Professional Documents
Culture Documents
6
PRODUCTION
Production involves the generation of output by means of combining various goods and services. Such
combinations rarely come together with spontaneity: the combinations are usually organised in a certain
way by some individual or individuals – who are known as entrepreneurs.
The manner in which such organisation occurs is not only of economic importance, but often of social
and political significance too.
Whether a company in a capitalist society like Singapore is attempting to maximise profits, or a collective
project in North Korea which is attempting to maximise agricultural output, both will be probably be
seeking to minimise costs. Thus both will have a deep interest in any technical relationships between
inputs and outputs. Both will be keen to harness new technology – if it is efficient and cost effective, and
it is available to them. Production processes and costs will be as important to managers in both types
of society.
In a capitalist society, production organisations come in various shapes and sizes: some are one-man-
businesses and some are huge multi-national corporations, – with organisations of every size existing
between these two extremes. The main categories of businesses, no matter what the size, are:
• one-man businesses
• partnerships of 2 or more individuals
• companies of limited liability, the shares of which aren't traded [a “Private” company]
• companies of unlimited liability, the shares of which are traded [a “Public” company]
One theory could not possibly cover such a wide variety of business concerns in explicit detail. The theory
of the firm therefore attempts to look for common ground: what are the essential properties and
features of production which hold as true for the manager of a hawker stall in Bukit Merah as they do
for SIM, or as they do for Shell Oil, or even as they do for a collective farm in North Korea.
The theory of demand (which we have just spent a month examining) is subjective. With consumer
choice, various assumptions were made, starting with the basic indifference curve. Supply, however, is
determined by objective technical factors.
The Theory of the Firm is concerned with the precise technical relationships involved with the
transformation of factor inputs [such as land, labour and capital] into outputs of goods and services
[such as computers, a meal at a restaurant, economics lectures, etc.].
As with consumer choice, though, it is useful to start with very elementary analysis - i.e. assuming only
two factor inputs, etc.
If a basic two dimensional model can hold, it is reasonable to assume that a multi-dimensional model will
hold true too.
Production should be regarded as the flow of factor inputs into a production process which results in a
flow of output.
So, for example, the factors of production which produced the Introduction to Economics lecture when
you were introduced to the ideas of production were:
• 18 capital hours: during the 6 hours that I was preparing, I used my computer
during the 3 hours of me presenting the lecture, I used a computer or visualisor
during the 3 hours of me presenting, I used the projector
during the 3 hours of me presenting I used the whiteboards
during the 3 hours of me presenting I used the lights
The lecture also needed raw materials: electricity for all the above equipment for all of those hours, the
board markers I used, the coffee I drank to keep me awake while I prepared the lecture, etc.
Even with all of this detail, the above is a vast oversimplification. The lecture needed land as well (space)
– my office (where I did my preparation), Hien’s office, the room in which the photocopier is situated, and
most important 3 hours’ usage of Room 1.
It also needed what is known as human capital – my expertise, accumulated through the years. If I
prepared for 6 hours for an Economics lecture, I could produce the quality of output which you are familiar
with, but if I was to spend 6 hours preparing a Sociology lecture, can you imagine how lousy that lecture
would be?
It also needed the time and ideas and business acumen – many years ago – of the entrepreneurs who
established BUV, and later, of those who organised that BUV run the University of London programme.
While they worked, they too needed space (land) and machines like telephones, etc. (capital)
It also needed the time and expertise of people like Dean Chris to plan everything, months in advance,
making sure that Room 1 was actually available on Friday mornings, and that there were not too many
students for the room. While he worked, he too needed an office (land) and machinery (capital).
It also needed the time and expertise of BUV’s marketing team, last year. How did you hear of BUV? If
none of you had heard of BUV or the University of London programme, none of you would be studying
here now, and the lecture would not have taken place. All those in the Marketing department needed
office space (land) and machinery (capital).
The same level of output could be achieved using different flows of inputs.
Production
Method 1
K
For example, a television set could be
produced by: 40 • 1 TV
Production Method 1
– an activity which combines say, 30
10 man hours with 40 machine hours,
or by: 20
Production
Method 2
Production Method 2 10 •1 TV
– an activity which combines say,
40 man hours with 10 machine hours,
30 L
10 20 40
A simple assumption to make is that if we were to double the amount of factor inputs using either
process, we would double the output [known as "constant returns to scale"].
Production
Method 1
K
160 • 4 TVs
120 • 3 TVs
80 • 2 TVs
Production
Method 2
40 • 1 TV •
•3 TVs 4 TVs
•
•1 TV 2 TVs
40 80 120 160 L
There is no reason why a firm could not mix production methods, producing say 5 TVs by method 1, and
5 by method 2; - or any other combination.
Production
Method 1
K
The “frontier” linking methods
400 • 10 TVs 1 and 2 is known as an
Isoquant. It shows various
combinations of production
methods – i.e. various
• 10 TVs combinations of factor inputs
200 – with which it is possible for
Production this particular firm to produce
Method 2 10 TV sets.
•10 TVs
200 400 L
PM1 PM2
K
PM3
400 •
Suppose there were 5 •
different production methods - PM4
(PM1, PM2, etc.)
200 •
PM5
•
•
200 400 L
• •
PM4 PM4
200 • 200 •
PM5 PM5
• •
• •
The Isoquant
L L
The isoquant is thus the locus of all of the most technically efficient combinations of factor inputs which
can be used to produce a given level of output.
q
2
q
1
q
0
L
Algebraically, the production function can be specified as:
Q = f(L, K)
As with indifference curves, it is normal to assume the function is monotonic - that is: if there is an
increase in L or an increase in K (or an increase in both), there will be an increase in Q.
The Average Product of a factor input per period is simply the output per period divided by the inputs
per period:
Q Q
APL = APK =
L K
The Marginal Product of a factor input per period is the change in output per period given a unitary
change in one factor input (holding the other input constant), i.e.:
dQ dQ
MPL = MPK =
dL dK
If inputs were perfect substitutes for one another, the production function would look like this:
[Extremely
unlikely!]
This would be a situation where man hours and machine hours were perfect substitutes for one another
– a situation which is hard to imagine.
[A husband might consider a “wife” and a “washing machine” to be perfectly substitutable for one
another, but after marriage he will soon discover that the washing machine is superior. It doesn’t shout
at him or nag him, thus putting him on a far higher level of utility than a wife would.]
Seriously: can YOU think of any situation, even given today’s advanced-technology environment, where a
machine has become a perfect substitute for a human?
With fixed proportion technology, the production function would look like this:
This would be a situation where man hours and machine hours were “perfectly complementary” – i.e.
always needed in the same proportions. This is not uncommon. You always need 1 man and 1 tractor
on a farm, for example. Giving 1 man extra tractors will not increase the agricultural output. In an office,
you need one word processor per secretary. Adding extra secretaries to the office without adding any
extra word processors will not increase the firm’s output.
Isoquants have many similar properties to indifference curves: they are downward sloping, you mustn’t
show them intersecting, etc.
However, objective values may be assigned to isoquants, whereas only subjective values of utility ["utils"]
may be assigned to indifference curves. [i.e. A few weeks ago, we said that if we reached a higher
indifference curve, we were “happier” – but we didn’t try to quantify that happiness. However, we CAN
quantify isoquants.]
Each point on the lowest curve would result in an output of 100 units. i.e. – each combination of factor
inputs on the lowest isoquant results in hundred units of output.
Each point on the second curve would result in an output of 200 units. i.e. – each combination of
factor inputs on the second isoquant results in two hundred units of output.
Just as with indifference curves (when further units of good x fail to yield further satisfaction) when an
isoquant becomes parallel with one of the axes, additional units of that factor of production fail to
increase output.
K
Extra units of Capital fail
to add anything to output
after this point.
6.03 THE M R T S
The gradient of the isoquant shows us the Marginal Rate of Technical Substitution between factor
inputs.
This is: how many machine hours are needed to replace one man hour while keeping output
at the same level (i.e. on the same isoquant).
of the isoquant: dK
• dL
Given a production function of the form: Q = aL K
Q
then: = aL-1K [the Marginal Productivity of Labour]
L
Q
and: = aLK-1[the Marginal Productivity of Capital]
K
Q Q
Invoking total differentiation: dQ = dL + dK
L K
-1 -1
Invoking total differentiation: dQ = [aL K ] dL + [aL K ] dK
But as we are moving along an isoquant, total output remains unaltered - i.e. dQ = 0
-1 -1
: [aL K ] dL + [aL K ] dK = 0
-1
dK aL K MPL
and hence: = -1
dL aL K MPK
The "returns to scale" experienced by a firm demonstrate the technical relationship between changes
in the quantity of inputs and the resultant change in the level of output.
If factor inputs are increased simultaneously by the degree n and as a result output rises by a
degree more than n, the firm is exhibiting increasing returns to scale.
q = f(L, K)
20
vq = f(nL, nK)
Q > 20
where v > n
10
Q = 10
0 10 20 L
Typically, increasing returns to scale do not occur at a consistent rate, rather they occur at an
increasing rate of increase:
90
80
70 q = 375
60
50
40
q = 75
30
20
q = 25
10
q = 10
10 20 30 40 50 60 70 80 90 100
If factor inputs are increased simultaneously by the degree n and as a result output rises by a
degree less than n, the firm is exhibiting decreasing returns to scale.
q = f(L, K)
20
Q < 20 vq = f(nL, nK)
where v < n
10
Q = 10
0 10 20 L
Typically, decreasing returns to scale do not occur at a consistent rate, either, but rather they
occur at an increasing rate of decrease:
v
100
90
80
70 q = 34.2
60
50
40
q = 28.5
30
20
q = 19
10
q = 10
10 20 30 40 50 60 70 80 90 100
If factor inputs are increased simultaneously by the degree n and as a result output rises by a
the same degree, n, the firm is said to be exhibiting constant returns to scale.
K
q = f(L, K)
20
vq = f(nL, nK)
Q = 20
where v = n
10
Q = 10
0 10 20 L
Constant returns to scale do not really exist – but there is a point (between low and high levels of output)
when a firm stops experiencing increasing returns to scale and starts to experience decreasing returns
to scale) and at the point, the firm can be said to be experiencing constant returns to scale.
v
100
90
80
70 q = 80
60
50
40
q = 40
30
20
q = 20
10
q = 10
10 20 30 40 50 60 70 80 90 100
If isoquants for fixed intervals of output (i.e. say: 10 units at a time) are getting closer and closer
together, this implies increasing returns to scale.
And if they're getting further and further apart, this implies decreasing returns to scale.
K
110 units of output
Decreasing
Returns to
Constant Scale
Returns to
Scale 100 units of output
90 units of output
80 units of output
Increasing 70
50 60 units of output
Returns to 40 units of output
30
Scale
20 units of output
10 units of output
L
The Law of Diminishing Returns states that if one factor input is increased marginally while all others
are kept constant, there will be diminishing marginal returns to that factor of production. [Diminishing
Marginal Productivity]
As more and more labour is put into a production ………which means that the marginal
process (holding capital constant) output will rise product is decreasing:
but at a decreasing rate of increase:
Total Marginal
Output Output
[Q]
Total [Q]
Product
Marginal
Product
Labour Labour
Mathematically, this is because the function is rising at a decreasing rate of increase. Although the
change in output (marginal productivity of labour) is positive, the rate of change – or the change in the
change – is negative:
Q
i.e. >0
L
but ²Q
<0
L²
This can easily be seen with reference to a simple production function (which exhibits constant returns
to scale). Assume initially that K* units of Capital hours and 1 Labour hour produce 10 units of
output [i.e. Q = 10 = f(1L, K*)]:
K
60 units
K*
50 units
45 units
40 units
30 units
20 units
10 units
1 2 3 4 5 6 7 L
Imagine if K is held constant at K*, but units of labour are increased incrementally. We can measure
the change in output given a change in L (the Marginal Productivity of Labour) on the above picture.
• holding capital constant, a second man hour adds 8½ units to output [i.e. MPL = 8½].
K
60 units
K*
50 units
45 units
40 units
30 units
20 units
18½ units
10 units
1 2 3 4 5 6 7 L
• still holding capital constant, a third man hour adds 7½ units to output [i.e. MPL = 7½]
K
60 units
K*
50 units
45 units
40 units
30 units
26 units
20 units
18½ units
10 units
1 2 3 4 5 6 7 L
• still holding capital constant, a fourth man hour adds only 7 units to output [i.e. MPL = 7]
K
60 units
K*
50 units
45 units
40 units
33 units
30 units
26 units
20 units
18½ units
10 units
1 2 3 4 5 6 7 L
• holding capital constant again, a fifth man hour only adds a further 6½ units to output
[i.e. MPL = 6½]
K
60 units
K*
50 units
45 units
40 units
39½ units
33 units
30 units
26 units
20 units
18½ units
10 units
1 2 3 4 5 6 7 L
• holding capital constant yet again, a sixth man hour only adds a further 5½ units to output
[i.e. MPL = 5½]
K
60 units
K*
50 units
45 units
40 units
39½ units
33 units
30 units
26 units
20 units
18½ units
10 units
1 2 3 4 5 6 7 L
• and holding capital constant yet again, a seventh man hour only adds a further 5 units to output
[i.e. MPL = 5]
K
60 units
K*
50 units
45 units
40 units
39½ units
33 units
30 units
26 units
20 units
18½ units
10 units
1 2 3 4 5 6 7 L
Another way of examining this phenomenon is to consider a production function (which also displays
constant returns), as below. Holding K constant at K*, you determine how many extra units of L it
takes to produce each marginal unit of output on the diagram below:
K
K*
4x
3x
2x
x
L
[You might like to assume that Q = 1x = f(10L,K*) to start off with]
Decreasing Returns to Scale occur with a change in all the factors of production employed. The Law
of Diminishing Returns is concerned with the effect of a change in just one of the factors employed.
Decreasing Diminishing
Returns to scale Returns to a Factor
1 1 100 1 1 100
2 2 250 1 2 200
3 3 400 1 3 300
4 4 600 1 4 400
5 5 725 1 5 490
6 6 840 1 6 560
7 7 900 1 7 580
8 8 920 1 8 585
The gradient of the isocost is determined by the Factor Price Ratio. An increase in the price of labour,
i.e. the wage rate w, causes the gradient of the isoquant to become steeper.
K
PK = the price of capital,
TVC
normally referred to as r
PK
(the rental cost of capital).
TVC TVC L
PL PL
1 0
Whereas in consumer theory, such a price …..an increase in the price of a factor does not
rise would cause the individual to be on a (necessarily) cause a drop in output, as the
lower level of welfare…… isocosts is not a constraint.
Beer K
M TVC
PB PK
U0
Q0
U1
The isocost is significantly more flexible than the individual's budget constraint. Firms find credit
terms easier to obtain, and loans more easily available for production purposes than individuals do for
consumption purposes.
Consumer Theory was all about welfare maximisation, subject to a budget constraint.
[The straight line was the constraint, and people tried to be on the highest indifference curve, given this constraint).
In a manner very similar to consumer theory, the A change in the Factor Price Ratio will
optimal position for the firm to produce at is where change the optimal manner in which the
the Marginal Rate of Technical Substitution is firm produces its output – i.e. a change in
equal to the Factor Price Ratio: its capital/labour input mix:
K K
TVC TVC
PK PK
MRTS’ = FPR’
MRTS = FPR K1 •
K0 • K0 •
Q0 Q0
TVC
PK
0
L0 L2 TVC L
PL
Notice: the optimal capital:labour ratio is determined by relative prices (the price ratio) rather than by
actual prices. Countries like the United States and Japan and Germany are capital intensive not because
capital is “cheap”, but because capital is relatively cheap compared to labour. Countries like India and
Kenya are labour intensive not because capital is expensive, but because there is abundant labour
supplies, and hence the price of labour is cheap relative to capital.
Germany: where wage costs are very high India: where wage costs are very low
K K
KG
LG
is big KI
LI
is small
KG
LG
KG
is small
LI
KI
is big
qG KI qI
LG L LI L
The firm's expansion path is the locus of all the technically efficient points of production at different levels
of output. It shows the path along which a firm should travel, if it gradually wishes to expand its output.
The gradient of the expansion path is the capital to labour ratio (K:L or K/L)
The Firm’s
Expansion Path
Q3
Q2
Q1
Q0
L
A change in the Factor Price Ratio causes a change in the Expansion Path. Try drawing it!
– 1 – Fixed Costs . . .
e.g. Rent, licenses, standing charges, etc. Fixed costs might be stepped. Say, in order to
No matter whether output is high or produce more than q* units of output, the firm
low – or even zero! – these costs of needs to rent another factory. In this case fixed
$F remain the same. costs will “jump” at levels of output exceeding q*.
$
$
F1 FC
F FC F0
q q* q
– 2 – Variable Costs . . .
e.g. Labour costs - wages (w), the rental of capital (r), raw materials, electricity, water, etc.
If the firm’s production function exhibited constant returns to scale, the Variable cost function
would be positively sloped and linear.
$
VC
q
But if the firm’s production function exhibits inconsistent returns to scale – typically increasing returns
to scale (which increase at a decreasing rate of increase) at low levels of output, followed by decreasing
returns to scale (which decrease at in increasing rate of decrease) at higher levels of output, then the
total variable cost would look like (see next page)…………
$
VC
q
Variable costs rise – but at a decreasing rate of increase - for a while; but then they begin to rise
exponentially.
TC = FC + VC
Diagrammatically, it is the vertical summation of the two functions, TFC (Total Fixed Cost) and TVC (Total
Variable Cost).
$
$
VC
F FC +
q q
$ TC
= F
It has a positive intercept on the vertical axis (the fixed cost component) and then rises in the same
manner as the variable cost function.
The total cost function, given that it is derived from points of optimality directly from the production
function, shows the cheapest possible manner in which a firm can produce any given level of output.
THE MANNER IN WHICH YOU DRAW THE TC FUNCTION IS CRUCIAL TO MOST AREAS WE WILL COVER
IN THE COMING WEEKS, SO MAKE SURE YOU CAN DRAW IT QUICKLY, EASILY AND ACCURATELY.
It must never look like this: Nor must it ever look like this:
$
$ $
q q
In other words: it must never be flat Total cost is never downward
in the middle (with a gradient of zero). downward sloping, over any range
It must always positively sloped. of output.
It must never look like this: Nor must it ever look like this:
$ $
$ $ $
q q q
The average cost of production is simply the total cost divided by units produced.
AC TC
Q
As such, it can be derived graphically directly from the Total Cost function. (I’ll show you this in a page
or two from now!)
The average cost, too, may be broken down into average fixed costs and average variable costs.
The average fixed cost starts at a very high level (imagine your total fixed costs are $1,000,000, and
you only produce one unit) but rapidly descends as output increases. At extremely high level of output,
average fixed cost become negligible. [Mathematically, the function would only touch the x axis at infinity;
but economically, AFC loses any significance when it’s down to a fraction of 1c, or something ridiculously
small like that.]
$ $
AFC
AFC
q q* q
Average variables costs are “u” shaped (shaped like the letter “u”]. [The reason for this should become
clear shortly – i.e. the next page or two].
AVC
The average total cost (ATC or AC) is the average fixed cost plus the average variable cost – summed
vertically:
$
$
AVC
+
AFC
q q
$
AC
=
q
If this doesn’t make sense, you try plotting the AC curve below………..
AVC
27 •
21 •
17
14
10
7
AFC
q
$ TC
$
TVC
q q
$ $
AC
AVC
q q
The principle underlying both is best understood by concentration just on the AVC. Imagine a TVC like
the first one I showed – bottom of page 138 – where the production function exhibited constant returns
to scale at all levels of output:
$ Hypothetical
TVC
q
What would the corresponding AVC look like?
Well, TCV/Q would give a constant – i.e. AVC would be a straight line:
Hypothetical
AVC
Imagine now that, for the first q* units of output, What would the corresponding AVC look like?
the actual TVC was higher than the straight line, Well, it would no longer be a constant, until the
but gradually getting closer to it: TVC became linear, but it would be gradually
approaching that constant:
$ Hypothetical
TVC $
Hypothetical
AVC
q* q q* q
Now imagine that after a certain level of output, q’, What would the corresponding AVC look like?
the TCV kinked upwards: Well, it wouldn’t be a constant any more: it
would start rising:
$ $
Hypothetical
TVC
Hypothetical
AVC
q* q’ q q* q’ q
Now imagine the whole process in a non-linear manner, with the TVC having a “constant” gradient at
only one point, and hence the AVC being “flat” at only one point.
Now look back at the diagram on the top of page 139, and take the “ray from the origin” as the
hypothetical TVC.
Marginal Cost is the increase in total cost by producing one extra unit of output.
The marginal cost is thus the gradient of the total cost curve.
MC dTC
dQ
As marginal cost is the gradient of total cost, it is thus minimised when the gradient of the total cost is
minimised. The gradient of the total cost curve is minimised just at the point where it stops rising at a
decreasing rate of increase, and commences to rise at an increasing rate of increase:
$ TC
$
MC
The Marginal Cost curve always crosses the Average Cost curve at the lowest point of the Average Cost
curve.
$
MC
AC
This is because if MC is below AC, it is pulling it down; and if it is above AC it is pulling it up.
Another way of demonstrating that the MC intersects the AC at it's minimum is by looking back to the
analysis on pages 143, and considering the following:
If the MC is the gradient of the TC (and thus, inevitably, the gradient of the TVC) it is must have the same
gradient as a ray from the origin at the point that that ray from the origin is tangential to the TC:
$ TC
• The ray from the origin
has the same gradient as
the TC function at point .
And that is the exact same point where the Average Cost is minimised.
$ TC
•
$
MC
AC
THINK ABOUT IT CAREFULLY!! It’s the intuitive understanding of tiny little principles like these which
determines the difference between people who pass and people who fail Introduction to Economics.
Economies of scale continue, of course, until the average cost curve is minimised. At higher levels of
output, per-unit cost of production (the average cost) starts to rise. Now the firm is experiencing what
is known as diseconomies of scale.
$
AC
Economies of scale should not be confused with returns to scale, discussed much earlier in this chapter.
Returns to scale show the relationship between changes in inputs and the resultant change in output –
and so influence the gradient of the total cost curve.
When a firm is experiencing increasing returns to scale, the total cost curve is rising, but getting flatter
and flatter (so marginal cost is falling). When a firm is experiencing decreasing returns to scale, the
total cost curve is getting steeper and steeper (so marginal cost is rising).
$
MC AC
Increasing Decreasing
Returns Returns
to to
Scale Scale
q
Be careful you don’t use the term “economies of scale” when you actually mean “increasing returns to
scale”, and vice versa!
If there is a change in the cost of factor inputs, this will inevitably lead to a change in all of the long-run
cost curves. If either capital or labour goes up in price, this will lead to an upward pivotal movement of
the total variable cost function. Ignoring fixed costs (because in the long run, all costs are variable):
$
TC1
You will recall earlier that I suggested that the TC function showed the cheapest way of producing any
given level of output. Given a change in factor prices, it is thus the new TC (TC1, above) which becomes
the relevant curve. TC0 is now gone! Irrelevant!
[In other words, everything that follows in the next few pages is with reference to the new TC.]
If there is a change in the price of one of the factors of production - say, an increase in the wage rate -
this will lead to a change in the optimal combination of factor inputs.
K
TVC
PK
K1 •
K0 •
Q0
TVC L0 TVC L
L1 PL PL
1 0
It is now optimal to use less labour – L1 instead of L0 – and more capital – K1 instead of K0. However,
it might not be possible - in the short run - to adjust to such a combination.
Capital may have to be imported - or even specifically built - and there might have to be careful
negotiations with trade unions before any reduction in working hours - or any redundancies - are
instituted.
In the short run, therefore, a firm might have to continue production with the old combination of factor
inputs - a situation which is no longer optimal.
Assume that a firm is initially producing an Then the price of labour rises (i.e. the wage
output of q0 at point below, using K0 units of rate goes up). The optimal manner in which to
capital and L0 units of labour. produce q0 now becomes point combining
K1 and L1.
K K
TVC0 TVC0
PK PK
K1 •
K0 • K0 •
q0 q0
But imagine that capital is fixed at K0 in the The only way that the firm can continue to
short run. produce q0 units is to continue with the old
capital/labour ratio at point , which is no
longer optimal.
K K
TVC0
PK
K1 •
• •
K0
FIXED K0
FIXED
q0 q0
L1 L0 TVC1 TVC0 L L0 L
PL PL
1 0
If an output of q0 is maintained, but combination is used instead of , the firm finds itself on an isocost
function which is further away from the origin - implying higher costs.
If an output of q0 is maintained, but combination is used instead of , the firm finds itself on an isocost
function which is further away from the origin - implying higher costs.
K
TVC0
PK
K1 •
•
K0
FIXED
q0
L1 L0 TVC1 TVC0 L
PL PL
1 0
TVC2
PL
0
In terms of the total cost function, (the new one, remember) is above .
$ TC1
•
•
q0 Output
•
•
q0 Output
In a more general case - specifically with respect to perfect competition - firms may have to change
their levels of output at short notice without being able to change their production process to the
optimal combinations of factor inputs in the short run.
Such a situation will lead to a whole new set of cost curves becoming effective in the short run.
K1 •
K0 •
q1
K2 •
q0
q2
L2 L0 L1 L
Assume the firm is producing an output of q0 at , using K0 and L0 of capital and labour inputs
respectively.
The technically correct way to expand from an output of q0 to q1 is to move to , by increasing capital
inputs to K1 and labour inputs to L1.
Alternatively, the technically correct way to contract the firm’s output from q0 to q2 is by moving to
decreasing capital inputs to K2 and labour inputs to L2.
Increased Production
Imagine that we have a situation where, in the short-run, capital is fixed at K0, and the firm is producing
an output of q0 technically correctly at .
K1 •
K0 •
q1
q0
L0 L1 L
Suppose that the firm suddenly receives an extra order from Japan, and wants to quickly increase its
level of output to q1. The optimal manner in which to expand would be to employ more labour [increasing
from L0 to L1] and more capital [increasing from K0 to K1] – i.e. a combination of labour and capital at
point where the MRTS is equal to the factor price ratio.
But if the firm’s level of capital is fixed at K0 in the short-run, it will be unable to do so.
The firm can increase it’s output, however, by using its existing capital and employing more labour –
extra overtime by workers, perhaps; or the recruitment of extra part-time workers.
Examine the above diagram, and see if you can determine yourself how the firm could produce q1 with
K0 before looking at the diagram overleaf……...
K1 •
FIXED
K0 • •
q1
K2 •
q0
q2
L2 L0 L1 L0* L
By employing extra labour [L0* L0] the firm can produce q1. This is, of course, using [L0* L1] more
labour than is technically efficient. But point lies on an isocost function which is further away from the
origin than the isocost on which the (optimal) point lies.
K1 •
FIXED
K0 • •
q1
K2 •
q0
q2
L2 L0 L1 L0* L
This implies that greater expenditure being incurred to produce q1 at than would have been
incurred at .
In terms of the total cost function, therefore, is above at the output level q1.
$ TC
$y •
Where $y > $x
$x •
•
•
q2 q0 q1 Output
$
AC
$y
q
1 •
$x •
q
1 • •
q2 q0 q1 Output
Decreased Production
Suppose the firm receives a cancellation of a large order from a customer in Kuala Lumpur. It is not a
permanent decrease in demand – the Malaysian company has had a fire at its warehouse, and needs
to make some structural repairs which will take 6 weeks to complete, and will have to shut down
operations until then.
The Singaporean firm obviously does not want to make units of output which it will be unable to sell, so it
will want to cut output – say to q2. The optimal manner in which to do so would be to reduce both labour
and capital inputs to L2 and K2 respectively, and produce at the technically optimal .
But if the firm has fixed capital inputs K0, if it reduces its level of output to q2 it will be forced - in the
short run - to produce at point rather than at the technically correct point .
K1 •
FIXED
K0 • •
q1
K2 •
q0
q2
L0** L2 L0 L1 L
Point , of course, is on an isocost function which is further away from the origin than the isocost on
which the (optimal) point lies.
K1 •
FIXED
K0 • •
q1
K2 •
q0
q2
L0** L2 L0 L1 L
This implies that greater expenditure being incurred to produce q2 at than would have been
incurred at .
In terms of the total cost function, therefore, In terms of the average cost function, is also
is above at the output level q2. above .
$ TC $
AC
•
• •
•
• • •
•
q2 q0 q1 Output q2 q0 q1 Output
If the firm is using K0 - which is the optimal level of capital with which to produce q0 - and K0 is fixed in
the short-run, then any variation from the output level q0 will cause the firm to incur higher costs than
would otherwise have been incurred.
q2 q0 q1 Output
$
In terms of the average cost
functions: the short-run AC is ACSR ACLR
tangential to long-run AC at
point (notice: short-run AC
is not minimised at this point). •
• •
if K0 is used to produce any •
output in excess of q0, or any •
output less than q0, then the
average cost of production will
be higher than if the technically
correct combinations of labour q2 q0 q1 Output
and capital had been used.
154 Introduction to Economics
BRITISH UNIVERSITY VIETNAM
$
In terms of the marginal cost MCLR
function: the long-run and short-
run MCs coincide at level of ACLR
output q0 (because – if you look to
ACSR
the top diagram on this page, the
•
short-run and long-run TCs are MCSR
tangential at this level of output,
and remember: MC is the • •
gradient of TC). •
•
And short-run MC goes through
the lowest point of short-run AC,
for the same mathematical
reasons that long-run MC passes
through the lowest point of long-
run AC. © Mark Harris 2013
q2 q0 q1 Output
In summary: assume that the firm has the correct level of capital, K0, to produce q0 (below). If it produces
q0, it will be on its long-run cost functions.
$ SRTC LRTC
q0 q
$ LRMC
SRMC
SRAC LRAC
q0 q
All of these short run curves were associated with a level of capital which was consistent with the optimal
combination of factor inputs to produce q0 units – K0.
If the level of capital changed - to K1, K2 or indeed any other level – and was then held constant in the
short run at that new level, a whole new series of corresponding short-run cost curves would ensue (for
each level of capital).
Following this analysis, we can actually deduce that the Long Run AC function is merely the combination
of a large number of short-run AC functions, what is known as the Average Cost Envelope:
$
ACLR
Output
The total revenue a firm receives by selling its output is the selling price of the good multiplied by the
quantity of units that are sold.
TR = P × Q
The Total Revenue curve can be derived from the demand curve, as we saw back in Chapter 2.
P $
TR
Q Q
If you did not do it a few weeks ago, make sure that you derive a TR function, step by step - i.e.:
P 10 P Q TR TR
9 10 1 ?
9 2 ?
8 8 3 ?
7 4 ?
7 6 5 ?
5 6 ?
6 4 7 ?
3 8 ?
5 2 9 ?
1 10 ?
4
3
2
1
D
0 1 2 3 4 5 6 7 8 9 10 q 0 1 2 3 4 5 6 7 8 9 10 q
Total Revenue TR
Average Revenue = or AR =
Quantity Q
It is equal to the price of the good - i.e. it’s another name for the demand curve.
TR
AR = = P
Q
D = AR
The marginal revenue is the change in total revenue by selling one extra unit of output. More
mathematically, the marginal revenue is the gradient of the total revenue curve.
dTR
MR =
dQ
D
Q
MR
(with “a” being the choke price [i.e. the vertical intercept], and “– b” being the gradient)
dTR
Marginal Revenue is the gradient of Total Revenue: MR = /
dQ
MR = a – 2bQ
Note: Both the Demand curve and the MR function have the same intercept, a
and the slope of MR [– 2b] is twice that of the Demand curve.
[LOOK BACK NOW TO YOUR NOTES IN CHAPTER TWO, AND CHECK THAT YOU UNDERSTAND THE
IMPLICATIONS OF THE LINK BETWEEN MARGINAL REVENUE AND THE PRICE ELASTICITY OF
DEMAND!]
P P
Where MR is positive,
demand is price elastic
Where MR is negative,
demand is price inelastic
D D
Q Q
MR MR
Where MR is zero,
demand is unit inelastic
D
•
Q
MR
PROFIT MAXIMISATION
6.18 PROFIT MAXIMISATION
Profit (which is typically denoted by the Greek letter pi []) is the difference between total revenue and
total cost. It can be measured by the vertical gap between the total revenue and the total cost curves.
$ TC
TR
TR
q0 q1 Q
Profit at any level of output between q0 and q1 can be measured by measuring the difference between
total revenue (money coming in from sales) and total cost (money being spent).
$
TC
{
TR
q0 q2 q1 Q
Profit will be maximised when this gap between the TR and the TC is maximised. The largest gap occurs
when the gradients of the functions are equal to each other.
$
TC
TR
q0 qmax q1 Q
Given that the Marginal Cost is the gradient of the Total Cost Curve, and that Marginal Revenue is the
gradient of the Total Revenue Curve, this means that profit maximisation occurs when Marginal Cost
is equated to Marginal Revenue.
MC
D
qmax Q
MR
THERE ARE ONLY TWO THINGS I WANT YOU TO LEARN ALL YEAR.
max: MC = MR
Another way of trying to understand this principle is by reminding ourselves what marginal revenue
and marginal coast actually are.
Marginal revenue: is the extra money received by selling one extra unit of output.
© Mark Harris 2013 St
Marginal cost is the extra money spent producing one extra unit of output.
Examine the diagram, below, and assume that a firm is producing 5 units of output. Imagine what
would happen to its profit if it produced a 6th unit of output . . . . .
P0
MC
P1
So, as long as MC is below MR, it is worth continuing to increase output, as every extra unit we
produce adds to total profit being made.
Now assume that the firm is making 9 units of output. What would happen to its profit if it increased
its output to 10 units?
MC
P2
P1
P
If we produced a 9 unit we
h
So, as long as MC is above MR, it is silly to continue to increase output, as every extra unit we produce
subtracts from the total profit being made.
From the previous page we saw that as long as MR is greater than MC, it is worthwhile to continue
increasing output, as we will add to profit. And from the above we can see that if MC is greater than
MR, we should not increase output, because we will subtract from profit.
It should be obvious, therefore, that the way to MAXIMISE profit is to keep on producing only until MR is
equal to MC.
So, no matter what the demand curve (and corresponding marginal revenue curve) looks like throughout
the next couple of chapters, a firm will maximise its profit by setting MC = MR.
Setting MC = MR
determines the profit MC
maximizing output:
P q0.
1
q0 q
MR
P
q1
Even if the nature of demand (and
corresponding MR) is very different, the
same principle applies.
P1 P1
Setting MC = MR determines the profit MC
maximizing output: q1.
Setting MC = MR determines
the profit maximizing output: P2
q2.