tugas review spm

8/10/2019 Tugas Review Spm

1/27

Management Accounting Research 14 (2003) 281307

Management control systems and trust in outsourcing relationships

Kim Langfield-Smith , David Smith

Department of Accounting and Finance, Monash University, Vic. 3800, Australia

Received 16 June 2002; accepted 10 June 2003

Abstract

Outsourcing is a form of strategic alliance that has increased in popularity over the past decade. However, there

has been limited research that studies the design of management control systems (MCS) and the role of trust in

such inter-firm relationships. This paper draws on a model by van der Meer-Kooistra and Vosselman [Acc. Organ.

Society 25 (2001) 51] to examine how control mechanisms and trust are used to achieve control in a single case study

of an electricity company and its outsourced IT operations. An analysis of the characteristics of the transaction,

environment and parties, indicated that the control strategy adopted appeared to be a trust based pattern of control,

rather than a market based or bureaucratic based pattern. Control was achieved through outcome controls and social

controls developing over time, and through the development of trust, particularly goodwill trust. This paper adds tothe growing knowledge of the design of control systems and trust in outsourcing relationships.

2003 Elsevier Ltd. All rights reserved.

Keywords:Outsourcing; Strategic alliance; Management control systems; Transaction cost economics; Trust

1. Introduction

With the advent of globalisation and enhanced levels of competition, many organisations have acknowl-

edged the difficulties of developing and maintaining the range of expertise and skills necessary to compete

successfully. Various forms of cooperative ventures and strategic alliances provide ways of gaining accessto the specialised skills and competencies that are needed to compete effectively in a globalised marketplace (Das and Teng, 2001a; Nooteboom et al., 1997).Strategic alliances are broad ranging relationshipsand can encompass joint ventures, franchises, joint research and development, joint marketing ventures,

long-term supply arrangements, and outsourcing relationships. The outsourcing of core and non-coreactivities are a form of strategic alliance (Nooteboom et al., 1997; van der Meer-Kooistra and Vosselman,

Corresponding author. Tel.: +61-3-9903-1472; fax: +61-3-9903-2577.

E-mail addresses:[email protected] (K. Langfield-Smith), [email protected]

(D. Smith).

1044-5005/$ see front matter 2003 Elsevier Ltd. All rights reserved.doi:10.1016/S1044-5005(03)00046-5


2/27

282 K. Langfield-Smith, D. Smith / Management Accounting Research 14 (2003) 281307

2000; Das and Teng, 2001b), particularly when the outsourced activity is of key strategic importance to

the organisation.Over the past decade, there has been increasing interest in the opportunities provided by strategic

alliances, and outsourcing in particular (Ring and Van de Ven, 1992; Nooteboom et al., 1997; Dasand Teng, 2001b). However, there is also a growing body of evidence of a high failure rate in sucharrangements. One cause of this is the high level of risk associated with alliances, compared to in-house

activities (Das and Teng, 2001a).Aspects that cause high risk include the difficulties inherent in gainingcooperation with partners who have different objectives, and the potential for opportunistic exploitationof the dependence relationship that exists between partners. Appropriate governance structures, including

management control systems (MCS) and the development of trust, may work to reduce risk and decreasefailure (Das and Teng, 2001a; Spekl, 2001).However, there has been only limited attention to the formof management control systems that are suited to strategic alliances, and we have only limited knowledge

of the role that trust plays in such relationships. There have been calls to extend the domain of MCS to

cover inter-firm relationships (Otley, 1994; Hopwood, 1996; Spekl, 2001), and research that considersexplicitly the design of MCS and trust in such situations has begun to appear (see, for example, Sealand Vincent-Jones, 1997; Seal et al., 1999;Das and Teng, 2001a,b). In particular, some research focuseson MCS design and outsourcing relationships (van der Meer-Kooistra and Vosselman, 2000; Mouritsen

et al., 2001; Chua and Mahama, 2002). The purpose of this paper is to build on this prior research, to addto our knowledge of the design of MCS in outsourcing relationships.

Outsourcing is the contracting of any service or activity to a third party (Drtina, 1994; McHugh et al.,

1995). Since the early 1990s, there has been considerable growth in outsourcing, in both the public andprivate sectors, with outsourcing being increasingly applied, not only to manufacturing activities, but

also to traditional in-house administrative and management functions. These include data processing andIT operations, human resource management services, accounting functions, internal audit and marketing

(Chalos, 1995). The practice of outsourcing is an international phenomenon. In the USA, Digital hasoutsourced its entire labour force and much of its production management to a labour hire firm. Theglobal clothing manufacturer Benetton is little more than a shell, having outsourced its product design,manufacturing, and merchandising to contractors. In Australia, the Commonwealth Bank has outsourced

its information technology, printing, record centres, supply functions and mail operations ( Long, 1998;Syvret, 1998).

Much has been written about the criteria that should guide decisions to outsource, and both successes

and failures in outsourcing are reported in the media. The financial press, in particular, tends to focuson the size of the contract, and the identity of the successful bidder. Outsourcing success or failure is

often judged by whether the outsourcer achieves cost savings, or experiences cost over-runs. Few reports,

however, consider how the inter-firm MCS can be designed to suit the particular characteristics of theoutsourcing relationship.

There are several, well-established models or frameworks for studying the design of MCS (see, forexample,Ouchi, 1979; Flamholtz, 1983; Merchant, 1985; Simons, 1995). However, these frameworkstypically focus on control systems within an organization. Only a few comprehensive models consider the

design of MCS in outsourcing relationships. For example, Spekl (2001)developed a model of controlarchetypes, based on transaction cost economics (TCE), andDas and Teng (2001a)modelled the rela-

tionships between MCS, trust and risk in various types of inter-firm relationships.van der Meer-Kooistraand Vosselman (2000)developed a comprehensive model of management control, which was based onprinciples of TCE, but which integrated the role of trust. These models will be drawn on in this paper.


3/27

K. Langfield-Smith, D. Smith / Management Accounting Research 14 (2003) 281307 283

The paper is organised as follows.Section 2provides a review of the literature on the design of MCS

and the role of trust in outsourcing relationships. In particular, the models ofvan der Meer-Kooistra andVosselman (2000),and to a lesser extent Spekl (2001), are drawn on to develop expectations of the

form of controls and trust that are suited in particular transactional situations.Section 3contains a casestudy of an electricity company that outsourced its information technology activities to a single specialistoutsourcer. The fourth section analyses the case study to determine the form of control pattern in place,

the nature of the control mechanisms and the role of trust in establishing close cooperative relationships.The final section concludes with a discussion of directions for further research.

2. Theoretical development

It is only recently that researchers have begun to consider the design of MCS, or governance structures,

in situations that span traditional organisational boundaries, including strategic alliances with suppliers,or outsourcing. These alliances are usually characterised as partnerships or relationships and may involvethe sharing of joint decision making in areas such as strategic planning, and new product and processdevelopment. They may also entail joint investments in relation-specific assets. Compared to arms-length

relationships, which are often well defined and contractually based, outsourcing relationships may en-compass a great deal of uncertainty or risk for both parties. The nature of the contract between the twoparties can be complex and cover many areas of interest that extend beyond the mere actions of supply

and receipt of goods and services. It has been argued that the complexity of such arrangements maypreclude the complete ex ante specification of detailed contracts (Heide, 1994).In addition, the need for

flexibility and adaptation in those partnerships may imply that control systems rely less on formal mech-anisms (Gietzmann, 1996; Ittner et al., 1999). However, in some outsourcing situations the institution

of a formal control system may enable greater control and transparency, which may not only impact onthe interorganisational relationship, but may have implications for strengthening control and increasinginsights within the outsourcing firm (Mouritsen et al., 2001).

A useful starting point in studying control systems in outsourcing relationships is to consider how

control systems within a single firm have been conceptualised.

2.1. Within-firm control systems

Control systems have been conceptualised and categorised in various ways: formal versus informalcontrols, behaviour versus outcome controls; mechanisticversus organic controls; bureaucratic versus clan

controls. However, these typifications are not distinct and there is some agreement that all organisationalcontrol systems consist of formal, explicitly designed controls, as well as the unwritten informal or socialcontrols that cannot be designed directly. Within formal designed controls, some researchers distinguishbetween outcome controls and behaviour controls (Ouchi, 1979; Eisenhardt, 1985). Outcome controls

measure and monitor the outputs of operations or behaviours, using techniques such as performancemeasurement. Behaviour controls, such as rules and standard operating procedures, specify and monitor

individuals behaviours.Clan or social controls are present in all organisations to varying degrees (Ouchi, 1979). Social controls

may develop from shared norms, values and beliefs, and may rely on the internalisation of goals, which

leads to organisationally desired behaviours. These controls cannot be designed explicitly, but can be


4/27


shaped and influenced by activities such as frequent interactions, meetings, negotiation of disputes, codes

of conduct, senior management attitudes and style, and rituals. There is some evidence that social controlsare also relevant in inter-firm relationships (Gietzmann, 1996; Nooteboom et al., 1997).

Ouchi (1979) highlighted the relationships between the mode of control, and the information char-acteristics of the task: the degree of output measurability and task programmability. Outcome controlswere said to be suitable in situations of high output measurability and low task programmability, whereas

behaviour controls suit situations of low output measurability and high task programmability. Where bothoutput measurability and task programmability are high, either behaviour or outcome controls may beused. Social controls emerge when both output measurability and task programmability are low.

2.2. Trust and control systems

Merchant (1985,p. 39) noted that almost every control system involves some degree of trust that the

individuals of concern will do what is best for the organization without any, or with only incomplete,monitoring of actions or results. However, it is only relatively recently that researchers have begun tostudy the specific relationships between trust and the MCS design, but there are inconsistencies. Some

researchers claim that control mechanisms and trust can be pursued simultaneously, or are complementary(Zaheer and Venkatraman, 1995; Goold and Campbell, 1987; Das and Teng, 1998), while others arguethat control mechanisms are detrimental to trust (see, for example, Lorange and Roos, 1992). However,

researchers have not always considered how specific forms of trust and specific types of controls relateor interact.

Several researchers have focused on the role of trust in governance relationships (see, for example,Zaheer and Venkatraman, 1995; Chiles and McMackin, 1996; Gietzmann, 1996; Nooteboom et al., 1997;Seal and Vincent-Jones, 1997). Trust may develop over time through processes of learning and adaptation,

which are essential to the strengthening of the relationship between partners, making the relationshipmore durable in the face of conflict and encouraging interactions between partners involving knowledgeexchange and promotion of each others interests (Johanson and Mattsson, 1987). Close relationships with

suppliers may involve the sharing of information, joint product and process development and joint costimprovement activities, and trust allows such alliances to flourish. It has been argued that certain minimumlevels of trust are essential in inter-firm relationships, as trust reduces the possibility of opportunistic

behaviour (Axelrod, 1984; Bradach and Eccles, 1989; Birnberg, 1998). In addition, trust may increasethe predictability of mutual behaviour through each party honouring commitments and allowing partners

to deal with unforeseen contingencies in mutually acceptable ways (Sako, 1992,p. 37).Trust is a difficult concept to study as it has been defined and classified in many ways. Most definitions

of trust focus on exposing oneself to vulnerability. A simple definition is that trust is having confidence thatones expectations will be realised (Luhmann, 1979). Other definitions suggest that trust entails positiveexpectations regarding the other in a risky situation (Gambetta, 1988), and includes adopting a belief,without having full information to confirm that belief (Tomkins, 2001). Several researchers have noted

the link between trust and information requirements (see, for example,Luhmann, 1979; Creed and Miles,1996; Wicks et al., 1999; Tomkins, 2001). Trust is characterised as an alternative uncertainty absorption

mechanism to providing increased information. This has led some researchers to suggest that there is aninverse relationship between willingness to trust and the need for information (see, for example,Wickset al., 1999), while others see the trade-off as more complex (see, for example, Seal and Vincent-Jones,

1997; Tomkins, 2001). The provision of information, such as cost reports or performance information,


5/27


is a component of MCS, and is certainly important in outsourcing relationships, where the sharing of

information and the generation of performance information is common.It has been argued that trust is particularly relevant to inter-firm relationships, as trust is only important

in situations where there is risk (Luhmann, 1979; Coleman, 1990; Sako, 1992), and risk managementis a critical aspect of these relationships (Ring and Van de Ven, 1992; Das and Teng, 2001b). Threedefinitions of trust relevant to managing outsourcing are contractual trust, competence trust and goodwill

trust (Sako, 1992). Contractual trustis based on the moral standard of honesty, and rests on an assumptionthat the other party will honour the agreement, whether the agreement is in writing or not ( Sako, 1992;van der Meer-Kooistra and Vosselman, 2000).The higher the level of contractual trust that a firm has in

an outsourcer, the less need there is to gather information to prevent or reduce opportunistic behaviour.

Competence trustfocuses on perceptions of ability and expertise, and is the expectation of technicallycompetent role performance (Barber, 1983,p. 14). In inter-firm relationships, competence trust relates

to a partners ability to perform according to the specified agreement or contract (Nooteboom, 1996). In

contrast, goodwill trustcan be defined as perceptions of a partners intention to perform in accordancewith those agreements (Ring and Van de Ven, 1992; Nooteboom, 1996). Goodwill trust is associatedwith integrity, responsibility and dependability (Das and Teng, 2001a). While these forms of trust maybe present to some extent in the early stages of an outsourcing relationship, they can also develop further

over time.

2.3. Transaction cost economics and control

A common framework for viewing the choice of governance structures in inter-firm relationships istransaction cost economics (see, for example,Gietzmann, 1996; Seal et al., 1999; van der Meer-Kooistraand Vosselman, 2000; Spekl, 2001). TCE is based on the notion that firms choose efficient organisational

forms or governance structures based on transactional issues, such as firm-specific investments, andexternal or internal uncertainty (Ittner et al., 1999).Governance structures can be characterised as one ofthree forms: markets, hybrids (including strategic alliances) and hierarchies (Williamson, 1991).TCE is

based on the idea that three aspects of transactions determine the appropriate mode of governance: the

frequency of the transaction, the uncertainty encompassed in those transactions, and the asset specificity ofthe transactions (Williamson, 1979).Asset specificity, in particular, is said to be of particular significance

in explaining the choice of governance structure.Asset specificity is the degree to which an asset can be redeployed to alternative use without sacrifice of

productive value. It is the opportunity loss associated with the early termination of a relationship (Spekl,2001).Asset specificity can arise in six ways: site specificity, physical assets specificity, human assets

specificity, brand name capital, dedicated assets and temporal specificity (Williamson, 1991). Under TCE,it is assumed that a high level of asset specificity creates dependency between the parties in a relationship,increases switching costs and leads to difficult governance situations.

Markets, hybrids and hierarchies rely on different control mechanisms to enable successful contracting.

Whereas markets rely on free competition to ensure control, andhierarchies rely on authority, hybridformsof control generally entail long-term contracts. Hybrid governance structures are long-term contractual

relations that preserve autonomy by providing added transaction-specific safeguards (Williamson, 1996,p. 378). The safeguards often include hostage arrangements to correct market asymmetries. For example,parties to the relationship may transfer assets that can only be fully recovered if the contract is executed

successfully. High levels of uncertainty associated with hybrids, may make it difficult to specify ex


6/27


ante performance criteria within contracts, and in the face of high asset specificity, alternative control

mechanisms such as specialised dispute resolution mechanisms may need to be put in place.TCE models have been criticised for not considering adequately thesocial contextwithin which trans-

actions are embedded. It has been argued that social embeddedness not only influences the design ofthe control systems, it also influences the relationship and each partys behaviour, including the levelof opportunism (Granovetter, 1985; van der Meer-Kooistra and Vosselman, 2000). A critical concern in

all inter-firm relationships is the attitudes and personal relationships between the two parties ( Faulkner,1995).In particular, trust has emerged as an important way of reducing opportunism, and as a factor inthe design and study of control systems (Ring and Van de Ven, 1992; Gietzmann, 1996; Nooteboom et al.,

1997; van der Meer-Kooistra and Vosselman, 2000). In the following section, two recent models that drawon TCE to depict control systems in outsourcing relationships will be outlined; one model integrates TCEand trust.

2.4. TCE-based models of management control

Spekl (2001)developed a TCE theory of management control to explain nine archetypes of control.

These control archetypes vary depending on three dimensions: task programmability, ex post informationimpactedness (output measurability) and asset specificity. Spekl (2001) specified two control archetypesfor outsourcing relationships: hybrid arms-length control and hybrid exploratory control. To date, Spekls

model has not been used in empirical research.van der Meer-Kooistra and Vosselman (2000)developeda model of control in inter-firm relationships that integrates TCE concepts and trust. They identified three

management control patterns relevant to outsourcing relationships: the market based pattern, bureaucracy

based pattern and trust based pattern. In the market based pattern, market mechanisms dominate therelationship, and this is similar to Spekls archetype of market control. However, Spekl does not see

market control as a suitable archetype for outsourcing relationships. The bureaucracy based pattern isanalogous to Spekls hybrid arms-length control, while the trust based pattern bears some similarities tohybrid exploratory control. While Spekls (2001) model focuses on the characteristics of the transaction

as a determinate of the control systems archetype,van der Meer-Kooistra and Vosselman (2000)providea more complete analysis by considering not only the transaction characteristics, but the transactionenvironment, the characteristic of the parties to the transaction, and the role of trust in achieving control.

The market based pattern suits transactions characterised by high task programmability, high mea-surability of output, low asset specificity and high task repetition. Many suppliers will compete for the

contract, and market prices will be directly linked to the quality of the output of the outsourcers activi-ties. Detailed contracts are not required and the possibility of returning to the market for competing bids

provides discipline for the current outsourcer to provide an efficient and effective output. In the face ofeffective market mechanisms, no specific control instruments are needed to manage the relationship. Theinstitutional environment is not relevant, and nor is supplier reputation, prior history of cooperation, or riskattitude. If one party to the relationship behaves opportunistically, another party can be chosen without

high switching costs, as there are no specific investments. The transaction environment is characterisedby low uncertainty and many available alternative suppliers.

Thebureaucracy based pattern, or hybrid arms-length control, suits transactions that have high taskprogrammability, high output measurability, moderate asset specificity and low to medium repetitiveness.The transaction environment has relatively low uncertainty and the future is fairly predictable. Controls

will be prescriptive and include detailed rules of behaviour and rigid performance targets. These will


7/27


be captured in detailed contracts, which are used to monitor performance. Comprehensive selection

criteria are set up and formal bidding is used to select a partner. Hostage arrangements can be used toensure compliance to contractual provisions, and arbitration may be used to resolve contract disputes

and to counter opportunistic behaviour in an environment of moderate asset specificity (Spekl, 2001).Contracts are specific and long-term, and the autonomy of the two parties is preserved. In this situation, acombinationof behaviour and outcome controls will be used, which is consistent with Ouchis prescription

for control systems in the face of high task programmability and high output measurability. Trust plays alimited role in the bureaucracy based pattern, but is important in the early stages of a relationship. Wherehuman knowledge and skills are critical to the quality of the work, the outsourcing firm must perceive that

the outsourcer has high levels of competence trust and contractual trust in order to select the outsourcerand to proceed with the contract.

Thetrust based patternofvan der Meer-Kooistra and Vosselman (2000)is characterised by low levels

of task programmability, low levels of output measurability and high asset specificity. Transactions are

not highly repetitive. The environment is highly uncertain and risky, so trust becomes the dominantmechanism for achieving control in this form of relationship, and this mitigates the risks associated withhigh asset specificity. The initial selection of the outsourcer is based on perceptions of competence trust,contractual trust and goodwill trust, which arise through friendships, former contractual relationships

and reputation. Initially, contracts are merely broad frameworks, which then develop further over time. Aseries of control devices, such as personal consultations and intense communications, are put in place todevelop competence trust and goodwill trust. In addition, the institutional environment can stimulate the

development of competence trust and contractual trust, through certification of the firms activities andlegal regulations. Goodwill trust becomes the solution to overcoming information asymmetry, and regular

personal contacts and an attitude of commitment can lead to its development. Control systems, in general,are more informal under these forms of relationships, and often take the form of social controls. Consistent

withOuchi (1979),behaviour controls are not suited in these situations of low task programmability, andwhen controls are formalised they tend to emphasise outcome controls, and these develop over timethrough the sharing of private information and the alignment of the parties performance expectations.Trust is necessary to achieve control, as activities and output cannot be measured with any certainty.

The transaction characteristics ofhybrid exploratory controlare similar to those of trust based control,except that asset specificity is moderate rather than high. Under the Spekl model (and under TCE ingeneral) high asset specificity (as found under a trust based pattern) cannot be tolerated in an outsourcing

situation as it increases the potential for opportunistic behaviour and information leakage, requiring theoutsourced function to be taken back in-house. However, firms do continue to engage in outsourcing in

situations of high asset specificity. The trust based pattern demonstrates that the development of goodwill

trust and contractual trust can mitigate opportunistic behaviour and the abuse of unequal bargaining power(Sako, 1992, p. 39;van der Meer-Kooistra and Vosselman, 2000). Similarly, under hybrid exploratory

control, exclusive contacts with suppliers are considered unacceptable, as they increase asset specificity,dependence and risk, in the light of incomplete contracts. However, organisations do enter into exclu-sive contracts with outsourcers. Again, goodwill trust and contractual trust will counter the potential

opportunistic behaviour.Table 1contains a summary of the characteristics of the transaction, transactionenvironment and parties for the three patterns of control, as well as the form of control mechanisms and

the role of trust in achieving control.Thus, depending on the characteristics of the transaction, the transaction environment and the parties,

we would expect the MCS of outsourced operations to follow one of the three control patterns outlined


8/27


Table 1

Control mechanisms and trust

Outsourcing

control pattern

Characteristics of the transaction,

transaction environment and parties

Control mechanisms The role of trust in achieving

control

Market based

pattern

Transaction

High task programmability

High output measurability

Low asset specificity

High repetition of transactions

No specific control instruments

required as market mechanisms

dominate

Not relevantswitching costs

are low

Competitive bidding at

periodic intervals

No detailed contracting

Market prices linked to

standardised activities and

outputs

Transaction environment

Many potential parties

Market price contains all the market

information

Social embeddedness and

institutional factors not relevant

PartiesNot important

Bureaucratic

based pattern

Transaction

High task programmability

High output measurability

Moderate asset specificity

Low to medium repetition of

transactions

Outcome and behaviour

controls, focused on direct

intervention by outsourcing

party

In selecting the outsourcer,

when human knowledge and

skills are important to the

quality of the work, the

outsourcing firm must perceive

high levels of competence trust

and contractual trust in the

outsourcer

Rigid performance targets

Detailed rules of behaviour

Detailed contracts

Comprehensive selection

criteria and formal bidding

Transaction environment Hostage arrangements

Future contingencies known

Medium to high market risks

Institutional factors influence

contractual rules

Parties

Competence reputation

Medium risk sharing attitude

Asymmetry in bargaining power

Trust based

pattern

Transaction

Low task programmability

Low output measurability, that tends

to increase over time

High asset specificity

Low repetition of transactions

Outcome and social controls

develop over time

Broad non-specific contracts

that develop time

Performance assessed

through broad emergent

standards

High levels of information

sharing and communications

Perceptions of competence

trust, contractual trust and

goodwill trust may determine

the selection of outsourcer,

and must be assessed in

advance

The institutional environment

can stimulate competence

trust and contractual trust


9/27


Table 1 (Continued)

Outsourcing

control pattern

Characteristics of the transaction,

transaction environment and parties

Control mechanisms The role of trust in achieving

control

Opportunistic behaviour and

information asymmetry will

be overcome by developing

goodwill trust and

contractual trust

Regular personal contacts,

intense communications and

an attitude of commitment

can stimulate competence

and goodwill trust

Transaction environment

Future contingencies unknown

High market risks

Social embeddedness

Institutional factors influence the

relation

Parties

Competence reputation

Experience in networks

Experience with contracting parties

Risk sharing attitude

No asymmetry in bargaining power

above. In a market based pattern, there are no explicit control mechanisms and trust is not relevant inachieving control. Under a bureaucratic based pattern, outcome controls and behaviour controls are the

prime control mechanisms, and trust plays a minor role in achieving control. Under a trust based pattern,outcome controls and social controls emerge over time, and trust plays a significant role in achieving

control.

3. Case descriptionthe outsourcing of IT&T at Central Energy (Central)

Central Energy is an Australian company operating in the electricity industry, which outsourced its

information technology function. The company was selected as a research site, as it was known to havewell-established outsourcing contracts.

Data were collected through semi-structured interviews with key managers in the electricity firm whowere involved in the decision to outsource the operation, and in the ongoing management of the outsourcedfunction. These managers worked in the areas of finance, information technology, shared services and

contracting.1 Interviews lasted two to three hours each. Access was also gained to relevant companydocuments. Two researchers were present at each interview to enhance the reliability of the interpretationof interview material and conclusions drawn. All interviews were tape recorded and transcribed. The

completed write-up of the case study was sent to the interviewees for comment and to the company for

final approval.2The purpose of the interviews was to understand the nature of the relationship between the company

and the outsourcer and type of control system used in the outsourcing relationship. To provide a con-text for the study, the following two issues were investigated: the motivation for the firm to consider

outsourcing, and the criteria used to select the outsourcer. To investigate the form of MCS, the focus

1 Four different managers were interviewed (one was interviewed twice) at the end of the first two years of the outsourcing

contact.2 Even though all company names are disguised, as part of gaining access to the company the researchers agreed to allow the

company final approval over any material published.


10/27


was on understanding the nature of the formal contract entered into with the outsourcer, and the pro-

cesses used to achieve control over the first two years of the relationship. To assess the appropriatenessof the control activities and to obtain a perspective of the managers perceived success or otherwise

of the outsourcing venture, managers were asked to reflect on the benefits that they had gained fromoutsourcing.

3.1. Background

Central Energy was formed in 1995, through the merger of two electricity authorities, Hilton Energy(Hilton) and Woodside Electricity (Woodside). Central distributes and retails electricity and value-addedenergy services to more than a million people across a large, diverse area. Central also sells major electrical

contracting services throughout Australia and retails energy services to major industrial and commercialclients across two states of Australia.

Deregulation of the Australian electricity industry led to considerable changes for organisations likeCentral. Increased competition occurred through the privatisation or corporatisation of state-owned mo-nopolies. In Centrals home state, the industry was broken up into a number of separate corporatisedentities that manage electricity distribution, generation, and the electricity grid. Adding to the increased

competitive pressure, all customers were able to buy power from any of the states electricity distribu-tors. Thus, distributors can seek customers from other states, making the customer base more vulnerable.

Improved customer service and reduced costs have become prime considerations. To remain competitivein this environment, Central established four separate businesses and a holding company. Each businesswas required to make a profit. When services were provided from one business to another, they were

supplied on a commercial arms-length basis.During the merger, the executive steering committee decided to outsource the information technology

and telecommunications (IT&T) function. This occurred within two months of the merger.

3.2. Motivation for outsourcing IT&T

Centrals stated policy was to . . . ensure that all internal services are provided effectively and effi-ciently to the core Network and Energy Services businesses. Where internal services did not meet the

level of quality or cost of external suppliers, competitive tenders were sought and evaluated. Thus, costeffectiveness and strategic issues were primary considerations in outsourcing decisions. However, the

specific factors that motivated the decision to outsource IT&T were broader and included the search for asolution to the problem of merging two very different IT cultures; the need to improve cost management

of IT&T; greater access to technical knowledge and expertise; and the need to bring more discipline andcontrol to IT&T.The merger of the two electricity authorities brought together two IT&T functions that were very

different in size and nature, as well as focus and direction. Developing a single, effective IT&T function

was seen as a major managerial issue and an important determinant of Centrals ability to compete inthe new contestable electricity environment. Hilton had 168 IT&T staff, compared with Woodsides 33.

Hilton was a mainframe environment with large-scale applications, and frequent cost over-runs. Woodsidehad a very low cost infrastructure: its systems were small, not run on mainframes, and there were limitedfunds for software development. Many of Woodsides IT staff took voluntary redundancies with the

merger.


11/27


Before the merger, cost management of the IT&T function was poor, particularly at Hilton. Cost

over-runs were common, and there was little control over the initiation of new projects. Hiltons IT&T wasstrongly customer-focused so that a plethora of operating systems was supported and new applications

were developed continually. There was limited accountability for spending on IT resources. A seniormanager described the situation:

A business would ask IT&T for something to be done, but they were not being charged the true cost.There was no discipline in the management of costs. One of the drivers in the move to outsource the

IT&T area was to introduce that discipline, so that if you do ask for something to be modified orchanged, someone comes back to you and says You know its going to cost you $20,000.

The transfer pricing system used to charge users of IT&T services sent the wrong signals, and led todysfunctional decision making. Favours and funny prices led to a lack of accountability and an absence

of commercial reality. Managers hoped that outsourcing of IT&T would help curb spending, as real

money would be seen to be leaving the organisation. An IT manager described past practices:The IT cost was like a bottomless bucket. The projects were actually controlled by the IT people, notby the users. They might have been initiated by the users, but basically, the users lost control. So the IT

people started to dominate the situation. Now that is the fault of the users, not a fault of the IT people.The organisation recognised that it did not manage that well.

The cost of in-house IT&T for the two electricity distributors before the merger was estimated at $42million per year. Many managers, however, believed that this was overstated. Nevertheless, IT&T was

clearly a significant cost to the business and so provided opportunities for cost improvement and achievingmore control over IT developments.

Another reason for engaging in outsourcing was that the organisation would need to gain access to

high levels of IT expertise to compete effectively. It was considered difficult and impractical to supplythis in-house. One manager described the thinking:

I think that at the end of the day, you really have to understand what your core competencies are. Thatis, what are the things that you are good at, that you do have an edge on, and that you really know the

ins and outs of it. No matter what business you are in, there are some things that you can do better thanothers. It is a matter of identifying and really understanding those things. If you go back a few years,we probably did not really have a feel for thatwe continued to do things because we had always done

them. It is a matter of sitting down and saying, what are we good at? What do we do that adds valueto our organisation? And if there is some lineball exercise or some clear-cut situation where we arenot adding value, then we have to think seriously about doing something else.

Central understood that its core business was energy services. The environment in which it was op-erating placed greater demands on the business for smarter systems, which would involve high use of

sophisticated IT. Thus, it was envisaged that there would be greater reliance on the IT&T function in thefuture, and it was of critical importance for survival. An external provider was believed to offer several

advantages over the in-house department. A manager described this:

The market is changing very dramatically and we do not have the depth in the IT group to be able

to provide IT solutions. We wanted to work with a partner whose core business was IT and thereforewas in a better position to search the world, to provide us with the right solutions at the right time.Now I do not want to appear to be negative towards our IT groups, because they did carry us a long


12/27


way forward. The question is, did they get too expensive? We lost control of the IT spend, but also the

market had moved a lot quicker, and we dont believe the IT groups were dynamic enough to providethe necessary solutions to go forward.

Thus, a strong motivation for outsourcing was managers perceptions that the use and development ofIT within Hilton was out of control. One senior IT manager described the situation that had existed at

Hilton:

In the past, utilities were fairly wealthy organisationscash rich and managed predominantly by

engineers who liked to build gold-plated systems. We now have redundant microwave links, multiplecomputer sites, you name it. It was technology gone mad. Whereas the old Woodside was managed bya man whose single focus in life was to build low cost solutions.

Some businesses within Central had started to develop their own IT groups and this hastened theneed to engage a highly competent IT outsourcer to bring IT planning and operations under central

control.

3.3. Criteria used in the outsourcing decision

In mid-1996, Central called for tenders for the IT&T function. Bids from two large global companieswere considered, along with an in-house bid. All bidders needed to demonstrate that they could offer value

within the dynamic electricity environment. The criteria for assessing the tenders included the following:the cost of providing the service; clear cost reduction paths; staff transition issues; the approach to

strategic planning of IT; what the tenderer could offer in terms of their experience, skills base andcompetence; how they planned to manage the contract; and the nature of the billing arrangements. Eachcriterion was weighted in the final analysis, with the final decision being made by the implementation

committee.The successful tenderer, Global Systems (Global), was chosen for many reasons, including its reputation

and company values, the quality of its proposal, the ability to respond speedily to changes in the market,

leverage and wide international knowledge base. Other factors thatfavoured the tendererwere the time-lineon cost reduction paths, and the way Global intended to deal with staffing issues. The in-house proposalwas said to be a clear third, largely because the team were unable to provide convincing arguments that

they were able to provide the necessary IT solutions for the future.Interestingly, no IT staff were on the implementation committee as it was felt that they would have

too much self-interest in the projects outcome. However, they were consulted in the decision. Globalwas physically located on-site, in the same area as Hiltons previous IT&T department. Central retained

ownership of the IT equipment.Once the decision was made to outsource IT&T to Global, a team was set up to work with staff ontransition issues. Existing IT&T staff were given three choices: take a voluntary redundancy package,transfer their employment to Global, or continue at Central in a different role. About half of the IT&T staff

left, with most of the remainder transferring to Global. Counselling and career transition training wereprovided for some staff. Many employees that transferred to Global saw better opportunities for career

advancement in an organisation that had IT as its core business. Some initial problems were experienced,with specialised knowledge of the IT systems walking out the door. As in many companies, whenredundancy packages are offered, it is often the most talented staff with best employment opportunities

that choose to leave. Many of the IT applications at Central were custom-built and not well documented,


13/27


so often only a few employees knew how particular software operated. In addition, over time some of

Centrals IT staff who had transferred to Global moved to other Global sites.

3.4. Managing the outsourcing relationship

The first twelve months of the outsourcing relationship were described by one manager as rocky.Indeed, it took about eighteen months for positives to clearly emerge. The key issues to managewere: agreeing on the baseline level of service to be included in the contract; verifying the cost of

the baseline services; managing false expectations of staff; managing differences between the cul-tures of Central and Global; the implementation of communication processes; developing trust; achie-

ving cost reduction; and implementation of the risk-reward system. These issues are considered in turnbelow.

The contract contained an agreed price for baseline services and additional payments for discretionary

projects that were above that baseline. Baseline services were defined as . . .

the cost of maintainingall infrastructure and systems in the organisation at the level at which they were at the point that thecontract was signed. This definition was difficult to interpret. The problem in establishing the baseline

involved agreeing on which projects were discretionary (over and above baseline services and hence anextra cost to Central) and which projects were included in the baseline service. During the first yearof the contract there were conflicts between Central and Global, as each had different ideas as to what

constituted baseline services.Some managers at Central believed that clear contract specifications were critical to ensuring a smooth

relationship:

People have to be very clear about their service level requirements. Not say 95 per cent uptime

or something like that. None of this arbitrary nonsense. Be very specific about what those ser-vice level requirements arespecify as much as you can get down. And people may say Oh, its

not necessary because youre going to work these things out together. You can, but that meansthere needs to be extreme trust and that only occurs when you are really partners. But you do notstart as partners. You start with the specifications; you work to it and the supplier/customer rela-

tionship. And that is where we came from. In the beginning it was an unhappy customer/supplierrelationship.

Some managers believed that these problems could have been averted if more time had been spentnegotiating the contract. However, at the time of the merger there was a great sense of urgency to solve

the IT problem, and to get an outsourcer onboard.

Another issue that dominated the early days of the relationship was an extensive verification of the costof baseline services, which had not been specified in the contract. A manager explains why this was thecase:

With the merger of the organisation, redundancy programs were going on, and with uncertainty inIT staff, there was a need to do something to get some stability in the process. The cost of putting

some trust in Global Systems during that six to nine months process was seen to be less than the costof delaying the contract verification prior to final contract signing. This was the cost of provision ofbaseline services. We entered into a contract that basically said, This is what we believe it is at the

moment, but it will be subject to verification in this next six months. That should not have occurred.


14/27


We should have spent more time as a business, chapter and verse, detailing what our expectations of

service were, what the limitations were, what they were really going to get out of it.

In the early days of the relationship many staff at Central were disappointed with Globals performance.There was an expectation that Global would deliver massive improvements in service levels and costsimmediately. There were also memories of how good things were before the partnership.

Improvements were slow to emerge for a number of reasons. Some of the most valuable IT&T staff hadleft the company, so the remaining staff did not have the skills to run some systems. Conflicts emerged

from the different objectives of the two parties: the outsourcer was clearly trying to make a good returnfrom the account, while Central was trying to manage costs. It took some time before the two partnerswere able to work towards shared objectives and targets for various projects. Also, managers within the

businesses were now being charged the full costs of their IT&T services, whereas they had previouslyhad much greater direct access to IT&T services, which carried few charges.

The decision to outsource IT&T was made within two months of the merger. In itself, this was a difficult

blending of two very different cultures, but Central and Global also had very different organisationalcultures. Central was a merger of two traditional public sector organisations, but it was moving quickly toadopt commercial practices and principles. Global was highly commercial and controlled in its operations.

One senior manager contrasted the two organisations:

The way Central works is to have the process in place, but to empower people to do what they need todo. The account is probably worth 40 to 45 million dollars a year to Global Systems. My equivalent,

the number one person in Global Systems, has signing authority to only $5,000everything else goesup the line into the bureaucracy. You can talk about bureaucracy in the public sector, but it is nothingto what you see in the way Global controls what you can and cannot do. That is the way they work.

But having said that, they have introduced a rigour into this organisation, which was far too free with

its IT spend. That is a difficult thing to do with an internal IT shop.

After the first eighteen months these differences became less important. One manager believed thatthis might have been due to Global learning the Central way.

The manner in which Central related to Global changed over the first two years of the contract. Ini-tially, Centrals IT&T OutsourcingManager was the sole contact point for Central, handling all day-to-dayproblems and trying to build strong relationships. However, the four businesses of Central were distinct

and had very different needs and foci. Over time, Global assigned different service managers with re-sponsibility for each business, to handle their own specific day-to-day issues. This allowed the IT&T

Outsourcing Manager to focus more on broader relationship matters. A greater focus came to be placedon open communication channels and addressing issues up-front. Senior managers from both Central and

Global met monthly as part of an IT steering group, to work through the strategic plan, to discuss majorprojects and to review spending to date. There was also a working group to handle applications for newIT capital projects.

The issue of developing trust between Global and Central was seen as important to managers at Central.

Over time, the managers came to understand that Global was committed to the relationship and theyrespected to high levels of skills and experience that Global brought to the relationship. A manager

outlined this viewpoint:

Central still wears by far the biggest risk in this relationship. If you assume the absolute worst, that

Global Systems delivers a system late, over budget and with poor quality, sure it might lose out on


15/27


a couple of million dollars profit, but the cost to Central would be significantly more. So there is an

element of trust in this relationship, and Global Systems has a reputation to sustain in the marketplace.It cannot afford to walk away from something like thisand it will not. These large organisations will

not and there is a good relationship between our CEO and their CEO to ensure that it simply wouldnot happen. There is too much win-win potential to come out of this.

Both parties worked towards building high levels of trust between the two organisations. Where there

were incompatibilities, Global replaced some of its staff who were directly involved with Central. Amongsome of the individual businesses within Central, one source of mistrust arose from businesses beingcharged for IT applications. Some managers felt that Global was ripping them off and was engaging in

point scoring. The source of this problem, however, was that, for the first time, the businesses were nowbeing charged directly for using IT&T services.

The contract with Global was for the provision of management services, while Central continued toown the IT infrastructure. One of the reasons for not giving total control to Global was the risk that if the

relationship were to fail with no control over IT assets, it would be difficult to recommence IT servicesin-house or with another outsourcer. A senior manager explained the relationship between trust and risk.

The more you trust, the more you are prepared to hand over. The more you feel Hey things couldbreak down. What recourse do we have? What backups do we have? How do we rebuild if there is a

divorce? There is some hesitancy in relation to the ownership of assetsbut there is a gradual moveaway from that thinking.

New management systems were implemented in Central to achieve greater cost control and managementof IT developments. Each of the businesses was required to submit an annual plan for new projects. Each

project needed to have strong business justifications, and was reviewed by a committee that prioritised the

projects. An overall strategic plan for IT, formulated by Central and Global, was developed. This systemis a clear contrast to the pre-merger practice of ad hoc project development.

Another aspect that encouraged cost management was the direct charging of IT&T costs to the busi-nesses. These costs were major overhead items that could run to many millions of dollars for each business.

The new systems encouraged an environment of accountability and cost consciousness. Managers withinthe business units began to look for ways that they could break down the costs into manageable elementsthat people in the operational areas could influence.

There were no performance measures in the initial outsourcing contract. Eighteen months into thecontract, a risk-reward scheme was devised to monitor Globals performance. The first application of thiswas in relation to a new customer service system (CSS), a discretionary project. The system was developed

after several weeks of negotiation between the two parties. When Global undertook a discretionaryproject, it would usually levy a charge that included direct costs, overhead and profit margin. Under therisk-reward system Global would charge only direct costs to Central. Bonuses would then be paid, based

on performance in three areas: cost, quality and time. Table 2provides some examples of the types ofmeasures that were used.

Thus, Central awarded Global a score out of 25 based on how prompt they were with delivering theproject on time. Depending on how the new system affected business continuity (which might be testedover several months), Global would receive a score of up to 20. (Business continuity was concerned with

whether the new system resulted in interruptions to business, such as computer downtime, interruptedaccess data, and external customer problems.) The targets that were set for each of the performance


16/27


Table 2

Performance measures used in the risk-reward scheme

Criteria Performance measures Weightings (%)

Cost Under- or over-budget 25

Quality Performance to test plan 20

Survey of end users 10

Business continuity 20

Time Delivery to schedule 25

Total 100

measures were very challenging. The score that Global received was then linked to a profit multiplier. An

overall score of 70 would return a normal profit for the project to Global, whilst a score of 90100 wouldgive it up to 150% of the normal profit. Global also conducted its own customer satisfaction surveys andthese data were shared with Central.

3.5. Benefits of outsourcing

Managers had high expectations of the improvements that would result from outsourcing the IT&T

function. However, benefits were slow to materialise, and it was not until after the first eighteen monthsof the contract that improvements could be clearly identified by Centrals managers. These benefitsincluded the benefits of access to Globals expertise and enhanced services; increased accountability and

cost consciousness within Central; and a greater discipline in IT planning.

The new competitive electricity environment placed increased demands on Central to gain access tohigh IT skills and technologies. There was a strong motivation for outsourcing IT&T and Global was ableto provide those benefits to Central. A senior manager described the difference Global made in this area:

I would say that they have brought into the organisation, in key areas, skills that we could not have hopedto bring in. The year 2000 problem is a classic case. They have a big contract with a state government,

and the skills that they built up there and that they have available to them worldwide are very, verygood. They brought two specialists in. We paid big bucks for it, but we got it done in a fraction of thetime that it would have taken us to do it internallyto do the reviews, and put the processes in place

to allow us to then roll out our year 2000 plan. The ability to do that and put those processes in placein an internal shop and get the right people for short periods of timealmost impossible!

The association with Global provided Central with the creative IT solutions that it was looking for.Another manager confirmed the importance of Globals international networks:

And that is the way Global Systems operates, they have a client here, and a client there and if we have

a certain issue, what they do is bring someone in. They fly someone in from the States or locally tosolve that specific problem because they have addressed it before in other organisations.

While IT&T was a major cost for the company, managers at Central were not overly concerned aboutreducing these costs. Instead, they were more focused on ensuring that the resources were not wasted. To

that end, the company put in place new systems to help manage IT activities.


17/27


To encourage responsible IT use, the cost of the baseline services needed to be redistributed to the

businesses that were using the services. A manager explained the philosophy with relation to the desktopproject, which was a project within the baseline.

What we have agreed to on the desktop project is based on actual charges per desktop per month fora total service. We are putting the incentive clearly back on the business to manage the number of

PCs. Prior to this theres been absolutely no incentive for them to do anything smarter about managingthose costs because it was a lump sum. Now we can say If you go from 500 PCs to 350 PCs, you will

save yourself $200+dollars per month per PC. We have taken a chunk out of the baseline cost andsubstituted it with a variable cost. As long as the total number of PCs in the organisation does not fallbelow 1200, these costs will be valid.

The costs of discretionary projects were charged directly to the businesses to which they related.

Central, however, was still attempting to develop equitable ways of charging the baseline cost to each

business. This activity involved determining the drivers of the cost of discretionary projects. A seniorfinance manager outlined the difficulties:

We had a sub-committee look at it in several ways, because you are not sure what the cost drivers are.

The helpdesk is one of the more expensive items. They decided to just divide the cost equally acrossthe four businesses and holding company. Which means the holding company pays too much. We have30 staff and the Contracting business has 1200 staff. We have tried to find a reasonable basis, and

sometimes that was just dividing it equally across all businesses, and that will remain until we have abetter understanding of what drives the costs.

As explained in a previous section, each business was required to submit plans of its IT requirements atthe start of each year, for approval. This provided control over costs as well as over the ad hoc development

of projects that had once existed. Having IT&T managed by an arms-length provider, which itself hadstrict, bureaucratic systems in place, made the implementation of spending controls simpler and supported

Centrals attempts to control spending. Control over ongoing costs was also achieved through the monthlyIT charges being managed by the IT&T Outsourcing Manager. All charges were reviewed and checkedand only charges for the baseline and approved projects were paid.

Once Global was established as the IT&T provider, more systematic controls over new developmentswere implemented. Businesses were required to present justifications for all IT capital projects to theworking group. A project would only be approved if there was a strong justification and if it met the needs

across the business.

3.6. Summary and conclusion

The outsourcing of IT&T was seen by managers at Central as an effective response to the problems thatexisted within its IT&T function, and these were essentially problems related to cost control and control

of IT planning and operations. It also provided a means for gaining access to world-class IT&T expertise,which was essential for competing in the new contestable electricity markets. These problems were

essentially solved through the outsourcing of IT&T to Global. However, the relationship was consideredto be very difficult in thefirst eighteen months. The initial contract provided only a broad framework,whichdeveloped over time to incorporate more specific provisions of specific responsibilities and performance

targets, through processes of negotiation and communication with the outsourcer. While some managers


18/27


saw the incomplete contract as a source of conflict that could have been avoided, this was not a universal

view. A strategic alliance with a cable supplier was undertaken two years after commencement of theIT&T contract. This agreement was undertaken without a full written contractsimply a heads of

agreementas the managers negotiating the agreement had seen the problems that arose with the IT&Tcontract. There was also a belief among some managers that a contract could be a source of mistrustbetween parties. This new relationship was to be characterised by an open book approach to the sharing

of information. Performance was to be tied to risk-return, and a steering committee consisting of managersfrom both firms was to develop performance indicators to support areas such as growth, delivery timeand cost. A manager described how the new relationships would proceed, compared to the IT&T project.

Its virtually the same selection process, but were going to implement it in a different way. Were not

going to tie it to a legal contract; we will try and build it on trust from day one. Now we could fall flaton our face but some experiences that Ive had suggests the legal agreement does introduce a senseof mistrust from day one, because all the is have got to be dotted and the ts have got to be crossed.

I have to admit our lawyers are not exactly ecstatic about this direction but it have accepted it with afew clauses that weve had to put in to keep them happy. But what were saying is that were not goingto be tied to a legal agreement. This thing we want to have is an arrangement and partnership built on

trust.

4. Analysis and discussion

In this section, we consider the characteristics of the transaction (task programmability, output mea-surability, asset specificity and frequency of the transaction), the transaction environment and the parties,

and assess whether the type of control mechanisms in place and the role of trust in achieving controlmatches any of the three control pattern presented in Table 1.

4.1. The inter-firm control model

4.1.1. Characteristics of the transaction

In general, it was not easy to achieve a high level oftask programmability at Central. High levels ofuncertainty arose from the complexity of the technology, the range of the IT tasks, and the ambiguity

of the IT market. While there were some routine IT tasks involved in the outsourcing contract, it wasstill difficult to specify ex ante procedures and processes needed to ensure success. The high levels of

uncertainty made it difficult to develop specific contract provisions prior to engaging the outsourcer, andin the early days of the relationship.The level ofoutput measurability was low initially, but increased over time as Central worked with

Global to assess the quality of the delivered outcomes of the outsourcer. This was possible due to the high

level of IT expertise that was retained within the electricity firm, which was developed further after theIT functions were outsourced, as well as through the mechanisms that were used to increase information

sharing during the execution of the contract. In the case of more discrete IT projects (such as the customerservice system), there was a lower level of uncertainty and output measures were pre-specified. Overtime, both parties to the relationship gained some experience in specifying output measures and targets,

as evidenced by the development of the risk-reward scheme.


19/27


The level ofasset specificityis a complex issue to assess. Ownership of the physical IT assets and thesoftware were retainedby Central, which managers believed safeguarded them from being over-dependenton the outsourcers. However, the granting of exclusivity to Global created a high level of dependency

on that one supplier. Global developed considerable levels of knowledge, skills and experience thatwere tailored to Centrals needs to allow them to execute their contract effectively (high human assetspecificity). Also, the outsourcers were located on-site (high site specificity). The two parties to the

relationship invested considerable time in establishing good working relationships with individuals andgroups in either party. Switching costs were highany withdrawal of service by the outsourcer would

have had a devastating impact on the ongoing operations of either party. Thus, the relationship wascharacterised by high asset specificity. Under a bureaucratic pattern of control, dysfunctional behaviourarising from high asset specificity would be protected by contractual rules, while under a trust based

pattern, high levels of trust would mitigate any opportunism or power plays resulting from high assetspecificity.

The transactionsincluded in the outsourcing relationship consisted of a mix of repetitive transactions,and one-off projects. This mix of transactions increased the level of uncertainty and contributed to lowtask programmability and low output measurability.

Thus, in the early days of the relationship, the characteristics of the transaction indicate that the out-sourcing relationship between Central and Global largely following a trust based pattern; low task pro-grammability, low levels of output measurability that increased over time, high asset specificity, and a

mix of low (and high) repetitiveness of transactions. However, as task programmability and output mea-surability were gradually becoming higher toward the end of the research period, this could signify amove towards a more bureaucratic pattern of control.

4.1.2. The characteristics of the environment

The environment within which the outsourcing relationship operated was one of high risk and highuncertainty, particular in the early days of the contract. Future developments in IT technology were dif-ficult to foresee and forecast with any accuracy, and a changing and uncertain competitive electricity

environment made it difficult to predict competitor and customer actions. Deregulation of the electricityindustry resulted in increased competition as well as increased opportunities for Central. However, theimplications of these changes for the future business of Central were unclear. The high risk and un-

certain setting also led to the relationship placing greater reliance on social embeddedness, through thedevelopment of competence trust and goodwill trust as means of control (this will be discussed further

in a later section of this paper). In summary, the characteristics of the environment within which theoutsourcing relationship developed are those associated with a trust based control pattern, as outlined

inTable 1.

4.1.3. The characteristics of the parties

A high level of competence, outstanding reputation and extensive experience as an outsourcer were

highly important characteristics that Central sought in their outsourcer. Due to the high level of risk anduncertainty in the outsourcing activity, Central could not risk engaging an outsourcer that did not have

such qualities. While risk sharing and low information asymmetry did not seem to be evident in the earlierdays of the relationship, as the relationship matured, Central came to share more information with Global,and through the risk-return system the sharing of risk was explicitly encouraged. Thus, theparties to therelationship came to display characteristics closely associated with those of a trust based pattern.


20/27


4.1.4. Summary

Thus, the three contingent factors of the characteristics of the transaction, environment and the partiespoint towards a trust based pattern of control. This is particularly the case in the early months of the

relationship. However, as the relationship matured, the levels of task programmability and outcomemeasurability increased through joint activities and the increased knowledge that developed between thetwo parties. This is more characteristic of a bureaucracy based pattern.

4.2. Control mechanisms

Trust based patterns of control are characterised by outcome and social controls that develop over

time. There is no reliance on behaviour controls, and trust forms the prime means for achieving control.Reliance on formal outcome controls is low at the start of the relationship, but increases over time. Therelationship is characterised by relatively unspecific general thrust contracts, performance assessment

based on broad emergent standards, and information sharing due to the participatory, interactive natureof the process of contract execution. In this section, we examine these control mechanisms and the role

of trust in achieving control in the case study.

4.2.1. The contract

Poorly specified contracts have been cited as a major reason for the failure of some outsourcing rela-

tionships, particularly in the IT area (Robertson, 1998; Domberger, 1998). However, there are situationswhere high levels of uncertainty make complete contracts unattainable. Some managers at Central be-

lieved that the lack of detail in the contract specifications was a barrier to establishing smooth relationshipswith the outsourcer, particularly at the start of the relationship. These problems related to the ambiguousspecification of the baseline services that needed to be maintained to achieve the fixed payment, and the

precise cost that the outsourcers would charge Central for those baseline services. There was also anabsence of performance measures and targets, which were then negotiated over the first few years of thecontract. Central prepared their contract in haste and this clearly led to dissatisfaction among managers

with the lack of precision in the provisions of the contract.Being risk adverse, it is understandable that managers may desire contracts to be tightly written,

particularly given the critical nature of the function that was outsourced and the size of the contract.

However, the complexity and uncertainty encompassed in the outsourced IT tasks made it unrealistic toexpect that all important aspects of the operation and relationship could be specified and incorporated

into a contract (Spekl, 2001).In these cases, important aspects of the relationships can only evolve overtime (Das and Teng, 2001b).As the relationship develops, partners may come to recognise specific needs

and requirements, which were difficult to anticipate at the start of the relationship. In the first eighteenmonths of the relationship, Central was able to work with Global to resolve perceived inadequacies inthe initial contracts, as both parties together improved their knowledge of the tasks and their execution.More sophisticated provisions were incorporated into the contract to cover areas such as performance

measurement, occupational health and safety standards, better assigned responsibility for costs, andspecifications for the sharing of cost improvements.

Some managers at Central saw the incomplete contract as a limitation and a source on conflict thatimpacted unfavourably on the relationship with Global. However, it could be argued that even if more timehad been spent on formulating the contract, given the high levels of task uncertainty and low outcome

measurability encompassed in the delivery of the base IT services, the contract may still have been


21/27


incomplete. While the conflict and negotiations with Global may have been regarded as time consuming

and dysfunctional, some managers acknowledged that this was an important way of developing sharedunderstandings and objectives. The move towards the formulation of the new strategic alliance based on

the heads of agreement was recognition by some managers that in some situations contracts cannot beeasily formulated and an attempt to do this may lead to initial mistrust.

4.2.2. Performance measurement

It has been suggested that a firm may gain control over the function being outsourced through ongoing

monitoring of work performance, as well as monitoring aspects of the outsourcing relationship throughoutthe contract period (McFarlan and Nolan, 1995).This may be achieved using performance measures andbenchmarks, which focus on areas such as customer satisfaction, delivery responsiveness, product quality

and cost, and either included within the contract, or negotiated at a later date.High levels of uncertainty in the specification of the outsourced function made initial performance

standards difficult to specify. Through numerous discussions and meetings between managers at Centraland Global, performance measures and targets were developed to provide the most appropriate incentivesfor the outsourcer to deliver quality services. A risk-reward scheme was introduced to encourage theoutsourcer to achieve more profits when undertaking a new discretionary project, while also delivering

cost savings to Central. The Central case emphasises the way in which incentive systems can be structuredto benefit both parties in the relationship, effectively encouraging the developing of shared goals, and

encouraged goodwill trust. Thus, the two parties were able to develop outcome controls as their sharedknowledge of the processes increased over time.

4.2.3. Information sharing through participatory, interactive processes

There were many activities undertaken by managers at Central and Global that involved participatory

processes and knowledge sharing.Key issues in establishing protocols for effective communications between the firm and its outsourcer

included the number of points of contact within the firm, and the regularity of formal meetings. An impor-

tant issue for managers at Central was establishing the point of contact to manage the relationship. Initiallythe IT&T Outsourcing Manager handled day-to-day problems that emerged between the outsourcer andthe various businesses of Central. One manager at Central suggested that in managing the transition, this

manager should have focused solely on establishing the relationship, leaving the day-to-day running ofthe process to subordinates. Regular meetings were held between managers and the outsourcers to review

performance and to discuss future plans.Establishing formal communication protocols are important in the early days of the relationship with

an outsourcer, when there are incomplete contracts, to establish the ground rules and expectations of eachparty. This may involve single or multiple contact points, depending on the characteristics of the functionthat is being delivered by the outsourcer and the complexity or critical nature of the service.

The mechanisms that Central and Global used to develop performance indicators also became an

important part of the participatory information sharing processes. First, the processes provided a forumfor interaction between the two parties, increasing the number of joint dealings and increasing familiarity.

The performance indicators also provided an efficient form of communication of expectations betweenthe two parties (Jarillo, 1988). Third, the design of the risk-reward system allowed both parties to sharein the rewards when new discretionary IT&T projects met expectations of effectiveness and efficiency,

thus, contributing to the strengthening of the relationship.


22/27


4.2.4. Summary

The means by which Central used outcome and social controls to manage the outsourcing relationshipare largely consistent with the requirements of a trust based pattern of controlrelying on outcome and

socials controls, broad non-specific contracts that develop time, performance assessed through broademergent standards and high levels of information sharing and communications. This compares with abureaucratic based pattern that relies on outcome and behaviour controls, which include comprehensiveselection criteria and formal bidding, rigid performance targets, comprehensive rules of behaviour and

detailed contracts. However, there was reliance on formal bidding procedures to award the outsourcingcontract, and as the relationship progressed, there was a clearer specification of performance targets and

of the contract itself.

4.3. Building high levels of trust

Trust is vital in achieving control in a trust based pattern of control, but is of lesser importance ina bureaucratic based pattern, and has no relevance in a market based pattern. In a trust based controlpattern, key forms of trust that work with control systems to reduce risk in outsourcing relationships are

contractual trust, competence trust and goodwill trust. Some levels of contractual trust and competencetrust are essential for both bureaucratic and trust based patterns, to allow the initial selection of theoutsourcer and for some form of contract to proceed. In a trust based pattern competence trust will

develop further over time and form a part of the control framework. Goodwill trust is difficult to achievein an outsourcing relationship (Das and Teng, 2001b), but is essential as it reduces the likelihood of

opportunistic behaviour. We have argued that there was a high dependency between Central and Global,due to the exclusive nature of the supply contract, and the strategic nature of the information that wasshared between Central and Global. Goodwill trust will work with outcome controls and social controls

to reduce risk (Das and Teng, 2001a).

4.3.1. Building competence trust

There are several mechanisms that increase competence trust in an outsourcing relationship, which in

combination with outcome controls and social controls will reduce risk. The major competence-buildingmechanism is proactive information collection (Das and Teng, 2001a). Central was careful to ensurethat the outsourcing company that they engaged had a strong reputation in the IT industry and strong

technical competence. Global was perceived by managers at Central as having high credibility, status andcompetence in the IT industry. On close examination of the interview transcripts it seems that the managersat Central made limited reference to the technical competence of the chosen outsourcercompetence

trust was not considered a concern with Global. However, developing high levels of goodwill trust was aconscious strategy followed at Central.

4.3.2. Building goodwill trust

It is clear from the quotations in the case studies that the outsourcing relationship commenced with alow level of goodwill trust. Some managers at Central believed that this was a function of starting with an

incomplete contract and conflicting priorities. It could also have be a function of the gap in knowledge,insights and control that can appear immediately following the outsourcing of a key function (Mouritsenet al., 2001). In addition, there were distinct differences in the cultures of the two organisations and this

made initial communications difficult. Managers were also very conscious of the need to develop trust


23/27


24/27


of perspectives and approaches, and is of particular importance in situations where contracts are not

precisely specified (Ring and Van de Ven, 1994).

4.3.3. SummaryEssential to achieving control in the relationship between Central and Global was the development of

trust. Unlike a bureaucratic based pattern or a market based pattern, control was difficult to achieve in thisrelationship as the difficulties in task programmability and output measurability meant that formal controlswere difficult to implement, particular in the early stages of the relationship. Trust, particularly goodwill

trust, played a role in working with control mechanisms to ensure control, and in allowing the operationsto proceed in the absence of tightly specified rules. Interestingly, goodwill trust continued to exist, andmay even have strengthened, in the face of the development of more rigid performance expectations and

the developing contract specifications. This runs contrary to the arguments of some prior research, whichhas argued that formal accounting and contracting may conflict with the development of trust (see, for

example, Seal and Vincent-Jones, 1997; Seal et al., 1999). The development of goodwill trust also alloweda relationship where there was high asset specificity, and hence high interdependencies and high risk ofopportunism, to continue and to remain under control.

5. Conclusion

This paper focuses on a case study of an outsourcing relationship, and how control mechanisms and trust

worked together to achieve control. The characteristics of the outsourcing transactions seemed to meetthe requirements of a trust based pattern of control, and control was achieved through the developmentof outcome controls and social controls and through the development of trust, particularly goodwill trust.

This paper contributes to the literature in several different ways. First, this paper draws on the threepattern of control specified by van der Meer-Kooistra and Vosselman (2000), to add to the growingknowledge of control systems in new organisational forms. Second, focusing on the three contingent

factorscharacteristics of the transaction, environment and the partiesto determine the form of controlpattern highlighted how the control pattern may gradually change over the life of a relationship, as thosecontingent factors change. In the case study, as the performance standards became more specified and

as the contract developed, the trust based pattern of control began to move towards a bureaucratic formof control. Finally, the paper provides some evidence that the development of trust may be compatible

with the development of tighter accounting controls and contracts, if trust is already well-established andthose controls develop in a supportive and cooperative manner, involving both parties.

The findings of this study are of interest to managers who enter outsourcing relationships, and toresearchers who endeavour to understand the nature of control systems and trust associated with neworganisational forms. However, the limitations of the study must be acknowledged, and future researchdirections considered. First, a key aspect associated with outsourcing is risk (Das and T

tugas review spm

Documents