in project management – a structuration perspective
MSc BA Change ManagementFaculty of Economics and Business University of Groningen
February 2017
Pascalle Albers Student number: 2114542 First supervisor: Dr. C. Reezigt
Word count: 13.736 (85.941 with appendices)
Abstract
This study aimed for a better understanding of the processes concerning how change agents are influenced on combining hard and soft practices in project management. Structuration theory was used as the theoretical lens for this study. Data for this study was collected in three different cases. In total fourteen semi-structured interviews were held with project members and managers. Within-case analyses and a cross-case analysis resulted in three main findings. First, the margin an agent receives and claims influences which practices are performed. This margin seems to be influenced by the hierarchical structures of the organization. Second, agents in organizations with highly-valued hierarchical structures seem to be less able to change the domination structures. Last, this study showed that while methods can be perceived as consisting of hard or soft practices, culture of an organization overrules procedures of methods. Thereby the culture of an organization influences how soft and hard practices are combined. This study contributes to the existing literature by taking an actor-perspective to study the underlying processes that influence changes agents’ decision-making on combining hard and soft practices in project management.
2
Table of Content
Introduction ...3
Literature Review ...5
Hard and Soft Practices in Project Management ...5
Structuration Theory ...7 Methods ... 11 Research Design ... 11 Case Description ... 11 Case X. ... 11 Case Y. ... 12 Case Z. ... 12 Data Collection ... 12 Data Analysis ... 13 Quality Criteria ... 13 Results ... 115 Within-Case Analysis... 15 Case X. ... 115 Case Y. ... 127 Case Z. ... 128 Cross-case analysis ... 20
Discussion and Conclusion ... 23
Discussion and Propositions ... 23
Margin... 23
Primary and secondary structures. ... 24
Culture. ... 26
Theoretical and Managerial Implications ... 288
Theoretical implications... 28
Managerial implications... 28
Limitations and Future Research ... 288
Limitations. ... 28
Future research. ... 29
References ... 230
Appendices ... 36
Appendix 1: Overview of Respondents Profiles ... 36
Appendix 2: Interview Guide ... 36
Appendix 3: Codebook ... 38
3
Introduction
“In the center of increased globalization is the need for project managers to have flexibility in a project system in order to be able to adjust constantly to emerging challenges and opportunities.”
(Fernandez & Fernandez, 2009, p. 10).
The globalization of business environment was one of the reasons why hard project techniques are no longer sufficient in project management (Pant & Baroudi, 2008). According to Koskela and Howell (2002), project management as they knew it was not sufficient for management of uncertainties and change. While project management used to focus on hard practices, currently soft practices are gaining popularity (Fernandez & Fernandez, 2009). Soft practices seem to be more appropriate in ill-defined situations where humans and culture are considered important (Checkland, 2000; Liu, Chen, Chen & Sheu, 2011). This differs from hard practices, as hard practices are applicable in managing well-defined situations (Crawford, Costello, Pollack & Bentley, 2003).
Although there is a growing interest for soft approaches in project management literature, scholars such as Sirkin, Keenan and Jackson (2014) and Cowie (2003) claim that both hard and soft skills are essential for project management. According to the latter, project management should involve soft practices and hard practices instead of using them separately. Ruuska and Vertiainen (2004) argue that hard practices are important for project success, however soft practices are needed to create a mutually understanding of projects and goals. Hard and soft practices are not mutually exclusive, as change agents use practices in a complimentary manner (Crawford & Pollack, 2004). Combining hard and soft practices seems to be a subject of interest in the field of project management. Nevertheless, there is a lack of literature on how change agents are influenced when it comes to combining these practices.
Several authors have acknowledged the existence of this theoretical gap. One of the studies that indicate this gap is a research from Gustavsson and Hallin (2014). While in reality a combination of hard and soft skills is more common, in literature attention is paid to using hard or soft practices independently. According to Karlström and Runeson (2006), in real-life projects soft and hard practices are not isolated activities. They mention that project management seems to be at best when practices are combined. However, they do not elaborate on how exactly these decision-making processes are influenced. This is underlined by Conforto and Amaral (2016), who find little research about what affects the combination of project management practices. By examining this literature gap, this research aims to enhance knowledge about what influences change agents on how they combine hard and soft practices in project management.
4
These structures influence how change agents make decisions. Structuration theory is used to examine social systems from an actor-perspective.
This theory provides a theoretical framework for interplay between agent and structure. According to structuration theory social reality is constructed by structure and agency (Giddens, 1984). Social structures are created by human agents and at the same time actions of human agents are created through social structures. Giddens (1984) refers to this as duality of structure. In this duality, agent’s actions are constrained but also enabled by these shared structures (Busco, 2009; Orlikowski & Robey, 1991). In this study social systems refer to these patterns of structuration in projects, whereas social practices refer to how change agents combine hard and soft practices in project management.
Little is known about how change agents are influenced on combing soft and hard practices in project management. To close the gap this study will focus on what affects this process. More explanatory research will contribute to this immature literature field. This study contributes by laying foundations for understanding how structure and agency affect project management. Further, the study provides a new perspective by applying structuration theory to examine social systems of projects. Structuration theory is used because of its actor-perspective, which corresponds to the aim of explaining how an agents’ decision-making process is affected. Besides this theoretical contribution, this study could also be of managerial interest. A better understanding of how change agents are influenced on combining soft and hard practices in project management will provide insights in this process. It is important to enhance this understanding of managing projects, since better understanding will help to improve project management processes and outcomes of projects. This results in the following research question:
How are change agents influenced on combining hard and soft practices in project management?
5
Literature Review
Hard and Soft Practices in Project Management
The business environment of the twenty-first century is a turbulent and continually changing environment in which different skill sets are required to achieve business success (Basl & Doucek, 2012; Kalina, Smutný & Řezníček, 2013; Pant & Baroudi, 2008; Schein, 1996). In order to create competitive advantages, employers increasingly need their personnel to be more adaptable, teachable and responsible. Communication within the team and network is necessary to successfully pursue collective goals (Pant & Baroudi, 2008). Different situations demand different management practices. According to Crawford and Pollack (2004), practices of hard and soft management are not mutually exclusive, but can be used in a complimentary manner. They argue that it is important in project management to combine practices. How this combination of practices is made, is determined by change agents. Change agents combine hard and soft practices according to the organizational context they operate in. This context thus creates structure to and influences this process. Structuration theory helps to explain this interplay between agent and structure. Before this study elaborates on structuration theory, hard and soft practices are clarified in order to understand which practices are referred to in this study.
Hard practices are linked to well-defined technical problems and are in search for objective knowledge (Checkland, 2000; Crawford & Pollack, 2004; Liu et al, 2011). Contrarily, soft practices emphasize inter-subjective creation of knowledge in ill-defined situations where humans and culture are considered important. According to Crawford and Pollack (2004) hard and soft practices are two opposing ends of a continuum. Projects are viewed neither as only hard nor soft, but as having characteristics of hard and soft side of the continuum (Atkinson, Crawford & Ward., 2006). Crawford and Pollack (2004) use their framework to clarify differences between projects. They propose seven dimensions on this continuum:
Goal clarity – How clearly defined project goals are.
Hard practices work with the assumption that goals are clearly defined previous to the project (Atkinson et al., 2006; Crawford et al., 2003; Crawford & Pollack, 2004). Soft practices focus on exploring the problem definition and acknowledge ambiguity in
their goals (Atkinson et al., 2006; Crawford et al., 2003). Goal tangibility – How tangible project goals are.
Tangible goals can often be defined in measurable and clear terms, this fits hard practices.
Soft practices have goals that rely on subjective interpretation and cannot be measured easily, these goals are intangible (McElroy, 1996).
6
Hard practices are measured through quantitative measures, which results in objective realities (Crawford & Pollack, 2004).
Soft practices are measured by qualitative measures, as the problem is not yet discovered (Atkinson et al., 2006; Crawford et al., 2003; Crawford & Pollack, 2004). Outcomes of qualitative measures are subjective interpretations.
Project permeability –Extent to which risks from the environment can influence the project. Hard practices are isolated from their environment, while they focus on management of
defined issues. Therefore, goals, processes, and outcomes are less sensitive to influences of the environment.
Soft practices create situations in which permeability can lead to project success (Belzer, 2001; Crawford & Pollack, 2004). By being permeable to environmental influences, deliverables can be altered during the project (Jackson, 2000).
Number of solutions – Project approach to examine and refine goals.
Hard practices focus on most efficient delivery of predetermined solutions; there is one best way for each project.
Soft practices tend to have multiple alternative options, which are discovered through debate, participation and discussion. It is supposed that there is no one best way, but there are several possibilities each valued differently by stakeholders (Crawford et al., 2003).
Degree of participation and practitioner role – Roles of project agents in managing the project.
Hard practices are non-participative. Agents of a team are experts in their area and their roles are clearly defined. Agents are aware of the boundaries between tasks that they or others have to perform.
Soft practices concern teambuilding and participation (Crawford & Pollack, 2004; Pant & Baroudi, 2008). People are encouraged to cross professional borders to attain many ways to solve issues.
Stakeholder expectation–Whatstakeholdersconsider legitimate use of project management. Hard practices focus on logical relations between project elements. There is little
interaction between stakeholders during the project (Crawford & Pollack, 2004).
7
According to Crawford and Pollack, these factors are the main hard and soft practices in project management. However, in academic literature more practices are pointed out. These include hard practices such as: cost- and schedule control, measuring work performance, monitoring of quality and risk analysis (Cleland & King, 1997; Frame, 2002; Kerzner, 2006). Moreover, practices which are considered soft concern negotiation, communications, change management, uncertainty, stakeholder management, community perception, legal acceptability, benefits, value management, and political and social impacts (Frame, 2002; Howell, Windahl & Seidel, 2010; Jaafari, 2001; Thiry, 2002).
Hard and soft practices are often related to certain project management methods. Methods consisting of many hard practices are for example Waterfall and PRINCE2 (Conforto & Amaral, 2016; Conforto, Salum, Amaral, Da Silva & De Almeida, 2014; Ghosh, Forrest, DiNetta, Wolfe, & Lambert, 2015). These methods operate linearly with detailed steps and long-term planning. Furthermore, these methods fit in nicely with hierarchical organizations. On the contrary, SCRUM and Agile project management are characterized by short cyclical planning, flexibility, and a horizontal approach. These are characteristics of soft practices (Beck et al., 2001; Dingsøyr, Nerur, Balijepally & Moe, 2012; Hoda & Murugesan; 2016, Juricek, 2014; Lee & Yong, 2010; Qumer & Henderson-Sellers, 2008; Sharify & Zhang, 2001).
To understand project management as intended in this study, the definition of projects should be defined. Projects are considered to be temporary organizations (Lundin & Söderholm, 1994), which implies that projects have their own social systems (Bresnen et al. 2004; Floricel et al. 2014). However, projects are not ‘islands’ in the organization, but they are interrelated to the environment in which they operate (Engwall, 2003; Grabher & Thiel, 2015). Social systems consist of reproduced practices, which are patterns of social relations across time and space (Giddens, 1984). Structuration theory provides the theoretical framework to examine social systems from an actor-perspective.
Structuration Theory
8
change and maintain structures, actors are also enabled and constrained by structures. Agents are therefore guided by structures, but at the same time create structures. Structuration theory helps with understanding how this works for change agents.
Social systems are not affected by human actions or institutional forces separately (Orlikowski & Robey, 1991). Structuration theory explains this relationship between individuals and society (Giddens, 1984; Jones & Karsten, 2008; Orlikowski & Robey, 1991; Yang, 2012). Structuration theory is a theoretical perspective in which duality of structure is central concern (Bresnen, 2016; Giddens, 1984; Jones & Karsten, 2008; Orlikowski & Robey, 1991). In Giddens view social reality is constituted by both structure and agency. Therefore, social phenomena cannot be the result of structure or agency separately, but rather the result of both of them combined (Jones & Karsten, 2008; Orlikowski & Robey, 1991; Yang, 2012). Roberts and Scapens (1985, p.446) argue:
Through being drawn on by people, structures shape and pattern interaction. However, only through interaction are structures themselves reproduced. This is the 'duality of structure'; it is in this way that structures can be seen to be both the medium and the outcome of interaction.
Duality of structure refers to the idea that social structures are created by human agents and at the same time these social structures create the actions of human agents. Social structures condition social actions by having contextual rules and resources, which enable human actors to make sense of their own and other’s actions (Busco, 2009; Jones & Karsten, 2008; Orlikowski & Robey, 1991; Yang, 2012). According to Giddens (1984) structures are abstract properties of a social system. Without human actors they cannot exist, since they do not possess any material characteristics (Orlikowski & Robey, 1991). Nevertheless, an agent’s actions are both enabled and constrained by these shared structures (Busco, 2009; Orlikowski & Robey, 1991).
Agents are constantly reflecting on their actions, therefore “the seed of change is there in
every act which contributes towards the reproduction of any ‘ordered’ form of social life.” (Giddens,
9
Figure 1: The three pillars of structuration
Source: Adapted from Giddens (1984) p. 29
Structures, modalities and interactions form the basis of the structuration theory, as is shown in figure 1 (Bresnen, Goussevskaia & Swan, 2005; Jones & Karsten, 2008; Manning, 2008). Modalities function as catalysts between everyday actions and social structures of organizations. Giddens specifies interactions between structures and human agency by proposing three modalities: interpretative schemes, facilities and norms (Bresnen et al., 2004; Orlikowski & Robey, 1991). Interpretative schemes can be seen as stocks of knowledge that agents use to interpret and reflect in order to produce or reproduce actions (Orlikowski & Robey, 1991). Such schemes include rules of signification, which enable or inhibit communication processes (Bresnen et al., 2005; Mauerer & Nissen, 2014). This can be illustrated by an example: clothes that people wear to work reflect this influence of social structures produced by agents’ conformation. When meeting someone in a professional environment, we draw on signification structures. Experiences and interpretative schemes influence translation from clothes to a person’s role. Therefore, if we meet a person in a white coat we assume this person to be a doctor or scientist, depending on if we are in a hospital or laboratory. The next modality is linked to structures of domination and includes facilities. Facilities are the means through which agents exercise power or accomplish their goals (Jones & Karsten, 2008). These means are mediated by resources that constitute organizational domination (Mauerer & Nissen, 2014). Resources are comprised of allocative and authoritative resources. Where allocative refers to generating command over materials, objects, or goods, authoritative refers to commands over actors or persons (Giddens, 1984). The last modality concerns norms, which are organizational rules regarding legitimate behavior and actions (Orlikowski & Robey, 1991). Through interaction of actors, rules of legitimate conduct are created when sanctions are used. To summarize, through interactions of these three modalities, human agents reproduce or change existent behavioral structures (Giddens, 1986).
10
theory bridges these analytical levels (Orlikowski & Robey, 1991). Interaction between agency and structure creates social systems (Manning, 2008). Social systems are:
The pattering of social relations across time-space, understood as reproduced practices. Social systems should be regarded as widely variable in terms of the degree of ‘systemness’ they display and rarely have the sort of internal unity which may be found in physical and biological systems(Giddens, 1984, p. 377).
Social systems consist of different levels. These levels can concern organizational structures, but also projects are considered social systems (Bresnen et al. 2004; Floricel et al. 2014). Each project has its own social structures forming a temporary social system. Projects are embedded in multiple contexts and different frames of reference influence the establishment of projects (Dubois & Gadde, 2002; Floricel et al., 2014; Manning, 2008; Mauerer & Nissen, 2014). These systematic boundaries enable temporary systems, but also restrain interaction (Bresnen et al. 2004; Floricel et al. 2014; Manning, 2008). The social structures of a project depend on rules and resources adopted by actors (Bresnen, 2016). Project agents have certain roles in which new relations, tasks, and routines are included. Structuration theory focuses on the interplay between actions and structures and this provides insight in the relations between the project agents (Bresnen, 2016; Manning, 2008).
11
Methods
The aim of this study is to explore underlying mechanisms of combining hard and soft project management practices. To answer the research question: How are change agents influenced on
combining hard and soft practices in project management?, a qualitative case study will be conducted.
This chapter includes the research design, followed by case descriptions, detailed explanation of data collection and analysis and it concludes with an elaboration on the quality criteria of this study.
Research Design
Such an approach enlarges the understanding of complex phenomena (Eisenhardt, 1989). The focus of this study is on how a deeper understanding of processes concerning the combination of project management practices can be created. To explore such processes, three descriptive case studies have been conducted. Through case studies insight can be gained in phenomena in a real-life context (Eisenhardt, 1989; Yin, 2003). Furthermore, case studies are used to generate theories (Aken, Berends & van der Bij, 2012). The cases used in this study are selected based on a theoretical sampling accordingly with the research question (Eisenhardt, 1989). Consequently, the selected cases meet the following requirements: (1) organizations work with project-based procedures and (2) respondents are all project agents. Individual employees from three organizations are selected as cases to be studied and compared. This makes it possible to examine influences on combining hard and soft practices in project management.
Case Description
For this research cases were selected in which respondents work with projects on daily basis and worked on a project at the moment of the interview. This ensured that the agents’ knowledge about processes could be easily recalled. To find cases that suite these descriptions, contact was made with managers of several organizations. This resulted in introductory conversations with three companies. In these conversations research objectives were explained to the managers. The managers, with whom the first conversation was held, were not interviewed for the research, but they appointed other managers and members to participate in this study. The disadvantage of this selection process is that it can be biased due to the fact that agents were selected by the manager. However, an advantage is that it ensured that agents were willing to cooperate and were currently working on a project.
To investigate influences on combining soft and hard practiced at every case, interviews were conducted with project managers and members. Unfortunately, not the same number of both roles could be interviewed at every organization. This could affect results of the study. In total fourteen interviews were held. An overview of the respondents’ profiles is presented in appendix 1. In the next section short descriptions of the organizations are provided.
Case X. The first selected case of this study is a Dutch organization which creates websites,
12
growing. It works with small, interchangeable teams consisting of four or five agents. This organization makes use of a rotation system, so that for each project the most suitable agents can be deployed. The project content of the projects is generated by the client in dialogue with the project manager and occasionally a member of the team. Clients of this organization are external parties who give this organization an assignment for making an online request. The first case uses Waterfall method as their approach to project management. In order to analyze this case from multiple perspectives, five interviews were conducted. Two managers were interviewed and three members.
Case Y. The second case concerns a Dutch organization which is responsible for mail and
logistic solutions. This organization is the link between senders and receivers of online and physical networks. It is one of the largest employers of the Netherlands and counts more than 49.000 employees. The company is divided in many departments and interviews were held with the business process department, which is primarily responsible for IT-related projects. Project teams consist of six fixed members, but are often larger depending on the project. Most of the time project assignments are delivered by an internal client, often a manager or director of the company. In rare cases the client is an external party, but the respondents were occupied with projects from an internal client. Content is then thought of by a small group. First a preliminary investigation is executed to see what the project would cost and if a profitable business case can be made. The organization used to apply the Waterfall method for each project, but since a couple of months the SCRUM method has been introduced. To analyze the second case, five interviews were held. Four out of these five respondents sometimes had the role as a member and sometimes as a manager of the project. The other respondent only had the role of member.
Case Z. The last case of this research concerns a Dutch organization operating in the public
sector. This organization is the subcontractor on social insurance funds in the Netherlands. Approximately 3200 employees work for this organization. This organization is divided in several departments and agents from multiple departments work together during a project. Projects arise from internal clients and are mainly IT-related projects. Teams mostly consist of eight to ten people per team, but this can differ per project. This organization makes use of internal and external project managers, but external managers often work at the organization for several projects. There are two project management methods at this company. These include SCRUM and PRINCE2. Four interviews were held at this case. Two of these interviews were conducted with project managers, of which one was an external project manager hired by case Z. The other two respondents were project members.
Data Collection
13
confidentiality and anonymity were assured and permission for recording was asked. All interviews were then recorded. The interviews lasted between 28 and 53 minutes and were held in Dutch, which is the native language of all the respondents.
All interviews started with general questions about respondents functions. This was followed by elaborating on how projects are managed and influences on this process. The main focus of the interviews lay on the experiences of the respondents with regard to the project management process. The researcher asked about what factors influence their actions. By asking about daily practices a clear view was obtained from the project processes.
Data Analysis
Several steps were taken to analyze the data. First, the interviews were transcribed in exact words and made anonymously. These transcripts were send to the respondents to be checked and all transcripts were approved by the respondents. Secondly, the transcripts were analyzed on the basis of within-case analyses, of the three cases separately, and on a cross-case analysis (Eisenhardt, 1989). During this process Atlas was used as an analytical tool. The within-case analysis helped to gain a rich understanding of each case. This enabled the examination of unique patterns and helped to become familiar with each case separately. After the within-case analysis, a cross-case analysis was performed. Data of the cases were combined and cross-case patterns were observed.
To generate codes, transcripts were analyzed in line with grounded theory of Glaser and Strauss (1967), in which theory is generated by comparative analysis. There were three stages of coding used during this study, namely open coding, axial coding and selective coding (Corbin & Strauss, 2008). The first stage consisted of using the open coding technique. This is the process of labeling views and thoughts of the agents into categories. The next step was to identify relationships among open codes through axial coding. Codes from the open coding process were reviewed and related to each other, so that new codes could be formed. In the last stage selective coding was applied. This resulted in selection of core categories (Hennink, Hutter & Bailey, 2010; Strauss, 1978). Through the analysis process, inductive and deductive codes were created. Inductive codes were generated throughout the analysis and emerged from contextual matters. On the other hand deductive codes were based on literature to find preliminary factors of influence. The codes are defined in a codebook, which can be found at appendix 3.
Quality Criteria
14
Reliability is enhanced in several ways. Standardization attends to control for researcher bias. An interview guide was developed for the semi-structured interviews and explicit procedures for data collection, data analysis and interpretation were determined before research. This leaves less influence of characteristics of the researcher on research findings (Yin, 2003). Also, dependency on the researcher is reduced due to the use of Atlas as an analysis tool. Furthermore, respondent bias is constrained due to combining perspectives. Interviews were held with respondents from three different organizations. Respondents had the role of member and/or manager in projects, as can be found in appendix 1, therefore all agents of a team are represented. At last, circumstance bias was reduced by creating similar circumstances for each interview. Therefore all interviews were held face-to-face, in a closed room at the organization of the respondents and interviews were held on different times of the day for all three organizations.
15
Results
In this chapter the three cases are first discussed separately, in a within-case analysis. This is followed by a cross-case analysis in which the focus is on the differences and similarities between the three cases. From the coding process two main topics are identified. First, there are structures that influence the change agent. Secondly, agents themselves influence the process as well. Following within-case analyses are structured accordingly. An overview of the coding scheme used for data analysis is provided in appendix 3.
Within-Case Analysis
This section covers the results of the within-case analyses for the three cases separately.
Case X. Based on the interviews with agents at case X, there are several noticeable factors. An
interesting result of the interviews is that case X concerns an organization with strict rules and a hard method, but there is a soft approach to these hard procedures. How these variables are noticed is explained in the next sections.
Structures. First influential features are the structures of case X. These structures exist of multiple hard practices, which are influenced by the importance of the culture of case X. The first process in which this can be observed is the monitoring process. There is a high amount of procedures for monitoring progress and quality of the project. This is monitored strictly by the project manager, as X1 mentions:
The project manager has to keep an eye on the hours spend. Actually, our intranet does that. Our employees fill in their hours and the amount of hours used is tracked by the system. Therefore, we know exactly how much time we spent on what tasks.
The organization built a supervision tool in their intranet to help project managers with keeping an eye on the progress: “We can check if tasks exceed the approved hours. When that happens, we [project managers] receive an e-mail.”(X4). Even though these procedures show the importance of monitoring, the culture of case X affects this process. Although the responsibility of planning is with the managers, members are supposed to cooperate. Member X3: “The monitoring of the planning is done together.
When you foresee delays, you have to inform the project manager and work it out together.”
The second process, in which hard practices are affected by soft characteristics of case X, is shown by the perception on success. The success of a project is measured through an efficiency score. Project member X3: “We measure how many hours we made in relation to the money we get for it.
(…) This results in a percentage that reveals how well we have done.” Although this hard
16
went. However, almost always there is also some education money [this means that in some cases a
project that exceeds its budget still can be of value to the organization, because by working extra hours the agents gain more experience] in the project.” This qualitative measure again shows importance of culture on project management practices.
Another process concerns the determination of responsibilities. Practitioner roles are clearly defined and in accordance every role has its own responsibilities. Although hierarchy is not considered important within case X, responsibility for the final product is borne by the managers. However, in concordance with the cooperative culture, when mistakes are made by members they will be held responsible. In the role as representative of the client, the managers’ decisions are made from another perspective than the members’ decisions. Members X5:
Sometimes there are situations where the project manager wants you to finish the task quickly as it costs too much money. But the developer wants to deliver something nice and neat. And more hours are needed to deliver that kind of product. That gives conflicts, as one wants it to be done quickly while the other needs more time.
Consequently, this results in different decision-making processes. Although the managers hold decisive power, there is room for discussion. Formal titles and job descriptions are not strictly enforced. Member X2 feels no threshold to go to project managers: “Usually you just walk over to
your project manager and fix it. It is not really hierarchical here.” The minimal importance of
hierarchy has a major role in the project management.
Agency. In addition to the structures of case X, agents themselves also have an influence of the process in different ways. The degree that agents can influence project management results from values of case X. To illustrate, case X has created an environment with room for learning. As member X3 mentions: “Sometimes you just go on as a programmer, until you get stuck. You should not do that
and have to learn from this. You have to engage in conversation.” The experiences of an agent thus
affect which practices are exercised.
Besides experience, motivation of agents is also relevant. As noticed earlier, managers have an organization oriented motivation. However, a member’s motivation is based on the level of satisfaction with the product and process: “It [the project] should give a good feeling with the team. I
made something beautiful and I am happy with the product. You have that as a criterion for success.”
17
Case Y. There are several noticeable factors that emerged from interviews at case Y. What
attracts attention is that case Y introduces a soft approach to project management, but this is adapted to fit the hierarchical organization. In the following paragraph these variables are further explained.
Structures. Remarkably, case Y is an organization with many procedures and hierarchical levels. These structures have a great influence on the project management practices. An illustration of how a hierarchy structure affects the process is by the limited authority of project agents. In a project management process there are four hierarchical levels involved: client, steering committee, project manager and project members. The lowest level, project members, is authorized to make decisions about how to execute tasks within projects. One level above is the manager, who supervises performance of members. Project manager Y7: “I want to know what is going on. I want to see ‘CC’s’
of e-mails, in order to supervise the schedule.” The manager can make decisions about the process and
methodologies, but as soon as decisions on ‘what’ or resources of the project are made, authority is with the steering committee. This committee consists of stakeholders and delegates of those authorized to decide. Consequently, it is possible that the steering committee is not authorized to make a decision and is obliged to shift the decision to a higher level. Y9: “Often there is another level in between with
people to whom you have to give account. In a sense, it is a rather stratified company.” These
structures are “deeply rooted within the organization of company Y” (Y10) and show the importance of control and monitoring for case Y.
Case Y operates in a dynamic environment and the Waterfall method, which it uses, was considered to be insufficient. As the projects’ primary purpose is costs reduction and the creation of economic gains, changes in the project management were required. To meet the clients’ wishes, SCRUM was introduced. Since comparable organizations implemented this method for similar reasons, SCRUM was considered to be the appropriate solution. Introduction of SCRUM created a new set of rules, as this method consists of more soft practices than Waterfall. It is mentioned by the agents that with SCRUM members have more responsibilities, it is easier to redirect the project as it is a more flexible method, there is an increase in communication between client and team and the last change is the intensified involvement of stakeholders. The introduction of SCRUM meant at the same time the formal introduction of soft practices in case Y. Agents mentioned these differences in SCRUM projects, but they also mentioned that the method was adapted to fit rules of the organization. This emphasizes the importance of the hierarchical structure. The roles that SCRUM attributes to members, differs from those that were attributed using the Waterfall method, which resulted in friction due to the hierarchical structure of the organization. As manager Y8 says: “In a SCRUM environment
there are no project managers anymore. There [with SCRUM] you have different roles.” These new
18
the method. Manager Y8: “There has to be some sort of direction. (…) either the SCRUM master or
the product owner is the project manager.” Therefore the role of manager is kept and remains to have
a controlling function. Besides maintaining managers for the overall picture, also steering committees are maintained to monitor the team. The rest of SCRUMs practices were used as intended by this method.
Besides structures of hierarchy, there are other structures within case Y that are related to rules. However, unlike the strict compliance of the hierarchical structures, these structures are handled less strict. Agents change structures as they were intended by handling differently. To illustrate this, within case Y there are agreements on project deadlines. In principle, these agreements are made to be met and agents are working towards these deadlines. Nevertheless, in reality deadlines are rarely met. Whenever a deadline is exceeded without delivering a project, this has little consequences. According to Y10: “Someone will be angry, but the next day everything is fine. So if a deadline cannot be met, it
cannot be met.” Agents put responsibility with the steering committee, as they have power over
resources, like budget and manpower. The steering committee is therefore considered to take action when they attach value to meeting deadlines. However, the steering committee consists of several departments and parties and this makes it difficult to determine “who is eventually final responsible.” (Y6). Moreover, intended rules on deadlines are altered by agents’ actions and interpretations of these rules.
Agency. In addition to structures, agents also influence the combination of soft and hard practices. The preference and experience of agents affect this process. However, this influence is limited to hierarchical structures of case Y. As Y7 mentions: “You will propose several choices to the
steering committee and argue these choices. They choose one, after you presented your preference and the benefits. (…) As long as you stay within your margin, decisions can be made by yourself.” Also,
the hierarchy structures result in members having less room for practicing based on preference than managers have. Members can suggest what they would prefer, but the manager has power to decide if members may act on their preference. When members act on their preference without getting approval first, it will come back to them in their personal review. As member Y9 mentions: “In fact, it is
arranged such that you have to comply.” This shows how the margin, in which agents can affect
project management, is limited to the structures of case Y.
Case Z. What stands out at case Z is that every agent mentions structures how they are
19
Structures. In case Z there is a difference between intended structures and the enforcement of structures. Rules are not considered strict requirements, but are rather like norms. The following quote of manager Z12 will illustrate this: “When it is green [sign of approval], it is good and everybody
knows that. So everyone just makes sure that it is green and you get no hassle.” Because of the
enactment of these control procedures, room is created for agents to act as they consider appropriate. Agents interpret structures in such a way that they make it easy on themselves. Instead of aiming for the best quality of projects, agents target their actions at getting approval. Therefore, the intention of this structure exceeds its purpose.
Another example can be shown by the importance contributed to deadlines. Although time is money for case Z, deadlines are rarely met. Deadlines are not so much perceived as hard deadlines. As manager Z12 mentions: “It is successful when you deliver the project within time and budget. Or not
too much over it, since delays are self-evident.” Exceeding a deadline is considered legitimate.
Whenever deadlines are exceeded, there are no actions undertaken. According to manager Z12 this is a matter of culture, since agents have the mentality to resign oneself to facts as they are. Member Z14 says: “In the meantime they do a lot of talking, but just nothing happens. (…)There is actually no one
really judged for it. It is considered more or less normal.”
Another procedure that shows influence of agents on structures is the creation of project plan documents. Member Z14 says about this document: “If it’s right, then first a project plan is created. In
reality, it is often that the project plan is finished when the project is almost finished. It is a document that is constantly changing.” Creating these documents is meant to be a guiding process throughout the project, but that is not how it is used by the agents. Projects are constantly changing and agents use this as a reason to not conform to intended structure, but rather to create a new structure. Member Z14: “Then they receive something different from what they had asked for, but that is rarely the case. That
they ask something and that they receive exactly that.” Agents abuse the margin due to a lack of
monitoring of these documents. In fact, case Z practices soft but in accordance with hard formal characteristics, they actually disguise the hard characteristics as soft.
Projects are significantly influenced by the hierarchical structures of case Z. There are four hierarchical levels involved in projects and every level corresponds with a certain level of power. One of the formal processes concerns the allocation of resources. Every decision with regard to budgets needs to go to agents with more power. However, rules regarding the allocation of resources are not that strict. A quote of member Z14 exemplifies this: “They just ask if someone has time and if someone
wants to be part of that project.” However, this process is limited to the hierarchical structures. For
example, whenever agents do not voluntarily join a project, managers designate agents to become members of the team. So in that case intended hierarchical structures emerge.
20
case Z’s demand for control resulted in changing the rules of SCRUM to fit the organization. Therefore, the rules of the method are not strictly followed. This is done by maintaining the role of the steering committee and project manager, both conform to the roles within PRINCE2, while in SCRUM there are no managers and steering committees. According to manager Z12 only terms linked to SCRUM are used, but the practice of this method is still in its infancy. She says: “I expect that the
need for control always be like it is now. (…) Therefore the organization will never come to hardcore SCRUM.” Agents Z11 and Z12 belief that case Z is searching for an intermediate method of SCRUM
and PRINCE2. This results in a margin for managers, which is used to adapt the method to their desire. This makes methods more like a guideline: “There is not a method, it is about
improvising.”(Z12). Managers are free to adjust the method as they want, but again boundaries of this
margin are set by hierarchical structures.
Agency. In addition to structures that case Z contains, there is also a role for agents to affect practices. It is noticed that mainly managers can influence the process, but members rarely have a saying. This results from differences in hierarchical position. Managers influence the process in two ways. The first factor stems from the way project managers view their tasks themselves. Z12:
You can be a project manager that holds on to lists very strictly. But you can also be part of the team. I am responsible and we do it all together. To me, the role depends on what is needed for that project.
The second way is through the preference of managers. Besides the fact that project managers look at most appropriate practices for each individual project, personal preference also plays a role in the decision-making process. Some decisions are based on what fits project managers best. According to member Z11: “(…) but how it is determined, (…) I think also in part by the preferences of those who
exert influence.” As mentioned before, practices can be chosen by the manager, but as long as it fits
structures of case Z.
Cross-case analysis
In this cross-case analysis, previous within-case analyses are compared. This analysis has resulted in some similarities and contradictions between the three cases. The cross-case analysis has the same main structure as the within-case analysis. Main topics are structures and the way agents themselves influence the process. The next sections elaborate on these topics.
Structures. When the within-cases are compared there are several noticeable results. What
21
Waterfall method, than cases Y and Z. Nevertheless, case X approaches these practices in a soft way. This can be illustrated by the following examples. At case X success is measured in a quantitative way, there is a high level of monitoring and the practitioner roles are clearly defined. However, culture of case X affects these practices, because culture is considered really important. Even though hierarchy in terms of title exists, in practice this is rarely noticed. This results in shared responsibility between members and managers. The soft approach is also shown as participation between agents is encouraged and success is also measured in a qualitative way. Cases Y and Z on the contrary apply soft methods of project management, despite a high level of hierarchy in their structure. Both organizations introduced the SCRUM method, including its soft practices. This has resulted in many soft practices, like short sprints, more flexibility and increased stakeholder involvement. Nevertheless, the SCRUM method is adjusted to the existing structures of cases Y and Z. Particularly strong hierarchical structures have affected elements of this method. For example, even though the role of managers and the steering committee do not exist in SCRUM projects, at cases Y and Z these roles remain by incorporating them in the SCRUM method. Due to this, the horizontal approach of SCRUM is stretched vertically to fit current hierarchical structures.
The second situation indicating differences between case X and cases Y and Z is how the agents’ margins are determined. At case X margins are determined by the values of the organization and agents. Rules concerning planning and method are strictly monitored and this is considered an important aspect in project management. However, all agents contribute to the monitoring and thereby share the responsibility. Again, case X approaches these hard practices in a soft way. This creates certain margins for agents in which they can operate. Agents seek for extensibility of these margins by engaging in discussion with other agents. Case X has a culture in which discussions with both members and managers are encouraged. The culture therefore influences the extensibility of margins within agents operate. The process of determining and extending margins differs in cases Y and Z. In these cases structures as they were intended by the organization are changed by actions of agents. This is a result from the lack of strict enforcement on those rules. For example, at both cases it is considered legitimate that deadlines are not met and no consequences result from this. There is no enforcement of intended structures and therefore no incentives for agents to comply with those structures are present. Agents make use of this lack of enforcement by extending the margins as they see fit. This results in the emergence of new structures. For the previous example, new structures emerged wherein it is accepted to not comply with pre-arranged deadlines. However, the extensibility of structures is restricted by hierarchical structures of cases Y and Z. This can be illustrated by the lack of authority managers and members receive at both cases, as all decisions about resources and the ‘what’ of a project needs to go through steering committees. This restricts decision-making power of project agents. Thereby the importance of hierarchy cultures of both cases is manifested.
22
results in soft practices being adapted to hierarchical structures. Moreover, as cultures differ between case X and cases Y and Z, these differences results in different approaches to their project management practices.
Agency. Structures create certain margins for agents to operate in. As mentioned before,
agents’ margins vary within the cases. What is noticeable is that agents at case X receive more responsibility and thereby have more influence based on preference and experience than agents in cases Y and Z. In both those cases agents have a more restricted authority because of hierarchical structures. Agents have certain margins, however structures of the organization restrict the width of these margins. At case Y margins’ width depends on different requirements. Namely, restrictions are based on the purpose of cost reduction, as this is the primary purpose for case Y’s projects. Width also depends on whether practices fit within hierarchical structures at cases Y and Z. This restriction results in different margins for members than for managers. Managers, who have higher hierarchical positions, have some margin while members rarely have this. This results in members needing to conform more to structures determined by the organization. The presence of this margin for agents’ practices and magnitude of these margins seem to influence decision-making of change agents.
23
Discussion and Conclusion
This chapter will summarize, interpret, and discuss the findings and will aim to answer the research question: How are change agents influenced on combining hard and soft practices in project
management? Findings will be compared to existing literature with the aim to develop propositions.
Furthermore, this chapter will present the theoretical and managerial implications of the findings. Finally, the limitations of this research and suggestions for future research will be presented.
Discussion and Propositions
Margin. One of the findings of the cross-case analysis conducted in this research was how
margins, which agents receive and claim, are of influence on the process of combining practices. In all of the three cases, margins are determined by the structures within the organization in which project teams operate. This is in accordance with the structuration theory (Giddens, 1984), which claims that projects and their institutional contexts interact recursively (Manning, 2008; Sydow & Staber, 2002; Tukiainen & Granqvist, 2016). The agents constantly seek to alter the existing structures. This is, for example, illustrated by the change in structures concerning deadlines in two of the cases. Agents’ actions have changed these structures, since it is now acceptable when deadlines are not met. This can be explained by Handy (1986), who states that agents have a propensity to act on their own interest and preferences. In the cases the margins’ width for agents is related to and restricted by the hierarchical position of an agent. Those occupying the higher hierarchical positions within organizations have more authority and wider margins than, for instance, project members. This can be illustrated in two of the cases by the steering committees, which exist of agents possessing higher hierarchical positions than the member of the project teams, as this comes with an increase of decisive power. This can be clarified by Pfeffer (1992), who argues that power can stem from three sources: control over information, formal authority to act, and control over resources. In commercial organizations, such as those discussed in this research, the allocation of resources is particularly important (Parsons & Giddens, 1980). In all of the three cases the allocation of resources lies more in the hands of those agents occupying the more hierarchical positions. In one case this responsibility lies with the manager, while at the other two cases the responsibility is bounded to even higher positions, such as steering committees. Margins at higher hierarchical positions are therefore wider. Consequently, hierarchical structures limit agents’ freedom of altering structures during projects.
Although the hierarchy in the investigated cases creates borders for agents to operate, agents are capable to change structures. As illustrated in the previous paragraph, agents can change certain structures. Such power supports the theory of Giddens (1982), in which he recalls that even agents in subordinate positions can: “carve out ‘spaces of control’ in respect to their day-to-day lives and in
respect of the activities of the most powerful.” (p. 197-198). This is referred to as ‘dialectic of control’.
24
within the structures of social systems (Mumby, 1987). Lukes (1974) states that power is perpetuated by: “the socially structured and culturally patterned behavior of groups, and practices of institutions.” Agents can refuse to comply and therefore, power creates potential liberation and alteration of structures (Stones, 2005). The agents of the cases studied in this research do not use their power to change domination structures. Rather, the agents accept to remain within their present hierarchical structure. This will be elaborated later in this chapter.
Proposition 1: All agents possess a transformative capacity, however the width of the margin is
constrained by the hierarchical structures of the organization.
The findings of this research suggest that structures are both formalized and non-formalized. Agents in all the three cases have mentioned differences between intended structures and the enforcement of structures. For instance, in one case, agents agree on what acceptable project efficiency scores are. However, in practice this score can deviate from the acceptable score, for example when hours are exceeded for educating agents. These are not formal structures, but these structures appear nevertheless as if they are accepted by the organization. In another case intended structures exist with the purpose to create guidance throughout the process of the project. Formally these documents should be made at the beginning of a project, in order to guide the efforts of the team. Nevertheless, agents have created new structures which allowed the documents to be finalized at the end of the project instead of beforehand. Structures are related to the frames of reference of agents. Therefore, structures emerge during interaction between agents, since all agents can have different frames of reference. Also, there is an overlap in structures as the rules of legitimation consist of the norms of the organization and agents themselves. Interactions are routinized, thereby creating both implicit and explicit procedures and norms for projects (Floricel et al. 2014). By reproducing structures and taking them for granted, structures of domination, signification, and legitimation become institutionalized according to agents’ comprehension about project networks. Sometimes agents, in interaction with other agents, liberate themselves from formalized structures. In such situations structures are created that change the agents’ margin. As illustrated above, this was noticed at all cases.
Proposition 2: Implicit and explicit structures emerge in interaction between agents and these
structures are susceptible to the norms of both the organization and the agents.
Primary and secondary structures. The hierarchical organizations in the cases studied in this
25
performance and potential benefits that could result from changing to a softer method. As other organizations successfully implemented SCRUM, the organizations studied in these cases copied this change. The institutional theory can clarify why these organizations changed to SCRUM. First, this phenomenon can be explained by mimetic behavior, which is one of the mechanisms of institutional change (DiMaggio & Powell, 1983). Mimetic behavior occurs as a response to uncertainty. When organizations perceive similar organizations as more successful or legitimate, they tend to model themselves after these organizations (Barreto & Baden-Fuller, 2006; DiMaggio & Powell, 1983). In both cases the organizations were encouraged by the uncertainty created by the environment. As a result, they imitated other organizations by introducing SCRUM in order to gain legitimacy. Furthermore, institutional theory views organizations as consisting of institutions, with rules, practices and structures. Within institutions taking action is conditioned since: ‘departures from them [actions]
are automatically counteracted by social controls that make deviation from the social order costly.’
(Lawrence & Shadnam, 2008, p. 2289). Nonconformity is associated with an increase in risk and cognitive demands and a decrease in legitimacy and resources. This is connected with costs expansion. Institutional contexts, networks of institutions, provide cognitive frameworks for institutions. By implementing new structures that are conforming to norms of the institutional context, organizations demonstrate that they act properly (Meyer & Rowan, 1977). In the studied cases the primary objective is economic gain. Changes are considered to be beneficial for economic efficiency and are therefore legitimated (Dillard, Rigsby & Goodman, 2004). This could clarify why both the hierarchical organizations transferred to soft project management methods, even though these methods do not fit within hierarchical structures.
Proposition 3: In uncertain situations mimetic behavior will attribute to the implementation of
management practices to gain legitimacy.
Proposition 4: Changes in economic institutions will be legitimate as long as economic gain is the
outcome.
26
Therefore, resistance from power exercising parties can be one of the reasons why domination structures remain.
Signification rules include interpretative schemes through which agents make sense of the situation. When SCRUM was introduced, interpretative schemes about roles of agents changed accordingly. However, in both cases the method was adapted to hierarchical structures and therefore new roles have emerged. These new roles include the domination structures as these structures are implemented in the SCRUM method. This has changed the interpretative schemes about what the roles of the agents are again. When interpretative schemes are subjected to this ambiguous situation, agents refer to situations they are familiar with. In both cases, agents were used to project managers having direction over the team. Therefore, in the new situation without a project manager, agents may behave in accordance with the old situation, which results in interpretative schemes that include the known domination structures. These interpretive schemes unfold in accepting power differences. The agents’ acceptance to maintain higher hierarchical positions that have authority over agents therefore legitimates power exercising parties to remain dominant (Courpasson, 2000). In this way, agents do not liberate themselves from domination structures, but rather reproduce the existing structures. Therefore, radical alterations to domination structures do not arise. This also happened in both the hierarchical cases, as the roles of managers and steering committees remained as they were, including the differences in authority as was before introducing SCRUM.
Proposition 5: When methods with soft practices are introduced in hierarchical organizations the
method has to adapt to the organization rather than the hierarchy that adapts to the procedures of the method.
27
Proposition 6: In economic organizations domination structures are dominant over structures of
legitimation and signification.
Proposition 7: Changes in structures of domination are less likely to occur in organizations with
highly valued hierarchical structures.
Culture. In all the three cases culture has a large influence on the way practises of a method
are carried out. In one of the cases, a hard project management method, called Waterfall, is applied. However, the approach of this method is rather soft. For instance, all agents mention the efficiency score as a success measure for the organization. However, the members’ personal success measure is the amount of satisfaction during the project and it is mentioned that the organization provides space to act on this motivation. In the other two cases the soft method SCRUM was implemented to introduce soft practices in the organizations. Nevertheless, this soft method was adapted to the hierarchical structures. These hierarchical structures are the basis for the culture of both cases. It is therefore suggested that hard and soft methods can be both approached in a hard or soft way, depending on the culture. According to Quinn and McGrath (1985), there are four organization cultures: clan, hierarchy, adhocracy, and market culture. One of the cases corresponds with the clan culture and the other two with hierarchy cultures. Clan culture is described by Quinn and McGrath (1985) as emphasizing participation, personal satisfaction, internal cohesiveness and values agents over financial goals (Ashkanasy, Wilderom & Peterson, 2011; Stoica, Liao, & Welsch, 2004). The structure of this culture is decentralized and leaders are considered to be mentors and facilitators of interaction (Buenger, Daft, Conlon & Austin, 1996; Denison & Mishra, 1995). In this study the clan culture makes use of a hard method and performs hard project management practices. However, project management is approached in a soft way, in accordance with the culture of this case. As illustrated above, this case emphasizes that personal satisfaction of the agents are valued above financial goals. The importance of culture is also seen in the other studied cases. In those cases hierarchy culture is combined with the SCRUM method, which contains many soft practices. These practices are approached by focusing on efficiency through procedures, rules, and structures. Also, managers are responsible for monitoring project progress. This corresponds with the description of hierarchy cultures, as described by Quinn and McGrath (1985). Hierarchy cultures represent top-down organizations, which exist of vertical hierarchical structures (Denison & Mishra, 1995). Within these studied cases, a soft method is introduced, but this method is altered in order to fit their hierarchy culture. This was done by maintaining the functions of managers and steering committees. The findings suggest that culture can result in soft methods to be approached in a hard way, and conversely hard methods to be approached in a soft way. The culture within organizations seem to exceed methods or practices within their project management.
Proposition 8: When cultures and hard or soft practices are not aligned, the culture of an organization
