be hired. Notice that the firm cannot choose anything else here
- prices are assumed to be taken as fixed for the firm, x̄2 is fixed and output is just a function
of x1 and x̄2 i.e. once we choose x1 we have chosen y as well.
The firm’s problem is then:

max pf (x1 , x̄2 ) − w1 x1 − w2 x̄2

x1

In words, the above says - maximize profits (=revenue (py) minus costs (w1 x1 + w2 x̄2 )) by
choosing the optimal quantity of input 1 (x1 ) to hire. It looks like the consumer problem but
there is no constraint. Why? Aren’t firms constrained? Yes, they are, the constraint is actually
there - it is just plugged in into the profit function - remember that the constraint on firm’s
behavior is put by nature - it is the production function, saying that output, y must equal
f (x1 , x̄2 ) - i.e. that y cannot be chosen independently of x1 .

How do we solve it?

This is a simple maximization problem of one variable so we just need to take the first
derivative of the function that is being maximized and set it equal to zero. We get:

pf1 (x1 , x̄2 ) − w1 = 0 (1)

where we denote by f1 (x1 , x̄2 ) the derivative of the production function with respect to x1 .
Notice that x̄2 is treated as a parameter, a constant which it is.
The above equation can be solved for the optimal quantity of factor 1, x∗1 that the firm
will use to achieve highest profits. We call x∗1 the factor demand for input 1. Just as in the
consumer theory, it will be a function of the prices in general, i.e. x∗1 = x∗1 (p, w1 , w2 ).

This equation has a very nice economic interpretation. Remember what is f1 (x1 , x̄2 ) (the
derivative of the production function with respect to x1 ) − it is simply the slope of the pro-
duction function at x1 , or in other words - the marginal product of factor 1.
Thus equation (1) is saying that at optimum we must have:

pMP1 = w1 (2)

2
 What does this mean? Remember what was the marginal product (MP1 ) - it was the
increment in output obtained by hiring one more unit of input 1. Multiply this by the price of
output and the left hand side above can be interpreted as the value of marginal product of
input one, i.e. the increment in revenue (money) that we would obtain if hiring one more unit
of input 1. What is the right hand side then? It is simply the cost of hiring an additional
unit of the factor. Thus optimality requires that we set the additional revenue we get from
hiring one more unit of the input equal to the cost of obtaining that unit.
Why is this optimal? Well, suppose that pMP1 > w1 at our profit maximizing optimum,
i.e. that the additional revenue obtained from hiring one more unit of the factor is larger than
the cost of doing so. Then we can increase our profits by hiring more of the factor until pMP1
decreases to a level equal to w1 (remember hiring more units of the input holding the other input
constant decreases its marginal product (MP1 ) due to the law of diminishing marginal product).
Since we are able to increase profits we must not have been profit maximizing. Similarly, if we
suppose that pMP1 < w1 at the profit maximizing point we see that if we reduce our usage of
input 1 we can actually increase our profits since this last unit brings us less value (pMP1 ) than
what we have to pay for it (w1 ).
Let us compute the optimal choice of x1 (the factor demand) for the Cobb-Douglas produc-
tion function f (x1 , x2 ) = xα1 xβ2 . The firm’s problem is:

max pxα1 x̄β2 − w1 x1 − w2 x̄2

x1

Setting the first derivative with respect to x1 we get:

pαxα−1
1 x̄β2 − w1 = 0

since x̄2 is just a constant. We can solve the above equation for the factor demand,
x∗1 (p, w1 , w2 ). We have:
w1
xα−1
1 =
pαx̄β2
1
or (rasing both sides to power α−1
)
w1 α−1 1
x∗1 = ( β
)
pαx̄2

Notice that for the Cobb-Douglas function the factor demand for input 1 depends on w1
and p but not on the price of the second input, w2 .

Numerical Example (different from class)

Let us now consider a particular example with a specific production function and prices.
1/2 1/2
Assume that f (x1 , x2 ) = x1 x2 , w1 = 2, w2 = 1, p = 4 and x̄2 = 1.

Proceeding as above, the firm’s problem is (just substituting the numbers):

1/2
max 4x1 (1)1/2 − 2x1 − (1)(1)
x1

3
 Taking the derivative and setting to zero:
1 −1/2
4 x1 −2=0
2
or,
−1/2
2x1 =2
or
−1/2
x1 =1
i.e.
x∗1 = 1
(you could have obtained the same expression by substituting the specific numerical values
used here for p, x̄2 , w1 , α, β in the general expression from before).

Having the optimal quantity demanded of the input we can compute the optimal amount
of output that will be produced by simply plugging x∗1 into the production function. We get,
y ∗ = f (x∗1 , x̄2 ) = (1)1/2 (1)1/2 = 1. We can also find the maximized profit, π ∗ = py ∗ − w1 x∗1 −
w2 x̄2 = 4(1) − 2(1) − 1 = 1. (everything turns out to be equal to one in this example :-) )

Profit Maximization in the Long Run

Now consider the long run - i.e. when all factors are variable and hence can be chosen by
the firm when deciding how to maximize profits. The difference from before in our example is
that both x1 and x2 can now be chosen. Thus the firm’s profit maximization problem in the
long run looks like:
max pf (x1 , x2 ) − w1 x1 − w2 x2
x1 ,x2

Notice that there is no x̄2 anymore as both inputs are to be chosen.

How to solve it?

Now we have a function of two variables and (unlike the consumer’s problem) there is no
constraint to use to express x2 in terms of x1 or vice versa. So what do we do? Well, same as
before: take derivatives and set to zero. The difference is that now we have two variables so we
have to take the derivative of the function with respect to each of them holding the other
constant. This is simply taking the partial derivatives with respect to each variable and
setting them to zero.
The partial derivative with respect to x1 (holding x2 constant) is obtained in the same way
as before:
pf1 (x1 , x2 ) − w1 = 0
∂f (x1 ,x2 )
where f1 (x1 , x2 ) = ∂x1
is the partial derivative of f with respect to x1 . Similarly, for x2
we get
pf2 (x1 , x2 ) − w2 = 0
∂f (x1 ,x2 )
where f1 (x1 , x2 ) = ∂x2
is the partial derivative of f with respect to x2 .

4
 Once again the partial derivatives are simply the marginal products of the two factors, i.e.
we have at optimum:

pMP1 = w1
pMP2 = w2

The interpretation is the same as before - the value of marginal product of each factor must
equal its price at optimum otherwise factor usage can be either decreased or increased yielding
higher profits.
The above represents a system of two equations that can be solved for the optimal factor
demands, x∗1 and x∗2 (notice that MP1 and MP2 are functions of x1 and x2 ).

Let us look at our Cobb-Douglas example and derive the factor demands as we did in the
SR case. The long run firm’s problem is:

max pxα1 xβ2 − w1 x1 − w2 x2

x1 ,x2

Taking the partial derivatives and setting them to zero we get:

pαxα−1
1 xβ2 = w1
pxα1 βxβ−1
2 = w2 (3)

Notice that these are two equations in two unknowns (x1 and x2 ). How to solve them?
The easiest way is to divide them through (i.e. divide both sides of the first equation by the
respective sides of the second):
pαxα−1
1 xβ2 w1
β−1
=
pxα1 βx2 w2
Cancelling the p0 s and collecting the powers of x1 and x2 we get:
α x2 w1
= (4)
β x1 w2
β β−1
(since xα−1
1 /xα1 = x−1
1 = 1/x1 and x2 /x2 = x2 ).
From above we can express x2 in terms of x1 :
βw1
x2 = x1 (5)
αw2

Now use this expression and plug in the first derivative equation (pαxα−1
1 xβ2 = w1 ) to get:
βw1 β β
pαxα−1
1 ( ) x1 = w1
αw2
Notice that we can
or,
βw1 β α+β−1
pα( ) x1 = w1
αw2

5
 Then we have:
w1
xα+β−1
1 = βw1 β
pα( αw2
)
1
or, raising to power α+β−1
:

w1 1 pα 1−β pβ β
x∗1 = ( βw1 β
) α+β−1 = ( ) 1−α−β ( ) 1−α−β
pα( αw2 ) w1 w2

i.e. some messy function of w1 , w2 and p. We can then use (4) to get x∗2 as an even messier
function of the prices.

Alternative Way to Solve (less messy):

Let’s now look at the derivatives once again. Remember that y = xα1 xβ2 , i.e. we can write
the system (3) as:
y
pα = w1
x1
y
pβ = w2
x2
Thus we have that at optimum:
pαy pβy
x1 = and x2 =
w1 w2

Remembering that y = xα1 xβ2 we get an equation for the optimal output y ∗ :

pαy ∗ α pβy β
y∗ = ( ) ( )
w1 w2
or,
pα α pβ β
(y ∗ )1−α−β = ( ) ( )
w1 w2
α β 1−β β α 1−α
i.e. y ∗ = ( wpα1 ) 1−α−β ( wpβ2 ) 1−α−β and hence: x∗1 = ( wpα1 ) 1−α−β ( wpβ2 ) 1−α−β , x∗2 = ( wpα1 ) 1−α−β ( wpβ2 ) 1−α−β .
The advantage in this method is that everything is more symmetrical and it stresses the inter-
dependence between the inputs and the output.

The above is general but pretty incomprehensible (although you can simplify it a bit). Let’s
look at a:

Numerical Example:
1/3 1/3
Assume that f (x1 , x2 ) = x1 x2 , w1 = 1, w2 = 2, p = 3. Then the firm’s problem is:
1/3 1/3
max 3x1 x2 − x1 − 2x2
x1 ,x2

6
 Taking the partials and setting them to zero we get:
1 −2/3 1/3
3 x1 x2 = 1
3
1/3 1 −2/3
3x1 ( x2 ) = 2
3
As in the general case we have two equations in the two unknowns, x1 and x2 . The simplest
way to proceed is once again to divide the equations through:
x2 1
=
x1 2
i.e. x2 = x1 /2. Substitute that into the first equation to get:
−2/3
x1 (x1 /2)1/3 = 1

or
−1/3
(2)−1/3 x1 =1
raising both sides to power 3:
x−1
1 = 2

or, x∗1 = 1/2. Thus x∗2 = x∗1 /2 = 1/4. Output is: y∗ = (1/2)1/3 (1/4)1/3 = (1/8)1/3 = 1/2,
optimal profit is: π∗ = 3y ∗ − x∗1 − 2x∗2 = 1/2.