Journal of Service Science and Management, 2011, 4, 52-58
doi:10.4236/jssm.2011.41008 Published Online March 2011 (http://www.SciRP.org/journal/jssm)
Inter-Group and Intra-Group Externalities of
Two-Sided Markets in Electronic Commerce
Lixiang Li, Yueting Chai, Yi Liu
Department of Automation, Tsinghua University, Beijing, China.
Received November 11th, 2010; revised December 23rd, 2010, accepted December 24th, 2010.
ABSTRACT
The e-marketplaces, which play an important role in facilitating transa ction s and aggrega ting informatio n in electronic
commerce, show positive inter-group externalities where one group’s benefit from receiving a service depends on the
number of the intermediary services consumed by the other group, and negative intra-group externalities where the
surplus is destroyed because members within the group compete with each other. In this paper, with a different ap-
proach from the emerging two-sided markets theory and the traditional microeconomic theory, we analyze a monopo-
listic e-marketplace owned by a third-party firm by substitu ting the size of consumers with the number of intermediary
services. Moreover, we pay close attention to solving the following problems: 1) When these inter- and intra-group ex-
ternalities exist, are both the demand curves and the pricing strategies of intermediary services different from those of
traditional goods? 2) How to price intermediary services to maximize profit in the e-marketplaces? 3) How do these
network externalities affect the management strategy of platforms? Finally, we exemplify such analytical results as
pricing strategies of platforms with managerial practice of electronic commerce.
Keywords: Electronic Commerce, E-Marketplace, Two-Sided Markets, Network Externalities
1. Introduction
During the last decade, many business-to-business (B2B)
and customer-to-customer (C2C) e-marketplaces, such as
ALibaba, GlobalSources, BuyerZone, eBay, and Taobao,
continue to flourish, and are generating enormous vol-
umes of trades. However, some famous commercial
websites like Commerce One had gone into bankruptcy.
One common feature of these platforms is that, besides
trade services, they also provide such information ser-
vices as industry news and economic policies, consulting
services for management decisions, and other services. In
fact, these information services can fascinate sellers and
buyers and help the platform form a virtual community to
disseminate new message. Thereafter, this fascinates
more sellers and buyers. Furthermore, most of these
e-commerce platforms usually provide free service for
sellers or buyers. A principle problem facing platform
owners is how to price trade services to maximize their
profit in electronic commerce? How do network exter-
nalities affect the management strategy of platforms?
Which economic theory explains these economic phe-
nomena? In fact, many literatures in two-sided markets
begin to speculate these problems, but there is few sys-
tematic research results in electronic commerce.
According to the emerging theory of two-sided mar-
kets, e-marketplaces are a typical two-sided market.
Rochet and Tirole [1] strictly defined two-sided markets
as ones in which the price structure affects the economic
outcome (volume, profits, and welfare). Some markets fit
this description: B2B e-marketplaces, credit cards, media
markets, real-estate agents and shopping malls. Evans [2]
provided more cases and discussion of such two-sided
markets. In these markets, intermediaries or “platforms”
enable interactions among distinct groups of agents and
attempt to gain the two sides with rational price to each
side. This interaction exhibits inter- and intra-group ex-
ternality: 1) inter-group externalities, or indirect network
externalities, as far as the benefit enjoyed by a member
of one group increases with the number of members of
the other group; 2) intra-group externalities, or competi-
tion externalities, as far as the benefit enjoyed by a
member of a group decrease because members within the
group compete with each other. For instance, industrial
B2B e-marketplaces shares these characteristics: 1) the
more services consumed by sellers in e-marketplaces, the

Inter-Group and Intra-Group Externalities of Two-Sided Markets in Electronic Commerce 53
higher the benefits for buyers connected with platforms,
and vice versa; 2) sellers compete for the trade of buyers,
and vice versa. Therefore, business strategies of interme-
diaries are shaped by inter- and intra-group externalities.
In the presence of inter-group externalities, Caillaud and
Jullien [3] put ‘divide-and-conquer’ strategies subsidiz-
ing the participation of one side (divide) and recovering
the loss on the other side (conquer). Intra-group external-
ities, however, blur the picture.
This paper naturally relates to the very recent literature
on two-sided markets, pioneered by Rochet and Tirole
[1,4], Caillaud and Jullien [3], and Armstrong [5,6],
based on the theories of network externalities initiated by
Katz and Shapiro [7]. Rochet and Tirole [1,4] emphasize
the importance of relative price elasticity of demand on
the two sides of the platform in determining its pricing
structures with general approaches. Besides, most analy-
ses apply to specific industries: payment systems [8,9],
Internet-related intermediation [3], software and video
games [10,11], media markets [12]. Jullien [13] and
Rochet and Tirole [1] propose useful road maps to this
flourishing literature. The main emphasis of these papers
is on the effects of inter-group externalities on the design
of pricing structures, the competition between platforms,
multi-homing vs. single-homing decisions, platform own-
ership, exclusive contracts and so on. Intra-group exter-
nalities are either abstracted away or not central to the
analysis. The paper closest in spirit to mine is Yoo et al.
[14] and Belleflamme and Toulemonde [15]: they study
how the two types of network externalities jointly shape
intermediary strategies in two-sided markets. However,
all this papers essentially consume that the number of
intermediary services consumed by an agent is either one
or zero, but not directly affected by the consumer’s
budget and platforms’ prices. That is to say, these papers
don’t consider sellers’ and buyers’ decision problems to
choose the consumption quantity of services, but simply
focus on whether agents will connect with platforms. In
our model the consumption choices with budget con-
straints play an important role, because they affect the
demand and prices of services consumed by sellers and
buyers and therefore profits and social welfare. Contrib-
uting to the recent two-sided market theory, we introduce
a novel mechanism 1) that models an industrial e-mar-
ketplaces with the theory of consumer behavior and the
theory of the firm, 2) that illustrates the shape of the de-
mand curves of intermediary services in platforms joined
by two-sided network externalities, 3) that deduces the
optimal pricing strategies of intermediary services when
the impact of inter- and intra-group externalities is sig-
nificant, 4) that shows the impact of such parameters as
network externalities on optimal prices.
The rest of this paper is arranged as follows. Firstly,
we model e-marketplaces joined by inter- and intra-group
externalities in Section II. Secondly, we analyze the im-
pact of the parameters that define sellers and buyers on
price levels in Section III. Finally, section IV concludes.
2. Model
There are a monopoly platform and two types of con-
sumers: sellers and buyers (denoted 1 and 2 in turn). As
can be seen in Figure 1, an agent of one group pays at-
tention to the quantities of intermediation services con-
sumed by agents of the other group and competition
among members within the group.
For simplicity, suppose the quasi-linear utility function
of a member of sellers or buyers is determined as follows:
when the platform admits m1 sellers and m2 buyers, the
quantities of the intermediary services consumed by the
ith (i = 1, 2, , m1) group-1 agent and the jth (j = 1, 2,
, m2) group-2 agent are x1i and x2j separately, each
consumer’s problem of choosing her most preferred con-
sumption level given service prices p1 or p2 and wealth
levels w1i or w2j can be stated as the following utility
maximization problems:





1
21
111211111
1121111 1
11
11 111
2
2
..1, ,
iiii i
x
nn
ij ii
ji
ii i
Maxuxb XrXxBy
i
x
bxr xxBy
st pxywim

 




 


 (1)



2
12
2221122 22
22122 222
11
22 222
2
2
..1, ,
jjjj j
x
nn
j
ijj
ij
jj j
Maxuxb XrXxBy
j
x
bxr xxBy
stp xywjm







 


 (2)
where X1 (X2) is the aggregate demand of services to
sellers (buyers), y1i (y2j) is the budget surplus to a mem-
ber of sellers (buyers). The parameter B1 (B2) describes
the net fixed benefit, which equals information values
minus usage costs to sellers (buyers) in the platform. The
inter-group externality parameter b1 (b2) weighs the
Figure 1. Inter- and intra-group externalities of e-market-
places.

Inter-Group and Intra-Group Externalities of Two-Sided Markets in Electronic Commerce 54
benefit a member of sellers (buyers) enjoys from inter-
acting with each member of sellers (buyers) because in-
termediaries save the transaction cost in direct transac-
tion, and the intra-group externality parameter r1 (r2)
weighs the loss a member of sellers (buyers) suffers from
competing among members within the group each other.
Generally, there are b1, b2 > r1, r2 in e-marketplaces.
Using the first-order conditions of (1) and (2), we get
the inverse demand functions to the two groups
11211
pbXrXB
1
2
 (3)
22122
pbXrX B (4)
Equation (3) describes that the aggregate demand X1
decreases with the price p1 when X2 is fixed, and the
same to (4). It is consistent with the demand-price char-
acter of general goods or services. Simultaneous Equa-
tions (3) and (4) have a unique solution characterizing
the demand X1 and X2 as functions of (p1, p2):
21 12 21 12
1
12 12
21 1221 12
2
12 12
rpbprB bB
Xbb rr
bprpbB rB
Xbb rr








(5)
Expression (5) shows that the aggregate demand X1 (X2)
increases with the prices (p1, p2) of services in the pres-
ence of inter-group and intra-group externalities. That is
to say, the aggregate demand curve of services slopes
upwards because network externalities make the two
groups in the platform obtain more value from consump-
tion of intermediary services when the network scale is
more large.
From the cost side, we assume the monopoly platform
will incur the marginal costs C1 and C2 for serving sellers
and buyers respectively. Consequently, the platform’s
profit is equal to:



1112 2
pCX pCX   2
(6)
where the sunk cost is not considered implicitly.
Substituting (3) and (4) into (6), we obtain the first-
order conditions of the profit maximization problem:


11221111
21212222
2
2
XbbXrXBC
XbbXrXBC


 

  


0
0
 (7)
From (7), the optimal aggregate consumption levels to
the two groups are


 

 

where and when the following condi-
tions hold:
*
10X* 20X



 

2111222
1221 211
20
20
rCBb bCB
rCBb bCB

 






(9)
with derived from b1 > r1 and b2 > r2.

2
12 1
4bb rr
2
We assume that expression (9) is satisfied for the fol-
lowing sections and the second-order conditions of the
profit function for (X1, X2) are also satisfied as follows:

22 22
11 22
2
222 2
12 12
22
12 12
20,2 0
40
Xr Xr
bb rr
XX XX







 

 


(10)
Substituting (8) into (3) and (4), we get the optimal
prices to maximize the platform owner’s profit:

 

 




 

22
112 121212121
*
12
12 12
1212 2
2
12 12
22
212 122 112 122
*
22
12 12
2121 1
2
12 12
22
4
4
22
4
4
bbbrrC bbbrrB
pbb rr
bbrC B
bb rr
bbbrrC bbbrrB
pbb rr
bbrCB
bb rr
 






 







(11)
These outcomes can be recorded as follows:
Proposition 1: The optimal prices and the
optimal aggregate demands of intermediary services

**
12
,pp



**
12
,nn , which maximize the platform owner’s profit,
are characterized by (11) and (8).
From (11), we can get that the optimal prices


**
12
,pp
vary with different marginal service costs (C1, C2). Intui-
tively, when the marginal cost for serving a group is
higher, the price of intermediary services is expected to
be higher in order to recover the loss of the platform. But,
in e-marketplaces joined by inter- and intra-group net-
work externalities, we deduce some unusual results about
the relation between optimal prices and marginal costs
for serving the two groups.
Proposition 2: In e-marketplaces connected with two-
sided network externalities, the price level to a group is
lower than the marginal cost for serving it, whereas the
price level to the other group is higher than the marginal
cost for serving it.

2111222 *
12
12 12
1211122 *
22
12 12
2
4
2
4
rCBb bCB
Xbb rr
bbCBrC B
Xbb rr







(8)
Proof. With total optimal price level 2 , minus
the total marginal cost for serving the two groups C1 + C2,
we obtain
*
1
pp*

 
*** *
1212 111222
0ppCC rbX rbX

   (12)

Inter-Group and Intra-Group Externalities of Two-Sided Markets in Electronic Commerce 55
Expression (12) shows that not every price level to a
group is greater than the marginal cost for serving it.
Moreover, the maximal profit of the platform is not nega-
tive, namely

** ***
111 222
0pCXpCX   
(13)
Expression (13) shows that not every price level to a
group is smaller than the marginal cost for serving it.
Therefore, there is either (2
) or (
2
) within the monopoly platform joined by inter-
and intra-group network externalities.
*
11
pC, *
2
pC*
11
pC,
*
2
pC
(Q.E.D.)
According to Proposition 2, the monopoly platform
can recoup a loss or make a profit not by charging both
higher prices, which exceed marginal costs to the two
groups, but by charging a higher price to a group whereas
providing price subsidy for the other group in e-market-
places.
3. Comparative Static Analysis
From (11), we can find the relation between optimal
prices and these parameters defined in this paper.
3.1. Impact of Inter-Group Externalities on
Service Prices
Optimal prices of trade services vary with the magni-
tudes of inter-group network externalities to the two
groups. The impact of positive inter-group externalities
of both sellers and buyers on optimal service prices are
stated as follows:
Proposition 3: When the magnitude of the inter-group
network externalities to group 2 is not smaller than that
to group 1 (b2 ≥ b1), the service price to group 1
becomes higher, but the service price to group 2
becomes lower with the rise in the benefit of the in-
ter-group externalities to group 1 b1. Oppositely, if the
magnitude of the inter-group network externalities to
group 2 is not bigger than that to group 1 (b2 ≤ b1), the
service price to group 2 becomes higher, but the
service price to group 1 is lower with the rise in the
benefit of the inter-group externalities to group 2 b2.
*
1
p
*
2
p
*
2
p
*
1
p
Proof. Many composite effects exist in e-marketplaces
owing to several parameters, so the relation between the
change of the optimal service price to each group and the
relative change of inter-group network externalities (b1,
b2) is obscure. However, while the relative value between
the inter-group externality parameters to the two groups
is known, the changes of and become distinct.
The first-order derivatives of optimal prices to the two
groups, and , on inter-group externality parame-
ters (b1, b2) in expression (11) are stated in expression
(14). Since all
*
1
p*
2
p
*
1
p*
2
p
*
11
pb terms are positive when b2 ≥ b1
and all *
12
pb

 terms are negative when b2 ≤ b
1,
*
11
pb is positive when b2 ≥ b1 and *
1
pb

2
is nega-
tive when b2 ≤ b1.
 











2
*21 211211
1
2
2
112 12
222 *
1121221222
2
22
12 12
2
*2121 1211
1
2
2
212 12
222
112121222
12
4
4
42
0, 0
4
4
4
42
bb bbrrCB
p
bbb rr
rbbrrb bC Bp
b
bb rr
brrb bbCB
p
bbb rr
rrr bbbbCB
bb








 








 






 



*
2
2
21
12
0, 0
4
p
b
rr






(14)
(Q.E.D.)
*
22
pb

 is the same to *
11
, whereas pb *
21
pb


is the same to *
12
pb

; thus we only provide scenarios
for *
11
pb and *
12
pb


. If the benefit of inter-group
network externalities to a group is bigger, the utility of
intermediary services to that group is bigger. Moreover,
the service price to that group can be higher (*
11
pb

 >
0, *
22
pb

 > 0). However, inconceivably, the optimal
service price to a group is lower while the benefit of in-
ter-group network externalities to the other group is big-
ger (*
1
pb
2

 < 0, *
21
pb

 < 0).
3.2. Impact of Intra-Group Externalities on
Service Prices
The relations between changes in optimal service prices
and changes in different sizes of negative intra-group
network externalities are showed as follows:
Proposition 4: If the benefit of the inter-group net-
work externalities to group 2 is bigger than that to group
1 (b2 > b1), the price to group 1 (group 2) will be lower
(higher) along with a rise in the benefits of the in-
tra-group externalities (r1, r2). Conversely, if the size of
the inter-group network externalities to group 2 is less
than that to group 1 (b2 < b1), the price to group 1 (group
2) will be higher (lower) along with a rise in the benefits
of the intra-group externalities (r1, r2).
Proof. The first derivatives of intra-group externality
parameters (r1, r2) on optimal prices of each group,
and in expression (11) are provided in expression
(15). If b2 > b1 and (9) is satisfied, both
*
1
p
*
2
p
*
11
pr and
*
1
r
2
p

 are negative, but both *
22
and pr *
21
pr


are positive. Oppositely, if b2 < b1 and (9) is satisfied,
both *
11
pr

 and *
12
pr

 are positive, but both

Inter-Group and Intra-Group Externalities of Two-Sided Markets in Electronic Commerce 56
*
22
pr and *
21
prare negative.

 






 
22
*12 2111222
1
2
2
112 12
22
*211221211
2
2
2
212 12
*
1121 221211
1
2
2
212 12
*221211122
2
1
2
4
2
4
22
4
22
bb rCBbbCB
p
rbb rr
bb rCBbbCB
p
rbb rr
rb brCBbbCB
p
rbb rr
rbbrCBb bCB
p
r














 
















2
2
2
12 12
4bb rr





(15)
(Q.E.D.)
We maybe hope that due to the bigger negative exter-
nalities among members within each group and the ulti-
mate negative impact on the benefits of group 1 and
group 2, the service prices to the two groups will at all
time be lower. However, if b2 > b1 (the attachment of
group 1 is more worthy to group 2 than the attachment of
group 2 to group 1), the service price to group 1 will be
lower and the service price to group 2 will be higher
when r1 or r2 becomes bigger. Similarly, if b2 < b1 (the
attachment of group 1 is less worthy to group 2 than the
attachment of group 2 to group 1), the service price to
group 1 will be higher and the service price to group 2
will be lower when r1 or r2 becomes bigger.
3.3. Impact of Information Benefits on Service
Prices
The optimal price to each group is different with varia-
tion of the information benefits (B1, B2). Instinctively,
when the benefit of information services is greater, the
expectation utility of the platform and the resulting price
will also be higher. However, in some circumstances, we
can find that the lower prices vary with the higher infor-
mation benefits. The influence of information benefits to
each group over the optimal prices is showed as follows.
Proposition 5: 1) If information benefit to group 1, B1
(group 2, B2) becomes bigger, the service price charged
to group 1 (group 2, ), will be higher.
*
1
p*
2
p
2) If the size of network externalities to group 2 (group
1) is bigger than that to group 1 (group 2), b2 > b1 (b1 >
b2), the service price charged to group 1, (group 2,
), will be higher when the information benefit to
group 2, B2 (group 1, B1), becomes bigger,. Conversely,
if the size of network externalities to group 2 (group 1) is
smaller than that to group 1 (group 2), b2 < b1 (b1 < b2),
the service price charged to group 1, (group 2, ),
will be lower when the information benefit to group 2, B2
(group 1, B1), becomes bigger in the same way.
*
1
p
*
2
p
*
1
p*
2
p
Through the first derivatives of the information bene-
fits (b1, b2) on optimal service prices and in
expression (11), we can easily prove the above proposi-
tion. If the information benefit to group 2 becomes big-
ger, the utility of the platform to group 2 will become
bigger, and the service price to group 2 will also become
higher (
*
1
p*
2
p
*
22
pB

 > 0). Therefore, we hope that the ser-
vice price to group 1 will also become higher simply
because the bigger value of intermediary services to
group 2 increases the benefit for group 1 with the indirect
externality. However, inconceivably, the information
benefits for group 2 is higher, the optimal price for group
1 is lower when b2 < b1.
All results in the above propositions are summarized
in Table 1.
4. Managerial Applications
In this paper, we have explored a general model of neu-
tral e-marketplaces where the benefit to one group from
inter-group externalities based upon the amount of ser-
vices consumed by the other group and the loss to each
group from negative intra-group externalities based upon
competition among members within it. We conclude
some useful results stated in the above propositions,
which can help managers to determine the optimal pric-
ing strategies of platforms connected with two-sided
network externalities. Moreover, the significance of the
relative size between the positive inter-group external-
ities to the two groups is emphasized in these proposi-
tions. In practice, information about the perceptual char-
acter of these platforms may provide managers some
Table 1. Summary of propositions.
Parameter
(Increase)
*
1
p
*
2
p
Assumptions
b1 + – when b1 ≤ b2
b2 – + when b1 ≥ b2
– + when b1 < b2
r1
+ – when b1 > b2
– + when b1 < b2
r2
+ – when b1 > b2
+ – when b1 < b2
B1
+ + when b1 > b2
+ + when b1 < b2
B2
– + when b1 > b2

Inter-Group and Intra-Group Externalities of Two-Sided Markets in Electronic Commerce 57
useful clues in recognizing which inter-group externality
is bigger.
Generally, a forward-auction e-commerce platform, i.e.
seller’s market, may have higher inter-group network
externalities to sellers compared to a reverse-auction
e-commerce platform, i.e. buyer’s market. According to
the above results, the service price to buyers may be very
lower or zero in supplier-favored marketplaces con-
trasted with that in buyer-favored marketplaces. For ex-
ample, such neutral platforms as alibaba.com, made-in-
china.com, globalsources.com and taobao.com run buyer’s
market using posted prices or forward auctions. In com-
parison, the price charged to sellers will be lower in re-
verse-auction compared to that in forward-auction. For
instance, FreeMarkets runs a buyer-favored B2B e-mar-
ketplace using reverse-auctions, most of the market-
making costs are charged to buyers [16]. Most of plat-
form’s benefits come from sellers with charging mem-
bership fees or commissions, whereas these trade ser-
vices are free for buyers. However, 315.com, which is a
neutral B2B block trading e-marketplace in the oil indus-
try, chemical industry, steel industry, plastics industry,
and rubber industry, charges commissions according to
the trading volume between buyers and sellers in the
platforms.
5. Conclusions
In summary, our theoretical analyses enrich the emerging
theory of two-sided markets with the introduction of
negative intra-group externalities within members of the
two groups. Based on our model, we derive the pricing
strategy that an independent intermediary can use to
maximize its profit. Based on our assumptions, the ag-
gregation demand levels of participants in electronic
markets as well as the prices charged from sellers and
buyers in the electronic marketplace are dependent on the
defined parameters of this paper.
For simplicity, we assume that equal prices are
charged from all sellers and all buyers, whereas the cost
of information services are all fixed costs that do not
affect the pricing decisions of e-marketplaces as long as
the total profit is positive. Also, we consider a monopo-
listic e-marketplace with a single-period model. Next, we
hope to extend this analysis to a dynamic model with
multiple periods. In future research, we will also analyze
a duopoly model, which incorporates platform competi-
tion between intermediaries in electronic commerce.
6. Acknowledgements
This research is supported by the Key Projects in the
National Science & Technology Pillar Program in the
Eleventh Five-year Plan Period (Project Numbers: #
2009BAH48B03).
REFERENCES
[1] J. Rochet and J. Tirole, “Two-Sided Markets: A Progress
Report,” Rand Journal of Economics, Vol. 37, No. 3,
2006, pp. 645-667.
doi:10.1111/j.1756-2171.2006.tb00036.x
[2] D. S. Evans, “The Antitrust Economics of Multi-Sided
Platform Markets,” Yale Journal on Regulation, Vol. 20,
No. 2, 2003, pp. 325-381.
[3] B. Caillaud and B. Jullien, “Chicken & Egg: Competition
among Intermediation Service Providers,” Rand Journal
of Economics, Vol. 34, No. 2, 2003, pp. 309-328.
doi:10.2307/1593720
[4] J. Rochet and J. Tirole, “Platform Competition in
Two-Sided Market,” Journal of the European Economic
Association, Vol. 1, No. 4, 2003, pp. 990-1009.
doi:10.1162/154247603322493212
[5] M. Armstrong, “Competition in Two-Sided Markets,”
Rand Journal of Economics, Vol. 37, No. 3, 2006, pp.
668-691. doi:10.1111/j.1756-2171.2006.tb00037.x
[6] M. Armstrong and J. Wrigh, “Two-Sided Markets, Com-
petitive Bottlenecks and Exclusive Contracts,” Economic
Theory, Vol. 32, 2007, pp. 353-380.
doi:10.1007/s00199-006-0114-6
[7] M. L. Katz and C. Shapiro, “Network Externalities, Com-
petition, and Compatibility,” American Economic Review,
Vol. 75, No. 3, 1985, pp. 424-440.
[8] J. Wrigh, “Optimal Card Payment Systems,” European
Economic Review, Vol. 47, 2003, pp. 587-612.
doi:10.1016/S0014-2921(02)00305-7
[9] J. Wrigh, “The Determinants of Optimal Interchange Fees
in Payment Systems,” Journal of Industrial Economics,
Vol. 55, 2004, pp. 1-26.
doi:10.1111/j.0022-1821.2004.00214.x
[10] G. G. Parker and M. W. V. Alstyne, “Two-Sided Network
Effects: A Theory of Information Product Design,” Man-
agement Science, Vol. 51, No. 10, October 2005, pp.
1494-1504.
doi:10.1287/mnsc.1050.0400
[11] A. Hagiu, “Pricing and Commitment by Two-Sided Plat-
forms,” Rand Journal of Economics, Vol. 37, No. 3, 2006,
pp. 720-737. doi:10.1111/j.1756-2171.2006.tb00039.x
[12] M. Reisinger, L. Ressner and R. Schmidtke, “Two-Sided
Markets with Pecuniary and Participation Externalities,”
Journal of Industrial Economics, Vol. LVII, No. 1, March
2009, pp. 1-26.
[13] B. Jullien, “Two-Sided Markets and Electronic Interme-
diaries,” CESifo Economic Studies, Vol. 51, No. 2-3,
2005, pp. 233-260. doi:10.1093/cesifo/51.2-3.233
[14] B. Yoo, V. Choudhary and T. Mukhopadhyay, “A Model
of Neutral B2B Intermediaries,” Journal of Management
Information Systems, Vol. 19, No. 3, 2002, pp. 1-12.

Inter-Group and Intra-Group Externalities of Two-Sided Markets in Electronic Commerce 58
[15] P. Belleflamme and E. Toulemonde. “Negative Intra-
Group Externalities in Two-Sided Markets,” Interna-
tional Economic Review, Vol. 50, No. 1, February 2009,
pp. 245-272. doi:10.1111/j.1468-2354.2008.00529.x
[16] V. K. Rangan, “Free Markets Online,” Harvard Case
Study 9-598-109, Harvard University, Cambridge, Feb-
ruary 1999.