Lecture 2, Costs and

technology
 Costs and industry structure
• Cost structures are a key determinant of industry
structure and pricing behavior.
• Definition: A firm’s “technology” is the production
relationship it uses to turn input goods into outputs.
A firm might be able to make 1 unit of output with (3,1,1)
units of inputs (x,y,z), or it might be able to make 1 unit
of output with (1,2,5) units of inputs (x,y,z).
The technology available to a firm, plus input costs,
gives us a cost function.
• We represent a firm’s technology with a production
function:
q = f(x1,x2,x3,…xk)
 Cost minimisation problem
• A firm’s cost minimisation problem is to find the
cheapest way of producing some output level
given technology and input prices.
• We represent this problem as:

By solving this problem for different levels of

we obtain the minimum cost of each possible
production level.
 Example
• Suppose we have a firm with 2 inputs, x1 and x2.
Suppose the firm has the production function f(x1,x2) =
x11/2x21/2.
Suppose that input costs w = (3, 2)
What is the cheapest way to produce q = 4?
• Solve: Minx (3x1 + 2x2) s.t. x11/2x21/2 = 4
Solve by substitution.
x11/2x21/2 = 4
x11/2 = 4x2-1/2
x1 = 16x2-1
Min: 3[16x2-1] + 2x22
Foc: -48x2-2 + 4x2 = 0
x2 = 121/3
x1 = 16/(121/3)
 Cost functions
• We typically represent the firm’s cost function with the
expression C(q) + F.
This lets us find:
• Fixed costs, F.
• Average costs, AC(q) = [C(q) + F]/q
We may separate this into average fixed costs F/q and
average variable costs C(q)/q.
• Marginal cost: MC(q) = dC(q)/dq.
• Some fixed costs may be “sunk costs”, which are fixed
costs which cannot be recovered, even if a firm does not
produce anything or exits the market. Entry costs are
often (partly) sunk.
 Costs and output decisions
• Recall that profit maximisation implies that marginal
revenue = marginal cost. Thus, a firm will produce at the
intersection of marginal cost and marginal revenue as
long as it produces any output at all.
• A firm will produce positive output (in the short run) as
long as the price exceeds AVC. A firm will shut down if
the price falls below AVC.
• A firm will exit the industry in the long run if the price falls
below its long run ATC.
• A firm will be willing to enter an industry only if the
present value of entry exceeds the sunk entry costs.
 Minimum efficient scale
• Recall that when average costs are falling, we have
economies of scale; when average costs are rising, we have
diseconomies of scale.
• Economies of scale can come from fixed costs, or from
specialization or more efficient production tecniques.
Diseconomies of scale typically occur because of increasing
costs of coordination and management at larger scale.
• We can measure scale economies by the measure S =
AC(q)/MC(q). Recall that MC < AC means AC is falling, MC >
AC means AC is rising. So S > 1 implies economies of scale
and S < 1 implies diseconomies of scale.
• Definition: Minimum efficient scale is the lowest level of
output at which economies of scale are exhausted; ie the
lowest q* at which S ≤ 1 for all q > q*.
 Natural monopoly
• Suppose that market demand is such that the maximum
market demand is less than the MES. Then, economies of
scale are global in such a scale, so we have a natural
monopoly.
• Definition: An industry is a natural monopoly if the average
cost of producing any quantity q* is less than the average cost
of producing any q’ < q*.
• Formally, we require this to be true for every q*, but typically
we call industries natural monopolies as long as this holds
true for the q* equal to observed demand, even if average
costs would eventually rise at some point past this.
• Natural monopolies are particularly common for utilities; these
often have large fixed costs for building a network (eg
electricity transmission, cable TV, landline phones), and so
strongly decreasing average costs.
 Efficient number of firms
• The efficient number of firms in an industry for producing
a particular output level is that which does so at
minimum total cost. If the minimum efficient scale is
large, it would be inefficient to have too many firms
serving the market.
• In general, the larger the economies of scale, the more
concentrated we would expect to see the market.
• There can be a tradeoff here; higher concentration
reduces welfare by increasing market power, but with
economies of scale higher concentration (ie fewer firms)
can increase welfare.
• The best attainable solution might have firms exercising
some market power and firms operating below minimum
efficient scale.
Sunk costs and market structure
• Recall that firms will only enter an industry with sunk entry
costs if they expect to break even. This means they must
earn positive profits per period to offset the entry cost.
• Example: Suppose entry costs in an industry are $100 million.
Suppose that a fair economic return on investment is 10%.
Then, firms must expect to earn excess (economic) profits of
$10million per year to be willing to enter this industry. If
expected profits are above this, more firms will enter
(reducing market power and driving down profits).
• Thus, in an industry with large sunk entry costs, we should
expect to observe a positive price markup (ie P > MC) through
exercise of market power. We should expect to observe
higher concentration in these industries.
 Sunk costs and excess entry
• We can sometimes observe an interesting asymmetry in
industries with sunk entry costs.
• Suppose that firms initially enter to the point where the
present value of industry profits equal industry costs; this is a
long-run equilibrium.
• Suppose that costs unexpectedly fall; now profits are higher
than that needed to cover entry costs, so more firms enter the
market, driving profits
• Suppose instead that costs unexpectedly rise, reducing profits
(but leave them still non-negative). This will not lead to firms
exiting: the entry costs are sunk, and cannot be recovered, so
there is no marginal incentive to exit.
• Does this describe the US airline industry?
 Economies of scope
• Recall that in general a firm may produce many different
outputs.
• We say that an industry has “economies of scope” when it is
less costly to produce some set of outputs in one firm than it
would be to produce the set in two or more firms.
• Example: Suppose the cost producing two goods q1 and q2
are C(q1, q2). Scope economies exist if:
C(q1, 0) + C(0, q2) – C(q1, q2) > 0
• Example: it may be cheaper to have a rail company that
provides both passenger and freight services than it is to have
two separate companies, one that provides each product.
• Economies of scope arise for two main reasons:
a) Outputs share common inputs. (Eg: railways, cereal)
b) Cost complementarities. (Eg software design and
consulting)