Changes and change management in construction and IT projects

Automation in Construction 20 (2011) 767–775

Contents lists available at ScienceDirect

Automation in Construction j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a u t c o n

Changes and change management in construction and IT projects Jan Bröchner ⁎, Ulrika Badenfelt Department of Technology Management and Economics, Chalmers University of Technology, SE-412 96 Göteborg, Sweden

a r t i c l e

i n f o

Article history: Accepted 8 January 2011 Available online 5 February 2011 Keywords: Change management Project management Industry comparisons

a b s t r a c t Modelling for change management in construction might benefit from experiences from other industries. This study compares reasons for change as well as change practices in interorganizational relationships in construction and information technology. Sixteen contractual relationships have been studied through a questionnaire and interview survey conducted among both clients and providers. Project-type relations are contrasted with more continuous service support contracts. More than their providers, clients tend to ascribe high change frequencies to an increase in technical understanding during projects. Early provider participation is associated with lower change frequencies, while the strength of incentives is less influential. The ease of transforming changes into new projects varies between technologies. Irreversibility is important, as well as relations between pre-investigations and the successive revealing of initial conditions, together with the life-cycle of formal and informal communication. Findings can be used for transfer of agile practices. © 2011 Elsevier B.V. All rights reserved.

1. Introduction How changes in construction projects originate, are negotiated and resolved has attracted many investigators. A growing number of empirical studies of why and how contracts are renegotiated throw light on the reasons for change and on the patterns of change management in individual projects. Researchers increasingly see a potential for model-based systems to make the handling of project changes more efficient [1–4]. A successful design of change models needs a basis for understanding why changes arise and current ways of managing change. Better insight into the relations between underlying technologies and changes should support the interpretation and transfer of change management procedures between different types of projects and between different industries. The project typology developed by Shenhar and Dvir [5] and Shenhar [6] partly relies on degrees of technological uncertainty associated with a low-technology to high technology scale, where construction belongs to the low end of the scale and the software industry at the other end, but this simplification might not be efficient in a project change perspective. In their proposal for modelling building projects, Isaac and Navon [4] argue for learning from how requirement traceability is practiced in software development projects. However, there is more to be learnt from a comparison of IT and construction. Construction projects usually result in goods that are immobile and durable, and this is reflected in their typical production processes. There is an issue of hidden faults that might be discovered only after several years of building use.

⁎ Corresponding author. Tel.: +46 31 772 5492. E-mail address: [email protected] (J. Bröchner). 0926-5805/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.autcon.2011.01.001

Construction technology changes slowly over the years. There are significant environmental consequences of most construction projects. The organization of production is typically based on numerous and specialized subcontractors, who might work together for the first time. Carried out outdoors, construction projects are subject to many types of external disruptions. IT projects, on the other hand, have results that often are highly and globally mobile even if tailored to a specific and localized use. There are similarities with construction: many software projects leave durable results, and some hidden defects may take time before they emerge as evident to users. However, information technologies develop rapidly; access to novel technological solutions can be a reason for the existence of a software project, and sometimes, technology shifts so quickly that it presses for change in ongoing projects. An increased depth of the industry comparison of what different technologies imply can be achieved by having a control population of corresponding business services within the two industries. Since parties to project contracts may expect to meet in future projects, the difference between project contracts and support service contracts, as well as the corresponding models for change management, should not be exaggerated. Therefore, in the present investigation, the aim has been to compare change management (reasons for change and change practices) in contractual relationships in construction and information technology, relying on pairwise interviews with both clients and providers. To support the analysis, there has been an equal number of parallel interviews in contractual relationships for the corresponding business services: management of built facilities, in order to match the construction projects, and IT support services to match the software projects investigated. The structure of this paper is as follows. In Section 2, the literature is reviewed, locating the interesting gaps in knowledge and

768

J. Bröchner, U. Badenfelt / Automation in Construction 20 (2011) 767–775

identifying the questions that led to our survey design. The survey method, which generated a set of sixteen contractual relationships, is described and explained in Section 3. Here, our first empirical step is to study relative frequencies of various reasons for changes in contracts. In the following step, the procedures used by clients and contractors are scrutinized. Results are presented in Section 4. Next, these findings are discussed against the background of earlier research, highlighting the dynamic issues raised by the survey results. Conclusions, both as to promising ideas for research and implications of a more practical nature, are drawn in Section 6. 2. Literature review The review of the literature begins with studies of project change in particular industry or sector contexts, and then studies of reasons for project change and how these reasons can be classified are brought in. Some investigators have looked at the dynamics of contractual relationships over the life cycle, and there is a stream of literature that explores the importance of the degree of complexity in contracts. The existence and consequences of incentive elements in project contracts have been analysed by many researchers. Fewer publications report studies of capabilities and their influence both on initial specification of project requirements and on subsequent contract renegotiations. 2.1. Industry contexts Surveying the literature on change and flexibility, there is a number of studies of construction projects with a focus on flexibility and change [7–11]. In a more normative vein, Ibbs et al. [12] have pioneered a systematic approach for managing project change, stressing the communication and documentation of critical project success factors. There are slightly fewer useful studies of change that include or exclusively deal with IT projects [13–15], although the emerging literature on ‘agile’ software development is relevant here [16,17]; e.g., the agility taxonomy proposed by Conboy [18]. However, there are many empirical studies that concern IT service contracts [19–28]. Contractual change in facilities management has drawn attention from fewer researchers [29–31]. Industry comparisons with a focus on project changes are unusual. Dvir and Lechler [13] included construction and software projects together with machine tool manufacturing and a few other project types in their study of how changes affect project success, but did not compare results and discuss how underlying technologies lead to different results. A recent survey, albeit not primarily with a focus on changes, fails to show significant overall differences in how people perceive project management performance when viewing construction against a broader background [32]. Olsson [33] included five types of projects, dominated by construction, whereas Barthélemy and Quélin [34] investigated outsourcing in five industries; in their statistical analysis, a simple 1/0 manufacturing/services dummy control variable failed to reach significance. Nevertheless, there are a few project analyses which have taken industry-specific production systems as their starting point, and Chang and Ive [35] have done so for their study of bargaining power in construction projects, introducing the concept of ‘process specificity’. If there are technology-based differences between styles in specifying requirements, in choosing contractual incentives and in renegotiating, it should be possible to argue that these differences reflect specific needs for change and change management models. 2.2. Reasons for changes A number of authors have suggested change typologies, as opposed to project typologies. In their knowledge-based decision support system for management of construction variation orders, Arain and Pheng [1] identify types of control that they link to variation causes, with an

emphasis on design-related issues. Motawa et al. [3] argue for the necessity of including change causes in change identification as an element in an integrated system for construction change management. Taking a broader view of projects, Dvir and Lechler [13] distinguish between plan changes and goal changes, where plan-changes are by definition lacking impact on project goals and are typically induced by the project environment. Hsieh et al. [36] studied public works projects and they split the causes of change orders into technical and administrative. A recent overview of earlier investigations of change in construction projects suggests that there are three main types of change causes: external, organizational and project internal causes [11]. Six types of unexpected events in defence-related projects were identified by Geraldi et al. [37]. In a classification to be used for a wide range of both project contracts and service contracts, it seems fruitful to distinguish between client origins, supplier origins and external origins of reasons for changes. Within these three major groups, further categories can be identified. The first category of changes originating with the client is due to high initial search or information costs for the client; as a consequence, and unintentionally, the original contract documentation from the client may contain errors, omissions and contradictions in specifications [29,36]. The second category is when the client finds that the primary business to be supported by the project or service contract has changed its needs in a way unforeseen in the original contract [13–15,38–41]. This category is often referred to as scope issues. In an IT case study with incentives in the contract, the parties faced a problem of determining whether scope changes implied ‘new’ features [42]; in this case, the authors distinguished between three categories of scope increases: fixes, clarification and enhancements. Their first category, ‘fixes’, are close to ‘dealing with errors in original documentation’. ‘Clarification’ is close to improved core understanding, and ‘enhancements’ appear to combine features from core change and improved core understanding. A supplier related category of reasons for change is when the provider begins to argue that the client's primary business would be supported better by another set of specifications or contractual regulations during the execution of a project or a service contract [15]. Another supplier change category is due to changes in delivery technology that were wholly or partly unforeseen when the contract was signed, rendering the original choices of technology obsolete [21,22,43]. A third supply side category is instantiated when a supplier to the primary supplier contacts the client directly and proposes a change of technology. In addition, there is a category of changes that have a joint client and supplier origin: as when a provider discovers during the contractual period that the state-of-nature or the technical state of the initial system is not as expected [14,36], and when both parties find that the mutual communication routines set up for the duration of the contract are inefficient [19,20,22]. The third main group of change categories refers to those arising from causes that are external to the parties to the contract [36,38]. External causes are thus a narrower concept than ‘environmental uncertainty’, as used by Barthélemy and Quélin [34] in their study of outsourcing in several industries, and who consider uncertainty from the viewpoint of the provider and thus also include all client-initiated changes in the environment of the contract. External causes can be of natural origin or societal. So there may be unexpected natural events, although many such risks are routinely allocated to the client in a force majeure clause of the contract and thus in most cases will not lead to any contractual changes. Finally, there are the external disruptions of societal or human origin, such as unexpected legal changes, political turmoil and labour unrest. Only a few studies have thrown light on how various reasons for change are related to each other. This is an important complication for model based change management. The Reuer and Ariño investigation [38] based on a contractual renegotiation model for strategic alliances in a range of Spanish industries relied on two categories of ex-post

J. Bröchner, U. Badenfelt / Automation in Construction 20 (2011) 767–775

contingencies and then found environmental change to be less significant as a renegotiation antecedent than strategic change. Again, Barthélemy and Quélin [34] looked at four types of reasons for change due to ‘environmental uncertainty’ related to difficulties in assessing outsourcing clients' future needs: technology, activity level, performance level and vendor skills. They concluded that greater uncertainty about the client's future needs went along with more elaborate contract clauses to deal with unexpected contingencies. This is clearly related to contract complexity (Section 2.4), and the study by Mayer and Bercovitz [26] of an IT software and hardware provider looked at contingency planning clauses in contracts, listing changes in technology, input prices, government regulations, and product requirements may stand as examples of what transacting parties often plan for. Summing up, the present investigation operates with nine categories of reasons for changes. Two of these reasons originate with clients: errors, omissions and contradictions in the original specification documents; and changes in clients' core activities (or in demand from their customers). There are three reasons that can be associated with the supply side: better understanding of client needs, changes in provider technology, and changes suggested to the client by a subcontractor with technology change (supplier to the provider). There are also two reasons for change with a joint client and provider origin: the ex ante unknown state of the system (IT, geotechnical conditions, etc.), and lack of communication between parties during the contract period. Finally, two types of external reasons for changes related to environmental uncertainty should be distinguished: unexpected natural events and unexpected societal influences on the project. 2.3. Dynamics of contractual relationships The dynamics of contractual relationships are both those which arise throughout a chain of contracts between the same parties [44,45] and processes that occur within the time frame of an individual contract [41,46]. Taking a life cycle view of the process of the individual contract, the organization of the pre-contractual requirements definition stage and in particular the involvement of future users of the project or service affects the future climate for managing contract changes, even if not entirely eliminating the need for changes [22,30,47]. Very often, the importance of contract history, the phenomenon of recurring contracts with the same pair of companies has been brought up and sometimes been the chief focus of an investigation. Thus Crocker and Reynolds [48] found that successive agreements between two parties tend to become more complete over time, although it is dubious whether this observation is applicable generally [25]. More detailed analysis of how complexity develops over time indicates that there are several factors to be taken into account [14,19,21,23,26,27,30,38,49,50]. Already in the 1960s, the fundamental contrast between the waterfall model for software development projects, with its emphasis on having thorough, correct and clear specifications before development, and iterative models with incremental development, often referred to as agile models, was recognized [17,51]. This contrast has been less readily identified for construction projects, which regularly show waterfall ambitions, although the subsequent work patterns often do contain iterative elements. With a waterfall type of development process, needs for change are likely to emerge during the contract period [41], but one should not assume that iterative and incremental processes are set up in a manner that precludes contractual change entirely. The effects of learning have been analysed in order to understand the dynamics of contractual relations in the personal computer industry, an industry characterized by high technological uncertainty. By studying how contracts between a provider of customized programming and a hardware producer evolved over a nine-year period, it was found that a succession of contracts was developed incrementally in response to actual problems experienced, to prevent similar disputes from arising in the future [14]. The contracts grew in length, the administrative sections more than the technical requirements; they could be viewed as

769

knowledge repositories. Furthermore, contingency planning and task description have been seen to behave as complements in contractual design [25].

2.4. Complexity of the contract Changes and change management in contractual relationships can be assumed to be influenced by the degree of tailoring of a contract to a particular context and also by the detail and complexity of the contract. Contingencies can be dealt with by having precise specification type (substantive) clauses, or there may be procedural clauses that prescribe the renegotiation processes and participants when managing particular changes. While Poppo and Zenger [21] measured contractual complexity by asking informants only whether they considered the formal contract to be highly customized and having required considerable legal work, Reuer and Ariño [50] quantified complexity as the weighted sum of the presence in a contract of eight types of provisions, which they were able to reduce to two main factors of contractual safeguards: enforcement provisions and coordination provisions. In the survey by Barthélemy and Quélin [34], a significant relation between contract complexity and ex post transaction costs was claimed after analysis of questionnaire responses from five industries with manufacturing and service outsourcing. Complexity was measured with simple scales for features of clauses related to control, incentives, price, flexibility and end of contract. One strategy for reducing complexity of contracting is to rely on standard forms of contracts [14,48]. For construction project contracts, there is a long tradition of using standard forms [15]; this is a more recent phenomenon for facilities management contracts and less widespread there [31]. For facilities management contracts, an analysis based on transaction cost economics found that a lack of standard contracts contributed to disputes; the contractual relationships concerned Hong Kong commercial buildings and were of a spot, lump-sum type [29]. Thus frequent subjects of dispute originated in the scope of the work, workmanship, and noise, nuisance, pollutions and interference, while few contracts had provisions for contingencies. These findings could be interpreted as symptoms of an immature market for building-related support services, as opposed to construction projects, where there is a tradition of standardization and reliance on widely known standard contracts, which reduces the complexity of at least the relationship [7]. However, Sweden is an early adopter of standardized contracts and specifications in the field of facilities management [52]. For IT outsourcing contracts, it is usual to have a framework agreement with general terms and conditions and then service level agreements for particular services [22]. Service level agreements, primarily for IT outsourcing, use various methods for defining performance targets and their measurement [47]. In recent years, it is probable that IT service contracts have been influenced by increasingly standardized routines for change management [28]. Poppo and Zenger [21] in their study of information systems outsourcing proposed a plausible linkage: contract complexity, understood as customization of contracts, increases relational governance, which in its turn increases exchange performance, measured by satisfaction with cost, quality and vendor responsiveness for the service in question. The issue of change frequencies has been raised by Arain and Pheng [2] as part of what they refer to as trending in their model for variation management in building projects. Both the rate of variations (=design changes) per month and total cost of variations appear to increase with construction project size, complexity and duration [8], and also for IT contracts, size of the contract has been seen as a variable that influences change behaviour [24]. In IT development projects, there are strategies that simultaneously reduce and increase complexity, by modularizing software production so that the resulting models can be produced and tested separately; but also here, as

770

J. Bröchner, U. Badenfelt / Automation in Construction 20 (2011) 767–775

Abrahamsson et al. [16] have pointed out, there seems to be an upper limit to project and team size for agile methods. 2.5. Incentive elements The influence of various types of contract incentives has been studied or at least taken into account in many studies [8,10,19,23,25, 29,31,34,48,49,53–55], but the relation between change management models and incentive mechanisms in contracts remains to be developed. Without an understanding of why a particular form of contract incentive was selected, the mechanism of its influence on change management could be misunderstood. Badenfelt [55] noted the importance of change-induced target cost adjustments in relationships based on target cost contracts. Details related to the incentive systems in project or service contracts are also prominent in a few other studies [20,33,56]. For agile software development in the context of target-cost pricing, there is the case study by Eckfeldt et al. [42], where it emerges as crucial that both contractual parties anticipated considerable changes in scope and therefore chose target-cost together with collaboration, in order to share risks and incentives. 2.6. Capabilities, from specification to renegotiation Many types of capabilities in project management are recognized [57–60]. A few capabilities appear to be particularly relevant in the present context. In general, the individual experience of managers in the relevant field has been acknowledged by earlier investigators as a background variable that has to be taken into account in studies of change [21,38]. A number of studies have looked into the types of discussions and negotiations that arise when change suggestions are made [7,9,21,29–31,38,53,55]. This corresponds to a special type of capabilities to deal with contract complexity. Several investigators have argued more precisely that the wording of change management clauses in the contract is influential [14,21,22,25,27,29,34,38,40,53]. A further topic is that of how the parties react to cost change clauses and in particular depending on who is to carry the cost for a particular change [10,14,25,29]. Taking a life cycle view of projects, there is the question of how to mobilize which capabilities in contract design [61] and in the early stages of requirement formulation. 3. Methodology Since it can be suspected that there is an element of social construction when people understand and describe their relations under contracts [62], it is an obvious advantage to have data and opinions from both parties to the contract. Earlier large questionnaire surveys have suffered from being one-sided, usually because they have relied only on responses from providers (and often with low response rates, which call into question the representativeness of findings). In the present situation, each interviewee was asked to assess the frequency of (each of the nine) change reasons on a fivedegree scale, between ‘never’ (=0) and ‘very often’ (=4) during the contractual relationship, before responding to a set of semi-structured questions. Thus data for this investigation were collected in 2008 and 2009 in Sweden through a total of 32 face-to-face semi-structured interviews with practitioners, preceded in the interview context by filling out a questionnaire with both categorized answers to six background questions and nine 0–4 Likert scales for indicating the frequency of occurrence for each of the nine reasons for change (a–i) already identified in Section 2.2: a) errors, omissions and contradictions in the original specification documents b) changes in clients' core activities (or in demand from their customers)

c) better understanding of client needs d) changes in provider technology e) changes suggested to client by a subcontractor with technology change (supplier to the provider) f) ex ante unknown state of the system (IT, geotechnical conditions, etc.) g) lack of communication between parties during the contract period h) unexpected natural events i) unexpected societal influences. Then there were eight semi-structured questions: 1. Who participated in the initial process of formulating requirements? 2. Which were the reasons behind the choice of a particular pricing formula? 3. Which incentives or penalties are used, if any? 4. How are contractual clauses for managing and pricing changes formulated? 5. Is it always clear how change clauses should be interpreted and applied, e.g. who is to carry the cost for a particular change? 6. Which is the origin of these change clauses? 7. Which discussions tend to arise between the parties when a change has been suggested? 8. When are change costs settled? At the end of interviews, respondents were encouraged to reflect on what they saw as important in general in contractual relationships of the type selected. A pilot version of the questionnaire was tested on two firms and afterwards modified and expanded. The main objects of investigation were four IT software project contracts, typically for e.g. development of business systems, system modules and similar types of software, including migration of functions to a new platform and four construction project contracts. In order to check for the relative strength of underlying technology influences, there were also four IT support service contracts and four facilities management contracts (services for built facilities). All sixteen contracts were required to have left the initial stage, so that the relationship between the parties would have been stabilized, during a period of at least six months after inception. Also, the client should be a recurrent client in the market, implying a certain level of experience and capability. Relying on publicly available web information and general industry knowledge, ongoing contracts and their contact persons were found. E-mails were sent to these contact persons, and in cases of no reply, telephone calls were made after a few days. In a few cases, access to IT contracts was denied; reasons given were mostly that sensitive negotiations were ongoing, and only occasionally, participation in the survey was declined because of ‘lack of time’ or because of a policy rooted in potential legal consequences, which occurred once with an international provider of support services. For some contracts, the client initiated provider contacts, and in other cases, it was inversely the provider who suggested a client representative. Obviously, these procedures would tend to exclude difficult and dysfunctional contractual relationships, thus introducing bias in favour of more harmonious projects. Once the contracts were identified, separate interviews were booked with a provider/contractor representative and a client representative. Each interview lasted about one hour. The questionnaire was distributed in advance to respondents who explicitly asked for it, which occurred in about half the cases. 4. Survey results After characterizing the contractual relationships according to the background data provided by respondents, the relative frequencies of reasons for changes are given, followed by what the semi-structured interviews revealed. The general reflections expressed at the end of interviews are also reported here.

J. Bröchner, U. Badenfelt / Automation in Construction 20 (2011) 767–775

4.1. Background to the contractual relationships Contract sizes, including subcontractor efforts, ranged from a few man years to more than fifty man years. A clear majority of clients were in the private sector, but for each of the four contract types, there were one or two public clients. It was found that respondent experience of similar contractual relationships was around ten years, with a few exceptions among the four IT software projects, where personal experience was shorter, within a one to five year range. Contract sizes, as measured in man year equivalents, including subcontractor manpower inputs, were usually in the range of ten to fifty man years, here again with the exception of a few software projects that were smaller. For all eight service support contracts, there was at last one earlier contractual relation between the provider and the client, but for the software and construction projects, there were examples of this being the first contractual relationship that the provider and the client had entered with each other. For each of the four types of contract, there was at least one example of an obvious incentive element in the contract, and again for all four types, there was at least one relationship with no elements of a bonus or other positive incentives. Further, for each contract type, there was at least one contract that referred to a standard form of contract; the rest were based on in-house forms.

4.2. Reasons for changes The categories of reasons for renegotiating and revising original contractual regulations and requirements, as explained in Section 2.2, occur with varying frequencies depending on the category and the type of contract, as shown in Table 1 for all sixteen contractual relationships. When comparing frequencies of reasons for contractual changes, it is possible to see that ‘increased understanding of client core activities’ and ‘increased technical understanding of prior systems’ are the two most often occurring reasons for IT projects. External disruptions, regardless of whether these are caused by nature or human action, are distinctly less common as a source of contractual change in IT projects. The construction projects stand out because of their high frequency of change due to errors and contradictions in the original project documentation. For the IT support service contracts, ‘lack of communication between the contractual parties’ is a prominent reason for change. For facilities management contracts, changes in the client's core business rank highest among reasons for contractual change. Considering only the industry type—IT related or building related— and including also the service contracts, technological innovation plays a slightly greater role for contractual change among the building related contracts. The perception that IT related contracts suffer more

Table 1 Reasons for contractual changes in the 16 contractual relationships, according to contract type [average reported frequencies; scale 0 = never; 4 = very often]. Reason

IT projects

Construction projects

IT services

Facilities management services

Errors in original documentation Client core activity change Better understanding of client needs Provider technology change Subcontractor technology change Better understanding of original state Lack of communication External disruptions, natural External disruptions, societal

1.13 1.56 2.25

2.63 1.38 1.50

1.56 1.25 1.44

0.50 2.63 1.69

0.75 0.75

1.50 1.38

1.25 0.25

1.88 0.75

2.25

1.50

1.38

1.00

1.38 0.13 0.19

0.63 0.75 0.38

1.69 0.44 0.63

0.38 0.25 0.69

771

than building related contracts from contractual changes necessitated by poor communication between the parties stands out clearly. Obviously, the questionnaire data reflect perceptions among respondents, who might perceive contractual relations depending on their own roles. In general, there is no great discrepancy between the views of clients and providers. Nevertheless, there are differences: clients are more apt than providers to attribute contractual change to better provider understanding of technical conditions. Also, subject to external disruptions, providers are less likely to indicate changes due to environmental reasons and more likely to do so for governmental regulatory change and other societal reasons. It seldom occurs that contracts are actually changed because of external disruptions, since these matters will have been covered in the original contract and the associated risk clearly allocated, usually to the client, but not requiring any change in the contract itself if and when the risk materializes. Otherwise, there is a broad consensus among both clients and providers, including the issue of ‘lack of communication’ as a category of change reasons. 4.3. Dynamics of the relationships Several respondents expressed their view of contracts in terms of a multistage model of contract relations, where the formal contract plays a crucial role initially but should be, according to respondents, and often is, effectively shelved later on when routines and informal relations have been developed and established. This view was more pronounced for the long term service contracts, which could be understood as having a first stage that has a strong similarity to a short-term project and evolves from the wording of the contractual clauses; once this stage has been left, the informal relations tend to dominate. When renewal of a long term contract is imminent, a new stage is entered, again with strong elements of a short-term project. 4.4. The contracts: complexity and routines Looking at all sixteen contractual relationships, we find that for six contracts, the provider had been involved in the initial formulation of requirements; here, it turned out that frequencies of changes were consistently lower for each change reason, compared to the ten contracts where the provider did not participate early on (Table 2). Without exaggerating the importance of different ratios for particular reasons of changes, given the small number of cases, it is noteworthy that provider participation is associated with change rooted in new technologies suggested by providers during the contractual period itself. On the other hand, the least difference in frequencies is found for ‘understanding of original state’, which is also surprising, since here it could have been expected that initial participation would lower the frequency of such changes. Although the data give a strong impression of early provider participation leading to greater robustness against contractual change, regardless of the category of change reason, the fundamental causality might be more complex. Thus a history of earlier and harmonious contractual relations or that the client has expected that the contractual relationship would be stable might explain why the provider was in a position to influence the formulation of initial requirements. It is typical to have monthly meetings between the contractual parties; in particular for construction projects, there may be a highly developed hierarchy of specialized meetings with frequencies of their own and also a hierarchy of groups for these meetings. Throughout the sixteen contractual relationships, there was no evidence of the parties trying to reduce their total transaction costs by postponing the financial settlement to later stages of the contractual period. However, in construction projects, there might be a supply chain delay, caused by the need for receiving subcontractor accounts necessary for billing the client. For the two types of service contracts, the common pattern was that there would be monthly settlements, sometimes combined

772

J. Bröchner, U. Badenfelt / Automation in Construction 20 (2011) 767–775

Table 2 Reasons for contractual changes in the 16 contractual relationships, according to provider participation in initial formulation of requirements [average reported frequencies; scale 0 = never; 4 = very often]. Reason

Provider participated

Provider did not participate

Frequency ratio participants/ non-participants

(N = 6)

(N = 10)

Errors in original documentation Client core activity change Better understanding of client needs Provider technology change Subcontractor technology change Better understanding of original state Lack of communication External disruptions, natural External disruptions, societal

1.08 1.17 1.38

1.68 2.03 1.93

0.65 0.58 0.71

0.75 0.50

1.70 0.95

0.44 0.53

1.42

1.60

0.89

0.75 0.25 0.42

1.18 0.48 0.50

0.64 0.53 0.83

with a fixed annual mechanism, typically taking the form of a predefined annual percentage fee reduction that is intended to reflect productivity increase within the provider organization. When comparing the IT software projects and the construction projects, it seems evident that the software development process is more structured, although it was only one IT project that followed a highly structured set of routines, the Rational Unified Process (RUP). Mention was made of the growing influence of ITIL change management methods partly reinforced by the international ISO/IEC 20000 standard for IT service management. In one other project, a much more flexible process was followed, Scrum, being so agile that in a sense, it is often doubtful whether it is meaningful to speak of contractual change when requirements shift successively during the development process. Opinions appear to be divided as to whether the contractual rules for handling changes are clear and easy to apply. Except for a few comments related to payment principles, those who were engaged in construction or IT software projects reported little difficulty in applying the contractual rules. However, facilities management contracts often suffer from a difficulty to reach consensus on which activities are to be classified as value-enhancing investments and which activities should be treated as part of operational tasks. Thus it emerges as easier to identify and launch a separate project in order to manage software changes in an IT service relationship than in a building based facilities management relationship. For construction projects, the early stages of the life cycle of a built facility sometimes give rise to discussions. Varying duration of embedded equipment complicates the issue of guarantees; two years and five years can be an important difference in terms of risk distribution between the two parties, while arising from the contractor's further relations with subcontractors and component manufacturers in the project supply chain. 4.5. Effects of contractual incentives According to information given by the respondents, eleven contracts had strong incentives and five lacked clauses that defined bonus and/or important penalties (Table 3). Here the frequency pattern is much more varied than in Table 2. The overall picture is however that contractual relationships based on strong incentives will experience higher frequencies of change. It should be noted that the frequencies are about the same for ‘better (provider) understanding of original state’, largely independent of incentives. Identifying the causal link between change frequencies and incentives is difficult, since it is possible that the client has chosen

Table 3 Reasons for contractual changes in the 16 contractual relationships, according to strength of contractual incentives [average reported frequencies; scale 0 = never; 4 = very often]. Reason

Errors in original documentation Client core activity change Better understanding of client needs Provider technology change Subcontractor technology change Better understanding of original state Lack of communication External disruptions, natural External disruptions, societal

Strong incentives Weak incentives in contract in contract

Frequency ratio strong/weak incentives in contracts

(N = 10)

(N = 10)

1.57

1.20

1.31

1.89

1.30

1.45

1.86

1.40

1.33

1.50

1.00

1.50

1.00

0.30

3.33

1.50

1.60

0.94

1.14 0.50

0.75 0.15

1.52 3.33

0.41

0.60

0.68

a set of incentives in anticipation of a relationship that would be subject to a range of sources of changes. Time incentives related to completion time for projects were mentioned occasionally by respondents, but were perceived to introduce complications when a client wishes to coordinate a construction project with the successive installation of production equipment. In addition, cost incentives in the shape of target cost contracts were considered to raise transaction costs significantly for both parties because the fee mechanism leads to a frequent need to re-estimate costs. In service contracts with Service Level Agreements, mechanisms such as provisional penalties for substandard performance were found repeatedly. A measured substandard performance in one period leads to a provisional penalty, which is not transformed into an actual reduction of payment to the service provider if a higher than a predefined minimum standard of performance is reached in one or more subsequent periods. This mechanism is easier to implement in IT service contracts, where the technology makes feasible the continuous and automatic logging of performance. The incentive element in one of the eight service contracts was designed as an explicit option of prolongation, conditional on good performance by the service provider. 4.6. Communication: a recurrent topic At the end of our interviews, respondents were encouraged to reflect freely on any topic that they felt was related to flexibility in contractual relationships. These reflections had many similarities, and although there were slight differences between the views aired by clients and providers, it is hard to find differences associated with industry type or according to the prior length of the relationship. Frequently, it was ‘communication’ that was mentioned first by respondents; communication should be early and rapid, it was thought. Communicating intents and goals was often felt to be necessary to secure mutual understanding. Furthermore, it was said that is not only between the immediate parties to the contract where communication is needed; there is also a wider need for making the benefits of the chosen service arrangements understood in the client organization, from top management down to the current users of the services provided. The role of documentation was emphasized by some respondents, and the spread of standard agreements as well as successively raised competence among clients was mentioned as important today. More than one respondent stressed the relation between communication, understanding and trust. Only

J. Bröchner, U. Badenfelt / Automation in Construction 20 (2011) 767–775

one of the clients, who belonged to the public sector, said that the difficulty of monitoring services constituted an issue. 5. Discussion The survey findings give rise to five topics for further analysis. First there is the nature of the project development process and how this reflects fundamental differences between information technology and construction technology. The second leads to the question of the potential for decomposing the process and the resulting product, and in the case of contracts for more continuous services, the ability to break away and to projectify changes. Third, there is also another question of projects within projects for the continuous service contracts, seen as distinct stages of the contract life cycle and which has to do with the degree of formality and shifting nature of communication between the parties to a contract. The fourth is that for a typical ‘project’ contract, there are different perspectives on contractual incentives depending also on underlying technologies. It is a matter of (costly) identifiability in a coproduction context; it is not only the client and the provider who must reach agreement on which party is the source for a (costly) change, it may also involve other organizations present on a project (physically, as production equipment suppliers, in construction) and who might affect the metrics to which incentive payments are linked. When the need for change arises in IT software contracts—or in IT support service contracts—it is clearly easier in comparison to construction projects, and a more frequent practice, to projectify the change. IT is a technology that is more decomposable than construction is; as a consequence, as Eckfeldt et al. [42] claim, “big [IT] projects are really lots of smaller projects”. There is a greater gap in skills and tools between construction projects and facilities management, whereas the set of resources is more similar for IT software and support contracts, also typically on a higher common level of training and education in the IT context. This observation is related to the fact that it is unusual for construction contractors to enter the market for facilities management services [63], and it also provides a clue to why survey respondents emphasized the change negotiation topic of where to draw the line between services covered by the contract and value-adding maintenance and repair action. The difference in basic requirements on the artefact that a project results in contributes to the difference between IT and construction projects. Most built facilities are required to be robust against external impacts from climate, noise, earthquakes and fire; this corresponds to a production process that while admitting prefabricated components has to ensure that the end product does not easily decompose physically and revert to its modules when subject to external disruptions. Production is typically a combination of assembly and transformation processes in construction projects; robustness is achieved largely through the physical and chemical transformations involved. In their turn, the transformation processes carry with them costly elements of irreversibility; the same characteristics that make them robust protection for the future users of the facility are also responsible for very high costs associated with iterations after a construction project has left the building design stage. The life cycle of relations over a contractual period presents both similarities and contrasting features when comparing the typical project contracts with the support service contracts. There is a flow from formality to informality in the early stages; communication routines, mutual understanding and trust emerge. There appears to be a stronger effect in the case of support service contracts of reduced dependence on the exact wording of the contractual documents. As the end of the contractual period approaches and renewal is imminent, the service contract relationships take on more of a project nature and formality returns. These phenomena, which are common to the IT and the construction domains, have not been emphasized by earlier researchers, but the life cycle pattern can be understood as the

773

outcome of attempts, conscious or not, to reduce transaction costs [64]. A challenge for model developers is to support a rapid initial, more formal stage and to accelerate the development of less burdensome routines for communication and handling of changes. A problem specific to the construction projects is that the end of the life cycle may strain both the formal and the informal mechanisms for coordination, as when heavy equipment for the client's core production is to be installed in parallel with the activities of the construction contractor. For all four types of contractual relationships, there is at least some degree of coproduction with client involvement in the delivery of projects and services [65]. This introduces ambiguities when there is a combination of changes and contractual incentives. Conflicts that are hard to resolve may arise late in the life cycle of a relationship because the parties fail to agree on how to allocate the responsibility for faults that are detected by the client and are difficult to trace unambiguously and where changes have been decided on in joint meetings. Given this link between incentives and changes, it is necessary to analyse the figures in Table 3, which show that contracts with strong incentives undergo slightly more changes than those with weak incentives. Looking at differences for the various reasons for changes, there are only two low-frequency categories (subcontractor initiatives; natural events), where the incentive relation seems to be powerful, but this observation is of dubious significance, given the small number of survey replies. There might be an unobserved underlying factor that creates the association between incentive strength and change frequencies, namely that clients faced with a project or service contract where they anticipate many changes for a range of reasons and that at least some of these reasons also make them think that nonfinancial incentives will be weak in the provider relationship. This would be similar to the phenomenon of ‘endogenous matching’ identified by Ackerberg and Botticini [66] as a caveat for those who are quick to draw causal conclusions without a thorough understanding of how clients decide about contractual arrangements and select their providers. The IT projects appeared in Table 1 to suffer more than construction projects from communication problems, and judging from interview responses, it is informal more than formal communication that is vital here. Instead, the construction contracts were more likely to experience needs for change because of insufficient initial documentation, which translates into the need for formal communication at the outset of the project or even before the contractor has been selected by the client. This dissimilarity between communication needs in the two industries can be given a technical explanation. For construction projects the initial state of nature can be comparatively costly to investigate and to document; soil conditions can be studied through geotechnical analysis, and the hidden structural features of an existing building that is scheduled for refurbishment, but client optimization of investigation efforts over the life cycle of a project very often leads to a need for successive changes as the true conditions are discovered as a by-product of ongoing work carried out by the contractor. Nevertheless, communication in a broad sense has been found important in construction projects [10], and it is obvious that the capability to communicate is important for project managers [60]. Do these isolated observations of underlying differences between industry technologies allow us to predict which elements of highly agile project development processes can be transferred from the IT to the construction domain? For this we need a typology of project work, a parallel to what has been proposed for the manufacturing industry long ago. Thus the Scrum development process can be said to be akin to small batch and unit technology in Woodward's [67] manufacturing typology, whereas construction projects are closer to continuous process production. Typical construction projects follow the waterfall method of producing project requirements, although there appear changes due to process adjustments necessitated by the range of reasons identified in the survey. With an analogous application of

774

J. Bröchner, U. Badenfelt / Automation in Construction 20 (2011) 767–775

Woodward's manufacturing typology, the continuous provision of support services resembles her ‘large batch and mass production’, and this similarity is not least evident in the service contracts being more dependent on routine behaviour and standard procedures. This is clearly another basis for differentiation than the influential project typology proposed by Shenhar and Dvir [5], which they based on system scope and technological uncertainty. 6. Concluding remarks A set of reasons for contractual change has been identified here and applied to contractual relationships in a survey of construction and information technology projects. By comparing change frequencies and as supported by interview responses, it has been possible to discover industry-specific features of relationships and changes. The effects of early provider participation in formulating requirements and of the presence of contractual incentives have been highlighted in the analysis. While the causality is found to be complicated, it is evident that a new typology for projects would be useful. Developers of model-based systems for handling project change who wish to test and transfer practices from one industry to another, in particular practices that lead to a higher degree of project agility, would benefit from recognizing industry patterns of irreversibility, relations between pre-investigations and the successive revealing of initial conditions, as well as the project life-cycle of formal and informal communication. A widespread implementation of models for change management appears to require features that support renegotiation of contractual incentives. It should be noted that the methodology relied on for the survey and the geographical region investigated may have introduced a bias that favours harmonious contractual relationships, since respondent firms were able to nominate the projects and the Swedish setting is believed to imply a less adversary and litigious culture than is common in the UK and the US. The first source of bias is somewhat mitigated by the unusual access to information from both clients and providers; the extent of the second and cultural source of bias can only be gauged by conducting further studies in other regions. Acknowledgements Support for this investigation from the Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (Formas) is gratefully acknowledged. References [1] F.M. Arain, L.S. Pheng, Knowledge-based decision support system for management of variation orders for institutional building projects, Automation in Construction 15 (3) (2006) 272–291. [2] F.M. Arain, L.S. Pheng, Modeling for management of variations in building projects, Engineering, Construction and Architectural Management 14 (5) (2007) 420–433. [3] I.A. Motawa, C.J. Anumba, S. Lee, F. Peña-Mora, An integrated system for change management in construction, Automation in Construction 16 (3) (2007) 368–377. [4] S. Isaac, R. Navon, Modeling building projects as a basis for change control, Automation in Construction 18 (5) (2009) 656–664. [5] A.J. Shenhar, D. Dvir, Toward a typological theory of project management, Research Policy 25 (4) (1996) 607–632. [6] A.J. Shenhar, One size does not fit all projects: exploring classical contingency domains, Management Studies 47 (3) (2001) 394–414. [7] A. Kadefors, Institutions in building projects: implications for flexibility and change, Scandinavian Journal of Management 11 (4) (1995) 395–408. [8] A.O. Akinsola, K.F. Potts, I. Ndekugri, F.C. Harris, Identification and evaluation of factors influencing variations on building projects, International Journal of Project Management 15 (4) (1997) 263–267. [9] S. Senaratne, M. Sexton, Role of knowledge in managing construction project change, Engineering, Construction and Architectural Management 16 (2) (2009) 186–200. [10] S.-O. Cheung, W.Y. Tak, O.K. Chiu, The aggressive–cooperative drivers of construction contracting, International Journal of Project Management 27 (7) (2009) 727–735.

[11] M. Sun, X. Meng, Taxonomy for change causes and effects in construction projects, International Journal of Project Management 27 (6) (2009) 560–572. [12] C.W. Ibbs, C.K. Wong, Y.H. Kwak, Project change management system, Journal of Management in Engineering 17 (3) (2001) 159–165. [13] D. Dvir, T. Lechler, Plans are nothing, changing plans is everything: the impact of changes on project success, Research Policy 33 (1) (2004) 1–15. [14] K.J. Mayer, N.S. Argyres, Learning to contract: evidence from the personal computer industry, Organization Science 15 (4) (2004) 394–410. [15] E.T.G. Wang, P.-H. Ju, J.J. Jiang, G. Klein, The effects of change control and management review on software flexibility and project performance, Information Management 45 (7) (2008) 438–443. [16] P. Abrahamsson, O. Salo, J. Ronkainen, J. Warsta, Agile Software Development Methods: Review and Analysis, VTT Publications, 478, VTT Technical Research Centre of Finland, Espoo, 2003. [17] C. Larman, V.R. Basili, Iterative and incremental development: a brief history, Computer 36 (6) (2003) 47–56. [18] K. Conboy, Agility from first principles: reconstructing the concept of agility in information systems development, Information Systems Research 20 (3) (2009) 329–354. [19] J.-N. Lee, Y.-G. Kim, Effect of partnership quality on IS outsourcing success: conceptual framework and empirical validation, Journal of Management Information Systems 15 (4) (1999) 29–61. [20] T. Kern, L. Willcocks, Exploring relationships in information technology outsourcing: the interaction approach, European Journal of Information Systems 11 (1) (2002) 3–19. [21] L. Poppo, T. Zenger, Do formal contracts and relational governance function as substitutes or complements? Strategic Management Journal 23 (8) (2002) 707–725. [22] E. Beulen, P. Ribbers, IT outsourcing contracts: practical implications of the incomplete contract theory, Proceedings from the 36th Hawaii International Conference on System Sciences (HICSS'03), 2003. [23] A. Kalnins, K.J. Mayer, Relationships and hybrid contracts: an analysis of contract choice in information technology, Journal of Law, Economics and Organization 20 (1) (2004) 207–229. [24] K.J. Mayer, R.M. Salomon, Capabilities, contractual hazards, and governance: integrating resource-based and transaction cost perspectives, Academy of Management Journal 49 (5) (2006) 942–959. [25] N.S. Argyres, J. Bercovitz, K.J. Mayer, Complementarity and evolution of contractual provisions: an empirical study of IT services contracts, Organization Science 18 (1) (2007) 3–19. [26] K.J. Mayer, J. Bercovitz, The influence of inertia on contract design: contingency planning in information technology contracts, Managerial and Decision Economics 29 (2–3) (2008) 149–163. [27] J. Goo, R. Kishore, H.R. Rao, K. Nam, The role of service level agreements in relational management of information technology outsourcing: an empirical study, MIS Quarterly 33 (1) (2009) 119–145. [28] C. Pollard, A. Cater-Steel, Justifications, strategies, and critical success factors in successful ITIL implementations in U.S. and Australian companies: an exploratory study, Information Systems Management 26 (2) (2009) 164–175. [29] J.H.K. Lai, F.W.H. Yik, P. Jones, Critical contractual issues of outsourced operation and maintenance service for commercial buildings, International Journal of Service Industry Management 17 (4) (2006) 320–343. [30] P.J. Luciani, An Investigation into the Cycle and Prediction of Organisation Facility Management Procurement, Thesis submitted for the Degree of Doctor of Facility Management, University of Technology, Sydney, 2007. [31] A. Kadefors, Contracting in FM: collaboration, coordination and control, Journal of Facilities Management 6 (3) (2008) 178–188. [32] D.J. Bryde, Is construction different? A comparison of perceptions of project management performance and practices by business sector and project type, Construction Management and Economics 26 (3) (2008) 315–327. [33] N.O.E. Olsson, Management of flexibility in projects, International Journal of Project Management 24 (1) (2006) 66–74. [34] J. Barthélemy, B.V. Quélin, Complexity of outsourcing contracts and ex post transaction costs: an empirical investigation, Journal of Management Studies 43 (8) (2006) 1775–1797. [35] C.-Y. Chang, G. Ive, The hold-up problem in the management of construction projects: a case study of the Channel Tunnel, International Journal of Project Management 25 (4) (2007) 394–404. [36] T.-Y. Hsieh, S.-T. Lu, C.-H. Wu, Statistical analysis of causes for change orders in metropolitan public works, International Journal of Project Management 22 (8) (2004) 679–686. [37] J.G. Geraldi, L. Lee-Kelley, E. Kutsch, The Titanic sunk, so what? Project manager response to unexpected events, International Journal of Project Management 28 (6) (2010) 547–558. [38] J.J. Reuer, A. Ariño, Contractual renegotiations in strategic alliances, Journal of Management 28 (1) (2002) 47–68. [39] N.O.E. Olsson, External and internal flexibility — aligning projects with the business strategy and executing projects efficiently, International Journal of Project Organisation and Management 1 (1) (2008) 47–64. [40] Y. Zou, S.-H. Lee, The impacts of change management practices on project change cost performance, Construction Management and Economics 26 (4) (2008) 387–393. [41] S. Collyer, C.M.J. Warren, Project management approaches for dynamic environments, International Journal of Project Management 27 (4) (2009) 355–364. [42] B. Eckfeldt, R. Madden, J. Horowitz, Selling agile: target-cost contracts, Proceedings of the Agile Development Conference (ADC'05), Denver CO, 2005, pp. 160–166.

J. Bröchner, U. Badenfelt / Automation in Construction 20 (2011) 767–775 [43] I. Segal, Complexity and renegotiation: a foundation for incomplete contracts, The Review of Economic Studies 66 (1) (1999) 57–82. [44] M. Engwall, No project is an island: linking projects to history and context, Research Policy 32 (5) (2003) 789–808. [45] T. Coltman, T.K. Bru, N. Perm-Ajchariyawong, T.M. Devinney, G.R.G. Benito, Supply chain contract evolution, European Management Journal 27 (6) (2009) 388–401. [46] H. Ness, S.A. Haugland, The evolution of governance mechanisms and negotiation strategies in fixed-duration interfirm relationships, Journal of Business Research 58 (9) (2005) 1226–1239. [47] N. Beaumont, Service level agreements: an essential aspect of outsourcing, Service Industries Journal 26 (4) (2006) 381–395. [48] K.J. Crocker, K.J. Reynolds, The efficiency of incomplete contracts: an empirical analysis of air force engine procurement, The Rand Journal of Economics 24 (1) (1993) 126–146. [49] K.S. Corts, J. Singh, The effect of repeated interaction on contract choice: evidence from offshore drilling, Journal of Law, Economics and Organization 20 (1) (2004) 230–260. [50] J.J. Reuer, A. Ariño, Strategic alliance contracts: dimensions and determinants of contractual complexity, Strategic Management Journal 28 (3) (2007) 313–330. [51] B.E. Rowen, Software project management under incomplete and ambiguous specifications, IEEE Transactions on Engineering Management 37 (1) (1990) 10–21. [52] A. Kadefors, J. Bröchner, Building users, owners and service providers: new relations and their effects, Facilities 22 (11/12) (2004) 278–283. [53] S. Saussier, Transaction costs and contractual incompleteness: the case of Éléctricité de France, Journal of Economic Behavior and Organization 42 (2) (2000) 189–206. [54] M. Furlotti, There is more to contracts than incompleteness: a review and assessment of empirical research on inter-firm contractual contract design, Journal of Management and Governance 11 (1) (2007) 61–99. [55] U. Badenfelt, The selection of sharing ratios in target cost contracts, Engineering, Construction and Architectural Management 15 (1) (2008) 54–65.

775

[56] A. Kadefors, U. Badenfelt, The roles and risks of incentives in construction projects, International Journal of Project Organisation and Management 1 (3) (2009) 268–284. [57] R. Atkinson, L. Crawford, S. Ward, Fundamental uncertainties in projects and the scope of project management, International Journal of Project Management 24 (8) (2006) 687–698. [58] J. Söderlund, A.L. Vaagaasar, E.S. Andersen, Relating, reflecting and routinizing: developing project competence in cooperation with others, International Journal of Project Management 26 (5) (2008) 517–526. [59] C. Sauer, B.H. Reich, Rethinking IT project management: evidence of a new mindset and its implications, International Journal of Project Management 27 (2) (2009) 182–193. [60] D.H. Stevenson, J.A. Starkweather, PM critical competency index: IT executives prefer soft skills, International Journal of Project Management 28 (7) (2010) 663–671. [61] N. Argyres, K.J. Mayer, Contract design as a firm capability: an integration of learning and transaction cost perspectives, Academy of Management Review 32 (4) (2007) 1060–1077. [62] D.J. Bryde, L. Robinson, Client versus contractor perspectives on project success criteria, International Journal of Project Management 23 (8) (2005) 622–629. [63] J. Bröchner, Construction contractors integrating into facilities management, Facilities 26 (1/2) (2008) 6–15. [64] G.M. Winch, Governing the project process: a conceptual framework, Construction Management and Economics 19 (8) (2001) 799–808. [65] S.L. Vargo, R.F. Lusch, Evolving to a new dominant logic for marketing, Journal of Marketing 68 (1) (2004) 1–17. [66] D. Ackerberg, M. Botticini, Endogenous matching and the empirical determinants of contract form, Journal of Political Economy 110 (3) (2002) 564–591. [67] J. Woodward, Industrial Organization: Theory and Practice, Oxford University Press, Oxford, 1965.