Production Theory

2
Session Objectives
What is a production function?
How is the economic short run distinguished from long run?
What are average and marginal products? How are they
related?
What are economies of scale and how are they different from
returns to factors of production?
3
Firms Objective
The economic goal
of the firm is to
maximize profits.
4
Firms Objective
Profit = Total revenue Total cost
the amount a
firm receives
from the sale
of its output
the market
value of the
inputs a firm
uses in
production
Total Revenue = Price Quantity
5
Costs: Explicit vs. Implicit
Recollect Principle #2: The cost of something is what you give up
to get it.
A firms cost of production includes all the opportunity costs of
making its output of goods and services.
Total opportunity costs include both implicit and explicit costs :
Explicit costs input costs that require an outlay of
money by the firm
Implicit costs input costs that do not require an
outlay of money by the firm
6
Costs: Explicit vs. Implicit
Example :
You need Rs.1,00,000 to start your business. The interest rate
is 5%.
Case 1: borrow Rs.1,00,000
explicit cost = Rs.5000 interest on loan
Case 2: use Rs.40,000 of your savings, borrow the other
Rs.60,000
explicit cost = Rs.3000 (5%) interest on the loan
implicit cost = Rs.2000 (5%) foregone interest you could have
earned on your Rs.40,000
In both cases, total (exp + imp) costs are Rs.5000.
7
Economic Profit vs. Accounting Profit
This example shows that an important implicit cost is the cost
of capital, the foregone returns you could have earned had
you used your savings to buy bonds or other assets instead of
investing them in your business.
Accounting profit
= total revenue minus total explicit costs
Accountants keep track of how much money flows into and
out of the firm, so they ignore implicit costs.
Economic profit
= total revenue minus total costs (including explicit and
implicit costs)
Economists study the pricing and production decisions of firm,
which are affected by implicit as well as explicit costs
Accounting profit ignores implicit costs, so its higher than
economic profit.
8
Economic Profit vs. Accounting Profit
9
Economic Profit vs. Accounting Profit
Quick Activity:
The equilibrium rent on office space has just
increased by Rs. 5000/month. Compare the effects
on accounting profit and economic profit if
a) you rent your office space
b) you own your office space
10
Economic Profit vs. Accounting Profit
Answers:
The rent on office space increases Rs.5000/month
a) You rent your office space.
Explicit costs increase Rs. 5000/month. Accounting profit &
economic profit each fall Rs.5000/month
b) You own your office space.
Explicit costs do not change, so accounting profit does not
change. Implicit costs increase Rs.5000/month (opportunity
cost of using your space instead of renting it), so economic
profit falls by Rs.5000/month.
11
The Production Function
Production is the process of transforming inputs into output.
The production function indicates that maximum level of
output the firm can produce for any combination of inputs.
12
The Production Function
For simplicity, assume that all inputs or factors of production can be
grouped into two broad categories, labour (L) and capital (K). The
general equation :
Q = f (K, L)
This function defines the maximum rate of output (Q) per unit of time
obtainable from a given rate of capital and labour input. Output may
be physical units such as automobiles or washing machines, or it may
be intangible, as in the case of medical care, transportation and
education.
Why stress on maximum?
Suppose the firms fail to organize or manage resources efficiently and
produce less than the maximum output for given input rates. In a
competitive environment, such firms are not likely to survive because
competitors using efficient production techniques will be able to
produce at lower cost, sell at lower prices, and ultimately drive
inefficient producers out of market.
13
The Production Function
The production function
indicates the maximum output
that can be obtained from a
given combination of inputs
that is, we assume the firm is
technically efficient.
14
The Production Function for Auto Parts
Consider a multi-product firm that supplies parts
to major U.S. auto manufacturers. Its production
function is given by Let
Q = F(L, K)
where Q is the quantity of specialty parts
produced per day, L is the number of workers
employed per day, and K is plant size (measured
in thousands of square feet).
15
The Production Function for Auto Parts
This table shows the quantity of output that can be obtained
from various combinations of plant size and labour
Number of Pl ant Si ze (000s)
Workers 10 20 30 40
10 93 120 145 165
20 135 190 235 264
30 180 255 300 337
40 230 315 365 410
50 263 360 425 460
60 293 395 478 510
70 321 430 520 555
80 346 460 552 600
90 368 485 580 645
100 388 508 605 680
16
Short Run vs. Long Run
The short run refers to the
period of time in which one
or more of the firms inputs is
fixedthat is, cannot be
varied
Inputs that cannot be varied in
the short run are called fixed
inputs.
Inputs that can vary are called
(not surprisingly) variable inputs The long run is the period of
time sufficiently long to allow
the firm to vary all inputse.g.,
plant size, number of trucks, or
number of apple trees.
17
Total, Average and Marginal Product
Start with Short-Run Production Using Single Variable Input
Total product of labour is the maximum rate of output
forthcoming from combining varying rates of labour input with a
fixed capital input.
Denoting the fixed capital input as K
o
, the total product of labour
function is TP
L
= f (K
o
,L).
Average product is the total product per unit of the variable
input and is found by dividing the rate of output by the rate of
the variable input. The average product of labour function is AP
L
= TP
L
/L
Marginal product of any input is the increase in output arising
from an additional unit of that input, holding all other inputs
constant. Marginal product of labor MP
L
= Q / L. The marginal
product is the slope of the total product curve.
18
Total, Average and Marginal Product
Production of specialty parts, assuming
a plant size of 10,000 square feet
Number of Total Margi nal
Workers Product Product
10 93
20 135 4.2
30 180 4.5
40 230 5
50 263 3.3
60 293 3
70 321 2.8
80 346 2.5
90 368 2.2
100 388 2.0
110 400 1.2
120 403 0.3
130 391 -1.2
140 380 -1.1
In our example, the
marginal product of labor
(MP
L
) is the extra output of
auto parts realized by
employing one additional
worker, holding plant size
constant
19
Law of Diminishing Returns
As units of a variable input are added (with all other inputs held
constant), a point is reached where additional units will add
successively decreasing increments to total outputthat is, marginal
product will begin to decline.
Notice that, after 40 workers are employed,
marginal product begins to decline
20
500
400
300
200
100
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
20,000-square-foot plant
10,000-square-foot plant
Number of Workers
The Total Product of Labor
Total
Product
21
The marginal product of labor when plant size is
10,000 square feet
5.0
4.0
3.0
2.0
1.0
0
10 20 30 40 50 60 70 80 90 100 110 130 140
Marginal
Product
Number of Workers
1.0
2.0
120
22
Another Example
Example:
A farmer grows wheat.
He has 5 acres of land.
He can hire as many workers as he wants.
23
Total, Average and Marginal Product
Labour
Input (L)
TP
L
AP
L
MP
L
0 0 - -
1 20 20 20
2 50 25 30
3 90 30 40
4 120 30 30
5 140 28 20
6 150 25 10
7 155 22 5
8 150 19 -5
24
Total, Average and Marginal Product
Observations:
When output is zero, TP is also zero; so that TP curve starts from the
origin and AP and MP curve have the same vertical intercept (for all
practical purposes!!!).
TP increases at an increasing rate over the range of L=0-3, and then
increases at a decreasing rate.
Initially the input proportions are inefficient there is too much of fixed
factor, capital. As the labour input is increased from 0 to 3, output rises
more than in proportion to the increase in labour input and as a result
MP of labour rises.
As labour input is increased beyond 3 units, diminishing marginal
returns set in and marginal product declines; the additional units of
labour still result in an increase in output, but each increment to output
is smaller.
When labour has increased to 7, TP reaches a maximum, and then, at 8
units, the amount of labour has become excessive and slows the
production process, with the result that total product actually declines.
25
Total, Average and Marginal Product
MP at first increases, reaches a maximum at L=3,
which corresponds to an inflection point on the total
production curve. At the inflection point, total
product function changes from increasing at an
increasing rate to increasing at a decreasing rate.
MP intersects AP at the maximum point on the AP
curve. This occurs at L=4. Whenever MP is above the
AP product, the AP is rising. When the MP is below
the AP, the average is falling. The intersection occurs
at the maximum point of AP.
MP becomes negative at L=8, when TP starts
diminishing.
Illustration of the Law of Diminishing Returns
to a Factor
26
Illustration of the Law of Diminishing Returns
to a Factor
27
Relationship Between AP
L
and MP
L
AP
L
= Q/L, by definition. That is, Q = AP
L
x L.
Differentiating both sides w.r.t. to L, we get
dQ/dL = (dAP
L
/dL)L + AP
L
The expression on the LHS is nothing but MP
L
. Hence,
rearranging terms, we get
(dAP
L
/dL) = (MP
L
- AP
L
)/L
This means that
(a) MP
L
> AP
L
implies (dAP
L
/dL) >0
(b) MP
L
= AP
L
implies (dAP
L
/dL) =0
(c) MP
L
< AP
L
implies (dAP
L
/dL) <0
28
Relationship Between AP
L
and MP
L
When the marginal product of labour is greater than the
average product of labour, the average product is rising; when
it is equal, the average product remains unchanged and when
it is lower, the average product is falling.
The intuition behind this relationship can be quickly captured if
we think of exam grades. Suppose that till now, Ms Ahana is
getting an average grade of B on the papers she has taken.
She sits for another exam, say in English. If she gets a grade
above B in English, this marginal grade pulls up her average
grade and makes it higher than B. If she gets B in English, her
average grade remains unchanged at B. If she gets less than B
in English, then her average grade falls below B.
29
Application From Cobb-Douglas Function - 1
Consider the Cobb-Douglas production function in the short-run:
Q = AK
a
L
b
where A >0, 0<a, b<1.
Questions:
Derive the average and marginal product of labour? What is the
relationship between the two?
Draw the average, marginal and total product of labour curve
for this (short-run) production function.
30
31
Optimal Employment of a Factor of Production
To maximize profit, a firm should hire labour as long as the
additional revenue associated with hiring of another unit of
labour exceeds the cost of employing that unit.
Suppose that the marginal product of an additional worker is 2
units of output and each unit of output is worth Rs.2000. Thus
the additional revenue to the firm will be Rs. 4000 if the worker
is hired. If the additional cost of a worker (i.e., wage rate) is
Rs.3000, the worker will be hired because Rs.1000, the
difference between additional revenue and additional cost, will
be added to profit.
If however, the wage rate is Rs.4500, the worker should not be
hired because profit would be reduced by Rs.500.
32
Optimal Employment of a Factor of Production
The principle is that additional units of the variable input should
be hired until the marginal revenue product (MRP) of the last
unit employed is equal to the cost of the input.
The MRP is defined as marginal revenue times marginal product
and represents the value of the extra unit of labour, i.e., MRP =
MR x MP.
Thus labour is hired until MRP
L
equals wage rate (w), i.e., MRP
L
= w, or, MR x MP
L
= w. For a price-taker firm, MR =P (given),
so that we get P x MP
L
= w. The term P x MP
L
is the value of MP
of labour and is denoted as VMP
L
. Thus for a price taker firm,
labour will be hired until VMP
L
=w
33
The Example of Firm Dealing with Auto Parts
Example:
The firm has estimated that the cost of hiring an additional
worker is equal to $160 per day, that is, MC
L
= $160.
Assume the firm can sell all the parts it wants at a price of
$40. Hence, MR = $40
Thus the MRP
L
= (MR)(MP
L
) = ($40)(MP
L
)
34
The Example of Firm Dealing with Auto Parts
Number of Total Margi nal Margi nal Margi nal
Workers Product Product Revenue Product Cost
10 93 160
20 135 4.2 168 160
30 180 4.5 180 160
40 230 5 200 160
50 263 3.3 132 160
60 293 3 120 160
70 321 2.8 112 160
80 346 2.5 100 160
90 368 2.2 88 160
100 388 2.0 80 160
110 400 1.2 48 160
120 403 0.3 12 160
130 391 -1.2 -48 160
140 380 -1.1 -44 160
35
Quick Activity - 1
Let the production function be given by:
Q = 120L L
2
The cost function is given by
C = 58 + 30L
The firm can sell an unlimited amount of output at a price equal to
$3.75 per unit
1. How many workers should the firm hire?
2. How many units should the firm produce?
Application - 1
How to determine the optimal horsepower for an Oil Pipeline?
Crude oil is carried by pipelines from oil fields and storage
areas over hundreds of miles to urban and industrial centres.
The output of such a pipeline is the amount of oil carried per
day, and the two principle inputs are the diameter of the
pipeline and the horsepower applied to the oil carried.
Leslie Cookenboo of the Exxon Corporation estimated the
production function with a 10-inch diameter to be as follows:
Q = 286H
0.37
where Q is the amount of crude oil carried per day, and H is
the horsepower.
36
Application - 1
Questions:
(a) Derive a formula for the marginal product of horsepower
(b) Do increases in horsepower results in diminishing marginal returns?
(c) Derive a formula for average product of horsepower.
(d) If the marginal revenue from an extra unit of crude oil carried per
day is $2, what is the marginal revenue product of horsepower?
(e) If an oil-pipeline firm can add all the horsepower it wants at a price
of $30 per unit of horsepower, what is the marginal expenditure on
horsepower?
(f) Under the circumstances described above, what is the amount of
horsepower that an oil-pipeline firm should use? [Answer: 22.22 units]
37
The Production Function With Two Variable Inputs
Up to this point, we have been concerned with the case where
there is only one variable input. Now take up the general case
where there are two variable inputs.
For simplicity, we assume that there are only two inputs, in
which case the situation is the long run, since all the inputs are
variable.
If X
1
is the amount of the first input and X
2
is the amount of
the second input, the production function is
Q = f (X
1
, X
2
) . (i)
where Q is the firms output rate. The marginal product of the
first input is Q/ X
1
; the marginal product of the second input
is Q/ X
2
38
The Production Function With Two Variable Inputs
The production function, represented by equation (i), does not
include many of the different ways in which a given output can
be produced because it includes only efficient combinations of
inputs.
For example, if 2 units of labour and 3 units of capital can
produce 1 unit of output, this combination of inputs and output
will not be included in the production function if it is also
possible to produce 1 unit of output with 2 units of labour and 2
units of capital. The former input-output combination is clearly
inefficient, since it is possible to obtain the result with the same
amount of labour and less capital.
39
Isoquants
An isoquant is a curve showing all possible (efficient)
combinations of inputs that are capable of producing a certain
quantity of output.
40
Several isoquants, each
pertaining to a different output
rate is shown in the figure. The
two axes measure the quantities
of inputs that are used.
Let labour and capital be the
relevant inputs in this case. The
curves show the various
combinations of inputs that can
produce 5000, 10000 and 30000
units of output.
Isoquants
41
For example, consider the
isoquant pertaining to 100 units
of output per period of time.
According to this isoquant, it is
possible to attain this output rate
if 10 workers are hired and 2
machines rented per period of
time.
Alternatively, this output rate can
be attained if 6 workers are hired
and 3 machines are rented per
period of time.
The Marginal Rate of Technical Substitution
There are ordinarily a number of efficient ways that a particular
output can be produced. The marginal rate of technical
substitution (MRTS) shows the rate at which one input can be
substituted for another input so as to keep the level of output
constant.
If X
1
is the amount of the first input and X
2
is the amount of
the second input, so that the production function is
Q = f (X
1
, X
2
)
the MRTS is equal to dX
2
/dX
1
, given that Q is held constant
(and X
2
is on the vertical axis of the isoquant graph and X
1
is
on the horizontal axis).
Geometrically, MRTS is the negative of the slope of the
isoquant. Its value is also equal to MP
1
/MP
2
. [Prove this
relationship]
Input Factor Substitution
Input Factor Substitution
Figure (a) shows isoquants for an electric power plant with boilers
equipped to burn either oil or gas. Power can be produced by burning
gas only, oil only, or varying amounts of each. In this instance, gas
and oil are perfect substitutes and the electricity isoquants are
straight lines.
At the other extreme of substitutability lie production systems in
which inputs are perfect complements: exact amount of each input
are required to produce a given quantity of output. Figure (b)
illustrates isoquants for bicycles in which exactly two wheels and one
frame are required to produce a bicycle. Wheels cannot be
substituted for frames, nor vice versa. Pants and coats for mens
suits, engines and bodies for trucks, and chemicals in specific
compounds for prescription drugs are further examples of
complimentary inputs. Production isoquants for perfect complements
take the shape of right angles or L- shaped.
Input Factor Substitution
Figure (c) shows a production process in which inputs can be
substituted for each other within limits.
A dress can be made with a relatively small amount of labour (L
1
) and
a large amount of cloth (C
1
). The same dress can also be made with
less cloth (C
2
) if more labour (L
2
) is used because the dressmaker can
cut the material more carefully and reduce waste. Finally, the dress
can be made with still less cloth(C
3
), but workers must be so
extremely painstaking that the labour input requirement rises to L
3
.
Although a relatively small addition of labour, from L
1
to L
2
, reduces
the input of cloth from C
1
to C
2
, a very large increase in labour, from
L
2
to L
3
, is required to obtain a similar reduction in cloth from C
2
to C
3
.
The substitutability of labour for cloth diminishes from L
1
to L
2
to L
3
.
Most labour-capital substitutions in production systems exhibits this
diminishing substitutability.
45
Application From Cobb-Douglas Function - 2
Consider the Cobb-Douglas production function:
Q = AK
a
L
b
where A >0, 0<a, b<1.
Questions:
Write the equation of the isoquant for this production function.
Show that the isoquants are negatively-sloped and convex to
the origin.
Define MRTS. What is the value of MRTS for this type of
production function?
46
Application - 2
A production function for agricultural crops
Crops can be produced using different methods. Food grown on large
farms in the United States is usually produced with a capital intensive
technology, which involves substantial investments in capital, such as
equipment, and relatively little input of labour. However, food can also
be produced using very little capital (a hoe) and a lot of labour.
One way to describe the agricultural production process is to show one
isoquant that describes the combination of inputs that generates a
given level of output.
Consider an isoquant (associated with a production function for wheat)
corresponding to an output of 13, 800 bushels of wheat per year. The
manager of the firm can use this isoquant to decide whether it is
profitable to hire more labour or use more machinery.
47
Application - 2
Suppose the farm is operating at a point on the isoquant, with labour
input (L) of 500 hours and a capital input (K) of 100 machine-hours.
The manager decides to experiment by using fewer hours of machine
time. To produce the same crop per year, he finds that he needs to
replace the machine time of 10 machine-hours by adding 250 hours of
labour. Thus, the manager finds that the marginal rate of technical
substitution is equal to 0.04 (WHY??)
The MRTS tells the manager the nature of the trade-off between
adding labour and reducing the use of farm machinery. Because the
MRTS is substantially less than 1 in value, the manager knows that
when the wage of labour is equal to the cost of running the machine,
he ought to use more capital. (At his current level of production, he
needs 250 units of labour to substitute for 10 units of capital)
In fact, he knows that unless labour is substantially less expensive
than the use of machine, his production process ought to become
more capital-intensive.
48
Application - 2
The decision about how many labourers to hire and machines
to use cannot be fully resolved until we know the costs of
inputs. However, this example illustrates how knowledge about
production isoquants and the marginal rate of technical
substitution can help a manager.
It also suggests why most firms in United States and Canada,
where labour is relatively expensive, operate in the range of
production in which MRTS is relatively high (with a high capital
to labour ratio), while farms in developing countries like India,
in which labour is cheap, operate with a lower MRTS ( and a
lower capital to labour ratio).
49
Rational Limits of Input Substitutions
From the law of diminishing returns to a factor of production, we
know that as more and more units of inputs are used, its marginal
product keeps on falling.
It is irrational for a firm to combine resources in such a way that the
marginal product of any input is negative because this implies that
output could be increased by using less of that resource. [This is
technically correct only if the resource has a positive cost. A firm
might employ additional workers even though the marginal product of
labour was negative if it received an employment subsidy that more
than offset the cost of output decision.]
For a production isoquant to be positively sloped, one of the input
factors must have a negative marginal product. Input combinations
lying along a positively sloped portion of a production isoquant are
irrational and should be avoided by the firm.
50
The Optimal Combination of Inputs
If a firm wants to maximize profit, it will try to minimize the
cost of producing a given output or maximize the output
derived from a given level of cost.
Assumptions:
There are two inputs labour and capital, that are variable in the
relevant time period.
The firm takes input prices as given.
Question:
What combination of capital and labour should the firm choose
if it wants to maximize the quantity of output derived from the
given level of cost?
The Optimal Combination of Inputs
To begin to answer this question, lets determine the various
combinations of inputs that the firm can obtain for a given
expenditure.
For example, if capital and labour are the inputs and the price of
labour is w per unit and the price of capital is r per unit, the input
combinations that can be obtained for a total outlay C are such that
wL +rK = C.
where L is the amount of labour input and K is the amount of capital
input.
Thus, the various combinations of capital and labour that can be
purchased, given w, r and C, can be represented by a straight line
capital is plotted on the vertical axis and labour on the horizontal axis.
The line, which has an intercept on the vertical axis equal to C/r, an
intercept on the horizontal axis equal to C/w and a slope w/r, is
called an isocost line.
The Optimal Combination of Inputs
53
5 10 15
0
Units of labor per month
5
10
U
n
i
t
s

o
f

c
a
p
i
t
a
l

p
e
r

m
o
n
t
h
Slope = -w/r = -Rs.1,500/Rs.2,500 = -0.6
Each isocost line
- Combinations of labor and capital
that can be purchased for a given
amount of total cost
-Slope is negative wage divided
by the rental cost of capital
Higher costs: isocost lines farther from origin
The Optimal Combination of Inputs
If we super impose the relevant isocost curve on the firms isoquant
map, we can determine graphically which combinations of inputs will
maximize the output.
The firm should pick the point on the isocost curve that is on the
highest isoquant, for example, R in the figure (next slide). This is a
point where the isocost curve is tangent to the isoquant.
Thus, since the slope of the isocost curve is the negative of w/r and
the slope of the isoquant is the negative of MP
L
/MP
K
, it follows that the
optimal combination of inputs is one where MP
L
/MP
K
= w/r. Or put
differently, the firm should choose an input combination where
MP
L
/w= MP
K
/r.
Thus the firm will maximize output by distributing its expenditures
among various inputs in such a way that the marginal product of a
rupees worth of any one input is equal to the marginal product of a
rupees worth of any other input used.
The Optimal Combination of Inputs
55
Q
1
Q
2
Q
3
f
0
Units of labor per month
U
n
i
t
s

o
f

c
a
p
i
t
a
l

p
e
r

m
o
n
t
h
Isocost curve
a
R
R: isoquant Q
2
is tangent
to the isocost line
How should a firm choose the optimal combination of inputs
if its objective is to minimize cost subject to a given level of
output?
Application From Cobb-Douglas Function - 3
Consider the Cobb-Douglas production function:
Q = AK
a
L
b
where A >0, 0<a, b<1.
Questions:
Derive the optimal level of capital and labour that a firm facing
this form of production function shall utilize to maximize output
subject to a given cost outlay.
56
Quick Activity - 2
Consider the Beiswanger Company, a small firm engaged in
engineering analysis. Beiswangers president has determined
that the firms output per month (Q) is related in the following
way to the number of engineers used (E) and the number of
technicians used (T):
Q = 20E E
2
+ 12T 0.5T
2
The monthly wage of an engineer is $4,000 and the monthly
wage of a technician is $2,000. If Beiswanger allots $28,000
per month for the total combined wage of engineers and
technicians, how many engineers and technicians should it
hire?
[Answer: E = 4; T = 6]
Quick Activity - 3
Consider the Miller Company, for which the relationship
between output per hour (Q) and the number of workers (L)
and the number of machines (K) used per hour is:
Q = 10(LK)
0.5
The wage of a worker is $8 per hour and the price of machine
is $2 per hour. If the Miller produces 80 units of output per
hour, how many workers and machines should it use?
[Answer: L = 4; K = 16]
Returns to Scale
In the short run, one input is fixed and we examined what
happened to the output when more of the other input is
applied. We discussed the law of diminishing marginal returns
in that context.
In the long run, all inputs are variable.
Question:
If all the inputs are increased in the same proportion, does
output increase in the same proportion?
The relationship between output change and proportionate
changes in both inputs is referred to as returns to scale.
59
Returns to Scale
60
Returns to Scale
Assume that the usage of all inputs increases by 25 percent. If output
increases by exactly 25%, the production function exhibits constant
returns to scale. If, however, output increases by more than 25%,
the production function exhibits increasing returns to scale.
Alternatively, if output increases less than 25%, the production
function is characterized by decreasing returns to scale.
Many production systems exhibit increasing, then constant, then
decreasing returns to scale. The region of increasing returns is
attributable to specialization.
61
Sources of Increasing Returns to Scale
62
1. Specialization of plant and equipment
As output increases, specialized labour can be used and efficient large-scale
machinery can be used in the production process. As more workers and machines
are used, it is possible to subdivide tasks and allow various inputs to specialize.
Example: Large scale production in furniture manufacturing allows for application
of specialized equipment in metal fabrication, painting, upholstery, and materials
handling.
2. Economies of massed reserves
Example: A factory with one stamping machine needs to have spare 100 parts in
inventory to be prepared for breakdowndoes a factory with 20 machines need
to have 2,000 spare parts on hand?
Sources of Decreasing Returns to Scale
Bigger is not always better; managers can experience decreasing
returns to scale.
The most cited reason is the difficulty of coordinating a large
enterprise. Beyond some scale of operation further gains from
specialization are limited and coordination problems emerge. It can be
difficult even in a small firm to obtain the information required to make
important decisions; in a large firm, difficulties tend to be greater. it
can be difficult even in a small firm to be certain that managements
wishes are being carried out; in a large firm, these difficulties too tend
to be greater.
When coordination expenses more than offset additional benefits of
specialization, decreasing returns to scale set in.
Whether there are IRS, CRS or DRS in a particular situation is an
empirical question that must be settled case by case. There is no
simple, all-encompassing answer. In some industries, the available
evidence may indicate that IRS are present over a certain range of
output. In other industries, DRS or CRS may be present.
63
Output Elasticity and Returns to Scale
To measure whether there are increasing, decreasing, or
constant returns to scale, the output elasticity can be
computed. The output elasticity is defined as the percentage
change in output from a onepercent change in all inputs.
If the output elasticity exceeds 1, there are increasing returns
to scale; if it equals one, there are constant returns to scale; if
it is less than one, there are decreasing returns to scale.
Thus, returns to scale can be analyzed by examining the
relationship between the rate of increase in inputs and the
quantity of output produced.
64
Quick Activity - 4
Consider the Lone Star Company, a maker of aircraft parts,
which has the following production function:
Q = 0.8L
0.3
K
0.8
where Q is the number of parts produced per year (measured
in millions of parts), L is the number of workers hired, and K is
the amount of capital used.
Compute the output elasticity and comment on the returns to
scale for the above production function.
65
Returns to Scale and Homogenous Production
Functions
For certain production functions, called homogenous production
functions, when each factor is multiplied by a constant t, the
constant can be completely factored out of the production
function expression. A homogenous production function of
degree k if
f (tK, tL) = t
k
f (K, L)
where k is a constant and t is any positive real number.
If both inputs are increased by the same factor t, output is
increased by the factor t
k
. Returns to scale are increasing if
k>1, constant if k = 1, and decreasing if 0<k<1.
66
Application From Cobb-Douglas Function - 4
Consider the Cobb-Douglas production function:
Q = AK
a
L
b
where A >0, 0<a,b<1.
Questions:
Is the production function homogenous? If yes, compute the
degree of homogeneity and comment on the returns to scale
for this production function.
Compute the elasticity of output with respect to labour and
elasticity of output with respect to labour. Show that the sum
of the elasticities is equal to the degree of homogeneity of the
production function.
67
Application - 3
68
Application - 3
The table shows the estimated returns to scale in 18
manufacturing industries in United States in 1957 using the
Cobb-Douglas Production Function.
A value of 1 refers to CRS, a value greater than 1 refers to IRS,
and a value less than 1 refers to DRS.
The table shows that for doubling (i.e., with a 100 percent
increase) of all inputs, output would rise by 111 percent in the
furniture industry, by 109 percent in chemicals, but only by 95
percent in petroleum. Although only the textile industry seems
to face CRS exactly, most other industries are very close to it.
69
Application - 3
70
Application - 3
The table reports the estimated output elasticities of capital
(a), production workers (b), and nonproduction workers (c) for
the same 18 manufacturing industries.
The fourth column shows that the returns to scale (which is
also the degree of homogeneity of this type of production
function) is the sum of the input elasticities of output.
The values of a, b and c reported in the table were estimated
by regression analysis with cross-sectional data for each of 18
industries for the year 1957, using a Cobb-Douglas production
function extended to three inputs and transformed into natural
logarithms. All the estimated coefficients, with the exception of
c for the rubber and plastics industry, were positive, as
expected.
71
Application - 4
Should two Punjab wheat firms merge?
Agricultural economists have estimated the production function for
many types of farms, here and abroad. Suppose that a study came up
with the following result for a particular type of Punjab wheat farm:
Q = ZA
0.1
L
0.1
E
0.1
S
0.7
R
0.1
,
where Q is the output per period, A is the amount of land used, L is
the amount of labour used, E is the amount of equipment used, S is
the amount of fertilizers and chemicals used, R is the amount of other
resources used, and Z is constant.
The owner of a Punjab wheat farm of this type is concerned that his
farm may be too small to compete effectively with larger wheat farms.
His farm is of below-average size, and he is troubled by the possibility
that larger farms may be more efficient than his. He is considering the
merger of his farm with a neighboring farm that is essentially the
same (in size and other characteristics) as his own, and he hires you
to advise him on this score. Whats the answer?
72