iBusiness, 2010, 2, 295-299
doi:10.4236/ib.2010.23037 Published Online September 2010 (http://www.SciRP.org/journal/ib)
The Evolution of Contestable Markets: A
Computing Simulation
Zhenguo Han1, Hui Zhang2, Minrong He1
1Southwest University of political Science and Law, Chongqing, China; 2Communication University of China, School of Science,
Beijing, China.
Received January 19th, 2010; revised April 29th, 2010; accepted June 21st, 2010.
ABSTRACT
This paper introduces a computing simulation of the evolution of contestable markets. The results show that the evolv-
ing outcome of contestable markets is a monopoly when the market demand is fixed. This computing simulation also
tells us how information to influence the evolution of contestable markets. In the end, some advices are provided.
Keywords: Stochastic Pattern, Contestable, Computing Simulation
1. Introduction
People who study economics always regard “perfect
competition” from neoclassic theory as a standard coor-
dinate of market structure. In the model of perfect com-
petition economic agents have no confliction in a direct
and intentional way. A perfectly competitive market rests
on four basic assumptions: 1) consumers and producers
are price taker. In other words, in the market there are
many independent firms and independent consumers who
believe correctly that their decisions will not affect prices;
2) product is homogenous. All products of the firms in a
market are perfectly substitutable with one another; 3) all
factors can be entry and exit freely and 4) the buyers and
the suppliers in the market have complete information.
Competition can ensure ‘the invisible hand’ taking effect,
so the perfectly competitive markets are ‘Pareto efficient’
[1]. But in a real market structure the perfect competition
is not the reality because incomplete information exists.
In the early 1980s, Baumol (1982) advanced the con-
testable markets theory, in which the competitive press
from potential entry firms limits existing firms very
strongly. The contestable markets theory emphasizes the
effective restrictions for the behavior of existing firms
[2]. These restrictions come from competition which is
from potential entry firms not from existing firms. In
other words, the contestable markets are completely
similar to the perfectly competitive markets because of
two reasons. One reason is that the role of the potential
entry firms can be regarded as the role of many existing
firms. As a result, the behavior of existing firms looks
like theirs in a perfectly competitive market though the
number of existing firms is not too much. The other rea-
son is that in this market the press from potential compe-
tition can also compel the existing firms to improve their
conductivity and to make their cost equal their revenue
because resources can enter and exit freely. As a result,
this market equilibrium is the same as the equilibrium of
the perfectly competitive market. If a firm can enter
without too large sunk cost, then this firm can be re-
garded as market participator. Therefore, many firms can
make a market become a perfectly competitive market,
but it is not essential condition.
The traditional economic theory emphasizes static
equilibrium more in a market. When the market realizes
a static equilibrium, many firms will coexist if each indi-
vidual firm sells a sufficiently small proportion of total
market output. The purpose of this model is to test how
stochastic consumption pattern influences the evolving
outcome of contestable markets by computing simulation.
The result of computing simulation shows that stochastic
consumption pattern will make existing firms exit from
the market and be inclined to increase market concentra-
tion, and in the long run, the market will become monop-
oly. Under the product homogeneity, stochastic con-
sumption pattern is a natural power to push a market to
concentration even if the competitive mechanism does
not take effect completely. The evolving trend of a mar-
ket can be expressed as follows:
Perfect competition→Monopolistic competition→oli-

The Evolution of Contestable Markets: A Computing Simulation
296
gopoly→monopoly
2. Case 1 [3,4]
This model makes use of NetLogo running circumstance.
NetLogo is a programmable modeling environment for
simulating natural and social phenomena. It was authored
by Uri Wilensky in 1999 and is in continuous develop-
ment at the Center for Connected Learning and Com-
puter-Based Modeling. We divide background into a
number of patches which present buyers. Each patch is
painted distinct colors which present products produced
by different firms, and thus buyers and sellers together
form a market. That means the number of firms is limited.
When the color of a patch changes, it means that this
buyer denoted by the patch chooses other product which
is produced by the corresponding firm denoted by the
new color. If one color disappears, it means that the cor-
responding firm exits from the market. The more patches
are painted one color; the higher percent the products of
the firm corresponding to the color occupy the market.
This model simulates the situation of contestable markets
perfectly. The reasons are as follows:
Any of buyers in the market can obtain products made
by any existing firms in time when they need, and firms
can supply their products quickly. In this way, buyers
and firms become price taker, and sellers and buyers do
not affect price, so the first assumption of a perfectly
competitive market holds. For the second assumption,
because the role of colors is only to tell that difference of
firms and customers’ selection is random, we can believe
product homogeneity. In our given market, buyers can
buy products made by any existing firms, and firms can
supply their products without any limitation when buyers
need, so resources are entry and exit freely. Though we
define other firms can not enter after the market is
formed, the third assumption of a perfectly competitive
market is satisfied because of the competitive press from
potential entry firms. This model regards the selecting
scope of buyers and the supplying scope of firms as in-
formation, so that buyers can select any existing color
and firms can supply any patches. This means complete
information for buyers and firms. The fourth assumption
of a perfectly competitive market holds.
The number of buyers contained in this model is 1089.
First we select the number of firms, and then we let the
model run. During the running process, the behavior of
buyers who select product is random and buyers select
varied product by changing patches’ color. The different
color reflects the different firm which supplies different
product. We assume buyers randomly select products in
this term as their consumption in next term. In other
words, buyers take their action according to stochastic
consumption pattern. Because of the assumption of the
contestable markets, the competition of potential entry
firms will compel price to keep constant.
2.1. The Original Firms Distribution does not
Affect the Evolving Outcome of Contestable
Markets Based on Stochastic Consumption
Pattern
The buyers who have complete information select their
product randomly from all of existing firms, and the
firms which have complete information supply their
products in time when buyers need. When the model
begins running, the same simulation will be repeated six
times because of the stochastic original distribution and
the stochastic action among agents. Then we analyze the
outcome of computing simulation (note: different color
denotes different firm). For example, we use N1-N10 to
denote ten varied firms. We simulate the evolving proc-
ess of contestable markets when the number of firms is
ten, eight and six respectively. Under the same condition,
we do six simulations. See Table 1.
Table 1 lists the most superior firm and the most infe-
rior firm before simulation begins under the original
product’s random distribution.
In the simulation process, the evolving outcome shows
that either the most superior firm (such as the fourth
simulation when the number of firms is eight or the fifth
simulation when the number of firms is six) or the most
inferior firm (the fifth simulation when the number of
firms is ten) was winner, but there are more intermediate
cases. The conclusion is that the original product’s dis-
tribution does not affect the final evolving outcome. In
other words, the final winner does not depend on the
original product’s distribution.
When the number of firms is eight, the third evolving
process can show that N6 and N4 which have been supe-
rior once exit from the market successively, and N2 beats
N4 by selection based on stochastic consumption pattern
after long time. The conclusion is that the superior firm
may be driven out from the market so long as competi-
tive firms exist.
2.2. Buyer’ Incomplete Information can Slow the
Evolving Speed of Markets
Two cases are designed to test how buyer’ incomplete
information to affect the evolving speed of markets: 1) a
buyer randomly selects any of four products consumed
by his upper, lower, left and right neighbors in this term
as his consuming product in next term, and 2) a buyer
randomly selects any of eight products consumed by his
eight neighbors in this term as his consuming product in
next term. In order to weak the effect of the original dis-
tribution greatly; we simulate the evolving process 50

The Evolution of Contestable Markets: A Computing Simulation 297
Table 1. The evolution of market with the complete information.
Number First Second Third Fourth Fifth Sixth
Superior firm at first1 N8 N1 N3 N8 N2 N10
Inferior firm at first2 N7 N5 N2 N3 N7 N9
Simulating result N1 N6 N5 N5 N7 N3
Ten firms
Number of buyers selection 1201 427 1192 1660 577 1175
Superior firm at first N7 N8 N6 N3 N1 N2
Inferior firm at first N2 N7 N8 N1 N6 N7
Simulating result N3 N1 N2 N3 N4 N5
Eight firms
Number of buyers selection 400 1688 1156 822 1564 923
Superior firm at first N2 N7 N4 N5 N7 N2
Inferior firm at first N1 N5 N1 N1 N5 N7
Simulating result N4 N6 N7 N4 N7 N8
Six firms
Number of buyers selection 963 1064 408 562 836 1316
1Superior firm at first is the firm which its products is the most proportion in the market when the simulation begins; 2Inferior firm at first is the firm which its
products is the lowest proportion in the market when the simulation begins.
Table 2. Influence of buyers’ information to markets evolving speed.
Firm number
Buyers information 10 8 6
Complete information (times) 1096.6 1084.7 954
Eight neighbors information (times) 965.9 1002.3 851.2
Four neighbors information (times) 1195.8 1290.4 1193.4
times respectively when the number of firms is ten, eight
or six.
Table 2 shows the mean number of the evolving proc-
ess when information is complete and incomplete.
Table 2 shows that stochastic consumption pattern did
not affect the evolving process of market more according
to eight neighbor’s information or complete information.
This conclusion can be explained as follows:
Because the firm’s information is complete and the
number of firms is not obviously dominant compared
with the number from the neighbors, eight neighbors’
information can be regarded as complete information. In
other words, buyers can obtain the same information
from their eight neighbors as complete information when
the number of firms is similar to the number of neighbors.
As a result, the market’s evolving speed is almost equal.
When the information decreases half, the number of
firms is obviously dominant compared with the number
from the neighbors. Four neighbors’ information which
can be regarded as incomplete information influences the
market’s evolving speed. Comparing to complete infor-
mation, in the evolving period the incomplete informa-
tion increases about 15% when the number of firms is 8
or 10, and about 30% when the number of firms is 6.
Though the data is random, the conclusion that the
buyer’s incomplete information can slow the market’s
evolving speed is reasonable.
The number of original firms how to influence the
market’s evolving result
Table 2 shows that the market’s evolving outcome
seems to be independent of the number of original firms
and the market’s final evolving outcome is monopoly.
This conclusion should be checked because the number
of firms changes too small (from 10 to 6), so the Case 2
was designed to study.
3. Case 2
In Case 1, the firm’s information is complete and each
firm always supplies their product quickly when buyers
need. Now we assume that the market aggregate demand
is constant in each term, that firms’ information is in-
complete, and that each firm increases their output ran-
domly. In order to find how the number of original firms
and the promotion of firms’ sales to influence the mar-
ket’s evolution, Case 2 was designed. In Case 2, each
firm is denoted by an agent with an ID, and the number
of products made by one firm is denoted by the number
of the same ID. Less than 10% firms randomly increase
their output which is 0% - 3% of the market aggregate
demand in order to extend their market proportion. If the
whole output exceeds the aggregate demand, the market
will randomly eliminate unnecessary products. When a

The Evolution of Contestable Markets: A Computing Simulation
298
Figure 1. The market proportion of firms during the first
stage of simulation.
Figure 2. 9 firms in the market during the later stage of
evolution.
firm’s product does not exist in the market, the firm will
exit from the market. The market aggregate demand is
designed 1089 so that it is comparable with Case 1.
Figure 1 shows that the market proportion of firms
with different ID is almost the same during the first stage
of simulation. The market will randomly eliminate un-
necessary products when the whole output exceeds the
aggregate demand. And when a firm’s product does not
exist in the market, the firm will exit from the market.
Figure 2 shows the evolving process from 500 firms to 9
firms in the market. Though the number of firms is in
variety during the original market, the evolving outcome
is also oligopoly, and then monopoly. The winner is in-
dependent of original firms’ distribution.
Table 3 shows the average time period of 12 times’
evolution from the number of original firms which is 500,
250, 100, 50, 25 respectively to number of firms which is
10, 5, 3, 2, 1, and the average evolving period from 10
firms to 5, to 3, to 2, and to 1.
Table 3 shows that the evolving period from 500, 250,
100, 50 and 25 firms respectively to 10 firms and firms
under 10 increases progressively. The result depends on
the market demand. When the market demand is constant,
the more the number of firms is, the more furious com-
petition is, and the faster the superfluous firms exit from
the market. But as the number of firms decreases, the
period which drives firms out from the market will pro-
long. For example, in the evolving process we find that
the average evolving period from 3 firms to 2 firms is 1.5
times of that from 5 firms to 3 firms, and the average
evolving period from last stage to 1 firm will increase 3.6
times. In addition, the less the number of original firms is,
the less the number of products by randomly increasing
0-3% proportion of the market demand in each term is,
and the less competitive press is. If the number of firms
is 25, only 0-2 firms (less 10%) increase their output, and
the number of increasing products is only 0-65 (0-3% of
the market demand). As a result, the evolving speed is
very slow. The conclusion is that monopolistic competi-
tion and oligopoly will be last fairly long time before
monopoly though the final evolving outcome of markets
is monopoly.
The mode can explain the evolving outcome of elec-
trical equipment markets in China. In a not long period,
the market demand does not change too much, and the
channels that buyers obtain information are more than
the number of manufacturers. In this case, if the number
of manufacturers is very much, the evolving outcome
should be oligopoly in a not long period. In fact, under
the years’ competition in the 1990s from the main elec-
trical equipment markets, the proportion of markets is
concentrated to several brands quickly. Air-conditioner’s
market can be as an illustrative example. In 2000, the
number of Air-conditioners’ brand was about 400 in
electrical equipment market in China, and about 140 in
2003. It means that the rate of elimination is about 30%
every year. In 2004, the number of main brands is only
about 50; the rate of elimination is about 60% in this year.
Table 3. The evolving result of model 2.
Decrease to
(number)
number of original firms
10 5 3 2 1  10 10  5 5  3 3  2 2  1
500 32.6 67.6 109.3 163.2 276.3 32.6 35 41.8 53.8 113.1
250 38.8 76.7 122.8 174.8 294.9 38.8 37.9 46.2 52 120.1
100 54.4 111 184.7 267.1 623.8 54.4 56.6 73.7 82.4 356.8
50 50.6 127.9 250.3 386.3 1063.2 50.6 77.3 122.3 136 675.9
25 62.8 127.6 311.3 672.9 1207.9 62.8 64.8 183.8 361.6 1207.9
Average evolving period: 47.8 54.3 93.6 137.1 494.8

The Evolution of Contestable Markets: A Computing Simulation 299
In five years, nearly 90% air-conditioner’s manufacturers
exit from the market. In other electrical equipment mar-
kets in China, the situation is similar-over 60% propor-
tion belonging to the first ten brands. In the retail market
of electrical equipment, the market concentration is in-
creasing, and the growth rate of turnover of the main
retail traders, such as SuNing, GuoMei, are over 50%
every year. The evolving outcome is fit for the Case 2.
The current situation of electrical equipment market
has some difference compared with the model in this
paper. The main reason is that personal income increases
more and the price of electrical equipment decreases
quickly. These factors make the market demand change;
therefore the market concentration is lower than that of
the model simulation. This tells us how to effectively use
the surplus capacity of electrical equipment trade. This
conclusion gives an answer-adopting vigorous measure
to exploit the rural market and to increase export.
The life of a product how to influence the evolving
speed
The number of buyers’ selection times the life of the
product equals the evolving time from the contestable
markets to monopoly. So if other factors are fixed, the
longer the life of the product is, the slower the evolving
speed is. Even if the life of a product is only one day, the
evolving time from the contestable markets to monopoly
needs at least 1.36 years (Case 1), and 2.6 years (Case 2).
If the life of a product is one week, the evolving time
needs 9.5 years to 18 years. The conclusion is that mo-
nopoly is impossible in the contestable markets if the
only reason of the evolution is that consumers stochasti-
cally select products or the market stochastically elimi-
nates unnecessary products.
4. Conclusions and Suggestions
The final evolving outcome of the contestable markets is
monopoly though this process is very long. The incom-
plete information of buyers and the incomplete informa-
tion of firms have different influence to the evolving
speed. Competition can short the evolving period. Fur-
thermore, the longer the life of a product is, the more
difficult the monopoly forms. The stochastic difference
of the original factors cannot explain the difference of the
evolving outcome. It means that the high proportion of
one firm owning today cannot ensure that the firm will
become winner tomorrow. So each firm should keep the
sense of crisis, and never release.
Those conclusions can instruct government to make
policies. A government which believes that competition
can gain maximum welfares should help new firms to
enter the markets if the life of products is very short, and
should make potential entry firms become real entry
firms. This is an effective measure to stop monopoly.
REFERENCES
[1] Y. D. Yu, Y. Y. Zhang and B. W. Zheng, “Western Eco-
nomics,” Economic Science Press, Beijing, 1997.
[2] N. D. Steven and E. B. Lawrence, “The New Palgrave
Dictionary of Economics,” Economic Science Press, Bei-
jing, Vol. 2, 1996.
[3] U. Wilensky, “NetLogo GenDrift P Global Model Center
for Connected Learning and Computer-Based Modeling,”
Northwestern University, Evanston, 2001.
[4] U. Wilensky, “NetLogo GenDrift P Local Model,” Center
for Connected Learning and Computer-Based Modeling,
Northwestern University, Evanston, 2001.