Voting, spatial monopoly, and spatial price regulation.
Meng-Hua Ye ; Yezer, Anthony M.J.
VOTING, SPATIAL MONOPOLY, AND SPATIAL PRICE REGULATION
Regulations often require that local public utilities engage in
high rates of freight
absorption. These regulations, often mandating uniform pricing,
are shown to arise
logically as a consequence of self-interested voting behavior. We
specifically consider
the case of a single-plant spatial monopoly which is regulated by
consumers distributed
around the plant. Consumers may change their delivered price by
voting to require a
rate of freight absorption which differs from the
profit-maximizing rate. Voting
outcomes under a median voter model predict the high rate of
freight absorption often
observed in practice.
I. INTRODUCTION
While the literature on spatial monopoly is large, and research on
regulated firms larger still, very little attention has been given to
the study of regulations affecting the spatial monopolist.(1) Regulated
spatial monopolies could include either public or private production of
telephone service, natural gas, cable television, waste removal, water,
sewage service, etc. Indeed the analysis could be adapted to mass
transportation systems providing access to an employment center. It is
common to find that regulations governing the spatial monopolist require
uniform delivered pricing in which the firm must absorb most or all
transportation costs.(2) It appears that regulated freight absorption by
the spatial monopolist is one of those cases that everyone knows about
but which no one has bothered to analyze in detail.
An economic analysis of regulated freight absorption requires that
the reaction of a spatial monopolist to the full range of requirements
for freight absorption, from zero to full absorption, be available. But
there is no such treatment in the literature. The first objective of
this paper is to fill this gap by analyzing the effects of continuous
variation in freight absorption on prices, output, profits,
consumers' surplus and total welfare. The results which we obtain
are rather surprising. For example, the unregulated profit-maximizing
spatial monopolist is found to be a knife-edge case in which
second-order conditions are not satisfied. In addition, welfare
maximization is achieved by regulating the rate of freight absorption so
that it lies between zero and the rate practiced by an unregulated firm.
Having determined the effects of alternative freight absorption
regulations on the firm, we then construct a voting model in which
spatially distributed consumer-voters choose the rate of regulated
freight absorption. In this spatial voting model, the spatial
distribution of consumer surplus is used to generate voter preferences
regarding the regulated level of freight absorption. This analytical
approach, which we call spatial surplus surface analysis, shows that the
median voter prefers a high level of freight absorption. This explains
the popularity of regulations requiring high rates of freight
absorption, the regulatory outcome most often observed. These results
have significant welfare implications. Consumer voters choose freight
absorption regulations with higher required rates of freight absorption
than practiced by an unregulated spatial monopolist engaged in spatial
price discrimination. The voting outcome lowers both total welfare and
total consumers' surplus below the levels obtained under
unregulated spatial monopoly.
II. MODEL OF THE SPATIAL MONOPOLIST: ASSUMPTIONS
Following the conventional stylized assumptions for single-plant
spatial monopoly, there is a given linear (one-dimensional) market
characterized by location along X, a subset of (- [infinity],
[infinity]), with a single plant located at point 0.(3) There is a
single firm producing and distributing a homogenous product, whose
quantity is denoted by q, to consumers distributed at uniform density
across the market. Resale or transportation of the product by consumers
is not possible, so that demand at any distance, r, is independent of
the price charged at alternative distances. Total quantity demanded at
distance r is given by(4) (1) q(P(r)) = [Beta] [P.sub.2] - [Beta] P(r),
q([Beta]) = 0, where P(r) is the delivered price and [Beta] is a
parameter equal to the choke price.(5) Subsequent measures of consumer
surplus will be based on the area under this demand function. Use of
such a measure implicitly assumes the marginal utility of income is
constant and, with a homogenous population, distributional questions are
not considered.
The spatial price function, P(r), takes the particular form P(r) =
P(0) + etr where P(0) is price at r = 0, t is the transportation cost
per unit of output per unit of distance and e is the fraction of freight
cost passed on to consumers, (1 - e) is the fraction of freight cost
absorbed by the firm. In the case of mill or Free on Board (FOB)
pricing, e = 1 and there is no freight absorption. Uniform pricing
occurs when e = 0 and there is complete freight absorption.
The monopolist produces output subject to increasing returns, which
are traditionally represented by fixed cost for an essential input and
constant marginal cost. Transportation cost is proportional to distance
and quantity transported. The total cost function of the monopolist is
given by (2) C = [C.sub.T] + F + [C.sub.V] = [C.sub.T] + F + mQ where C
is total cost, [Mathematical Expression Omitted] is total transportation
cost, F is total fixed cost, [Mathematical Expression Omitted] is the
total variable cost, m is constant marginal cost, q(P(r)) is total
output sold at distance r, and Q is total quantity sold. The market
extends in two directions from the plant covering a total distance of
2R.
The welfare measure is the sum of monopoly profit and
consumers' surplus. The voting behavior of consumers is motivated
by an attempt to maximize their personal consumer surplus.
III. STANDARD RESULTS ON THE BEHAVIOR OF THE SPATIAL MONOPOLIST
First consider the production and pricing decision of the
unregulated spatial monopolist maximizing profit in the environment
described in section II. The monopolist chooses a market radius R, so
that the actual market served [-R, R] is a subset of X, selects P(0)
which determines the price at r = 0 and determines e, the fraction of
transportation cost passed on to consumers. Pricing over space is
determined by P(r) = P(0) + etr, r [Epsilon] [0, R], but the monopolist
must choose P(R) [is less than or equal to] [Beta], the choke price.
Thus the general pattern of spatial pricing is a function of P(0) and e,
which vary independently of R except when the choke price is
binding--i.e., in the case where P(R) = [Beta]. When P(R) < [Beta],
P(0), e, and R are chosen independently. Recalling that the total cost
at distance r is given by equation (2), and letting total revenue = Z,
profit may be written as (3) [Pi] = Z - [C.sub.T] - [C.sub.V] - F
[Mathematical Expression Omitted] As is well known, for linear demand
the level of e which maximizes profit is 0.5, i.e., half of
transportation costs are absorbed.(6)
Welfare maximization is based on the sum of [Pi] and
consumers' surplus. Consumer's surplus for households located
at r, s(r), is given by (4) [Mathematical Expression Omitted] Overall
consumers' surplus is (5) [Mathematical Expression Omitted] Welfare
maximization implies maximization of [Pi] + S, which has been shown to
required e = 1,(7) or marginal cost pricing. Other well-understood
results involve the case of e = 0, uniform pricing or complete
transportation cost absorption, which is often prompted by regulation.
IV. REGULATING FREIGHT ABSORPTION BY THE SPATIAL MONOPOLIST
The major object of this study is the effect of regulations
controlling the level of transportation cost absorption. Given that
welfare is maximized at e = 1 (zero absorption or mill pricing) and firm
profit is maximized at e = 0.5 (spatial price discrimination), it is
curious that regulations commonly require that the monopolist set e = 0
(uniform delivered price). It is always possible that such regulations
result from ignorance, cultural notions of fairness, or problems
implementing distance-based pricing. Our analysis is designed to
determine if voter self-interest and knowledge of the effects of e on
consumer surplus can lead to regulations requiring low values of e.
To determine the effect of variation in e on the decision
variables, we treat e in equation (3) as a parameter and use standard
maximizing techniques. The first-order necessary conditions of the
optimal P(0) and R for the maximization of the total profit [Pi] given
by equation (3) are (6) P(0) = ([Beta] + m)/2 + (1 - 2e)tR/4, and (7)
q(P(R))[P(0) - m - (1 - e)tR] = 0. Whether equation (7) implies q(P(R))
= 0 or [P(0) - m - (1 - e)tR] = 0 depends crucially on the concavity of
[Pi] over R, i.e., on the second-order conditions for R: (8)
[Mathematical Expression Omitted] We find, applying equations (6) and
(7), that [Mathematical Expression Omitted] for q(P(R)) = 0 and e
[Epsilon] [1/2, 1], and that ([Delta].sup.2] [Pi])/[[Delta]R.sup.2]) =
(1 - e) [Delta] [t.sup.2] R(2e-1) [is less than or equal to] 0 for P(0)
- m - (1-e)tR = 0 and e [Epsilon] [0, 1/2]. Consequently, we have the
optimal solutions as follows. For e [Epsilon] [0, 1/2], (9) P(0) = [2(1
- e) [Beta] + m]/ (3 - 2e), and (10) R = 2([Beta] = m)/t(3 - 2e). For e
[Epsilon] [1/2, 1], (11) P(0) = ([Beta] + 2em)/(1 + 2e), and (12) R =
2([Beta] - m)/t(1 + 2e). This results in the following series of
propositions. For purposes of comparison with results already in the
literature, outcomes obtained for specific values of e = 1/2, e = 1, and
e = 0 will be noted using subscripts 1/2, 1 and 0 respectively and the
subscript [Omega] will indicate a result that maximizes welfare. First
we develop three propositions about the effects of varying e on the [0,
1] interval which are implicit in the standard literature on spatial
monopoly.
PROPOSITION 1. A single-plant spatial monopolist under transportation
cost absorption regulation will charge the choke price at r = R if
regulated e [Epsilon] [1/2, 1]; otherwise it will charge a price less
than the choke price at r = R if regulated e [Epsilon] [0, 1/2).
Proof. Substituting equations (11) and (12) into the spatial
pricing definition P(R) = P(0) + etR, we obtain that P(R) = [Beta] for
all e [Epsilon] [1/2, 1]; and substituting equations (9) and (10) into
the spatial pricing equation at r = R, we obtain that P(R) < [Beta]
strictly for all e [Epsilon] [0, 1/2).
PROPOSITION 2. Regarding the maximum market radius of the
single-plant spatial monopolist under transportation cost absorption
regulation, [R.sub.W] [is greater than or equal to] [R.sub.1] =
[R.sub.0]; and [Delta] R/ > 0, [[Delta].sup.2] R/[[Delta] e.sup.2]
> 0 for all e [Epsilon] [0, 1/2), and [Delta] R/[Delta] e < 0,
[[Delta].sup.2] R/[[Delta] e.sup.2] > 0, for all e [Epsilon] (1/2,
1].
Proof. The rank of R with respect to welfare maximization, and
profit maximization with e = 0, 1/2, and 1 is determined simply by
plugging the respective values into equations (10) and (12). The
monotonicity results are obtained by taking derivatives of equations
(10) and (12) with respect to e and the concavities are given by the
second-order derivatives of R with respect to e. These results are
illustrated below in Figure 1.
Note that the choice of market radius is continuous in e but not
differentiable at the profit-maximizing radius [R.sub.1/2]. The first
condition, derived from equation (7), is switched from one to another as
the sign of the second-order conditions is flipped around this point.
This switching arises for a variety of demand curves, convex, concave,
and linear, but it does require that there be a choke price. In such
cases, there are two factors which can produce a market boundary at R.
First, if the rate of transportation cost absorption is high, marginal
cost of production plus transportation may rise to equal price, or m +
tR = P(R). In such cases, P(r) is below the choke price and q(P(r)) >
0 for r < R, and P(r) becomes infinite and q(P(r)) = 0 for r > R.
Second, if transportation cost absorption is small (1 [is greater than
or equal to] e > 1/2), delivered price reaches the choke price at the
market boundary and q(P(R)) = 0, although the marginal cost of
production and transportation is less than delivered price at the
boundary, m + tR < P(R). This dual nature of the conditions
characterizing the market boundary, not noted in previous studies,
appears to have serious consequences for the shape of the welfare
surface and consequent regulatory decisions.
PROPOSITION 3. Concerning total output, [Q.sub.w] > [Q.sub.1/2]
> [Q.sub.1] = [Q.sub.o]; and [Delta] Q/[Delta] e [is greater than or
equal to] 0, [[Delta].sup.2] Q/ [[Delta]e.sup.2] [is less than or equal
to] 0 for all e [Epsilon] [0, 1/2], and [Delta] Q/ [Delta] Q/ [Delta] e
[is less than or equal to] 0, [[Delta].sup.2] Q/ [[Delta]e.sup.2] [is
less than or equal to] 0 for all e [Epsilon] [1/2, 1].(8)
These results are fairly obvious from the literature. The
profit-maximizing spatial monopolist sets e = 1/2 and produces a large
output because sales extend over the largest market radius excepting the
welfare-maximizing case.
Now we come to a series of three new propositions, not implicit in
previous work, which provide the basis for statements about welfare and
for voting analysis conducted in the next section.
PROPOSITION 4. Total consumer surplus generated by the single-plant
spatial monopolist under alternative transportation cost regulations
follows [S.sub.w] > [S.sub.1] > [S.sub.1/2] > [S.sub.0]; and
[Delta] S/ [Delta] e > 0, [[Delta].sup.2] S/ [[Delta]e.sup.2] > 0
for all e [Epsilon] [10, 1/2], and [Delta] S/ [Delta] e > 0,
[[Delta].sup.2] S/[[Delta] e.sup.2] < 0, fall all e [Epsilon] [1/2,
1].(9)
PROPOSITION 5. Profit of the single-plant spatial monopolist under
alternative transportation cost regulations ranges from [[Pi].sub.1/2]
> [[Pi].sub.1] = [[Pi].sub.0] > [[Pi].sub.w] = 0; and [Delta]
[Pi]/[Delta] e [is greater than or equal to] 0, [[Delta].sup.2] [Pi]/
[[Delta]e.sup.2] [is less than or equal to] 0, for all e [Epsilon] [0,
1/2] and [Delta] [Pi]/[Delta] e [is less than or equal to] 0,
[[Delta].sup.2] [Pi]/[[Delta]e.sup.2] [is less than or equal to] 0, for
all e [Epsilon] [1/2, 1]. (As expected, profit rises with e until it is
maximized at e = 1/2 and then it falls steadily.)(10)
PROPOSITION 6. Welfare associated with the regulated single-plant
spatial monopolist varies as [W.sub.w] > [W.sub.0.683] >
[W.sub.1/2] > [W.sub.1] > [W.sub.0]; and [Delta] W/[Delta] e >
0, [[Delta].sup.2] W/[[Delta]e.sup.2] [is greater than or equal to] 0
for all e [Epsilon] [0, 1/2], [Delta] W/[Delta] e = 0 for e = 0.683, and
[[Delta].sup.2] W/[[Delta]e.sup.2] [is less than or equal to] 0 for all
e [Epsilon] [1/2, 1].(11)
The relation between e and W is displayed in Figure 3. There is the
familiar result that a global maximum of welfare is obtained at an
extreme point through strict marginal cost pricing at [W.sub.w]. Of
course, this special case assumes some non-distorting scheme which can
be devised to compensate the spatial monopolist for operating where
profit is negative (specifically [Pi] = -F). The most surprising result,
to our knowledge not obtained elsewhere, is that, ignoring the extreme
point maximum under marginal cost pricing, welfare is maximized at e =
0.683. It is useful to distinguish this as the "internal welfare
maximum" which will be noted as [W.sub.m] hereafter.
Overall these results appear substantially inconsistent with
regulatory decisions to require high rates of transportation cost
absorption by the regulated spatial monopolist. Both consumer surplus or
total welfare are minimized under uniform pricing. It appears that
consumers should instruct regulators to require that transportation cost
be fully reflected in price.
V. CONSUMER VOTING AND REGULATED TRANSPORTATION COST ABSORPTION
Consumer-voters base decisions on the regulated rate of
transportation cost absorption based on the consumer surplus generated
by the spatial monopolist. Following Mueller [1989], if consumer welfare
can be reduced to a single metric, voting outcomes on regulation are
determined by the median voter.
The spatial distribution of consumer's surplus for different
values of transportation cost absorption is given by the s(r, e)
function displayed in Figure 4 below.(12) Total surplus, which was
discussed in Proposition 4 and displayed as a function of e in Figure 2,
is the integral of s(r, e) on the interval r [Epsilon] [0, R].
The spatial surplus surface formed by s(r, e) is quite irregular.
Voting behavior of consumers located at a particular r is motivated by
an attempt to maximize surplus generated by different rates of
transportation cost absorption. Finding the value of transportation cost
absorption, e*(r), which maximizes consumer surplus for voters at a
given r involves rather long and cumbersome manipulation. Conceptually,
we are slicing the spatial surplus surface along a given r and
determining the level of e, noted e*(r), which maximizes the function
thus obtained. It is necessary to find e*(r) separately on the interval
[10, 1/2] and on [1/2, 1]. Comparison of the maxima generated on each of
these intervals allows one to compute the e*(r) associated with a
maximum of surplus on the entire interval e [Epsilon] [0, 1]. Figure 5
shows the relation between location and e*, the surplus-maximizing level
of transportation cost absorption.(13)
Location is expressed in terms of the maximum market radius,
[R.sub.1/2], shown in Figure 1. There are two significant groups of
consumer voters, one with e* = 1 and another with e* = 0. The median
voter preference is for e = 0.33, as illustrated in Figure 5.(14) The
voting outcome which prefers e [is greater than or equal to] 0.33
consists of consumers located beyond [0.357R.sub.1/2] and extending
nearly to [R.sub.1/2]. While voters consider only their own self
interest, the final outcome appears to have an obvious element of
cross-subsidy. Because regulated e is below 1/2, the market will end
with P(R) < [Beta], the choke price, and consumers living near
[R.sub.1/2] will be cut off from service. Voters living from point 0 to
R = [0.357R.sub.1/2] are forced to subsidize those living past
[0.357R.sub.1/2].
The high rate of freight absorption selected by the median voter is
inconsistent with a maximum of profit, welfare, or consumers'
surplus. Indeed consumers' surplus is lower at the e = 0.333 level
selected by voters than when the unregulated profit-maximizing firm
engages in spatial price discrimination and sets e = 0.5. This
surprising voting outcome is generated in part because the median voter
selects such high freight absorption that consumers near the maximum
market radius, i.e., near [R.sub.1/2 [prime]] are cut off from service.
Thus far the one-dimensional representation of the market has been
adequate. However, analysis of voting behavior should take place in a
two-dimensional market where, under the assumption of uniform consumer
density, the number of consumers increases with the square of r. If the
results in Figure 5 are modified, essentially by "spinning"
Figure 5 about the origin to sweep out a circle and hence consider the
number of consumers a function of r in a circular market with uniform
consumer density, the level of e* characterizing the median voter is e =
0.297.(15)
Thus a spatial model of consumer voting on the level of
transportation cost absorption to be imposed on the regulated spatial
monopolist suggests that high rates of absorption will be required.
voting outcomes which set e = 0.297 (or e = 0.333 for the
one-dimensional case) imply that regulators will be compelled to require
that the spatial monopolist absorb transportation costs at rates which
greatly exceed those associated with either profit maximization by the
firm, surplus maximization for consumers, or overall welfare
maximization.
V. SUMMARY AND CONCLUSIONS
It is quite common to find a regulatory requirement that spatial
monopolists absorb most, if not all, transportation cost, i.e., that
they engage in uniform delivered pricing. Given that this is a clear
departure from marginal cost pricing, the conclusion that it is not
consistent with welfare maximization follows easily. The results
presented here extend this easy conclusion in several ways and, based on
the spatial surplus surface, provide a voting model which explains the
political pressure for uniform pricing schemes.
By merging theoretical models of the spatial firm with regulatory
and voting analysis in a manner not previously reported in the
literature, we obtain a variety of original and, we hope, interesting
results. First, analysis of the effect of varying e over the entire
choice set e [Epsilon] [0, 1], shows that uniform pricing, e = 0,
minimizes both firm profits and total consumer surplus. Second, internal
welfare maximization is achieved with fairly low levels of
transportation cost absorption, i.e., with e = 0.683 for the specific
case considered here. Third, the spatial distribution of consumer
surplus generated by alternate levels of transportation cost absorption
is such that the median voter prefers regulations requiring very high
levels of absorption, with e = 0.297 for the case analyzed. Thus the
spatial distribution of benefits from regulatory decisions may result in
voting outcomes which are far removed from those which maximize either
the profits of firms or the collective welfare of consumer-voters.
Indeed, these voting outcomes produce levels of total welfare and even
total consumers' surplus which are below those which would be
obtained if the spatial monopolist were not regulated at all. Given the
popular notion that uniform delivered pricing is "fair,"
politicians may find that both cultural pressures and self-interested
voting outcomes force them to require high rates of freight absorption
by firms subject to regulation.
APPENDIX
If we spin the linear market around the origin, a circular market
is generated. The total profit for such a circular market is
[Mathematical Expression Omitted] Following the same approach we
developed to deal with the linear market, we have for e [Epsilon] [0,
1/2], tR = 3([Beta] - m)/2(2 - e) and P(0) = [3(1 - e) [Beta] + (1 +
e)m]/2(2 - e); and for e [Epsilon] [1/2, 1], tR = 3([Beta] - m)/2(1 + e)
and P(0) = [(2 - e) [Beta] + 3em]/2(1 + e). A graph similar to Figure 5
can then be drawn accordingly: from 0 to [0.375R.sub.1/2 [prime]]
consumers prefer e = 1; from 0.375[R.sub.1/2] to 0.525R, consumers
located at r prefer an e that satisfies 2 [(1 + e).sup 2] r = 3; from
[0.525R.sub.1/2 [prime]] to [0.750R.sub.1/2 [prime]] consumers all
prefer e = 0; and finally, from [0.750R.sub.1/2 [prime]] to
[1.00R.sub.1/2 [prime]], consumers located at r prefer an e that
satisfies tr = 3([Beta] - m)/2(2 - e).
Therefore, we find that 50 percent of the consumers prefer e >
0.297 and 50 percent of the consumers prefer e < 0.297; 27.6 percent
prefer e > 1/2 and 72.4 percent prefer e < 1/2; 14.1% consumers
prefer e = 1 and 28.7% consumers prefer e = 0; and consumers at the
middle of the market (i.e., r = [0.5R.sub.1/2]) prefer e just above 1/2.
[Figures 1 to 5 Omitted]
(1)Analyses of spatial pricing and output decisions by firms have
traditionally considered three cases: mill pricing with zero freight
absorption, profit-maximizing spatial price discrimination, and uniform
pricing with complete freight absorption. A number of significant
results on the behavior of the spatial monopolist under these specific
alternative freight absorption schemes have been obtained beginning with
Singer [1937] and Hoover [1937] and extending to papers by Holahan
[1975], Beckmann [1976], Heffley [1980], and Hsu [1983]. In addition,
papers on spatial competition by Norman [1981], Villegas [1982] and
Capozza and Van Order [1977] have considered the effects of mill pricing
vs. spatial price discrimination on price, output, and welfare under
spatial oligopoly. An excellent summary of results is found in Greenhut,
Norman, and Hung [1987]. (2)Often the price of public services and
utilities does not vary with location of the consumer. Such nominal
uniform pricing does not automatically imply that all transportation
costs are absorbed. Differences in maintenance, service frequency,
repair service, emergency response time, etc. may result in higher total
cost to consumers located further from point of production. As in other
cases, economic analysis should go beyond statutory provision to actual
performance measures in measuring the degree of transportation cost
absorption. (3)We will "spin" the line around the origin to
discuss cases for circular markets below in section V. (4)A general form
of the demand function can be specified as q(P) = ([Alpha] - 1)
[P.sup.2] - [Alpha] [Beta] P + [[Beta].sup.2], which is well behaved because the choke price is [Beta] and the maximal demand is
[[Beta].sup.2] no matter what [Alpha] is, while [Alpha] controls the
concavity of the demand function: [Alpha] > 1 implies that [Alpha] is
convex, [Alpha] < 1 concave and [Alpha] = 1 for the linear case we
are studying. (5)As price rises to the choke price, quantity demanded
goes to zero. (6)From equation (3), the first-order condition, [Delta]
[Pi]/[Delta] e = 0, gives P(0) = ([Beta] + m)/2 + (1 - 2e) [Iota] R/3.
Combining this equation with the other first-order condition, equation
(6), we obtain that e = 1/2. It also can be shown that, for the general
demand function defined in footnote 3, we have the optimal e for profit
maximization is greater than 1/2 if the demand function is convex
([Alpha] > 1) and is less than 1/2 if the demand is concave ([Alpha]
< 1). (7)From the first-order conditions, [Delta] W/[Delta] e = 0 and
[Delta] W/[Delta] P(0) = 0, we have, respectively, P(0)-m = 2(1-e)tR/e,
and P(0)-m = (1-e)tR/2. Solving these equations simultaneously, we
obtain e = 1. (8)These results can be derived by defining [Mathematical
Expressions Omitted] and applying the optimal values of P(0) and R given
by equations (9) to (12). (9)The total surplus is calculated according
to equations (4) and (5). These results are illustrated in Figure 2.
(10)These results are derived from equation (3). (11)These results are
derived by adding [Pi] and S together. (12)A detailed derivation of this
graph is available from the authors upon request. (13)A detailed
derivation of this graph is available from the authors upon request.
(14)From Figure 5 we can see that 50 percent of the consumers along the
market prefer e > 0.333 and 50 percent prefer e < 0.333. About
33.3 percent of all consumers prefer e > 0.5 and 66.7 percent prefer
e < 0.5. Finally, approximately 33.3 percent of the consumers prefer
e = 0 and 22.2 percent prefer e = 1. Consumers located at the middle of
the market (where r = 0.5R1/2) prefer e = 0. (15)See the appendix for a
derivation.
REFERENCES
Beckmann, Martin J. "Spatial Price Policies Revisited."
Bell Journal of Economics, December 1976, 619-30. Capozza, Dennis R. and
Robert Van Order. "Pricing Under Spatial Competition and Spatial
Monopoly." Econometrica, December 1977, 1329-38. Greenhut, John and
Melvin L. Greenhut. "Spatial Price Discrimination, Competition and
Location Effects." Economica, November 1975, 401-19. Greenhut,
Melvin L., George Norman, and Chao-Shun Hung. The Economics Of Imperfect
Competition: A Spatial Approach. New York: Cambridge University Press,
1987. Heffley, Dennis R. "Pricing in Urban Spatial Monopoly: Some
Welfare Implications for Policies Which Alter Transportation
Costs." Journal of Regional Science, May 1980, 207-25. Holahan,
William L. "The Welfare Effects of Spatial Price
Discrimination." American Economic Review, June 1975, 498-503.
__."Spatial Monopolistic Competition versus Spatial Monopoly."
Journal of Economic Theory, January 1978, 156-70. Hoover, Edgar M.
"Spatial Price Discrimination." Review of Economic Studies,
January 1937, 182-91. Hsu, Song-ken. "Pricing in an Urban Spatial
Monopoly: A General Analysis." Journal of Regional Science, May
1983, 165-75. Mueller, Dennis C. Public Choice II. New York: Cambridge
University Press, 1989. Norman, George. "Spatial Competition and
Spatial Price Discrimination." Review of Economic Studies, January
1981, 97-111. Singer, H. W. "A Note on Spatial Price
Discrimination." Review of Economic Studies, January 1937, 75-77.
Villegas, Daniel. "Comparative Performance of Spatial Models of
Economic Markets With Linear Household Demand." Southern Economic
Journal, December 1982, 893-903.
MENG-HUA YE and ANTHONY M.J. YEZER, Department of Economics, George
Washington University. Helpful comments were made by an editor, referee,
and a number of our colleagues at department seminars.