E5 Managerial Economics

Module 5

Production analysis
Introduction
This module moves from the decisions of utility-maximising households
to examine the factors governing the behaviour of perfectly competitive
profit-maximising producers. It is a transition module that introduces
many central topics, among which are productivity, costs and economic
profits. The most profitable way to employ the firm’s resources to
produce a given product requires an understanding of production
function. Production function is simply an input-output relationship
between one or more factors of production (input) and the good or service
produced (output). This relationship is analysed and quantified during a
production study to determine the most economical combination of input
resources to obtain a given level of output. A study of production
functions is also fundamental to cost analysis. Once production function
has been identified, its cost function can be derived from the production
function. Hence the production affects the firm’s status in its industry.

Upon completion of this module you will be able to:

 outline the major decisions that firms must take.

 distinguish the short run from the long run.
 describe the relationship between average product, marginal
product and total product.
Outcomes
 apply the Iso-quant approach to production.
 distinguish (diminishing) returns to factor from (diminishing)
returns to scale.
 describe the rule of costs minimisation.

Long-run: All factors may be varied.

Production function: An input-output relationship between one or more

factors of production (input) and the good or
service produced (output). It is an equation or table
Terminology or graph showing the amount of output of a
product that a firm can produce per period of time
for each set of inputs. Both inputs and outputs are
measured in physical rather than in monetary units.

Short-run: The quantities of some inputs are variable while

97
Module 5

others are in fixed supply.

Production analysis
Once demand for a given product or service has been determined,
management decides the most profitable way to employ the firm’s
resources to produce that good or service. Such decisions involve an
understanding of production functions.

A production function is simply an input-output relationship between one

or more factors of production (input) and the good or service produced
(output). This relationship is analysed and quantified during a production
study to determine the most economical combination of input resources to
obtain a given level of output. Or conversely, it may involve the
determination of the maximum output obtainable from a given level and
mix of inputs. A production study may pertain not only to the production
of goods, such as planes, computer games, scanners, ice cream, but also
to the production of services, such as Internet services, hospitality
industry or health care.

The production function

Production function is an equation, table, or graph showing the amount of
output of a product that a firm can produce per period of time for each set
of inputs. Both inputs and outputs are measured in physical rather than in
monetary units. Technology is assumed to remain constant during the
period of the analysis. Technology summarises the feasible means (know-
how) of converting raw inputs, such as steel, labour and machinery, into
an output such as a car.

A study of production functions is also fundamental to cost analysis.

Once a firm’s production function has been identified, its cost function
can be derived from the production function, provided the market prices
of the input factors are known. Hence, the production function strongly
affects the firm’s status in its industry, and the study of production
functions is even more basic than the study of cost functions.

Like a demand function, a production function can be expressed as a

schedule, graph, or an equation such as
Q = f (K, L) (1)

where Q, a specific output, is a function of the input factors, K and L.

Note that the general ideas presented here are valid for any two inputs,
not just for labour and capital. It is important to remember that a
production function relates to some given level of technology. If the
technology changes through the upgrading of labour, materials,
machinery, equipment, processes, or management, the production
function changes accordingly.

98
 E5 Managerial Economics

The short run versus the long run

In production and cost theory, the distinction is made between the short
run, in which the quantities of some inputs are variable while others are in
fixed supply, and the long run in which all factors may be varied.
Consequently, it is useful to classify the inputs on the basis of whether or
not they are variable in the short run. Labour is typically regarded as
variable and capital as fixed in the short run.

It is important to understand that the long run does not refer to a long
period of time. It is a peculiarity of the economists’ jargon that the term
has no direct connection with time at all, and that the firm is likely to be
in a long-run situation for relatively short periods. When intending to
change its scale of production, the firm must continue to operate in a
short-run situation until its most-fixed factor becomes variable.

Production in the short run

As a manager, your job is to use the available production function
efficiently. This effectively means that you must determine how much of
each input to use to produce output. In the short run, some factors of
production are fixed, and this limits your choices in making input
decisions. For example, it takes several years for Honda Motor Company
to build an assembly line. The level of capital is generally fixed in the
short run. However, in the short run, Honda can adjust its use of inputs
such as labour and steel; such inputs are called variable factors of
production.

Therefore, the short run by definition is that period of time during which
at least one input is fixed. Variable inputs are those inputs whose
quantities are directly related to the level of production.

Mathematically, this function is captured by the following relationship:

Q = f ( K , L) (2)

where, K , refers to the fixed level of input of capital. The horizontal bar
in this equation indicates that the input factors to its left are regarded as
fixed in the production process under analysis, while the factor to its right
is variable. The quantity of the output product, Q, is the result of
combining a variable quantity of input factor, (skilled labour), with fixed
quantities of other input (buildings or equipment).

Measures of productivity
The productivity of a factor of production refers to the amount of output
that can be produced by that input, holding constant the input of all other
factors of production. Obviously, an input, such as human resources, can
be more productive if it works with modern mechanical and computer-

99
Module 5

assisted equipment and high-quality raw materials. Similarly, the plant or

equipment can be more productive if it is being operated by highly skilled
and well-trained workers. We use the phrase the state of technology when
we refer to the quality of the resources involved in the production. Table
5-1 is a numerical representative of this function.
Table 5-1 Alternative short-run production functions

Row Capital inputs Labour inputs (person month)

1 (machine month) (L)

2 (K) 0 1 2 3 4 5 6 7 8

3 1 0 15 35 77 112 130 145 155 162

4 2 0 22 48 80 125 161 188 208 223

5 3 0 35 72 120 180 255 310 358 387

6 4 0 500 102 162 230 305 385 444 493

Note that this table in fact represents four short-run production functions
and not just one. Column 2 shows the amount of capital used per month.
Accordingly, there are four possible fixed sizes of capital, varying from 1
to 4, K  1 , K  2 , K  3 and K  4 , with 1 being the smallest and 4
the largest plant. Furthermore, each K is combined with the variable
input of labour, varying from 0, column 3, to 8, the last column. The
shaded area, rows 3, 4, 5, and 6, shows the level of output (total product).
For example, row 3 represents a production function that combines 1 K
with different Ls, 0 to 8, while row 4 represents a different production
function that combines 2 units of K (a bigger plant) with labour, 0 to 8
units. Clearly, the manager has several plant options.

Demonstration problem
Referring to Table 5-1, suppose the firm is currently producing 160 units
of output and using a production function that employs 2 K. Now suppose
that the demand for the firm’s output increases permanently to, say 240,
with a possibility of a further 10 per cent increase in the following year.
How many ways can this firm meet the market demand?
Answer:

The firm can produce 240 units of output using its current plant
by stretching it almost to its limit. This will be done by combing
2K with 8L. Alternatively, the firm can produce the same output
by expanding its plant (K) using a production function that
combines either 3K (the next plant size) with 5L, or 4K with 4L.
However, when we consider the possibility of the additional 10
per cent increase in demand, the manager is left with only two
choices, plant 3, K  3 , and plant 4, K  4 .

100
 E5 Managerial Economics

Law of diminishing returns

Table 5-2, extracted from Table 5-1, shows that, in a given state of
technology and keeping the input of capital constant, K  2 , additional
units of labour (the variable input) yield increasing output per unit of
input up to a point. But eventually a point is reached beyond which
further additions of labour yield diminishing returns per unit of input. In
this illustration, the point of diminishing returns is reached at an input of
five units of L. Put differently, the relative productivity of the marginal
units of the variable inputs (L) initially increases (up to L = 4) and
diminishes subsequently. Hence, the initial increasing returns to the
variable input gives way to diminishing returns.

These, which are short-term phenomena, can be stated more broadly as

the law of variable proportions. They state that as more and more of the
variable factors are added to a given quantity of the fixed factors, the
increment to output attributable to each of the additional units of the
variable factor will increase at first, will later decrease, and will
eventually become negative.
Table 5-2 The law of variable proportions

Units of input Units of output Increment to Returns to the

(L) (for K  2 ) output variable input

0 0 - -

1 22 22 Increasing

2 48 26 Increasing

3 80 32 Increasing

4 125 45 Increasing

5 161 36 Diminishing

6 188 27 Diminishing

7 208 20 Diminishing

8 223 15 Diminishing

Total product, average product and marginal product

Total product (TP) is simply the maximum level of output that can be
produced with a given amount of inputs, Column 2, in Table 5-2.

Average Product (APL), the most popular measure of productivity, is

defined as total product divided by the quantity of the variable input (L)
used, APL = Q/L.

101
Module 5

Marginal Product (MPL) is the change in total output attributable to the

Q
last unit of the input of labour, MPL  .
L

Note that the average and the marginal product concepts can also be
defined for the fixed input of capital. Accordingly, APK, the average
product of capital, is defined as total product divided by the quantity of
the fixed input (K), whereas the marginal product of capital (MPK) is the
Q
change in total output divided by the change in capital, MPK  .
L

Figure 5-1 shows graphically the relationship among total product,

marginal product and average product. The first thing to notice about
these curves is that as the use of labour increases between points 0 and A,
the slope of the total product curve increases (becomes steeper). That is,
output increases at an increasing rate. Since the slope of the TP curve is
MPL, the marginal product curve will be increasing over this range. The
range over which marginal product increases is known as the range of
increasing marginal returns.

Figure 5-1 also shows that marginal product reaches its maximum at point
A, where four units of labour are employed. This is indicated on the TP
curve by the inflection point, point A. That is, the point at which the
curve changes from concave upward to concave downward. As the usage
of labour increases from the fourth through to the eighth unit, total output
increases, but at a decreasing rate. This is why marginal product declines
between four and eight units of labour but is still positive. The range over
which marginal product is positive but declining is known as the range of
diminishing marginal returns to the variable input.

Marginal product becomes negative when more than nine units of labour
are employed. After a point, using additional units of input actually
reduces total product, which is what it means for marginal product to be
negative. The range over which marginal product is negative is known as
the range of negative marginal returns.

102
 E5 Managerial Economics

Figure 5-1

Q C

B TP

140 A

AP
0 4 5 9 MP L

The relationship between marginal product and total product, exhibited in

Figure 5-1, can be summarised as:
1. As long as the MPL curve is rising, the total product curve increases
at an increasing rate and is convex to the horizontal axis.
2. The quantity of input L at which the TP curve changes its curvature
corresponds precisely to the point at which the MPL curve peaks. This
occurs at approximately four units of input, as shown by point A.
3. When the total-product curve reaches point C, its maximum MPL is
zero. Beyond this point, the marginal product is negative and the total
product declines.

The law of variable proportions discussed above can be also discussed in

the context of the average revenue. The average product of labour, APL,
can be extracted from Table 5-2 as follows:

L APL

0 0

1 22

2 24

3 26.67

4 31.25

5 32.2

6 31.33

7 29.7

8 27.87

103
Module 5

APL is initially subject to the law of increasing returns, reaching a

maximum at approximately 5 units of labour. Beyond that, it becomes
subject to the law of diminishing returns. This relationship is also
exhibited in Figure 5-1. To sum up:
1. The average product increases as variable input increases as long
as the marginal product exceeds the average product.
2. When the marginal product is less than the average product, the
average product decreases as variable input increases.
3. When the average product is at a maximum, the average product
and marginal product are equal.

Figure 5-1 also illustrates the three stages of a typical production

function.

Stage 1: This stage extends from zero input of the variable factor to the
level of input where the average product is at a maximum. In this stage,
the fixed factors are excessive relative to the variable input.
Consequently, output can be increased by increasing the variable input
relative to the fixed input.

Stage 2: This stage extends from the end of Stage 1 (the point where the
marginal product and the average product are equal) to the point where
the marginal product is zero and the total product is at a maximum, point
C in Figure 5-1. Stage 2 is a rational stage in which relatively good
balance has been achieved between the variable and fixed inputs.

Stage 3: In this stage, in which input is greater than nine units, the
variable-input factor is excessive relative to the fixed factors, the
marginal product is negative, and the total product is falling. It is
completely irrational to produce in this stage.

The optimal level of employment of a variable input

To determine the optimum input-output relationship for maximum profit,
we must shift the analysis from physical input-output relationships to
economic relationships. The precise amount of the input variable needed
to produce maximum profit depends upon the revenue and cost associated
with hiring an additional unit of the variable input, where the cost side
depends on the price of the input variable, while the revenue side is a
function of marginal product of the input variable as well as the selling
price of the output.

The additional revenue associated with hiring an extra unit of the input of
labour equals the value of output produced by this additional hiring.
Being referred to as the value of marginal product (VMP), this is defined
as MPL x P, where MPL = Q/L and P is the price of the product sold.
On the other hand, the additional cost of hiring is the wage rate, W.
Therefore, optimality requires that:

104
 E5 Managerial Economics

MP L x P = W (3)
Figure 5-2

Recalling that the rational stage of production is stage 2, where MPL is

diminishing (Figure 5-2), the firm should hire labour up to a point where
the value of its marginal product equals its marginal cost, L0.
Alternatively, equation (4) expresses the employment condition in real
terms. The left hand side of this equation is the marginal physical product
of labour, whereas the right hand side is the real wage:

MP L = W/P (4)

Demonstration problem
Suppose the production function is given by the following equation
Q = 20L – 5L 2

where L is defined in thousands of units of labour. What is the expression

for the marginal product and average product of labour function? At what
level of labour does TP reach its maximum? If P = $5 per unit and W =
$2 per hour, what is the optimal level of employment of labour?
Answer:

dQ
MPL   20  10 L.
dL
Q
APL   20  5 L
L
Note ‘d’ in the MP expression is the mathematical notation of
derivative. The marginal product is the derivative of the total
product with respect to quantity. That is, it shows the amount of
change in output when the input of labour changes
infinitesimally. TP is maximised when MPL is set equal to zero:
20 – 10L = 0, hence L = 2 (000).

VMP L= W, $5(20 –10L) = $2, and L = 1.96 (000) = 1960

105
Module 5

The production function of multiple variable inputs (Isoquants and

Isocosts)
The firm’s production problem is to choose the optimal levels of all
inputs, not just a single input, that enter the production process such that
output is maximised for a given budget constraint. Therefore, our next
task is to examine the optimal choice of capital and labour in the long run,
when both inputs are free to vary. In the presence of multiple variables of
production, there exist various combinations of inputs that enable the
manager to produce the same level of output.

The basic tools for understanding how alternative inputs can be used to
produce output are isoquants and isocosts. An isoquant is a line joining
combinations of inputs (K, L) that generate the same level of output. The
word isoquant comes from the Greek word iso meaning equal and the
Latin word quantus meaning quantity. As with indifference curves, there
will be a family of isoquant curves, with higher curves preferred to lower
curves. Similarly, the curves do not cross, and they are convex to the
origin. Figure 5-3 depicts a typical set of isoquants. Because input
bundles A and B both lie on the same isoquant, each will produce the
same level of output, namely Q2 units. Input mix A implies a more
capital-intensive plant than the input mix B. As more of both inputs are
used, a higher isoquant is obtained. Thus as we move in the northeast
direction in the figure, each new isoquant is associated with higher and
higher levels of output,
Q 2 > Q 1 > Q 0.
Figure 5-3

Two points need to be appreciated.

1. Unlike indifference curves, isoquant curves are not negatively
sloped throughout their length. Isoquants may bend back, as
shown in Figure 5-4, at both relatively high and relatively low
ratios of capital to labour. The reason for this difference is that in
consumer theory we assume that marginal utility would never be
negative, whereas in producer theory we make no such restrictive
assumption concerning marginal products.

106
 E5 Managerial Economics

Figure 5-4

To understand this better, we should distinguish between

technical efficiency, whereby there is no wasted resources, and
economic efficiency, whereby there is no waste of expenditure.
Points A and B on the isoquant Q1, in Figure 5-4, are technically
inefficient points. For example, the firm at point B is using too
much labour relative to point C. By moving from B to C, the
firm, while maintaining its output level (Q1), can save (LB – LC)
in labour. Note that in reducing the input of labour, employment
of K remains unchanged. Similarly, point A is also technically
inefficient. The two perpendicular (bold) lines drawn against the
vertical and horizontal axis, separate the space into the economic
and uneconomic region. Points above and the to the right of these
lines (outside the box) such as B and A, are in the uneconomic
region, while points below and to the left of these two lines
(inside the box), such as C, are in economic region.
2. Furthermore, isoquants are convex. The reason isoquants are
typically drawn with a convex shape is that inputs such as capital
and labour are not perfectly substitutable. In Figure 5-3, for
example, if we start at point A and begin substituting labour for
capital, it takes increasing amounts of labour to replace each unit
of capital that is taken away. The slope of an isoquant at any
point can, therefore, be expressed as the ratio of the amount of
capital that can be subtracted from the production process to the
amount of labour that is added to the production process, such
that output level remains constant. This ratio, known as the
marginal rate of substitution (MRTS) is the rate at which labour
and capital can substitute for each other:
K
MRTS   (5)
L

107
Module 5

where ∆K is the decrement to capital that will just allow output to

remain unchanged given a one-unit increment, ∆L, to the labour
input.

Shapes of Isoquants
The possibility of substitution of input L for input K and the degree of
substitution between them determine the shape and slope of the isoquant.
In addition to the normal shape of isoquants illustrated above, there are
peculiar shapes associated with perfect substitutes and perfect
complements, as shown in Figure 5-5. Panel (a) shows the right-angled
isoquants resulting from two inputs that are perfectly complementary; that
is, input of a single variable by itself will not produce any output. To
obtain output, both Y and X must be input in a fixed ratio. For example,
the manufacturing of a car requires an input of one body and four wheels,
a ratio that never changes. Therefore, if two bodies are input, eight wheels
must also be input to obtain two units of output, because the two inputs
are complementary. Note that in panel (a), Y = body casting, and X =
wheels.
Figure 5-5

(a) (b)

Panel (b) shows the isoquants resulting from two inputs that are perfect
substitutes for one another. Suppose that a firm is choosing between two
types of computers to store company data. One has a high-capacity hard
drive that can store 10 gigabytes of data, while the other has a low-
capacity hard drive that can store five gigabytes of data. If it needs to
store 100 gigabytes of data, it could either purchase 10 high-capacity
computers and no low-capacity computers (the vertical intercept), or it
could purchase no high-capacity computers and 20 low-capacity
computers (the horizontal intercept). Or it could purchase five high-
capacity computers and 10 low-capacity computers. Note that in panel
(b), Y = high-performance computers, while X = low-performance
computers.

For the fixed-proportions production function (perfect complements), the

expression for the production function is:
Q = min (4W, casting) (7)

108
 E5 Managerial Economics

On the other hand, the production function for the linear (perfect
substitutes) isoquants is represented by

Q = 10H + 5L (8)

For this production function, the slope is constant and the MRST does not
change as we move along the curve.

Cobb-Douglas production function

This is an intermediate between a linear production function and a fixed
proportions production function. This production function is known as the
Cobb-Douglas production function, and it is given by the formula
Q  AK a Lb (9)

where A, a, and b are positive constants. A is the scale variable, where a

and b (the exponents) are the elasticity of output with respect to capital
and labour respectively.

In the Cobb-Douglas production function, capital and labour can be

substituted for each other. Unlike a fixed proportions production, capital
and labour can be used in variable proportions. Unlike a linear production
function, though, the rate at which labour can be substituted for capital is
not constant as you move along an isoquant. Note, however, that this type
of production function is not the only one that allows substitutability
between L and K. It is, however, one that has several friendly
mathematical properties.

Isocosts
An isocost line is a locus of combinations of inputs that require the same
total expenditure. Let us express the firm’s expenditure on inputs as
TC = K x PK + L x PL (10)

where TC is the total dollar expenditure; PK and PL are the unit prices of
capital and labour, respectively; and K and L are the number of physical
units of capital and labour that are to be employed in the production
process. This can be rearranged to appear as
TC PL
K  L (11)
PK PK

in which form it is perhaps more recognisable as a linear equation

explaining K in terms of L and three known values (TC, PL, and PK). It
can therefore be plotted in the same space as the isoquant curves. The
intercept of the isocost line on the capital axis occurs when L = 0, and this
intercept value of K is simply the total expenditure divided by the price of
the capital units, resulting in a particular physical quantity of capital units,
TC/PK. As we purchase units of labour, it is evident that our purchase of

109
Module 5

capital units, for the same expenditure (isocost) level, is drawn down in
the ratio of the price of labour to the price of capital.
Figure 5-6

Figure 5-6 shows the graph of isocost lines for a variety of different total
cost levels, TC1, TC2, and TC3, where TC3 > TC2 > TC1. In general, there
are an infinite number of isocost lines, one corresponding to every
possible level of total cost.

Demonstration problem
Suppose due to a general slowdown in the economy, the price of all
factors of production, labour and capital, decline by the same percentage.
How would this change impact the isocost line?
Answer:

A proportionate decrease in the price of both factors leaves the

slope of the isocost line, PL/PK, intact while shifting it to the
right in a parallel fashion

Cost minimisation (economic efficiency)

For each output level, there will be a minimum-cost combination of the
factors necessary to produce that output level. In Figure 5-7, we show
three isoquant curves representing 10, 20, and 30 units of output. The cost
of producing 10 units is minimised at point A, since any other
capital/labour combination producing 10 units, such as at A', will lie to
the right of the isocost line JK and would thus require a larger total
expenditure to purchase, J'K'. Recall that the intercepts on the capital and
labour axes are equal to TC/PK, where PK and PL are presumed to remain
unchanged. Hence, larger total expenditures are represented by higher
intercept points and higher isocost lines. Similarly, the output level 20 is
produced at the least-cost by the input combination represented by point
B, as 30 units of output is produced at the least-cost by the input
combination at point C.

110
 E5 Managerial Economics

Figure 5-7

Input cost is minimised for a given output level, where the isoquant
representing that output level is just tangent to the lowest attainable
isocost line. Equivalently, output is maximised for a given input
expenditure level where the isocost line representing that expenditure
level is just tangent to the highest attainable isoquant. At the point of
tangency between the isoquant and the isocost curve, the slopes of these
curves are the same. That is, optimality requires that the rate at which the
firm can technically substitute labour for capital equals the rate that the
market allows it to.
MPL PL
MRTS   (11)
MPK PK
Rearranging terms, we have
MPL MPK
 (12)
PL PK

This alternative arrangement holds that cost minimization or output

maximisation requires that the ratios of the marginal products to prices of
all inputs be equal. This is useful in that this way the rule of optimisation
can be generalised to a larger number of inputs, 1, 2, 3,…..
MP1 MP2 MP3
   .... (13)
P1 P2 P3

Demonstration problem
Suppose MPL = 1, MPK = 3, whereas PL and PK are respectively $1 and
$2. Is the firm that faces these factors optimising information?

111
Module 5

Answer:

MPL PL MPL 1
Optimality requires  . Plugging in:  ,
MPK PK MPK 3
PL 1
whereas  . Therefore, the rate by which the firm is able
PK 2
to substitute labour for capital, 1/3, is lower than the rate the
market allows it to, 1/2. As such, the firm is not minimising its
costs. Specifically, the firm is using too much labour and too little
capital. In terms of a diagram, Figure 5-8, below, portrays this
MPL
situation. Note that, at point A, MRTS   1 / 3 is the
MPK
PL
slope of the isoquant, and  1 / 2 is the slope of the isocost
PK
line. The firm needs to substitute K for labour. Doing so will
increase MPL and decreases MPK , and hence will increase
MRTS toward 1/2. The process of substitution stops when
the MRTS equals the price ratio (1/2).

Figure 5-8

Demonstration problem (mathematically challenging)

Suppose the production is represented by a Cobb-Douglas production
function, Q  K 0.5 L0.5 . Also, suppose that PK and PL are $10 and $5,
respectively.
a. Determine the optimal combination of inputs.
b. If the manager has $500 to spend on labour and capital, how many
units of K and L should the firm acquire?

112
 E5 Managerial Economics

Answer:

a. First we need to take the (partial) derivative of Q to find the

marginal products of labour and capital:
MPK  (0.5) K 0.5 L0.5 and MPL  (0.5) K 0.5 L0.5 . Hence,
MPL 0.5K 0.5 L0.5 K P 1
  . Setting this equal to L  , we
MPK 0.5K 0.5 L0.5 L PK 2
K 1
have:  . Meaning that for every one unit of K, the firm
L 2
needs to hire two units of L in order to minimise its costs, L =
2K.

b. The isocost line is represented by

TC  PK xK  PL xL  10 .K  5 .L  $ 500 . Substituting in for
either L or K: $500 = 10(K) + 5(2K) = 20K, hence K = 25, L =
50.

In Figure 5-7, the line that connects A, B, and C is referred to as the

expansion path (EP). The expansion path, or more precisely, the long run
expansion path, is a locus of the tangency points between various
isoquant and isocosts that shows the least-cost combinations of labour and
capital a firm would choose as it expanded its output level. This assumes
constant price of labour and capital and constant state of technology.

Input substitution
Economic efficiency depends on the relative factor prices. If the price of
one factor changes, all else being kept constant, the profit-maximising
firm will attempt to substitute away from the factor that has become
relatively more expensive and in favour of the factor that has become
relatively less expensive. Suppose the initial situation in Figure 5-9, is
point A, where output level Q0 is being produced economically efficiently
by a combination of K0 and L0.

Now suppose that labour prices rise, for example, because of a new
agreement with the labour union. This causes an increase in the cost of
labour and the isocost line rotates down (clockwise) from EF to EF'. If
the firm wishes to maintain its current expenditure, it cannot produce Q0.
Alternatively, if the firm wishes to maintain its output level at Q0 to hold
its market share, it will need to spend more money on the inputs.

In the long run, the firm will substitute capital for labour, that is, the firm
will increase its plant size to K 1 and decrease its labour input to L1,
point B. The increased input price ratio, PL/PK has caused the production
process to become relatively more capital intensive. Note that E'F'', the
new isocost line that is tangent to the isoquant Q0 at point B, is parallel to
EF'.

113
Module 5

Figure 5-9

Returns to scale
Figure 5-10 illustrates three sets of isoquants for different production
processes. In each panel, the points on the ray from the origin show the
proportion by which output increases when both inputs are increased
proportionately. The distance between the successive isoquants brought
about by a proportionate increase in inputs is a reflection of how output
responds to changes in inputs. For example, doubling all inputs means
doubling the distance from the origin along the ray. We refer to changing
all inputs by the same proportion as a change in scale. What is the effect
on output of a change in scale?

In panel (a), we find that doubling inputs leads to a doubling in output.

That is, the move from point A to point B doubles the distance from the
origin and also doubles output from 100 to 200. The property of output
increasing proportionately to the scale of inputs is called constant return
to scale, because the amount of output per bundle of inputs is always
constant.

A production process exhibits constant returns to scale if increasing all

inputs proportionately increases output in equal proportion.

Unlike the proportionately spaced isoquants in panel (a) of Figure 5-10,

panel (b) has isoquants that bunch closer together and panel (c) has
Isoquants that spread out as the distance from the origin along a ray
increases. In panel (b), we find that doubling inputs leads to a more than
doubling in output (tripling). That is, the move from point A to point B
triples output and hence less than doubles the distance between 100 and
200 from the origin. The property of output increasing more than
proportionately to the scale of inputs is called increasing return to scale,

114
 E5 Managerial Economics

because the amount of output per bundle of inputs increases with the
scale.

Panel (c) shows the case of a production function in which output

increases less than proportionately to scale. That is, the move from point
A to point B more than doubles the distance between 100 and 200 from
the origin and also less than doubles output (50 per cent increase). The
property of output increasing less proportionately to the scale of inputs is
called decreasing returns to scale, because the amount of output per
bundle of inputs falls as scale increases.
Figure 5-10

Why might some production processes exhibit constant returns to scale

and others exhibit increasing or decreasing returns? One answer is that a
larger scale of operation often allows workers and managers to specialise
in different tasks. Henry Ford’s assembly line, for example, could
produce more cars than the same workers and equipment could do if
production had to be done one car at a time in separate garages, with each
worker designing, machining, assembling and finishing the cars. On the
other hand, large scale can also introduce inefficiencies, often associated
with breakdown of coordination and critical information flows.

From a policy perspective, increasing returns to scale seem to imply that

large firms are desirable and should be encouraged, because they are able
to produce more efficiently than small ones and thus should be able to sell
their product at a lower price. Yet, as we will see later, large firms often
have greater market power that would allow them to charge high prices,
and so large firms are sometimes broken into smaller, competing firms to
encourage lower prices. Alternatively, as in the case of regional electric
utilities, large firms may be regulated by government bodies.

Another way to describe the effects of different returns to scale is to

graph output against scale. Assuming that we have fixed a particular ratio
of inputs (that is, we have limited ourselves to a particular ray from the
origin), scale is measured on a graph like Figure 5-10 as the distance from
the origin. Plotting scale versus output leads to Figure 5-11. The product-
to-scale curve for decreasing returns to scale is concave with respect to
the bottom of the graph. For increasing returns to scale, the curve is
convex; and for constant returns to scale, it is a ray from the origin. The

115
Module 5

positioning of the curves in this Figure has no significance, as each type

of curve will be positioned according to the equation that represents the
expansion path.
Figure 5-11

Scale (K, L)

Testing production function for returns to scale

If the production function is known, it can be analysed algebraically for
returns to scale. Suppose we have a Cobb-Douglas production function
Q  AK a Lb . This is a general expression assuming no prior knowledge
of returns to scale. Let us double the inputs of K and L
Q'  A(2 K ) a (2 L) b
or
Q'  (2) a b ( AK a Lb )  (2) a b .Q

Hence, if (a + b) = 1, Q' = 2Q, which is the case of constant returns to

scale. If (a + b) >1, Q' > 2Q, which is the case of increasing returns to
scale, and if (a + b) <1, Q' < 2Q, which is the case of decreasing returns to
scale.

Technological changes
So far, we have treated the firm’s production function as fixed; that is, it
remains stationary over time. But as knowledge in the economy evolves
and as firms acquire know-how through experience and investment in
research and development, a firm’s production function will change. The
notion of technological progress captures the idea that production
functions can shift over time. In particular, technological progress refers
to a situation in which a firm can achieve more output from a given
combination of inputs, or equivalently, the same amount of output from
fewer inputs.

116
 E5 Managerial Economics

We can classify technological progress into three categories: neutral

technological progress, labour-saving technological progress, and capital-
saving technological progress. Figure 5-12 illustrates neutral
technological progress. In this case, an isoquant corresponding to a given
level of output (10 units) shifts inward (indicating that lesser amounts of
labour and capital are needed to produce a given output), but the
isoquants shift so as to leave MRTS unchanged along any ray (e.g., OA)
from the origin. Under neutral technological progress, in effect, the firm’s
entire map of isoquants is simply relabelled, each one now corresponding
to a higher level of output, but the isoquants themselves retain the same
shape.
Figure 5-12

Figure 5-13 illustrates labour-saving technological progress. As before,

the isoquant corresponding to a given level of output shifts inward, but
now along any ray from the origin, the isoquant becomes flatter,
indicating that the MRTS is now less than it was before. You should
recall that MRTS = MPL/MPK, so the fact that the MRTS decreases
implies that under this form of technological progress the marginal
product of capital increases more rapidly than the marginal product of
labour. This form of technological progress would arise when technical
advances in capital equipment, robotics, or computers increase the
marginal productivity of capital relative to the marginal productivity of
labour.

117
Module 5

Figure 5-13

Figure 5-14 depicts capital-saving technological progress. Here, as an

isoquant shifts inward, MRTS increases, indicating that the marginal
product of labour is increasing more rapidly than the marginal product of
capital. This form of technological progress would arise if, for example,
the education or skill level of the firm’s actual (and potential) work force
rose, increasing the marginal productivity of labour relative to the
marginal product of capital.
Figure 5-14

118
 E5 Managerial Economics

Module summary
Production of any goods requires input of at least two basic factors. In the
short run, there has to be at least one fixed factor of production and one
variable. This results in the law of variable proportions: diminishing and
increasing. In the process of production, as the input of the variable factor
Summary (labour) increases, the marginal product of labour first increases
(increasing returns), and then decreases (diminishing returns).

An isoquant shows the various combinations of two inputs that can be

used to produce a specific level of output. Its slope, the ‘marginal rate of
technical substitution (MRTS),’ is equal to the ratio of the marginal
products of the two inputs. In order to minimise production costs or
maximise profit, the firm must produce where an isoquant is tangent to an
isocost.

Returns to scale is a long-range planning concept that enables

management to determine the effect of increasing all inputs by the same
proportion. ‘Constant,’ ‘increasing,’ and ‘decreasing’ returns to scale
refer to the situation where output changes, respectively, by the same, by
a larger, and by a smaller proportion than inputs. Increasing returns to
scale arise because of specialisation and division of labour and from
using specialised machinery. Decreasing returns to scale arise primarily
because as the scale of operation increases, it becomes more and more
difficult to manage the firm and coordinate its operations and divisions
effectively. In the real world, most industries seem to exhibit near-
constant returns to scale.

119
Module 5

Assignment
1. The manager of a plant calculated the cost at different output levels.
The result is in the table below:
Units of Total Product
Labour (Units)
0 850
Assignment 10 1,700
20 3,500
30 6,900
40 10,000
50 11,500
60 12,600
70 11,550
80 10,400

Find the average product and marginal product of labour for

different levels of labour input.
2. A firm uses only labour to produce to output Q. The production
function is:
1/2
Q = 10 x L
If labour costs $ 25.00 per hour, then find the firm’s total cost,
average cost, and marginal cost functions.

120
 E5 Managerial Economics

Assessment
1. Cost of labour is $100 per unit and capital is $200 per unit. If the
marginal product of labour is 1,000 and marginal product of capital of
is 2,500, then is it optimal to use the combination of 10 labour and 30
capital? If not, then how should the owner revise the combination?
Assessment
2. A firm makes an optimal production decision based on its technology
and prices of inputs. The inputs that are used are labour (L) and
capital (K). Suppose labour cost is $20 per unit and capital cost is $10
per unit. What is the MRTS for production of 500 units? If prices of
inputs remain the same, will the MRTS change for production of
1,000 units? If the firm employs 100 units of labour and 100 units of
0.5 0.5
capital with the technology K L , then is the firm optimising its
production decision?

121
Module 5

Assessment answers
1. Wage (w) = $100, Rent (r) = $200

MPL = 1,000, MPK = 2,500

MPK/r = 2,500/200 = 12.5 > MPL/w = 1,000/100 = 10

Therefore more capital (or less labour) needs to be employed until

MPK/r = MPL/w.

2. In equilibrium (when optimal decisions are made) MRTS = Input

Cost Ratio = 10/20 = 0.5. MRTS at an output level of 500 should
not differ from MRTS at an output level of 1,000 if the input costs
do not change.

Technology = K0.5L0.5, MPK = 0.5K-0.5L0.5 and MPL = 0.5K0.5L-0.5

Therefore, MPK/MPL = L/K. When optimal decisions are made,

L/K = 0.5, and
L = 0.5K

If 100 L and 100 K are used to produce any level of output, since
L/K = 1 > 0.5.

122
 E5 Managerial Economics

References
Baye, M. (2002). Managerial Economics and Business Strategy. Irwin:
McGraw Hill.
Besanko, D., & Braeutigam, R. (2002). Microeconomics: An Integrated
References Approach. New York: Wiley & Sons.
Douglas, J. E. (1992). Managerial Economics: Analysis and Strategy.
Upper Saddle River, NJ: Prentice Hall.
Pindyck, R. S., & Rubinfeld, D. (2001). Microeconomics, 5th Edition.
Upper Saddle River, NJ: Prentice Hall.
Salvatore, D. (1993). Managerial Economics in a Global Economy.
Irwin: McGraw Hill.

123