Chapter 8

COST FUNCTIONS

Copyright ©2005 by South-western, a division of Thomson learning. All rights reserved. 1

 Definitions of Costs
• It is important to differentiate between
accounting cost and economic cost
– the accountant’s view of cost stresses out-
of-pocket expenses, historical costs,
depreciation, and other bookkeeping
entries
– economists focus more on opportunity cost

2
 Definitions of Costs
• Labor Costs
– to accountants, expenditures on labor are
current expenses and hence costs of
production
– to economists, labor is an explicit cost
• labor services are contracted at some hourly
wage (w) and it is assumed that this is also what
the labor could earn in alternative employment

3
 Definitions of Costs
• Capital Costs
– accountants use the historical price of the
capital and apply some depreciation rule to
determine current costs
– economists refer to the capital’s original price
as a “sunk cost” and instead regard the implicit
cost of the capital to be what someone else
would be willing to pay for its use
• we will use v to denote the rental rate for capital

4
 Definitions of Costs
• Costs of Entrepreneurial Services
– accountants believe that the owner of a firm
is entitled to all profits
• revenues or losses left over after paying all input
costs
– economists consider the opportunity costs of
time and funds that owners devote to the
operation of their firms
• part of accounting profits would be considered as
entrepreneurial costs by economists
5
 Economic Cost
• The economic cost of any input is the
payment required to keep that input in
its present employment
– the remuneration the input would receive in
its best alternative employment

6
Two Simplifying Assumptions
• There are only two inputs
– homogeneous labor (l), measured in labor-
hours
– homogeneous capital (k), measured in
machine-hours
• entrepreneurial costs are included in capital costs
• Inputs are hired in perfectly competitive
markets
– firms are price takers in input markets
7
 Economic Profits
• Total costs for the firm are given by
total costs = C = wl + vk
• Total revenue for the firm is given by
total revenue = pq = pf(k,l)
• Economic profits () are equal to
 = total revenue - total cost
 = pq - wl - vk
 = pf(k,l) - wl - vk
8
 Economic Profits
• Economic profits are a function of the
amount of capital and labor employed
– we could examine how a firm would choose
k and l to maximize profit
• “derived demand” theory of labor and capital
inputs
– for now, we will assume that the firm has
already chosen its output level (q0) and
wants to minimize its costs
9
Cost-Minimizing Input Choices
• To minimize the cost of producing a
given level of output, a firm should
choose a point on the isoquant at which
the RTS is equal to the ratio w/v
– it should equate the rate at which k can be
traded for l in the productive process to the
rate at which they can be traded in the
marketplace

10
Cost-Minimizing Input Choices
• Mathematically, we seek to minimize
total costs given q = f(k,l) = q0
• Setting up the Lagrangian:
L = wl + vk + [q0 - f(k,l)]
• First order conditions are
L/l = w - (f/l) = 0
L/k = v - (f/k) = 0
L/ = q0 - f(k,l) = 0
11
Cost-Minimizing Input Choices
• Dividing the first two conditions we get
w f / l
  RTS (l for k )
v f / k

• The cost-minimizing firm should equate

the RTS for the two inputs to the ratio of
their prices

12
Cost-Minimizing Input Choices
• Cross-multiplying, we get
fk fl

v w

• For costs to be minimized, the marginal

productivity per dollar spent should be
the same for all inputs

13
Cost-Minimizing Input Choices
• Note that this equation’s inverse is also
of interest
w v
 
fl fk
• The Lagrangian multiplier shows how
much in extra costs would be incurred
by increasing the output constraint
slightly
14
Cost-Minimizing Input Choices
Given output q0, we wish to find the least costly
point on the isoquant
k per period
Costs are represented by
C1 parallel lines with a slope of
C3 -w/v
C2

C1 < C2 < C3
q0

l per period
15
Cost-Minimizing Input Choices
The minimum cost of producing q0 is C2

k per period This occurs at the

tangency between the
C1

C3
isoquant and the total cost
curve
C2

k* The optimal choice

q0 is l*, k*

l per period
l* 16
Contingent Demand for Inputs
• In Chapter 4, we considered an
individual’s expenditure-minimization
problem
– we used this technique to develop the
compensated demand for a good
• Can we develop a firm’s demand for an
input in the same way?

17
Contingent Demand for Inputs
• In the present case, cost minimization
leads to a demand for capital and labor
that is contingent on the level of output
being produced
• The demand for an input is a derived
demand
– it is based on the level of the firm’s output

18
 The Firm’s Expansion Path
• The firm can determine the cost-
minimizing combinations of k and l for
every level of output
• If input costs remain constant for all
amounts of k and l the firm may
demand, we can trace the locus of cost-
minimizing choices
– called the firm’s expansion path
19
 The Firm’s Expansion Path
The expansion path is the locus of cost-
minimizing tangencies
k per period The curve shows
how inputs increase
E
as output increases

q1

q0

q00
l per period
20
 The Firm’s Expansion Path
• The expansion path does not have to be
a straight line
– the use of some inputs may increase faster
than others as output expands
• depends on the shape of the isoquants
• The expansion path does not have to be
upward sloping
– if the use of an input falls as output expands,
that input is an inferior input
21
 Cost Minimization
• Suppose that the production function is
Cobb-Douglas:
q = kl
• The Lagrangian expression for cost
minimization of producing q0 is
L = vk + wl + (q0 - k  l )

22
 Cost Minimization
• The first-order conditions for a minimum
are
L/k = v - k -1l = 0
L/l = w - k l -1 = 0
L/ = q0 - k  l  = 0

23
 Cost Minimization
• Dividing the first equation by the second
gives us
w k  l 1  k
  1 
   RTS
v k l  l

• This production function is homothetic

– the RTS depends only on the ratio of the two
inputs
– the expansion path is a straight line
24
 Cost Minimization
• Suppose that the production function is
CES:
q = (k  + l )/
• The Lagrangian expression for cost
minimization of producing q0 is
L = vk + wl + [q0 - (k  + l )/]

25
 Cost Minimization
• The first-order conditions for a minimum
are
L/k = v - (/)(k + l)(-)/()k-1 = 0
L/l = w - (/)(k + l)(-)/()l-1 = 0
L/ = q0 - (k  + l )/ = 0

26
 Cost Minimization
• Dividing the first equation by the second
gives us
 1 1 1/ 
w  1 k  k 
     
v k  l l

• This production function is also

homothetic

27
 Total Cost Function
• The total cost function shows that for
any set of input costs and for any output
level, the minimum cost incurred by the
firm is
C = C(v,w,q)
• As output (q) increases, total costs
increase

28
 Average Cost Function
• The average cost function (AC) is found
by computing total costs per unit of
output
C(v ,w , q )
average cost  AC (v ,w , q ) 
q

29
 Marginal Cost Function
• The marginal cost function (MC) is
found by computing the change in total
costs for a change in output produced
C(v ,w , q )
marginal cost  MC(v ,w , q ) 
q

30
 Graphical Analysis of
Total Costs
• Suppose that k1 units of capital and l1
units of labor input are required to
produce one unit of output
C(q=1) = vk1 + wl1
• To produce m units of output (assuming
constant returns to scale)
C(q=m) = vmk1 + wml1 = m(vk1 + wl1)
C(q=m) = m  C(q=1)
31
 Graphical Analysis of
Total Costs
With constant returns to scale, total
Total
costs
costs
are proportional to output
AC = MC
C

Both AC and
MC will be
constant

Output
32
 Graphical Analysis of
Total Costs
• Suppose instead that total costs start
out as concave and then becomes
convex as output increases
– one possible explanation for this is that
there is a third factor of production that is
fixed as capital and labor usage expands
– total costs begin rising rapidly after
diminishing returns set in
33
 Graphical Analysis of
Total Costs
Total C
costs

Total costs rise

dramatically as
output increases
after diminishing
returns set in

Output
34
 Graphical Analysis of
Total Costs
Average
and MC is the slope of the C curve
marginal
costs MC
If AC > MC,
AC AC must be
falling

If AC < MC,
min AC
AC must be
rising
Output
35
 Shifts in Cost Curves
• The cost curves are drawn under the
assumption that input prices and the
level of technology are held constant
– any change in these factors will cause the
cost curves to shift

36
 Some Illustrative Cost
Functions
• Suppose we have a fixed proportions
technology such that
q = f(k,l) = min(ak,bl)
• Production will occur at the vertex of the
L-shaped isoquants (q = ak = bl)
C(w,v,q) = vk + wl = v(q/a) + w(q/b)
v w 
C(w ,v , q )  a  
a b  37
 Some Illustrative Cost
Functions
• Suppose we have a Cobb-Douglas
technology such that
q = f(k,l) = k l 
• Cost minimization requires that
w  k
 
v  l

 w
k   l
 v
38
 Some Illustrative Cost
Functions
• If we substitute into the production
function and solve for l, we will get
 /  
1/   
l q   w  /  v  /  


• A similar method will yield

 /  
1/   
k q   w  /  v  /  

39
 Some Illustrative Cost
Functions
• Now we can derive total costs as
C (v ,w , q )  vk  wl  q 1/  Bv  /  w  /  

where
B  (  )   /    /  
which is a constant that involves only

the parameters  and 

40
 Some Illustrative Cost
Functions
• Suppose we have a CES technology
such that
q = f(k,l) = (k  + l )/
• To derive the total cost, we would use
the same method and eventually get
C(v ,w , q )  vk  wl  q 1/  (v  / 1  w  / 1 )( 1) / 

C(v ,w , q )  q 1/  (v 1  w 1 )1/ 1

41
Properties of Cost Functions
• Homogeneity
– cost functions are all homogeneous of
degree one in the input prices
• cost minimization requires that the ratio of input
prices be set equal to RTS, a doubling of all
input prices will not change the levels of inputs
purchased
• pure, uniform inflation will not change a firm’s
input decisions but will shift the cost curves up

42
Properties of Cost Functions
• Nondecreasing in q, v, and w
– cost functions are derived from a cost-
minimization process
• any decline in costs from an increase in one of
the function’s arguments would lead to a
contradiction

43
Properties of Cost Functions
• Concave in input prices
– costs will be lower when a firm faces input
prices that fluctuate around a given level
than when they remain constant at that
level
• the firm can adapt its input mix to take
advantage of such fluctuations

44
 Concavity of Cost Function
At w1, the firm’s costs are C(v,w1,q1)
If the firm continues to
buy the same input mix
Costs Cpseudo as w changes, its cost
function would be Cpseudo
C(v,w,q1)

C(v,w1,q1)
Since the firm’s input mix
will likely change, actual
costs will be less than
Cpseudo such as C(v,w,q1)

w1 w 45
Properties of Cost Functions
• Some of these properties carry over to
average and marginal costs
– homogeneity
– effects of v, w, and q are ambiguous

46
 Input Substitution
• A change in the price of an input will
cause the firm to alter its input mix
• We wish to see how k/l changes in
response to a change in w/v, while
holding q constant
k 
 
l
w 
 
v 
47
 Input Substitution
• Putting this in proportional terms as
( k / l ) w / v  ln(k / l )
s  
(w / v ) k / l  ln(w / v )

gives an alternative definition of the

elasticity of substitution
– in the two-input case, s must be nonnegative
– large values of s indicate that firms change
their input mix significantly if input prices
change
48
Partial Elasticity of Substitution
• The partial elasticity of substitution between
two inputs (xi and xj) with prices wi and wj is
given by

( x i / x j ) w j / w i  ln( x i / x j )
sij   
(w j / w i ) x i / x j  ln(w j / w i )
• Sij is a more flexible concept than 
because it allows the firm to alter the
usage of inputs other than xi and xj
when input prices change 49
Size of Shifts in Costs Curves
• The increase in costs will be largely
influenced by the relative significance of
the input in the production process
• If firms can easily substitute another
input for the one that has risen in price,
there may be little increase in costs

50
 Technical Progress
• Improvements in technology also lower
cost curves
• Suppose that total costs (with constant
returns to scale) are
C0 = C0(q,v,w) = qC0(v,w,1)

51
 Technical Progress
• Because the same inputs that produced
one unit of output in period zero will
produce A(t) units in period t
Ct(v,w,A(t)) = A(t)Ct(v,w,1)= C0(v,w,1)
• Total costs are given by

Ct(v,w,q) = qCt(v,w,1) = qC0(v,w,1)/A(t)

= C0(v,w,q)/A(t)
52
 Shifting the Cobb-Douglas
Cost Function
• The Cobb-Douglas cost function is
C(v ,w , q )  vk  wl  q 1/  Bv  /  w  /  

where
B  (  )   /    /  
• If we assume  =  = 0.5, the total cost
curve is greatly simplified:

C(v ,w , q )  vk  wl  2qv 0.5w 0.5

53
 Shifting the Cobb-Douglas
Cost Function
• If v = 3 and w = 12, the relationship is
C(3,12, q )  2q 36  12q

– C = 480 to produce q =40

– AC = C/q = 12
– MC = C/q = 12

54
 Shifting the Cobb-Douglas
Cost Function
• If v = 3 and w = 27, the relationship is
C(3,27, q )  2q 81  18q

– C = 720 to produce q =40

– AC = C/q = 18
– MC = C/q = 18

55
Contingent Demand for Inputs
• Contingent demand functions for all of
the firms inputs can be derived from the
cost function
– Shephard’s lemma
• the contingent demand function for any input is
given by the partial derivative of the total-cost
function with respect to that input’s price

56
Contingent Demand for Inputs
• Suppose we have a fixed proportions
technology
• The cost function is
v w 
C(w ,v , q )  a  
a b 

57
Contingent Demand for Inputs
• For this cost function, contingent
demand functions are quite simple:
C(v ,w , q ) q
k (v ,w , q ) 
c

v a
C(v ,w , q ) q
l (v ,w , q ) 
c

w b

58
Contingent Demand for Inputs
• Suppose we have a Cobb-Douglas
technology
• The cost function is

C(v ,w , q )  vk  wl  q 1/  Bv  /  w  /  

59
Contingent Demand for Inputs
• For this cost function, the derivation is
messier:
C  1/    /     /   
k (v ,w , q ) 
c
 q Bv w
v   
 /  
 1/    w 
 q B 
 v 

60
Contingent Demand for Inputs
C 
l (v ,w , q ) 
c
  q 1/  Bv  /  w  /  
w   
  /  
 1/    w 
 q B 
 v 

• The contingent demands for inputs

depend on both inputs’ prices

61
 Short-Run, Long-Run
Distinction
• In the short run, economic actors have
only limited flexibility in their actions
• Assume that the capital input is held
constant at k1 and the firm is free to vary
only its labor input
• The production function becomes
q = f(k1,l)
62
 Short-Run Total Costs
• Short-run total cost for the firm is
SC = vk1 + wl
• There are two types of short-run costs:
– short-run fixed costs are costs associated
with fixed inputs (vk1)
– short-run variable costs are costs
associated with variable inputs (wl)

63
 Short-Run Total Costs
• Short-run costs are not minimal costs
for producing the various output levels
– the firm does not have the flexibility of input
choice
– to vary its output in the short run, the firm
must use nonoptimal input combinations
– the RTS will not be equal to the ratio of
input prices
64
 Short-Run Total Costs
k per period
Because capital is fixed at k1,
the firm cannot equate RTS
with the ratio of input prices

k1

q2
q1

q0

l per period
l1 l2 l3
65
 Short-Run Marginal and
Average Costs
• The short-run average total cost (SAC)
function is
SAC = total costs/total output = SC/q
• The short-run marginal cost (SMC)
function is
SMC = change in SC/change in output = SC/q

66
 Relationship between Short-
Run and Long-Run Costs
SC (k2)
Total SC (k1)
costs C

SC (k0)
The long-run
C curve can
be derived by
varying the
level of k

Output
q0 q1 q2 67
 Relationship between Short-
Run and Long-Run Costs
Costs

SMC (k0) SAC (k0) MC The geometric

AC relationship
SMC (k1) SAC (k1)
between short-
run and long-run
AC and MC can
also be shown

Output
q0 q1 68
 Relationship between Short-
Run and Long-Run Costs
• At the minimum point of the AC curve:
– the MC curve crosses the AC curve
• MC = AC at this point
– the SAC curve is tangent to the AC curve
• SAC (for this level of k) is minimized at the same
level of output as AC
• SMC intersects SAC also at this point
AC = MC = SAC = SMC
69
 Important Points to Note:
• A firm that wishes to minimize the
economic costs of producing a
particular level of output should
choose that input combination for
which the rate of technical substitution
(RTS) is equal to the ratio of the
inputs’ rental prices

70
 Important Points to Note:
• Repeated application of this
minimization procedure yields the
firm’s expansion path
– the expansion path shows how input
usage expands with the level of output
• it also shows the relationship between output
level and total cost
• this relationship is summarized by the total
cost function, C(v,w,q)
71
 Important Points to Note:
• The firm’s average cost (AC = C/q)
and marginal cost (MC = C/q) can
be derived directly from the total-cost
function
– if the total cost curve has a general cubic
shape, the AC and MC curves will be u-
shaped

72
 Important Points to Note:
• All cost curves are drawn on the
assumption that the input prices are
held constant
– when an input price changes, cost curves
shift to new positions
• the size of the shifts will be determined by the
overall importance of the input and the
substitution abilities of the firm
– technical progress will also shift cost
curves 73
 Important Points to Note:
• Input demand functions can be derived
from the firm’s total-cost function
through partial differentiation
– these input demands will depend on the
quantity of output the firm chooses to
produce
• are called “contingent” demand functions

74
 Important Points to Note:
• In the short run, the firm may not be
able to vary some inputs
– it can then alter its level of production
only by changing the employment of its
variable inputs
– it may have to use nonoptimal, higher-
cost input combinations than it would
choose if it were possible to vary all
inputs
75