1.

Efficiency, Coordination and Economic Organization

Trade is very beneficial for both side of trading relationship and for social welfare. Individuals and
economics organizations will be analysed in this course and transaction between and within them.
Economic organizations are created entities within and through which people interact to reach
their goals. The highest-level organization is the “Economy” as a whole; “Markets” (and the way
transactions are governed, managed and carried out) are lower-level economic organizations;
“Firms” and other formal entities (e.g. labor unions, government/regulatory agencies,
associations, etc.) are economic organizations that are formed and interact with individuals in
markets. There are at least five interesting elements in this definition that are in some extent
hidden:
1. Economic organization are instruments. They are created in order to reach some goals. They are
a tool to achieve some goals
2. As long as they are instruments to reach goal, these economic organizations have to be judge in
their capacity to achieve these goals.
3. since we are talking about economic organizations, when we are talking about goals, we are
talking about economic goals and the satisfaction of economic needs
4. In order to do all this, individuals who are actually responsible of these economic organizations,
have to know their preferences and therefore their goals and be able to rank different possible
allocation of resources. They have to know the utility functions and how to maximise them.
5. Basically we need all these because economic system work in a scenario of scarcity of resources
that mainly means two things:
- if you want to consume more spare time you have to reduce the consumption of another good.
We deal with individual trade-off. If you want more of one thing you have to give up the possibility
to have more of another thing.
- But it may also be a scarcity between individuals and economic entities in the sense that if an
individual want to achieve a greater level of utility, these may become the level of expenses of
another individual or institutions.

The whole economic system can be defined as a collection of different markets. And market
mainly interact with two actors: firm on one side and individuals on the other side.
Markets coexists and interacts with the other. The input of one firm can just be the output of
another one who is on the top of the supply chain.
Individuals and firms made interrelated choices. As a consumer, if I want to consume more of one
product I should probably work more to earn that income that enable me to increase the demand
for that product or if I consume more one good may imply the consumption of less of another
good.
The existence of these mutual inter-relationships between firms and individuals across a very large
number of markets make almost infinity set of choices and possible allocations.

There is an important criterion which help us to understand when an allocation of a resources is

better than another: Pareto efficiency
An allocation of resources A is inefficient if there is some other available allocation B that
everyone concerned likes at least as A and that one person strictly prefers. In such a case A is
Pareto dominated by B (B is Pareto superior to A) and it is clearly wasteful from a society point of
view. Otherwise A is said to be Pareto efficient (or Pareto optimal).
Suppose to have two allocation: A and B (products, services or whatever). You have just 5
individuals in A. A is Pareto dominated by B if in moving from A to B the first four individuals gain

1
the same level of utility, so the level of satisfaction is the same, but the fifth one gains more in
moving from A to B.
If we have a situation in which first, second, third and fourth individuals gain more but the fifth
gain less, we cannot say anymore that allocation B is Pareto superior, since in moving from A to B
the fifth individuals gain less. Both allocations are efficient, we can’t discriminate between the
two, it doesn’t contemplate the application of the majority rule.
1. Notice that to give all resources to a single insatiable and completely selfish individual
would be efficient (ethics is not contemplated).
2. Moreover, there are typically many efficient allocations for a given collection of resources.
Thus, the efficiency criterion may be weak on ethical grounds and as a predictor of
outcomes.
It is also fair to say that its predictive power is not totally absent: this is because of the efficiency
principle.
The efficiency principle: If people are able to bargain together effectively and can effectively
implement and enforce their decisions, then the outcomes of economic activity will tend to be
efficient (at least for the parties to the bargain). Which means that, in a given time horizon only
the Pareto efficient solution will survive.
Indeed, since efficient choices and allocations are less vulnerable, we should expect inefficient
arrangements being supplanted over time, while efficient ones survive.
Basically, if we have A and B and the first four individuals in B stay the same as level of utility and
the fifth one increases his own utility. B Pareto dominate A and B will emerge overtime and A will
be disregarded because the fourths individuals have no interests in moving from A to B, the first
four will achieve the same level of utility.

One useful instrument to better highlight the gain that actor may obtain from trade is the
Edgeworth Box.
It starts to a very simplified framework: we have two consumers, A and B. Their endowments:

(endowment allocation)
for examples the total quantities available are

In the Edgeworth Box the width is made by the total number of good 1 available (8) and the height
is the total number of good 2 (6). The box identifies all the quantities available of the two goods.
The BOX includes all feasible allocations of the goods between the 2 consumers. Of course,
including the before-trade allocation (endowment allocation). The consumer A is represented on
the lower left corner and consumer B is represented on the upper right corner.
Endowment allocation represent only one among other feasible and possible allocations. An
allocation is feasible in the extent it is contained in the box.
Which allocations will be blocked by one or both consumers? Which allocations make both
consumers better off? In order to answer we have to introduce the concept of indifference curves.
In economics, an indifference curve connects points on a graph representing different
quantities of two goods, points between which a consumer is indifferent. That is, any
combinations of two products indicated by the curve will provide the consumer with
equal levels of utility, and the consumer has no preference for one combination or
bundle of goods over a different combination on the same curve. One can also refer to
each point on the indifference curve as rendering the same level of utility (satisfaction)
for the consumer. One important characteristic is that curves cannot intersect each
2
other because indifference curve means that allocation x is indifferent to allocation y. That means
that they will be allocated on the same curve.
We can have different type of indifferent curves. For perfect substitutes indifferent
curves will be straight lines with inclination -1. You really do not care about one
rather than the other good. As long as you can have more level of the
other, the level of utility increase. So, the level of utility increase going
up on the right corner.
For complementary goods (as right and left shoes) level of utilities still increase going
on the upper side on the north-west.
We will deal with monotonic and convex preferences.
Monotonic means that the more you have of one good and the better it is, always, consumer will
never achieve a satiety point. Convex preferences mean that you prefer a mixed bundle of two
goods rather than the extremes. Balanced quantity of the two goods is always better.

In the endowment allocation we have to illustrates the two indifferences curve (one
for consumer A and one for B). there is a space for improving the welfare for both
parties, it is represented by the free space between the two indifferences curves. An
allocation that improves the welfare of a consumer without reducing the welfare of
another is a Pareto-improving allocation.
Trade improves both A’s and B’s welfares. This is a Pareto-improvement over the
endowment allocation. i.e. A reduces consumption of good 2 and sell it to B who in
exchange reduce consumption of good 1 and sell it to A. Further trade cannot
improve both A and B’s welfares.
All these movements will keep on until there will be no more possibility of
improvement à when the two indifference curves are tangent, trade can’t improve the welfare
neither of A nor B without deteriorating the welfare of the other side. Pareto-optimality have
been achieved.
Of course, in the Edgeworth box there could be many Pareto-optimal allocations. The contract
curve is the set of all Pareto-optimal allocations. The Core is the set of all Pareto-optimal
allocations that are welfare-improving for both consumers relative to their own endowments.
Rational trade should achieve a core allocation.
Trade is also useful for welfare
WELFARE THEOREMS
1. From any endowment point, trade will lead to a competitive equilibrium that is a Pareto
efficient. If you give the two people the initial endowment and you let them trade, you will
reach the Pareto efficiency. To competitive equilibrium to Pareto equilibrium
2. From Pareto efficiency to competitive equilibrium. If you want a specific Pareto efficient
allocation to emerge into the economic system, there would be a particular endowment
allocation that could be reached by actors through trade.

It is also important because it leads to more goods and services in the economy (to be traded).
These occurs for two basic reasons:
- Specialization (theory of comparative advantage)
For explaining it let’s start with an example. There are two individuals: Bob and Ann. In 24h
Bob produce 10 fishes or 10 bananas, while Ann 30 fishes or 10 bananas.
1st scenario: NO TRADE, Bob à 5 F and 5 B; Ann à 15 F and 5 B. As a result, total
production will be 20 F and 10 B.
2nd scenario: TRADE, trade enables specialization. It enables individuals to specialize in the
3
 production where they have a comparative advantage with the other actors. Anna is more
capable to produce fishes. Bob, compare to Ann, is a comparative advantage to Ann.
The fundamental concept of opportunity cost is the value that one can obtain by using a
resource in the best alternative use. When Bob produce 1 banana, he has the opportunity
cost of 1 fish. For Ann, the opportunity cost of producing 1 banana is the fact that it gives
away the opportunity to produce 3 fishes. That’s why Bob should specialize in producing
banana. à Bob 10B and Ann 30F as the total production.
Both actors have a convenience in trading banana and fishes. The agreement would be 1
banana for 2 fishes because Bob will be keen on trading if he can have more than one
fishes for a banana while Ann will be interested on trading if she can give less than 3 fishes
for one banana. This agreement would make everyone happy. Bob will consume 5B and
10F while Ann 5B and 20F. they have achieved a greater level of utility.
- Specialization increases productivity
Find your talent and try to bring to perfection this talent because this won’t increase only
your own level of utility but will also increase the social welfare of all of us. If you specialize
yourself in something very specific you became smarter in doing that, you increase your
productivity. For instance, Bob could improve the number of bananas produced from 10 to
20 and Ann from 30 fishes can improve to 40, 50 and so on

But productivity and specialization require coordination among actors. In order to specialize our
self, we need coordination mechanisms in place which enables us to confidently specialize our
skills. Always remember that specialization requires coordination. Time and efforts of specialists
are wasted unless a) they can be sure about the fact that the other specialists are doing their part;
and b) they will be able to buy on the market what is necessary for their needs.
To better coordinate we need information. We can get all this relevant information from:
-centralized planning à intention to collect all information about preferences of people, centralize
all this information and in a hierarchic way dictate what should be produced and consumed.
-assigning this role to markets à the role of markets is to make interacting individuals and firms.
Markets enable autonomous decentralized decision. Markets are costless mechanism to achieve
efficient allocations because PRICEs act as “information vehicles” by signaling scarcity. But this is
100% valid only if MARKETS ARE PERFECTLY COMPETITIVE: only in this case prices signal the true
benefits and costs for the use of resources by the economic system.

Best possible general economic equilibrium possible in the whole economic system (best scenario
for a static point of view for a policy perspective).
It is the collection of markets that act in a competitive way and perfect competition is taking place.
But perfect competition may exhibit “market imperfections”:
-Market power
-Externalities
-Asymmetric information
-Transaction costs

2.Insights on Competitive Structures

All different market forms can be ranked from perfect competition to monopolies in terms of
market power. Market power is the power of a firm to set prices over marginal costs. That
measure the power of a company in the market. That power will be at its lowest point in perfect
competition and will reach its maximum level for what concerns monopoly.
4
Perfect competition
One important premise is that in all these analyses we will adapt the paradigm of the structure-
conduct-performance paradigm (SCP). According to the structure–conduct–performance
paradigm, the market environment has a direct, short-term impact on the market structure. The
market structure then has a direct influence on the firm's economic conduct, which in turn affects
its market performance. Therein, feedback effects occur such that market performance may
impact conduct and structure, or conduct may affect the market structure.

Def. Firms do not have market power. A perfectly competitive market is a market where firms are
price-taker, i.e. they do not determine the price to which sell their products, price is settled by the
market, i.e. by the interaction of demand and supply.
There are 5 central assumptions, necessary for the behave of the market
• Atomicityà firms have to be really small compared to the market demand
• product homogeneity à all firms should produce exactly the same product or service
• perfect information (every agent, firms and consumers) know the price charged by every
firm.
• Firms have access to all production technologies (for simplicity one can assume the
extreme form of technology symmetry, but it’s not necessary).
• No entry and exit barriers (free entry and exit)
If these assumptions are met, firms are price-taker. Why?
• Products are homogenous
§ If the company raises the price, consumers will buy the products from competitors
and the demand for the company is null (perfect information). Therefore no
(rational) firm will raise the price.
§ Firms cannot collude, given their high number (atomicity)
§ If the company reduces the price, consumers will try to buy all the products from
that company. But the company is not able to serve the entire market due to its
limited production capacity (atomicity)
• Why no extra-profits?
§ If a firm makes extra-profit, this extra-profit will attract other firms to acquire the
technology required (equal access to technology) and enter into the market (no
barriers).
§ This induced competition will erode in the long run any possibility of extra-profits
for the firms with the adoption of the best technologies possible by firms

Perfect competition equilibrium (at least in the long run): p = MC = AC min

price will be equal to marginal costs and to the minimum to the average cost curve.
The maximum quantity of the good produced at the lowest possible cost = Max Productive
efficiency
The maximum quantity of the good sold at the lowest price= Max Allocative efficiency
These two concepts of Productive and Allocative efficiency are quite important.
Productive efficiency. A configuration A is more efficient in productive terms than a configuration
B if, given the same level of quantity produced, A will use lower amount of inputs than in
configuration B.
Allocative efficiency. Starting from the same level of initial resources in configuration A and B
(inputs), A is more efficient than B in allocative terms if the quantity produced and traded is
bigger.
5
Perfect competition realises the best of these two concepts of efficiency. They represent the two
important elements of social welfare of a static point of view.

The long run equilibrium in graphical terms (p = MC = AC min)

Average costs (AC)
Marginal costs (MC) are increasing in terms of quantity
because we are moving ourselves in a contest of scarcity. The
most you produce, the more you earn.
If the MC curve intersects the AC curve by construction it has
to do it in the minimum of the AC curve, that is called
minimum efficient scale (quantity level in which the AC curve
stop decreasing).
All firms present in a perfect competition set the price at the
lowest possible. That’s why we have the maximum level of
allocative efficiency.

Price comes from the interaction of demand and supply. Let’s suppose that the demand curve
increase, there would be a new minimum price and the MC curve represent the individual supply
and every single firms will be able to sell if it can gain a price at least equal as the marginal cost.
Demand increase, price increase for the additional units because the additional cost for the
additional units equal the marginal cost. MC for each single firm represent, in a perfect
competition, the individual supply. The extra profit is exactly equal to tot revenues – tot costs.
Total cost is also equal to average cost to quantity. Quantity produced will increase until the old
equilibrium is restored.
Another important concept related to allocative efficiency surplus.
Consumer’s surplus:
§ Difference between the price an individual is willing to pay to have a certain good
or service and the market price for the same good or service
§ It measures the consumers’ welfare
Producer’s surplus:
§ Difference between the price of a certain good or
service paid to the producer and the price the
producer is willing to accept for selling the same
good or service
§ It measures producers’ welfare
Social welfare: consumer’s surplus + producer’s surplus. Sum of
variable profit off all the firms. Total Profit – Fixed Costs.
W = Sc+ Σvπ
Surplus in perfectly competitive markets
• Consumer’s surplus (red)
• Supplier’s surplus (green)
• Social welfare or total surplus (red + green)
In a perfect competition consumer’s and supplier’s surplus are at
their maximum, so basically transactions which are economic convenient, are effectively carried
on. The demand curve shows the willingness to pay by consumer, while the supply curve measures
the willingness to sell by a producer. In perfect competitive market social welfare is at its
maximum level.

6
A profit equal to zero in a perfect competition means that all your input in production could not
obtain more by being employed in alternative usage than what you are actually doing. Even in
perfect competition firms have an incentive to stay in the market.

(Short-run) Supply curve in p.c. equal to the horizontal

sum of marginal costs of individual firms (which number is
fixed in the short run)

Potential competition (Contestable markets)

Results are not exactly the same in terms of performances of the firm and social welfare, but they
are also not so different from perfect competition. Potential competition makes a market
contestable. A contestable market is a market where firms from other markets/sectors can
perform a hit and run competition with no costs of entry and exit.

Pre-requisites
- No requisites on the n° of firms in the market (even one-firm market)
- No entry and exit barriers (no sunk costs and non-redeployable investments). Of course, they
have to make investment but not redeployable.
- Perfect information for consumers (they are able to react immediately to price differentials
between companies). Consumers have to know everything.
- Time requested for the incumbent to retaliate to the entry of the new firms (by lowering price) is
superior to the time needed for the entrant to make all the investment necessary to operate in
the focal market.
Suppose that I am the only firm operating in the market. If I set a price equal to the average cost
curve. I’m not making any extra profit. If I fix a price above average cost, I create an extra profit for
possible new entrants. And considering the last requisite, the new entrants will not be able to
enter on time.

Results
- Incumbent Firm(s) are forced to settle a price near to the average cost in order not to “turn on”
the signal of extra-profits.
- As a matter of fact, every extra-profit will be captured and exploited by new entrants with a hit
and run competition
- If the market is contestable, the n° of firm is a poor predictor of the market power, and even a
market with only 1 firm may behave more similarly to perfect competition rather than monopoly

PERFECT COMPETITION and CONTESTABLE MARKETS difficult to observe in the real world. Only
useful as benchmark models for comparing results (in terms of efficiency and social welfare) of
more realistic market structures, where firms do have market power (oligopoly, monopoly)

MONOPOLY & DOMINANT POSITION

A Monopoly is a market where there are just one single firm and it is
not threatened by potential competition. Now price-taking is not
credible anymore, the monopolist is a price-setter.
Choosing the price is equivalent to choose the quantity to produce.
The monopoly wants to maximise total profit. Maximum profit will
arise when marginal revenues MR equal marginal costs MC
7
Max π = TR(q) – TC (q), p (q) · q – TC (q) MR = MC
The optimum price could be set by the elasticity of the demand.
And you can express the optimal decision by the monopolists in terms
of the mark-up. The monopolist will settle a higher price as long as it faces an inelastic
demand, while the more the demand curve is elastic the more it will fix a price close
to MC, and similar to the one that would emerge under perfect competition. You can
see that it in this formula because when epsilon is very high the (elasticity is very high) optimal
thing for the monopoly is to fix a price close to marginal cost. When epsilon is not very high
(demand is not very elastic) the price can diverge from marginal cost.

Given two hypothetical demand curve a and

b, the monopolist will set a higher price in
market A instead of B because in A it faces a
much more inelastic demand. Because any
increased price is much more punished, in
terms of quantity demand, by consumer in
contest B rather than A.

[Note that in perfect competition the

(perceived) DD is horizontal and elasticity
tends to infinity (P= MC)]
NOTES about monopoly
1. Dominant Firm. We can consider a market with are
characterized by a very large firm (A sort of shark) surrounded
by a large number of tiny firms (plankton). A part of the
demand will be served by these small firms. The analysis on
pure monopoly we conducted so far can also be applied to the
market context in which there’s only one very large firm and a
set of very small firms with a limited production capacity.
If K is the total production capacity of the set of small firms,
these latter typically fix a price only marginally inferior to the
one fixed by the large firm and produce a quantity such as their
capacity is saturated:

2.One monopoly is always better than 2

2 monopolies (wholesaler and retailer) will be worse in terms of welfare for both firms and
consumers than one monopoly that works in both. One monopoly increases the social welfare.
This is known as the “double marginalization” problem in economic. We can understand it
through an example. Example DMP
We have a monopoly in a retail market that is represented by this equation: p = 5 – (1/50) q,
no fixed costs and use of only 1 input for each unit of output (1 engine for 1 car).
Now suppose MC for using this input is constant and under perfect competition in the
wholesale market the retailer buys for 1 Euro. If the wholesaler would like to maximize profit
has to set MC = MR
MR = 5 – (2/50) q = MC = 1 à q = 100; p = 3

8
 Observe that the Marginal Revenue curve for the retailer is the Demand curve for the
wholesaler.

Now suppose that wholesale market is not anymore in perfect competition but instead is
dominated by a monopolist.
What is the price (and quantity) that the monopolist would settle?
The demand faced by the monopolist in the wholesale market is p = 5 –(2/50)q
MR = 5 –(4/50)q = MC = 1 q = 50; p = 3
What the monopolist in the retail market will do? Now its marginal cost is 3 and not 1 as before.
MR = 5 –(2/50)q = MC = 3 q = 50; p = 4

In the first scenario (wholesale market perfectly competitive): π = 200; SC = 100

In the second scenario (wholesale market in monopoly): πwholesaler = 100; πretailer = 50; SC = 25

3.Sometimes a monopoly is better than everything à NATURAL MONOPOLY it shapes the

organization and structure of important industrial sectors. Basically, all important network
industries are characterized (in some extent) by a natural monopoly. For instance, Gas, electricity
and telecommunications. From a conceptual point of view, they do not differ so much one from
another. The supply chain is divided in Production/generation à Transportation à Distribution à
Sale. the natural monopoly characterizes these sectors. In the past all these stages were
assimilated to be as a natural monopoly. Nowadays these sectors were liberalized (opened to
competition) restricted to some stages: production/generation and the sale stage. While the other
steps are still considered in natural monopoly. These sectors are regulated. Optimal solutions in
monopoly originates a loss in terms of social welfare (DWL dead weight loss)
The willingness to pay is higher than the actual price. It represents missed opportunities in trading
terms. A monopoly wants to maximize its profits and do not consider DWL. It has to be regulated.
Regulations forced the firm to set a price that maximize social welfare, or it is not very far from it.

The concept of Natural Monopoly is quite simple à it is a

production in which there is the subadditivity of cost function that
means [a]n industry in which multi-firm production is more costly
than production by a single firm (Baumol, 1977, p. 810)
Producing the giving quantity to a single firm is less costly that
making produce exactly the same quantity at more than one firm.

9
Note: Scale economies are a sufficient but not necessary condition to
prove subadditivity
(And as a consequence, to have natural monopoly) one typical cost function is the total cost
function TC = F + cq the average cost function AC = F/q + c MC= c
The average cost function is always decreasing with respect to quantity. If you have to represent it
graphically, average cost is always decreasing and tend to marginal cost. Subaddivity to the cost
function is always verified because if small firms produce a lower quantity will nevertheless pay
high fixed costs. The fact that the AC is decreasing is a sufficient condition but strictly not
necessary.
Even if AC is not always decreasing there could be subadditivity.
Ex. TC = q^2 +8q +9; AC= q +8 +9/q if we derive it, we find a MES at 3.
Minimum efficient scale is the quantity for which the AC curve stop decreasing. We can be in a
natural monopoly even in this situation.

Who decide the quantity to be produced? We choose the level of Q that is variable. We do that by
expressing our preferences to the demand curve. Q is the point where the demand curve (our
willingness to pay for one good) intersect the AC curve.

Public utilities
Natural monopoly was and still is the crucial economic concept behind public service utilities.
E.g. Natural Gas, Electricity, Water, Telecommunications, Railroads: all businesses that delivers an
essential good or service through a wide network infrastructure.
Before the beginning of the liberalization, all the supply chain was assigned to a single firm that
was a vertically integrated monopoly owned by the State and it was ruled by implicit regulations.
After liberalization new firm can enter in the first and the last stages. Private investors can enter in
these sectors so regulations should be explicated à regulatory agency.
Why Production and Sale were no more considered natural monopoly? Minimum Efficient Scale
MES shifted over the origin. It was convenient in these sectors also to set up these production
facilities even with a small firm (demand-side explanation). The other fact is that the demand has
increased a lot creating space for competition (supply-side explanation).

Economic Regulation
Firms should adopt the behavior decided from these regulatory agencies.
• Ex-post regulation: ANTITRUST (later in the course): collection of laws which regulates the
conduct and organization of business corporations to promote fair competition for the
benefit of consumers. 2 areas of great importance:
-Anticompetitive practices (e.g. cartels)
-Abuse of dominant position
à Policy intervenes to punish anticompetitive behavior

Ex-ante regulation: specific REGULATORY COMMISSIONs that right from the beginning set invasive
and pervasive “rules of the game” (e.g. including prices of final or intermediate services/products).
Regulate natural monopoly.
Regulation is really routed in natural monopoly but there are other two ancillary reasons: nature
of the demand and nature of the supply.

10
For what concern the demand, for most of the time the demand is inelastic, and it is very rigid.
Gas for instance is a necessity good with no or imperfect substitutes. That means that the
monopoly price will be a high price.
The second point that strength the need of regulation is the existence of “non-redeployable”
investment (sunk cost) that create barrier to entry and there are no more potential entrants.
Let’s suppose that you want to exit from the market, you will not obtain all the residual value of
the assets. A non-redeployable investment means difficult to assess how much residual value one
can get from the investment in case of exit from the market.

Why are network infrastructure often “scarcely-redeployable” investment? Their (future) value is
uncertain. Causes:
§ Long-life plants
q Returns from the investment embrace a long period of time
q Changes in the demand of consumers and shifts in their preferences are
always possible
§ Second-hand markets are highly imperfect
q Investments are specific to a given geographical and institutional context:
this reduces n° of credible acquirers: high bargaining power of potential
acquirer(s). Number of potential powers is lower, that means that the
bargaining power shift from the seller to the buyer
q Difficult from the other side (potential acquirer) infer the “true” value of the
asset: information asymmetries. When you are not sure about the quality
as an acquirer, you are available to offer a low price to acquire that good.

3. Firm heterogeneity, market power and price discrimination

Monopolistic competition and competitive selection considered a more realistic scenario than
perfect competition. Perfect competition considered profits of the firms all the same and equal
to zero. We also have a dynamic from the market where new firms enter when there is a
positive stock or exit from the market when the stock is negative. In the long run all firms will
have equal size. All firms have the same productivity. These results of perfect competition do
not really copy well the reality that characterized many firms that consider competition a
rivalry.
There are four stylized facts of firms operating in a given industry in the reality.
The first important effect is that profits are different across firms even in the long run. The
second point is that we often observe contemporaneous exits or entries in the same
industries. The third is that we observe firms with different sizes and finally we observe that
larger firms are more efficient in productive terms.
Competitive selection model Jovanovich 1982, Econometrica
The basic assumption of perfect competition is taken also in this case. There is heterogeneity
among firms in terms of productivity. Before entering in the market firms have only a vague
idea of what would be their productivity level.
Competitive MKt where each firm is charcaterized by θ (estimate of the value of own
productivity, capability):
q!
Π=pq - θ
In this simplified expression firms do not have fixed costs. The lower is θ, the higher are
variable costs They want to maximize this profit function. At the beginning of each period,

11
 each firm decides whether to remain active or not. Next, active firms decide how much to
produce, which they do by choosing the quantity that max profits.
!" #
1° order condition: p - θ = 0; so: q* = ! p θ this leads to the quantity that maximize profit.
Having obtain this quantity we can obtain the level of profit of each firm.
# # #
Maximum profit levels: Π = ! 𝑝! θ – (! 𝑝θ)! / θ ; so: Π* = $ 𝑝! θ
Firms are still price taker, but profit of each company is different. Period by period firms will
discover θ. Firms that discover low θ will exit the market. The higher is θ, the higher is the
quantity and hence the size. That means that higher firms are the more productive and also
with higher θ. (Stylized facts are demonstrated).
The result that we obtain is quite similar to perfect competition but not exactly the same
because now we consider some inefficiency that in the perfect competition are not
considered.

Monopolistic competition Chamberlin (1933, p. 57): “It is evident that virtually all products are
differentiated, at least slightly, and that over a wide range of economic activity differentiation.
Many industries where many firms are competing. There are not one equal to the other. Firms
are heterogeneous in the product they offer or the service they deliver that may change in the
way the product is packed or the item is flavored (small changes in characteristics).
is of considerable importance” (from The Theory of Monopolistic Competition)
5 central assumptions of perfect competition
• Atomicity
• Product homogeneity. à product differentiation
• Perfect information (every agent, firms and consumers) know the price charged by every
firm.
• Firms have access to all technologies.
• No entry and exit barriers (free entry and exit)
Firm became price-maker. Each firm faces a
downward-sloping demand curve.

Greater differentiation makes demand less elastic. If

the firm is a price maker, it will apply the price that
will make the maximized profit. Firms will gain some
positive extra profit. This extra profit here is a signal
that let more firms will enter to the market and the
perceived curve will shift to the origin until any firm in
the market will not make extra profit anymore. No
signal for entry will be exerted.

12
In the long run firms will produce less than he minimum
efficiency scale and we will have also inefficiency in allocative
perspective: price will be greater than the marginal costs.

Final taxonomy
P. C. : p min; Q max; Cs max; Null extra-profits; max W
Potential Competition: close (or not to distant) results to
(from) p.c.
Monopolistic competition (with moderate differentiation) and
Comp. selection: close (or not to distant) results to (from) p.c.
Oligopoly: p intermediate; Q intermediate; Cs intermediate;
Positive extra-profits; Intermediate W
Dominant Position: close (or not to distant) results to (from) monopoly (see below)
Monopoly: p max; Q min; Cs min; Extra-profits max; W min

Price discrimination.
Pricing policies that differentiate the price they charged for the product they sell. For a large
majority of cases this policy is preferred by companies. Price are set differently to different
consumers or for different products.
Why do firms want to “discriminate”?
We are referring to a world where firms have
market power. They percieved the demand
curve as downard sloping and they can set a
price higher than the marginal costs.
Marginal costs MC are constant.
Firms will set quantity that maximize profits.
This optimal price pose two problems: the
profit loss to buyers who are willing to pay
more (A, consumer’s surplus seen by the
perspective of the producer) and profit loss
with consumers who do not buy even though
there are gains from trade, value the goods
more (B).
Firms want to discriminate because they want
to recoup these two areas. We can observe three different tipologies of price discrimination.

Perfect price discrimination (I°) (Pigou, 1920):

This is an utopistic curiosity that is becoming more and more real thanks to internet. Cookies and
big data are helping to infer what is the willingness to pay of each consumer by tracking all the
purchase’s consumer history. It can occur in isolation (you do not know that the price differs from
other customers). Menu cost are null, everything is changeable with a single click. Firms will price
differently each consumer for each product sold. One consumer - One price for each unit sold
• Each customer is charged a different price – exactly matching exactly his/her willingness to
pay for each unit
• Maximize producer surplus but also Social Welfare. This maximum level of social welfare
goes all to the producer. It is pareto efficient, but we will have ethical problem
• Equity problems: no consumer’s surplus.
• Problems:
13
 – It is impossible to know which the willingness is to pay of each consumer
– Difficult to avoid arbitrage (absence of resale). The once that buy the good at a
lower price can sell it at a higher price. It may be difficult to prevent this resale
condition

3° Price discrimination means group pricing-

“Group” Pricing: different prices for different groups of consumers, same price within the
same group. Firms analysed its own demand and segment the market with a selection by
(exogeneous) indicators: Age, Occupation, Geography
• Examples: geographical market segmentation (books in India and UK); special discounts
(senior, student, etc)
Trick: apply elasticity rule to each market segment. Firms can create sub-markets with a
different elasticity. They behave as a Monopoly in each sub-Market. They can apply the
elasticity rule and maximize their own profit in each specific submarket. The producer has to
consider the resale possibility as absent. We will set different prices for different prices à
higher prices for less elastic market.
- In sub-market i: max πi = RT(qi) – CT (qi) o pi(qi) · qi – CT (qi)
- In sub-market j repeat the same and obtain:
Of course absence of resale possibilities is needed also in this case
– Implications:
• Rule: different elasticities = different prices.
• Specifically, higher prices in less elastic markets
We make the assumption of two market: uptown and downtown. In downtown the demand curve
is more elastic. It has to operate in a unique total market C. it is made by the horizontal sum of the
two demand curves. It will maximize profit in this scenario and it require MC curve = MR curve.
The price will be settled. In these graphs you observe that by applying this price you produce too
much for the uptown market and price should be resettled higher. In downtown market you will
produce less compared with the price discrimination.

No fixed costs, marginal costs constant and equal to 2. TC = 2q

According to the elasticity curve, in the uptown it will be -1 and in downtown -4

14
Price increasing of one will decrease the quantity of 1 in uptown and of 4 in downtown

Scenario A: price discrimination MR = MC = 2

Condition that maximize profit f.o.c in U: 8 – 2qu = 2; qu = 3; pu = 5; πu = 9
f.o.c in D: 5 – 2qd = 2; qd = 1.5; pd = 3.5; πd = 2.25
Total πud (in this case = Sp) = 11.25 total revenues – variable costs
Total Sc: Scu (= 4.5) + Scd (= 1.125) = 5.625
W = 16.875

Scenario B: no price discrimination firm prefer to act in both market

f.o.c in T: 6.5 – qT = 2; qT = 4.5; pT = 4.25; πT (again = Sp) = 10.125
Sc = 4.5 + [(4.5+3)*0.75]/2 = 7.3125
W = 17.4375
We obtain an increase of social welfare and an increase of consumer’s surplus and a reduction of
producer’s surplus.

2° Price discrimination à Self-selection by consumers

Firms may not be able to identify those indicators which are capable to segment the market in
submarket because elasticity di not really differ according to indicators. This discrimination let the
customer to self-select itself. This selection can be made on consumption of the same product or
different version of the same product (versioning).
- seller cannot directly identify consumer type but can still induce consumers to distinguish
themselves. This selection may be based on the willingness of consumers to consume:
- different quantities (so price paid by consumers depends on the quantity of the good
consumed: nonlinear pricing, bundling)
- Different versions of the same product (Versioning): (often) case of VERTICAL
DIFFERENTIATION (rather than HORIZONTAL)

It is an ancient pricing mechanism

“It is not because of the few thousand francs which would have to be spent to put a roof over the
third-class carriages or to upholster the third-class seats that some company or other has open
carriages with wooden benches... What the company is trying to do is to prevent the passengers
15
who can pay the second-class fare from traveling third class; it hits the poor, not because it wants
to hurt them, but to frighten the rich. And it is again for the same reason that the companies,
having proved almost cruel to third-class passengers and mean to second-class ones, become
lavish in dealing with first-class passengers. Having refused the poor what is necessary, they give
the rich what is superfluous.” (Jules Depuit, “On Tolls and Transport Charges”, 1849)

VERSIONING. Typically time is a crucial dimension, the ones that are willing to wait more are the
less precious type of consumers. The aim is to sell more than one version of the same product (e.g.
premium and basic) at different prices targeting different segments of consumers (experts and
beginners).those who are willing to spend more money should select the deluxe version and those
who are willing to pay more should select the basic one.
• This applies also when the low-quality version has the same cost of production of the high
quality one (or even higher):
– Software: Basic version obtained by degrading the premium through the
disablement (bearing some costs) of some functions;
– Information services about share prices: the “delayed” version is produced with
some additional costs with respect to the “immediate” version.
Leave that consumers choose the version they prefer (self-selection).

Versioning is a seller selling different version

of one product. In this case the dimension of
versioning is time. We have two version of this
product and two type of consumers.

In this framework there are no cost.

Profit with first degree price discrimination.
The seller can sell the immediate version to
impatient customer and the delayed version
to patient ones. Profit will be equal to 7000.
Uniform pricing. The seller has to choose one version and sell it at one price. He can sell the
delayed version to both typologies of customers. Profit will be 5000.

In the case of versioning the seller will set a price of 100 (willingness to pay of impatient) for the
40 costumers and a price of 30 for the patient ones. Profit will be 3000 (and not 5800) because it
will be calculated as (100*40) + (30*60). But also the impatient customers will buy it at 30. Why ?
The consumer surplus of impatient customers’ is equal to 0, he does not gain anything. Instead as
an impatient customer if I buy the delayed version I gain 10 in term of consumers’ surplus (40 -30 ,
that is the willingness to pay minus how much the effectively would pay). The consumers’ surplus
for the patient customers is 50 – 30 = 20. The loss of the firm will be 20*60.
The price of the immediate version should be at maximum 90 because at this price they are
indifferent to the immediate or the delayed version. The full price (100) minus the surplus I would
gain in the other alternative. In this way profit is not that high but it would be equal to 5400.
Because in this way is respected the incentive compatibility constrain.
Each consumer should select himself for the version that has be selected by the firm for such that
individual. As long as firms are able to meet this constrain the better versioning will be
implemented.

16
To make a profitable Versioning. Key constraint (incentive): you can't make the inexpensive
version too attractive to those willing to pay more (most precious consumers). Need to:
- Lower price of the premium version. (as in the example)
- Lower quality of the basic version
- Additional constraint 1 (participationcontraint): cheap version must be sufficiently cheap
that low types are willing to purchase. At the same time this basic version should have a
sufficient quality that even non-sophisticated consumers would buy it. Lage numer of client
as possible.
- Additional constraint 2 (good design constraint): it should be impossible for consumers to
transform the basic into the premium version. There are many example to firms that are
accused to cheat the consumers. One case in the windows one:

See Windows NT case (Shapiro & Varian, 1999, Information rules, p. 64, Chapter 3):
Microsoft offers two versions of its Windows NT software: the Windows NT Workstation, which
sells for about $260, and the Window NT Server, which sells for $730-$1,080, depending on
configuration. Workstation NT can run a Web server but accepts only ten simultaneous sessions;
the server version will accept any number of simultaneous sessions. According to an analysis by
O'Reilly Software, the two operating systems are essentially the same. In fact, the kernel (the core
component of the operating system) is identical in the two products and relatively minor tuning
can turn Workstation NT into Server NT. In response to O'Reilly's analysis, Microsoft claimed that
the two operating systems differ on more than 700 counts. According to one reporter:

"While the Big 'M' folks in Redmond maintain the products are vastly different, critics allege
Workstation can be switched into the Server version with a few easy tweaks. An official Microsoft
marketer suggests that's like arguing the only difference between men and women is a Y
chromosome. We think it's more akin to discovering your date is indeed a drag queen."

Versioning is not an easy strategy for the firms. It may be considered like a kind of art. You have to
find a trade-off between lots of dimensions. It is not prosecuted by law (antitrust) because it may
lead to a decrease by welfare but for the firms is difficult to use this strategy.
And what about consumer welfare? Is versioning good or bad for consumers?
Difficult to say a priori. The answer crucially depends on whether versioning is able to enlarge the
customer base or not:
a) If the number of consumers increases, CS may not be inferior with versioning, and social welfare
could be higher.
b) If the number of consumers does not increase, then versioning enables the producer to gain
more at the expense of consumers. In this case versioning leads CS to be lower, and social welfare
is unlikely to increase
General rule of thumb: if price discrimination increases output it can be beneficial. If output does
not increase welfare is reduced. If the number of consumer increase the social welfare could be
higher, otherwise it would lead in a lower social welfare.

Second degree price discrimination (versioning on quantity)

Annotation about second degree price discrimination. Sellers induce customers’ self-selection by
proposing different menus (combinations quality/price or quantity/price). One dimension can be
the quantity. A typical non-linear tariff is the so-called two-part tariffs. It is made by a fixed fee
that consumers have to pay independently on how much they will consume, and a second fee
based on the quantity they would consume. This is a sort of versioning because this structure of
17
the tariff, the typical consumer faces a trade-off between two versions. Dealing with version one
the consumer spends a low level of money in absolute terms but a lot in relative terms because
each unit of consumption cost him a lot. In the second version he consumes a lot of quantity,
spend a lot of money in absolute terms and little in relative terms because the single units would
be more amortized. In this last case the marginal expenditure is constant and equal to price.
Tariff entails a fixed entry fee (f) and a per-unit price (p). T = f + pq
Total average unit price depends on quantity
Customers self-select depending on their
preferences
Simplifying is like having 2 options: a) consume
few units (and spend less in absolute terms) but
paying a high average price; b) consume more
units (and spend more in absolute terms) but
paying a lower average price. à Beware of the
participation constraint (NOT AT THE EXAM)
To meet the participation constraint (each
consumer prefer to consume rather than not
consuming at all) the seller may also propose
different combinations (f,p)
How many versions should the firm puts to the
market?
No general rule because the firm is facing a trede-off in producing more versions.
More versions: More possibilities to capture all the value from consumers but also:
- More personalization costs (which presumably are convex in the number of versions)
- Risk of “cluttering” effect for potential consumers. You do not know what to choose because I do
not know what the best option for me is. So, you don’t choose.

Shapiro and Varian (1999, p. 72): “If you can't decide how many versions to have, choose three.”
3 can be better than 2 for the phenomenon of “extremeness aversion”: risk that 2 versions (heavy
and light) are felt by potential consumers as "too big" or "too small." with the risk that a high
percentage will opt for the light version /generating less revenue for the company. To add another
category (super-gold-premium) and make the previous-premium version as the medium one can
produce some advantages. E.g. Mc Donald’s beverage

Excursus1: note that with the “extremeness aversion” argument we are now diverging from the
rational “homo economicus” implied by the neoclassical theory who always knows exactly and
with no doubt what she/he wants.
Excursus2: if switching from 2 to 3 is not possible for whatever reason, note that a firm might
always artificially increase the number of versions and sell the preferred version exploiting the
cognitive biases of consumers à decoy effect.

The Economist subscription tested on 100 students from the Sloan School of Management
1. on-line subscription for one year and access to all issues from 1997 for US $59,00: 16 students
2. Print subscription for one year for US $125,00: 0
3. Print subscription for one year and on-line access to all issues from 1997 for US $ 125,00: 84

The Economist subscription tested on other 100 students from the Sloan School of Management
1. on-line subscription for one year and access to all issues from 1997 for US $59,00: 68 students
18
3. Print subscription for one year and on-line access to all issues from 1997 for US $ 125,00: 32
By dropping the second dominated option, the number of students opting for the last option were
less.

The Decoy effects

When a consumer face two choices and one is well
performing in one parameter but less in one other
(price/quality) and vice versa, the consumer will not
know what to choose.
To add a close (to increase comparability) and strictly
dominated alternative may help orientate undecided
consumers.

Prospect theory
Generally, individuals (and so consumers) are much
more sensitive to losses than to gains.
The decoy effect exploits this psychological trait: the
average consumer does not know if he “wins” by
choosing “3” but he is sure (or at least more confident)
that he does not lose by choosing that option.
2 other important types of price discrimination by self-
selection (2° PD): BUNDLING and AUCTION

Def. Bundling is an offering of two or more distinct products as a package at a single price. It refers
to products related each other.
Pure bundling: sale of the package but not of single components.
Mixed bundling: sale of the package and of single components (package price < sum of prices of
single components).
Type of a 2° price discrimination/versioning (example Microsoft office). You pay a different price if
you consume more or less of these goods.
Pure bundling logic ≈ Mixed bundling logic. The logic behind them is not so different.
Mixed bundling may lead to greater profit than pure bundling.
Mixed bundling may be more difficult to implement.
The price of mixed bundling that maximize profit may not be the same of the price of pure
bundling that maximize profits.

We focus on the pure mode to make general points

Profitable bundling = Reduce ‘willingness to pay’ dispersion
Two software and two different consumers. If you are forced to
sell them singularly you will set the price of 100 for both and you
would have a total amount of profit or 400.
But you can sell a bundle of these two software, you would price it
at 220, sell it at both consumers and gain 440.

When “to bundle” is especially profitable?

You should have heterogeneous and negatively correlated willingness to pay of consumers for
single components (someone has to prefer more one component, someone else more another

19
one) but overall, willingness to pay of consumers for the products should be similar. If these two
conditions are met, bundle would be profitable.

Dispersion of the sum of the willingness to pay single components (20 + 20 = 40 selling software
separately, 0 with the bundle) has to be greater than dispersion of the willingness to pay the
bundle.
N.B. When a small group of potential consumers show relatively a very low willingness to pay (the
bundle and single components), try to target these consumers and sell them the bundle may not
be convenient (caveat to the respect of the participation constraint)

AUCTION (not really price discrimination)

Types. Ascending (aka English), Descending (Dutch), First-price, Second price.
Why do AUCTIONs are often preferred to sale at fixed prices (even if they entail a cost, in setting
up, managing, etc.)?
Focus: Ascending. Suppose a seller has to sell an item. There are only 2 potential buyers. Buyers’
evaluations for the item are either both 100$ or both 150$. Buyers of course know their
evaluations, while the seller doesn’t. The seller only knows that each value is equally likely.

Scenario “Fixed price”: if the seller sets a price of 100$ he is certain to sell the item (gaining 100$).
If he sets a price of 150$ he has 50% probabilities to get 150$, and 50% to end up with 0$. This
results in an expected value of 75$, which is lower than 100$. Thus, in this scenario, he will opt for
setting p=100$.
Scenario “English auction”: He starts the auction by calling out 100$. If buyers value the item 100$
they will make a sign of acceptance. As the seller asks for higher bids, buyers will be silent, and the
auction ends with a winning bid of 100$ (let’s suppose that the quicker buyer wins). This occurs
with probability of 50%. However, if buyers’ valuation is 150$, they will continue outbidding each
other, until the price reaches 150$. This event will occur with probability of 50%. So, the expected
value, in this scenario is 125$ > p=100$.

When seller has only a vague idea on buyers’ potential valuations, auction is a mechanism to make
buyers pay a price that reflect their valuations
4.Oligopolistic markets: Interdependence and Duopolistic Classical models
The profit maximalization is the objective of the firm. MR = MC.
We have also another component Structure-conduct-performance paradigm: market structure is
very important to determine firm performance. Also, in the case of oligopoly is important to taking
account the interdependence between company.

In perfect competition there are infinite firms and very small: no market power, price taker. In
monopoly there is only one firm that has the power and it take into account the elasticity of the
demand to maximize profit.

The term Oligopoly means (from Greek): oligos=few; polein=sell

• The number of companies in the market, N, is small
• The behavior of each firm significantly affects the behavior of other firms, which leads to
strategic interdependence. In define the quantity and the price, you must take into
account what other firms do. What other companies choose affect your outcome.

What tools do we need?

20
 • Neoclassical profit maximization logic à profit maximization rule (already covered)
• Strategic interdependence à game theoretical insights

We need a forma framework to analyse strategic interdependence: Game theory is the formal
modelling of optimal decision-making in contexts of strategic interaction. In its simplest version
(simultaneous game, two players, pure strategies):
There are 2 players. Each player has a set of possible actions, which may be discrete or
continuous. Different combinations of actions unambiguously determine different outcomes. Each
outcome is unambiguously associated to a pay-off for each player. Players are perfectly rational
and perfectly informed. Players decide their actions simultaneously, aiming for pay-off
maximization.

The key concept in game theory is the notion of Nash equilibrium.

• Formally, a strategy profile (i.e. a configuration of strategies) is a Nash equilibrium if no
player can unilaterally deviate from its strategy and improve its pay-off. In a strategic
situation, players do the best they can taking in consideration what the others are doing.
You must also know that the other players are going to do the same.
• Intuitively, a Nash equilibrium is a situation where each player’s strategy is the best
response to the strategy of the other players.

there are just to player: prisoners and they are making an

offer. They can betray the other and have a discount of
years. If both prisoners stay silent, they would stay in
prison for 1 year. If they betray each other they would
stay 2 years in prison. If only one betray the other, he
would be free and the other would stay in prison for 3
years.

The optimal strategy is to betray in both cases (dominant

strategy = best strategy regardless the other prisoner is
doing). The Nash equilibrium is both serve 2 years.

This equilibrium is not Pareto optimal because it could be easily that both serve 1 year. They
cannot agree before, to a rational perspective they are going to assume that the other player is
going to choose what is the best for him.

Let’s consider a company that want to choose the best

production. The reaction function is a curve that tells me
which is the best response for each production the other
firm is going to produce.
My rival is going to have also a production function. The
Nash equilibrium is the point in which the two functions
intersect. Firm B knows that the best response for A is
going to be on the reaction function of firm A and vice
versa. Both of them know that the other knows.

21
In sequential games (simplest version):
All assumptions are the same, but players decide their actions sequentially, aiming for pay-off
maximization. The player acting first is known as the leader, while the other is known as the
follower. In oligopoly we will use sequential decisions.

Discrete sequential games are typically represented as decision trees.

Sequential games can be easily solved by backward induction: you start by determining the pay-
off maximizing strategy of the follower. Then, you base the strategy of the leader on this
information. The sequence of optimal actions obtained through backward induction is known as
Subgame perfect Nash equilibrium.

Two-stage sequential game of entry deterrence:

There is a decision on the part of the new entrants and then the
incumbent has to choose if start a price war or not. We have a
payoff of the new entrants if he decides to stay out of the market. -
5 for both in the case of price war; 5 in the case of acquiescence.
I have to think what the incumbent would choose. Of course, the
incumbent would choose the acquiescence because on the
contrary if he selects the price war, he would also lose 5.
The subgame perfect Nash equilibrium would be (5,5).
Models can be:
1. Collusive (e.g. cartel). Companies talk to each other and they get on agreement in order to
arrive to a solution that is better for both. They also can betray the agreement.
2. Competitive: no collusion (Bertrand, Cournot, Stackelberg).
• Simultaneous models (Bertrand, Cournot).
• Sequential models (Stackelberg).
The difference of these models are the Key variables due to the specificity of the model:
- Prices adopted by each company (Bertrand)
- Quantities offered by each company (Cournot, Stackelberg)

Bertrand Model. The Bertrand (1883) model analyzes firms’ behavior under conditions of
oligopoly, adopting price as the focal strategic variable. In its simplest form, it is based on the
following assumptions:
• Only 2 companies: duopolistic competition
• No potential entrants (closed markets)
• Firms produce homogenous good (perfect substitutes for each other)
• Player are perfect rationality; they do what is better
• Player hve perfect information
• Same cost function (same technology)
• Only 1 strategic variable: price!
• Price is decided simultaneously
Hypotheses
• 2 firms i and j producing the same good (perfect substitute)
• Price is the only strategic variable
• Firms simultaneously decide their price
• MC=AC=constant for both firms. There are no fixed costs and firms have the same MC
• Consumers demand the good from the company with the lowest price

22
What are the options for firm i? If i sets a price:
• Lower than j, it captures the entire market demand
• Equal to j, it shares the market demand with j
• Greater than j, it has a null market demand (consumers demand
the good from j)
The game is perfectly symmetric; firms are identical.

We assume a cost function of this type: TC = c*q

• Fixed costs are zero:
• Average Cost (AC) and Marginal Cost (MC) coincide:

• Thus, the profit will be:

i and j choose their price in order to maximize profits. The game is

simultaneous and competitive (each company tries to maximize its own
profit). Under our cost assumptions, profit functions are:

Nash equilibrium: Couple of strategies where none of the players find it

convenient to change strategy given the other’s strategy. No one can
unilaterally change its position and improve its situation à Each
company’s price maximizes profits given the other’s choice

It is possible to demonstrate that in the Nash equilibrium each firm

chooses a price equal to c: pi*=pj*=c
None of the two companies has an incentive to change its choice, given the other’s choice
• Price higher than c: loss of the entire demand
• Price lower than c: the firms makes losses instead of profits. P can’t be < c because
c=MC=AC
Given the market conditions, firms are identical, and the game is symmetric; the reasoning
developed for one player is perfectly applicable to the other
Three possible cases:
• Case I pi > pj > c; all consumers are going to go to pj because they are price sensitive
• Case II pi = pj > c; both firms know that lowering their price they will steal demand
• Case III pi > pj = c; same as case I
As long as at least one of the two firms set a price higher than c, there is no equilibrium. That’s
because the firm setting p > c is either getting none or half of the demand. In both cases it is
incentivized to lower the price to the point it is infinitesimally lower than the price charged by the
other firm.à Each firm has always the incentive to revise its price decision, unless the price for
both firms is equal to c=MC=AC
Thus, the Nash equilibrium is represented by the following couple of strategies:

23
 In the graph we can see how RF (reaction function) looks like. RF of firm 1 is
characterized by a cealing in correspondance to pmax and a floor in
correspondance of marginal costs because 1 can’t charge a price lower than
MC, it would make losses. The second floor is due to the price elasticity of
consumers. That pmax is the price that firm 1 would charge in condition of
monopoly.
This model is not realistic. In the real world, an increase in the number of
firms (beyond 2) normally implies a decrease in equilibrium price, whereas
the Bertrand model would predict no change in price. In reality, most
industries with only two competitors seem to make more than zero profits
Why? Customers are not only interested in price but also in the
characteristics of the product and of the geographical distribution of the firm.

Cournot model
Developed in 1838 by a French philosopher, mathematician and economist Antoine Augustin
Cournot. Symmetric Cournot duopoly assumptions:
§ Only 2 firms, duopoly
§ No potential entrants (closed markets)
§ Homogenous good
§ Perfect rationality. Perfect information
§ Same cost function (same technology)
§ Only 1 strategic variable: quantity (q)
§ Production levels are simultaneously decided
§ The price is determined by the market at a level where the demand equals the joint
production of the two firms
The strategic variable is quantity. Firms choose how much they want to
produce, and the price is given by the aggregated market demand (under the
hypothesis of standard goods, the DD has a negative slope):

The cost function is the same as in the Bertrand TC = c*q

I have just the power on q1, the quantity I have to produce. The strategic interaction is given by
the fact that there is a unique price for both firms. As firm 1 I can optimize my
quantity, but the missing part is given by the quantity of firm2, in which I have no
power. Firms’ profit functions are formalized as follows:
The two firms strategically interact by influencing the (unique) market price through
the quantity they set. Equilibrium: given the competitor’s choice, the firms choose the best
strategy to maximize their profits. Assume that:
• Firms can choose the quantity they prefer in the interval
• Both the profit functions can be differentiated in quantity
%& %&
Profit maximization for each firm entails that the first order derivative is : %"" = 0; %"! = 0. This
" !
ensure that the stationary point must be a maximum (under usual economic assumptions it always
is). In this case is not necessary to calculate the second order derivative.
Goal: derive the equilibrium (2 steps):
1. Determine the set of optimal choices of each firm given the rival’s behaviorà reaction
functions. I can only optimize my profits depends on what the other firm produces.
2. Put the two reaction curves together in order to find the combination of mutually
compatible decisions (i.e., the Nash-Cournot equilibrium of the game)

24
Def. Reaction function: locus of optimal actions undertaken by each firm in response to any given
action by the other firm
This graph gives us the intuition of how the
The dark blue line is the quantity produced in monopoly.
The intersection with our RF is the point in which q1 is
equal to zero. The optimal response of q2 is just
producing the quantity of monopoly. The same occurs
for the other point of intersection. The slope of the blue
line (-1) corresponds to different total quantities
indicating the same sum of quantities (quantity in
monopoly).
The RF of firm 1 is q1 as a function of q2 (red line), and
vice versa.
The green line tells us the quantity produced in perfect
competition. It is the quantity that is the most efficient
one, higher welfare. If q2 increases the quantity in perfect competition, q1 in forced to produce 0.
The nash equilibrium lays in between the quantity produced in monopoly and in perfect
competition. It is not so efficient like perfect competition, but it is also not harmful like the
monopoly case.

Given the following inverse demand function: P(Q) = a - bQ

Where Q is the total industry output, equal to: Q = q1 + q2
And the following cost functions are the
same (in this case!!): TC(q) = c*q.
Firm 1’s profit is:

Thus the first order condition is

Or simply:

à Firm 1’s reaction function

Given our hypothesis that the two firms are identical, we can find Firm 2’s
reaction function:
• The equilibrium is given by the couple of values
• In order to identify the equilibrium, Firm 1 decides its output on the basis of the
conjectures regarding Firm 2’s behavior; for example, if Firm 1 expects that Firm 2 will
produce the quantity , then Firm 1 will have to produce
- Equilibrium values - Expected values
I have to substitute the equilibrium variables with the expected values. In our case:

25
With symmetric firms (same Marginal Cost): c1 = c2 = c
That is the Nash-Cournot equilibrium (no player has an incentive to deviate unilaterally
from their chosen strategy)

Sequential models: 2-step competition Stackelberg model

Sequential competition; the competition articulates in subsequent steps
Simplest case: 2-step models
• Decisions are not simultaneous anymore
• The two steps are not independent
2-step model solution: backward induction
• The second step is contingent on the decisions taken in the first step: in the Stackelberg
model, the follower maximizes the profit given the leader’s choice.
• Then, the first step is optimized for the maximum pay-off: in the Stackelberg model, the
leader maximizes its profit given the follower’s profit maximizing reaction in the second
step.
It is based on the assumption that the leader:
• Knows the game rules (perfect information). Is rational
• Knows that the follower is rational. Knows that the follower in the second step will try to
maximize its profit given the leader’s decision
The model with a linear demand function and null costs: p(Q)= a – bQ

the follower’s profit is given by:

Applying the first order condition for a

maximum, we have:
i find the optimal q2 as a funtion of q1.
I found the reaction function.
The leader’s choice:
The leader considers the (future) follower’s choice when it chooses its own output level
Its profit maximization depends on

In maximizing my profit as a leader, I can

express the quantity q2 as a function of q1. I can predict the behavior of my followers.

Assuming a linear demand function, we have:

Assuming null costs and substituting the
reaction function

Applying the first order condition, the quantity chosen by the leader is:
The quantity chosen by the follower is
a a a
* a−b⋅ a−
a − bq 2b = 2 = 2 = a⋅ 1 = a
q2* = 1
=
2 2b 2b 2b 2 2b 4b
26
In the real world, firms’ strategic behaviors depend on several variables:
• Price (Bertrand model)
• Quantity (Cournot and Stackelberg models)
• R&D investments; Product features; Commercialization modes
• …
All these decisions imply strategic interaction and interdependence

5.Entry barriers, entry deterrence and limit pricing

There is a strong emphasis in strategic interdependence.
Entry: entry of a new firm producing a good that is a perfect substitute for the goods already
produced in that industry. Because otherwise the firm would not be competing with the
incumbents. This would increase competition. The dichotomy is between incumbents and the new
entrants. Note:
• The degree of substitutability depends on consumers’ preferences.
• A new entry does not imply the creation of a new firm (e.g. entry from another industry:
entry in industry A of a firm belonging to industry B (diversification)).
The entry decision depends on expected profits, that are a function of:
- Production costs.
- Demand conditions post entry à revenues. Indeed, the entry has an impact on the market
quantity and price. Thus, new entrants have to forecast the reaction of incumbents.
Demand conditions are unique: relevant for new entrants and incumbents. With new
entrants it would change.

In formal terms we have:

- Green: market demand (at industry level).
- Blue: demand for the entrant if incumbents are producing q1.
- Red demand for the entrant if incumbents is producing q2.

The new entrant will face a Potential Demand given by:

PD = MarketDemand – q(incumbent)
With q(incumbent) equaling the quantity that incumbents choose to produce
after the new entry. We have a market demand (green), you have a
correspondence in the market between the aggregated quantity of a products and the willingness
to pay from customers. When new entrant entry in a market, they will increase the quantity, and
the price is going to decrease in order to absorb the extra quantity.

Entry barriers are obstacles preventing new firms from entering a market and compete against the
incumbents. In a market with no entry (and exit) barriers:
• Every firm can enter the market and compete with the incumbents.
• In the long run P=AC. As a result of the entry of new firms and the increase of supply. Until
price is equal to AC, firms will continue to enter in the market. Entri barriers allow
incumbents to have the price higher than the average costs

27
 A) If p>AC, firms make extra-profits. Without entry barriers, new firms enter the market.
As a result of the increase in supply, price
goes down. New firms will continue to enter until P=AC.
B) If p<AC, firms make losses. Without exit barriers, the most inefficient firms leave the
market. As a result of the decrease in supply, prices goes up. Firms will continue to exit
until p=AC.

Entry barriers allow the incumbents to keep the price higher than the average cost.
However, the threat of new entry affects the price set by incumbents. After new entries,
incumbents risk incurring losses. Incumbents may tend to set lower prices in order to prevent
new entries.
TAXONOMY OF ENTRY BARRIERS
Institutional/legal barriers
- Administrative authorizations needed to conduct business. Imposed by policy maker. E.g.
in the taxi business you need a very expensive license
- Patents. Prevent other firms for using the innovation. If you want to use the technology,
you have to pay royalties at the inventor.
Structural barriers:
- Economies of scale (real or pecuniary), scope and learning. AC decrease while quantity
increase. You need to make certain capital investments.
- Customer loyalty (e.g. switching costs, brand loyalty)
- Access to key resources (e.g. distribution channels)
Strategic barriers
- Capacity investment. I might increase the capacity of my plant in order to demonstrate at
the new entrants that I am able to increase my production.
- Predatory pricing. The incumbents tries to push out the new entrant by lowering the price
- Proliferation of products. Incumbents are going to produce a wide range of products in
order to cover all the demand of the market.
Bain’s definition (1956): “Anything that allows incumbents to rise prices above competitive levels
without inducing entry”. It includes economies of scale and capital requirements. It is very vague
and general.

Stigler’s definition (1968): “Entry barriers are given by the cost that a firm entering new markets
has to bear, but that the incumbents do not have to bear”. It has a narrower scope with respect to
Bain’s definition. It is not tautological. But barriers are not relatet only to costs à strategic
barriers.

Obstacles preventing new firms from entering a market and compete against the incumbents.
Entry barriers allow the incumbents to keep the price higher than the average cost.

Sylos Labini postulate: New potential entrants behave as if they were able to forecast that the
incumbents will keep their production at the same level as before the new entry. If incumbents
produce 100 products they will keep producing 100 even after they entry the market.
Thus, the new entrant will assess whether entering or not considering:
- The residual demand diagram
- Its own cost function

28
How do incumbents react? Which are new entrants conjectures? How much will incumbents
produce after the new entry?
• Potential entrants hypothesize that incumbents will not vary their production after the
new entry.
• Incumbents do not vary their production after the
new entry.
à The potential demand for the new entrant is given
by the difference between the market demand and
the quantity already offered by incumbents.

P1 = price before the new entry

ACi = incumbents’ Average Cost
ACe = new entrants’ Average Cost
IF p1 – ACi > ACe - ACi
àThe demand of the new entrant is placed above the cost function and new firms enter the
market. If price is greater than the AC, the new entrants is going to be able to enter the market
and find it profitable. The greater the difference between incumbents’ costs and new entrants’
costs, the greater the possible difference between price and costs without incurring new entries

Hypotheses of Bain, Sylos Labini & Modigliani model:

A) Perfect information
B) Sylos Labini postulate (no production changes after the entry)
C) 2-steps competition (Stackelberg)
§ In t=1 the incumbent is the monopolist and it decides both price and quantity. It is
the leader.
§ In t=2 a new potential entrant decides whether entering in the market. It is the
follower.
D) Linear costs, constant average costs
E) Incumbent’s absolute cost advantage: incumbents’ costs are always lower than new
potential entrants’ costs.
Many factors may underlie new entrants’ cost disadvantage:
- Product differentiation (higher differentiation may be needed to compensate for switching
costs)
- Institutional barriers (e.g. payment of the royalties related with a certain patent)
- Less advantageous contracts due to lack of prior relationships with suppliers
The entry decision is taken considering the potential demand and the cost function (including the
opportunity costs of investing in other industries)
The potential entrant will actually enter the market if it can obtain positive profits:
Positive profits >incumbent’s cost advantage à p1 – ACi > ACe - ACi

If the entrant’s residual demand diagram has a part that is above the
average cost curve, positive profits can be made, and the new firm
may enter the market.
If p > ACe , there is a part of De allowing to the new entrant
positive profits.
à The new entrant enters the market and the price decreases.

There is no entry if the new entrant cannot make positive profits.

29
The higher the difference between ACe and ACi, the greater the possible difference between p
and ACi without incurring in any new entry.
à With a high difference between ACe and ACi , the incumbent can charge higher prices without
attracting new potential entrants. Entry barriers would be high. Incumbents have a high cost
advantage. à p - ACi is a proxy for the height of entry barriers.

Maximum efficiency for the incumbents at the limit price: price where p1 – ACi = ACe – ACi
The limit price is ACe. Highest price that can be charged without incurring new entries:
• For higher prices the entrant may enter the market.
• For lower prices the entrant cannot enter the market.

Model 2. Bain, Sylos Labini & Modigliani with ES

Economies of scale à decreasing average cost before the Minimum Efficient Scale.
à In the B-SL-M model with ES, the higher the economies of scale, the higher is the price limit
(i.e. the price enabling new entry deterrence).
Hypotheses:
A) B) C) as before
D) The two firms face the same cost function: no absolute cost advantages. The incumbents
are in a more better position in the learning curve.

Line of reasoning:
- SL postulate: the new entrant knows that the incumbent will keep production unchanged.
- The potential new entrant’s decision depends on residual demand (i.e. the difference
between the market demand and the incumbent’s production).
- Considering the average cost curve of the new entrant, the incumbent sets the quantity so
that residual demand will not allow any profits for the new entrant.
- As a result, the potential new entrant refrains from entering.

Where AC and the Residual Demand are tangent, the

profit for the new entrant is null

AC is higher than the residual demand De à the price is

lower than the AC à the entrant would incur losses.
- qL is the minimum quantity allowing new entry
deterrence
- PL is the Price Limit allowing new entry deterrence
We want to set a quantity such that residual demand of
the new entrants is exactly tangent to the average cost
curve. At most, with the optimal quantity, the new
entrants are to get 0.
The Price Limit is the highest price the incumbent can set in order to prevent new firms’ entry in
the market.
Charging a price lower than PL à extra profits reduction for the incumbent.
If the price is higher than the price limit, there is a part of the Residual Demand that is above the
Average Costà Positive profits for the new entrant

Limitation and critique:

30
The Sylos-Labini postulate implies irrational conjectures of the potential entrant regarding the
leader’s behavior:
• After the entry, the incumbent might not find it convenient anymore to produce the
quantity qL at the price limit PL.
• Ex-post, an accommodating strategy by the incumbent may be the most rational outcome,
as a price war would harm both players.

Dixit model (1982): removal of the SL postulate

• First step: the new entrant decides
whether to enter the new market or not.
• Second step: the incumbent decides
whether to engage in a price war or
adopt an accommodating strategy
If the potential entrant does not enter, the incumbent will make monopolist profits .
If the potential entrant decides to enter:
• Price war with low profits for both players equal to Õw
• Acquiescence à Cournot duopoly profits equal to Õc

1. As long as Õc is greater than Õw , the incumbent will choose an accommodating strategy

in the event of new entry, and the new entrant knows it.
2. Since the new entrant is aware of point 1 and it acts first, it is only a matter of selecting the
best alternative between no entry and entry à accommodating strategy.
3. No entry is preferable to entry à accommodating strategy if and only if the payoffs
associated with the former are. Thus, in this version of the model, if 0 is greater than Õc
(i.e. Cournot profits are negative).

The threat to engage in a price war is not credible. The monopolist does not have a rational
interest to implement the threat. The incumbent will produce the quantity maximizing the
monopoly profits Õm rather than the one corresponding to the price limit.
Entry is prevented if and only if Õc < 0
Can the incumbent alter the payoffs in such a way as to make its threat credible?

The second important market imperfection (market failure) and prevent an economic system to
maximize social welfare and to reach a Pareto efficient. The first big challenge is the presence of
market power: the capability of companies to set prices above costs.
Externalities
• An externality (spillover) is a cost or a benefit imposed upon someone by actions taken by
others (with no compensation). That’s the general definition.
• An externally imposed benefit is a positive externality
• An externally imposed cost is a negative externality
Examples of Negative Externalities
• Air & water pollution. All production activities that entail some waste material.
• Loud parties next door. People upstairs enjoy music over night while you are sleeping.
• Traffic congestion.
• Second-hand cigarette smoke. Especially if you are a non-smoker.
Examples of Positive Externalities

31
 • A well-maintained property next door that raises the market value of your property. You
can enjoy the view.
• Network externality. All communication technologies exploit network externalities. The
more people are owning the technologies more is the value of owning it.
• Vaccines à positive reflections also on the health of the other.

Types of externalities
• Consumption externalities
– Consumption of a good by agent A has a direct impact on agent B’s utility
– E.g., smoking, loud music, tidy garden.
– Network externality (2° part of the course)
• Production externalities
– Production actions by agent A have a direct impact on agent B’s utility/profit
– E.g., beekeeper and apple orchard, polluting firm and fisherman, etc.

àWelfare loss (blue triangle):

Marginal private (= social) benefit of
consumers (DD) exceeds marginal
private costs suffered from firms
(SS) but its inferior to the social
marginal costs suffered from the
society (which also includes external
costs): producing those units has a
cost which is larger than the benefit
from a social welfare perspective

Welfare loss (blue triangle): Marginal

private (= social) benefit of
consumers (DD) is below marginal private costs suffered from
firms (SS) but it is superior to the social marginal costs suffered from the society (which also
includes external benefits): producing those units brings a benefit which is larger than the cost
from a social welfare perspective
(opposite graphs for the consumption)
• Externalities cause welfare loss
– too much resources are allocated to an activity which causes a negative externality
(overproduction)
– too little resources are allocated to an activity which causes a positive externality
(underproduction).
à The problem is one of “missing” markets (Coase) and generally an inefficient Pareto
allocation

Example on negative consumption externalities: Roommates

• 2 agents A and B
• There are two “goods”:
• Stuff – i.e., money: mA and mB : Endowments = $100

32
 • Smoke – concentration: 0 ≤ s ≤ 1 (s = utility of smoke according to the
concentration)
• A is a smoker: uA(mA,s)
• B is a non-smoker: uB(mB,t), where t = 1-s (utility given by money and clean air)
• Note: s + t = 1
If B has the right to a smoke-free environment,
endowment is at
If A has the right to smoke as much as he wants,
endowment is at
Allow trade, or make A pay B per unit of smoke, or
make B pay A per unit of smoke reduction.
By allowing people to trade you can reach a Pareto
efficient solution. The trick is to assign rights.

Ronald Coase’s intuition was that most externality problems are due to an inadequate
specification of property rights and, consequently, an absence of markets in which trade can be
used to internalize external costs or benefits.

Coase’s Theorem(s)
• (weak version: efficiency proposition) When parties can bargain without cost (no
transaction costs), the resulting outcome will be efficient, regardless of how the property
rights are specified (e.g. Polinsky, 1974, pp 1665, ‘Economic Analysis as a Potentially
Defective Product: A Buyer’s Guide to Posner’s Economic Analysis of Law’, 87 Harvard Law
Review). Verified in most cases.
• (strong version: efficiency + invariance propositions): When parties can bargain without
cost, the resulting outcome will be efficient and the level of the externality generated the
same, regardless of how the property rights are specified. This is a much stronger
definition. It says that the level of smoke generated, in our example, will be the same.
Strong version generally not verified: A and B care about who gets the property rights and the level
of smoke generated changes (from a theoretical point need of quasi-linear preferences). Very
hardly verified.

Key message of Coase theorem is that Pareto improvements are possible to the extent that there
are ways that allow the economic system to internalize the externality through market
mechanisms à Bargaining of consumers (Edgeworth box example)
Negative production externalities by Merging of firms (steel mill vs. fishery example)
Steel mill & Fishery: price takers
Scenario A: separate
Steel mill: max π (S, x) = psS - Cs (S, x) à total revenues of steel – total cost (x= pollution)
Suppose Cs (S, x) = S2 + (x-4)2 and ps=12.
F.O.C1 (∂π/∂s) =0; S* = 6
F.O.C2 (∂π/∂x) =0; x* = 4 optimal quantity of pollution produced
πS = 36 optimal profit level
Fishery: max π (f, x) = pFf - CF (f, x), the more pollution produced, higher the costs
Suppose CF (f, x) = f2 + xf and pF=10.
F.O.C (∂π/∂f) =0; 10 -2f*- x = 0; f* = 5 - (1/2)x
Given the choice of x*=4 by Steel mill, f* = 3.
33
πF = 9
πS + πF = 36 + 9 = 45.

Scenario B: merged. Now a merged firm max π (s, f, x)

Merger: Steel mill + fishery: max π (S, f, x) = 12S + 10f - S2 - (x-4)2 - f2 - xf global profit function
F.O.C1 (∂π/∂s) =0; SM = 6
F.O.C2 (∂π/∂f) =0; 10 -2f*- x = 0; xM = 10 – 2fM
F.O.C3 (∂π/∂x) =0; -2(xM-4) – fM = 0
Substituting F.O.C2 in F.O.C3: xM = 2 and fM = 4
πM = 48 while in scenario A it was πS + πF = 36 + 9 = 45.
Intuition: Is the market through the profit signal that provides the incentive to the steel mill to
merge with the fishery. The merged firm by taking also care of the social cost of steel production
makes everybody better off.
Scenario A: S* = 6; x* = 4, f* = 3, πS + πF = 45
Scenario B: SM = 6; xM = 2, fM = 4 , πM = 48
The Steel mill now cares about the damage provoked on the fishery:
In scenario A, Steel mill F.O.C2 (∂π/∂x) =0 à (∂Cs(S, x)/∂x) =0
In scenario B, Steel mill F.O.C3 (∂π/∂x) =0 à (∂Cs(S, x)/∂x) +(∂CF(f, x)/∂x) =0

But Coase theorem hinges upon small numbers involved and absence of bargaining (transaction)
costs. What if numbers involved are larger (externalities affect society at large) and bargaining
costs are high? Public policy has a more invasive role than just assigning property rights or leave
the market exploit profit signals.
The three pillars of the theorem are:
- Property rights should be well defined. Without any possible doubt and misunderstanding.
- Number of agents involved should be small. Higher is the number and more difficult would
be to accommodate everyone’s preferences
- Transaction/bargaining costs should be small enough in order to enable agents to transact
effectively. Otherwise it become really costly to find the Pareto efficient solution.

What to do if Coase theorem breaks down. In order to solve or alleviate the problem of
externalities it is required that policy maker step in. we have three possible different approaches:
• Command & Control. Limits of pollution (negative production externalities) for different
firms. Everybody has to respect it. Policy maker decide specific limits of pollution and if the
limits are not respected, the firm would be punished.
• Pigouvian taxes (or subsidy). An attempt to reproduce social optimum demand curve and
we talk about taxes that shift them and try to move the optimal quantity produced on the
optimal level from a social welfare point of view. if you produce a certain number of
products your costs would be higher considering the tax. Each unit of production.
The exactly opposite is the Pigouvian subsidy and we can talk about it in the case we have
positive externalities. Pigouvian was an economist of the early nineteenth century.
• Tradeable permits also called Cap and trade (“artificial” markets in the spirit of Coase).
When applicable, it is preferable because both firsts approach apply burden limits on all
firms that are in the markets. But in the real world we can have heterogeneity and this
approach can exploit it by reducing total costs.

6.Public (and common) goods, What is a public good?

34
Non-excludability from consumption à once these goods are provided you can’t exclude other
people for them consumptions
Non-rivalry in consumption à the fact that one agent is using this good, does not exclude other
actors to use it. Public goods have to be provided for everyone in the same quantity
Examples: Defense, Clean air, Street Lighting
……An economic (rather than a political) concept:
its public nature depends on its intrinsic characteristics that forced the system that is the public
the right way to provide them
a) Do we have the technical capability to exclude non-payers from non-rival goods
consumption? No technically possible
b) If technically possible, is it economically feasible? (Pedestrian walks with gates?) possible
but not convenient from an economical point of view.

When to provide a public good? Example

- Suppose 2 roommates, 1 & 2
- Whether or not to purchase a TV
-Given the size of the apartment, TV in the living room: both roommates will be able to watch it.
TV is a public good.
Question: is it worth for them to buy the TV?
We suppose that the TV costs c, so in
order to purchase the TV, the sum of
the contributions
must be at least c:

Utility of Person 1:
Utility of Person 2:

Reservation price (i.e. willingness to pay) of Person 1 for the TV (utility he gets from
paying TV and watch it equal to utility he gets from not paying and not watching):
Reservation price of Person 2 for the TV: analogous

In this TV problem, there are 2 allocations of interest:

1) Each person spends his wealth only on his private consumption:
When the TV will be provided?
Surely when allocation 2) Pareto dominates allocation 1). This means:

Using reservation prices and budget constraint we can write:

Therefore, it has to be: Which in turn it implies:

Allocation 2) will be Pareto dominant if each reservation price is greater than

each payment and the total amount that the roommates are willing to pay is at least as large as
the cost of the purchase.

35
This is so straightforward you may ask why we have been through all these mathematical
formalisms for this intuitive result à In the real world, the reservation price (willingness to pay) of
each person depends on his/her personal wealth
Whether or not to provide a public good, will depend on reservation prices and in turn on the
distribution of wealth among members of a community
It is perfectly possible that for some wealth dbns: for others:
Example. Imagine a situation where one roommate really loves the TV and the other roommate is
nearly indifferent about acquiring it. Then if the TV-loving roommate had all of the wealth, he
would be willing to pay more than the cost of the TV all by himself. Thus it would be a Pareto
improvement to provide the TV. But if the indifferent roommate had all of the wealth, then the TV
lover wouldn’t have much money to contribute toward the TV, and it would be Pareto efficient not
to provide the TV.

What is the problem with the provision? The free-riding problem

Suppose reservation prices exceed personal payments and the sum of personal payments is
enough to cover the cost of the good. Thus allocation 2) is Pareto efficient. Are we sure that this
equilibrium will emerge? Of course not
Suppose each person has a wealth of 500 €, each person values the TV at 100 €, the cost of the TV
is 150 €. Suppose that there is no way for one of the roommates to exclude the other one from
watching (as before). But now suppose that each roommate has to decide independently from the
other whether or not to buy the TV
à Nash eq. (Don’t buy; Don’t buy)
Here the game is trivial but it makes the point: each player
has an incentive to free-ride on the expenditure of the other
(since by doing this, it enjoys the TV without paying
anything). But if both players think in this way the result is
that the good will not be provided, even if the scenario
where TV is bought, and the payment is equally shared is a
Pareto improvement.
In reality we can expect things not to unfold in this way: the 2
roommates contribute for a fraction of the TV and the TV will be bought. But what happens when
numbers become bigger (and so the possibility to free-ride)? No provision is a serious possibility

Who can solve the problem? Government can remedy:

Providing the public good and paying it with tax revenues (but still problems with free riders and
forced riders in paying taxes). Farced riders are these people that pay taxes but are not interested
in those public goods (military defense for pacifists, light for blind)

How much of a public good? How to decide? Complex issue

By definition the public good has to be provided in the same quantity for all individuals, even
though all individuals have different preferences regarding their “perfect” quantity. But “one size
should fit all”.
In principle, one could collect the willingness to pay/contribute of each individual for different
quantities (e.g. with exit polls) or present individuals with different options to choose from, for
then making them vote on the preferred size.
But still problems:
- How to collect trustable and credible “willingness to pay” by citizens?

36
 - Voting is not perfect (e.g. social preferences may not be transitive)
At the aggregated social level: x > y, y > z, z > x
This is a version of the “Condorcet Paradox” from which Arrow
started his reasoning that led him to the so-called
“Impossibility theorem”. Preferences should be complete and
transitive. Data can be manipulated

“Commons”
Those goods rivals in consumption but not excludable: common grazing land when the villagers
graze their cows on a common field.
– Individuals are free to graze their cows in the field with no restrictions
Hardin (1968): This situation typically determines the so-called tragedy of the commons, i.e. over-
exploitation of the field.
– Any villager is tempted to graze more than his / her (hypothetical) share
– If all succumb to the same temptation, the grass ceases to grow and the value of
the pasture to everybody disappears.
Each villager has to decide how many cows bring to graze.
c is the number of cows, each cow costs a. y (milk) = f(c) with f’(c) > 0 and f’’ (c) < 0.
Villagers are price takers: p of a liter of milk = 1 € MR = MP (talking about value or product is the
same)
A. Private property (one villager owns the field and decide how many cows enter
( π = f(c) – ac à f.o.c.: f’(c*) = a
𝑚𝑎𝑥
'
if f’(c) > a, a new cow will enter; if f’ (c) < a, a cow will exit, eq.: f’(c*) = a

B. Commons (none owns the field, everybody can freely enter with cows)
Suppose that there are c cows currently being grazed so that the current output per cow is equal
to f(c)/c. When a villager contemplates if adding a new cow, (s)he will compare f(c+1)/(c+1) > a. If
37
this inequality is verified, the new cow will be added; otherwise not. It follows that the total
number of cows will be a specific c^ that verifies f(c^)/c^ = a. This c^ is certainly greater than c*.

In its essence it is a problem of bad definition of property rights, and remedies can be searched
there at least in first that fix the problem.
- Private property provides such a mechanism. Indeed, if the field is owned by someone
who can control its use and, in particular, can exclude others from overusing it, then there
are by definition no externalities. The market solution leads to a Pareto efficient outcome.
- Of course, private property is not the only social institution that can encourage efficient use of
resources (Nobel prize Elinor Omstrom defined 8 managing principles for governing commons in
her famous book Governing the commons, 1990, freely available in Internet). For example, rules
could be formulated about how many cows can be grazed on the village common. If there is a
legal system to enforce those rules, this may be a cost-effective solution to providing an efficient
use of the common resource.
However, in situations where excludability mechanisms are not implementable and the law is
ambiguous/difficult to enforce, the tragedy of the commons can easily arise. Overfishing in
international waters is an example: each fisherman has a negligible impact on the total stock of
fish, but the accumulated efforts of thousands of fishermen results in serious depletion.
Then, there are circumstances where we want “commons”
An appeal to conscience […] (can produce) two communications, and that they are contradictory:
(i) (intended communication) "If you don't do as we ask, we will openly condemn you for not
acting like a responsible citizen"; (ii) (the unintended communication) "If you do behave as we ask,
we will secretly condemn you for a simpleton who can be shamed into standing aside while the
rest of us exploit the commons." à In these situations, social (i.e. solidaristic) values have to
prevail in order to make things work

38
7.Asymmetric Information
In purely competitive markets all agents are fully informed about traded commodities and other
aspects of the market.
◆ What about markets for medical services, or insurance, or used cars?
- A doctor knows more about medical services than does the buyer.
- An insurance buyer knows more about his riskiness than does the seller.
- A used car’s owner knows more about it than does a potential buyer.

Imperfectly informed markets with one side better informed than the other are markets with
asymmetric information.
◆ In what ways can asymmetric information affect the functioning of a market?
4 applications will be considered:
0 Adverse selection
0 Moral hazard
0 Signaling [remedy]
0 Incentives [remedy]

Model of Adverse Selection (Akerlof 1970, QJE)

◆ Consider a used car market: 100 people want to sell their used car; 100 people want to
buy.
◆ Two types of cars; “lemons” and “plums”. Everyone knows that 50 of the cars are plums,
50 are lemons.
◆ Each lemon seller will accept $1,000; a buyer will pay at most $1,200.
◆ Each plum seller will accept $2,000; a buyer will pay at most $2,400.
Suppose quality of the cars can be verified.
◆ If every buyer can distinguish a plum from a lemon, then lemons sell for between $1,000
and $1,200, and plums sell for between $2,000 and $2,400.
N.B. Gains-to-trade are generated when buyers are well informed.
Suppose that now quality can not be verified: no buyer can tell a plum from a lemon before
buying.
What is the most a buyer will pay for any car?
In this case the buyers have to guess about how much each car is worth. We’ll make a simple
assumption about the form that this guess takes: we assume that if a car is equally likely to be a
plum as a lemon, then a typical buyer would be willing to pay the expected value of the car. Using
the numbers described above this means that the buyer would be willing to pay:
(1/2)* 1200 + (1/2)* 2400 = $1800.
But who would be willing to sell their car at that price?
The owners of the lemons certainly would. The owners of the plums certainly not: they need at
least $2000 to part with their cars. N.B. At a price of $1800 only lemons would be offered for sale
◆ No plums are sold, and consumers soon realize that only lemons are sold
◆ They revise their expectations (lemons prob. = 1; plums prob. = 0)
◆ Lemons are the only cars sold: the equilibrium price will be somewhere in-between $1000
and $1200.
Even though the price at which buyers are willing to buy plums exceeds the price at which sellers
are willing to sell them, no such transactions will take place.
The problem is that there is an externality between the sellers of good cars and bad cars; when an
individual decides to try to sell a bad car, he affects the purchasers’ perceptions of the quality of
39
the average car on the market. This lowers the price that they are willing to pay for the average
car, and thus hurts the people who are trying to sell good cars. It is this externality that creates the
market failure.

In the used car market, producers do not choose endogenously whether to have “plums” or
“lemons”. But note that adverse selection may also occur when producers do have the option to
produce low-quality or high-quality goods and face different costs for producing the two

Adverse selection with quality choice. Suppose that each seller can choose the quality, or value, of
her product.
◆ Two umbrellas: high-quality and low-quality.
◆ Which will be manufactured and sold?
Suppose:
- Buyers value a high-quality umbrella at $14 and a low-quality umbrella at $8.
- Before buying, no buyer can tell quality.
- Marginal production cost of a high-quality umbrella is $11.
- Marginal production cost of a low-quality umbrella is $10

Suppose every seller makes only high-quality umbrellas.

- Every buyer pays $14 and sellers’ profit per umbrella is $14 - $11 = $3.
- But then a seller can make low-quality umbrellas for which buyers still pay $14, so increasing
profit to $14 - $10 = $4.
N.B. There is no market equilibrium in which only high-quality umbrellas are traded.

All sellers make only low-quality umbrellas.

Buyers pay at most $8 for an umbrella, while marginal production cost is $10.
N.B. There is no market equilibrium in which only low-quality umbrellas are traded.
Now we know there is no market equilibrium in which only one type of umbrella is
manufactured.
Is there an equilibrium in which both types of umbrella are manufactured?
A fraction q of sellers makes high-quality umbrellas; 0 < q < 1.
Buyers’ expected value of an umbrella is EV = 14q + 8(1 - q) = 8 + 6q.
High-quality manufacturers must recover the manufacturing cost: EV = 8 + 6q>= 11 à q >= 1/2.

◆ So at least half of the sellers must make high-quality umbrellas for there to be a pooling
market equilibrium. But then a high-quality seller can switch to making low-quality and
increase profit by $1 on each umbrella sold. Since all sellers reason this way, the fraction of
high-quality sellers will shrink towards zero -- but then buyers will pay only $8.
N.B. So there is no equilibrium in which both umbrella types are traded.
0 with just one umbrella type traded
0 with both umbrella types traded
So, the market has no equilibrium at all à Adverse selection has destroyed the entire market!

There are several markets where adverse selection mechanisms may arise: FINANCE of
INNOVATION
- There are good and bad innovation projects (bad projects = high risk of failure);
- Those who provide external finance (i.e. banks) cannot perfectly discern high risk vs. low
risk projects, and generally proponents are much more informed about odds of success.
40
 - To shield from the risk, banks pose unfavorable conditions for lending (i.e. very high
interest rates or credit-rationing).
- Only (often low-skilled and high-risk loving) “kamikaze” innovators will ask for money while
capable innovators (those with sufficiently good quality but at the same time realistic
projects) may prefer to give up searching for external debt finance (discouraged
borrowers) or search for other alternative financing sources.
- The market may soon shrink to 0 transactions.

Adverse selection in finance

A bank has to decide to grant an annual loan of 100,000 € to a firm.
The firm may either be a low-risk project proponent or a high-risk one. The bank does not know
what type the firm is. Specifically, the bank believes that with probability 10% the firm is a high-
risk project proponent. For a high-risk firm, the loan is worth 20,000 € (this is the expected profit
the firm wishes to obtain thanks to the loan), whereas for a low-risk firm, the value is 3,000 € (i.e.
expected profit). For the bank, the cost of granting the loan to a low-risk firm is 1,000 € (this is the
administrative/monitoring cost and nothing more since the loan will be repaid in due time). While
the cost of providing the loan to an high-risk firm is estimated to be equal to 30,000 € [this is the
administrative/monitoring cost + (the non-repaid loan – eventual residual value of the assets
bought by the firm) + the reputation and opportunity costs of a non-performing loan].
The average evaluation by consumers is: 10%*20K + 90%*3K = 4.7 K
The average cost suffered by the bank is: 10%*30K + 90%*1K = 3.9 K

Based on this average, the

bank is considering setting
two prices for the loan
(i.e. interest rates):
- r = 4.5% for an
annual interest
payment of 4.5 K
(that is below the
average valuation
but above the
average cost).
- r = 30% for an
annual interest
payment of 30 K
(an interest rate
that covers cost
regardless of firm
type).
Adverse selection is a
problem of hidden information
But we have a second problem originated by asymmetric information: moral hazard. This is caused
by an hidden action problem
Moral hazard in a transaction occurs when the party with more information about its actions or
intentions has a tendency or incentive to behave inappropriately from the perspective of the party
with less information.
If you have full bike insurance are you more likely to leave your bike unlocked?
41
1) Suppose no insurance is available
Consumers have an incentive to take the maximum possible amount of care (large
expensive locks, avoiding go in high-risk areas, etc.)
2) Suppose a full insurance is available
Consumers have no incentive to take any care. In case of theft, get money to the insurance
company and get a brand new bike. This lack of incentive to take care is what is called MORAL
HAZARD.

1) Moral hazard is a hidden action problem: if the amount of care is observable, there would be no
problem.
2) The insurance company faces a trade-off: full insurance (higher immediate returns) means too
little care will be undertaken by ensured (higher risk of bearing great costs later). How solved?
3) The insurance companies will not want to offer the consumers “complete” insurance. They will
always want the consumer to face some part of the risk. This is why most insurance policies
include a “deductible,” an amount that the insured party has to pay in any claim. By making the
consumers pay part of a claim, the insurance companies can make sure that the consumer always
has an incentive to take some amount of care.

Where is the market inefficiency here?

The possibility that moral hazard might occur can lead to less trade than the optimum
Consumers may want to buy more insurance, and the insurance companies would be willing to
provide more insurance if the consumers continued to take the same amount of care . . . but this
trade won’t occur because if the consumers were able to purchase more insurance they would
rationally choose to take less care!

There are several markets where moral hazard may arise: FINANCE of INNOVATION
- For an outsider (external investors) may be difficult to monitor the strategies,
choices, decisions taken by the prospective innovative managers/entrepreneurs;
- Managers/entrepreneurs may put in place actions not in the interest of the
investors, or that investors would not have agreed (e.g. divert funds to different
aims, exert less effort than required).
Banks: Use of collateral to secure debt
VC: “active investors” (i.e. work side by side with entrepreneurs), co-investment with
entrepreneurs as a guarantee of high effort, use of milestones.

8. Asymmetric Information: remedies to the market failure. We will analyze two type of
remedies: signal and incentives.
At a very broad level, quite intuitively, one can tackle (almost) any problem by: Solve it: try to
reduce information asymmetries by using signals.
Get around it (i.e. neutralize its negative effects): try to align incentives of the two parties.

Signaling
Adverse selection is an outcome of an informational deficiency. A way to deal to other selection
phenomena.
What if information can be improved by high-quality sellers signaling credibly that they are high-
quality? E.g. warranties, professional credentials, references from previous clients, etc.
42
A labor market has two types of workers; high-ability and low-ability. A high-ability worker’s
marginal product is aH. A low-ability worker’s marginal product is aL. à aL < aH.
A fraction h of all workers is high ability.1 - h is the fraction of low-ability workers.
Each worker is paid his expected marginal product. If firms knew each worker’s type they would
pay each high-ability worker wH = aH. Pay each low-ability worker wL = aL.
If firms cannot tell workers’ types then every worker is paid the (pooling) wage rate; i.e. the
expected marginal product: wP = (1 - h)aL + haH.
As long as the good and the bad workers both agree to work at this wage (and high ability
workers maintain their high level of productivity) there is no problem with adverse selection in the
eyes of the firm. But suppose the firm fears that the pooling wage scheme may depress high-
ability workers and be detrimental to its profit and therefore it is very much willing to look for a
separating equilibrium: wH = aH and wL = aL
How can this separating equilibrium be achievable?
If wH and wL is the wage scheme offered, how is possible to select the “right” types of workers for
the “right” wage?
How is it possible for high-ability workers being selected for wH and not wL? How can they credibly
demonstrate their nature?
SIGNAL: a costly method available to signal you are a good type such that the cost of signaling is
too high for bad types that they won't do it
3 characteristics for an efficacious signal: intentional, observable, costly
Workers can acquire “education”. Education costs a high-ability worker cH per unit and costs a
low-ability worker cL per unit. cL > cH.
Suppose that education has no effect on workers’ productivities (extreme assumption, just to
make the point and make it simple, obviously this should not be necessarily true in real life): i.e.,
the cost of education is a deadweight loss.
High-ability workers will acquire eH education units if
(i) wH - wL = aH - aL > cHeH, and
(ii) wH - wL = aH - aL < cLeH.
(i) says acquiring eH units of education benefits high-ability workers.
What they earn is higher than what they spend for acquire education.
(ii) says acquiring eH education units hurts low-ability workers.

Acquiring such an education level credibly signals high-ability, allowing high-ability workers to
separate themselves from low-ability workers.

Q: Given that high-ability workers acquire eH units of education, how much education should low-
ability workers acquire?
A: Zero. Low-ability workers will be paid wL = aL so long as they do not have eH units of education
and they are still worse off if they do.
Signaling can improve information in the market.
But it always comes with a cost, and for this reason, the equilibrium is sub-optimal with respect to
a full information scenario. [If the pooling equilibrium is accepted by high-ability workers (and
does not affect their high productivity) total output does not change and education is costly so
signaling may hurt market’s efficiency (to some extent resources are wasted in order to enforce a
separating equilibrium).]

Innovative start-ups may show their goodness to external investors:

-patenting
43
 -endorsement by a reputable alliance partner
Of course, these dimensions are not pure signals, but they may still have a signaling function

Incentives
They may be useful both in moral hazard and adv selection.
Principal-agent theory (Agency theory): Area of economics that deals with all situations where
there is a principal who wants an agent to act in the principal’s interest to achieve some goal but
possess less information than the agent, cannot monitor perfectly agent’s behavior and resulting
performance of agent’s action is noisy. In such situations, the agent can indulge in moral hazard,
but proper incentives set by the principal, by re-aligning objective functions between the two, can
help mitigate the problem
How the principal can mitigate the problem?
Choosing another form of remuneration, making the “agent” participate to the sharing of the
performance (i.e. part of the agent’s salary depends on the final performance)

From Sharecropping in agriculture to Stock options for managers in public companies (mezzadria)
Many corporations are owned by a large number of small shareholders. This considerably
decreases the incentives for each single shareholder to monitor the behavior of managers: the
effort necessary is very high compared to the benefit. Board of directors only partly defend
shareholders, and managers are better informed anyway. Managers are paid with a fixed fee and
a flexible one based on their performance.

Principal: output y = f(x) [p = 1 → output = value; x: agent’s effort]

Agent: payment s(y)
Principal’s π = y – s(y) → π = f(x) – s(f(x))
Agent’s utility u: s(y) – c(x) → s(f(x)) – c(x)
The agent will be willing to «work» for the principal as long as its u ≥ 0
Principal’s problem 𝑚𝑎𝑥
( π = f(x) – s(f(x)) s.t. s(f(x)) – c(x) = 0 maximize the production function
(
taking the utility of the agent at least zero.
( π = f(x) – c(x)
𝑚𝑎𝑥 MP(x*) = MC(x*)
(
Optimal incentive scheme: Marginal product of Agent’s effort = its marginal cost
If principal can observe the amount of effort exerted by the agent, wage works perfectly:
Agent’s problem 𝑚𝑎𝑥( u = wx + K – c(x) w = MC(x) principal sets w = MP (x*).
(
But if principal cannot observe x and y = f (x, ɛ), w is highly inefficient since received w, the agent
has an incentive to shirk (e.g. ↓x since ↓c(x)).
In these cases, «sharecropping» or similar incentivizing methods (Agent gets s = αf(x) and Principal
(1- α)f(x)) might be preferable. But bear in mind that they are not optimal compared to the full
information scenario:
Agent’s problem 𝑚𝑎𝑥( u = αf(x) – c(x) αMP(x) = MC(x) with x ≠ x* the level of effort would not
(
be equal to x because what the agent get from it is a percent of the output. It would not be the
optimal x from the point of view of the principal.

Of course, if signals are exogenously absent, firms may always endogenously look to put in place
some incentive schemes in order to make workers’ type reveals (“signals”) themselves.

44
With these solutions you can mitigate the problems, but it would never be optimal as in a perfect
information scenario.

Zappos’ case of incentives. It was founded in 1999 in Las Vegas for selling shoes on-line.
Management team: “customer obsessed culture”. Customer service: key asset.
365 day-return window, free-shipping, but in team’s view the Call centre 24/7 with no limits
talking (“protracted talk therapy”, as one observer noted) would have been their core advantage.
A call-centre job isn’t typically very desirable, nor does it pay well (e.g. 11$ per hour)….How could
they obtain that? Paying workers more was not an option àMore fun & power for customer-
service representatives
But Zappos management team needs to know early who really was engaged with the project and
share its customer-centric obsession culture and who is not.
To escape from a pooling eq. to reach a separating eq. where good customer-service
representatives are separated from bad ones, Zappos use the following incentive scheme:
From Levitt & Dubner (2014): “When new employees are in the on boarding period- they have
already been screened, offered a job and completed a few weeks of training- Zappos offers them a
chance to quit. Even better, quitters will be paid roughly 2000$ just for quitting”
This is the cost the firm is willing to suffer in order to tell good from bad workers.
Incidentally note:
“Fewer than 1% of new “hires” accept “The Offer”.
“In 2009, Zappos was bought by Amazon.com for 1.2$ billion.”

9. Transaction costs
Simon’s parallelism (1991, “Organization and Markets”)
Suppose that a mythical visitor from Mars approaches the Earth from space, equipped with a
telescope that reveals social structures. The firms reveal themselves, say, as solid green areas with
faint interior contours marking out divisions and departments. Market transactions show as red
lines connecting firms, forming a network in the spaces between them. As our visitor looked more
carefully at the scene beneath, it might see one of the green masses divide, as a firm divested
itself of one of its divisions. Or it might see one green object gobble up another. No matter
whether our visitor approached the United States or the Soviet Union, urban China or the
European Community, the greater part of the space below it would be within the green areas, for
almost all of the inhabitants would be employees, hence inside the firm boundaries. Organizations
would be the dominant feature of the landscape. A message sent back home, describing the
scene, would speak of "large green areas interconnected by red lines." It would not likely speak of
"a network of red lines connecting green spots.”

TC perspective (Coase 1937, Williamson 1975, 1981). In the neoclassical vision of markets, the
exchange of goods between firms and consumers is without frictions or costs in the market. If
markets are perfectly competitive, the economic system may reach Pareto optimal allocations.
The Transaction Cost Economics (TCE) doubts the existence of these “perfect” markets:
transactions are not instantaneous, the use of markets is costly: time & money to search for sellers
& buyers, negotiate exchange terms, write contracts, inspect results, enforce deals.
In many circumstances, alternative governance structures (e.g. the firm) can be more efficient
than the market.

The 3 pillars of the TCE

45
 • Bounded rationality
• Opportunism
• Relationship-Specific Investments
Without these three conditions the market works efficiently, otherwise the use of the market
implies substantial costs. All these three conditions may be put in place to make market a costly
system for transaction.

“Boundedly rational agents experience limits in formulating and solving complex problems and in
processing (receiving, storing, retrieving, transmitting) information“.
“Does bounded rationality mean that people (and therefore their actions) are irrational? Not at all.
People making choices are intendedly rational. They want to make rational decisions, but they
cannot always do so”. Incapacity to process information and to solve problem.
In decision making, rationality of individuals is limited by:
a) the information they have: the economic agent does not know perfectly the present state
of the nature and he is not able to forecast all the possible states of nature on which he
will finds himself after a decision (substantial limit caused by environmental uncertainty).
b) the cognitive limitations of their minds: even if it is possible to know the present and
forecast all the possible future states of nature, there is not an algorithm that enables to
find optimal solutions in a reasonable time. There is a “cognitive” and “time” problem
(procedural limit). Agents rely on heuristics (past actions that lead you to a satisfactory
result) and approximations for their behavior rather than maximize actions.
Substantial limit. It’s more like poker: The player has to play but he does not know the other
players’ cards and the sequence of the cards in the deck. He has to make a move, but he does not
know present and future states of nature. Uncertainty. You cannot assign probability.
Procedural limit: It’s more like chess game, the player probably knows all the possible states of
nature (rules are settled, distribution of figures in the chess is known, other’s player moves
perfectly observable), but it is impossible to find optimization’s algorithms (even with computers).
Risk (o forse no).
In the TCE the bounded rationality implies two important facts.
If 2 or more parties want to regulate a transaction by a contract:
• It is very difficult to include in the contract all possible details of interest for the object of
the contract: they cannot include every likely (and less likely) circumstance in the present
and in the relevant future that is of interest.
• Even if this could be possible, the parties are not able (or it would be too costly) to
negotiate every one of these single details.
Pillar of bounded rationalityà Contracts are incomplete: the parties of a contract have and can
exert a certain degree of discretion over the fulfillment of contractual clauses.

Opportunism
(Some) Economic agents:
Pursue selfishly their own utility and interest even if this is detrimental to the utility of interest of
someone else.
§ Not all economic agents are opportunists, but it is enough that some of them
behave opportunistically in order to incur in additional costs of transaction.
§ Ex-ante it is not possible to distinguish between opportunistic and not opportunistic
economic agents
§ Opportunistic behaviors can appear during the writing of the contract and also after
(Post-Contractual Opportunism).
46
Relationship-Specific Investments
Investments that are specific to the relationship between the parties and to the nature of the
counterparts
Examples (among the myriad possible).
- A supplier working for FIAT Chrysler specializes itself in the production of a car-component
that can be sold to FIAT Chrysler only.
- A supplier locates the plant close to one single specific customer firm in order to facilitate
just-in-time processes, but there are no other firms in the surroundings to whom the good
can be sold.
- A firm acquires a software and trains its employees for its usage, while this type software is
sold, and its usage taught only by a specific software house.
- I am invited to hold lectures in the small island of Neverland, if I accept, I have to translate
all my slides and learn how to speak Neverlandish.
- ….
The 3 more important types of relationship-specific investments:
§ Dedicated (often physical) asset specificity (supplier-buyer relationship along the
supply chain) refers to assets whose (physical or engineering) properties are
specifically tailored to a particular transaction. Fiat example of before (supplier has
to invest in machines who are specific for FIAT cars)
Glass container production requires molds that are custom tailored to particular
container shapes and glass-making machines.
Ports investing in assets to meet the special needs of some customers
§ Site specificity (high transportation costs) refers to assets that are located side-by-
side to economize on transportation or inventory costs or to take advantage of
processing efficiencies. Cement factories are usually located near lime stone
deposits. Can-producing plants are located near can-filling plants
Steel manufacturing: side-by-side location of blast furnaces, steelmaking furnaces,
casting units, and mills saves fuel costs (no need for re-heating)
§ Human asset specificity. Some of the employees of the firms engaged in the
transaction may have to acquire relationship-specific skills, know-how and
information. Clerical workers acquire the skills to use a particular enterprise
resource planning software. Salespersons have to acquire detailed knowledge of
customer firm’s internal organization in order to sell.

Relationship-Specific Investments
The more asymmetric these areas are between parties, the more the firm facing low relationship-
specific investment may “exploit” the firm facing high relationship-specific investment. There
could be different degree of relationship-specific investment. Given a full amount of investment, a
part of this investment is redeployable (general) and is large the more a firm would be reluctant to
exit from the relationship (once entered). Redeployable in case of exit from the relationship.
The relation-specific part of the investment is not recognized outside the relationship (non-
redeployable inv.) Lost in case of exit.

The combination of these three pillars leads to the famous HOLDUP problem. Example on rent,
quasi-rent and the holdup problem

47
 • Suppose your company contemplates building a factory to produce cup holders for Ford
automobiles. The factory can make up to 1 million holders per year at an average variable
cost of C dollars per unit.
• The construction of your factory is financed with a mortgage from a bank that requires an
annual payment of I dollars. The loan payment of I dollars thus represents your (annualized)
cost of investment in this plant. I is an unavoidable cost: You have to make your payment
even if you do not do business with Ford.
• You will design and build the factory specifically to produce cup holders for Ford. Total cost
of making 1 million cup holders is thus I + 1,000,000C dollars per year.
Of course, your expectation is that Ford will purchase your holders at a profitable price (you will
make non negative extra-profits out of this business).
• But suppose that you also want to consider outside options (i.e. Market). The “market
price” you can expect to get from selling your cup holders is Pm.
Suppose that:
• Pm > C.
• Variable profits (without considering fixed cost I): 1,000,000(Pm - C) >0
• But that the annual investment cost I > 1,000,000(Pm - C).àWhich implications?
1) Producing and selling to FORD is the best option
2) Since I is an unavoidable cost (has to be repaid) once in the market you would still keep
producing and selling cup holders even if the relationship with FORD breaks down: selling
to the market is better than not producing at all.
For example, if I = $8,500,000, C = $3, and Pm = $4, then the RSI is $8,500,000- 1,000,000(4 - 3) =
$7,500,000. Of your $8,500,000 investment cost, you lose $7,500,000 if you do not do business
with Ford and sell to the market

Now suppose that before you take out the loan to invest in the cup holder plant, Ford agreed to
buy 1 million sets of cup holders per year at a price of P* per unit.
You will accept these terms as long as extra-profits are non-negative:
1,000,000(P* - C ) – I ≥ 0 which implies that P* > Pm
This is what this stream of literature refers to as “Rent”: the rent is simply the profit a company
expect to get from a contractual relationship if everything goes as planned.
But is Ford willing to exploit after the contract the existence of that red area? And if yes, to what
extent?
To understand if and to what extent (and so the magnitude of the hold up problem) one has to
define the concept of quasi-rent

Quasi-rent is the extra profit that one

gets if the deal goes ahead as planned,
versus the profit one would get if he had
to turn to his next-best alternative (in
our case Market)
Suppose your company stipulate the contract with Ford with P* = $12 per unit, and like before Pm
= $4 per unit, C = $3 per unit, and I = $8,500,000.
At the original expected price of $12 per unit, Rent is: (12 - 3)1,000,000 -8,500,000 = $500,000 per
year. The contract leads you a positive profit, so you think “let’s do it”. But……
After contract signed, your quasi-rent is (12 - 4)1,000,000 = $8,000,000 per year.
Ford wants that quasi-rent.
Ford would like (if able) to renegotiate the contract down to, let say, Pnew* = $8.
48
 Ford would increase its profit by $4,000,000 (in reduction of costs)
You would still want to furnish Ford rather than the market
If you break up the relationship with Ford your profit is:
[1,000,000(Pm - C ) - I ] = [1,000,000(4 - 3) - $8,500,000]= -$7,500,000.
If you stay in the relationship with Ford (even with deteriorated conditions) profit is:
[1,000,000(Pnew* - C ) - I ] = [1,000,000(8 - 3) - $8,500,000]= -$3,500,000
-$3,500,000 > -$7,500,000 à STAY

Of course, Ford, if it could, would not stop at Pnew* = 8$ but it will fix the minimum possible price
given that you are still willing to serve Ford rather than the market
In other words, Ford, if had the possibility to do that, would fix a new price such that
Pnew* = Pm + ε = 4 + ε where ε is sufficiently low.
The fact that Ford is willing to behave in this way, i.e. “opportunistically”, has a great incentive to
do that, does not mean that it is capable to do it.
“We have signed a contract, with precise obligations from both parties…… we are safe………we are
sure that FORD can not behave in this way……………Are we really sure?”
I am afraid you can not be sure………
CONTRACTS ARE INCOMPLETE (and thus potentially ambiguous).
Ford could assert that, in one way or another, circumstances have changed and that it is justified
breaking the contract.
It might, for example, claim that competition for that particular model of cars you are producing
cup holders for is facing fierce competition, and if terms with suppliers do not change (enabling
FORD to settle lower prices for the final car) it will face bankruptcy risks.
Or it might claim that the quality of your cup holders fails to meet promised (potentially unwritten
or ambiguous) specifications and that it must be compensated for this lower quality with lower
prices.
You may want to consider the possibility to fight Ford in court for breach of contract.
1) This is itself a potentially expensive move
2) Contractual incompleteness leads almost by definition to ambiguity in contractual terms so
the outcome of the trial (even if you are in principle right) can not be taken for granted!!!!!
Actually, it can be highly uncertain.
Ford knows both 1) and 2) (that’s why it has “held you up”)

End of the story: Most of the time you are better off accepting Ford’s revised offer than not
accepting it (this of course also depends on the extent to which Ford wants to appropriate your
quasi rents, i.e. to what extent Pnew* is far or close to P*) .

49
This example makes evident how your market transaction with Ford was made problematic and
inefficient by the presence of:
- Specific-relationship investment you have to make with Ford
- Bounded rationality that leads to an incomplete contract between you and Ford
- Opportunistic behavior by Ford

• Bounded rationality (No?: all complete contracts)

• Opportunism (No?: no need of complete contracts)
• Relationship-Specific Investments (No?: no need to defend from opportunism, because
there is no loss in switching from a transaction to others)

Thus, if only one of these pillars is not met, the market will still work efficiently.
But if all pillars are working and you know that you will likely encounter all these problems in the
relationship with Ford……………………
Would you ever be willing to enter into a market relationship at these conditions with Ford?
And if anybody is not willing to do that who would provide cup holders to Ford?
One possibility is…………………..Ford itself (hierarchy)
Given that markets are not always perfect and their use may entail substantial costs the main aim
of the TC perspective is to individuate for each transaction the best governance structure possible.

3 alternatives:
Market which may present costs (related to the use of contracts)
Hierarchy (transactions are internalized and carried out inside the firm, costs arise from the
administrative bureaucracy that internalize and monitor the exchanges)
Hybrid forms
The most efficient solution (the market vs. planned solutions) is identified by a Darwinian
selection where the best option finally survives. Note that from this perspective, firm is seen as an
efficient response to a market failure
The famous Fisher Body-GM case study is emblematic in this respect.
2 determinants:
1) Relation-specific investment needed for carrying
out the transactions
2) Frequency of transactions
An increase in uncertainty of transactions (which may
affect the degree of incompleteness of contracts) is
likely to lead to less hybrid forms

Important from a managerial perspective:

- be able to spot areas of “contractual incompleteness”
- organize transactions accordingly, to limit the negative backlashes of opportunism

CASE OF STUDY: In 1980 three Finnish companies Nokia, Aspo, and Salora established a research
joint venture called Micronas in order to develop semiconductors. Underlying the project was the
transition in the semiconductor industry. Until the late 70s semiconductors were rather
homogenous products that were available from numerous manufacturers. In the late 70s it
became evident that in the future an increasing amount of semiconductors will be application-
specific.

50
The initial technology was bought and transferred from an American company, Micro Power
Systems Inc.(MPSI), which was selected amongst nearly twenty companies. The contract with
MPSI was detailed and complicated.
The transaction was twofold from the perspective of Micronas (Nokia):
1. transfer the relevant technological information as well as
2. to provide training in semiconductor design, manufacture (and plant construction).
Please, after reading the excerpt from the original contract (next slide), try to:
(a) Say if the contract is subject to contractual incompleteness, and why (yes or no)?
(b) Organize the training transaction in the most effective way, by paying specific attention to
the timing of payment.

10. Innovation and technological change: Market structure and innovation

Perfectly competitive market is the best option speaking about social welfare and in the long
period the equilibrium is achieved as p=MC=ACmin. The maximum quantity of the good produced
at the lowest possible cost= Max Productive efficiency (goods and services are produced using the
least cost combination of resources and technology)The maximum quantity of the good sold at the
lowest price= Max Allocative efficiency (resources are dedicated to the combination of goods and
services that best satisfy consumer needs. These factors are static, but things are more dynamic in
the real world.
Dynamic efficiency means the capacity of a market system to make improvement and generation
over time of new products and production techniques. It implies that the choices that we make
today can change the future.

“INNOVATION has become the industrial religion of the late 20th century. Business sees it as the
key to increasing profits and market share. Governments automatically reach for it when trying to
fix the economy.” The Economist, February 20th, 1999, Survey of Innovation in Industry

The Lisbona Strategy (2000) “To become the most dynamic and competitive knowledge-based
economy in the world”
Europe 2020 (2010) “A European strategy for smart, sustainable and inclusive growth”
ü Research and innovation are placed at the centre of the Europe 2020 strategy (Horizon
2020). This includes the headline objective of increasing spending on R&D to 3% of GDP by
2020. Innovation become an important topic. This change is due also to some economics:

- The “Solow residual” Solow (1957, RES) shows that for US between 1909-1947 technical
change accounted for 87.5% of economic growth. He demonstrates that the growth is
based on these factors

- The legacy of Schumpeter. He was the first one to identify the difference between
invention vs. Innovation vs. diffusion
• invention (telephone):
• (technological) materialization of an intuition (often) engendered by scientific
knowledge.
• Might be exogenous and largely random in nature; it could not be influenced by
market signals. Can be a process of “stumbling” over. Serendipity.
51
 • innovation (telecommunication service):
• transformation of an invention into a new product/service/process;
• it addresses new needs (often latent needs) on the part of consumers or firms;
influenced by economic signals (like pursuit of profit).
• diffusion (telecom penetration) spread process of an innovation into the economy. The
diffusion curve is like a logistic one. In the first phase the penetration rate is low due to
a lack of incentives and information. A firm will adopt the innovation if benefits of
adoption of the innovation are greater than costs of adoption.

A Taxonomy of Innovations
Product versus Process Innovations
• Product Innovations refer to the creation of new goods and new services, e.g., DVD’s and
cell phones
• Process Innovations refer to the development of new technologies for producing goods or
new ways of delivering services, e.g., robotics and CAD/CAM technology.
Radical (disruptive) versus incremental Innovations (often PRODUCT)
• Technological breakthrough that might create new markets
• Small adjustments over existing products
Drastic versus Non-Drastic Innovations (PROCESS)
• Drastic innovations have such great cost savings that they permit the innovator to price as
an unconstrained monopolist
• Non-drastic innovations give the innovator a cost advantage but not unconstrained
monopoly power

Drastic versus Non-Drastic Innovations

• Suppose that demand is given by: P = 120 – Q and all firms have constant marginal cost of
c = $80
• Let one firm have innovation that lowers cost to cM = $20
• This is a Drastic innovation. Why?
§ Marginal Revenue curve for monopolist is: MR = 120 – 2Q
§ If cM = $20, optimal monopoly output is: QM = 50 and PM = $70
§ Innovator can charge optimal monopoly price ($70) and still undercut rivals whose
unit cost is $80
• If cost fell only to $60, innovation is Non-drastic
§ Marginal Revenue curve again is: MR = 120 = 2Q
§ Optimal Monopoly output and price: QM = 30; PM = $90
§ However, innovator cannot charge $90 because rivals have unit cost of $80 and
could under price it
Innovator cannot act as an unconstrained monopolist:
§ Best innovator can do is to set price just under $80 and supply all 40 units
demanded.

What Schumpeter is most known for are his two different vision
- The paleo-schumpeterian entrepreneur: The theory of economic development (1912)
Individual capable of transform an invention into an innovation through an entrepreneurial
act. Incentive: more than just economics but economics play a role: expectation of supra
normal profit arising from being the first in the market. Absence of extra-profits as a pay-
off of innovative activity = low innovation rate. The extra-profit generated by an innovation
52
 stimulate imitation; over time the entrepreneurial rent of the innovator is eroded by
competition from imitators. The main actor of innovation is the entrepreneur due to the
ambition of success and the willingness to create.
In order to stimulate innovation, there should not be instantaneous imitation (Patents and
IPR protections). The process of “creative destruction”: competition through innovation
results in high level of dynamic efficiency à INNOVATION MAINLY FROM NEW FIRMS
- The neo-Schumpeterian vision: Capitalism, socialism and democracy (1942)
High innovative performance of the capitalist system characterized by the presence of
large oligopolistic companies. Key role of large scale techno-structures (R&D labs) in the
innovative process: R&D benefits from economies of scale. Capital market imperfections
for new firmsàINNOVATION MAINLY FROM BIG COMPANIES (IN CONCENTRATED
MARKETS)

1) Which market structure is more able to ensure innovation? Arrow (1962) vs. Schumpeter
(1942): Monopoly provides less incentive to innovate than competitive industry because of
the “Replacement Effect”. Competition is the most preferred market form to innovate. But
there is a sort of cannibalization. Arrow considered the possibility of new incentives (profit
added) while Shumpeter considered the capability (profit). An intermediate level is
preferred à oligopoly is the preferred market shape for innovation and dynamic
efficiency.
a) Views are not necessarily at odds: incentives to invest in R&D are different from
capabilities to do so.
b) Mixed evidence:
- inverted U-relationship between intensity of competition and innovation
- radical innovation introduced by new firms (SM1)
- incremental innovation more by large companies (SM2)
More often radical innovation are introduced by start-up while incumbents are more keen
on to introduce incremental innovation.

2) To what extent “innovation should affect market structure”? How much we want
innovation to affect market structure?
Appropriability: ability of the innovator to capture the benefits engendered by his
innovation to the detriment of other firms. Appropriability instruments :
§ Obtained exogenously through regulation of property rights: to the innovator is
assigned the right to exclude others to make unauthorized use of the innovation
(temporary monopoly).
§ Obtained endogenously by an innovative firm through the establishment of
strategic barriers to imitation.
Practical problems of patents:
§ You immediately diffuse of valuable information (spillovers) relating to the
technology in the moment you apply for a patent, even if it is confidential you need
to involve more people
§ When it became public, it can be easy to be coming out to something that is
completely legal, but it is clearly an imitation: inventing around;
§ It may lack of information in the paper and it could lead to difficult enforcement;
Main alternative strategic endogenous mechanisms to assure appropriability:
§ secrecy (e.g. Coca Cola formula), keep the secret about technical details;

53
 § lead time (Intel microprocessor), always ahead your competitors, when someone is
able to replicate your innovation, you will be able to launch a new innovation;
§ complementary investments in brand, sales & distribution and customer care
(Nespresso).

The innovation should be novel to be patented, should have an industrial applicability, (and other
two ) it has to be eligible (you cannot patent some field, a medical protocol or a mathematical), it
should be an advancement (not obvious)

Historically, 2 different views on patent lenght, breadth and enforcement:

- Incentives View(Shumpeter Mark 1). Patent are really necessary to have
- Openness View. They could be important in some sectors but in other not. It is more
important to diffuse it to heve an higher innovation.
This debate has been historically solved much more in favour of the incentives view rather than
the openness view. But the openness view is gaining momentum.

The first side effect is that number of patents are slightly increase.

The second one is that we pretty know that firms think that patents are not always effectively to
protect innovation. Cohen, Nelson and Walsh (2000), Protecting their intellectual assets:
Appropriability conditions and why US manufacturing firms patent (or not), NBER working paper,
n. 7752.- 1478 US manufacturing firms with R&D expenses. Firms were asked to report the
percentage of their product and process innovations for which each appropriability mechanism
had been effective in protecting the “firm’s competitive advantage from those innovations” during
the prior three years. Informal mechanisms were more used.

PATENT might not be used to defend own innovation but rather for preventing others from
innovate. In many sectors (like for example ICT) technologies are becoming more and more
complex: the number of patents on the same product is increasingly growing
Eg. 3G mobile phone (n° of patents to protect the standard is 7,796!!!!): PATENT THICKET
The third side effect is : Complexity of innovation à Almost impossible that a single firm owns all
IPRs on these different technological parts that uses (or wants to use).
Thus, having a patent on a specific and possibly key component of a technology might lead to:
a) block others’ innovation (“blocking patents”, Heller and Eisenberg 1998, Science).
b) Give more bargaining power in a licensing transaction
à Jaffe e Lerner (2004): “Rembrandts in the Attic” something valuable that you do not use
but you can sell it to someone else
Tragedy of anti-commons à ownership is too much fragmented

The last effect is the weak patent: the number of application for officer increase and it is possible
to make errors. Even a weak patent can be used to prevent others to innovate.

“[…] My own introduction to the realities of the patent system came in the 1980s, when my client,
Sun Microsystems--then a small company--was accused by IBM of patent infringement.
Threatening a massive lawsuit, IBM demanded a meeting to present its claims. Fourteen IBM
lawyers and their assistants, all clad in the requisite dark blue suits, crowded into the largest
conference room Sun had.
54
The chief blue suit orchestrated the presentation of the seven patents IBM claimed were
infringed, the most prominent of which was IBM's notorious "fat lines" patent: To turn a thin line
on a computer screen into a broad line, you go up and down an equal distance from the ends of
the thin line and then connect the four points. You probably learned this technique for turning a
line into a rectangle in seventh-grade geometry, and, doubtless, you believe it was devised by
Euclid or some such 3,000-year-old thinker. Not according to the examiners of the USPTO, who
awarded IBM a patent on the process.
After IBM's presentation, our turn came. As the Big Blue crew looked on (without a flicker of
emotion), my colleagues--all of whom had both engineering and law degrees--took to the
whiteboard with markers, methodically illustrating, dissecting, and demolishing IBM's claims. We
used phrases like: "You must be kidding," and "You ought to be ashamed." But the IBM team
showed no emotion, save outright indifference. Confidently, we proclaimed our conclusion: Only
one of the seven IBM patents would be deemed valid by a court, and no rational court would find
that Sun's technology infringed even that one.

An awkward silence ensued. The blue suits did not even confer among themselves. They just sat
there, stonelike. Finally, the chief suit responded. "OK," he said, "maybe you don't infringe these
seven patents. But we have 10,000 U.S. patents. Do you really want us to go back to Armonk [IBM
headquarters in New York] and find seven patents you do infringe? Or do you want to make this
easy and just pay us $20 million?"

After a modest bit of negotiation, Sun cut IBM a check, and the blue suits went to the next
company on their hit list.”

All these arguments have reinforced the “openness view” and also led to radical positions (e.g. see
Boldrin and Levine pamphlet (2008) Against Intellectual Monopoly).
“If national patent laws did not exist, it would be difficult to make a conclusive case for
introducing them; but the fact that they do exist shifts the burden of proof and it is equally difficult
to make a really conclusive case for
abolishing them.”

Innovation in network externalities

Network economics: the hardware/software paradigm, winner-takes-all-markets and
technological standard wars.

The concept of network market is broader than just the informatic technology one. A specific
feature of the network market is that the winner takes all after a bit of competition. For instance,
Android for smartphone obtain the 99% of market share. For auction sites, eBay is the winner
because its performance was higher than other sites but also because the it exploits the network
effect. The structure of the market lead to some performance of the market itself. We do not
often observe this dynamic; these results were determined thanks to the structure of the market.
The situation doesn’t change very rapidly, eBay is the persistent leader of the market, not only for
a while. Only Amazon in the years takes its place. eBay had a high position in a niche market and
then it started to grow. At a certain point eBay really took the lead and the market was dominated
by one firm only.
55
An introductive example: The story of the garbage bin. In the town there are six guys that like
playing videogames, Alan, Bud, Charlie, David, Eliah, and Frank.
Their utility function is characterized by externalities: their preferences depend on the intrinsic
value of the game console plus a bonus that depends on the number of other guys owning the
same type of console (why? because they can exchange games, challenge friends, and have a
greater variety of games) à Network externality
• Ui = Xk + w*Nk
• Xk = intrinsic value,
• Nk = number of consumers owning type k console,
• w = parameter (let us say, w=0.2) higher the value, higher the network externalities
• k = a, b, c (3 different consoles)
Each firm invested in R&D for its own console and managed to patent some product features. Each
firm also invested in marketing and advertising à standard war
Yesterday, the guys received the catalogues, illustrating technical features
Therefore, tonight the guys will read the catalogues, assign their preferences (intrinsic value Xk)
and make their consumption decisions.
They will assign different preferences to different consoles, since they value
different characteristics.
When they fall asleep, their structure of preferences is:
N.B.: this is only Xk (intrinsic value of the console), since there is no externality
until the first customer purchases the first console

What is market structure (potential) on the Sunday night when they fall asleep?
33% for each firm
Charlie’s cat is very hungry on the Monday morning and happens to turn the
garbage bin upside down. Charlie wakes up suddenly, it is 8 am. The other guys
in town are still sleeping.
Charlie thinks “since I am awake, I might go to the games shop and buy my
console”. He goes and his purchase decision is obviously to buy C.
From the very moment Charlie makes his buy, the utility of other customers is
changed. The second customer is very likely to buy C.
If he does so, the third will CERTAINLY buy C and the market is locked in.
à C wins and takes ALL the pie error: Frank = 0.8
FURTHER CONSIDERATION: Of course, there are missing elements in the story,
but you can already infer the role of past and casual events on today situation: how can we define
business processes where a cat and a garbage bin are so important for determining the final
outcome of the market? If the initial move would have been different, things could be completely
different, leading to completely different outcomes. Path dependent, history matter, non-ergoolic.
What happened at the beginning really determine where we are now.

COOPETITION: firms cooperate before the products is launched on the market. Then, they start to
compete.

Network markets
Definition of network externality: a good exhibit a network externality when the positive change in
the utility a consumer derives from it raises as the number of consumers that purchase the same
product increases. Examples: telephone, email, hardware-software, party, etc.
56
Difference between direct vs. indirect network externality:
Direct externality: the value of the good increases automatically as the number of users increases
(examples. phone or e-mail). Generally, the good ha no intrinsic value.
Indirect externality: the value of the good increases as the number of users increases but only in
the presence of economies of scale in production (of software) because of a greater offer of
complementary products (or a better quality) of the good (admitting that consumers care for
software variety). In this case, the good has often an intrinsic value (independently of network
externalities).

Distinction between a “two-way” and a “one-way” network (Economides, 1996, IJIO)

Dir ect. 2-way network: all the networks where service AB is different from the service
BA (they represent two different goods). Each knot of the network represents a user.
The value of the network is a function of the active links inside the network.
In a 2-way network composed by n knots, there are n(n-1) potential links. The entry of
a new user produces a positive externality on existing users, since she adds 2n new potential links.
Metcalfe’s Law: if a network is composed by n users and each user
assigns a value to the network which is proportional to the number of
users
à the value of the network V = f(n2-n)
à If n is large à V = f(n2) the value is approximated

The value of the network is a function of the users

Relationship between number of users and value of the

network implied by the law. It is posed by a convex, but it leads to estimate the value of
the network.

True relationship is a concave relationship. Metcalfe law could overestimate the value of
a network. It

Indirect. 1-way network: When one of AB or BA is unfeasible, or does not make economic
sense, or when there is no sense of direction in the network so that AB and BA are identical,
then the network is called a one-way network. hardware-software paradigm: the greater the
usage of a hardware, the greater its attractiveness in terms of developing “software”, an
increase in the num ber of software programs further raises the attractiveness of the
hardware and increases the number of new adopters and so on (“bandwagon effect” or “positive
feedback”). Ex. Netflix à Hardware: platform; software: tv series

The same logic applies to the IBM –Intel- Microsoft vs. APPLE case. IBM was dominating the
computer industry until the middle of the 80s. Computers were not available for everyone, then in
the 60s there were the era of minicomputer but still for very specific tasks. IBM didn’t believe for
the commercial appeal of having a personal computer for each individual. A complete
misjudgment. It allowed startups to enter in the market and commercialized computers. IBM
understood the potential of personal computers when apple create the Apple II. IBM was forced
to make a collaboration to other actors.
57
Three remarks:
1) Bandwagon vs. negative feedback
If someone is experiencing a positive feedback and has competitors, these latter are probably
experiencing a negative feedback
The higher is the number of users of a platform, the higher is the incentive for the developers of
complementary goods (sw) to create contents for such platform, the higher is the attractiveness
of the platform, the higher is the number of users that use the platform (bandwagon effect or
positive network feedback) and so on…
The lower is the number of users of a platform, the lower is the incentive for the developers of
complementary goods (sw) to create contents for such platform, the lower is the attractiveness
of the platform, the lower is the number of users that use the platform
(negative network feedback) and so on…

2) These dynamics also affect “Two-sided markets”

Rochet-Tirole (2006, RAND): “markets where one enable interactions
between end-users, and try to get the two sides “on board”
- cross-side externality àalmost always
- same-side negative externality à seller want to be by themself
and higher is the number of consumer, higher in the negativity (not so
frequent)
Examples: on-line auctions, credit cards, crowdfunding, eBay,
Amazon

3) EMPIRICAL ESTIMATES OF NETWORK EXTERNALITIES

Large number of studies in the economics-managerial literature that document the presence of
network externalities:
• Indirect. Software (Dranove and Gandal, 1999; Gandal et al., 1999), ATMs (Saloner and
Shepard, 1995), CD players (Gandal et al., 2000), Yellow Pages (Rysman, 2002), Videogames
(Clements and Ohashi 2005), Electrical Vehicles (Li et al. 2017).
• Direct Telecommunications (Majumdar and Venkataraman 1998)

How to estimate them?

From Koski and Kretschmer (2008, JICT):
The complementary goods approach (e.g. Gandal et al. 2000) derives a system of equations and
uses the number of software available as a variable in hardware adoption regressions and vice
versa.
[In the hedonic approach (e.g.Gandal, 1994, Sarnikar, 2002), the installed base is treated as a
product characteristic that will have a positive effect on prices if there are network effects].
[In the adoption approach (e.g. Koski, 1999, Gandal et al., 1999), the installed base at t-1 carries a
positive expected sign in the adoption or diffusion equation at t.]
[The timing approach (e.g. Saloner and Shepard, 1995) establishes that firms with higher expected
network effects will adopt a technology earlier and proxies expected network benefits by the
number of potential (internal) users of the technology]

58
Innovation in a network market: the start-up problem
The demand for a network good
Respect to the standard demand curve, the demand for a network good does not only depend on
price but also on the agents’ expectations about total consumers of the good in order to
understand if people are going to use the network or not. Have to care about what others are
doing.
à Some basic (and key) implications

- Everybody in the economy is convinced that no body will buy the network good: the
network good will be effectively unsold (very pessimistic expectations) irrespective of a
high or low price: “network failure”. Nobody is bought it because everybody think that.
- Consumers want to buy a hardware if many softwares are available for that hardware. But
software developers will write softwares only if a great number of people has already
bought the hardware since they are potential buyers: if nobody makes the first move the
network good will be effectively unsold (again irrespective of a high or low price): “network
failure” à Chicken-egg paradox equilibria. Consumer want to buy hardware if a lot of
software are available. Netflix (platform) with a lots of series tv (software). But who write
software want to write it only if there are enough buyers.
Incentives can solve this problem and vertical integration. Hardware producer own at least
part of the software (Netflix produce film also itself).
(note how these markets are characterized to a great extent by “self-fulfilling prophecies”,
the fact that we all think something, make this thing happened).
Network failure is really a possible outcome and expectation are important.
- The aggregate demand could be upward sloping (Rolfhs 1974, BJE): If a firm
is able to influence the expectations of potential consumers about the
penetration that the good will have (and so its value in the future), is totally
plausible that as the expected size of the network increases, consumers are
willling to pay a higher price for joining the network.
People could like the good in itself or not like it at all.
- Demand presents a critical mass effect (Rolfhs 1974, BJE): Critical mass in nuclear
engineering. Under radioactive decay conditions, the amount of uranium necessary to
start a self-sustaining process of production of neutrons that maintains unchanged its
quantity. Any larger amount of uranium will cause an explosive nuclear chain reaction. Any
smaller amount of uranium will cause nothing, and it will soon decompose.
Critical mass can be related to sociology. Schelling brought the concept to describe the
social phenomena. There can be 0, a big explosion and the different between the two is by
reaching a critical mass. Example of social phenomena that do not reach a critical mass is
the failed applause.
Suppose a University Professor is particularly nasty and decide to schedule a series of
seminars on Saturday afternoon. Attendance by students is not strictly mandatory, but of
course «being there» (or not being there) may favorably (or negatively) impress the
Professor.
Knowing all that, each student will decide whether to participate or not on the basis of a
minimum threshold of attendance to these seminars by the class. Students’ expectations
are adaptive over time (i.e. what students expect in terms of seminars’ participation in the
present week will strictly depend on the actual attendance registered in the previous
week). Needless to say, there is no possibilities for coordination among students.

59
 The class is composed by 10 students. Students are heterogeneous and they have different
minimum thresholds.
More specifically, let us assume that these thresholds exhibit a distribution close to the
normal Gaussian distribution. In particular, for 1 student the minimum threshold is equal
to 2 students; 1 student has 3 as minimum threshold; 2 students have 4; 2 students have 5;
3 students have 6; and lastly, 1 student has a minimum threshold of 8.

Analysis. 4 students represents the critical mass for this cycle of seminars: 4 students are willing to
participate only if the participants to the seminars amount to 4 students.
This is an equilibrium. But is it stable? Let’s try to perturb it.
Let us suppose that at the first seminar instead of 4, 5 is the number of attendees. The following
week (as you can see from the graphs of the previous slides), the number of students who will
attend the seminars will be equal to 6; the Saturday of the 3° week, students will become 9, and at
the Saturday of the 4° week, every student will attend the seminar.
Let us suppose that instead of 4, 3 is the number of attendees. The following week (as you can see
from the graphs of the previous slides), the number of students who will attend the seminars will
be equal to 2; on the Saturday of the 3° week only one student will attend, starting from the
Saturday of the 4° week, every student will not attend the seminar. The tipping point between
success and failure is represented by the achievement of the critical mass.

à Critical mass in network economics

For any given price charged by the firm(s) it is the minimal amount of consumers which join the
network and are satisfied of this choice. Any larger amount will trigger the bandwagon effect, any
smaller amount will bring to a network failure. Firm can impose the critical mass setting the price.
STYLIZED EXAMPLE in network economics. Firm A has invented and patented a network good
(exhibiting direct externalities) and is going to commercialize it under a monopolistic regime.
Trade-off. High price and a high number of people to convince. Low price and a small number of
people to convince. Suppose:
• p = 1€; at that price a representative agent may be willing to buy the good if it can be
enable to communicate with (say) 100 people. Lower number.
• p = 100 €; at that price one may be willing to buy the good if it can be enable to
communicate with (say) 1.000.000 people.ù

If the choice of the firm is p = 1, the firm has to convince 100 people
If the choice of the firm is p =100, it has to convince 1.000.000 people
Whatever the choice, if the firm wants to sell the good it has to attract the critical mass (which is
increasing in price).
If p =1€ but the firm convinces only 99 individuals, someone of these individuals will be unhappy
(someone who gives to the good a value of 1€ only if 100 individuals had joined the network); so
he will leave; the network size will shrink to 98 individuals, again someone of these individuals……
If p =1€ but the firm convinces more than 100 individuals, e.g. 101, the value of the network good
raises, and some more agents will want to buy the network good, network size becomes 102, so
the value of the good raises…….
à AIM: Attract Critical Mass so to innescate the BANDWAGON EFFECT

Aggregate demand for a network good: a stylized model (Rohlfs 1974, Bell J of Econ and Man; see
also Shapiro and Varian 1999)

60
Consider a network that is of interest for N potential agents. They are indexed by x which is
uniformly distributed on the interval [0,1]. For a very low X, the utility will be high also without a
lot of buyers. For someone that is not really interested (X tend to 0), the number expected has to
be high for convince him.
Each agent faces the binary decision of whether to buy the good or not. The good exhibits positive
network externalities. ne = number of people expected
U = (1 - x)n e - p If she buys; U = 0 If she doesn’t buy
Since a continuum of potential consumers exists in the interval [0,1], there will be therefore a
particular consumer, indexed by x*, such that she is indifferent between buying and not buying
the good. This consumer is found by: or

Hence all the consumers that have a higher willingness to pay for the service (x ≤ x*) will buy the
good, while all the agents that have a lower willingness to pay (x > x*) won’t.

Perfect foresight is assumed n = Nx* = n expected Remember self-fulfilling prophecies)

…and then the inverse aggregate demand curve for a ntwork good becomes

2 stable equilibria:
a) Network of zero size (“network failure” caused by pessimistic expectations)
b) Network of large size (CRITICAL MASS is reached and positive feedback is triggered)

Critical mass: for any given price, it is the minimal amount of consumers which join the network
and are satisfied of this choice
Also the y axe is implied in the shape of the curve. It is feasible that a
network can be unsold. Low X is dependent to the number of buyers.
3 potential equilibria Xl, Xm and eventually 0.

“[…]It is plausible to assume that when people are willing to pay

more than the cost of the good, the size of the market expands and,
when they are willing to pay less, the market contracts. Geometrically
this is saying that when the demand curve is above the supply curve, the quantity goes up and,
when it is beneath the supply curve, the quantity goes down.”
Translating it int dynamic terms and putting time in the x-axis we will find:

X = % users (penetration rate)

1° phase: launch (0 – Xl);
2° phase: rapid growth (positive feedback, Xl -Xh);
When the Xl is reached the positive feedback start
to create a rapid growth
3° phase: maturity (universal service policies, Xh –
100%).

61
Comparison with the diffusion path of a «standard» innovation
Differences on average the start-up phase could be
more difficult than a standard good. You have to
coordinate the expectation. The growth stage because
of the presence of network externalities, the growth is
generally much more rapid. A shorter period that
generally occur at a standard innovation. In the
maturity, network goods reach higher penetration
rate. It is difficult to find someone that do not have a
telephone. This is not the same for the microowen.
The penetration rate is also called merit goods. You
need to have this good to not been marginalized by the
society.
Trying to extent the usage of these goods to the population by government intervention is called
Policy for universal service. These interventions are more ethical than economics
When there are positive consumption externalities
Minitel risk à From Koski and Kretschmer (2008): […] The success of Minitel, an electronic
directory and rudimentary e-commerce service based on a closed system40 introduced in 1983 in
France, was by and large based on this kind of strategy: the French government sponsored millions
of consumers giving them the Minitel terminal equipment free of charge. Apart from the subsidy,
this did also come at a different kind of cost: It has been suggested that the diffusion and use of
the Internet has been slower in France than in other European countries because of the popularity
of Minitel.” Minitel was considered as a substituted by the French customers. In fact they show a
very low interest for it in the short term. Subsidizing something that has a lower utility from that
point is called “Minitel risk”

Some examples:
1) Video-telephone: network failure in U.S. in ’70 and Italy in ‘80
Launched in the early 70s by AT&T in the Chicago area. Price not too high (86 $ flat monthly rate)
Service quality not too bad. Major failure causes:
a) Privacy
b) Rate of return regulation (cost-plus scheme)
- If something new is technically feasible does not necessarily mean that will be appreciated
by consumers (invention does not imply innovation)
- Low incentives to invest in marketing and related expenditure to reach the critical mass.

2) Fax
-1843 invented by Bain ; 1865 commercial fax service between Paris and Lyons
-1925 AT&T commercialize the fax in U.S.; 1982/1990 rapid growth (c.a. 80% penetration rate
among offices)
Very difficult launch phase, then a very rapid growth stage.

3) Telephone (invented in 1870s)

US President Rutherford Hayes «It's a great invention, but who would ever want to use one?».
One to one Vs. Network nature
Exploitation of patents in the short-term, i.e. high prices. Rent of the equipment: $14/month
(equivalent to $330 in 2000s) vs. production cost: $4 [See Rohlfs 2003]
3) Whatsapp 4) VCR and CD (U.S.)
62
How to solve the start-up problem?
STRATEGIC POLICIES
● Compatibility with competitors: share the effort but be careful since interlinking can
generate a free-riding problem. Should ease the achievement of the critical mass
● Promotional prices (very often below costs). Critical mass is a function of the price.
● Advertising
● Freeware to “Trend setter”, tolerate piracy (H3G, Microsoft Excel vs. Lotus 1-2-3)
● Invoke public aid (Minitel Francia ’80, Broadband Italy ’00, decoders for digital tv in ‘05)

Some exception of a winner takes all:

- There are markets where externalities are geographically limited. bunded. Food delivery:
one company rather than another locally. We see a collection of local monopolies.
- Community of interest. In this case I do not care about the size of the network. I only care
that few individuals that I know are using this network.

Strategic choice between compatibility and incompatibility

First of all, let us define compatibility:
● 2 products are defined compatible if they “can work together”: the output of a given
brand can be used by other brands. If this is the case, we say that different brands adopt
the same standard. In the opposite case, we say that products are incompatible.
● There is downward compatibility when a new release of a product is compatible with the
old one, but it is not the other way around (example Windows Office). Not very well
readable with an older version of the software. If not, there would be no incentives to
people to migrate to the new version of the product.
● A product is “one-way compatible” if it can work together with the output of a rival brand,
but it is not the other way around (example: Linux-Windows). Linux offer full compatibility
with windows but not vice versa. For Linux is a surviving strategy because windows is the
leader. The leader has no interest to offer compatibility.

In order to better understand the consequences of the strategic choice (comp. Vs non-comp.) let
us focus on the characteristics of a network market
The success of a standard will depend on its capacity to solve the start-up problem (i.e. attract the
critical mass) and trigger the bandwagon effect
If there is competition between firms (non-compatibility)
- Network Markets “naturally” tends to monopoly (winner takes all position)
- Choices of early adopters are fundamental and can determine the victory of a standard
against the other standards (so early stages are extermely important for firms). The initial
phases are crucial in determining the victory of a standard and the gain of other possible
standards
- First mover advantage. The time to market is of crucial important. It is easier to achieve the
critical mass if there are no competitors.

Formally (see Farrell and Saloner 1985, EL; Cabral IO)

2 incompatible goods: A e B.
Population (N =1) is formed by agents who prefer A (a) and agents who prefer B (b) with a+b=1.
Utility of a: v(x) if buy A and v(x) – α if buy B
Utility of b: v(x) if buy B and v(x) – α if buy A

63
Let us suppose that market is perfectly competitive so firms charge the same price, and also
suppose that being the only one to acquire the good will bring a zero benefit.
Monopoly will be an equilibrium if v(1) – α ≥ 0.
Now suppose a sequential entry of consumers into the market.
Consumers b will buy A only if v(a) – α > v(b) and Consumers a will buy B only if v(b) – α > v(a).

The achievement of a critical mass in a standard war

1) Early adopters are more

important (for reaching the critical
mass) the higher are the network
externalities (high v’) and the lower
is the love for variety (low α).
2) Same logics applies to the first
mover advantage

Remark: the model also makes self-evident the risk that an inferior technology could emerge as
winner of a standard war.

This is a real possibility from a theoretical point of view, but (luckily enough) on an empirical
ground is very difficult to affirm that such risk has ever become reality.
Katz and Shapiro (1986). Technology Adoption in the Presence of Network Externalities. Journal of
Political Economy, 94, 822-841.
The only (controversial) case is the one of Qwerty vs Dvorak. This case is very controversial.
QWERTY is the standard and still used but is not the one that perform better in term of comfort
and speed. Questionable whether there was path dependent inefficiency or not…less questionable
is the fact that the process has shown a higher degree of path dependency since we are adopting
this system invented 150 years ago.
QUESTION(S): WHY A KEYBOARD IS A NETWORK GOOD? AND OF WHAT TYPE?
Wo sided market (platform). Users that develop their skills on a type of keyboard. Producers of PC
know that most of the users are common using the qwerty keyboard.

2 forces may spur standardization

Firms have to bear high costs in R&D to develop a standard and in large measure they are sunk
because highly specific. If i choose non compatibility and i lose the market war, thw non
redeployable R&D investments are lost
Note also that marketing expenditure and coordination complementors costs can be extremely
high when a firm choose incompatibility and opt for a go-it-alone strategy
à So, it is not unusual that in network markets standards are defined through international
organizations (UL, ITU, NIST in the USA) or alliances between firms (“coopetition logic”). Two
different type of standardization: formal and informal.Or firms choose an “open standard” policy
(very low licensing fees for the patented technology they developed).
Es. Coopetition. Open Mobile Alliance [Both “hardware” (Mobile operators) and “software”
companies (Application & Content Providers) for standard setting in Mobile telecom. Some
64
members (+200): Ericsson, NTT Docomo, Vodafone, AT&T, China Telecommunications, Verizon,
Motorola, Telecom Italia, Microsoft, Intel, Softbank Mobile, etc. )].

The logic of open standard through a game-theoretic approach (Grindley 1995, Pepall et al.
2009)
a) Battle of the sexes game
2 Nash Equilibria
Result: Both firms will prefer to cooperate rather than fight each other

b) The rather fight than switch game

1 N. E
Result: Both firms prefer to fight rather than cooperate

The equilibrium of the first game Pareto dominates the equilibrium of the second game
The difference between the two games stems from the greed for profit of the firm that develops
the standard
Lesson: This is an example in which firms’ are better off if are cooperative or not to greedy. Firms
can switch from the second to the first game if the leading firm:
- Set very low licensing fees (policy of open standard)
- Reduce absorption costs of the follower (in knowledge-intensive sectors these may be large, e.g.
The Micronas case, in lecture on
Transaction costs). Increase the
probability.

More generally, the strategic choice

“compatibility vs non-compatibility”
implies the following trade-off:

Sponsor/defend: development of a standard and restriction of the use towards concurrent firms
(high licensing fees)
Give away: possibility for the concurrent firms to use the standard without restriction or with low
licensing fees
License in: use of a standard developed by
another firm
Clone: use of an open standard without
restriction

65
CASES ON STANDARD WARS
Arthur (1996), Harvard Business Review (“Increasing returns and the new world of business”)
It is casino gambling, where part of the game is to choose which games to play, as well as playing
them with skill. We can imagine the top figures in high tech—the Gateses of their industries—as
milling in a large casino. Over at this table, a game is starting called multimedia. Over at that one,
a game called Web services. In the corner is electronic banking. There are many such tables. You
sit at one.
How much to play? you ask. Three billion, the croupier replies. Who’ll be playing? We won’t know
until they show up. What are the rules? Those’ll emerge as the game unfolds. What are my odds
of winning? We can’t say.
Do you still want to play? High tech, pursued at this level, is not for the timid.
In fact, the art of playing the tables in the Casino of Technology is primarily a psychological one.
What counts to some degree—but only to some degree—is technical expertise, deep pockets, will,
and courage. Above all, the rewards go to the players who are first to make sense of the new
games looming out of the technological fog, to see their shape, to cognize them. Bill Gates is not
so much a wizard of technology as a wizard of precognition, of discerning the shape of the next
game.
This story highlights the connection between the psychological and the technical part of this
market. Psychology is so important because we know that expectation is a crucial part of the
success. More than often the most rational strategy for a firm, may be to do an irrational
commitment to win the race.
N. B. As said before, in the case firms choose NO COMP, the most rationale strategy “may be to
signal an irrational commitment to win the race (dello standard) at all costs. This is analogous to a
well-known result of oligopoly theory; namely, the most rational strategy for an oligopolist may be
to signal an irrational strategy commitment to punish price cutters-regardless of the ruinous cost
that it may incur by doing so.” from Rohlfs, 2001, p. 45.
For the network we can say the same thing. If you are keen and eager to put at risk the survival of
your company (all in), this may sound irrational, but it can turn out to be the most rational
because you can put good expectation in the market and this is very important.

Factors that can positively influence the standard war.

Size in terms of:
- Financial resources
- Commercial strength (marketing and distribution channels)
- Brand and reputation

The IBM-APPLE case testifies that these factors matter, the VHS-Betamax case are an example of
how a small firm can beat a big company
1) VHS VS Betamax
A small firm had success over the large firms due to the particular strategies that this small firm
put in place.
PREMISE
• 1970 Partnership between Sony, Matsushita and JVC for developing a prototype of VCR for
scientific purposes (U- Matic on a patent-sharing agreement). They develop together a
software for video recording events and scientific experiments. In the large consumer
market, it was not possible to sell because it was not consumer friendly. It had to be
converted to the customer market.
both firms were not very happy from the collaboration: Sony wanted to be the leader and
66
 were afraid that JVC wanted to steal its knowledge. Nevertheless, they adapt this
technology for the consumer market. Sony follow a specific technological trajectory while
JVC another one. Sony start designing a high quality in term of sound and image, also the
design was better. This standard enables the fruition of cassette of one hour long à worse
result. JVC had a lower performance in term of design and quality but the fruition was of 2
hour long.
At that time the video cassette recorder was thought for the consumer for the time shifting
à record video show to see later on. For this propose the duration was crucial: movies
typically last more than one hour.
IMPORTANT DATES
• 1975 Launch of Betamax (Sony) in the U.S.
• 1976 Launch of VHS in the U.S (JVC/Matsushita and other partners)
• 1977 launch of a new version of Betamax with the duration of two hour. But in the same
year JVC increase the duration at 4 hours. At this time, customers, wanted to record not
only movies but also events with a higher duration: more than 2 hours. The video recorder
became a network when Andre Blay send the request to Hollywood to record video that
could be seen from the video camera. Some title of movie of the century fox was sold
through tis system. The response of consumer was overwhelming
Another important impulsive was given by another entrepreneur, George Atkinson, that
started to rent also movies in this way.
they had to make a choice: on which format produce the video cassette. Since the
community of VHS was larger, they decide to implement this software. Then the
bandwagon started
• 1986 almost all cassette was produced by JVC
• 1988 Sony adopted standard VHS

Three important takeaways

- Difference between invention and innovation. The difficulty in transforming an invention
into an innovation.
- The importance of software. The choice of software producer that originate the advantage.
There was no-coordination between hardware and software, but the software emerges
naturally
- In case you choose open standard strategy, you may have more success, but you have to
face more competitive pressure. Panasonic (JVC), given its small size, opt for an open
standard strategy.

2) The “fake” war between DVD and DIVX: the use of vaporware
Technology that replaced the video cassette recorder. It was not really a war. It exemplifies the
expectations of the market.
Now we are in the mid ’90. Three companies: Sony, Panasonic and Toshiba develop a new
technology together to avoid another war (temporary alliance). They collaborate for the DVD and
it was launched in 1996. They also collaborate with the Hollywood studios as Columbia. Early
adopters really like the possibility to see the movie with a high quality. Some other studios (in
particular Disney) that had concern about piracy since this DVD was a digital standard. It could be
really easy to duplicate it. There were in the DVD some technologies to prevent it but Disney and
Universal was sceptical with it. The DVD was commercialized in the US heavily by one important
67
firm in terms of electronic: Best-Buy. The main competitor at that time of Best-Buy, was Circuit
City. It tried to make some agreement with those studios that were not totally satisfied by DVD. CC
makes also some agreement with some hardware producer to launch a different product respect
to DVD. Circuit City announced the DIVX and launch it in the market in the 1998. The product was
retired from the market the year late.
This DIVX also offer one-way compatibility and this was not a great and ambitious move. This
typology of strategy can continue to produce their movie in DVD: there is no reason to buy the
other product. It does not pose any limitations to the DVD producer and this enable to unsuccess.
It could be better to made an agreement with other DVDs studios.

From Dranove e Gandal (2003, JEMS)

The only variable that have a negative and a significant impact was the announcement of the
product. Vapor ware is a strategic announcement in order to change the expectation of consumer.
Takeaways: strengths and importance of the vapor ware thanks to expectations

3) Sony “vendetta”: THE BLU-RAY VS. HD-DVD STANDARDS WAR

All actors involved wanted to avoid a standard war, but this attempt failed because the two
proposed formats were too different. The two products were commercialized at the same time.
The price was comparable for the two options and very high. Sony was more able to attract
studios to the exclusive production this was a huge cost for Toshiba.
Despite tremendous efforts and investments by both sides to gain the upper hand, sales of both
formats had been sluggish by any measure.
The slow consumer adoption was generally attributed to their reluctance to upgrade while the
format war was until ongoing.
Uncertainty was also hurting the movie studios: in 2006 one analyst firm had projected that media
companies stood to lose $16 billion in revenue over seven years if a single standard failed to
emerge. In January 2008 Warner tipped the market towards Sony and blue ray had a more huge
community of users.
Takeaways:
- it matters a lot to Sony to combine the network of blue ray to the Sony PlayStation 3. It
was commercialized with it.
- You can induce software producer with money to come to your platform rather than
another

“[They should] understand that this is the last physical format there will ever be. Everything
is going to be streamed”- Bill Gates, Microsoft chairman
Switching costs & technology replacement
Which factors determine the migration of consumers from an old to a new network technology?
You will base your decision based on these two dimensions: Price of new technology and the
performance of new vs old technology (what the new technology enable you to do compared to
the old one).
Network markets add to these frameworks two dimensions:
Individual Switching costs. Costs that users suffer in the pass to the use of a good to another good
independently from others’ choices. The investment that the user has made to use the new
technology may be really specific to the old technology and in some way non-redeployable.

68
Expectations of consumers about other consumers’ behaviour (collective switching costs). Your
decision to migrate cannot be done in “isolation”. You would have an expectation of how many
others made the same choice (collective switching costs).
INDIVIDUAL + COLLECTIVE SW costs produce very strong lock-in effects

Individual switching costs

Learning (sunk costs given by non-fully redeployable product-specific investments)
[Search costs]; [Loyalty costs (psychological)]
Examples: Car (low sw. costs): driving license; ICT (high sw. costs): learning; compatibility

Collective switching costs

The value of the new technology will depend on how many users will switch (because of network
externalities). Thus, a new chicken-egg paradox to solve (which may prove to be more difficult
than an ex-novo situation, since users already satisfy to some extent their needs with the extant
technology).
But we are only talking about inertia (as we know, at the end, in the long run, the best technology
is likely to prevail). Society adopt new technology too slowly in terms of social welfare. Everyone
would benefit on this network technology, but these costs may create a problem in coordination
of the migration. If I am the only one, my utility get worst. Ex. Migration from the AM frequency of
the radio to FM. It was a slow process because consumers did not buy FM for the fear that other
consumers would not adopt it.

Technology replacement (The adoption game, Farrell and Saloner, 1985, Rand Journal of
Economics)
They put forward this simple game where there are 2 users that should decide if they want to shift
from the old platform to the new one.
The payoff of each user depends on the decision of the other user because of network
externalities
Since we consider network externalities:
𝛼>𝛿,𝛾
𝛽>𝛿,𝛾
Nash equilibria: (new, new); (old; old)
𝛽<𝛼 and (old, old) is the market equilibrium à EXCESS INERTIA
(strong)
𝛽>𝛼 and (new, new) is the market equilibrium à EXCESS MOMENTUM (strong)

Things might be even more blurred… (Cabral, 2010 and 2018)

B: benefits in moving to a new technology
A: benefits to stay in the old technology
C: costs
If (b1-c1) = (b2-c2) > 0 [same analysis as before, 2 NE,
potential Excess inertia, or Excess momentum, depending on actual values]
But (b1-c1) ≠ (b2-c2). Suppose (b1-c1) > a1 and (b2-c2) < a2.
For example, b1 = 5, b2 = 3, c1= c2 = 2, a1 = a2 = 2.5
2 NE as before (N;N) and (O;O) but now none Pareto
dominates the other
àExcess (M or I) is limited to segments of population

69
(Weak) Excess momentum in a sequential game
b1-c1 (slightly) > a1; b2-c2 (slightly) > 0 and < a2 à N;N

But (b1-c1) + (b2-c2) < a1+a2 [the gain of “1” type does not
compensate for the loss of “2” type in changing]

Final thought
Always keeping in mind caveats and exceptions (e.g. community of interest, local nature of
network externalities, multi-homing phenomena), we know that the choice of “non compatibility”
(“standard wars”) can frequently lead to global monopolies. Thus, if WINNER TAKES ALL MARKET!!
What the LOSERS can do? First of all, once the firm is able to achieve and obtain the critical mass
at expense of all the other, the others can open new games. Be capable to lower the barriers to
entries to the network externalities that the leader was been able to exploit.
Because nothing is everlasting!!!!!!!
• Very often in network markets competition and rivalry is not within the market but is for
the market
• Monopoly can only be temporary, and they can always be interrupted by technological
breakthrough (schumpeterian competition) it may always face the threat of other
competitors
• Market shares can be a poor indicator of the degree of competition and rivalry that exist in
a network market

Alternative theories of the firm: contractual and holistic approaches

Why firm exists? what is the objective and its implications?

The neoclassical vision the firm is as a “black box” that have the only objective to maximize profit.
Critical issue is that it is not clearly specified how to transform input in output and not gives us
insight on why the firm exists. In the real world there is a high level of firm heterogeneity
In theory all transaction that increases social welfare could take place in the market. From a
theoretical point of view the existence, behavior and goals of the firm are explained in economic
and managerial theories.

Firms consist of different kinds of agents

• Owner(s)
• Providers of inputs: managers, workers
1. The contractual approach
• A firm is just a bundle of contracts
• Diverse objectives within. Assign different duty to different agents (there might be
different objective according to the different contracts)
• No firm-level objectives
2. The holistic approach (including the neoclassical view)
• Beyond linkages. Can be characterized as an independent entity.
• Anthropomorphic vision: more than the the sum of its parts
• There are clear firm-level objectives (no diverse objectives within)

70
The contractual approach can be divided in two subcategory: principal-agent problem and
contractual imperfections.

Principal-agent problem
We have typically a separation between manager and owner’s firm.
Owner-controlled firm: ownership = control
Managerial firm: ownership ≠ control
• Shareholders (principals) vs. managers (agents)
Perfectly informed world à same outcomes. Manager act in the best interest of the firm. We
would act the same outcome in both firms because we would not have information asymmetries.
Managers would not able to hide something to the owner.
Reality à asymmetric information à discrepancy. Managers do not care directly of the profit.
Objectives of managers could not be aligned whit those of the firms.
• Ownership fragmentation is negatively correlated with the incentive and ability to
overcome these difficulties. Most large firms are owned by a large number of
(relatively small) stakeholders (e.g. publicly-traded companies)
The ways to mitigate these issues are
• Monitoring (board of directors)
• Partial alignment of incentives (profit-contingent compensation). Give the manager
a compensation that is partially contingent with the profit of the firm. For intance
stock option is to buy shares at a fixed price. If i behave as a good managers the
value of the firm and of the shares increase and I have still the possibility to buy
them at that fixed price.

It is not a total alignment because first of all the owner may not be professional and risk prevent
you to doing the right decision. Human being is risk adverted. You might not be able to behave in
the right interest of the firm. On the other hand, if you have zero risk involved this is also
problematic à I have not risks and I not really care (managers).
It is important to find an optimal balance between risk and incentives (fixed wages + profit-
contingent compensations)

Managerial theory. Baumol’s model of sales maximization

Managers goals: tend to prefer sales maximization
Owners goals: profit maximization trying to maximize only revenue.
Why do managers prefer to maximize sales and not profits?
• Compensation is often linked to revenues
• It gives an illusion of personal prestige and visibility. Revenue are most impactive than
profit.
• It also allows strategic and financial freedom. Managers steer all the various activity of the
firm. If you have a lot of revenues, your set of possible action become wider and also
correlated costs. Managers like to have this freedom: expand the company…
• Profit maximization may entail cost-cutting, which is often a difficult decision. For example,
firing employee.
In a very famous research paper Baumol (1959, “Business Behavior, Value and Growth”)
• During the decision-making period, the CEO attempts to maximize sales
• Sales maximization is subject to the provision of a minimum required profit to ensure a fair
dividend to shareholders
• This way, the stability of CEO’s job is assured
71
 • Conventional cost and revenue
functions are assumed
• Total costs increasing
• Demand curve is downward sloping
The profit curve is given by the difference
by revenue and costs. Analytically i should
produce the quantity with the maximum
difference between these two curves.
Managers tend to consider only the
revenue curve, creating an inefficiency.
Managerial theory of the firm - implications
Profit rates are higher in owner-controlled
firms than managerial firms
The higher ownership concentration, the
higher profits
(not always true)
Empirical evidence is mixed.

Team productivity theory (the second principal-agent problem).

The firm is a bundle of contracts and these contracts define certain teams that perform a certain
task. The problem is that is difficult to ascertain the contribution of each individual. Individual
contribution cannot be isolated and measured. Free-riding tendency trying to exploit efforts of
other team members. There is no monitor that free riding could be optimal. Changes are that
teammates focus only on results and they are still going to accept the free rider. It brings ethical
consideration but if you are taking account only of pure rationality, free riding is optimal.
Firm should assume an “insider monitor” with special incentives (not to shirk) in order to prevent
the free rider tendency. “Residual claimant” à Monitor who receives what remains of the budget
after team payments have been made

Entrepreneur seen primarily as a “monitor”

• Similarly, to early vision of entrepreneur as “coordinator of factors of production”)
or “receiver of non-fixed and risky income”. It is a very partial view: entrepreneurs
are more than just a “monitor” device.
- Kizner (1979): entrepreneur is alert to hitherto unexploited opportunities that
only he’s able to recognize and exploit
- Knight (1921): entrepreneur is willing to tolerate uncertainty
- Schumpeter (1934): entrepreneur is a revolutionary, an innovator, overturning
tried and tested convention and producing novelty

Second theory of the contractual approach: contractual imperfections.

Transaction cost theory
Firm arise as a result of efficiency considerations (make-or-buy)
Production costs + coordination costs < Market price + Transaction costs

Market transactions are plagued by transaction costs:

• Search and information costs
• Bargaining costs
• Policing and enforcement cost
72
Firms arise whenever the cost of transacting through the market is greater than the cost of
organizing transactions internally. Relevant determinants:
• The more the bounded rationality, the higher transaction costs
• The higher opportunism, the higher transaction costs
• The higher relationship-specific investments, the higher transaction costs
Also, frequency of transactions affects transaction costs. Low frequency means higher costs.

Property rights theory

Contracts are necessarily incomplete: it is impossible to specify provisions for any possible state of
the world.
Renegotiations are likely to occur ex-post
Property rights influence the relative bargaining position of parties during renegotiations, as they
allow residual rights of control. When I enter in a contract, property rights don’t matter. But of
course, for all these situations that are not mentioned on the contract, property rights matter.
à Ex-ante incentives to transact may suffer, especially when a relationship-specific investment is
required (hold-up problem)

àTransactions may not happen even when mutually beneficial

Integration solves the problems arising from contract incompleteness (integrating activity over a
same umbrella; going to a firm from a bigger one or more vertical integrated)
à Property rights theory is well suited to explain both firm existence and vertical integration

The holistic approach: the resource-based view (RBV) of the firm

Profits are heterogeneous and sticky within markets and industries. If a firm is successful, it tends
to be successful overtime à SUSTAINABLE COMPETITIVE ADVANTAGE (SCA)
Ability of firms to achieve persistent high profit levels over time à RBV does not challenge the
neoclassical profit max assumption. Firms still try to achieve the maximum profit. The Resources
Based View is interested in
• what is a SCA
• how can SCA be achieved
The neoclassical approach is based on structure-conduct-performance paradigm that take into
account strategic interdependence and it leads to the optimal performance.
According to the RBV approach we consider the firms as a bundle of resources. The importance of
the structure is not so absolute. Resources (cash, building, brand…) and capabilities (engineering,
skills…) let the firm to compete and thanks to them it is going to have a higher or lower
performance.
Where does competitive advantage of stem from?
1. Possession of one resource, that other firms do not possess
2. Possession of a way to combine resources that other firms cannot replicate
3. Possession of one (or combination of several) resource(s) that create the best “strategic
fit” with the external environment
Types of resources/assets:
• Physical: plant equipment, location, access to raw materials
• Human: training, experience, judgment, decision-making skills, intelligence, relationships,
knowledge
• Organizational: Culture, formal reporting structures, control systems, coordinating
systems, informal relationships
73
1° point: external environment matters
• It sets what “game” firms are playing (on innovation, cost savings, etc.)
• For example: a gun in a Resort is less useful than a gun in a Jungle
2° point: some resources are key for a wide range of environments, others are not
A resource like a financial resource is extremely important in various environment, instead the
design capabilities is not important in all environment (steel industry). More versatile resources.
3° point: sometimes it is a matter of combining resources rather collecting resources
Ferrari is one of the most powerful brand in the world because it is able to combine resources and
value proposition that cannot be replaced.

What makes a resource (or a mix of resources) a firm possesses contribute to its competitive
advantage? Resources should be:
1. Valuable: must deliver some value
2. Rare!
3. Imperfectly imitable (isolating mechanisms to avoid reverse engineering imitation)
4. Non-substitutive (no strategically equivalent substitute) : possessed by other firms
à Need for some ISOLATING MECHANISMS:
• path-dependence of production processes and resources accumulation (e.g.
ecosystem, brand reputation). History matter. The sequence of past decision is
important. The brand, ecosystem, image that you are able to create overtime include
path-dependence (ex. Apple ecosystem; Amazon; Coke)
• intellectual property rights (IPR) mechanisms (e.g. patents, secrecy). By law you can
exclude other firms to use this product.
• Causal ambiguity – unclear where the advantage is coming from. (ex. Honda win a
competition à it is due to the engineering, mechanics, technicians, management or to
the driver?)

The holistic approach: the evolutionary view of the firm (Nelson and Winter, 1982, et al.)
Even assuming profit maximization as the primary goal of the firm you’re not actually able to
maximize profit: it is impossible. Because there is bounded rationality. Our information,
capabilities and time are bounded. Firms still have an objective, but maximization is unfeasible.
Need for a satisfactory (imperfect) criterion
• Develop overtime strategies and practices that lead to a satisfactory level of performance
• Over time, strategies and practices become ROUTINES so the firm is a bundle of routines
A routine is an instinctive pattern of behaviour that save energy and time. Internalize all the
optimal pattern of behaviour, you do not have to think about it. Routine plays an extremely
important roles in the contests of firms. The various activities are conducted according to
satisfactory criteria: routine. When you share your development in a project is, for instance,
another routine that help to reach a satisfactory level of performance. Overtime you would
develop (and adjust if necessary) routines in order to reach satisfactory level of performance.
Another important aspect is the concept of heuristic: a cognitive shortcut. Not take into account
all the possible point of view in a rational way. Heuristic may also lead to biases.
Search for new practices and strategies is constrained by the existing set of routines. When you
set up a routine is difficult to change it. You need to search for new strategies and practices.
This approach is based on how is readily available in your mind. Descriptiveness of firm behavior.
You do not calculate the actual probability. For intance, a plane accident scares you more than a
car accident even if the probability is lower.
The neoclassical approach is still useful because help you to formalize a very strictly behaviour.
74
The higher competition, the lower the market power. By studying in great detail this extreme
condition, I can infer the very simple and useful law. This is generally true.
Assume a change of driving conditions occurs, say, that the roads have become wet and slippery
and the fog has reduced visibility. The neoclassical theory is able to predict that traffic will be
slower and accidents more frequent. But it is not able to predict that any particular driver will
drive more slowly or have an accident because of heterogeneity.

Firm growth: diversification and vertical integration

Two categories of period: short and long run.
In the short run (less than one fiscal year) you are able to make small changes in production level;
little optimization and variation of the production mix; make-or-buy choices for contingent
reasons (e.g. demand’s peak). Your structure is already set, you care only about avoidable costs.

In the long run you are free to change everything;

• Organizational structure
• Adopted technology
• Plants size and new machinery
• New products or processes
• Entry to / exit from the industry
• Diversification
• Delocalization of some activities
You try to be more strategic.

Conventionally, we recognize horizontal boundaries of the firm. It has to do with the quantity of
products. (ex. Pizza restaurant. Open another pizza restaurant)
Lateral boundaries of the firm that is related to variety of products that means diversification.
Sometimes diversification is put in contrast of vertical boundaries. But it is has not to do with
horizontal boundaries. (ex. Related diversification opening a burger restaurant; unrelated
diversification opening a department store)
Vertical boundaries of the firm have to do with the supply chain. With the distinction between
inside activities vs. outsourced activities / left to the market (make-or-buy decision). Every product
is the result of lots of subsequent activities. Usually these chains are too long to be conducted by
one firm only. Different firms may participate in the same supply chain. The level of vertical
integration is due to the number of activities of the supply chain that a firm is able to do. (ex.
Acquisition of a mozzarella factory; a supplier of yours. You acquire a service of delivery.)
International boundaries of the firm that are related to the geographical span of firm activities
(e.g. offshoring). (ex. Expand in your county or in another country. Integrating forward you can
import tomato sauce)

Why do firm diversify?

Synergies happen when two different
resources create something that is more
than the two alone. Economy of scope
arise when you use the same resource for
different scope, it implies a cost saving.
Diversification also entails costs.
Is risk reduction alone a good reason to
diversify? How do firms finance it?
75
Internal capital market: it is generally easier; you can use the cash flow of stable business.
External capital market (debt and equity markets): they have very heavy requirement. It is not
easy to get it.
The choice to diversify ultimately depends on a cost and benefit analysis: are synergies and
economies of scope sufficiently high to overcome financial, entry and coordination costs?
Three conditions should be satisfied:
• Attractiveness test: the target industry should be structurally attractive. If you diversify in
an industry with a lot of synergies but with a high risk, it could be not a good decision.
• Cost of entry test: the cost of entry should not exceed (expected) future profits stemming
from the target industry. Always take into account costs.
• Better-off test: strong synergies should be present between the two businesses (2+2=7).
Ideally, there should be very strong synergies (7, not 5)
However, the cost-benefit analysis is often biased, and diversification happens much more often
than it should. Why?
Managers like to have high strategic and financial freedom and they can significantly benefit from
diversification:
1. Social prominence, public prestige and political power (building an empire)
2. Limit managerial risk: diversifying limits managers’ risk of achieving a poor profitability
overall àCorporate Governance issues

What mitigates corporate governance issues?

1) Monitoring (e.g. board of directors)
2) Incentives to managers (e.g. profit-based compensation)
3) Capital market discipline (e.g. takeovers risk) : If bad managers are running a company.
Management tends to overpay for diversifying acquisitions.
The stock market ends up internalizing this information, expecting managers to overpay
for additional acquisitions in the future.
The firm’s shares’ market price falls immediatelyà The opportunity arises for another
entity to try a hostile takeover and appoint its own managers, with the potential to
increase profitability thanks to a better management – “raider”à Capital market discipline
deters inappropriate management
4) Labor market discipline (e.g. managers reputation): managers’ reputation circulates in the
labor market. Potential employers know whether managers pursue personal goals or
organizational ones

Key takeaways
• Diversification allows the business to grow and benefit from synergies
• Diversification is optimal only when its benefits (i.e. economies of scope and other
synergies) outweigh its costs (both ex-ante and ex-post)
• Empirical studies show that the performance of diversified firms is, on average, lower than
the one of more focused firms: this is probably due to the fact that diversification happens
much more often than it should, due to self-interested

76
 Vertical integration
Make-or-Buy choice: implement business activities in-house (wholly owned solutions) or
through an external supplier (outsourcing).
Make advantages
• Lower transaction costs (search, negotiating, enforcement and monitoring costs)
• Strategic independence
• Better control on multiple business dimensions
• Know-how protection (protection against potential intellectual property losses and
technological leakage)
• Sheltering of competitive advantage
Buy advantages
• Higher technical efficiency: external providers usually benefit from strong economies of
scale and learning, due to specialization and demand aggregation of several customers
• No need to make additional investments
• Lower coordination costs

Property Rights Theory (PRT):

Integration is important because it determines
• Who controls resources
• Who makes decisions
• To whom profits are allocated
Appropriate ownership structure is conducive to efficiency
The residual right of control is the right to decide on all the situations that are not included in the
contract (typically firm owner)
• Residual rights of control over assets belong to the owner of such assets
• Vertical integration transfers the residual rights of control over the assets of the vertically
integrated firm to the vertically integrating firm
à Transaction costs decrease (higher strategic control, no hold-up problem)
à (But: Coordination costs rise)

77
Example
PepsiCo has two types of bottlers:
1. Independent (no PepsiCo authority on how operations are managed)
2. Company owned (PepsiCo has the ultimate authority over how the bottling assets
are deployed)
In case of a marketing campaign (e.g. Pepsi Challenge), production should match (increased)
demand
1. Independent bottlers can refuse to participate (no residual rights of control over
independent bottlers’ assets)
2. Instead, PepsiCo-owned bottlers can be replaced with a more cooperative team
Empirically, the degree of vertical integration differs:
• Across industries (e.g. firms in the aluminum industry are more vertically integrated
than firms in the tin industry)
• Across firms within an industry (e.g. Hyundai is more vertically integrated than
Honda)

Firm growth: alternatives to integration and internationalization

Reality is more complex than the traditional make-or-buy dilemma would suggest:

We have on the one hand of

the spectrum a wholly owned
integration and on the other
hand a complete outsourcing.
But in the middle, there are
lots of other possibility.
Intermediate forms
1. Tapered integration (making some and buying the rest)
2. Licensing (e.g. franchising, the right to use a firm's business model and brand for a
prescribed period of time)
3. Joint ventures (cooperation on a new joint firm) and strategic alliances (cooperation on a
joint project) in order to pursue one or more shared objective
- Tapered integration is a mixture of vertical integration and market exchange (i.e. co-
sourcing, making some and buying the rest) à a formal co-sourcing

78
 - Franchising
• Franchiser performs tasks involving substantial scale economies (e.g. purchasing
and branding) concede partial ownership rights in exchange for a fee
• Franchisees build and operate the business
• Underperformance is limited through tight quality controls, strict monitoring and frequent
surprise inspections.

Strategic alliance and joint venture

They are the most interesting intermediate form. The most difficult topic. Organizing complex
business transactions collectively without sacrificing autonomy:
Strategic alliance: cooperation, coordination, information and resource sharing for a joint project
by the participating firms. So, two firms agree to share information and resources for a joint
project. For this project the two firms work together to fulfill the goal (ex. Google and Luxottica for
Google Glasses). Does not necessarily presuppose a contract, more informal yhan the joint
venture and tends to be more flexible.
Joint venture: particular type of strategic alliance where a new independent organization is
created and jointly owned by the promoting firms. It is more structured and less risky than an
informal strategic alliance, but also less flexible and more time consuming. Agreement of
cooperation for a specific project. It leads to an organization that is going to be created. It needs a
contract; it is more time consuming and structured. Alliances and joint ventures can be:
• Horizontal (firms in the same industry)
• Lateral (firms in different industries)
• Vertical (firms at different stages of the supply chain)
• Alliances rely more on trust, reciprocity, cooperation and information sharing than arm’s
length contracts do
• Disputes are rarely litigated and tend to be resolved through negotiation. No formal
vehicles to solve a dispute.

Factors leading to strategic alliances and joint ventures:

1. Impediments to comprehensive contracting
2. Relationship-specific assets by both parties. Transaction costs tends to be very high. If the
interests are aligned, opportunism is reduced.
3. Complementarities between the resources and capabilities of the parties involved. Possible
synergies between the capabilities and competence of the two firms involved.
4. Advantages in terms of information, knowledge sharing and contextual adaptation (e.g.
internationalization). Not just a matter of physical assets.
5. Transitory nature of the collaboration opportunity .

Drawbacks:
• You need to trust the other party because there is a potential of opportunism. Risk of
losing control over proprietary information
• Intensive knowledge sharing necessary, some of this information may be property.
• Contracts not very well specified (and imperfect contractual specification).
• Agency (interests of the agent not aligned with the owner’s) and influence (trying to
influence key decisions) costs
• Coordination issues
• No formal mechanisms for making decisions
• No formal mechanisms for resolving disputes
79
 The internationalization process: stylized facts
In the 60’s the offshoring was only of the production activity. Progressively the internalization
landscape start to increase thanks to improved coordination and recombination capabilities due to
advances in ICT, modularization and standardization of complex tasks, iimprovement in the
capability of emerging economies.
Nowadays internalization improvement is thanks to knowledge. A great degree of knowledge is
still tacit.

The smile of value creation (MNEs, knowledge and

location) simply concept that tell you that activities
of a firm can be distinguished by the value that
they have.

R&D and marketing imply a very high value added.

The middle activities are generally offshored.

60s-80s Today
Manual and labor intensive Technology and capital intensive (the impact of ICT)
Orientation towards production Orientation towards customers
One product Product diversification
National business context International business context
Long product life cycle Short product life cycle
Loyal customers Unstable and demanding customers’ needs
Mere profit maximization Greater attention towards social, environmental and ethical issues

After offshoring, some companies decide to re-shore, backshore and near-shore their activities
Why?
• Changes in the business context, now customers and proximity are very important
• Changes in priorities (e.g. cost vs differentiation)
• Emergence of technological enablers (e.g. 3D printing)
• Managerial mistakes

Internationalization: firm involvement beyond the home country boundaries

• Export (simple trade flows between countries)
• Intermediate forms granting indirect presence in other countries (e.g. some types of
licensing agreements and strategic alliances)
• Foreign Direct Investment (FDI): establishment of a controlling interest into another
country
• Greenfield investments (new firm from scratch)
• Joint ventures (when the third entity being created is abroad)
80
 • Acquisitions (purchase of an existing local firm)
Firms engaging in Foreign Direct Investments (FDI) are called Multinational Corporations (MNCs)
or Multinational Enterprises (MNEs)
The theory of multinational enterprise is an extremely large field of economic.

Economic Theories of MNCs

1. Hymer’s seminal contributions (it based the way for many follow theories)
2. Dunning’s eclectic paradigm
3. Verbeke’s dynamic framework

Before 1960s (i.e. before Hymer’s seminal work) several economists investigated international
trade (export). These theories regarded trade and capital movements across borders (not FDI):
• the main determinant of movements of goods (trade) across borders is the difference in
factor endowments. The focus was mainly on trade: movement of goods and thoughts.
• the main determinant of movements of funds (capital) across frontiers is the difference in
interest rates. If a country is abounded in a factor (ex. labour) will specialize in it (labour
intensive goods). You focus in what you can do best and then you trade with other
countries.
à These theories were assumed to extend and apply to all types of investment
Hymer (1960s) distinguishes between:
1. Portfolio Investments (purely financial investments)
2. Movement of goods: exports and imports
3. Foreign Direct Investments (FDI) providing the firm control over the business activities
abroad
FDI are the essence of the internationalization process: while exports are simple market
transactions, FDIs allow firms to grow internationally.
FDI involves extra costs and risks (what Zaheer (1995) has defined the “liability of foreignness”:
being a foreign imply some disadvantages, you don’t start from 0 but from -1):
• Costs of communication and acquisition of information in a different cultural, linguistic,
legal, economic and political context
• Costs of international coordination
• Costs due to less favorable treatment given by host countries’ governments
• Risk of exchange rate fluctuations (monetary risks)
à The MNC’s competitive advantage must be so strong as to overcome the liability of foreignness
To sum up:
• Hymer is the first scholar defining the importance of FDIs as an autonomous category
• FDIs imply both potential to exploit multiple advantages and additional costs
• As always, FDI makes sense only if its benefits outweigh its costs

Dunning’s Eclectic Framework

Synthesis of previous approaches to study why firms become multinationals in a comprehensive
framework. He has provided a framework that sum up a variety of contribution that summarize
the internalization.
The degree and structure of foreign activities depend on the existence of 3 types of advantages:
• Ownership advantages, advantages that belong at the multinational enterprise (3 types)
1. Asset-based (technology, brand, human capital…) Factors of competitive advantage

81
 2. Transaction-based (benefits of common ownership) benefits of common
ownership, related to the ability of the multinational firm to coordinate the various
activity better than the competitors.
3. Institutional (ethics, vision, corporate culture…)
- Location advantages (geographical, political, resource-based). There must be a reason for
choosing a location rather than another.
• Internalization advantages (transaction costs evasion) stem from internalizing an
international economic activity into the boundaries of your firm under the assumption that
this integration gives you some advantages. You might prefer to internalize a certain kind
of knowledge rather than a licensing agreement.
à OLI eclectic paradigm
The three components of the paradigm give rise to four categories, based on the primary drivers
of international expansion:
1. Resource seeking: search for cheap or productive resources.
2. Market seeking search for new markets. Caused by the saturated demand in your
Country.
3. Efficiency seeking international division of production aimed at increasing efficiency
through selective exploitation
4. Asset seeking development of international presence aimed at acquiring strategic
assets
Problems:
• The original eclectic paradigm is static model, it gives you a snapshot of the reasons that
push a firm t to internalize but lack of consider the evolution and the renewal of
competitive advantage
• It is well-suited to give a snapshot of the drivers of internationalization, but it fails to
account for expansion dynamics
• It doesn’t consider the evolution of resources and capabilities thanks to international
presence
• Renewal of competitive advantage is neglected

Verbeke’s dynamic framework (strategic and dynamic point of view)

Main components:
• Internationally transferable firm-specific advantages (non-location-bound FSAs)
• Non-transferable firm-specific advantages (location-bound Firm Specific Advantages)
• Location advantages
• Resource recombination, and value creation through it
• Complementary resources of external actors
• Bounded rationality
• Bounded reliability
Non-location bound FSAs, location-bound FSAs and location advantages refer to the bundle of
resources in the firm’s possession. They can be distinguished based on
• Mobility (transferable vs non-transferable)
• Availability (firm-specific vs location-specific)
Location advantages can be exploited by any firm operating in that location. However, they do not
benefit every firm in the same way
By expanding internationally, a MNE relinquishes its location-bound FSAs and its home country
location advantages. However, it benefits from the host country location advantages

82
Recombination and complementary resources of external actors capture dynamic features such as
innovation, adaptation and evolution
Complementary external resources may be needed for effective deployment of FSAs, especially in
the initial phases of the expansion (dynamic adaptation)
Resource recombination is essential to reinvigorate the MNE’s competitive advantage, developing
new resources and capabilities by selectively integrating existing ones with newly acquired ones
Skillful resource recombination leads both to the upgrading of existing (non-location-bound and
location-bound) FSAs and the development of entirely new ones
Bounded rationality and bounded reliability capture uncertainty
Among the costs of international expansion, it is necessary to account for the fact that human
beings are boundedly rational, imperfectly informed and inherently unreliable
The concept of bounded reliability is very broad: besides including opportunism intended as “self-
seeking interest with guile” (Williamson, 1979), it includes sources of benevolent preference
reversal, such as reprioritization in good faith and failure to deliver on overcommitments

Advantages of the framework

• It highlights the importance of the transferability of (some) FSAs abroad and their degree
of complementarity with the host-country location advantages
• It accounts for environmental complexities (not only bounded rationality, but also
bounded reliability)
• It captures international determinants of the evolution of competitive advantage over time

To sum up
• Consider carefully the interplay between firm-specific (ownership) and location advantages
before expanding internationally
• Be aware you may need to borrow complementary resources, especially in the first phases
• Modern internationalization is mostly about strategic positioning: do not adopt a short-
term exploitation-oriented vision, but consider the possibility to tap into heterogeneous
sources of knowledge and recombine them
• Do not forget about bounded rationality, bounded reliability and the liability of foreignness

Industrial and competition policy- Public Policy towards business

Micro and not macro perspective (e.g. monetary policy and/or general fiscal policy)
We will exclude macro elements as monetary policy, general taxation that refers to all firms. There
are two different public policy: industrial policy and competition policy. Competition policy is
subdivided in regulation (ex-ante) and antitrust (ex-post regulation that want to punish the anti-
competitive that the firms may put in place in the market).

Classification
• “Industrial Policy is a nation’s official total effort to influence sectoral development and
thus, the national industrial portfolio” (Bingham in Handbook of public policy, edited by B.
Guy Peters and Jon Pierre, 2006, SAGE, p. 293).
• “Any type of selective intervention or government policy that attempts to alter the
structure of production toward sectors that are expected to offer better prospects for
economic growth than would occur in the absence of such intervention, i.e., in the market
equilibrium.” (Pack and Saggi 2006).
83
 • “Industrial policy means the initiation and coordination of governmental initiatives to
leverage upward the productivity and competitiveness of the whole economy and of
particular industries in it.” (Johnson 1984)

“Industrial policy is any policy that affects a subset of firms, firms’ activities and industries
differentially from the remaining group of firms, firms’ activities and industries. Any tax, subsidy,
trade and other policy measure that affects only a limited and specific domain of a nation’s
production system can be considered as an industrial policy intervention.” The government is
making a choice.
Examples:
1) Support to young innovative companies (L. 221 Italy 2012, MISE)
2) Automobile industry
-Import quota (or other trade restrictions) imposed by the Government to protect the national
industry
-Government subsidy (or guaranteed low interest rate loan) for buying new (eco-friendly) cars
-The set-up of a new public research center on autonomous self-driving car innovations in
partnership with private operators

a) Different types:
1) offensive (support to specific industries able to play a major role on the global value chains,
e.g. the concept of “smart specialization”) and defensive (“protectionism”, “save losers” or
“sunset industries”) policies, where defensive in the short-term may become offensive in
the long-run (successful “infant industries” policies).
2) Horizontal vs. vertical industrial policies. E.g. an horizontal example could be the previous
one: startup in Italy are supported in all sectors not a choice about the industry but about the
firms activity. An example of vertical policies is the automobile industry one, the government
decided to support a specific sector.
b) Different rationales: economics (e.g. market failures), social justice (e.g. employment
safeguard), national strategic aims (e.g. “national champions” policies), etc. as a country could be
problematic to have foreign operators, this could create problem to defend ouself.
c) Different outcomes: Political and lobbying influence [see public choice theory, Buchanan
2003],we cannot be always sure that the objective of the firms is the maximization of social
welfare rather than their specific interests. “Government failures” [apart from lobbying activities,
i.e. often effectiveness of policies depends “on get the implementation details right”, and this may
be tricky details (Duflo, 2017)] even with the greatest intentions it may result with a failure.

Focus on Policy Innovation (Industrial Policy, at least in its “offensive” component is more and
more coincident with its subset of Policy Innovation): e.g. Italy: Start-up Act and Industry 4.0 Plan
From the Lecture 12 on innovation:
“INNOVATION has become the industrial religion of the late 20th century. Business sees it as the
key to increasing profits and market share. Governments automatically reach for it when trying to
fix the economy.” The Economist”, February 20th, 1999, Survey of Innovation in Industry
Combination of normative and positive arguments (referred more to the actual outcomes)

84
We have always to keep in mind that policy maker
should have the objective to maximize social welfare.
Social welfare will consist to increase efficiency both in
the static domain and in the dynamic one. We already
know that these three cycles interact se there could be a
trade-off between these three activities.

Industrial Policy
Has public policy a role in the whole innovative process?
Of course, yes and of course yes at multiple levels. There are many aspects that may, even in
indirect way, influence the innovative process (e.g. patents). The innovative activity in economy is
strongly (directly) influenced by several forces, including competition policy, regulatory and law
regimes, patent system (among others) all contribute to shape the interested dynamics. Without
forgetting the institutional and cultural context (e.g. see Baumol, 1990, on this latter aspect on
JPE).

Rationales for public sustainment to R&D activities

There is a consistent body of agreement that public interventions in the innovation domain thanks
to externalities and the presence of asymmetric information between firms, bankers, suppliers...
Private firms (may) invest less than the social optimum for two main reasons (universally
acknowledge):
- Knowledge spillovers; thanks to network externalities, firms when investing in R&D exploit
new knowledge that benefits also other firms and competitors. Each single firm is
attempted to limit R&D investment for this reason, but it would lead to a lower investment
of public firms.
- Capital market imperfections. There could be a problem of hidden information or of hidden
action

KNOWLEDGE SPILLOVERS
Initial idea dates back to Schumpeter (1946, ch. 8), then developed by Nelson (1959, JPE) and
Arrow (1962 ch. In edited book) among the first:
“The primary output of R&D investments is the knowledge of how to make new goods and
services, and this knowledge is not rival- use by one firm does not preclude its use by another. To
the extent that knowledge cannot be kept secret, the returns to the investment in it cannot be
appropriated by the firm undertaking the investment, and therefore such firms will be reluctant to
invest, leading to the underprovision of R&D investments in the economy” (Hall 2002 OREP, p. 35).
The argument relies on the stylized facts that any formal and/or informal mechanism to protect
innovation is only partly efficient at the very best.
The underprovision of R&D is more severe for high-tech start-ups due to the higher relative costs
in protecting innovation (e.g. lack of complementary assets, see Teece 1986, Research Policy).

Critique to the Spillovers rationale

ü Existence of spillovers in research generally accepted, documented on an empirical ground,
even if obvious measurement problems exist. Often “localized” (But does the mere
existence of spillovers justify policy intervention?
85
The argument below it is considered a vision. Firm J can take the R&D profits of firm I. firm J can
benefits from other firms without spend anything. Reality is more likely to be similar of a no free
lunch vision. R&D produced by other firms is not immediately useful from firm J. it can be used
and useful by firm J only if the firm have enough capacity to take it, only if it is able to take the
knowledge and combine it with its own knowledge. In this case, the inequality that was posed
before can’t be taken for granted. Is the private level of R&D lower than the social optimum?

CAPITAL MARKET IMPERFECTIONS: Asymmetric information between firm and investor(s)

The hidden info problem (adverse selection)
Hidden action means that the financial part of the firm cannot control the managers in a proper
way. We can use collateral, but it is a remedy that cannot work very well. Asymmetric information
will often lead to capital market imperfection. (moral hazard)

Empirically analyses generally confirm greater financial constraints suffered from R&D
investments than other typology of investment
Cash Flow à R&D (e.g. Hall 1992, NBER; Himmelberg and Petersen 1994, RES;

Credit rating index à R&D (e.g. Czarnitzki and Hottenrott 2011, SBE)
• “Near universal recognition of the presence of market failure in the provision of finance for
high-tech start-ups”. They are the firms that will suffer most for these imperfection

In any case there is a specific type of firms that is suffering more: young, new and high-tech firm.

Taxonomy
Main policy measures
1. Fiscal incentives in R&D
2. Grant to R&D: amount of money that enable the winner of the project to carry on it.
3. Innovative entrepreneurship policy

Fiscal incentives can be considered “automatic” (provided that a firm fulfil the requisites these
firms are eligible for getting the incentives) and grants can be considered “automatic” (choice to
assign or not). They both have pros and cons, there is no one better than another.

Mayor cons of grants is selectivity, the policy maker has a high degree of freedom
• Possible errors and distortions (i.e. wrong choice of a technology) even if with the better
intentions. Member of the committee may not really want to maximize social welfare but
their own.
• Possible inefficient behavior of politician/policy makers:
§ “Cherry picking” (Lerner 1999, JB). Policy maker may channel money to those who
do not really needs (that are able to sustain R&D alone). Just to not pay taxes.
§ “Technology pork barrel spending” (Cohen, and Noll,1991, book). You may devote
money to advantage some specific group of people that may help you to be
reelected. You waste money.
Mayor pros
• No complications to the tax system
• No room for recipient firms to re-label as R&D activities that are not. Just to save money in
terms of taxes.
• Possible to send funds where most needed (from a social point of view)
86
 • Another main “indirect” advantage is that they may exert a stamp of approval by
recipients. You compete against other, if you win it can make your life easier. The winner
of the grants can be a signal in the eyes of possible shareholders. The “stamp of approval”
effect proves to be important also for high-tech start-ups in the European context (e.g.
Colombo et al. 2011 Economics Letters, 2013 Industrial and Corporate Change; Grilli and
Murtinu, 2018, Research Policy).
à Selective (grants) rather than automatic (fiscal incentives) proves to be beneficial
Worldwide implementation of these schemes at various levels (supra-national, federal/national,
regional/local) in the past, present and probably in the future.
But what do we know about their efficacy?

ü Crowding-in (additionality): public R&D subsidies increase private R&D expenditure. The
best possible scenario is represented by crowding in but at the end of the policy we return
on a low level. In the best scenario R&D and the policy measure has produced behavioral
additionality results and firms continuing to invest.
ü Neutral effect: public R&D subsidies do not stimulate neither depress private R&D
expenditure. On one hand, public subsidy increases overall R&D budget of the firm so it is
likely that R&D projects are undertaken that would have not materialize in the absence of
the subsidy. On the other hand, subsidies have to be financed (with taxes) so benefits are
uncertain.
ü Crowding-out (substitution): public funding mere substitute private funding and there is no
(or few) undertaking of new R&D project. Firms would have realized (many of) those
projects even in the absence of the subsidy.

Innovative entrepreneurship policy

it does not mean in any extent the entrepreneurship rate. It is not about quantity but about
quality of the firm created. Lowering entry barriers to become an entrepreneur is not a good idea:
entrepreneur enter expectation and they have an expectation. A subsidy may disturb this vision of
the expectation. Competitive MKt where each firm is charcaterized by θ (estimate of the value of
own productivity, capability):
q!
Π=pq - θ
At the beginning of each period, each firm decides whether to remain active or not. Next, active
firms decide how much to produce, which they do by choosing the quantity that max profits.
!" 𝟏
1° order condition: p - θ = 0; so: q* = 𝟐 p θ
# # 𝟏
Maximum profit levels: Π = ! 𝑝! θ – (! 𝑝θ)! / θ ; so: Π* = 𝟒 𝒑𝟐 θ

“[…] If the selection process is the outcome of a Bayesian process of learning, a subsidy may be
both useless (the more efficient entrepreneur does not need it, while the less efficient one leaves
the market once the subsidy ceases to be in operation) and harmful (less efficient entrepreneurs
are given an artificial seedbed, while market competition would have induced them to leave the
market). If the former situation is prevalent, the industrial policy supporting entry is affected by a
deadweight component; if the latter is prevalent a substitution effect arises”.

The second consideration is that when we talk about success in innovative entrepreneurship the
funder’s human capital is extremely important. “For a new, high-technology firm, the primary
assets are the knowledge and skills of the founders. Any competitive advantage the new firm

87
achieves is likely to be based upon what the founders can do better than others” the human
capital gives you the skills to put in practical the success. A policy intervention should really look to
incentivise people with high human capital. This was the aim of many start-ups in the last decade
starting in USA and then followed by other Countries.
à Specific vs. generic human capital (Becker, 1975: Human Capital. National Bureau of Economic
Research, New York)

Several recent interventions after the Great Financial Crisis tried to stimulate “innovative”
entrepreneurship with the tendency to overarch all these measures under a unique coherent
umbrella (as a proper industrial policy should do).
We have seen the blossom of Start-up Acts around the world

The Italian case

• Regulatory change in Italy (2012) – “The Startup Act” intended to spark the national
innovation ecosystem.
• Targeted Young Innovative Companies (YICs).
• Requirements:
• <5 years old;
• <€ 5m annual sales;
• not listed;
• no corporate spin-off;
• Innovative:
• Tangible IP rights (e.g. patent, license);
• R&D investments >15% of the revenues;
• >1/3 of employees/founders must hold a PhD or >2/3 must have a master
degree.

Regulation (quaderno)

Regulation of tariff dynamics in practice

• Cost Plus
• Price Cap
• Yardstick competition

Cost Plus Regulation

• The regulator defines a tariff to allow a volume of revenues equal to the sum of incurred
costs, which include a fair remuneration for the invested capital:
TR= Cost + r RAB
§ TR: allowed total revenues
§ Costs: Operating costs (labour, materials, services) + annual capital depreciation
(infrastructure, network, machineries)
§ r: rate of return on capital
§ RAB: Regulatory Asset Base (net invested capital)
There is the sum because it is the opportunity cost in investing money in other alternative uses

88
Procedure:
- based on firm accounting of the previous year, the regulator determines costs that can be
recovered
- Based on these costs and on expectation about future demand, total revenues are
determined
- Price structure is proposed to the
firm, which approves it
- Prices are implemented
Every year the regulator has to analyse
information and decide and calculate the
value and the tariff that should be posed
to the regulated firm. We have three
different means for efficiency:
productive, allocative and dynamic one.
Dynamic efficiency in this context must
be understood.

Example : Gas transportation in Italy (4°transitory 2018/2019 period – 5° from 2020)

For consumers: Allocative efficiency (minimum price but only with respect to observed costs)
• Possible productive inefficiency, due to information asymmetries
§ Price is based on accounting presented by the firm, not on an “efficient frontier” of
costs
§ No incentives to reduce costs
• High administrative costs
§ Frequent revisions
§ A lot of information needs to be elaborated
For the regulated firm:
• Financial-economic equilibrium
§ Costs are covered
§ Investors and creditors are remunerated
• Reduced organizational and technological dynamics
§ No incentives to implement best-practices
§ No incentives to adopt innovative systems
• High propensity to invest (it ensures remuneration)
§ Investments are not necessarily efficient, in terms of localization, technology,
dimension, etc.

For this tendency to (over-) invest in physical capital generally dynamic efficiency results are not
considered that bad with this regulatory scheme

Price Cap Regulation

A price cap simply sets a maximum allowed inter-temporal path for the price of a specific product.
The rules for the path are set in advance and only depend on factors that are beyond the control
of the regulated firm. The regulated firm sets the price for the service with the only requirement
of not exceeding the ceiling imposed by the regulator. The regulated firm maintains its natural
objective of maximizing profits. The basic price cap rule:

89
Procedure
1. At the beginning of the regulatory period: accounting analysis and costs’ estimation
§ It can be assumed that the cap is equal to the most recent price, or to costs
measured on a sample of similar firms (yardstick: see later)
§ Or simply repeat Cost Plus estimation
2. Regulatory period definition (normally, 3-5 years)
§ The length of the regulatory period allows the firm to keep gains from productive
efficiency improvements and, hence, creates an incentive to adopt efficient
behaviours
§ Weaker incentive at the end of the period: the firm knows that efficiency gains are
short-lived. This may lead to the so called Ratchet effect: the firm, knowing
regulator’s behaviour, slows down in its path towards efficiency as the end of the
period approaches
§ Time dynamics
§ Productive efficiency gains exceeding those predicted by the formula go to the firm
§ The firm could also incur in losses if cost reduction occurs at a rate lower than X
(but this situation may turn out to be undesiderable for any party involved:
regulated firm, regulator, citizens)
Profit sharing at the end of the regulatory period
§ In the revision year, the new base is not equal to the costs the firm had over time
§ The difference price-cost is distributed to consumers and firm
A percentage of this margin is included in the new base of the Cap, while the residual is left to the
firm (in Italy usually a 50%-50% rule is applied)
Basic formula augmented with……
Cost pass-through: the regulator can modify the Price Cap formula to pass-through consumers
those costs on which the firm has no direct control. Quality and other investment incentives:

Social welfare effects:

• Productive efficiency:
§ Incentives to cost reduction: beneficial to consumers
• At the end of the regulatory period, part of benefits are passed to
consumers, and part is a premium to the efficient behavior of the firm
• Possible allocative inefficiency:
§ During the regulatory period costs are not measured
§ Prices are normally higher than a firm’s costs
[but a likely win-win scenario at the end in comparison with cost-plus]
• Lower administrative costs with respect to Cost Plus

For the regulated firm:

• Organizational and technological dynamics
§ Continuous productivity gains
• Adoption of new organizational practices and new technologies
• “Mixed” propensity to invest
§ Higher for cost-reducing investments
§ Lower for quality-enhancing investments (alleviated if not solved by adjustments in
the formula)
§ Exposure to economic and financial risk
90
 § Costs are covered only if there are continuous productive efficiency gains
§ Variable remuneration for investors

Examples price cap regulation in Italy

Electricity
2° Regulatory period: x = 3.5 for distribution (2.5 for transmission) . In the 1°period it was higher
(c.a. 4% for both)
3° Regulatory period (2008-2011)
x = 1.9 for distribution; x = 2.3 for transmission
4° Regulatory period (2012-2015)
x = 2.8 for distribution; X = 3 for transmission
5° Regulatory period (2016-2023)
x = 1.9 for distribution; x = 1 for transmission

Yardstick Competition
• Model of regulation based on the principle of comparative competition between firms
(Yardstick Competition)
• Assumption: if firms operate under similar conditions, they should, in principle, have
similar costs
• The model can be applied to price regulation:
§ The regulator binds the price of a regulated firm to the average costs of other firms
§ By imposing separation between price and cost, we obtain a mechanism with a high
incentivizing power (analogous to price cap)
• The incentive to reduce costs is due to the fact that efficiency gains do not translate into
revenues reduction
§ If the firm reduces costs at a level equal or lower than the average cost set by the
regulator, the firm obtains an increase in profits.
• In general, by applying this mechanism to all firms, we get an overall cost reduction and, as
a consequence, a price reduction
In theory, this is true for firms operating in identical contexts

In practice, firms operate in different contexts (e.g., electricity distribution or water supply:
environmental diversity → population density, climate, altitude), which generate different unit
costs
• (Partial) solution: econometric/statistical methods to evaluate to what extent exogenous
variables affect costs

Implementation problems:
• a) Inefficiency can be due to unobserved variables
• b) what to do with inefficient firms that do not manage to reach the yardstick?
• To sum up, yardstick regulation introduces an indirect form of competition, but its
implementation is far from being trivial. In practice, regulators rarely use it in its “pure “
form. At maximum, they use its logic when they build comparative measures relatively to
some cost categories and use them to retrieve information within other regulatory
mechanisms (e.g., Service distribution is usually regulated through Price Cap but initial
costs of the local monopolistic utilities are computed adopting a “yardstick” spirit).
Effects (analogous to Price cap):
- Possible allocative inefficiency
91
 § Costs of the regulated firm can be different from the price calculated on costs
estimated on a sample of firms
- Productive efficiency
§ The firm benefits (temporarily) from cost reduction
§ During price revision, productivity gains enter average costs valuation and, hence,
in the definition of the new price → producers’ surplus extraction

Avoid “price squeeze” by incumbent on margins of new entrants

- Beside “access price” regulation through mechanisms we have already seen (e.g. price
cap), introduce unbundling.
- 4 types: accounting, functional, legal, ownership unbundling
- Trade-offs in terms of productive, allocative, dynamic efficiency

Phase 1: Monopoly vertically integrated

Regulation of final tariff so to limit market power of the only incumbent firm in the market. Before
the liberalization we were in a vertical integrated monopoly owned by the State and regulation
were implicit.
Phase 2: Opening to competition. Regulation is aimed at ensuring fair conditions of entry to new
firms but still final tariffs are regulated so to avoid exploitation of market power by the incumbent
ex-monopolist which is now a dominant player in a market where competition is still at its infancy
stage. Regulation becomes explicit. A product firm want to maximize profit so there should be a
regulatory agency.
Phase 3: Access. Regulation focus is on ensuring that firms are not discriminated in accessing
essential facilities still considered natural monopolies. The focus of the regulator is only on the
price of access of the infrastructure that is still in the end by incumbents and not only in the final
tariff. Final tariff are not regulated anymore or not so heavily as in the past. Entry of new firms is it
now possible with a price equal to costs. Competition become stronger and stronger and the focus
of the regulator could be relaxed because competition will lower tariff.

92
Concentration Indices
Why industry concentration matters?
• One of the most studied areas in the domain of market structure, particularly in the
industrial organization literature. It is very important to have a fair level of competition in
an industry. Perfect competition is better for social welfare than monopoly. The more we
have competition, the more we have social welfare. In monopoly we have a high DWL.
• When antitrust agencies are evaluating a potential violation of competition laws, they will
typically attempt to measure concentration within the relevant market.

Intuitively, concentration depends on the number and size of the firms belonging to the focal
industry
Concentrated industry typically has very few firms or many firms, but few of very large size
Dispersed industry has many firms and none of the firms has a dominant position. Equal market
share
In a nutshell, concentration
I. is widely studied in industrial economics
II. is useful to get a clear and immediate snapshot of the focal industry
III. has deep effects on firms’ behavior and offers indications on competition mechanisms
- Analysis of the causal relationship:
Is concentration leading to market power or viceversa? In reality the two things are
intertwined and both things are true. Also market power lead to concentration, if you have
market power you increase the likelihood that you stay in that market and increase your level
of concentration.

Oligopolistic industries are characterized by high concentration. The maximum level of

concentration is reached in the case of monopoly
Competitive industries are characterized by low concentration. The minimum level of
concentration is reached in the case of perfect competition.
Things are not always so clear, there might be some intermediates level of concentration. This is
why indexes are useful.

THE MARKET SHARE is the ratio between sales of firm i and total sales in industry j
Market share of firm i in industry j, where N firms operate:
𝑞,-
𝑠,- = .
∑,/# 𝑞,-
Where 𝑞,- is the sales of firm i in industry j
N firms (rows) operating in M industries (columns)

Focusing on one industry, the market share would be: si = qi/Q with Q=Siqi

The concentration vector lists market shares in descending order, in order to emphasize the
presence of large firms in a given industry. It is simply a matter of ordering. It constitutes a
convenient representation starting from the largest market share ending with the lowest. It might
be convenient to see immediately see where the concentration is. If also the largest firm has a low
market share, it is not a very concentrated market.

Concentration curve

93
A useful representation of concentration in the Cartesian plane (for a certain industry at time t) is
given by the concentration curve
• Horizontal axis: firms ordered in decreasing
order by size (from the largest to the smallest)
• Vertical axis: Cumulative Market Shares (CMS)
The sum of the market shares cumulative held at the
beginning is going to grow faster (first firm has the
higher market share). The sum must be 1. If it grows
very fast at the beginning it is a very concentrated
market
Concentration indices
• synthesize in one index the entire information contained in the concentration vector. You
can have an immediately idea of the concentration in an industry.
• facilitate the comparison between the degree of concentration in different periods and
industries.
They can be:
I.Absolute: weighted sum of market shares. With different weights, different concentration
indices are identified. Attached to each market share I
have a particular weight. There are different indices
with different rational

II. Relative: refer to the degree of inequality between firm sizes within a certain industry

Concentration index - concentration ratio

Sum of market shares of the first k firms, in decreasing order (vector)

The weight =1 for the first k firms

The weight =0 for the remaining N-k firms
As rule of thumb k is usually equal to 3, 4, 8 or 20
You can have many indices depending on k. that denotes how many firms you are going to put in
your concentration index. I take into account the cumulative market share in descending order. In
k= 1 I will take into consideration only the first firm with the higher market share.
Criticalities:
1) The choice of k is arbitrary, and this is a problem because it is going to influence the index
2) It provides the same result for industries with different concentration
Example:
§ Industry A: (0.1, 0.1, 0.1, …)
§ Industry B: (0.37, 0.01, 0.01…)
§ ® C4A=0.4 and C4B=0.4
While in A market shares are evenly spread, in B there is one firm that controls almost 40% of the
market: this crucial information is lost. The firs one looks kind of oligopolistic and in industry B
the situation is similar to a monopoly. We need to be careful for this reason.

94
Herfindahl index
The weight of each share is given by the share itself: the HI is given by the sum of
the square of the market shares
The larger a firm, the more it contributes to the value of the index. I’m simply
summing the square of market share. I’m giving more importance to larger
market share. I want a synthesis of the level of concentration. HI ranges from 0 to 1

If there are N firms with the same size, HI= 1/N

- In perfect competition HI=0 (N à infinite)
- In monopoly HI=1 (N = 1)

Entropy index
In the Entropy index, the weight of each share is equal to the logarithm of the
inverse of the market share
Smaller firms provide a greater contribution to the total amount of the index
(through the inverse of the logarithm) if I had a very small firm I’m going to have a very large value
for one market share. The larger it is do not means anymore the larger of concentration. What the
industry is telling me is the entropy of the system.
The entropy index provides a measure of disorder, ranging from 0
to infinity
• In an industry where there is only one firm, the entropy is
minimal:
• In an industry where there are several identical firms, the
entropy increases together with the increase in N,
theoretically approaching infinity in perfect competition
It is not so much used in the literature

The role of the degree of disaggregation

Concentration indices are used to compare concentration:
• Between different industries at the same time
• In the same industry over time
The degree of disaggregation needs to be the same in order to meaningfully compare results.
Indeed, the degree of disaggregation of the industry has an impact on concentration (typically,
increasing the degree of disaggregation, concentration increases as well)
e.g. Biscuit industry (disaggregated so more concentrated) à confectionery ind. à food industry
(very aggregated). I have to compare industries with the same level of disaggregation

• The degree of disaggregation is defined relying on the classification of production

activities
• There are several national and international classification of production activities that
allow for a shared definition of the industries
Example:
• ATECO - ATtività ECOnomiche (ISTAT, Italy)
• NACE - Nomenclature statistique des activités économiques dans
la Communauté européenne (Europe)
• SIC, Standard Industrial Classification (US)

95
Relative concentration indices
measure the degree of size inequality among firms
• Graphically à Lorenz Curve
• Numerically à Gini Index
These measures are traditionally adopted in labor economics in order to measure income
inequality and income distribution.

Lorenz Curve
Firms are put in increasing order
On the horizontal axis: cumulated share of the number of firms e.g. The 10% smallest firms…
On the vertical axis: cumulated market shares e.g. …produce 20% of the industry output
The upper bound of the Lorenz curve is given by the perfect equality line
• The perfect equality line is the shape the Lorenz curve would get assuming perfect
equality:
§ The first 10% of the firms
produce 10% of the output
§ The second 20% of the
firms produce 20% of the
output
§ …
• Note: the portion of plane
comprised between the Lorenz
curve and the perfect equality line
is called concentration area
In the reality there are always a certain
level of inequality. in this case if the curve rises very slowly ending rapidly, it is very concentrated
The perfect inequality is the case that you have one firm that produce for the entire market
(triangle area)

Gini index
Ratio between the concentration area and the area underlying the perfect equality line
The Gini index varies between 0 and 1:
• All the firms have the same size
§ The absolute equality line and the Lorenz curve coincide
§ The concentration area is null
§ The Gini index is equal to 0
• One firm produces for the entire industry
§ The concentration area coincides with area underlying the perfect equality line
The Gini index is equal to 1

Encava and Jacquemin qualitative classification

It synthesizes industry concentration in a finite number of categories:
1) Monopoly: there is one firm with a market share higher than 80%
2) Dominating firm: there is one firm with a market share between 50% and 80%, and the
others are much smaller
3) Duopoly: 2 firms of similar size control at least 80% of the market

96
 4) Asymmetric oligopoly: 3 or 4 firms control at least 80% of the market, the highest share
being around 40%
5) Symmetric oligopoly: 3 or 4 firms equally control at least 80% of the market
6) Asymmetric competition: the largest firm holds a market share between 20% and 50%
7) Symmetric competition: the largest firm controls at most 20% of the market

ANTITRUST
A practical application of all the previous chapters. It is a quite divisive topic. If you look at the
same practice put in place by a firm, some economists could think that this strategy is completely
fair while others not.
Mostly ex-post form of regulation. Three antitrust areas that deals with antitrust regulation:
• Agreements between companies that restrict competition (Art. 85 EU Treaty of Rome,
1958 [~Art. 101 TFEU, 2007]; US Sherman act 1890) Cartels or other unfair arrangements in
which companies agree to avoid competing with each other and try to set their own rules.
Anticompetitive practices: Cartels (Collusion)
• Abuse of a dominant position (Art. 86 EU Treaty of Rome, 1958 [~Art. 102 TFEU, 2007];
Sherman act 1890) and these two are the ex-post regulation.
A major player tries to squeeze competitors out of the market: Predatory pricing
• Mergers & Acquisitions (Art. Clayton act 1914; EU Council Regulation (EEC) No 4064/89 of
21 December 1989) ex-ante à antitrust want to be sure that M&A activity does not
constitute a sort of artificial dominant position.
Originated in the US in 1890. There were two section: agreement between companies and the
abuse of a dominant position. It has parallel norms in terms of EU.
What these practices mean and what antitrust does to punish and correct these behaviors?
The most emblematic activities are the collusion (1) and the predatory pricing issue (2).

COLLUSION à firm agree on adopting the same type of strategy in the market. the classical
example is that they fix the same price (monopoly price) and they divide the profit in 2. In this
case, the collusive agreement’s dimension is the price.
Strategic variable could also be the quantity according to Cournot. The two firms reduce the
quantity produced to have a higher price for this quantity for having a larger profit. If the firms are
homogeneous colluding is more convenient rather than competing.
p = a – b(q1 + q2)
Π1 = pq1 – cq1 = [a – b(q1+q2)]q1 – cq1
Π2 = pq2 – cq2 = [a – b(q1+q2)]q2 – cq2
dΠ1/dq1 = a – 2bq1 – bq2 – c = 0àq1* = [(a – c) /2b] – (1/2)q2*
dΠ2/dq2 = a – 2bq2 – bq1 – c = 0àq2* = [(a – c) /2b] – (1/2)q1*

Colluding is generally more convenient that competing for firms. Fixing a monopoly price (or
quantity) and divide it for 2, it is always better than competing for both firms. We compare similar
firms in terms of product and cost function they have. Colluding may be more convenient but once
the agreement is made a firm could also cheat, having incentives to deviate the agreement.

Sustainability of collusion depends on

1) n° of (identical/similar) firms in the market (if there are only 2 firms that decide to collude, they
split the monopoly profit in 2. If there are 4 firms that collude, the incentives to deviate the
97
agreement is greater because the hypothetical profit that the cheater could reach is higher than
the previous case. By fixing a price lower than the monopoly price a firm can gain much more).
The higher the number of firm involved, the higher is the incentive to cheat.
2) probability that no external factor will occur to change the game (e.g. entry of new firms,
obsolescence of the focal product overtime).
3) frequency of interactions between firms (i.e. the number of times firms play the same game)
2. Repeated interactions and 3. No game changer
We have to consider the possibility that there is a game
changer.
1. If the game is played only once, it is a prisoner’s
dilemma type of game: NE is (1,1) cheat and cheat. A collusive agreement is not
sustainable.
2. Now suppose that the game is repeated an indefinite length of time and the probability the
game will be played each time is 0 < p < 1. Also suppose that firms adopt a punishment
strategy for deviation from the collusive agreement of a grim type (“I collude until you
collude, if you cheat even once I will cheat forever”).
2
3a. If a firm cheat its payoff is: 3 + 𝑝 + 𝑝! + 𝑝0 + ⋯ 𝑝1 + ⋯ that is equal to 3 + #32.
!
3.b. if a firm colludes, its payoff is: 2 + 2𝑝 + 2𝑝! + 2𝑝0 + ⋯ 2𝑝1 + ⋯ that is equal to #32.
! 2 #
4. Both firms will collude if: #32
> 3 + #32 ; 𝑝 > ! ; it depends on the probability that no game
changer will occur, with a sufficient high level of probability, firms will decide to collude.
5. If «p» is high enough (so «no game changer» is at the horizon) firms will likely collude (or to
better say the collusive agreement is sustainable).
#
N.B. if |p| < 1, we have that ∑1/4 1 !
1/5 𝑝 = 1 + 𝑝 + 𝑝 + ⋯ = #32
The higher is the probability that there is no game changer, higher is the probability to collude.

Internet is becoming a transaction arena. E-commerce increases frequency of interactions and, in

doing so, enhance «collusion» possibilities. One emblematic example is represented by Latcovich
& Smith 2001 “Pricing, sunk costs, and market structure on-line: evidence from book retailing”
Oxford Review of Econ. Pol. Selling books in internet there were two major firms Amazon and Bn.
When a book was ranked in the best 15 in the New York Times, the price for both firms was the
same. When the same book exited the 15 positions, both firms increase the price.
àThese temporal dynamics of a market can reveal that “big players” are colluding...Why????
INTUITION:
Firms collude on price. In periods of:
- High demand à High incentive to deviate from the cartel (i.e. higher profits at the
expense of the other firms) à collusive price has to be reduced in order to reduce the
incentive to deviate and make sustainable the cartel.
- Low demand à Low incentive to deviate from the cartel (i.e. lower profits at the
expense of the other firms) à collusive price can be raised since incentive to deviate is low
and the cartel is still sustainable. Countercyclical movement of price means that the
collusive agreement may be enforced by the presence of other firms in the market.

Tacit vs. Explicit collusion

We are entering in a very complicated issue. Explicit collusion is always illegal. Tacit collusion is not
always illegal and for a regulator id difficult to prosecute it. The collusion could arise with
conscious parallelism, without any communication. Firms may understand a way to collude
98
observe the moves of competitors, following them. From these two extremes (explicit and
conscious parallelism) there are a lot of options to put in place concentred action, it is a blurred
area. The parties never end with a formal agreement. Even tacit collusion if it involves some
communication it may still give rise to some antitrust persecutions.
Premise: Dominant position is never punished if not artificial (see slide on drastic vs. non drastic
innovation) natural monopoly means that a product is really appreciated by consumers and/or
technologically superior. The market from being competitive it becomes a monopoly. What is
really punished is an abuse of dominant position. An example is predatory pricing: price an item so
low that competitors should left the market.

Suppose that demand is given by: P = 120 – Q and all firms have constant marginal cost of c = $80
Let one firm have innovation that lowers cost to cM = $20.
This is a Drastic innovation. Why? Better use of the input. We have benefits on these three
different aspects:
§ Marginal Revenue curve for monopolist is: MR = 120 – 2Q
§ If cM = $20, optimal monopoly output is: QM = 50 and PM = $70
§ Innovator can charge optimal monopoly price ($70) and still undercut rivals whose
unit cost is $80
Should the innovator being blocked because it will turn out into a monopolist?

Ex-post allocative inefficiency but not ex-ante: p now is 70$ and not 80$, 50 units traded rather
than 40.
Ex-post productive efficiency gain: now producing 1 unit costs 20$ rather than 80$
Dynamic efficiency
(a new technique available)

A major player tries to squeeze competitors out of the market: Predatory pricing
• Is there a dominant position?
To answer this question depends on the definition of a relevant market

What is the “relevant” market? (EU Commission definitions)

(a) A relevant product market comprises all those products and/or services which are regarded as
interchangeable or substitutable by the consumer by reason of the products' characteristics, their
prices and their intended use.
(b) A relevant geographic market comprises the area in which the firms concerned are involved in
the supply of products or services and in which the conditions of competition are sufficiently
homogeneous.

A relevant market deals with

• Demand side substitutability (customers) in terms of product characteristics
• Supply side substitutability (suppliers) if supplier can serve a given focal areas
Practical determination is far from obvious and quite complicated (e.g. survey data collection,
analysis, econometrics, etc.).
Several methods and techniques, all not-immune from critiques and problems. It can also become
some sophisticated from a implementation point of view.

“Small but significant non-transitory increase in price (SSNIP test)

99
- Start with smallest possible market and ask if 5% price increase would be profitable for a
hypothetical monopolist (market for bananas or market of concrete in the area x). That means
that people do not want to change banana with something else. That specific product has no
specific substitute.
- if not, then firm does not have sufficient market power to raise price. The antitrust analyze the
next closest substitute.
- Next closest substitute is added to the relevant market and test repeated (market for bananas &
kiwi or market for concrete in x&y).
- Process continues until the point is reached where a hypothetical monopolist could profitably
impose a 5% price increase (fruit market or market for concrete in x&y&z). Profitable. Also, to
compete for the area.
- Market then defined (fruit market/ market for concrete in x&y&z).

Cellophane fallacy: generally, the “prevailing” price is the starting point to apply the SSNIP test,
but this is not necessarily a competitive one. If it is already the one of monopoly, since the
monopolist optimally sets a price in correspondence of an elastic part of the demand curve
(remember the mark-up formula), in this case “an increase in the current level of prices might
induce customers to switch to other products that would not necessarily be substitutes under
competitive conditions (e.g. lower quality products)” (Willis 2005, Introduction to EU Competition
Law, p. 28).”

Once “relevant market” defined

What in practice constitute dominance?
(a) Market share (generally above 40% in EU; 50% in US)
(b) Length in time of that market share
(c) Differences in market shares with second competitors (high Herfindahl index)

• Is there an abuse of a dominant position?

What is an abuse? The company exploits this position to eliminate competition
How in practice? Example we can have different forms of it
(a) depriving smaller competitors of customers by selling at artificially low prices they can't
compete with (Predatory pricing)
(b) (b) obstructing competitors in the market (or in another related market) by forcing
consumers to buy a product which is artificially related to a more popular, in-demand
product
(c) (c) refusing to deal with certain customers or offering special discounts to customers who
buy all or most of their supplies from the dominant company

What makes a price a predatory price?

1) Areeda-Turner (1975) a price is a predatory price if this price is below its marginal costs.
2) Recoupment loss (Ordover and Willig 1981). The firm that set p < MC will incur in negative
short-run profit because doing that competitors would leave the market. In the long run can fix a
price that is much higher than marginal cost gaining long term profits. It is able to do that because
there is less competition and competitors left the market. Reasonably expectation to gain more in
the long run.
3) Predatory intent (proofs of that). The antitrust can enter to the firms documents and in many
antitrust cases, it can also show a clear predatory intent.
Note: Generally, for US are all necessary conditions, while in EU 1) might just be sufficient
100
Looking at social welfare maximization that should represent the aim of Antitrust, all this process
turns out to be really problematic once put into practice. WHY? Think to “predatory pricing”: price
down is the only sure thing of the whole process. Different views and different implementation
philosophies about Antitrust policy. Because of the degree of randomness, it is affected to
subjectivity. Not 100% sure that it is a predatory pricing strategy.
The Chicago antitrust school aiming at constructing a base of knowledge and building a series of
argument on each antitrust intervention. The intervention of the state may be revealing itself to
be decremental.
Antitrust legislation may end up harming rather than benefiting social welfare.
Over time, the Chicago School has built a theoretically sounded framework aimed at diminishing
the relevance of antitrust interventions.
See for example the “single monopoly” theory (Bork, 1978) to illustrate how monopolists do not
have interest for monopolizing adjacent markets, since there’s the possibility of only “one
monopoly profit”.
Suppose that we have homogenous consumers: each one has wtp = 10 for jars & lids.
Cost for producing a jar = 3
Cost for producing a lid = 3.
a) If a firm is a monopolist in the “jars & lids” market: Best unit profit = 4 (10 - 6)
b) If the two markets are separated, and jars offered at a competitive price (equal to cost, so
price of jar = 3), the remaining wtp of consumers for lid is equal to 7, thus again the
maximum unit profit is = 4. (10 - 3)

No interest for the lid monopolist to monopolize even jars, because profits will remain the same.
[The theory is valid, only under very specific assumptions, e.g. goods are consumed at fixed
proportions, homogenous preferences, etc. but it is an exemplification and explain how harmful
the antitrust regulation could be]. It can be decremental also to social welfare.

Example on predatory pricing

1) Firms may refrain from pricing aggressively so to avoid any risk of being accused of predation
(MC has to be estimated and one never knows how things turn out in legal courts). The
interventions of antitrust can be harmful also to consumers: everybody likes low price.
2) Predatory Pricing is very unlikely:
- It is extremely expensive (price reduction may also mean higher output to produce at very
high costs, if MC curve is increasing). And maybe unaffordable.
- Acquisition can be a much less expensive way to eliminate a competitor.
- If the prey exits easily, this means that some others could also easily enter into the market.
Low barriers to exit are also correlated with low barriers to entry.

Strategic use by inefficient competitors of antitrust infringement claims of efficient firms in a

“Tonya Harding-style competition” (Evans, 2009, The Middle Way on Applying Antitrust to
Information Technology). Tonya is a famous US skater that before the Olympics ask her husband to
break the leg of her major opponent. Unfair style of competition. The winner is a winner because
put in place practice that are not fair in perfect competition. The winners have to deals with all
these claims (antitrust issue) and this has the effect to distort the winner to be better using the
same amount of money investing in something more beneficial for consumer. If I can’t beat you on
the field, I ask for a third party (regulatory agents) to intervein the beat you in another field that is
not the market.
101
The second point is that different Philosophies but also different implementation:
Application of Rule of reason vs. Rule of law [“per se” violations]
All these norms (collusion, predatory pricing, abuse of dominant position) can envisage a violation
“per se”. All firms can be punished by the antitrust regulation. The rule can be applied in a very
strict rate or with more subjectivity (rule of reason). Case by case you have to decide.
Problem of Rule of Reason: subjectivity in the decision, lack of transparency, little predictability to
market players
Problem of RoL (per se): risk of wrong decision in terms of welfare , you always has the risk to
consider illegal someone that is not anti-competitive.
US started by RoL EU by RoR, but now they are quite close to the other. EU surpassed US toward a
Rule of Law.

Merger regulation
The focus of antitrust is on HORIZONTAL (see also Appendix 2 on Cournot oligopoly: more
concentration = more mkt power). From EU antitrust guidelines: “Horizontal mergers – i.e. those
between competitors on a particular market – can lead to a loss of direct competition between
the merging firms. By contrast, vertical and conglomerate mergers do not immediately change the
number of competitors active in any given market. As a result, the main potential source of anti-
competitive effects in horizontal mergers is absent from vertical and conglomerate mergers. They
are thus generally less likely to create competition concerns than horizontal mergers. In addition,
vertical and conglomerate mergers may also improve a company's efficiency by better ordinating
their different production stages.”

Historically, considerations related to dynamic efficiency were on the background and did not
enter invasively into the decision. Manly why the relationship between competition and
innovation is not monotonic. We can often observe a U-shaped correlation between them. Even
economic reasoning gives not precise indications on which market form is more conducive to
innovation. That is why antitrust as the very recent past refrain from making dynamic efficiency
entering into the picture and making affect the decision whether to consent for a merge or not.
See Denicolò and Polo (2018): “[…] the literature has shown that competition may be either good
or bad for innovation, depending on the circumstances. In the light of this, antitrust authorities
have generally refrained from taking extreme stances and have followed a cautious approach.
Intervention has been limited mainly to cases in which the merging firms’ innovative products are
close to the commercialization stage. In these cases, innovation outcomes have been regarded as
sufficiently predictable as to be amenable to the standard, static analysis.”
This is still very much the case, even if things may change in the (near) future. Dynamic efficiency
does not enter since the very recent past. From an academic point of view merges may also have
an affect from innovation rate. Dynamic efficiency is not very entering also today but things may
change in the future.

A recent decision by EU antitrust has cleared the 130 Billion Euro merger between Dow/DuPont
(agro-chemical sector) but only subject to structural remedies (i.e. the divestiture of major parts
of DuPont's global pesticide business, including its global R&D organisation) on the argumented
basis, then also formulated in the so-called “innovation theory of harm” (ITOH), that the eventual
merger (without remedies) would have led (in the view of the EU Commission) to a significant

102
decrease in the innovation rate among the involved parties that would exceed any possible gain in
the other efficiency dimensions.
The ITOH has triggered an academic (surely not “ivory tower”) debate on the merits of this (new)
approach in antitrust merger control, with many opponents, see Jullien/Lefouilli – Denicolò/Polo.

Mergers are scrutinized only if they involve sufficiently big actors (ex-ante or ex-post). See EU
thresholds in Appendix 1. Smallness is excluded from the attention of antitrust. Antitrust do not
scrutinize that kind of firm. If small firms want to merge, we will have a plus in productive
efficiency. We can also have a plus in allocative efficiency if the merge of the small firms reaches
the one of the bigger ones. Only merges that exceed a particular threshold would be scrutinized by
antitrust.

Is (ex-post) largeness automatically punished in horizontal mergers?

Not really (rule of reason). Every antitrust at every latitude, merged activity would never stop just
because it exceeds the threshold. But once it exceeds it, the regulator agency would investigate
and try to understand these points:
- Synergies? Economies of scale? Yes à plus on productive efficiency
- If synergies do exist can productive efficiency be transferred at least in part to consumers
(also more allocative efficiency in terms of lower price)? If the answer is yes, we have a
plus also in the allocative efficiency terms and the regulator would permit the merge

As a matter of fact, the main proceeding is the following:

103
SUM UP.
We started from the concept of trade and why it is beneficial for social welfare. We defined the
two form of static efficiency: allocative and productive. But there are market imperfections
(market power, externalities, asymmetric info, transaction costs) that help us to understand what
a firm is and its nature. Then we have dealt to firm growth and innovation (dynamic efficiency).
Given these strategies we analysed industrial and competition policy that try to maximize social
welfare.

104