Market Areasi
Market Areasi
Market Areasi
AND FIRM
LOCATION ANALYSIS
19
input-oriented firms pay more to transport inputs. Some firms cluster in a
region and thus influence the cost structure of all enterprises in the same
area, generally by lowering the production costs. Then, even more firms will
be attracted to the region because of the cost advantages resulting from the
agglomeration. Lower production costs that benefit firms clustered together
are called agglomeration economies.
This part examines various models of firm location behavior. Chapter 2
and Chapter 3 focus on what Fujita and Thisse (1997, 2002) call ―shopping
models.‖ Shopping models are appropriate for studying the competitive
behavior among firms that sell consumer goods, as opposed to shipping
models, which describe competition among producers. Shopping models
assume that the costs of labor and other inputs are the same everywhere.
The only factors that change with the location are the number of consumers
that the firm can attract and the number of rival firms.
A model of urban hierarchy that determines which cities more efficiently
provide what services stems from the simple model of market areas. Chapter
3 explores theories that explain the relationships among neighboring cities
of different sizes, that is, theories of hierarchies, networks of cities, and
central places. The concept of a central place is rooted in the fact that each
type of product requires a market area of a certain size. The combinations of
products sold in a city changes with its size. Consumers in smaller cities
need to travel to an urban center to find specialized products.
Chapter 4 and Chapter 5 explore ―shipping models.‖ In Chapter 4, when
we relax the assumption that the costs of production are equal everywhere,
we discover that the costs of the inputs and transportation costs produce a
predictable effect on the location of enterprises. Chapter 5 concentrates on
the effects of agglomeration economies on firm location. Agglomeration
economies, along with the theory of cumulative causality, help shape the
foundation of the endogenous growth theory, (New Growth Theory) that is
supported by development economists, international economists,
and macroeconomists.
Chapter 6 concentrates on spatial pricing patterns. These pricing patterns
refute the myth that monopoly is the only market structure capable of
practicing price discrimination. Price discrimination consists of selling the
same product at different prices and to different consumers without incurring
any increase in cost. The chapter also reveals the ineptitude of some antitrust
laws, such as the Robinson-Patman Act, which actually create and sustain
local monopolies while claiming to encourage competition. Chapter 6
highlights the inconsistency between a basing-point pricing system—which
is now illegal but which the American steel companies used until 1924—and
the ―legal by definition‖ basing-point pricing system that the U.S. federal
government uses to establish prices. The steel mills in Europe currently use a
basing-point pricing system despite Article 85 of the Treaty of Rome that, in
principle, forbids commercial restrictions in the European Union.
Chapter 2
21
the store price, also called the mill price (or gate price for agricultural
products), plus the transportation costs that are incurred in traveling to the
store. This transportation cost includes more than the cost of gas, oil,
insurance, and depreciation for an automobile, or than the cost of bus, taxi, or
subway fares. The most important component of transportation cost is the
opportunity cost of the time involved in traveling to and fro m the store.
We assume that the spatial monopolist imposes an f.o.b. mill price,
which means that consumers pay the mill price (the price that the store
charges) plus all transportation costs. Under f.o.b. pricing, if the consumers
do not go to the store to buy the product, they can order the product from
the dealer and pay all shipping charges. The opposite of f.o.b. pricing is
uniformed delivered pricing, also known as cost, insurance, and freight
(c.i.f.) pricing. Under uniform delivered pricing, the firms pay all trans-
portation costs. Firms that offer free delivery (for goods such as flowers,
pizzas, prescriptions, or groceries) employ a uniform delivered pricing
strategy, and thereby become price discriminators.
Panel A Panel B
Price-distance function
Price
Price
$10
$4
Demand
8 Distance 5 Quantity
The spatial monopolist cannot control all market area changes. Road
construction, considerably higher gasoline prices, or more traffic congestion
increase the transportation costs per mile, making the price-distance
function steeper. In Figure 2.3, for example, if the price of the product is
PD2
Price PD1 Price
PD3
$6
$4
$2
Demand
Distance
Figure 2.2 Price-distance functions and demand when the mill (store) price
changes.
$4, but the transportation costs per mile double to $1.50, the market area
radius shrinks from 8 to 4 miles. However, if one constructs a more efficient
road system that cuts down the time needed to drive to the store and the
transportation costs per mile fall by 50¢, then the market area radius
expands to 12 miles. (Details are in the Appendix.)
Rosser (1991)
PD2 PD1
Price Price
PD3
$4
Demand
4 8 12 5 Quantity
Distance
Figure 2.3 Price-distance functions and demand when transport costs change.
Finally, the market area changes when the demand curve shifts. If
demand for the product increases, pushing the limit price to $12 (rather
than $10), the market area boundary will increase from 8 miles to 10.67
miles. If instead the limit price falls to $6, the boundary decreases to 2.67
miles. The graph of this instance is left for the reader to draw and analyze.
(See the Appendix for a hint.)
In summary, the mill price, the cost of transportation, and the demand
for the firm‘s product determine the market boundary for a spatial monopo-
list. Market boundaries shrink if the firm charges a higher price, if travel costs
increase, or if demand decreases. Market area boundaries expand when the
firm lowers its price, when travel costs decrease, or when demand increases.
%DQ
. ð2:1Þ
. %DP .
Inelastic demand
Demand
Distance Quantity
The midpoint on a demand curve is the point of unit elasticity because the
percent change in quantity equals the percent change in price. Above this
point, the elasticity coefficients are greater than one, and the demand is
called ―elastic.‖ The demand is ―inelastic‖ when the elasticity coefficients
are less than one on the lower portion of the demand curve.
Because the limit price is ten and half of ten is five, at a price of $5 the
demand curve of Figure 2.4 is unit elastic. Therefore, because consumers
who live right next to the firm pay $4 for the product, their demand is
inelastic. If customers living 1 mile from the point of sale incur $1 in
transportation costs, they face a delivered price of $5, and consequently,
have a unit elastic demand. Consumers who live farther pay yet a higher
delivered price and have an even more elastic demand. Those who live 8
miles from the firm are the most responsive to price changes and face a
nearly perfectly elastic demand. Knowing this, firms try to decrease
delivered prices for distant consumers for whom the demand is elastic,
while increasing prices for local customers who have inelastic demand. We
will see in chapter 6 some examples of spatial price discrimination
conditional upon distance from the point of sale.
A B
second quartile
A B
Distance
Figure 2.6 Market areas for a duopoly under perfectly inelastic demands.
Hotelling also used the principle of minimum differentiation to justify
why ―buyers are confronted everywhere with an excessive sameness‖ in
retail products. If we consider ―location‖ with respect to the attributes of a
product, the principle of minimum differentiation establishes that firms
producing heterogeneous products can alienate certain groups of consu-
mers. Retailers, consequently, will tend to
$10
$4
16 8 0 8 16
A
Distance
Figure 2.7 Market area for a monopoly when the demand has a limit price.
$11
$9
$7
i Firm A k j Firm B
Q1 Q2 Q3
Figure 2.8 Market areas for duopolists when their demand curves have a limit price.
point i shops at Firm A. The consumer at point j prefers to shop at Firm B
and pay a delivered price of $7 rather than $11 at Firm A. Customers located
at the median point of the market area (point k) are indifferent between A
and B. Point k is therefore referred to as the break-even location. At either
store, the delivered price for consumer k is $9. The market area for FirmA is
on the left of the market and Firm B‘s market area is on the right.
From the previous diagram, it seems that consumers always shop at the
nearest establishment. This assertion is true if the stores sell identical
products at the same price and if their consumers face identical trans-
portation costs. However, Figure 2.9 shows one situation where consumers
may wish to avoid the nearest store and shop at a store farther from their
house. Why?
Firms C and D are located on quartiles of the bounded market area.
Both stores charge the same price ($4) for identical products. Trans-
portation costs for Firm D are $1.50 per mile, but only 25¢ per mile for C.
Firm D may have either inadequate parking or difficult access. Because of
the high transportation costs, only consumers within approximately 2
units of distance around Firm D will shop there. Consumers who reside
more than 2.34 units of distance to the right of D will actually save on
transportation costs by shopping at C. The market area of Firm D
resembles an island inside C‘s market area. (Calculations are in the
Mathematical Appendix.)
C D
Distance
Figure 2.9 Demand curves for a duopoly when one has higher transport costs.
B" C’ A B A" C B' A' C ''
Three’s a Crowd
Hotelling‘s model was extended by Lerner and Singer (1937) and Eaton and
Lipsey (1975) to explain how more than two firms (carts on wheels)
compete for the same market. If more than two firms sell identical products
in a linear market, competition forces them to eventually disperse. If three
firms try to locate next to one another, the firm in the center will have too
narrow of a market for survival. Because moving is costless, the firm in the
middle of the trio will move down the street, leapfrogging over the other
firms until the three share one-third of the market area equally
(Figure 2.10). In the same way, four firms would share one-fourth of the
market, and so forth. Entry will continue as long as each firm‘s share of the
market permits it to cover its costs including a normal rate of return. The
minimum market area needed to limit a firm to a normal profit is called its
threshold size.
What happens to entry if each of the firms faces a slightly greater than
normal rate of return but the sum of the excess profits is not enough to
sustain the coexistence of one more firm? The excess profits continue
until (1) the market grows large enough to permit an additional firm to
eke out a living, (2) a new firm benefits from lower production costs, or
(3) a new firm benefits from an owner who is content with less than a
normal profit.
Firm E Q2 Firm F
Q1 Q3
1
For more information about subgame perfect Nash equilibrium concepts in game
theory, Drew Fudenberg and Jean Tirole, Game Theory (and London: MIT Press, 1991).
$
Q1 E F Q3
model assumes that all goods are homogeneous and that all consumers buy
from the seller with the lowest price. Firm A chooses its price assuming that
FirmB will not react. Once FirmA sets a new price, FirmB has three options.
(1) It could keep a higher price (and lose all sales); (2) it could set an equal
price (and share the market), or (3) it could set an even lower price and
capture the entire market itself. In Figure 2.12, the last choice is the most
profitable and the one that FirmB will choose—again, assuming that FirmA
will not respond.
Of course, Firm A reacts and faces the same three choices, Firm A is
going to surprise Firm B by choosing to set an even lower price. The price
wars wage on until the point where all prices equal the marginal cost of
production. Prices may even dip below the marginal cost for short periods
of time if one firm tries to drive its rival out of business.
As the Bertrand theory suggests, stores along a well-traveled route practice
price shading, that is, they set a price just below that of their competitors.
Price shading increases market share as long as the price charged is higher than
the marginal cost. The store that begins this price war hopes that in the end
some firms will go out of business and others will not dare to enter the market.
As rival firms leave the market, demand for the remaining firms
temporarily increases until the supernormal profit attracts other entrants.
Firms will realize above normal profits, according to d‘Aspremont,
Gabszewicz, and Thisse (1979), only if they locate as far away from each
other as possible. The greater the distance separating firms the more control
they have over their own price because a longer distance makes comparison
shopping more difficult. The resulting dispersion is by no means nirvana,
however. Firms farthest from the market center discover that their customer
base is smaller and are lured back toward the center. At the market center,
they will again be pressured to disperse by fierce price wars. Firms will
never be satisfied with their location.
2
de Palma et al. (1985); Claycombe (1991).
products rather than transportation costs. 3 Consumers may not purchase a
good with the lowest delivered price if they are willing and able to pay more
for a good with a specific attribute. For example, a shopper looking for a
pair of shoes to exactly match the color of an evening gown will search
where there is a large number of shoe stores. The result will be larger
selections and lower prices.
3
Dudey (1990).
4
Thill (1992).
agglomerate. Antique dealers, shoe stores, and used
car dealerships tend to cluster together. Antiques are
objects that are very difficult to compare, and collectors
are often open to all possibilities and any good deal.
Likewise, finding the correct pair of shoes requires
physically inspecting them and trying them on. Used
automobiles are heterogeneous and a costly invest-
ment. Determining their quality requires both visual
inspection and testing.
On the other hand, Fischer and Harrington found that
video stores, movie theaters, gasoline stations and
supermarkets are more widely dispersed. For these,
there is little product differentiation and no reason to
comparison shop. Video stores may differentiate them-
selves by specializing in a specific type of film, but all of
them carry the same new releases. As for cinemas,
consumers who chose their movie, see it in the
closest theater. Customers who merely want to fill up
their gas tanks consider all vendors of gasoline as
sellers of homogeneous products. Likewise, those
who just need a gallon of milk will treat all stores that
sell dairy products, including service stations,
as homogeneous.
Chapter Questions
Price Price
Demand
Distance Quantity
2. Frank and Jessie James have started up convenience stores that are
almost identical. They are located at quartiles in a linear market area.
The only difference is that Frank‘s prices are higher than Jessie‘s, and
Frank‘s transportation costs are also greater than Jessie‘s. Diagram
this case and determine the market areas for both Frank and Jessie.
3. According to Harold Hotelling, in an election campaign ―Every
candidate ‗pussyfoots,‘ replies ambiguously to questions, and
refuses to take a definite stand in any controversy for fear of losing
votes.‖ Depict this phenomenon graphically using Hotelling‘s model.
4. Pat Jevons from Hardly Burgers has decided that because a large
enough market exists, it would be profitable to open a kiosk-
restaurant next to the university. Pat‘s problem is that Chris Señior
of Windful Sandwiches had also chosen a site in the same market area.
The demand for fast food around campus at noon is perfectly inelastic,
and surveys conclude that consumers consider Hardly Burgers and
Windful Sandwiches to be basically the same thing. Changing
locations is costless. Production costs for both firms are equal.
a. Using Hotelling‘s traditional theory, where should Pat locate?
Diagram and briefly explain your answer.
b. According to the argument made by d‘Aspremont, Gabszewicz,
and Thisse (1979), is this location a stable equilibrium? If not, what
if anything could Pat and Chris do to make their existence in the
market more stable?
5. Differentiate between multipurpose and multi-stop shopping. How
does commuting influence these shopping patterns? If you were
designing a shopping mall, where in your city would you suggest it
be located and what types of stores would you require to take
advantage of travel-to-shop behavior?
6. Claycombe and Mahan (1993) determined that beef prices are lower in
areas where commuting by automobile is highest. Would you predict
the same type of behavior for clothing, furniture sales, and automobile
sales? Why or why not?
Research Assignments
1. Check out the Web site for travel times to work and means of
transportation to work for your city or county through the census
at http://www.census.gov (QT-P23; from ST-3 data) and compare
these data to others in your class.
a. When do most people leave home to go to work in your area?
b. What is the mean travel time to work?
c. What proportion of commuters drive alone?
d. If your city is in a metropolitan statistical area (MSA), check out
the American Chamber of Commerce Research Associates
(ACCRA) Cost of Living Index for the average price of a T-bone
steak in your area and compare it with other MSAs that the class is
studying. Do you notice a trend in beef prices and ―proportion of
commuters who drive alone‖ when comparing your data to that
of others in the class?
2. Describe a retail center in the city that you are studying. Which stores
sell complementary goods? Which stores sell substitute goods? How
could your retail center increase the marginal benefit for its
customers?
Mathematical Appendix
Algebraically, Panel B of Figure 2.1 is a demand function for the product.
The limit price, the maximum price that consumers will pay for any quantity
of the good, is $10. The equation for the demand function for this example is
QZ5K1⁄2 P, where Q is the quantity demanded and P is the price. By solving
for P, we will get the inverse demand function: PZ10K2Q. The inverse
demand function allows one to easily graph a demand function with price
on the vertical axis. If the mill price is $4 and transport costs are 75¢ per mile,
the price-distance function as graphed in Panel A of Figure 2.1 is PZ4C
0.75D, where D is the distance from the store to the consumer‘s residence.
To find the radius of a market area, set the price-distance function equal to
the limit price of the demand function and solve for D. In this example,
where $10 is the limit price, 4C0.75DZ10 and DZ8. Thus, the farthest
distance that people will travel at the current mill price is eight units.
The equation for PD1 in Figure 2.2 is PZ4C0.75D. When the mill price
increases to $6, the equation for PD 2 becomes PZ6C0.75D. A drop in the
mill price to $2 changes the equation for PD3 to PZ2C0.75D. Setting each
of these three equations equal to the limit price of $10 generates the market
area radii of 8, 5.33, and 10.67, respectively.
In Figure 2.3, the equation for the price-distance function for PD1
remains at PZ4C0.75D, but the equation for PD2 becomes PZ4C1.5D
because transport costs per mile are $1.50 and people will no longer travel
even 4 miles to the store. However, if transport costs per mile decrease to
50¢, the equation for the price-distance function becomes PZ4C0.5D, as in
PD 3. The market area radius expands to 12 miles.
Market areas also change with demand. When the equation for the price-
distance function (PZ4C0.75D) equals a limit price of $12 instead of $10,
the radius of the market area extends to 10.67 [(12K4)/0.75]. But when the
limit price is $6, the radius drops to 2.67 [(6K4)/0.75].
The price-distance function for the monopolist in Figure 2.7 is still PZ
4C0.75D. Setting this equation equal to $10 limit price, 4C0.75DZ10,
creates a market area radius of 6/0.75Z8 units of distance.
In Figure 2.9, if the price-distance functions for C and D intersect at $6.55,
then the break-even point is 10.2 units of distance right of C (where 4C
0.25DZ2.55 or DZ2.55/0.25Z10.2) and 1.7 units to the left of D (where
$4C1.5DZ2.55 or DZ2.55/1.5Z1.7). Similarly, the price-distance func-
tions equate to the right of D at $7.57, a point 2.38 units to the right of D, and
14.28 units right of C.
Questions to Mathematical Appendix
1. Assume that the demand curve is QZ5K1⁄4 P. The mill price is $5 and
transportation costs are 50¢ per mile.
a. Find the inverse demand function.
b. Find the price-distance function.
c. Find the radius of the market area.
2. What is the new radius of the market area in Question 1 if the
transportation costs increase to $1.00 per mile? What if they decrease
to $0.25 per mile?
3. What is the new radius of the market area in Question 1 if the mill
price drops to $2.50? What if it increases to $7.50?
4. What is the new radius of the market area [(12K4)/0.75] in Question 1
if demand changes to QZ7.5K1⁄4 P ?
5. What is the new radius of the market area if demand for the firm in
Question 1 changes to QZ2.5K1⁄2 P ?
References
Claycombe, R. J. 1991. Spatial retail markets. International Journal of Industrial
Organization 9 (2):303–313.
Claycombe, R. J., and T. E. Mahan. 1993. Spatial aspects of retail market structure:
beef pricing revisited. International Journal of Industrial Organization 11
(2):283–291.
d‘Aspremont, C., J. J. Gabszewicz, and J. F. Thisse. 1979. On Hotelling‘s ―Stability in
competition‖. Econometrica 47 (5):1145–1150.
de Palma, A., V. Ginsburgh, Y. Y. Papageorgiou, and J.-F. Thisse. 1985. The principle
of minimum differentiation holds under sufficient heterogeneity. Econometrica
53 (4):767–781.
Dudey, M. 1990. Competition by choice: the effect of consumer search on firm
location decisions. American Economic Review 80 (5):1092–1104.
Eaton, B. C., and R. G. Lipsey. 1975. The principle of minimum differentiation
reconsidered: some new developments in the theory of spatial competition.
Review of Economic Studies 42 (1):27–49.
Fischer, J. H., and J. E. Harrington. 1996. Product variety and firm agglomeration.
RAND Journal of Economics 27 (2):281–309.
Fudenberg, D., and J. Tirole. 1991. Game Theory. Cambridge, Mass: MIT Press.
Fujita, M., and J.-F. Thisse. 1997. Economie géographique, problèmes anciens et
nouvelles perspectives. Annales d’Economie et de Statistique 45 (37):87.
Fujita, M., and J.-F. Thisse. 2002. Economics of Agglomeration: Cities, Industrial
Location, and Regional Growth. New York: Cambridge University Press.
Hanson, S. 1980. Spatial diversification and multipurpose travel. Geographical
Analysis 12:245–257.
Hotelling, H. 1929. Stability in competition. Economic Journal 39 (153):41–57.
Lerner, A. P., and H. W. Singer. 1967. Some notes on duopoly and spatial
competition. The Journal of Political Economy 45 (2):145–186.
O‘Kelly, M. E. 1981. A model of the demand for retail facilities, incorporation multi-
stop multipurpose trips. Geographical Analysis 13:134–148.
Rosser, J. B. Jr. 1991. From Catastrophe to Chaos: A General Theory of Economic
Discontinuities. Norwell Mass: Kluwer Academic.
Thill, J. C. 1992. Spatial duopolistic competition with multipurpose and multi-stop
shopping. Annals of Regional Science 26 (3):287–304.
Chapter 3