ECO2003 – Intermediate

Microeconomics

Production in the short run

Geraint Van der Rede

Geraint.vanderrede@uct.ac.za

1
An Introduction to Production
We generally think about production as “raw materials
being turned into final goods”

We see production as a tangible thing, but it is so much

more than this!

Production in economics is defined as

“any activity that creates present or future utility” or
“the process of transforming inputs into outputs”

So, production could be anything from actually making

goods, to telling a joke, to teaching.
2
The Production Function
It makes more sense to think of the production
function in the context of the second definition of
production.

The production function is the relationship by

which inputs are combined to produce outputs.

Inputs = generally factors of production

e.g. land, labour, capital, knowledge, technology

Outputs = the objects that provide us with utility

3
The Production Function
Think of this like a recipe … but instead of eggs and flour and butter, we use capital
and labour

Obviously, capital and labour by themselves are not enough to make output – you
need raw materials as well.

You can’t just make bread out of bakers and ovens – you need flour, yeast, salt, etc.

But for simplicity of analysis, we will not model raw materials into our analysis
separately, but there is no reason why you can’t include these intermediate
products in the model.
4
The Production Function
2(2)(2)=8 2(3)(3)=18

As an example, consider the production

function given by: Q=2KL K

1 2 3 4

How much output do you get when you 1 2 4 6 8

combine 4 machine-hours with 5 person-
hours?
2 4 8 12 16
L

3 6 12 18 24
How many machine-hours would you
need to create 60 units of output with 6
person-hours? 4 8 16 24 32

5
 Fixed and Variable Inputs
Different factors take different lengths of time to change.

For example, does it take the same amount of time to buy and
set up a new factory as it does to hire a new worker?

This introduces the idea of the short run and the long run

VS

6
Fixed and Variable Inputs

Short run: The longest period of time during which at least one of the inputs cannot be varied

Long run: The shortest period of time required to alter the amounts of all inputs in the process

If the input can be changed in the short run, then it is a variable input.

If the input cannot be changed in the short run (or if it can, it costs an enormous amount of
money), then it is considered a fixed input.

In the long run, all inputs become variable.

In the production of good grades for this course, what would the inputs be? Menti.com -
530646

7
 Production in the Short Run
In the short run we assume that capital is a fixed input.

In other words, we say that K is fixed at some value of K0.

Consider the example from earlier: Q = 2KL

In the short run, we might only have 1 machine, so K=1 and the
production function simplifies to Q = 2(1)L = 2L
Similarly, if another company has 3 machines, then their
production function becomes Q = 2(3)L = 6L

8
Production in the Short Run
Let’s do an example:
Draw the short run production function that corresponds to where capital is fixed at K 0 = 4.

𝐹 ( 4 , 𝐿 )= √ 4 𝐿=2 √ 𝐿

Output

Labour
9
 Production in the Short Run

As you saw in the example, we aren’t limited to straight line production functions – in fact, one of
the most common production function shapes is non-linear.

It still represents 3 general properties of short run production functions:

1. The production function should pass through the origin
2. Initially, adding extra units of variable input should increase output at an increasing rate
3. After some point, we should see extra units of the variable input increasing output by smaller
and smaller increments

10
 Production in the Short Run

Note how each extra unit of labour gives us

less and less extra output after L=4, and
output decreases after L=8.

This is the law of diminishing marginal

returns:

If other inputs are fixed, the

increase in output from an
increase in the variable inputs
must eventually decline

11
Production in the Short Run
Some things to think about:
Why might you start off by seeing increasing levels of the variable input increasing output at an
increasing rate?

Maybe there is something to do with benefits from division of tasks – when you’re making
sandwiches, maybe having two people instead of one person means you get specialisation.

Does the concept of diminishing marginal returns make sense in reality? Do we see it in practice?

It does seem to make sense. Think about our group of sandwich makers. Does adding a third person
really make such a difference to our total production? What about a fourth person?

At what point do we see everyone getting distracted and sitting around drinking tea? 12
 Technology in the Production
Function
In 1798, Thomas Malthus said that the law of diminishing marginal returns basically meant that because land is
fixed, having more farmworkers would ultimately add less and less to food production and we would all starve to
death. He predicted that average food consumption per capita would be driven down to starvation level.

But he didn’t account for technological advances, which might not spell doom and gloom for the human race.

BUT we still saw diminishing returns – it’s just that the growth in technology outstripped the effect of diminishing
returns.

Unfortunately, the logic of Malthus’ prediction still holds: if the population continues to grow, the limits on the
amount of land we have will spell food shortages.

Technology is embedded in the construction of the production function.

How does technological growth impact the production function? 13

Technology in the Production
Function
Technological advancement will shift our
production function upwards.

Why?

Basically, at the same level of variable

input, we can produce more because of
technological advances.

Think about how much 10 workers could

harvest by hand vs. with combine
harvesters.

14
Total, Marginal and Average
Products
We are now moving on to define some important concepts in the theory of production:

Total product curve: A curve showing the total amount of output as a function of the amount
of variable input (think about this as the graph of the production function)

Average product: Total product divided by the number of units of the variable input used in the
production process.

Marginal product: Change in total product due to a one-unit change in the variable input.

15
 Marginal Product
Geometrically, marginal product is the slope of the
total product curve.

Algebraically, we can calculate the slope in one of two

ways:

Discrete MP:
Continuous MP:

Marginal Product reaches its maximum at the point of

inflection of the TP curve.

16
Marginal Product

Q=TP Continuous MP – the slope of

the tangent line at any given
point

Discrete MP – the slope of the

line joining two points on the TP
curve

L
17
Average Product

Graphically, Average Product (AP) is Q=TP

the slope of the ray from the origin to
any point on the TP curve.

Algebraically:

We can see that AP begins by

increasing, and then starts decreasing

L
18
Average Product and Marginal Product

The basic idea is this:

If you have MP>AP, then your AP curve is increasing.

If you have MP<AP, then your AP curve is decreasing.

MP and AP intersect at the maximum of the AP curve

Think about what happens when there is a study group of

students whose average is 65%.
Another student joins the group; what happens to the overall
average if their mark is 85%?
And if it is 50%
19
Average Product and Marginal Product

Note that for values of L < 6, the slope of the AP

line is smaller than the slope of the TP curve 
MP>AP

For L > 6, the slope of the AP line is greater than

the slope of the TP curve  AP>MP

At small values of L, slope of ray and slope of TP are

indistinguishable, and flat  MP=AP=0

At L = 6, the slope of AP must be the greatest it can

be. It also happens to be tangent to TP at that point
 AP=MP at its maximum
20
Practical Considerations of MP
and AP

Suppose you have a fleet of fishing boats and you have 2 boats on each end of a lake. You get
100kg of fish from each boat sent to the East end of the lake, and 120kg of fish from each boat
sent to the West side of the lake.
Should you change your allocation of boats?

Not necessarily – it depends on what gains you get from sending the third boat to the West
side.

The question hinges on marginal product, but most people seem to answer this question by
thinking about average product.

21
Practical Considerations of MP
and AP
Number of East End West End
Boats

TP AP MP TP AP MP

0 0 0 0 0

1 100 100 100 130 130 130

2 200 100 100 240 120 110

3 300 100 100 330 110 90

4 400 100 100 400 100 70

Here, it makes no sense to send the third boat to the West End as the marginal gain from doing so is not
worth it. 22
Practical Considerations of MP
and AP
Number of East End West End
Boats

TP AP MP TP AP MP

0 0 0 0 0

1 100 100 100 120 120 120

2 200 100 100 240 120 120

3 300 100 100 360 120 120

4 400 100 100 480 120 120

Here, it does make sense to send the third (and fourth) boat to the West End as the marginal gain from
doing so outstrips what they could earn you in the East End
23
Practical Considerations of MP
and AP
Basically, you have to consider how to best allocate your resources – how
do you do it?

Economically, the most efficient way of allocating your resource is to

allocate the next unit of input to the process in which its marginal
product is highest.

If you have perfectly divisible inputs, and for processes where the MP of
one isn’t always higher than the MP of the other, you should allocate this
input between production processes until the marginal product is equal
across both processes.

24
 Example
Given the following function for baking loaves of
bread To maximise output, we need to hire 7 labourers

2. 215
Where L= hours spent baking loaves The maximum output produced when the firm hires 7
labourers

1. Where is Q maximised? 3.
This can be used to find the marginal product of each
2. What is this maximum value of Q? respective worker. To find the
marginal product of the 5 worker,
th
simply substitute 5 in for L.
3. Write an expression for MPL.
4.
4. Write an expression for APL.
We could use this function to find the average product of
5. What is the MPL and APL when L=2? labour for any number of workers

5.

25
 ECO2003 – Intermediate
Microeconomics

Production in the long run

Geraint Van der Rede

Geraint.vanderrede@uct.ac.za

26
Production in the Long Run

Recall: the long run is the shortest period of time it takes for all
inputs to become variable.

So … Now, everything can change.

We no longer assume K to be fixed, so we can vary the level of K.

So, how do you draw the production function Q = 2KL ?

27
Production in the Long Run
How do we graph production in the long run?

We draw what are called isoquants

Isoquant – “iso” means same; “Quant” for quantity.

So, we choose a value for Q, say Q0, and plot our graph in L-K space.

Does this remind you of something? Maybe how we represent 3D utility functions with
indifference curves.

28
 Production in the Long Run
So, for Q = 2KL:
Assume Q=Q0=16

Assume Q=Q1=32

Assume Q=2C (Arbitrary constant)

This is a rectangular hyperbola defined for
positive K, L and C.

So, you choose an output level, and then solve

for K as a function of L.

Remember that numerical labels on isoquants 29

have meaning here.
 Marginal Rate of Technical Substitution

In consumer theory, we asked ourselves “how much of

x would we substitute for y to remain at the same
level of utility?”

In the world of the firm, why don’t we ask an

analogous question:
“How much labour could we substitute for capital in
order to continue producing the level of output we
were before?”

The answer to this question is the Marginal Rate of

Technical Substitution (MRTS)
30
 Marginal Rate of Technical Substitution

If we were to give up of capital, then to remain on the

same isoquant, we need to hire units of labour.

So, MRTS is the absolute value (positive value) of the

slope of the isoquant.

We can define this at the point A as

If MRTS=5, this means I am substituting 5 units of capital

for 1 unit of labour.
31
 Marginal Rate of Technical Substitution
So, we know that if we give up some of one input
(K), we have to increase the amount of the other
input (L) to stay on the same isoquant.
Also, we can work out that the loss from the
decreased K must offset the gain from the
increased L, in production terms.
How much do we lose from decreasing K?
How much do we gain from increasing L?
So, we know these terms are equal:

Rearranging, we get:

32
Thus, we can write
 Marginal Rate of Technical Substitution

The MRTS effectively tells us how a firm is willing to

substitute between inputs (for the purposes of this
course between K and L)

Note that if you have a lot of a particular input, you

would be willing to give up a lot of it for a little bit of
the other one.

Note how this supports the diminishing marginal

returns assumption from earlier.

33
Returns to Scale

Returns to scale are a long run phenomenon

This is because they require a change in all of the inputs (this means all the inputs
need to be variable  long run)
Returns to scale have to do with how production responds to a proportional
increase in all inputs.

Increasing Returns to Scale – a proportional increase in all inputs leads to a more

than proportional increase in output
Constant Returns to Scale - proportional increase in all inputs leads to an equally
proportional increase in output
Decreasing Returns to Scale – a proportional increase in all inputs leads to a less than
proportional increase in output 34
Returns to Scale
(Mathematically)

Assume that we increase all of our inputs by a

factor of c.
Then, the three types of returns are characterised
by:

Increasing Returns to Scale

Constant Returns to Scale

Decreasing Returns to Scale 35

Diminishing vs. Decreasing
Returns

It is very important to remember the distinction between diminishing returns and decreasing
returns:

Diminishing marginal returns: a short run concept which talks about how output changes due
to an increase in a specific factor.

Decreasing returns to scale: a long run concept which talks about how an increase in every
factor affects total production.

Firms with increasing, decreasing or constant returns to scale will probably still exhibit
diminishing returns.
36
 Returns to Scale vs. Diminishing
Returns
Your returns to scale depend crucially on the
production function you are dealing with.

It is possible for one production function to exhibit

increasing returns to scale at some point, constant
returns to scale at others, and even decreasing
returns to scale.

BUT NOTE: The function shows diminishing returns at

all levels of production – this can be seen from the
curved isoquants

In fact, diminishing returns is a pretty universal

characteristic of production functions 37
Types of Production Functions

There are 3 basic types of production functions you will come across:
1. Cobb-Douglas function (this is the most commonly used function)

Also, If as well, then diminishing marginal returns to that factor.

If , then increasing marginal returns to that factor.
2. Leontief (or fixed proportions) production function

This is the function used when the inputs are perfect complements
3. Straight line production function

This is the function used when the inputs are perfect substitutes 38
The Leontief (Fixed Proportions) Production Function

Example:
MRTS does not exist
Assume K=1 and L=1.

MRTS=/undefined
Thus, Q = min{2(1), 3(1)} = min{2, 3} = 2
Does it help us at all if K increases and L doesn’t?

We need K and L to increase in the same

proportion to increase output.

These inputs are thus perfect complements.

MRTS=0
The line of efficient combinations of inputs lies
along the line where 2K=3L.
Solving out gives that
39
The Straight Line Production
Function
Example
Shell
K and L are related in a linear way, and when you solve for
an isoquant, you will get the equation of a straight line:

This is a relationship between perfect substitutes.

The MRTS is constant at all levels of inputs – exactly

because of the straight line relationship.

Examples: Whether you get petrol from Shell or Caltex, Caltex

you’ll still be able to drive the same distance
40
The Cobb-Douglas Production Function

Let’s look at this function in terms of returns to 1 1

scale. 2
𝑄= 𝐾 𝐿 2

Let’s increase all of our inputs by a factor of c:

So, our different types of returns depend on the

parameters and
If then increasing returns to scale
f then constant returns to scale
f then decreasing returns to scale 41
Cobb-Douglas and MRTS
How do we calculate the MRTS for a Cobb-Douglas production
function?
Well, we make use of the formula

So, now assume we have a Cobb-Douglas function in its most general

form:
To derive the MRTS, we need MPL and MPK:

By dividing these out, we get:

42