Join one of the NEED-based projects of Let's Go Africa! Internship and volunteer opportunities in 12 different African countries.
Psychology - Pedagogy - Medicine - Sports - Psysiotherapy
This summary is based on the first edition (2010)
Industrial Organization is concerned with public policy towards business behavior.
We want to understand how firms interact, and how that affects social welfare.
We also want to understand the rationale for certain behavior, to be able to judge whether government should intervene.
The structure-conduct-performance (S-C-P) paradigm is the organizing framework of research in industrial economics. It consists of the following elements:
Basic conditions: Supply and demand.
Market structure: Number of sellers and buyers, barriers, cost structures
Firm conduct: Pricing behavior, product strategy, advertising
Market performance: Productive and allocative efficiency, full employment
In the neoclassical model of perfect competition the following assumptions are made:
Many small buyers and sellers
Complete and perfect information
Free and easy entry and exit
When looking at the market demand, there is a negative relationship between price and quantity. Total market demand equals p = f(Q)
A tool to measure the sensitivity of demand to price is the price elasticity of demand:
εqp = - p/Q * ∂Q/∂p
It is the proportional change in quantity demanded in response to a given proportional change in price.
Marginal Revenue (MR): The change in total revenue as a result of a small change in output. It is the revenue received from selling one unit of output extra.
∂(pQ) / ∂Q = p + Q * ∂p/∂Q
If we combine the equations of the price elasticity of demand and the marginal revenue we get an equation that shows the relationship between those two factors;
MR = p + Q * ∂p/∂Q = p(1+Q∂p / p∂Q) = p(1-1/ εqp )
Consumer surplus: The difference between the maximum amount consumers would pay for a given amount of output and the amount that they actually pay. In the figure below one can see the consumer and producer surplus.
A firm can have different kinds of costs:
Opportunity costs: The costs of using the factory’s services for its own production. Instead of using the factory itself the firm could also rent it to another firm in the market in return for factory services.
Average variable costs (AVC): Variable cost per unit of output
Average costs (c): Cost per unit of output
Marginal costs (MC): The change in total cost per unit change in output
Fixes costs (F): The costs of the services of fixed inputs
Sunk costs: The cost of investing in a tangible or intangible capital asset is said to be sunk if the value of the asset cannot be recovered by resale upon exit from the market.
The general equation of a firm’s cost function is;
C(q) = F + c(q)
Figure 2. Firm cost curves
In a perfect competitive market the firm maximizes profit when MC=P, assuming that P> AVC. If P
In the short-run equilibrium the number of firms that supply the market is fixed. The industry supply function is the sum of all the supply functions of the firms in the market. The short-run equilibrium price is the price at which the short-run industry supply curve and the demand curve intersect.
In a short-run time period firms can make economics losses or profits which affect the long-run equilibrium.
If P>average costs the firms make economic profits. This will attract new firms to enter the market. The industry supply will increase and therefore the industry supply curve will shift to the right.
Firms keep entering en exiting the market until the equilibrium price equals the minimum value of average costs. In this case all firms produce at p=mc and so they all maximize profits. Then there is no incentive for firms to exit the market or for new firms to enter the market, the long-run equilibrium is reached.
In a monopoly there is only one firm active in the market and it is impossible for other firms to enter the market. With perfect competition all firms are price takers but in a monopoly the firm is a price setter. A monopolist maximizes profits when MC=MR.
A monopolist charges a higher price and a lower output than would be in the long-run equilibrium of a perfectly competitive industry. People who buy the monopolized product pay now more than they would pay in a perfectly competitive industry. This extra amount is an income transfer from consumers to producers. Because of the higher price consumers who buy the product, have less income to spend on other goods. People who do not want to pay for the higher price lose the surplus value that they would have had in a perfectly competitive market.
This lost consumer surplus is called the deadweight welfare loss (DWL); the reduction in consumer welfare that is not balanced out by an increase in the income of the owners of the firm.
The DWL is the consumer surplus in the long-run equilibrium in a perfectly competitive market minus the sum of the monopoly consumer surplus and the monopoly profit. The costs to the consumers are the monopoly profits plus the DWL.
A generalized measure of market performance for a single industry can be written as:
W= θ1π + θ2CS
θ is a variable that indicates the weight of a factor to the market performance. When θ1= θ2=1 profits and consumer surplus are weighted equally in measuring market performance.
There is no theory in economics to indicate what the value of θ must be. They often assume that θ1= θ2=, because people argue that we treat the welfare of firm owners and consumer not differently.
Compensation principle: If a change in the market reduces one group’s welfare and increases the income of other groups, the government can restore the original income distribution by taxing the group of who the income has gone up and using the receipts to make the people who lost income as well of as before.
The Lerner index for market power: The proportional difference between monopoly prices and competitive prices. It is states as follows;
(p – MC) / p
In a monopoly a firm maximizes profits where MR=MC. The value of the Lerner index for this monopolist then equals;
(p – MC) / p = 1 / εQp
There could be a situation where there is a single supplier active in a market but another firm, which may be located in another region, will try to enter the market. The cost curve of the alternative supplier will probably be higher than the cost curve of the local supplier due to transportation costs. If the local firm charges a price above the price of the alternative firm and consumer have perfect information, consumers will buy the products from the alternative firm and the local firm sells nothing.
To prevent the alternative firm from entering the market, the local firm should charge the monopoly price that maximizes profits by charging the highest price possible that will make entry for the other firm not profitable. If the local firm does not succeed and the alternative firm will enter the market, the market will become an oligopoly.
If a firm has a good position of market power in the market the firm will try to keep this position. They will invest many resources because these make the costs of entry higher. As a result it is more difficult to entry for new firms. All these activities to keep or to reach a certain position in market power are called rent seeking. The resources devoted to this are a social cost of market power. Rent sharing is the payment of more than competitive returns to input.
Allocation efficiency arises if some industries are imperfectly competitive.
X-inefficiency: monopolist may not have an incentive to be as efficient as a perfectly competitive firm.
To see how demand and supply work in the real world we often use experimental markets. In these experiments redemption values are used to value each unit if output that can be supplied or demanded. In the experimental markets is looked at how short and long run equilibria are reached. Smith has done several of experiments. In one experiment he used the double action institution, which permits buyers to bid the prices they want and sellers to offer the prices they want.
The double action institution is often used in a model of perfect competition. Another institution that could be used is the posted offer institution. The supply side states at which price and quantity it will produce and then consumers will decide how much en from which supplier they are going to buy. Transaction prices eventually reach the competitive price and there are more trading periods necessary for the market to reach the equilibrium than with double action trading mechanism. The posted offer institution is especially used in the retail market.
Basic game theory
A game is a stylized model that depicts situations of strategic behavior, where the payoff for one agent depends on its own actions as well as on the actions of other agents. The choice of one player depends on what it expects the other players to do.
A game consists of:
A set of players. Who is involved?
A set of rules. Who moves when? What do they know when they move? What can they do (strategy space)?
A set of payoff functions. The utility each player gets as a result of each possible set of actions.
All this is common knowledge.
To get a solution of a game we use a prediction of how the game will be played between rational, utility-maximizing players.
A player has a dominant strategy whenever he has a strategy that is strictly better than any other strategy regardless of what strategy the other players will choose. If a player has such a strategy and is rational we expect the player to choose the dominant strategy.
Both players are better of choosing Defect instead of Cooperate. So choosing Defect is a dominant strategy for both players. This leads to the circled outcome. Both players have to be rational and have to assume that the other player is also rational. The point where both players choose Defect is called a Nash equilibrium.
When we look at the graph it would be more profitable for both firms if they both would choose to Cooperate. However that is not credible, unless the game is repeated. If you repeat the game over and over again the number of strategies increases because you then also can consider what has happened in the past.
A player should cooperate if everyone in the past has done so. If not the player should defect. Suppose the following game:
If you decide to always Cooperate, you will get 1 in each future period.
What happens if you Defect?
You will get 2 in this period. That’s good.
You will get 0 in each future period. Because the other player then will also play defect. That’s not so good.
Payoff of always Cooperate:
1 + 1∙δ + 1∙δ2 + 1∙δ3 + 1∙δ4 + 1∙δ5 + … = 1/(1 – δ).
Payoff of Defect:
2 + 0∙δ + 0∙δ2 + 0∙δ3 + 0∙δ4 + 0∙δ5 + … = 2.
So we have a Nash equilibrium whenever 1/(1 – δ) > 2 or δ > ½.
That is, whenever players are sufficiently patient.
This really is a Nash equilibrium. This is known as a trigger strategy: a single defection immediately triggers non-cooperation forever.
If the condition does not hold cooperation is not stable. Players know that if they would cooperate, everyone would immediately have an incentive to cheat. Hence, no one will cooperate in the first place.
The Cournot model
The most common model of an imperfectly competitive market is the Cournot model of quantity-setting oligopoly. In this model there are two firms active and each firm decides how much it will produce. Prices adjust so that the quantity suppliers supply equals the total quantity demanded.
In this model we assume that the firms know everything about the market, the product is standardized and the two firms have identical and constant average costs. The market inverse demand equals
p = a – bQ
Each firm acts independently and wants to maximize its own profits, given that the other firm is producing the equilibrium output. Each firm does what’s best for him given what the other firm does; this is called the Nash equilibrium:
First the best response function of both firms has to be determined. This function consists of all the pairs of outputs that maximize one firm’s profits for an arbitrary output level of the other firm.
When firm 1 has decided how much it will produce, there will be a part of the demand that still has to be supplied. The amount that is left for firm 2 to supply is given by the residual demand function.
The value of the Cournot equilibrium can be found by solving the equations of the best response functions of firm 1 and firm 2. This gives the coordinates where the two functions intersect. The total output is the sum of the quantities produced by firm 1 and firm 2.
In a Cournot model the Lerner index for market power, with ci marginal cost of firm i and si market share, is;
(p-ci) = si / εQp
The Herfindahl index (H) is the sum of squares of market shares of all firms in the industry. We use this index to calculate the industry-average Lerner index, with market share weighted marginal cost:
(p-^c^)/p = H / εQp
Cournot argued that firms choose an output level and that prices adjust to this. Many years after Cournot, Bertrand criticized these assumptions. He argued that in imperfectly competitive markets firms choose prices instead of output levels and that they sell the quantities demanded at those prices. Keeping all other things of the Cournot model the same, there is one other thing that also differs; the residual demand curve.
If both firms in the market charge the same price, consumers will buy random from a firm. In this case people buy from both firms and thus each firm serves halve of the market and they both have positive profits.
But if one firm charges a price that is slightly lower than the price the other firm charges, its profit margin will almost remain the same, while the sales of this firm will almost double. This leads to higher profits.
When each firm sets a price equal to marginal cost, to maximize profits, it assumes that all other firms also set prices equal to marginal costs. But if every firm does this they make zero profits. This is called the Bertrand Paradox.
With homogeneous products, the reaction curves (best-response curves) of firm 1 and 2 look like:
There are two kinds of product differentiation:
Horizontal differentiation: Some consumers prefer one variety, some prefer another variety. There is no universal preference ranking, there are differences in tastes.
Vertical differentiation: There is a universally accepted quality ranking of varieties. One specific variety is more preferred to another variety for all people.
If we combine horizontal differentiation with the price-setting duopoly model we get the inverse demand functions:
p1= 100 – (q1 + θq2)
p2= 100 – (q2 + θq1)
θ is a number between 0 and 1 that measures the degree of substitutability between the two varieties.
If θ=0, the quantity of variety 2 has no influence on the price of variety 1, and vice versa.
If θ=1, the varieties are perfect substitutes. The two demand functions become identical and reduce the homogeneous product demand function.
The reaction curves of differentiated products look like:
The prices are strategic complements for the firms; an increase in one price makes profitable an increase in the other price. But in case the firm does not state prices but it states output, then output is a strategic substitute for the firm; an increase in one making profitable a reduction in the other.
The Bertrand equilibrium is found by solving the reaction equations of the firms. The greater the degree of product differentiation, the greater the equilibrium price-cost margin. As long as products are differentiated, equilibrium price-cost margins fall as the number of firms rises.
Bertrand competition is more competitive than Cournot competition. The reaction curves (best-response curves) in Bertrand competition are upward sloping instead of downward sloping, like with Cournot competition.
We are going to take a look at a duopoly model where one firm (1) has an informational advantage over the other firm (2). This model is often called the model of Stackelberg. In this model the follower, firm 2, makes its output decision as he would do in the Cournot model and the leader, firm 1, knows this. Firm 2 produces the output that maximizes its own profits given the output produced by firm 1. Firm 1 knows how firm 2 will act and will therefor take this into account when deciding how much to produce. Firm 1 will produce another output than it would produce as a Cournot duopolist. The reaction function of firm 2 in the Stackelberg model is different from the reaction function in the Cournot model. An example:
If the inverse demand curve is:
p= 100 – (q1 – q2)
and the cost function is:
c(q) = 10q
Then the reaction function of firm 2 is:
q2= 45 – 0.5 q1
The residual demand curve of firm 1 will be: p = (100 – q2) – q1
If we substitute the reaction curve of firm 2 in this equation, the residual demand curve of firm 1 will be:
p = (100 – (45 – 0.5q1)) – q1 = 55 – 0.5q1
To maximize profits, firm 1 will set Marginal costs equal to marginal revenue.
MR= 55 – 2(0.5q1) = 55 – q1
MR = MC gives 55 – q1 = 10 q1 = 45
If we know q1 we can calculate q2 q2 = 45 – 0.5 x 45 = 22.5
The total output produced under the Stackelberg leadership is 45 + 22.5 = 67.5
And the equilibrium price is 100 – 67.5 = 32.5
The Stackelberg output is higher than in the Cournot equilibrium and the price is lower. The total profits made exceed the profits made in Cournot equilibrium.
The equilibrium profit in imperfectly competitive markets is lower than the maximum monopoly level. To come closer to the monopoly level, firms can raise prices or restrict output. Output restrictions can be a result of collusion. Collusion is a collaboration between firms to act together as one firm to maximize profits. It is prohibited by the European law.
If output restrictions are the result of an agreement between independent firms, the firms have colluded. If the restrictions are the result of independent decisions by independent firms, the firms have not colluded.
There are different forms of collusion. One form is tacit or non-cooperative collusion. It includes a tradeoff between current and future profits. If the industry output is less than the non-cooperative output level and the firm increases output, rivals will notice this and will eventually do the same. This results in a lower price and all firms will have lower profits than before instead of higher profits.
So before increasing its output a firm must consider how much profit it loses if all firms increase output, how fast rival firms will follow and how much the profit loss in the long run is relative to profit raise in the short run.
The trigger strategy is used to model the tradeoff between payoffs received at different times. It states that:
Each firm produces its share of monopoly output in the first period and continuous doing this in the following periods if all firms do so.
If in a period monopoly price is not equal to the period price, the firms will produce at their Cournot output for the rest of the periods.
The following equations are given:
P = 100 – Q and c(q) = 10q MC=10
Then the payoffs are:
The monopoly profit per period ( with p=55 and q=45) is;
π= (55-10)(45) = 2025
Each firm produces half of the monopoly output (45/2=22.5) and therefor earns half of the monopoly profit = 2025/2= 1012.5
If the Cournot price = 40 and the equilibrium output = 30, then the profit per firm in this Cournot duopoly is:
π = (40-10)(30) = 900
If firm 2 produces half of the monopoly output, so it produces 45/2=22.5. Then the residual demand curve of firm 1 will be:
p = (100 – 22.5) – q1 = 77.5 – q1. This leads to the MR curve equal to 77.5 – 2q1
Firm 1 wants to maximize profits and sets MR=MC 77.5 – 2q1 = 10
This gives q1= 33.75. This output is higher than the monopoly output 22.5
The price will be p = 77.5 – 33.75 = 43.75. This price is lower than the monopoly price 55.
The profit for firm 1 is (43.75 – 10)(33.75) = 1139.0625. This profit is higher than the profit made if both firms produce the restricted output, 1012.5.
The extra profit made by firm 1 only holds for one period. To measure the consequences of this expansion of output one can use the present-discounted value. The equilibrium present-discounted value of a firm under trigger strategy is:
Vts = 1/(1+r) πm + 1/(1+r)^2 πm + … = 1/r πm
If all firms follow the trigger strategy, then after one period of high profits all firms will revert to produce their one-period Cournot equilibrium output forever. The defecting firm will also produce this output. The present-discounted value of the defecting firm is:
Vbrf = 1/(1+r) πbrf +1/(1+r)^2 πCournot
The trigger strategy stability condition is the condition that must be satisfied in order for output restriction to be a value-maximizing strategy for a firm. This stability condition is:
(1/r) >= (π(brf) – π(m)) / (π(m) – π(Cournot))
There could be uncertainty about the location of the demand curve. There can be unexpected shocks in the demand curve due to changes in consumer tastes or changes in other industries. Assume that the demand is for instance 100 and the value of the shock is equal to ε. The new demand is then 100 + ε or 100 – ε. Green and Porter included this uncertainty in the trigger strategy. The extended version consists of three parts:
The outputs produced by each firm
Threshold price: price below expected price and above lowest possible realized price, pts+ ε
In this generalized trigger strategy, if the realized price is equal or below the threshold price, firms revert to Cournot behavior for R periods and restrict output again. If the equilibrium conditions of this generalized strategy are met no firm will ever defect.
A business cycle is a pattern of economic activity characterized, at the level of a single market, by a series of demand shocks in the same direction.
A boom: Series of high demand
A slump: Series of low demand
A generalized trigger strategy allows firms to restrict output but it does not restrict output to the profit-maximizing level. Then collusion is used. We here assume that collusion involves the spending of money to put a private enforcement mechanism in place. With collusion cost K per period. The collusive value of a single firm is:
Vcol = (π(m) – K) / r
A firm has to options in an uncertain environment:
Tacitly collude: sometime go through periods of output expansion and low profit
Explicitly collude: keep profits high while helping to pay the cartel’s monitoring and enforcement costs. This one is more likely to be profitable.
It can be that a public policy prohibits collusion and imposes fines on the firms caught of collusion. Then the firm’s collusive value is:
Vcol = (π(m) – K – τ1τ2F) / r
F is the fine imposed
τ1 is the probability that collusion is detected
τ2 is the probability that the colluding firms are convicted if collusion is detected
τ1 is larger when there are many resources of enforcement agencies. τ2 depends on the nature of the legal system. When a policy imposes fines, the collusion value will be lower than without fines.
If demand is stable then collusion is stable. If demand is unstable, firms may increase their value by colluding even though there are costs associated with it. There are many different factors that affect the stability of a cartel. We distinguish between external and internal factors.
These factors relate to the firms that are part of the cartel.
Cost differences make it more difficult for cartel members to come to an agreement. This is because the members then have different outputs that make MR equal to MC. Firms with high costs make larger losses if the price goes down. When all costs among the members are the same the cartel is more likely to be successful.
Product differentiation. Wide product differentiation often leads to different views on what the members should do. Firms that produce high-quality varieties will probably favor higher prices. While firms that produce low-quality varieties will probably favor lower prices. Because of this a price schedule is necessary. The price will be somewhere in the middle of what both groups want. So a cartel is more favored when the product is relatively standardized.
Rates of time preference. A low discount rate facilitates tacit collusion. If the rates of time preference between countries are about the same, they are more likely to agree on a way to trade off present and future income.
Multimarket contact. If firms are diversified and meet in many different markets tacit collusion is more likely.
Seller concentration. If there are many firms in an industry it is easier to reach a common view on what should be done. Joint exercise of market power is more likely to happen when seller concentration is high.
Vertical integration. Vertical mergers make it easier to facilitate tacit upstream collusion.
These factors relate to the market conditions external to the cartel.
Buyer concentration. Output restriction can be affected by the demand side of the market. Large buyers may bargain low prices and force other firms to cut price which results in a rivalry. This theory is called the theory of countervailing power. If firms in an industry buy and sell inputs from/to many small suppliers tacit collusion is more likely to be stable.
Entry. The entry of new firms in the market destabilizes collusion. Entry and expansion of output form firms outside the cartel will lead to cartel collapse. So if entry costs are high, new firms are not likely to enter the market and collusion is more likely.
Industry growth rate. If the industry growth rate is low it is not attractive for new firms to enter the market. Thus collusion is then more likely than with a high industry growth rate.
Size and frequency of individual transactions. If there are frequent and relatively small sales in an industry, the profit gained by short-run defection from output restriction will be relatively small. Collusion is likely to be stable when sales are small and frequent.
Open bidding and other types of public policy. If a firm defects from collusion and a government buyer reveals the author of the low bid, it also exposes the low bidder to immediate retaliation. When there is open bidding collusion is more likely.
The Stackelberg quantity leadership model can be used to model entry deterrence. But another model can be used for this, namely the model of limit pricing. In this model the entry is limited; an incumbent firm makes choices that affect the incentives of potential new firms to enter the market. We set a model in which firm 1 is an incumbent firm and firm 2 is a potential entrant. If a firm wants to enter the market it deals with a fixed and sunk entry cost e. We take that the inverse demand curve is p = 100 – (q1 + q2).
When a firm enters the market its profit depends on the output of the incumbent firm, firm 1.
π2 = (p-10)q2 = (90 – q1 – q2)q2
After the entry the market will be a non-cooperative quantity setting duopoly. The entrant will maximize profit and therefore produces the output given by its best response function:
q2 = 45 – 0.5 q1
p – 10 = 90 – q1 – (45 – 0.5 q1) = 45 – 0.5 q1
The discounted present value of the entrant is:
V2 = 1/r (45 – 0.5 q1)2 – e
The first part of the equation represents the all profits in the future discounted at rate r. The second part represents the sunk entry cost that the entrant must pay to enter the market.
An incumbent firm can commit in advance the output it will produce. The commitment could be effective if the firm can choose a technology that makes its cost fixed and thus its marginal cost almost zero. The firm can produce at the highest possible capacity level.
The incumbent firm can choose a level that makes profits for the entrant in the new market zero to keep the entrant out of the market. This output is called the entry-deterring output.
V2 = 0 1/r (45 – 0.5 q1)2 – e = 0 q1 = qL = 90 - 2√re
To determine whether entry-deterring is profitable depends on the entry cost. We distinguish between two extreme cases:
Blocked entry. In this case the entry is blocked and the market is a natural monopoly. Entry cost is so large that the entry-deterring output is less than the monopoly output. The incumbent firm will then produce the monopoly output and it is not profitable for the other firm to enter the market.
Contestable market. In this case entry is costless and the market is a contestable market. The incumbent firm is forced to produce the output that would be produced in the long-run equilibrium of a perfectly competitive market.
In-between. There is also a case between these two extremes. The incumbent firm can decide to produce slightly more than the limit output so that the entrant will stay out of the market.
pL = 10 + 2√(re)
When entry costs increase, the entry-limiting price pL will also increase.
πL = (pL – 10) qL = 2√re (90 - 2√re)
The discounted present value of the incumbent firm if it deters entry is:
VL = (2√re (90 – 2 √re)) / r = 2√ (e/r) x (90 - 2√re) = 180 √ (e/r) - 4e
An incumbent has another choice besides entry-deterring. It could let the entrant enter the market and then compete as a Cournot duopolist. The Cournot duopoly value is 9000.
If the entry cost is 350 and the interest rate r is 1/10, then the entry-deterring output level is:
qL = 90 – 2√(350x0.1) = 78.17
And the present discounted value, if it deters entry, is 180 – √(350/0.1) – 4 x 350 = 9248.9
This value is higher than the Cournot present discounted value. So in this case it is profitable to deter entry.
But entry cost could also be lower, for instance 300. Then the entry-deterring output level is:
qL = 90 – 2√(300x0.1) = 79.05
And the present discounted value, if it deters entry, is 180 – √(300/0.1) – 4 x 300 = 8659.
This value is lower than the Cournot present discounted value. In this case the incumbent should not deter entry but instead it should accommodate entry.
In the previous cases we assumed that there is perfect and complete information. However it can also be that the entrant is uncertain about what kind of market in enters or which incumbent it will face in the market. In this case actions of the incumbent can deter the entrant to enter.
Suppose that e=8000, r=1/10 and that the marginal cost of the entrant is 10. The incumbent could have two kinds of marginal cost:
The marginal cost of the incumbent can be 10. If the firm then enters the market, there will be a Cournot duopoly with the firms having the same marginal cost. In this case the present discounted value of the entrant when it enters is positive and therefor the firm will enter the market.
The second possibility is that the marginal cost of the incumbent is 1. If the firm then enters the market there is a Cournot duopoly but this time with unequal marginal cost. Because the marginal cost of the entrant is much higher than the cost of the incumbent, the entrant will have a negative discounted value when it enters. Thus the firm will not enter the market.
Suppose the marginal cost of the incumbent is 10 but there is imperfect information so the entrant does not know this. The incumbent is a monopolist in the pre-entry period and will produce the monopoly output. After this period the cost of the incumbent is revealed and so the entrant knows what the cost is. Because the cost of the incumbent is the same as of the entrant, the entrant will enter the market and the market will become a Cournot duopoly.
Another case is that the incumbent still has marginal cost of 10 but in the first period it produces as if it has a marginal cost of 1. He produces as a monopolist at cost of 1. The entrant does not gain any information in this case because it is still not totally clear what the cost of the incumbent is. The entrant’s uncertainty about the kind of cost the incumbent has make it profitable for the incumbent to expand output and to deter entry.
A dominant firm that follows a predatory pricing strategy cuts price below rivals’ average cost to drive rivals from the market. Both firms will make a loss; the loss of the dominant firm is even higher than the loss of the following firm. This action of the dominant firm is called a predatory campaign. When a firm does not survive a predatory campaign usually the consumer will become the victim.
Once rivals go out of business, the predator raises price and collects enough economic profit to more than balance out any short-term losses. This strategy is especially helpful if it gives the dominant firm a reputation that it is tough and always fights entry.
Predation is a strategy that is used after entry has occurred – but it can also serve to deter future entry. For this to occur, we need that the predator has more financial resources to survive a price war than the prey (long pocket).
Selten’s chain-store paradox argues that predation does not work in a market with perfect and complete information.
When a firm does not want to deal whether there will be predation or not it has three options. It can choose for collusion, for the purchase of a rival or for a third alternative. This third option is to impose higher costs on the rival which lowers the price and lowers the cost of the dominant firm. This theory is called the theory of raising rival’s costs (RRC). As a result of the higher cost, the rival will probably lower output. If the cost is high enough it can even shut down the firm entirely. The dominant firm will quickly expand output or raise its price to gain from the cost rise of the rival.
It may even be profitable for the dominant firm to accept some increase in its own cost, if this can result in an even larger cost increase on smaller rivals.
Price discrimination: Identical units of a good or service are sold at different prices, either to the same customer or to different customers. There are three types:
First-degree price discrimination.
The firm charges each consumer his highest amount she would be willing to pay. This is called the reservation price of the consumer.
First-degree price discrimination is more possible in theory than in the practice. When it is possible it allows a monopolist to produce the maximum output possible. Consumers dot not lose or gain from this because each one of them pay its own reservation price. There is no output restriction in this market but the output increase of the monopolist causes output restrictions in other markets. Even though there is output restriction in other markets, the total welfare increases.
Second-degree price discrimination.
The firm cannot distinguish between different types of consumers, but designs products in such a way that output is divided in lots. Each lot being sold at the highest price at which the whole lot will be purchased.
Businesses use fixed price structures and product characteristics in such a way that they can gain information about the mind construction of the utility-maximizing consumers. So they can find out what the upper limit is that a consumer would pay for a particular type of product. Usually when a new product is introduced the price is high but over time the price will fall and stabilize. If the production of the product will increase the marginal cost will decrease and thus the firm can charge a lower price. A reason for this price to drop in time can be intertemporal price discrimination.
An example: there are two groups of consumers, techies Nt and others Ne. Techies get utility αt and others αe and there are two time periods. Techies want to buy the latest cellphones so for this product αt> αe. A cell phone is a durable good and therefore each consumer buys at most one. If a techie buys a cell phone in period 1 his net discounted utility is:
αt + α(t) / (1+r) – p1 = (2+r) / (1+r) αt – p1 the reservation price is Rt1 =(2+r) / (1+r)αt
The reservation price for non-techies is calculated with the same formula:
Re1 = (2+r) / (1+r)αe
The price that techies are willing to pay is higher than the price of non-techies. So in the first period techies will probably be able to by the cell phones at the reservation price of non-techies.
If a firms sets a price to maximize value to sell to all consumers, its value will be:
V1 = (2+r) / (1+r) αe (Nt + Ne)
Because techies can buy at a lower price than they would be willing to pay they enjoy consumer surplus:
St1 = (2+r) / (1+r) (αt - αe ) Nt
When we sum up the firm’s value and the techies’ consumer surplus we get the net social welfare:
NSW1 = (2+r) / (1+r) (αt Nt - αe Ne )
Techies will buy in the first period if the net surplus they get by doing so is at least as great as the discounted surplus from waiting one period. If the price in the first period is higher than the reservation price of non-techies the techies will not buy in the first period. They will wait until the second period when prices are reduced. Both consumer surplus and NCW would be greater without intertemporal price discrimination. So intertermporal price discrimination is privately profitable but socially harmful.
Of some products there are different qualities of the same product. For instance, in an train a firm can offer seats in the first or second class. Each class has its own price so that consumer can choose. A train company can choose to offer only one class but it can also choose to offer both classes.
Third-degree price discrimination.
The firm can distinguish between different groups, can charge them different prices and there is no consumer arbitrage (resale from one group to another). Consumer arbitrage is impossible for goods sold in different regional markets due to transportation costs.
Each different group of consumers has a different price elasticity of demand and therefor will a monopolist find it profitable to charge each group a different price. Consumers with high elasticity are charged lower price than those with low elasticity. They are more sensitive to price.
Each consumer group has its own inverse demand curve. These curves form together the combined inverse demand curve. For instance:
p1 = 200 – 2Q1 and p2 = 100 – Q2
Together they form Q = (200-p)/2 + 100 – p
A monopolist will charge a higher price to one group and a lower price to the other group compared with a case of no price discrimination. So there is a group that loses consumer surplus but there is also a group that gains consumer surplus. All the profits and consumer surplus that result from third-degree price discrimination will increase social welfare.
Price discrimination makes it possible to supply to small markets that otherwise would not have been supplied.
If you want to switch from one software to another software you have to pay some costs for this, called switching costs. In this case these costs are inherent in the nature of the product. However, the costs can also be rooted in product design or be contractual. When you have a DVD-player and wants to buy a blu-ray means that you not only have to pay for this new player, you also give up the ability to watch all bought DVDs. Switching costs create some kind of loyalty and gives suppliers some market power over consumers who would have to pay when switching to another supplier. Because of the increased market power the firm can charge a higher price.
A firm can offer repeat customers a loyalty discount. The firm can attract consumers with this. In the first period the firm will price all reservation prices but in the second period it offers a discount to the repeat group en therefore builds op market power. A firm will have the incentive to compete more in the first period than in the second period, because in the first period it can ‘win’ the most consumers.
There are two types of network externalities:
Direct: The utility you get from using a product increases the number of other people using it. For instance, every additional telephone that is added to the network creates an additional possibility of a connection.
Indirect: The product itself does not generate utility. It is only useful with other complementary goods. For instance, a DVD-player cannot give you utility or be used without DVDs.
A firm can make a commitment to potential producers of compatible products. He wants the potential producers to enter the market so that the supply of the compatible product will increase. This will also increase the production of the firm itself because there are not more compatible products available on the market.
Many argued, like Karl Marx, that the control of production is concentrated in the hands of a only few producers. However modern economists do not agree with this. They argue that there will be a high seller concentration in technological industries but they also say that this could change. The market structure between countries and industries can differ very much.
The market structure in developed countries has changed over time. The capital-intensive manufacturing sectors have increased in the developed countries. This allowed the countries to shift resources to labor intensive, high-value added service industries. In the US the value of the manufacturing sector decreased while the value of the service sector increased. In the EU we see the same pattern in this period.
The side of the market structure that is mostly analyzed was the supply-side. By analyzing this we can explain differences of firm concentration in different industries. These explanations can be categorized into three parts:
Entry conditions. Entry conditions are directly based on supply and demand conditions. Markets where entry is difficult are concentrated and markets where it is easy would be unconcentrated.
Organizational capabilities. According to Chandler a market will become concentrated if there are certain technology or demand characteristics and there is a potential efficiency advantage to large-scale operation.
Endogenous sunk costs. Sutton argues that some industries depend on R&D, product design and marketing. In some markets a firm can create a strategic advantage. Markets in which vertical product differentiation can be improved by marketing usually only have a few large firms to contribute to the development of new products.
We take a simple model of Cournot competition with identic firms and with:
P = a – Q
c(Q) = F + cq + dq2
MC(q) = c + 2dq
AC(q) = F/q + c + dq
d reflects the diseconomies of scale. A firm wants to produce at minimum efficient scale. This is the level of output that gives the lowest cost per unit. It is where the marginal cost is at its minimum, or where marginal cost crosses average cost. This output level can be calculated with the following formula:
qmes = √(F/d)
If we substitute this equation in the average cost curve we get:
ACmin = c + 2√(dF)
But when there is more than one firm in the market, each firm will maximize profits by setting MR equal to MC. The best-response curve for firm 1 is:
q1 = (A – c – Q-1) / (2(1+d))
We assumed that firms are identical and therefor they produce the same output in the equilibrium. The output produced in this Cournot equilibrium is:
q = (A – c) / (n + 1 + 2d)
The profits in this equilibrium are:
ΠCour = (1 + d)((A-c) / (n + 1 + 2d))2 – F
In stage 1 firms will enter the market as long as is profitable. The number of firms in the long-run equilibrium is:
nCour = (A-c) √((1+d)/F) – (1 + 2d) = (A-c) / qmes √ ( 1 + (1/d) ) - (1 + 2d)
The equilibrium number of firms:
increases in the size of the market A;
decreases in the minimum efficient scale qmes;
decreases in diseconomies of scale d.
So far we have assumed that the firms are single-plant. But what happens if there are multiplant firms? Suppose the cost function of a single-plant is the same as before, namely c(Q) = F + cq + dq2. If a firm operates at two identical plants, so they have the same cost function, the firm-level cost function will be:
Q(q) = 2 [ F + c(q / 2) + d(q/2)2 ]
When producing at two plants, it is always most efficient to divide production equally between the two plants. The firm minimizes costs by opening up another plant. In the graph below you can see the average cost curve with optimal number of plants. It becomes flatter as output increases. The AC is measured on the y-axis and the output Q is measured on the x-axis.
Mergers have impact on market and firm structure. Mergers are created because people want corporate control. There are three kinds of mergers:
Horizontal mergers; Such a merger can create a market leader or concentrate the supply in the hands of only a few firms.
Vertical mergers; This is a strategic merger to be as cost-saving as possible and to let the costs of the rival rise.
Conglomerate mergers; These mergers can allow a firm to diversify risk by evening out fluctuations in income.
Over the last centuries the US economy has had a lot of mergers. All these mergers are divided into five different merger waves:
The first wave: The turn of the 19th century
The first wave began in 1893 with a depression, peaked in 1901 when United States Steel was formatted and ended in 1903 with another depression. The mergers in the first wave were especially horizontal mergers. There are a few factors that led to the first wave:
There was market integration and although this already had started a decade before this wave, it still contributed to the wave.
There was intense price competition. Firms merged together to avoid the big disadvantages of this.
Some firms wanted to escape the pressure of competition and merged to avoid this.
Industrial securities had become an effective device to raise large amounts of financial capital form the private sector. This made it possible to carry out mergers without using a high amount of cash.
The second wave: The roaring 20s
The second wave started after World War 1 ended. The market was recovering from a great depression. It ended with the stock market crash in 1929. In this wave declining-dominant firm industries were transformed into oligopolies. There was a lot of technological change that made competition more intense.
The third wave: The conglomerate 60s and unconglomerate 80s
A significant different thing compared to the waves before is that the third merger wave only consists of conglomerate mergers. An explanation for this is that there was a US antitrust policy in this period. Firms were seeking for market power and decide to merge.
The fourth wave: Unconglomerate 80s
The fourth merger wave undid the conglomerate mergers of the third wave. The conglomerate mergers were value-maximizing in the third wave but something changed over time so that it was not profitable anymore. In this period the financial markets also changed, they became more efficient.
The fifth wave: Turn of the 20th century
Somewhere early in 1990 there was a very low point of merger activity. This activity slowly rose from then to 2001. After this it dropped and then rose again. It ends with the worldwide collapse of financial markets in 2008.
If we look at horizontal mergers in a Cournot model, we can derive an inverse demand function:
P = a – b (q1 + q2 + …. qm + qm+1 + … + qn)
The best-response function for firm 1 is:
2q1 + 2q2 + …. qm + qm+1 + … + qn = (a-c) / b
Equilibrium price an output are:
P = c + 1 / (n + 1) * (a – c) / b
q = 1 / (n + 1) * (a – c) / b
The profits for the firm in equilibrium are:
π = b ( 1 / (n + 1) * (a – c) / b )2
A merger might be profitable for a survivor firm if it is less efficient than the parent firms. But such a merger is likely to reduce social welfare and consumer welfare. If the survivor firm is more efficient than the parent firms, then the merger could be profitable.
As we saw earlier, in a Cournot model, you benefit if you can commit to be more aggressive.
But a merger only commits you to be less aggressive
Adam Smith said that trade is profitable to all the parties that are involved in the trade. It allows countries to specialize in a specific product and exchange the surpluses of this product this with foreign countries. The country would gain from this trade. Smith also introduced the invisible hand; leading individuals to show the interest of the society while thinking they only pursue their own self-interest. The classical theory began with the assumption that product markets are perfectly competitive and relies on the exercise of market power in the export market.
The comparative advantage theory explains the exchange between two countries of two goods from different industries. But there is also trade of two-way flows of goods from the same industry. This kind of trade is called intraindustry trade and is a major part of total trade. It especially has a major part among industrialized countries.
Intraindustry trade can only occur if the monopoly price-cost margin is higher than the transportation cost from one market to another. Suppose we have a market with a Cournot duopoly model with firm 1 based in country 1 and firm 2 based in country 2. The price in each country only depends on the sales in the country itself and therefor output decisions are also taken separately.
When firm 2 sells in country 1, it has to pay production cost and transportation cost. The greater the transportation cost, the smaller the output that firm 2 will export, the closer the down-sloping best-response curve of firm 2 will be to the origin. If the costs are so high that the curve of firm 2 is below the best-response curve of firm 1, then firm 1 will sell the monopoly output. So a condition for intraindustry trade to take place is that the transportation cost must be lower than the monopoly price. If this condition holds and the countries trade with each other, then this will give a lower equilibrium price which results in a consumer surplus for country 1. Although country 1 experiences this consumer surplus, it is worse off in its home market.
A firm can also be a price-setter instead of a price taker. With price-setters we usually look at products that are imperfect substitutes. The price best-response curves slope upward. We are still looking at the same firms in country 1 and 2. If transportation cost is high the price firm 2 will charge is also high which leads to a higher best-response curve for firm 2. The domestic firm exercises less market power if there are foreign competitors. Intraindustry trade leads to more welfare for consumer when products are differentiated and firms are price-setters. If a country is involved in trade the consumers have access to both domestic and foreign varieties which increases welfare.
A country can apply a strategic trade policy by giving a subsidy to the exporting country to cover the transportation cost. A subsidy has the same effect on consumers as an output restriction. When firms produce strategic substitutes, the direct and strategic effects of the subsidy are to increase the profit of the subsidized firm at the expense of rivals.
In a model with price-setting firms, the firm that receives a subsidy can change its behavior which results in a reaction of the other country. This leads to a gain for the subsidized firm and therefor indirectly benefits the country that is granting the subsidy.
A government should choose wisely which industries to subsidy and which not. Because if it subsidies the wrong industry it can be worse of in the end. However, the government does not always seek for social welfare.
A country can also impose tariffs and quotas on exports and imports. It are inward looking policies, they are undertaken because of their effect on the home market. This is that quotas and tariffs can protect domestic producers form foreign competition, increasing their profit and leaving consumers worse off.
We will now look exchange rate fluctauations. We assume that the currency of country 1 is the dollar and of country two the euro. The exchange rate reflects the number of euros needed to buy a dollar. The exchange rate does not affect firm 1’s profit on sales in country 1. However, it does affect the profits of firm 2 supplying in country 1. When the exchange rate increases, a depreciation of the euro, the unit cost of firm 2 will decrease. The best-response curve of country 2 will shift outwards. Only a small part of exchange rate fluctuations can be seen in domestic prices.
Concentration effect: Lower price-cost margins within intraindustry trade raise the possibility that the equilibrium number of firms will fall because of free trade, economizing on fixed costs.
There are different options to supply in a foreign market:
Licensing a foreign firm to produce the product in a foreign market.
Establishing a joint venture with a local partner.
Purchasing an existing plant in a foreign market.
Setting up a new plant in a foreign market (greenfield investment).
Two countries can form an export cartel to make export easier but most policies forbid collusion.
Dumping: When a foreign firm sells its export products at a price that is below its home market price. It is often claimed that dumping should be a policy concern because dumping is unfair. Some countries use an antidumping policy to counter this.