This chapter discusses methods for measuring market concentration, and develops models that are useful for evaluating entry issues. In particular, we can measure the "competitiveness" of an industry by looking at the existing market structure (concentration), and by looking at conditions of entry (potential competition).

Defining the Market: All produces or enterprises with large long-run cross-price elasticities of either supply or demand should be combined in the same market.

But what about potential entry, as in airlines, where actual entry has not happened and so no elasticity data are available? Scherer argues that we can also include existing capacity that can be shifted in the short run.

Measuring Market Concentration:

Concentration Ratio: N-firm concentration ratio is the sum of the market shares of the N largest firms (largest in terms of sales shares in the market). Problem: One gets the same 4-firm concentration ratio from a structure where the 4 largest have equal 20 percent shares, and one in which the 4 largest have 77, 1, 1, and 1 percent shares.

HHI: Sum of squared market shares of all firms in the market…. Compute for case above….

Remember the Structure-Conduct-Performance Paradigm: Cournot theory, collusion theory, and empirical evidence suggests that highly concentrated markets tend to have prices elevated above marginal cost. Demsetz: Differential Efficiency Hypothesis: Concentration may not cause high price-cost margins. More efficient firms (cost or quality) naturally achieve large market shares, and the cost or quality advantage allows firms to charge a premium above their marginal cost. This latter result can be generated in a Cournot setting where firms have different marginal costs. Note, however, that if the Differential Efficiency Hypothesis is correct, then antitrust intervention will not enhance efficiency. Empirical evidence tends to support the Differential Efficiency Hypothesis.

Scale Economies are another reason why some markets are more concentrated than others. Remember natural monopoly and the reason why it exists….

Entry Conditions: Entry conditions determine the extent of potential competition; existing competition is a function of current prices, and if prices rise, potential competitors become important.

Issues Associated With Potential Entry:

• How many prospective firms have the ability to enter in some reasonable period of time?
• How long does it take to enter the industry?
• How costly is entry?
• Will a new firm be at a disadvantage relative to incumbent firms?
• Will potential entrants have access to the same technology, products, and information as the incumbents?
• Is it costly to exit?

Examples of Entry Costs: Costs of acquiring a production facility, licenses, advertising and distribution channels, expertise, etc.

Noncooperative Oligopoly Equilibrium With Entry

Suppose we have an identical-firm Cournot industry with entry cost K. Then the equilibrium number of firms N is satisfied by:

Profit(N)/r - K > 0 > Profit (N+1)/r - K

Note that this does not include any of the usual incumbent advantages we usually associate with real-world industry.

Barriers to Entry

Traditionally, industrial organization (IO) suggests that serious and persistent market power occurs when (i) markets are sufficiently concentrated that it is relatively easy for firms to monopolize or collude, and (ii) there are high "barriers" or entry costs that prevent new competitors from entering and competing away rents.

But what is a barrier to entry? Controversial issue. Father of IO Joe Bain: A barrier to entry reflects the extent to which, in the long run, established firms can elevate price above minimum average cost (marginal cost) without inducing potential entrants to enter the industry. Examples: Scale economies, government-induced restrictions like patents and licenses, and access to lower-cost inputs or technology or simply higher product quality.

Note that Bain's definition of an entry barrier can be consistent with Demsetz' Differential Efficiency Hypothesis.

Stigler: An entry barrier is a cost borne by entrants and not by incumbent firms-differential costs. Examples include the advertising and other costs associated with launching a new brand in a consumer products market. Thus the existence of "brand loyalty" may serve as a type of barrier to entry.

Note that fixed costs are not usually seen as a barrier to entry, because in the long run incumbent firms must also replace their capital.

The Theory of Contestable Markets and Sunk Costs:

The theory of contestable markets is that if entry is costless, then market power will not occur even if there are such enormous economies of scale that a single firm dominates the market. The reason is that, in the extreme, contestability assumes that an entrant can come in and take away the entire market from the incumbent if the incumbent exercises market power. This perfect contestability assumes that entrant firms have access to (i) the same production technology, inputs, and information as the incumbent(s), (ii) ZERO sunk costs (costs that cannot be salvaged upon exit), and (iii) the entry lag is smaller than the price adjustment lag (i.e., a new firm can enter before the incumbent can lower price). Perfect contestability is pretty unrealistic, but if it exists, then the market equilibrium will be efficient regardless of the number of firms. Contestability extends the perfect competition result to all other market structures.

A key problem is that almost always there are sunk costs, which cannot be salvaged upon exist, implying an exit cost. The existence of sunk costs imply exit costs, which in turn implies a risk to entry, and thus can be considered an entry barrier. Example: New Coke.

Even though perfect contestability is unlikely, it has refocused economists and antitrust authorities on issues associated with entry, and reduced emphasis on static market concentration analysis.

Dominant Firm Theory: Much more common to have a dominant firm in a particular U.S. industry than to have a pure monopolist. So how would a dominant-firm model work?

Static Analysis: von Stackelberg model:

Assume we have a "competitive fringe" that are price-takers, and thus their output is reflected along a supply curve. The dominant firm, however, monopolizes that part of market demand that is residual of the market served by the competitive fringe. Moreover, assume the dominant firm acts strategically, and knows how the fringe's supply behavior will react to his output and price.

The dominant firm's residual demand is Dd(P) = D(P) - S(P).

Define P(hat) to be the "choke-price" at which the fringe supplies the entire market, implying that Dd(P(hat)) = 0. As price falls below P(hat), the dominant firm's residual demand D(P) - S(P) grows.

Assume further that the dominant firm has a cost advantage over the fringe, and so at P0, S(P0) = 0, at which point the dominant firm monopolizes the entire market demand. Thus the residual demand starts at the vertical intercept at P(hat), and becomes the market demand curve at P0. The residual demand has its own MRd curve, and the dominant firm's optimal output occurs where MRd = MC. Note that the price is lower than under pure monopoly.

Dynamic Analysis: Limit Pricing Model

Dynamics are important. For example, US Steel's dominant position eroded from 65 percent in 1901 to 24 percent in 1967. In other cases, like AT&T, dominant positions can persist. Why?

The size of the fringe changes over time. For example, if fringe firm capacity is funded through retained earnings, then the higher is the dominant firm price (the market price), the faster will the fringe increase their capacity and the sooner will the dominant firm lose its dominant position. Thus the optimal price path for the dominant firm takes into account the effect of today's price on future fringe capacity (and thus on future price and dominant firm profit).

Define a limit price to be a price sufficiently low as to limit fringe expansion is zero. Assuming that the dominant firm has a marginal cost advantage, at some point in fringe expansion it will pay to set the limit price.

The dynamically optimal price path is a downward trajectory that eventually hits the limit price, and depends on factors such as discount rate and the capacity of fringe firms.

Strategic Entry Deterrence:

Limit-Entry Pricing: Bain-Sylos Model

Bain-Sylos Postulate: The entrant believes that, in response to entry, each incumbent firm will continue to produce at its pre-entry output rate.

Assume a cartel or monopoly. The incumbent(s) can manipulate the residual demand by way of its pre-entry price and output. In particular, a higher incumbent output (and lower price) would imply a smaller residual demand. If the incumbent can shrink down residual demand (by lowering price) until it is below the potential entrant's AC curve, then the price at which residual demand is tangent to the potential entrant's AC curve is the limit price.

Problem: The Bain-Sylos model assumes that the potential entrant believes that the incumbent will continue at the limit price and output level even if entry actually occurs. But the incumbent firm(s) lack credibility-there is nothing to commit them to the limit price in a post-entry environment.

Modern, Strategic Theories of Limit Pricing: In order to limit entry, the incumbent firm must be able to take an action prior to entry that truly affects the profitability of the potential entrant were they to actually enter. Possibilities: manipulate market demand, the incumbent firm's cost, or the entrant firm's cost.

Example 1: Output Adjustment Costs

What if there is a cost to adjusting output? Then if the incumbent selects a high rate of output prior to entry, the adjustment cost may be such that it will be best off to continue to produce at a high rate even if entry occurs. This is credible commitment….

Example 2: Switching Costs

What if it is costly for buyers to acquire the necessary information to switch sellers?

Example 3: Uncertain Post-Entry Demand or Incumbent Marginal Cost

What if the potential entrant does not know post-entry demand or the incumbent's marginal cost? Then the incumbent can affect the entrant's beliefs by its own price and output choices.

Capacity: Dixit Model

Fixed cost K required for production. Capacity cost "r" per unit of capacity stock "x". Variable cost "w" per unit "q" produced.

C(q) = K + rx + wq if q no larger than x

K + (w + r)q if q is greater than x

Given preexisting capacity stock x, a firm has fixed (and sunk) cost of K + rx. Thus once capacity is installed, production of "q" units involves a total variable cost of wq, but to produce output in excess of x, the firm incurs a capacity cost of (x - q)r in addition to wq.

3-Stage Game:

1. The incumbent has a first-mover advantage and sets capacity x.
2. The potential entrant observes x and decides whether or not to enter at cost K.
3. If entry occurs, the entrant selects x. All firms in the market set "q" in a Cournot setting.

Key insight: The incumbent firm can deter entry by committing to sufficient capacity x. If entry occurs, the subsequent Cournot game features a cost advantage for the incumbent (w rather than w+r), implying that the new entrant will have a smaller share of the market and lower profits. Entry can be deterred if these post-entry profits are smaller than the sunk cost K of entry.

(Go Through Figure 6.11)

Preemption and Brand Proliferation:

Suppose there are two types of goods-X and Y-which are imperfect substitutes for one another. If a new firm enters and produces good Y, there will be two firms-one producing X and the other Y, and they will engage in price rivalry in the way we saw last chapter under product differentiation. If one firm could produce both goods X and Y as a monopolist, it could coordinate the price of those two goods and be more profitable. Thus production of good Y is more valuable to the incumbent than to the entrant. Before it is profitable for an entrant to produce Y, an incumbent will already do so. This is the strategy of brand proliferation, and it deters entry and reduces consumer surplus.