Sie sind auf Seite 1von 35

This article was downloaded by: [Ume University Library]

On: 01 April 2015, At: 23:00


Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Management Information


Systems
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/mmis20

The Impact of Cloud Computing: Should


the IT Department Be Organized as a
Cost Center or a Profit Center?
a

Vidyanand Choudhary & Joseph Vithayathil


a

Merage School of Business, University of California, Irvine

Department of Management, Information Systems, and


Entrepreneurship (MISE) at the College of Business, Washington
State University
Published online: 08 Dec 2014.

To cite this article: Vidyanand Choudhary & Joseph Vithayathil (2013) The Impact of Cloud
Computing: Should the IT Department Be Organized as a Cost Center or a Profit Center?, Journal of
Management Information Systems, 30:2, 67-100
To link to this article: http://dx.doi.org/10.2753/MIS0742-1222300203

PLEASE SCROLL DOWN FOR ARTICLE


Taylor & Francis makes every effort to ensure the accuracy of all the information (the
Content) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions

The Impact of Cloud Computing: Should


the IT Department Be Organized as a
Cost Center or a Profit Center?

Downloaded by [Ume University Library] at 23:01 01 April 2015

Vidyanand Choudhary and Joseph Vithayathil


Vidyanand (VC) Choudhary is an associate professor of information systems at the
Merage School of Business, University of California, Irvine. He received his Ph.D.
in management and an M.S. in industrial engineering from Purdue University, and a
B.Tech. in computer science and engineering from the Indian Institute of Technology, Delhi. His research interests are broadly in the area of economics of information
systemsindustrial organization and game theory applied in the context of information
systems. Specific research topics include governance of IT, information goods pricing
and product design, price discrimination and versioning strategy, and cloud computing.
His research has been published in a number of journals, including Management Science, Information Systems Research, and Journal of Management Information Systems.
He is a cofounder of the Workshop on Theory in Economics of Information Systems
(TEIS), which is in its third year. His research papers have won best paper awards at
the Twenty-Second International Conference on Information Systems (ICIS) and the
Tenth Workshop on Information Technology and Systems (WITS).
Joseph Vithayathil is an assistant professor in the Department of Management,
Information Systems, and Entrepreneurship (MISE) at the College of Business, Washington State University. He received his Ph.D. in management from the University
of California, Irvine, an MBA from Harvard University, an M.S. in systems science
and mathematics from Washington University in St. Louis, and a B.Tech. in electrical
engineering from the Indian Institute of Technology, Madras. His research interests
are broadly in the area of economics of information systems and the application of
economic theory, agency theory and the related issues of moral hazard, adverse selection, information asymmetry, and incentives in the context of information systems.
Specific research topics include governance of IT, IT-enabled corporate governance,
and the impact of cloud computing and social media on organizations. He has significant professional and entrepreneurial experience in the IT industry and in executive
management.
Abstract: How does the adoption of cloud computing by a firm affect the organizational structure of its information technology (IT) department? To analyze this question, we consider an IT department that procures IT services from a cloud computing
vendor and enhances these services for consuming units within the firm. Our model
incorporates the competitive environment faced by the cloud vendor, which affects the
price of the cloud vendor. We find that when the cloud vendor faces intense competition, the cost-center organizational model is preferred over the profit-center model.
Infrastructure services such as basic storage, email, and raw computing face intense
competition, and our results suggest that such services be offered as a free corporate
resource under the cost-center organizational structure. When the cloud vendor has
Journal of Management Information Systems / Fall 2013, Vol. 30, No. 2, pp. 67100.
2013 M.E. Sharpe, Inc. All rights reserved. Permissions: www.copyright.com
ISSN 07421222 (print) / ISSN 1557928X (online)
DOI: 10.2753/MIS0742-1222300203

68

Choudhary and Vithayathil

pricing power, a profit-center organizational structure is likely to be preferred. Our


results suggest that highly differentiated services such as cloud-based enterprise-wide
enterprise resource planning or business intelligence be offered under the profit-center
structure. Finally, the profit-center structure provides greater internal quality enhancement to cloud-based IT services than the cost center.

Downloaded by [Ume University Library] at 23:01 01 April 2015

Key words and phrases: chargeback, cloud, cloud computing, cost center, IaaS, IT
governance, PaaS, profit center, SaaS, supply chain.

Cloud computing is a disruptive technology that provides an alternative or is an


adjunct to in-house information technology (IT) services. For example, Dropbox
and Amazon.com are disrupting the market for traditional storage providers while
Carbonite and Box are disrupting the market for backup services. Forrester Research
predicted that the public cloud computing market will grow from $26billion in 2012
to $160 billion in 2020 [9]. This extraordinary growth forecast represents a compounded annual growth rate (CAGR) of 25percent. International Data Corporation
(IDC)[23] forecasted that public IT cloud services would grow from $21.5 billion in
2010 to $72.9billion in 2015, which represents a CAGR of 27.6percent. IDC further
noted that the cloud services growth rate surpasses the overall IT services global
growth rate of 6.7percent. There are cloud services available for all major enterprise
systems. For example, cloud-based enterprise resource planning (ERP) services are
available from IBM, Microsoft, Oracle, and SAP. Similarly, customer relationship
management (CRM) services are available on the cloud from Salesforce.com, Sage,
and NetSuite.
The availability of on-demand IT services and the pay-as-you-go pricing model
makes it attractive for firms to source many IT services from the cloud. The role of
the IT department within firms that source most of their IT services from the cloud is
likely to be very different from the current model where the IT department provides
captive services. In the former, the IT department is likely to oversee vendor selection, determine and enforce corporate standards for security and privacy, and provide
enhancement to the incoming cloud services such as application and data integration.
For example, the Avis Budget Group and Dell were using on-premises CRM solutions
that were executed and maintained by their IT departments. Now they use Salesforce.
com as a cloud-based CRM vendor. The IT department at the Avis Budget Group
provides limited integration with legacy systems.
This changing role of the IT department should affect the governance structure
of the IT department. Prior information systems (IS) literature and evidence from
industry show that two forms of IT governance are prevalent: the profit center and the
cost center. In this paper we develop a stylized model to examine the impact of cloud
computing and software as a service (SaaS) on the choice of organizational form. We
address the following research questions:

The Impact of Cloud Computing

69

RQ1: When is it better for a firm that uses cloud-based services to organize the
IT department as a cost center versus a profit center?
RQ2: What are the factors that drive the choice of organizational structure in
this context?

Downloaded by [Ume University Library] at 23:01 01 April 2015

RQ3: How much quality enhancement is provided by the IT department under


the different organizational structures?
The National Institute of Standards and Technologies (NIST) [28] has characterized
cloud computing into three building layers: infrastructure as a service (IaaS), platform
as a service (PaaS), and Saas. While there are other definitions proposed, the NIST
definition is widely used. Amazon.coms EC2 (Elastic Compute Cloud) offering is an
example of an IaaS, and Microsofts Azure is an example of PaaS, while NetSuite and
Salesforce.com are examples of SaaS. Armbrust et al.[2, 3] and Mell and Grance[28]
provide the structure and characteristics of cloud computing; they note that cloud
computing consists of software applications offered as a service as well as computing
and storage infrastructure services. A distinctive benefit to users of cloud services is
the pay-as-you-use attribute. A second important financial feature enabled by cloud
computing, as noted by Armbrust et al.[2], is the ability for the firm to eliminate the
fixed cost of a captive data center and only incur usage-based pricing for IT services
from the external cloud vendor. Choudhary[8] shows that the faster dissemination of
new features under cloud-based SaaS can lead to higher quality of IT services.
Gurbaxani and Kemerer [21, 22] and others note that IT departments are commonly
organized as either a profit center or a cost center. Under the profit-center model, IT
services come with a pricing schedule that internal users pay. In contrast, the cost center
offers internal IT services at no charge to the internal users. Gurbaxani and Kemerer
raise the issue of information asymmetry between the IT department, internal IT consuming functional units, and the firm. Jensen and Meckling[24] argue that managers
within a firm possess private knowledge that leads to information asymmetry. Such
information asymmetry can make it difficult for the firm to determine the optimal price
for IT services. Our description of a cost center fits with their expense center, where
IT services are provided free of charge to internal consumers. The prior literature on
the organizational structure for the IT department is in the context of captive IT services. The question regarding the impact of the IT department procuring cloud-based
IT services on the choice of organizational structure has not been addressed.
Internal pricing of IT services is analyzed in the prior literature by Dewan[12],
Dewan and Mendelson [13], Mendelson [29], Pick and Whinston [30], Wang and
Barron[35], and Whang[36]. Wang and Barron[35] also analyze the profit center
and the cost center. They consider computer usage in their analysis, as well as the IT
departments objective to maximize budget allocation to IT. They conclude that the
profit center is unable to overcome the rent that the IT department can extract from
information asymmetry; therefore, the cost center is always the preferred organizational
structure. In contrast, we demonstrate that both organizational forms are viable and use
an alternative approach to that of Wang and Barron by evaluating benefits to the firm

Downloaded by [Ume University Library] at 23:01 01 April 2015

70

Choudhary and Vithayathil

Figure 1. Cloud Computing Model and Supply Chain Model

from the use of IT. Under this approach, the IT department maximizes profits when
operated as a profit center and determines internal price and quality of IT services;
whereas when operated as a cost center, the firm determines the quality of IT services.
Jensen and Meckling[24] and Wang and Barron[35] consider the objective of the
cost center to be cost minimization. A cost-minimization strategy may not account
for the benefits that accrue from higher-quality services. We employ a model of the
cost center where the objective of the firm is to maximize the benefits from IT. Jensen
and Meckling[24] and Wang and Barron[35] consider information asymmetry on
the cost of IT, to which we add information asymmetry of the demand profile for IT
services by the internal consumers.
The use of cloud-based services as an input to the IT department may be viewed as
a supply chain. The IT department adds value to the externally procured cloud-based
IT services and supplies such enhanced services to the internal consuming units.
There are some similarities and there are areas of differences between the traditional
supply chain model and the model presented in this paper. Figure1 illustrates the
two models.
Traditional supply chain analysis considers coordination problems that also exist in
the cloud computing model presented in this paper. Such problems consist of, first,
double marginalization, which can be addressed by vertical integration [33] or with
two-part pricing[10]. Our cloud computing model exhibits double marginalization
under the profit center because the IT department marks up the price from the incoming
cloud vendor. However the cost center addresses double marginalization by offering
IT services at no charge. Second, apart from the conflicting incentives of the various
entities, information asymmetry is another issue in supply chain coordination, which
may cause the bullwhip effect[26]. Demirkan et al.[11] show the benefits of supply
chain coordination among SaaS vendors, and Sen et al.[31] show that information

Downloaded by [Ume University Library] at 23:01 01 April 2015

The Impact of Cloud Computing

71

sharing in IT services outsourcing is welfare enhancing. In our model, the cost center is characterized by information asymmetry, whereas in the profit center, the IT
department possesses cost and demand information and is responsible for pricing,
thus alleviating the information problem. Third, supply chain analysis also considers
inventory optimization, which has solutions such as revenue-sharing contracts[6, 7],
information sharing[19, 27], and quantity discounts[15]. In contrast, this paper is
centered on IT services without any physical goods. Fourth, quality is mostly absent
from this literature and the primary considerations in the prior literature are price and
quantity. However, there are limited exceptions such as Economides[16], who examines supplier quality, where vertical integration can result in quality improvement.
Our cloud computing model conceptually resembles a supply chain structure. However, there are important areas of differences. First, we include the firm as the governing body over two of the supply chain entities: the IT department and the consuming
unit. Furthermore, the firm has the authority to impose the operating organizational
structure, which is either the cost center or the profit center. The firm optimizes over
the two supply chain entities under its purview, whereas the traditional supply chain
models consider each entity as optimizing its own benefit separately, or the models
optimize overall supply chain benefit. Second, the effect of double marginalization
under the profit-center model is different from traditional supply chain setting because
the profits from the IT department remain within the firm. Third, the cloud computing
model in this paper considers IT services and therefore does not encounter supply chain
issues for physical goods such as stock-outs or excess inventory. Fourth, we include
the quality of IT services supplied as an important characteristic of IT services under
the cloud computing model. Moreover, the well-used news vendor model employed in
the supply chain literature does not apply here because demand is modeled as deterministic and as private information of the consuming unit. Furthermore, traditional
supply chains have characteristics such as long-term relationships, negotiated pricing,
and revenue sharing, which are also present in outsourcing contracts as described later.
In contrast, cloud computing services are defined in this paper by the characteristics
of arms-length, take-it-or-leave-it standard pricing that is based on usage.
Most IT services have marginal cost associated with the services. We draw a distinction between such IT services and information goods. Digital information goods
such as music, books, and software have zero or almost zero marginal cost, whereas
typical IT services provided within a firm have nonzero marginal cost. Networking
and storage services have a cost per unit such a megabits per second or megabytes,
respectively, whereas other communication services have subscription costs. Allen[1]
supports the notion of marginal cost of IT services, while Dewan[12] analyzes pricing and suggests that computer services be priced at marginal cost. In this paper, the
marginal cost has two parts: (1)a base marginal cost and (2)marginal cost of quality
that is increasing in quality. The IT department enhances the quality of IT services,
where quality can be defined as the service-level responsiveness or a low error rate in
transaction processing. Quality improvement under such a definition is likely to have
a fixed cost component as well as a quality-related marginal cost component. As a
simple example, when the number of internal IT consumers grow, the support required

Downloaded by [Ume University Library] at 23:01 01 April 2015

72

Choudhary and Vithayathil

for maintaining the service-level quality may require a new fixed cost infrastructure as
well as an increased, variable labor cost. In contrast to a fully captive IT department,
under our cloud computing setting in this paper, it is the cloud vendor who provides
the initial quality of IT service. This initial quality may be enhanced by the IT department. Our marginal cost function captures the notion that modest enhancements can
be accomplished within reasonable costs, but major enhancements are likely to be
prohibitively expensive.
Traditional IT outsourcing has characteristics that are materially different from that
of cloud computing. Outsourcing possesses one or more of the following characteristics: (1)medium- to long-term contracts; (2)a single customer per contract; (3)a
contract structure that may be one or more of time and materials, performance-based,
or incentive-based contracts; (4)customized covenants; (5)negotiated pricing; (6)a
single vendor; (7)a custom product; and (8)is relationship specific. Furthermore,
the IT system or service development is fully customized; therefore, the constraints
on quality improvement under outsourcing are similar to that faced under captive IT
service or system development.
In contrast, cloud computing vendors offer a standardized product to all customers with take-it-or-leave-it pricing that has a metered or a pay-for-usage structure.
Standardized cloud-based IT services have a mass-market orientation, thus limiting
the customization or customer-specific quality enhancements that are possible under
traditional outsourcing. Hence, the amount of customization to improve the elements of
IT quality may be more limited under cloud computing when compared to traditional
IT outsourcing or fully internal IT development. Therefore, internal quality enhancements to cloud-supplied services need to be incremental. We capture this dimension
of cloud computing by normalizing the marginal cost of quality improvement by
the base level of quality supplied by the vendor. Under this setting, higher internal
IT quality levels will require a sufficiently high base level of quality from the cloud
vendor because the cost of a substantial quality improvement is prohibitive. Because
of these significant differences between traditional outsourcing and cloud computing,
this paper focuses on cloud-sourced IT services.
The IT department is modeled as executing its task within each organizational
structure without shirking. We also assume that the internal IT-consuming functional
units need the IT department to add value to cloud-based IT services before the final
services are delivered to them. The value added to IT services procured from the cloud
vendor may take the form of quality improvement, monitoring, and ensuring vendor
performance, selection, and administration of vendors, as well as adding software
and other application functions on top of the procured services. For example, if the
procured services are SaaS, the IT department may add value by administering the
services for internal clients or it may add user-specific applications on top of the SaaS
offering by the use of application programming interfaces (APIs).
The effect of cloud computing on the choice of organizational model has not been
studied previously. We examine whether a cost-center or profit-center organizational
structure is better for the firm when the IT department obtains services from a cloud
vendor at a certain price and quality. We find that the pricing power of the cloud ven-

Downloaded by [Ume University Library] at 23:01 01 April 2015

The Impact of Cloud Computing

73

dor affects the choice of organizational structure for the IT department. In addition,
we find that the choice of organizational structure affects the level of internal quality
enhancement of the IT services that are procured from a cloud vendor.
We find that the profit center is the preferred organizational structure for the IT department when the cloud vendor is characterized by high pricing power as this structure
generates greater benefits from IT for the firm than the cost-center structure. When
the cloud computing services are infrastructure oriented or commodity oriented, it is
beneficial for the firm to organize the IT department as a cost center. Under the NIST
description of cloud computing, IaaS provides the least value-added services within
the gamut of cloud computing offerings. According to Knipp[25], IaaS has become a
commodity, and IaaS services are subject to intense competition. Cloud services that
constitute IaaS offering are services such as basic storage or raw computing resources.
We find that such services should be offered as a free corporate resource under a costcenter organizational structure.
In contrast, when the IT department deploys higher value-added or differentiated
cloud-based services, the results indicate that a profit-center organizational structure
is preferred. In the NIST cloud computing model, PaaS and SaaS constitute higher
value-added services such as ERP, CRM, and business intelligence(BI).
We also find that the profit center offers higher internal quality enhancement when
the cloud vendors pricing power is high, otherwise the cost center provides higherquality enhancement. We find that quality enhancement and consumption of IT services
decrease with increasing pricing power of the cloud vendor for the profit center as
well as the cost center. Under the profit-center organizational structure, increasing
pricing power of the cloud vendor can lead to a reduction in the price charged to the
internal consuming units.
The sections that follow include the following: a description of the conceptual model
followed by the development of the analytical model, which examines whether the costcenter or profit-center organizational structure is preferred. The analysis is followed
by a discussion of the results, managerial implications, and concluding remarks.

Conceptual Model
This paper focuses on the organizational structure of the internal IT department when
the IT department (1)procures cloud-based IT services and (2)adds value to these
procured services for delivery to internal consumers. We examine which organizational
form for the IT department, the cost center or the profit center, maximizes benefit to
the firm from the use of IT. We examine the impact of cloud-specific factors such as
the competitive environment faced by the cloud vendor and the price and quality of
services provided by the cloud vendor.

IT Organizational Structure
There are four parties in our model: (1)the IT department that provides IT services to
the internal consumers within the firm; (2)the internal consumers or consuming units

74

Downloaded by [Ume University Library] at 23:01 01 April 2015

(a)

Choudhary and Vithayathil

(b)

Figure 2. (a) Cost-Center Model and (b) Profit-Center Model

of such IT services, aggregated into one unit for exposition and tractability; (3)the
cloud vendor that the IT department procures services from under a given price and
quality; and (4)the firm. The firm chooses from two IT organizational structures: a
cost center or a profit center. Under the cost-center organizational structure, the firm
determines the quality of services that the IT department will offer. The IT department follows this quality mandate determined by the firm. Under the profit-center
organizational structure, the IT department determines the quality and price for the
internal IT services.
Conceptual models of the cost-center and profit-center organizational structures are
illustrated in Figure2. As noted earlier, the IT department provides its services at zero
price to the internal consumers under the cost-center organizational structure shown in
Figure2a. The firm determines the quality enhancement level and generates benefits
from IT through the internal consumer, which generates a surplus. The cost of IT services is borne by the firm. The internal consumer has a private demand profile for IT
services, and the amount of IT services consumed is based on this demand profile.
Figure 2b illustrates the profit-center organizational structure. The IT department
controls the decision-making process and therefore determines the price and internal
quality enhancement for IT services. The internal consuming entities pay the IT department for its services. The quantity of IT services consumed by the internal entity is
determined by its demand profile for IT services. The IT department bears the cost of

Downloaded by [Ume University Library] at 23:01 01 April 2015

The Impact of Cloud Computing

75

providing IT services and receives the revenue from the internal consumers. Finally,
the total benefit to the firm from the use of IT is the sum of the IT department profits
and the surplus generated by the internal consumer.
Information asymmetry is present in our model with regard to the cost of IT and
internal demand for IT services. Consistent with Jensen and Meckling[24] and Wang
and Barron[35], the cost information is private to the IT department. The internal ITconsuming unit knows its demand and the willingness to pay (WTP) for IT services,
and this is private information in the cost-center organizational structure. However,
when the IT department is organized as a profit center, the buyer and seller relationship enables the IT department to learn the aggregate demand profile of the internal
consuming entity. The firm can estimate the cost of IT and the consuming entitys
demand for IT services.

Model
The analytical model we develop is stylized and includes simplifying assumptions
that are necessary for tractability and clarity of exposition of the key results. We
now describe the details of the model about the IT department. The IT department
procures IT services from the cloud vendor, enhances the cloud vendors incoming
IT quality, and supplies the enhanced IT services to the internal consuming units. We
refer to the internal consuming units as the internal customer. The quantity of IT
services provided to the internal customer is denoted as x, the quality of IT services
provided to the internal customer is denoted as q. The IT department procures IT
services from the cloud vendor at a quality qc and at a price pc charged by the cloud
vendor. The IT department enhances qc by an amount q and supplies the resulting
quality q=qC+ q.
We assume that demand, cost, and profit functions are continuous, and the payoff
functions are twice differentiable and concave. The demand for IT services by the
consuming unit is decreasing in price charged for the service and increasing in quality
supplied. The consuming unit is assumed to have exogenously determined its demand
for IT services for any given price and quality. The base level of quality necessary to
operate the firm is zero, and the base level of demand for IT services is at zero quality
and price. The demand profile of the internal consumer is such that at a sufficiently
high price, there will be no consumption of IT services.
When the IT department operates under the cost-center organizational structure, the
information asymmetry with regard to cost and demand renders the firm incapable of
determining an optimal quality level based on full information. Therefore, under such
information asymmetry, the firm estimates its optimal quality level, and we use q*CC to
denote the estimated quality. Firms use various methods such as benchmarking[20]
to estimate quality. We capture the effect of information asymmetry in our model by
introducing an error term in the firms estimate of demand and cost functions.
When the IT department operates under the profit-center organizational structure, the
IT department sets the price and quality level for internal IT services to maximize its
profits IT. As noted in the introduction, the IT department is informed of the demand

76

Choudhary and Vithayathil

function of the internal consumer as well as its own costs. The price and quality set
by the IT department operating as a profit center are p*PC and q*PC. The IT department
sets the price p*PC and quality q*PC in order to maximize its profits IT.
Next, we describe the details of the model as they pertain to the internal consumer of
IT services. The internal demand for IT services x is characterized by a linear demand
function. The quantity of IT services consumed is a function of the price and quality
of the internal IT services as follows:

Downloaded by [Ume University Library] at 23:01 01 April 2015

x = k a p + b q.

The intercept for demand is denoted by a parameter k. The parameters and


denote the marginal effect of price and quality of IT services on the internal consumer,
respectively. This linear demand specification is consistent with the prior literature
(e.g., [4, 5, 14, 17, 18, 32, 34]). In order to capture the effect of information asymmetry, we assume that the firm is not fully informed about the value of the demand
parameter . The firm believes that this parameter is drawn from a uniform distribution
U(,+), where is a noise parameter for the estimate.
The WTP for IT services by the consumer is specified by the inverse demand function
(k+bqx)/a. Thus, we can derive the surplus generated by the internal consumer
by integrating the inverse demand:
x

k + q x
dx x p.

SI =

Now we describe the model details as they pertain to the cloud vendor. The cloud
vendor serves the broader market for cloud services and will have many customers.
The price and quality of its cloud-based IT services such as IaaS, PaaS, and SaaS are
determined by the market forces facing this vendor. Such market forces include competition, demand, and cost. The IT department of the focal firm in this paper procures
x quantity of cloud-based IT services at a quality level qC from the cloud vendor.
The price charged by the cloud vendor is denoted by pC, which depends on the quality of their service and various other factors such as competition, efficiencies from
multitenancy, and scale and scope. We treat these competitive factors as exogenous,
and they are captured in our model by a parameter 0. Larger values of result
in the increased pricing power of the cloud vendor. The price charged by the cloud
vendor is pC=c+qqC. Pure competition in the cloud is represented by =0, which
yields pC=c.
The IT department procures IT services at quality qC and price pC charged by the
cloud vendor. The IT department adds value to these cloud services by (1)monitoring
the vendors and (2)endogenously optimizing and improving IT service quality by an
amount denoted as q to an enhanced level q=qC+Dq; (3)delivering IT services at
this quality q to the internal consuming units; (4)incurring the marginal cost of quality improvement bDq/qC, which results in total marginal cost MCIT=pC+b(Dq/
qC); and (5) incurring a convex fixed cost of quality improvement expressed as
FCITQ=a(q2qC2). The ratio Dq/qC ensures that large quality enhancements to the
cloud vendors service quality are substantially more expensive to the IT department.

The Impact of Cloud Computing

77

In order to capture the effect of information asymmetry of IT cost, we assume that the
firm is not fully informed about the value of the cost parameter a. The firm believes
that this parameter is drawn from a uniform distribution U(ae,a+e), where e is a
noise parameter for the estimate.

Assumptions

Downloaded by [Ume University Library] at 23:01 01 April 2015

We now state the assumptions in our model.


Assumption 1 (Regularity Conditions): Standard mathematical regularity conditions are assumed. The demand, cost, and profit functions are continuous, and
the payoff functions are twice differentiable. Demand is increasing in quality
and decreasing in price: D/q=b>0, D/q=a<0. The cost of quality is
increasing in quality: dFCITQ/dq=2aq>0, d2FCITQ/dq2=2a>0. Therefore,
the cost function is convex. The IT profit and firm value are evaluated net of IT
cost and are concave functions. The inverse demand function p=(k+bqx)/a
exists and p/q=b/a>0.
Assumption 2 (Intercept): For each quality level, there is a threshold of maximum
consumption denoted by x=k+bq when the price is zero. Similarly, at a sufficiently high price denoted by p=(k+bq)/a, there will be no consumption
of IT services. The demand for IT services at the base quality level of q=0 is k
units.
Table 1 provides a list of parameters and variables with a brief description. The
sections that follow analyze the effect of organizing the IT department as either a cost
center or a profit center. The two organizational structures are individually analyzed
in each of the subsections. The individual analysis is followed by a comparison of
the results from the analysis of the two organizational structures. The proofs for the
lemmas and propositions that follow are presented in the Appendix.

IT Organized as a Cost Center


When IT is organized as a cost center, the internal consuming unit is provided with IT
services at no charge. The total surplus available for the consuming business unit is

k + q*CC

SICC

) k + (q*CC ) z

dz.

The firm determines the optimal level of quality enhancement by maximizing its
expected benefits from IT, Maxq{E[B{FCC]}, where B{FCC=SICCxMCITFCITQ. The
sequence of events is shown in Table2.
Lemma 1 (Interior SolutionIT Organized as a Cost Center Under Cloud Computing): Under a cost-center organizational structure with cloud computing, the
firm sets quality offered by the IT department to

78

Choudhary and Vithayathil

Table 1. List of Model Variables, Functions, and Parameters

Downloaded by [Ume University Library] at 23:01 01 April 2015

Symbol
pC
qC
c + qC
q
e
q
pIT
BF
SI
Dq
x
a
a
b
k
b

Description
Price charged by the cloud vendor for IT services
IT service quality provided by the cloud vendor
Cloud vendors marginal cost of IT services
Competitive environment of the cloud vendor
Maximum error in firms estimation of b, a
Quality of IT services provided by the IT department
IT department profit under the profit-center organizational structure
Benefit to the firm from IT
Surplus generated by the internal IT consumer
Internal quality enhancement by the IT department
IT services quantity consumed by the firm
Fixed cost of quality parameter
Marginal effect of internal price on IT service consumption
Marginal effect of quality on IT service consumption
Intercept of the consumer demand function
Marginal cost of quality enhancement parameter

q*CC =

3 b (qC k ) + qC k ( c + qC )

6 ( aqC + b ) qC 3 +
2

)) ,

and the benefit to the firm is

( (

k + q*CC
*CC
*CC
BF*CC =
qC k + 2 ( b c qC ) + q 2bq
2qC

a q*CC

qC2 .

Lemma 1 reports the optimal quality that the firm would set under the cost-center
organizational structure. This is the quality level that the IT department would be
directed to offer to the internal consuming units. It can be established that the benefit
to the firm is decreasing in the marginal effect of price and also decreasing in the
marginal cost of IT quality enhancement. Similarly, an increase in marginal cost of
quality enhancement will result in decreased benefits to the firm.
The parameter e captures the degree of information asymmetry. When =0, the
firm has perfect information. We find that the optimal quality set by the firm under
information asymmetry is greater than the quality that would be set when the firm has
perfect information. This occurs because the firm knows that any misestimation of the
quality parameter leads to reduction of benefits to the firm, and the firm compensates
by increasing internal quality enhancement.
Proposition 1 (Comparative Statics for the Cost Center with Cloud Computing):
Increasing pricing power of the cloud vendor results in a reduction in quality

The Impact of Cloud Computing

79

Table 2. Sequence of Events When the IT Department Is Organized as a Cost Center


Step
1
2

Downloaded by [Ume University Library] at 23:01 01 April 2015

Action
The firm estimates the optimal quality of IT services q*CC and informs the IT
department of the quality of the IT services that will be offered.
The IT department procures IT services from the cloud vendor at price
pC and quality qC and enhances the quality by Dq to deliver the internal
quality level q*CC.
The internal IT consumer determines the quantity of services consumed
and generates its surplus.
The firm realizes net benefits from IT which is the surplus of the internal IT
consumer less the total cost of providing the services.

enhancement by the IT department and lower consumption of IT services by the


consuming unit: (Dq*CC)/q<0, x*CC/q<0.
Proposition 1 reports the effect of competition faced by the cloud vendor on the
internal quality enhancement of IT services by the IT department. At first glance, the
relationship between the level of competition in the cloud and the level of quality
enhancement provided by the internal IT department is not obvious. If the cloud vendor
enjoys greater pricing power, then it charges higher prices. The firm could direct the IT
department to increase the level of internal quality enhancement to justify the higher
cost and provide increased surplus for the consuming unit. However, Proposition1
shows that the firm will instruct the IT department to reduce quality enhancement
when the cloud vendor has increased pricing power.
To understand this result, we begin by noting that under the cost-center organizational
structure, IT services are offered as a free corporate resource to the consuming unit.
Hence, there will be overconsumption of such IT services. When the cloud vendors
prices are increased, the firm finds such overconsumption to be even more detrimental
to the benefits derived from IT. Under a cost-center structure, the firm can limit consumption by reducing internal quality enhancement. Consequently, it chooses to reduce
internal quality enhancement when the cloud vendor has increased pricing power.
Differentiated and specialized cloud computing services such as cloud-based BI
software services exemplified by IBMs Cognos Express are likely to have increased
pricing power. Such services may require ongoing technical support or customization
by the IT department. Our results indicate that the internal consumers of such services are likely to get minimal support from the IT department under the cost-center
model.

IT Organized as a Profit Center


Organized as a profit center, the IT department has the ability to charge for internal
IT services, bear the cost for IT services, and make a corresponding profit. The profit
generated by the IT department constitutes part of the benefit to the firm from the

80

Choudhary and Vithayathil

Table 3. Sequence of Events When the IT Department Is Organized as a Profit


Center
Step

Action

The IT department provides IT services at a price and quality denoted


as {p*PC,q*PC} based on knowledge of the internal consumers demand
function.
x*PC units of IT services are demanded as a function of the price and
quality offered by the IT department, and a surplus to the consuming unit
of SI*PC is generated net of price for IT services.
A profit of pIT is generated by the IT department.
The firm realizes benefits from IT that is the sum of the net surplus of the
internal IT consumer and the profit of the IT department.

Downloaded by [Ume University Library] at 23:01 01 April 2015

3
4

deployment and use of IT. The internal consumer obtains value in the form of surplus
from IT services and pays the IT department for such services. The IT department
is able to learn the demand profile of its internal consumer and procures IT services
from the cloud vendor at a base level of quality qC and at a price pC. The IT department determines the optimal internal price and quality enhancement by maximizing
IT profit: Maxp,q{pIT}, where pIT=xpPCMCITFCITQ. The sequence of events is
shown in Table3.
Lemma 2 (Interior Solution: IT Organized as a Profit Center Under Cloud Computing): Under a profit-center organizational structure with cloud computing, the
price and quality of IT services that maximizes benefit to the firm are
p* PC =

)) (

)( (

2 a ( qC ) k + b + c + ( qC ) b ( qC ) b k + ( qC ) ( qC ) c + ( qC )
2

a ( qC ) b ( qC )
2

q* PC =

qC (qC b ) ( k + ( b c ) qC )

4 a ( qC ) b ( qC )
2

))

)2

and the benefit to the firm is

k p* PC + q* PC * PC
BF* PC =
( 2b ) + qC k + 2 (b c qC ) + p*PC
q
2

a q* PC

))

qC2 .

Lemma 2 reports the results from the solution to the first-order conditions when
the IT department maximizes its profits. The benefit to the firm from the use of IT is
the sum of the profits generated by the IT department and the net surplus generated
by the internal consumer. It can be seen that the benefit to the firm is decreasing in
the marginal cost of IT quality enhancement and the fixed cost of quality. The two

The Impact of Cloud Computing

81

organizational structures for the IT department are compared to understand which


organizational structure is favorable in terms of benefits to the firm from the use of
IT. The next section performs this analysis.

Downloaded by [Ume University Library] at 23:01 01 April 2015

Proposition 2 (Comparative Statics for the Profit Center with Cloud Computing):
(i)Increasing pricing power of the cloud vendor results in reduced internal IT
prices, provided the cloud vendors quality is sufficiently high: p*PC/q < 0,
qC>bab/(b22aa), and increased internal IT prices otherwise. (ii)Increasing
pricing power of the cloud vendor results in decreasing quality enhancement by
the IT department and decreasing consumption of IT services by the consuming
unit (Dq*PC)/q<0, x*PC/q<0.
Proposition 2 reports the effect of increasing pricing power of the cloud vendor
on internal prices for IT services, internal quality enhancement of IT services, and
consumption of IT services under the profit center. As reported in the first part of the
proposition, the effect of the cloud vendors pricing power on internal IT services is
nonintuitive because it can increase or decrease internal price.
The intuition for this result is as follows: (1)when the cloud vendor supplies IT
services of sufficiently low quality and the cloud vendor increases price, the IT department compensates by correspondingly increasing price. This in turn enables the IT
department to restrain the more costly internal consumption of IT services by the
consuming unit, whereas (2)when the cloud vendor supplies IT services of sufficiently
high quality and the cloud vendor raises price, the IT department decreases internal
quality enhancement and passes the cost savings from the lower-quality enhancement
to the consuming unit in order to ensure profitable consumption of IT services.
The comparative statics results in the second part of Proposition2 with regard to
quality enhancement and consumption are similar to that of the cost-center results.
However, under the profit-center structure, the IT department has the ability to recover
the increased payments to the cloud vendor through internal pricing. This leads to
lower demand, making it difficult for the IT department to sustain the fixed cost of
internal quality enhancement. Thus, higher payments to the cloud vendor under the
profit-center structure cause the IT department to reduce quality enhancement when the
external cloud vendors price increases. It follows that reduction in quality enhancement will reduce consumption.

Comparison of the Profit-Center and Cost-Center Organizational


Models
The results from Lemmas 1 and 2 can be used to generate a comparison between the
two organizational structures. The comparison leads to the following propositions:
Proposition 3 (Competition in the Cloud and Choice of Organizational Structure):
(a)The profit-center organizational structure is preferred if and only if the cloud
vendors pricing power q in the interior is (i)1<<2 when Z4>0 and (ii)<2
or >1 when Z4<0. (b)The cost-center organizational structure is preferred if

82

Choudhary and Vithayathil

and only if the cloud vendors pricing power in the interior is (i)<1 or >2
when Z4>0 and (ii)2<<Z1 when Z4<0, where 1, 2, Z4 are as defined in
the Appendix, and the feasibility conditions for interior solutions are satisfied.

Downloaded by [Ume University Library] at 23:01 01 April 2015

Proposition 3 demonstrates that cloud services such as SaaS affects the choice of IT
organizational structure. Since SaaS and other cloud services are a relatively recent
developments, prior literature such as Gurbaxani and Kemerer[21, 22], Jensen and
Meckling[24], and Wang and Barron[35] did not consider a cloud-based environment. Consequently, our result in Proposition 3 contributes to the understanding of the
impact of cloud-based services such as SaaS on the choice of organizational structure
for the IT department.
Observation 1 (Competition in the Cloud and Choice of Organizational Structure): The profit-center organizational structure is preferred when >1 and the
cost-center organizational structure is preferred when <1.
Observation 1 is derived using numerical methods and combines the results of
Proposition3 with the feasibility conditions for interior solutions reported in Lemmas1 and2. We examine the interior region of the profit center and cost center across
the parameter set and find that the profit center is preferred when the cloud vendors
pricing power is sufficiently high. Using SaaS as an example of a cloud-based IT
service, the intuition is that strong pricing power enables the SaaS vendor to charge
higher prices. Because of higher pricing, under the cost-center organizational structure,
the IT department reduces quality enhancement. In contrast, under the profit-center
structure, the IT department is able to offer greater quality enhancement because of
its ability to charge the internal users for such value addition. Whereas the profit center can observe changes in demand due to changes in price and quality, the decision
maker in the cost center can observe the range for optimal quality but cannot pinpoint
the precise level of quality to be offered. Figure3 illustrates the effect of the cloud
vendors pricing power from Proposition3 and also illustrates the effect of the fixed
cost of quality enhancement on the choice of organizational structure.
Proposition 4 (Competition in the Cloud and Internal IT Service Quality Enhancement): Under cloud computing, the profit center provides greater quality enhancement of IT services than the cost center if and only if the cloud vendors pricing
power is sufficiently strong:

2
2

3bk ( b qC ) 4 a ( qC )
> qC (b c + ( k / )) +

6 aqC ( b + qC ) + ( b qC ) 3b 2 qC

/ ( q )2 .
C

Proposition 4 demonstrates that the organizational structure of the IT department


under cloud computing affects the quality of internal IT services. The cloud vendors
pricing power, which is partly determined by the competitive environment of the
cloud vendor, affects the magnitude and direction of the effect of internal IT services
quality. The prior literature has not addressed the question of cloud vendor quality or
the internal quality enhancement of IT services as reported in Proposition4. Intuition
may suggest that a profit center is likely to provide higher-quality enhancement than

Downloaded by [Ume University Library] at 23:01 01 April 2015

The Impact of Cloud Computing

83

Figure 3. Cloud Computing Vendors Pricing Power, Fixed Cost of Quality Enhancement,
and Choice of IT Organizational Structure

a cost center in the context of internally generated IT services. Proposition4 builds


on such intuition by adding the dimension of cloud computing and the dimension of
quality enhancement of IT services supplied by the cloud vendor. A cloud vendor with
low pricing power will supply low-priced IT services, and the cost-center structure is
able to spread the fixed cost of internal quality improvement over a larger consumption base. Therefore, an IT department organized as a cost center can provide greater
quality enhancement than the profit center. Cloud-based storage services such as
Carbonite are characterized by numerous vendors and intense competition. In such
instances, the cost center is likely to provide greater quality enhancement. In contrast, when the cloud vendor has strong pricing power, the IT department organized
as a profit center offers greater quality enhancement because the profit-center model
allows the IT department to charge for its higher-quality services. The high quality of
IT increases internal consumption of IT services even though the consuming unit has
to pay for such services. This is likely to be the case for higher-level services such as
SaaS because such services are well differentiated with lower levels of competition
and increased pricing power when compared to IaaS. Figure4 illustrates the effect
of the cloud vendors competitive environment on quality enhancement by the IT
department from Proposition4.
Observation 2 (Impact of Cloud Vendors Quality on the Preferred Organizational
Structure): The crossover threshold q1 defined in Proposition3 (i)decreases with
increasing cloud vendor quality qC when the cloud vendors quality is sufficiently
low qC<0.53 and (ii)increases with increasing cloud vendor quality qC when the
cloud vendors quality is sufficiently high qC >0.71 when the parameter values
are {k=10, a=3.5, c=0.5, a=3, b=10, b=1, e=0.1].

Downloaded by [Ume University Library] at 23:01 01 April 2015

84

Choudhary and Vithayathil

Figure 4. Cloud Computing Vendors Competitive Environment and Quality Enhancement by


the IT Department

Observation 2 demonstrates that the quality of IT service provided by the cloud


computing vendor affects the threshold of the cloud vendors pricing power q1, which
determines the choice of organizational structure. This observation is illustrated in
Figure5. The direction of this effect is such that increasing quality of IT services by
the cloud vendor favors the profit center when the incoming cloud vendors quality is
sufficiently low, as seen on the left side of Figure5. To understand this observation,
note that when the cloud vendors quality is low, then internal IT quality enhancement
plays a greater role in generating benefits. When the cloud vendors pricing power is
sufficiently high, then the profit center provides greater quality enhancement than the
cost center. Therefore, when the cloud vendors quality and pricing power are low, then
the cost center is preferred. However, when the cloud vendors pricing power increases,
the profit center provides greater quality enhancement and is preferred.
We now explain the right-hand side of Figure 5. When the incoming cloud vendors
quality is sufficiently high, the internal quality enhancement plays less of a role in
generating benefits to the firm from the use of IT, and the cost center encourages
consumption with zero pricing, which leads to increased surplus by the consuming
unit and, therefore, greater benefits to the firm.

Discussion
This paper examines whether a firms reliance on cloud-based IT services affects
the choice of organizational structure for the IT department. Our analysis considers
factors that affect the choice of organizational structure for the IT department such

Downloaded by [Ume University Library] at 23:01 01 April 2015

The Impact of Cloud Computing

85

Figure 5. Preferred Organizational Structure Crossover Threshold q1 and Cloud Vendor IT


Quality

as the cloud vendors pricing, quality, and competitive environment. In our model,
the IT department procures services from a cloud vendor and enhances the quality
of cloud-sourced IT services prior to offering the services to internal consumers.
When a firms IT department uses cloud computing services that are infrastructure
or commodity oriented, such as email, document collaboration, basic storage, or
raw computing, our results indicate that it is beneficial for the firm to organize the
IT department as a cost center. In contrast, when the IT department deploys higher
value-added or differentiated cloud-based services such as ERP, CRM, or BI, the
results indicate that a profit-center organizational structure would be the preferred
structure for the IT department.
Mapping our results to the NIST description of cloud computing, IaaS provides the
least value-added services from the gamut of cloud computing offerings. Cloud services
that fit the IaaS model are services such as basic storage services or raw computing
resources. According to Knipp[25], IaaS has become a commodity and is subject to
intense competition. Table4 provides examples of both commodity services that are
IaaS offerings, differentiated high value-added services that fit the SaaS model, or
new and emerging services that are built on top of the SaaS model.
Consider the specific examples of cloud vendors from Table4 such as Gmail from
Google, which is a commodity service. Many organizations and institutions are
adopting cloud services for basic IT services such as email. Cadillac Fairview, one
of the largest North American commercial real estate companies, adopted Google
for cloud-based email services for all its employees. Similarly, the Roche Group, a
global pharmaceutical and health-care leader, moved from a captive infrastructure to
Googles cloud-based email, document storage, and document sharing for all their

86

Choudhary and Vithayathil

Table 4. Examples of Cloud-Based IT Services

Downloaded by [Ume University Library] at 23:01 01 April 2015

Highly competitive,
commodity services
E-mail
Gmail
Yahoo! Mail
Hotmail
Raw storage
Carbonite
Dropbox
Google Drive
Raw computing
Amazon EC2
SmartCloud
Rackspace Cloud

Differentiated, higher-priced
cloud services
Enterprise CRM/ERP
Unlimited CRM
Oracle CRM
RightNow CRM
NetSuite
Microsoft Dynamics
Business intelligence in the cloud
Cognos Express
Tibco Spotfire
Clarity Business Intelligence
MicroStrategy Cloud
Plex Online

Cloud
vendor
Google
Yahoo!
Microsoft
Carbonite
Dropbox
Google
Amazon.com
IBM
Rackspace

Cloud
vendor
Salesforce.com
Oracle
RightNow
NetSuite
Microsoft
IBM
Tibco Software
Cloud9 Software
Micro Strategy
Plex Systems

employees. When the IT department sources commodity services such as email from
Google, basic storage and backup services from Dropbox or Carbonite, or raw computing services through Rackspace or IBM SmartCloud, our findings suggest that the
cost-center organizational structure would be the preferred organizational structure
for the IT department.
The apex of the value chain in the NIST model constitute SaaS offerings. Such higher
value-added, differentiated services that fit the SaaS model from Table4 are Oracle
CRM, NetSuite ERP, BI services from Tibco such as Spotfire, and MicroStrategy
Cloud. Other institutions such as Gartner add another layer above the NIST SaaS
model to capture emerging cloud-based services such as BI. Cloud vendors of such
services are likely to enjoy higher pricing power and lower direct competition. Our
results suggest that firms would benefit from organizing the IT departments deploying
these services as a profit center.

Downloaded by [Ume University Library] at 23:01 01 April 2015

The Impact of Cloud Computing

87

Vendors of SaaS services such as enterprise CRM and BI, with the characteristics
of differentiated products and high value addition, are likely to enjoy greater pricing
power than cloud vendors offering commodity services such as storage, backup, and
email. The profit-center structure is preferred when the former services are deployed
because we find that the use of IT under the profit center generates greater benefits to
the firm than the cost center when the cloud vendor has high pricing power.
We model the quality of the internal IT services as an important factor that affects
the benefits that the firm derives from the use of IT. In the context of internally generated or captive IT services, prior literature has noted that the profit center is likely to
offer higher quality. This paper builds on the prior literature by analyzing the impact
of the rapidly growing cloud-based sourcing of IT services. In contrast, under cloud
computing, the result from prior literature holds only when the cloud vendors pricing power is high. Otherwise the cost center organizational structure provides higher
quality enhancement than the profit center.
Many cloud vendors, including Salesforce.com, offer APIs to enable in-house
customization and quality enhancement. However, cloud vendors with proprietary
technology may limit the API set, thereby increasing the cost of quality enhancement.
Alternatively, cloud vendors with emerging technology such as advanced business
analytics may have insufficient APIs, resulting in an increased cost of IT quality
enhancement. When internal quality enhancement is desirable, the client firm should
consider the profit-center organizational structure if the cloud vendor has strong pricing power.
In our model, IT services are abstracted to have one set of characteristics. In practice,
though, IT services have varying characteristics. Our results allow for a hybrid organizational model where some services are offered at no cost to the user under a cost-center
structure and others are offered under a profit-center structure. For example, Table4
shows commodity services that are characterized by intense competition and the low
pricing power of the cloud vendor, and high value-added, differentiated services that
are characterized by a high pricing power. When a firm offers a menu of services that
cover both types of services, our results would indicate that the commodity services
be offered as a free corporate resource, whereas the value-added services be offered
under a profit-center model. Specifically, when a firm procures email services and
CRM from cloud vendors, our results suggest that the cloud-based email service be
offered under the cost-center model and the CRM service be offered under the profitcenter model. As an example, the University of California, Irvine offers email and
other services at no charge, but there is a charge for other IT services such as software
and support for backup and recovery.
The use of cloud computing by the IT department also introduces some characteristics that are similar to that of a supply chain. However, there are certain important
differences. In our setting, the firm is making a decision on organizational structure
instead of making decisions on inventory and related issues. The effect of double
marginalization under the profit-center model is different from traditional supply
chain models because the IT department and the consuming unit are both part of the
firm. The profits from the IT department remain within the firm unlike an external

88

Choudhary and Vithayathil

vendor in a supply chain. This paper considers IT services only and therefore we do
not consider supply chain issues such as physical goods and inventory.

Downloaded by [Ume University Library] at 23:01 01 April 2015

Conclusion
We employed a stylized model to analyze the impact of cloud computing on the
choice of organizational structure for the IT department. While the model makes
simplifying assumptions, it produces a variety of results that add to the literature and
have managerial implications. Clearly, any abstraction that is necessary for an analytical model is limited in its ability to capture the complexity of IT governance. Future
work can relax some of the assumptions. For example, we assumed that all services
are intermediated by the IT department. However, future work could examine the case
where some services are disintermediated such that the IT consuming units procure
services directly from cloud vendors. Future work could also include a hybrid cloud
structure, where some services are procured from a public cloud and others are served
from a private cloud. Large companies with multiple users would be able to generate
efficiencies comparable to that of a public cloud because they have sufficient diversity of internal consuming units to benefit from multitenancy and shared resources.
As a result, efficiency arguments for using a public cloud become less compelling,
and other factors such as control, security, and proprietary features may be the more
dominant considerations. Commodity services are likely to be procured from a public
cloud resulting in a hybrid structure. Future research may need to consider such a
hybrid structure.
The adoption of cloud computing is likely to have other consequences for the IT
department. As more services are procured from cloud vendors, the need for functions within the IT department that serve to administer, monitor, and maintain the
IT infrastructure will be considerably diminished or even eliminated. Under cloud
computing, the IT department may need to increase the focus on adding value to the
incoming cloud-based services. The mission of the IT department would transition
to determining the means by which the IT personnel can ensure that enhancements
to cloud-based services fit the current and future needs of the consuming units. The
impact of such changes could be the subject of future research.

References
1. Allen, B. Make information services pay its way. Harvard Business Review, 65, 1
(JanuaryFebruary, 1987), 5763.
2. Armbrust, M.; Fox, A.; Griffith, R.; Joseph, A.D.; Katz, R.H.; Konwinski, A.; Lee,G.;
Patterson,D.A.; Rabkin,A.; Stoica, I.; and Zaharia, M. Above the clouds: A Berkeley view of
cloud computing. Department of Electrical Engineering and Computer Science, University of
California, Berkeley, February 2009.
3. Armbrust, M.; Fox, A.; Rean, G.; Joseph, A.D.; Katz, R.; Konwinski, A.; Lee,G.; Patterson,D.; Rabkin, A.; Ion, S.; and Zaharia, M. A view of cloud computing. Communications
of the ACM, 53, 4 (2010), 5058.
4. Banker, R.D.; Khosla, I.; and Sinha, K.K. Quality and competition. Management Science,
44, 9 (1998), 11791192.

Downloaded by [Ume University Library] at 23:01 01 April 2015

The Impact of Cloud Computing

89

5. Bhargava, H.K., and Choudhary, V. Information goods and vertical differentiation. Journal
of Management Information Systems, 18, 2 (Fall 2001), 89106.
6. Cachon, G.P. Supply chain coordination with contracts. In A.G. de Kok, and S.C. Graves
(eds.), Handbooks in Operations Research and Management Science, vol. 11. Amsterdam:
North-Holland, 2003, pp. 229340.
7. Cachon, G.P., and Lariviere, M.A. Supply chain coordination with revenue-sharing contracts: Strengths and limitations. Management Science, 51, 1 (2005), 3044.
8. Choudhary, V. Comparison of software quality under perpetual licensing and software as
a service. Journal of Management Information Systems, 24, 2 (Fall 2007), 141165.
9. Columbus, L. Roundup of cloud forecasts and market estimates. CloudTech, 2012 (available at www.cloudcomputing-news.net/blog-hub/2012/jan/18/roundup-of-cloud-computingforecasts-and-market-estimates-2012).
10. Corbett, C.J.; Zhou, D.; and Tang, C.S. Designing supply contracts: Contract type and
information asymmetry. Management Science, 50, 4 (2004), 550559.
11. Demirkan, H.; Cheng, H.K.; and Bandyopadhyay, S. Coordination strategies in an SaaS
supply chain. Journal of Management Information Systems, 26, 4 (Spring 2010), 119143.
12. Dewan, S. Pricing computer services under alternative control structures: Tradeoffs and
trends. Information Systems Research, 7, 3 (1996), 301307.
13. Dewan, S., and Mendelson, H. User delay costs and internal pricing for a service facility.
Management Science, 36, 12 (1990), 15021517.
14. Dixit, A. A model of duopoly suggesting a theory of entry barriers. Bell Journal of Economics, 10, 1 (Spring 1979), 2032.
15. Dolan, R.J. Quantity discounts: Managerial issues and research opportunities. Marketing
Science, 6, 1 (Winter 1987), 122.
16. Economides, N. Quality choice and vertical integration. International Journal of Industrial
Organization, 17, 6 (1999), 903914.
17. Gal-Or, E. Information sharing in oligopoly. Econometrica, 53, 2 (1985), 329343.
18. Gal-Or, E., and Ghose, A. The economic incentives for sharing security information.
Information Systems Research, 16, 2 (2005), 186208.
19. Gavirneni, S.; Kapuscinski, R.; and Tayur, S. Value of information in capacitated supply
chains. Management Science, 45, 1 (1999), 1624.
20. Gordon, S. Benchmarking the information systems function. Working Paper Series 94-08,
Center for Information Management Studies (CIMS), Babson College, November 1994.
21. Gurbaxani, V., and Kemerer, C.F. An agency theory view of the management of end-user
computing. Working Paper, Sloan School of Management, MIT, Cambridge, 1989.
22. Gurbaxani, V., and Kemerer, C.F. An agent-theoretic perspective on the management of
information systems. In Proceedings of the 22nd Annual Hawaii International Conference on
System Sciences. Kailua-Kona, HI: IEEE Computer Society Press, 1989, pp.141150.
23. IDC cloud research. International Data Corporation, Framingham, MA (available at www
.idc.com/prodserv/idc_cloud.jsp).
24. Jensen, M.C., and Meckling, W.H. Divisional performance measurement. In M.C. Jensen
(ed.), Foundations of Organizational Strategy. Cambridge: Harvard University Press, 1998,
pp.345361.
25. Knipp, E. Dont bring a differentiated knife to a commodity gun fight. Gartner, Stamford,
CT, March 6, 2012 (available at http://blogs.gartner.com/eric-knipp/2012/03/06/dont-bring-adifferentiated-knife-to-a-commodity-gun-fight).
26. Lee, H.L.; Padmanabhan, V.; and Whang, S. Information distortion in a supply chain:
The bullwhip effect. Management Science, 43, 4 (1997), 546558.
27. Lee, H.L.; So, K.C.; and Tang, C.S. 2000. The value of information sharing in a two-level
supply chain. Management Science, 46, 5 (2000), 626643.
28. Mell, P., and Grance, T. The NIST definition of cloud computing. Special Publication800-145, National Institute of Standards and Technology, U.S. Department of Commerce,
January 2011.
29. Mendelson, H. Pricing computer services: Queueing effects. Communications of the
ACM, 28, 3 (1985), 312321.
30. Pick, R.A., and Whinston, A.B. A computer charging mechanism for revealing user
preferences within a large organization. Journal of Management Information Systems, 6, 1
(Summer 1989), 87100.

90

Choudhary and Vithayathil

Downloaded by [Ume University Library] at 23:01 01 April 2015

31. Sen, S.; Raghu, T.S.; and Vinze, A. Demand information sharing in heterogeneous IT
service environments. Journal of Management Information Systems, 26, 4 (Spring 2010),
287316.
32. Singh, N., and Vives, X. Price and quantity competition in a differentiated duopoly. Rand
Journal of Economics, 15, 4 (Winter 1984), 546554.
33. Spengler, J.J. Vertical integration and antitrust policy. Journal of Political Economy, 58,
4 (1950), 347352.
34. Vives, X. Duopoly information equilibrium: Cournot and Bertrand. Journal of Economic
Theory, 34, 1 (1984), 7194.
35. Wang, E.T.G., and Barron, T. Controlling information system departments in the presence
of cost information asymmetry. Information Systems Research, 6, 1 (1995), 2450.
36. Whang, S. Alternative mechanisms of allocating computer resources under queuing delays.
Information Systems Research, 1, 1 (1990), 7188.

The Impact of Cloud Computing

91

Appendix
Proof of Lemma 1

Downloaded by [Ume University Library] at 23:01 01 April 2015

The firm is not perfectly informed of the demand parameter and the cost parameter
a. The firm estimates these parameters as b {, a[. These estimates are assumed to have
an uniform error distribution U(be,b+e), U(ae, a+e). Marginal cost and fixed
cost of quality enhancement are

MCIT = pC + b (Dq/qC)

(A1)

FCITQ = a[ (q2qC2).

(A2)

The total surplus available for the consuming business unit when the firm sets IT
quality to be q*CC is obtained from integration of the inverse demand function as
follows:

k + q*CC

SICC

) k + (q*CC ) z

( ( ))

dz = k + q*CC

(A3)

/ 2.

0

The estimated benefit to the firm from IT is the surplus in Equation(A3) minus the
cost of IT from Equations(A1) and (A2), as follows:

B FCC = SICC x MCIT FCITQ , xCC = k + q*CC


k + ( q )) ( k + ( q + q )) ( bq + q
(
=

*CC

qC

(c + qC ))

(A4)

a q ( q + 2qC ) ,

where Dq = (q*CC qC). The firms problem is to maximize the expected benefit from IT
expressed as E[BFCC]=E[SICC(xMCIT+FCITQ)]. The expected benefit is computed
by integrating over the uniform probability density functions for b {,a[ and we have:
a + +
1 1
E BFCC = BCC
d da
f
2
2
a

( )

k ( q + qC )
k2
2
+ k + ( q + qC ) b ( c + qC )
bk kqC ) + a ( qC ) (A5)
(

2
qC

)(

( q + q )2
C
+
q 32 + 2 6 ( b + aqC ) .
6qC C

( (

92

Choudhary and Vithayathil

The problem statement and the first-order condition are as follows:


a + +
1 1
Max BCC
d da ,
f
q
2
2
a

( )

2
2

k ( q + qC ) 3 +
bk
+
2 a ( q + qC )
3
qC

+ (b c )

2b ( q + qC )
qC

Downloaded by [Ume University Library] at 23:01 01 April 2015

(A6)

qC = 0.


The optimal cost-center quality enhancement to offer is the solution to the first-order
condition in Equation (A6) for (q), as follows:
q

*CC

))
(3 + )

qC 3 ( k c ) + qC 2 + 3 ( ) 6 a 3b ( k + qC )
6 ( aqC + b ) qC

(A7)


We also have q*CC = (Dq*CC + qC), which when substituted into Equation (A7) generates the optimal quality under the cost center as determined by the firm:
q*CC =

3 b (qc k ) + qC ( c + qC ) k

6 ( aqC + b ) qC 32 + 2

)) .

(A8)


Substitute Equation (A8) and optimal quality enhancement from Equation(A7) into
the firm benefit expression(6), into the functional unit surplus expression(A3), and
into the demand function x=(k+bq*CC). Simplifying and rearranging terms generates
the additional results:

(x )
=

*CC 2

BF*CC


q*CC
x*CC c + b
+ qC q*CC

qC

2
( qC ) ,

q*CC = q*CC qC

(x )
=

(A9)

*CC 2

SI*CC
x*CC =

(
(3

6ak qC + 3b ( k + qC ) qC k 2 + 32 ( c + qC )
6 ( aqC + b ) qC

+ 2

).


Concavity condition for an interior solution: The second derivative of the objective
function(A6) is

6 ( aqC + b ) qC 32 + 2

3qC

).

(A10)


Concavity requires that the determinant (A10) is negative. This in turn requires that
the numerator (6a(aqC+bb)qC(3b2qC+e2))>0. This expression is the denominator
of q*CC and q*CC, which is therefore positive from the concavity requirement.

The Impact of Cloud Computing

93

Positive quality and quality enhancement: Since the denominator of q*CC is positive from the concavity requirement, the numerator of (A7) is required to be positive,
which generates the requirement b>bka/qC(k+a(bcqqc)), qC<(k+a(bc))/
aq, and c<b+(k/a). For quality enhancement to be positive, we set q*CCqC to be
positive, which yields the following:
b<

qC 3 ( k c ) + qC 3 ( ) 6 a + 2
3 ( k + qC )

Downloaded by [Ume University Library] at 23:01 01 April 2015

Positive quantity: Setting x


requirement
<

*CC

in Equation (A9) to be positive results in the

k 2 qC

3 2 akqC + b ( k + qC ) qC ( c + qC )
k>

)) .

2 qC ( c + qC b )
2 aqC + b

))

and
b < (c + qqC).
Comparative statics with marginal cost of quality enhancement and fixed cost of
quality: BFCC/b = (k + bqCC)((Dq)/qC) and BFCC/a = ((k + b(qCC))2/2a2) are
always negative.
This completes the proof.

Proof of Proposition 1
The effect on quality enhancement by the IT department from of an increase in the
cloud vendors pricing power is obtained by taking the derivative of Equation(A7)
with respect to as follows:

q*CC

)=

3 ( qC )

6 ( aqC + b ) qC 32 + 2

(A11)

From Lemma 1, the denominator of (A11) is positive from concavity requirements


for an interior solution. The numerator is positive. Therefore, the derivative(A11) is
negative and the effect of an increase in the cloud vendors pricing power is to reduce
quality enhancement by the IT department.
The effect on the consumption of IT services is obtained by taking the derivative of
x*CC from Equation(A9) with respect to as follows:
32 ( qC )
x*CC
=
.

6 ( aqC + b ) qC 32 + 2
2

(A12)

From Lemma 1, the denominator of (A12) is positive from concavity requirements


for an interior solution. Hence, the derivative (A12) is negative.
This completes the proof.

94

Choudhary and Vithayathil

Proof of Lemma 2
The IT department sets optimal price and quality, where its profit is given by
IT = x ( p MCIT ( q )) FCITQ , x = k p + q.

(A13)

Substitute the cost of IT from Equations (A1) and (A2) into (A13). The IT departments maximization problem and first-order conditions are as follows:

Downloaded by [Ume University Library] at 23:01 01 April 2015

)(

k p + ( q + qC ) bq + qC ( c p + qC )
Max a q ( q + 2qC )
p,q
qC

bq
k + c 2 p + q +
+ ( + ) qC = 0,
qC

) ,

(A14)

b k p + 2q )
( 2 a + ) qC = 0.
( p (b + c )) 2aq ( q

C

Solving the first-order conditions for optimal price and quality generates the following solution:

q* PC
=

qC b ( k + c ) + qC k + ( ( b + c ) + b ) + ( 4 a + ( )) qC
4 aqC2 ( b qC )

))

(A15)

p* PC =

2 aqC2 k + ( c + qC b ) + (qC b ) bk qC ( c + qC b )
4aqC2 ( b qC )

).

(A16)


Substituting Equations (A15) and (A16) into the expressions for IT surplus and
consuming unit surplus and quantity from the demand function, and simplifying, we
obtain the following results:
q* PC =

*ITPC =

qC ( b qC ) ( k + ( b c ) qC )

b2 2 2bqC + 2 4 a qC2

2 k + 4 a ( qC ) + 2 k ( b c qC )
a ( qC )

( 2b c q q ) ( c q + q )

C
C
C
C

4 a ( qC ) ( b qC )
2

(b qC ) 4a (qC )
2

,
(A17)

2 a 2 ( qC ) k + ( b c qC )
4

* PC
SIT
=

2 a ( qC ) ( k + ( b c ) qC )
2

* PC

4 a ( qC ) ( b qC )
2

The firm benefit is the sum of IT profits and business unit surplus, pIT*PC+SI*PC, and
after rearranging the terms, is expressed as follows:

The Impact of Cloud Computing

BF* PC =

4 2
2
2

16 a ( qC ) + ( b qC ) ( k + ( 2b c ) qC ( + ))

2
2
a ( qC ) ( k + c qC ( )) + 2 a ( qC )

3k 2 b2 2 4 ( q )2 2 + 3 2 ( c + q )2

C
C

+ 6k ( b c qC ) 2b (3c 4qC + 3qC )


((b q ) 4a (q ) )
2

Downloaded by [Ume University Library] at 23:01 01 April 2015

95

(A18)
.

Concavity condition for an interior solution: The Hessian matrix of the objective
function(A13) is

+ b

qC

b
+ q
C

2b
2a +
qC

(A19)

The determinant is as follows:


4 a ( qC ) ( b qC )
2

(qC )2

and concavity requires that the determinant be positive, and therefore its numerator is
positive, 4aa(qC)2(babqC)2>0, because the first principal minor 2a is negative.
This in turn requires a>(babqC)2/4a(qC)2.
Positive quality and quality enhancement: Setting q*PC in Equation(A17) to be positive results in the requirement b>ba/qC. For quality enhancement to be positive, we
set (q*PCqC) to be positive, which yields the following:
b<

qC b ( k ca + qC (b a )) 4 aaqC
a ( k ca + qC (b a ))

).

Positive quantity: Setting x*PC in Equation(A17) to be positive results in the requirement b>c((k/a)+qqC).
Comparative statics with marginal cost of quality enhancement and fixed cost of quality: BFPC/b=((xPC)(Dq)/qC) and BFPC/a=(Dq(qPC+qC)) are always negative.
This completes the proof.

Proof of Proposition 2
(i) The effect on internal prices charged by the IT department from of an increase in the
cloud vendors pricing power is obtained by taking the derivative of Equation(A16)
with respect to as follows:

96

Choudhary and Vithayathil

p* PC ( qC ) 2 aqC + ( b qC )
(A20)
=
.
2
2

a
q

4
(
(
)
)
C
C


From Lemma 2, the denominator of (A20) is positive from concavity requirements
for an interior solution. Therefore, it is clear that the sign of the derivative in(A20)
is determined by the sign of the numerator. The condition for a negative numerator is
determined by solving for (qC)2(2aaqC+b(babqC))<0, which results in the following condition:
2

qC > bab/(b2 2aa).

Downloaded by [Ume University Library] at 23:01 01 April 2015

(A21)

When (A21) is satisfied, (A20) is negative and an increase in the cloud vendors
pricing power will reduce the internal prices charged by the IT department. Conversely,
when (A21) is not satisfied, (A20) is positive and an increase in the cloud vendors
pricing power will increase internal prices.
(ii) The effect on quality enhancement by the IT department from of an increase in
the cloud vendors pricing power is obtained by taking the derivative of Equation(A15)
with respect to as follows:

q* PC

)=

( qC ) ( b qC )
2

4 a ( qC ) ( b qC )
2

(A22)


From Lemma 1, the denominator of (A22) is positive from concavity requirements
for an interior solution. We also have b>ba/qC from the requirement for positive
quality in Lemma2. Therefore, the derivative (A22) is negative and the effect of an
increase in the cloud vendors pricing power is to reduce quality enhancement by the
IT department.
The effect on consumption of IT services is obtained by taking the derivative of x*PC
from Equation(A17) with respect to as follows:
2 a 2 ( qC )
x* PC
=
2

4 a ( q ) ( b q
2

2
C)

(A23)


From Lemma 2, the denominator of (A23) is positive from concavity requirements
for an interior solution. The numerator is positive. Hence, the derivative (A23) is
negative.
This completes the proof.

Proof of Proposition 3
The difference in benefit to the firm for the two governance structures is expressed
as

(B

* PC
F

BF*CC > 0,

(A24)

The Impact of Cloud Computing

97

where BF*CC is provided in Equation (A9) and BF*PC is provided in Equation(A18). It can
be verified that (BF*PCBF*CC)=0 is a quadratic function of . Therefore, it yields two
solutions that can be written in the standard form for roots of a quadratic equation:
m2 m22 4 m1m3

2 m1

Solving Equation(A24) for and rearranging the terms, we obtain the following
two roots for Equation (A24):

(
= (q kZ

( Z )) / (q ) Z
+ ( Z )) / ( q ) Z ,

Downloaded by [Ume University Library] at 23:01 01 April 2015

1 = qC kZ 2 ( qC ) ( b c ) Z1
2

2
C
2 ( qC ) ( b c ) Z1

and Z1, Z2, Z3, Z4 are as follows:
Z1 =

qC 2 a 6 a ( qC ) ( ba qC )
2

((b q ) 4a (q ) )

2 2

)+

92 2 aqC + ( 2b qC )

(q 6a + 6b q (3
C

Z3 =

(A26)

2 2

+ 6b qC 3 +

6 ( b qC )

))

6 aqC 6b + qC 32 + 2

(A27)

72 a 3 ( q ) 3 2 aq + ( 2b q ) 48a 3 ( q ) 3 2
C
C
C
C

+ 4 a 2 q 4 2 6 a q 2 b q 2 + b q 4 4
( C)
( C) (
(
C )
C )

(qC )

9 ( b + qC ) 2 aqC + ( 2b + qC )

(6aq

))

+ 2

(4a (q ) + (b q ) )
C

4 a 2 ( qC ) 2 6 a ( qC ) ( b qC ) + ( b qC )

(A25)

6 ( qC a + b ) qC 32 + 2
4

Z2 =

((b qC)2 4a (qC )2 )

(6aq + 6b q (3
C

+ 2

))

(A28)

98

Choudhary and Vithayathil

12 a 2 ( qC ) 2

2 2

4 a ( qC ) + ( b qC )
2

( 6aq 6b + 3q
C

b2 2
+ 2 aqC
4 a ( qC )2 + ( b qC )2

Downloaded by [Ume University Library] at 23:01 01 April 2015

(qC )2
2

4 a ( qC )2 + ( b qC )2

) (
2

+ qC 2

2bqC

2 2

4 a ( qC ) + ( b qC )
2

(A29)

+
.
2
6 aqC 6b + 3qC 2 + qC 2

9

) (
2

32 6b + 3qC 2 + 2qC 2

Z4 =

Note that Z4 is the denominator of the two roots of the solution. We examine
two cases: one where the denominator (qC)3aZ4 is positive and another where it is
negative:
Case 1: Positive values for the denominator
Step 1: When this denominator is positive, it is easy to see that 2>1 because
2 1 = 2

Z 3 / ( qC ) Z 4 > 0.
3

Step 2: Determine parameter values that generate positive Z4:


{k = 10, a = 3.5, c = 0.5, a = 3, b = 10, b = 1, e = 0.1, qC = 0.5}

Z4 = 1.95793, q1 = 0.5755, q2 = 8.44731.

(A30)

Step 3: Evaluate three parameter values to determine the preferred organizational


structure in each of three ranges (,1), (1,2), and (2,). The selected values are
=0.4, =1.0, =9.0, in conjunction with {k = 10, a = 3.5, c = 0.5, a=3, b=10,
b=1, e=0.1, qC = 0.5}.
Step 4: Evaluate the benefits to the firm from profit center and cost center for each
parameter set listed in step 3. This generates the following values for the benefits to
the firm:

q = 0.4, firm benefit from cost center = 30.3018,

firm benefit from profit center = 29.7866

q = 1.0, firm benefit from cost center = 24.0457,

firm benefit from profit center = 25.2093

q = 9.0, firm benefit from cost center = 3.3.5593,

firm benefit from profit center = 1.84837.

(A31)

Step 5: Comparing the results in step 4, it is clear that the profit center is preferred
if and only if 1 < < 2.
Case 2: Negative values for the denominator:
Step 1: When this denominator is negative, it is clear that 2 < 1.

The Impact of Cloud Computing

99

Step 2: Determine parameter values that generate negative Z4:


{k = 10, a = 2, c = 0.5, a = 3, b = 10, b = 1, e = 0.1, qC = 0.5}

Z4 = 1.25, q1 = 26.798, q2 = 7.20204.

(A32)

Downloaded by [Ume University Library] at 23:01 01 April 2015

Step 3: Evaluate three parameter values to determine the preferred organizational


structure in each of three ranges (,2), (2,1), and (1,). The selected values are
=7.0, =10, =27, in conjunction with {k = 10, a = 2, c = 0.5, a=3, b=10,
b=1, e=0.1, qC=0.5}.
Step 4: Evaluate the benefits to the firm from profit center and cost center for each
parameter set listed in step 3. This generates the following values for the benefits to
the firm:
q = 7.0, cost center firm benefit = 0.1875,
profit center firm benefit = 1.125
q = 10.0, cost center firm benefit = 19.1719,
profit center firm benefit = 8.15625
q = 27.0, cost center firm benefit = 525.187,
profit center firm benefit = 526.125.

(A33)

Step 6: Comparing the results in step 4, it is clear that the profit center is preferred
if and only if < 2 or > 1.
Therefore, under all the conditions, the result that the profit center is the preferred
structure if and only if 1< < 2 when Z4>0 and <2 or >1 when Z4<0 holds
for positive values of 1,2. The converse is immediately true that the cost center is the
preferred structure if and only if <1 or >2 when Z4>0 and 2<<1 when Z4<0.
These results hold regardless of whether the parameters generate feasible solutions.
Our region of interest is only in the interior where both structures are feasible.
This completes the proof.

Proof of Proposition 4
The difference in optimal IT quality enhancement between the two governance structures is expressed as follows:

( q

* PC

q ( b q ) ( k + ( b c ) aq )

C
C
C
q*CC =
+ qC
b2 2 2bqC + 2 4 a qC2

))

3 b (q k ) + q k ( c + q )

C
C
C
+ qC = 0.

6 ( aqC + b ) qC 32 + 2

(A34)


Solving (A34) for q and rearranging the terms, we obtain the result that (Dq*PC
Dq*CC)>0 when

100

Choudhary and Vithayathil

2
2

3bk ( b qC ) 4 a ( qC )

> qC (b c + ( k / )) +

6 aqC ( b + qC ) + ( b + qC ) 3b 2 qC

Downloaded by [Ume University Library] at 23:01 01 April 2015

This completes the proof.

/ ( q )2 . (A35)
C

Das könnte Ihnen auch gefallen