First Exam for Econ 221 Industrial Organization Spring 2020
Instructions
The exam is out of 45 points. The exam is due by 5 pm on 3/27 (48 hours total). To submit your
exam, please email me at paul.lombardi@sjsu.edu, a pdf copy of your answers. The exam is
open book/note, but I will not tolerate plagiarism. Email me if you have clarifying questions.
Good Luck!!
Problems
Question 1: Explain one of the two ways we proved setting MR=MC maximized firm profits.
(2 pts)
Supply: Q = 4p
Demand: Q = -2P +12
1
Total Costs: TC= 𝑄2
2
Question 2a: Find the competitive market equilibrium price and quantity. (2 pts)
Question 2b: Find the profit-maximizing quantity produced by an individual firm in the above
market. (2 pts)
Question 2c: How many firms will produce in this market? (1 pt)
Question 2d: Calculate consumer and producer surplus in the above market. (2 pts)
Question 3a: Now assume the above market has a single producer (i.e., is a monopoly). Find the
market equilibrium price and quantity. (4 pts)
Question 3b: Calculate consumer and producer surplus in the above market. (4 pts)
Question 4a: In the dominant firm with a competitive fringe model, the dominant firm behaves
like what kind of firm? (1 pt)
Question 4b: In the dominant firm with a competitive fringe model, the firms within the
competitive fringe behave like what kind of firm? (1 pt)
Question 4c: When the dominant firm raises prices, what are the two reasons their quantity
produced decreases? (2 pts)
Question 4d: Explain one of the factors that determine the dominant firm’s market power.
(2 pts)
Question 4e: In the dominant firm with a competitive fringe model, firm behavior varies across
three price ranges. Explain this pattern in words and graphically (i.e., your explanation should
include three separate graphs). (3 pts)
Question 5a: What is the Coase Conjecture? (2 pts)
Question 5b: What are two strategies firms use to counter the Coase Conjecture? (2 pts)
Question 5c: Broadly, what is the goal of strategies firms use to counter the Coase Conjecture?
(3 pts)
Question 5d: Under what market conditions is the Pacman Strategy more likely to hold than the
Coase Conjecture? (2 pts)
Question 6a: Describe two sources of Dead Weight Loss (i.e., inefficiency) associated with
market power discussed in class. (2 pts)
Question 6b: How is X-Inefficiency related to the Economies of Scale benefit of monopolies?
(2 pts)
Question 7a: In Monopoly Quality Degradation and Regulation in Cable Television, why did
companies with multiple bundles increase prices and quality relative to areas with a single
bundle? (3 pts)
Question 7b: In Monopoly, Quality, and Regulation, what characteristic of the theoretical
framework made government regulation difficult? (3 pts)
Monopoly Quality Degradation and
Regulation in Cable Television
Gregory S. Crawford University of Arizona
Matthew Shum Johns Hopkins University
Abstract
Using an empirical framework based on the Mussa-Rosen model of monopoly
quality choice, we calculate the degree of quality degradation in cable television
markets and the impact of regulation on those choices. We find lower bounds
of quality degradation ranging from 11 to 45 percent of offered service qualities.
Furthermore, cable operators in markets with local regulatory oversight offer
significantly higher quality, less degradation, and greater quality per dollar,
despite higher prices.
1. Introduction
In many markets, firms choose not only the prices but also the qualities of their
products. In many cases, this is the primary dimension on which firms compete,
as in pharmaceutical, media, and professional services and many high-technology
markets. Theorists have long recognized that in the presence of imperfect competition, offered qualities can be distorted from the social optimum because
firms equate private instead of social marginal benefits and marginal costs (Dixit
and Stiglitz 1977; Spence 1980). This induces a welfare loss analogous to that
from price distortions. Indeed, aspects of a firm’s product offerings, and not
pricing, have been the focus of recent highly contested antitrust cases (for example, Microsoft, GE/Honeywell).
The tendency of firms with market power to distort quality has been most
clearly formulated in the monopoly nonlinear pricing literature, in which it is
shown that the firm’s products suffer from quality degradation (Mussa and Rosen
1978; Maskin and Riley 1984). Because products of different qualities are substitutes, a monopolist cannot simultaneously offer each consumer his or her
efficient quality and also extract his or her full surplus, even with a fully nonlinear
We would like to thank Gary Biglaiser, Silke Januszewski, Eugenio Miravete, and seminar participants at Northwestern University, the University of California, Los Angeles, the National Bureau
of Economic Research 2002 winter program meeting, the Society for Economic Dynamics 2002
meeting, the 2004 Kiel-Munich Workshop on Network Industries, and the 2004 Centre for Economic
Policy Research conference Competition in the New Economy for helpful comments.
[Journal of Law and Economics, vol. 50 (February 2007)]
䉷 2007 by The University of Chicago. All rights reserved. 0022-2186/2007/5001-0007$10.00
181
182
The Journal of LAW & ECONOMICS
tariff. Instead, under standard assumptions, quality for all but consumers with
the highest tastes for quality is distorted downward. Furthermore, consumers
with low preferences for quality may be excluded entirely from the market.
Regulation, by either minimum quality standards or price caps, generally reduces
distortions but can have ambiguous effects on prices and welfare (Besanko,
Donnenfeld, and White 1987, 1988).
Despite the widespread acknowledgment of the potential for quality degradation, measures of its extent and implications for outcomes in real-world markets are few. In this paper, we analyze quality degradation in a market long
thought subject to its effects: the cable television industry. To do so, we introduce
an empirical framework based on the Mussa-Rosen model that exploits the
optimality conditions for the monopolist’s quality choice problem to recover
measures of the quality of the monopolist’s offerings. This permits us to directly
measure how much cable monopolies degrade quality relative to a competitive
alternative. It also allows us to measure the impact of local regulatory oversight
on ameliorating monopoly quality distortion.
We present two main results. First, we find evidence of substantial quality
degradation in the cable television industry across a variety of specifications.
While some firms offer two or three goods, most offer just a single product
quality. Furthermore, offered qualities are at least 11.1 percent and 30.3 percent
less in three-good markets and 44.7 percent less in two-good markets than what
would be provided in a competitive market offering the same number of goods.
Second, we find that local regulatory oversight—in the form of certification
by the local franchise authority to cap cable prices—has important ameliorative
effects. Systems in franchise areas where the local franchise authority was certified
offer an estimated 25.1 percent more services, 24.1 percent higher quality for
low- and medium-quality goods (where offered), and greater quality per dollar
to consumers despite higher prices. These results are consistent with the impact
of minimum quality standards and could be of significant interest to policy
makers concerned about the effectiveness of past regulatory interventions in the
industry but troubled by continued growth in cable prices.1
The rest of this paper is organized as follows. In Section 2, we survey the
canonical Mussa and Rosen (1978) model of monopoly quality choice that forms
the foundation of the empirical analysis. We also present extensions to this model
developed by Besanko, Donnenfeld, and White (1988) to allow for quality choice
in the presence of regulation. In Section 3, we describe the cable television
industry and discuss its suitability for this empirical analysis, followed in Section
4 by the empirical model and algorithm for recovering quality measures. Section
5 presents the results, and Section 6 concludes.
1
The most recent report on cable prices by the Federal Communications Commission (FCC)
found prices increased by 5.4 percent for the 12 months ending January 1, 2004, slightly less than
the 5-year compound annual increase of 7.5 percent from 1998 to 2003 and far higher than the 1.5
percent increase in the Consumer Price Index over the same period (FCC 2005a).
Regulation in Cable Television
183
2. The Incentives to Degrade Quality
In this section, we discuss the quality degradation result from the theory of
monopoly nonlinear pricing using a simple, two-type version of the model of
Mussa and Rosen (1978).2 Consider a monopolist selling two goods, q1 and q2,
whose qualities can be freely varied over Q p [0, Q]. Consumers are assumed
to be differentiated by a type parameter that takes on three distinct values, t0,
t1, and t2 (t 0 ! t 1 ! t 2), with respective probabilities fi (with f0 ⫹ f1 ⫹ f2 p 1) and
k
associated cumulative distribution function Fk { 冘jp0 fi. Type 0, t 0, is included
to allow for the possibility that some consumers prefer not to purchase either
of the firm’s products.3 For convenience, we assume the hazard function for the
type distribution, fi / (1 ⫺ Fi), is increasing in i.4 The monopolist is assumed to
be able to offer a nonlinear tariff specifying a different total price per quality
variant offered, P1 and P2. The firm knows the distribution of types in the
population and selects the tariff that maximizes its expected profit (with the
expectation taken over consumer types).
Consumer preferences are assumed to be quasi-linear in money, ui { u(q,
ti) p v(q, ti) ⫺ P(q). A consumer of type ti is assumed to choose that bundle,
q i, which maximizes his or her utility, so that
q i { arg max u(q, ti),
i p 1, 2.
(1)
q苸{q1,q 2}
Furthermore, given that no consumer can be forced to participate in the contract,
the monopolist’s choice of qualities and prices must be such that the consumer
voluntarily chooses to accept the contract, which requires
u(q i , ti) ≥ 0,
i p 1, 2.
(2)
Equations (1) and (2) are the incentive compatibility (hereafter IC) and individual rationality (IR) constraints.
The firm’s optimization problem is then to maximize expected profits,
冘
2
max E[p] p
P(q)
fi[P(q i) ⫺ C(q i)],
(3)
ip1
subject to optimal behavior by consumers, as encompassed in the IC and IR
2
Since the derivations in this section are standard, we omit a number of technical details; see, for
example, Laffont and Tirole (1993, chap. 2) for complete details. Furthermore, the Mussa-Rosen
model has recently been extended by Rochet and Stole (2002) to allow households random private
values for the outside option, with interesting implications for the extent of and patterns in quality
degradation. We explore the differing implications of these models in ongoing work (Crawford and
Shum 2005) and note that the results we present here are conditional on our assumed form for
household preferences.
3
This “outside type” is generally not included in the typical theoretical exposition. We include it
here to facilitate empirical implementation of the model, in which there are always some consumers
who purchase the “outside good.”
4
This rules out bunching of types at a single quality variant. Wilson (1993, chap. 8.1) presents a
detailed discussion of the conditions under which this assumption is likely to be violated.
The Journal of LAW & ECONOMICS
184
constraints. The term C(q i) is the firm’s cost function, which is assumed to be
purely additive across consumers.5 Define the total surplus function S(q, ti) {
v(q, ti) ⫺ C(q). Using a common trick from the screening literature, we can rewrite
profits as the difference between the total and consumer surplus:
冘
2
max E[p] p
u(q)
fi[S(q i , ti) ⫺ u(q i , ti)].
(4)
ip1
In this reformulated problem, the monopolist solves for the optimal utility
quality schedule and determines optimal prices (given utilities) from the binding
IC constraints. Under standard assumptions, we can use the IC constraint to
rewrite the objective function, which yields
max E[p] p f1[S(q 1,t 1) ⫺ u 1] ⫹ f2{S(q 2 , t 2 ) ⫺ [v(q 1, t 2 ) ⫺ v(q 1, t 1)] ⫺ u 1}.
(5)
q1,q 2 ,u1
This problem is solved by setting the utility of the lowest type to zero,
u 1 p 0, and maximizing the resulting unconstrained objective function with
respect to q 1 and q 2. The corresponding first-order conditions are
Sq(q 1, t 1) p
1 ⫺ F1
[vq(q 1, t 2 ) ⫺ vq(q 1, t 1)] and Sq(q 2 , t 2 ) p 0,
f1
(6)
where vq { ⭸v/⭸q. Quality degradation for the low type (i p 1) is visible from
equation (6). The socially optimal quality for each type, denoted q**
i , is that
which sets the derivative of the total surplus function to zero, Sq(q, ti) p 0. In
equation (6), however, we see that q 1 is chosen so that Sq(q, t 1) 1 0, which implies
that q*1 ! q**
1 : quality is degraded to low types. However, there is no degradation
at the top for the higher type t 2. Given optimal qualities from equation (6),
optimal prices fall out naturally from the IR and IC constraints. Since u 1 p 0,
p*1 p v(q*,
1 t 1) and p*
2 p v(q*,
2 t 2 ) ⫺ [v(q*,
1 t 2 ) ⫺ v(q*,
1 t 1)] p p*
1 ⫹ [v(q*,
2 t 2) ⫺
v(q*,
t
)]
.
1
2
Figure 1, which is adapted from Maskin and Riley (1984), demonstrates graphically the solution for the one-dimensional case with N p 2. At this point, we
focus only on the solid curves in that figure. The monopolist would like to
extract all consumer surplus by offering product qualities q**
and q**
and
1
2
charging prices p**
and
p**
,
but
with
such
an
offering
the
high
type
would
1
2
prefer to mimic the low and select q**
(note for a given quality, consumer utility
1
is higher the lower on the figure they can locate). The constrained optimum is
given by variables with single asterisks. As above, the high type continues to
consume the efficient quality (and pays a lower price), but quality to the low
type is degraded, from q**
to q*1 .
1
5
We make the usual curvature assumptions v1 1 0, v11 ≤ 0, v2 1 0, c 1 0, and c 1 0 , as well as the
normalization that v(0, ti) p 0 for all i. Furthermore, we maintain the standard single-crossing
condition that uqt 1 0 , which implies that higher types have greater willingness to pay (WTP) for
quality at any price or that consumers may be ordered by their type, t.
Regulation in Cable Television
185
Figure 1. Quality degradation with two types adapted from Maskin and Riley (1984)
2.1. Continuous Types but Discrete Qualities
The theory described in the previous section applies also to the case of continuous types but to discrete qualities. To see this, suppose instead that consumer
types are continuously distributed on [T, T] with probability density function
f(t) but that the monopolist has decided to offer just two qualities regardless.
He or she may do so for a number of reasons. There may be fixed costs associated
with the design, production, or marketing of products of different qualities. Or
there may be incremental (especially marketing) costs of offering numerous
goods. If these are large, the monopolist will offer only those products that can
cover his or her fixed costs, limiting the number of products in the market
(Spence 1980; Dixit and Stiglitz 1977).
Suppose the firm offered arbitrary qualities q 1 and q 2. Who would buy these
goods? All consumers for whom u(q 2 , t) ≥ u(q 1, t) and u(q 2 , t) ≥ 0 would buy
good 2. Because of the structure of the problem—notably the single-crossing
condition—only the first of these constraints would bind. Let t 2 denote the
consumer type that is just indifferent between purchasing the two goods and
t 1 denote the analogous consumer type just indifferent between purchasing good
1 and the outside (or no) good. Then the share of the distribution of consumer
types that purchase each good, fi , is given by the integral under the distribution
186
The Journal of LAW & ECONOMICS
Figure 2. Continuous types and discrete qualities
between the type cut points: fi p ∫t i⫹1
f(t)dt (defining t 0 p T and t 3 p T). Figure
i
2 presents a graphical representation of this framework. In that figure, type tA
lies between the cut types t 1 and t 2 and so consumes the lower bundle. Type tB
lies above the larger cut type t 2, and like that type consumes the higher bundle.
For both types tA and t B (and for all types other than the cut types t 1 and t 2),
both the participation and incentive constraints hold strictly. The key result is
that given these qualities q 1 and q 2 and associated shares f0 , f1, and f2, the
monopolist’s profit is described by equation (5) just as in the discrete-type case.6
An important consequence of continuous consumer types is that quality distortion will generally occur for almost all consumers. In particular, only the
highest cut type t 2 will consume an efficient quality (q *2 p q **
2 ). All other types
t 1 t 2 that also purchase the high-quality good (like tB) will necessarily receive
inefficiently low qualities. Similarly, while quality will still be degraded to the
lower cut type (q *1 ! q **
1 ), it will be lower still for other, higher, types (like tA)
that also purchase the low-quality good, t 1 ! t ! t 2. This is also illustrated in
6
This is a subtle point. Were we to specify a particular continuous distribution of consumer types,
solving the firm’s problem for the optimal cut types, t ’s, is a challenging problem requiring more
sophisticated techniques than those employed here (Crawford and Shum 2005). The insight is that
even if firms are making these more sophisticated calculations, the discrete-type first-order conditions
must hold for the cut types ultimately chosen by firms.
Regulation in Cable Television
187
Figure 1, in which the two dashed curves are indifference curves for the types
tA and tB in Figure 2. Type tA, who consumes the same bundle as type t 1, has
an efficient bundle that lies to the right of type t 1’s efficient bundle, which implies
that the quality distortion to type tA is higher than that to type t 1. Similarly,
there is a positive distortion to type tB, even though he or she consumes the
same bundle as type t 2, to whom there is no distortion.
The theory described above applies analogously for an arbitrary number n of
offered qualities. For any n, equation (6) continues to hold, with associated
degradation for all but the highest offered quality q n. However, when the type
distribution is continuous but the monopolist offers only discrete qualities, the
cut types t 1 and t 2, as well as n, the number of offered qualities, are also choice
variables. In this paper, while we do not use the monopolist’s optimality conditions for these variables in recovering quality measures, we do briefly analyze
the number of goods offered by firms in the empirical analysis.
Finally, note that it is typical in models of this type to make additional assumptions on the distribution of consumer types to ensure the optimal prices
and qualities are monotonically increasing in types. Because, however, we restrict
our attention to the implications of the model for a discrete number of qualities,
we do not have to do this. Indeed, it could be the case that in some market n
the inverse hazard function of types, [1 ⫺ Fn(t)] /fn(t), is nonmonotonic in t (as
in Figure 2). If the firm in market n were to offer a fully nonlinear price/quality
schedule in such a case, it would require sophisticated solution techniques involving pooling of types at particular qualities (Wilson 1993). With discrete
qualities, however, pooling obtains regardless of the shape of the type distribution.
For our purposes, it is convenient if the inverse hazard function d ...

Purchase answer to see full
attachment