Project autonomy in complex stakeholder environments

case study in non-profit service R&D



Helsinki University of Technology, Industrial Management


Helsinki University of Technology, BIT Research Centre


Project autonomy has recently emerged in project management literature from two different perspectives. First, autonomy has been considered as a possible success factor in projects and teams, with somewhat contradictory findings (Gemünden, Salomo, & Krieger, 2005; Hoegl & Parboteeah, 2006). Second, autonomy has been suggested as a contextual factor that should be taken into account when defining a project's strategy (Artto, Kujala, Dietrich, & Martinsuo, 2007; Artto, Martinsuo, Dietrich, & Kujala, 2007). Such research indicates that projects differ in their degree of autonomy and that project researchers and managers should be aware of and perhaps even make efforts to promote a project's autonomy in its environment.

A project as a temporary endeavor has traditionally been perceived as autonomous or even isolated. It is purposefully decoupled from its context, given sufficient autonomy for execution, and again recoupled to its parent organization afterwards. According to Lundin and Söderholm (1995), the execution phase in temporary organizing is about planned isolation. It refers to the minimization of any disturbance that could threaten the actions of the temporary organization.

At the same time, however, research suggests that projects should not be considered as isolated entities, but rather they should be studied and managed in connection with their context. Particularly when projects are understood as temporary organizations, their relations with the surrounding environment are identified as an important focus for research (Lundin & Söderholm, 1995; Lundin & Steinthórsson, 2003; Turner & Müller, 2003). Projects may differ in terms of the degree to which they are integrated with their surrounding environment (Johansson, Löfström, & Ohlsson, 2007), which may have implications to their management and to their success. For example, Engwall's (2003) findings suggest that there is greater probability of smooth and efficient project execution when a project's purpose and employed practices are aligned with the ideas, structures, and behavioral patterns of the surrounding organization(s).

Johansson et al. (2007) propose that Lundin and Söderholm's (1995) concept of planned isolation might not be suitable for implementing development projects. They concluded that project implementation depends on how the distinction between the project and the permanent organization is defined. Those projects that are deeply integrated in the permanent organization might be easier to execute, but their potential for radical organizational change is modest. Projects that are implemented according to the traditional project management model can be innovative and creative, but implementation of their results may be difficult, and it may be hard for the permanent organization to benefit from them (Johansson et al., 2007).

Many projects cannot operate in isolation due to the external stakeholders that they need to collaborate with. Literature on project stakeholder networks has shown that different stakeholders direct different expectations and interests towards projects, thereby creating a need to repeatedly redefine ends and means (Hellgren & Stjernberg, 1995) and bringing additional complexity to projects (Cova, Ghauri, & Salle, 2002). The complexity of the stakeholder environment increases along with the number of different stakeholders, their different and even conflicting needs, and their mutual interactions and dependencies (Waldrop, 1992; Thiry, 2004).

While project network literature typically focuses on delivery projects only, supplier interaction and joint projects have also been considered in product development literature. Agreement exists on the need to communicate externally (Tushman & Katz, 1980; Katz, 1982) and to integrate suppliers, customers, and other relevant stakeholders in the research and development (R&D) project (Petersen, Handfield, & Ragat, 2005; Koufteros, Cheng, & Lai, 2006). However, such literature typically assumes the project as an isolated entity with some intra-and extra-organizational connections, rather than perceives the project as a network.

Recent literature on project strategy has pointed out that project autonomy is mainly discussed in relation to one strong parent organization (Artto, Kujala, et al., 2007; Artto, Martinsuo, et al., 2007). Although large, complex projects have been assumed to have high degrees of autonomy, the nature of their autonomy has not been explored, explained, or debated in their stakeholder environment.

The purpose of this paper is to increase understanding on project autonomy and its formation in a complex stakeholder environment. In particular, the study investigates how project autonomy is used and enabled. We examine project autonomy in a complex service R&D context, develop a framework for characterizing project autonomy in its context, and thereby bring new evidence of project autonomy in complex environments.

Literature Review

Project Autonomy

The concept of project autonomy has recently raised interest among project management scholars, but literature is not unanimous on its content and formation. According to Lampel and Jha (2004), project autonomy refers to the degree to which the project is allowed to evolve without constant report and input from the parent organization. Autonomy of a project organization as a social system has been defined as authority to set its own goals, its social identity, and boundaries with other social systems, its resources to complete its task, and freedom to organize the behavior of its members (Gemünden et al., 2005). It is argued that the project team's autonomy increases as the team has a right to make more decisions that are relevant to its task. Such decisions are related to setting technical and business specifications, determining product and process design content, scheduling and budgeting, obtaining resources, coordinating with stakeholders, and monitoring the progress and evaluating performance (Gerwin & Moffat, 1997a). Our focus is on the autonomy of a project (as a social system, e.g., team) and not that of individual persons.

Earlier research emphasizes that there are two sides to autonomy: autonomy taken and used by a project (or team) and autonomy given or withdrawn by managers external to the project. Autonomy has been investigated through a team's right to make decisions in areas pertinent to its task and related authorizing or deauthorizing processes (Gerwin & Moffat, 1997a), external influence and internal equality of influence in project decisions (Hoegl & Parboteeah, 2006), and how managers may withdraw a team's autonomy by engaging in decision making in different decision areas (Gerwin & Moffat, 1997b). Gemünden et al. (2005) have studied project autonomy through four different dimensions: goal-defining autonomy, structural autonomy, resource autonomy, and social autonomy.

Different projects need and use different degrees of autonomy. In particular, the more innovative, complex, and technically novel the projects are, the more relevant is autonomy claimed. The use of autonomy appears, e.g., in the form of different organization designs and success factors and measures. Larson and Gobeli (1988) and Hobday (2000) both found that (autonomous) project-based organizing was particularly suited for very complex projects, whereas matrix forms suited projects with any degree of complexity (Larson & Gobeli, 1988). Shenhar (2001) and Griffin (1997) have suggested that different degrees of technology uncertainty or novelty will require different success factors and measures for projects. Gemünden et al. (2005) found that certain types of autonomy were associated with higher degrees of innovativeness. Project autonomy has been connected particularly with technology novelty (Tatikonda & Rosenthal, 2000), i.e., the more technology novelty, the more the project needs autonomy. Also, studies on radical innovations have emphasized the necessity for a project's disruptive and independent nature, even to a degree that they may generate demand not recognized by the firm and its customers before (Garcia & Calantone, 2002). For instance, scientific work is considered to require professional autonomy, i.e., the freedom to choose focuses for research and pursue them independently (Bailyn, 1985; Varma, 1999).

One reason for investigating project autonomy has been its alleged connection with project performance. However, empirical evidence has so far been conflicting. For example, Hoegl and Parboteeah (2006) studied connections between external influence, internal team autonomy, and teamwork quality in 145 software development teams. They found evidence on the positive implications of autonomy: external influence over operational decisions was negatively associated with teamwork quality and cohesion, whereas team-internal equality of influence over project decisions was positively associated with teamwork quality and cohesion. In turn, Gemünden et al. (2005) studied the link between different dimensions of project autonomy and project success, and also considered the project's innovativeness as a moderating variable in a survey study with over 100 firms in different industries. Their study did not find such links (except with locational autonomy) and rather suggested that the link between autonomy and success is conditioned with multiple factors. Gerwin and Moffat (1997a, 1997b) reported conditions in which team autonomy can be successfully implemented and suggested that autonomy is dynamic rather than a stable feature of the teams. An important part of maintaining autonomy was the managers' willingness to support the team rather than interfere with its decision making (Gerwin & Moffat, 1997b). These and other studies indicate that autonomy is a relevant feature in projects, it can vary from one project to another and even during a project, and its link to project success is contingent on other factors and conditions.

Organizational Enablers and Barriers to Autonomy

Organizations promote and hinder projects' autonomy in different ways. For instance, organization design, structure, and processes may be used to create or limit autonomy. Product development literature largely agrees on many enablers of project autonomy: the integrated (concurrent) nature of the development process that assembles capabilities from everywhere in the organization to a project (Tatikonda & Rosenthal, 2000); the use of strong, independent heavyweight teams and managers (Wheelwright & Clark, 1992); project-based organizing (Larson & Gobeli, 1988; Hobday, 2000); and the use of common project management methods and tools (Tatikonda & Rosenthal, 2000; Gerwin & Moffat, 1997b). Earlier research has strongly emphasized the communication between product development teams and their environments (Tushman & Katz, 1980; Katz, 1982), but it has simultaneously noticed the need to isolate the project, guard its boundaries through, e.g., gatekeepers, and at the same time ensure sufficient information and resource flows between the project and its environment (Reid & de Brentani, 2004). In this sense, product development literature has attempted to build a balance between autonomy and support.

Also, the surrounding organization may use direct authorizing processes to give the project team the right to do its work (Gerwin & Moffat, 1997a). The project team may need to engage in different negotiations and collaboration to get authority for its own decisions. For example, the project team may need to consult with upper management or external persons for operational decisions (Hoegl & Parboteeah, 2006).

Besides driving and supporting autonomy, the organization may generate barriers to projects' autonomy. The surrounding organization may withdraw a project's autonomy if the environment changes, if there is no sufficient shared understanding of the new product development process, or if managers develop negative attitudes towards the project because of their lost authority (Gerwin & Moffat, 1997b). The structures, processes, systems, and authorizing practices may all constrain the projects' autonomy.

Studies on product development and organizational design tend to favor connecting the project with the parent organization due to the need for information and resource sharing. However, earlier product development and organizational design research suggests that autonomy in project management should always be allowed to some degree, and the project's successfulness is contingent on contextual and other variables. Earlier research suggests that high degrees of autonomy would be most beneficial with complex projects with high technology uncertainty. The previously mentioned literature on project autonomy primarily examines the autonomy of a project with one parent organization and implicitly makes an assumption that a project's external environment includes only one strong organization.

Research Questions

Due to the increased complexity in the projects' environments, the concept of project autonomy should be extended to projects operating in complex stakeholder networks. In this study, project autonomy is considered as the degree to which the project is allowed to evolve without constant report and input from both the parent organization and the surrounding stakeholder network (in line with Lampel & Jha, 2004).

Projects are often implemented through a collaboration of a network of stakeholder organizations (suppliers, customers, legislators, interest groups, and partners); therefore, project autonomy should be considered with regards to such a complex stakeholder environment. In fact, many modern project types such as those in open innovation and standardization involve multiple stakeholders, each with their unique needs and interests (Artto, Martinsuo et al., 2007). There may be several strong organizations taking part in the project or influencing the project in other ways, different organizations may act as equal partners, and one clear leader organization may not always be identified. Prior research does not explain how a project gains and maintains autonomy in a complex stakeholder environment, and what the role of project autonomy is in such context. Therefore, we focus our research on the following questions.

RQ1. How (through what kinds of practices) does a project take and use autonomy in a complex stakeholder environment?

RQ2. How does the parent organization promote and hinder a projects' autonomy?

RQ3. How do projects differ in their autonomy in a complex stakeholder environment?

Research Methods

We employ a qualitative research strategy in an embedded single-case setting. We sought an organization that would have multiple projects, each operating with multiple, different stakeholders. With this case selection strategy, we wanted to ensure that all projects had a sufficiently similar context and were complex but still were unique in their stakeholder environment. “Consortium” (a pseudonym), a national network of service providers, and its central agency here called “Center” with its portfolio of R&D projects, was discovered and selected as a suitable research setting. Our focus is, therefore, on nonprofit service R&D projects each funded, designed, and delivered through a different network of stakeholders within Consortium and its external partners.

Two different viewpoints are taken to Center's R&D projects. First, we examine 11 different projects in the portfolio based on the experiences of their project managers (RQ1 and RQ3). Second, we examine the organizational enablers and barriers for project autonomy through the experiences of six unit managers (RQ2).

Case Organization and its R&D Projects

Consortium is a nonprofit consortium of organizations in multiple service businesses. The entire consortium consists of the central agency Center and some hundreds of member organizations. Our research data was collected within Center and its subsidiaries. Center is organized into 10 units, each of which is relatively strong and independent. It employs 300 experts, its seven subsidiaries employ another 500 people, and Consortium member organizations employ altogether tens of thousands of persons.

Center has three main forms of activities. The first is to supervise the interests of entire Consortium by, for example, representing it in different forums. The second task is to provide various kinds of expert services to the member organizations by their request. The third task is to conduct research and development activities that are guided by the strategic long-term objectives of Consortium and typically implemented in a project form. Traditionally, Center has had little cooperation between units, and the culture is largely characterized by strong individual expertise.

The R&D portfolio of Center consists of about 40 projects, whose topics reflect Center's vast arrow of expertise areas. The projects are long in duration, lasting typically from one to several years. R&D projects are usually at least partly funded by external financiers, and the projects bring together several stakeholders. Government agencies, public associations, and other instances provide financing, and generally these organizations all have their representatives in project steering groups. Typically, several member organizations of Consortium participate in a project, and they often provide some financing, too. The role of member organizations in projects varies: they may be active partners, or they may just participate in some project-related events, such as seminars or workshops. In some of the projects, Center is the clear project leader, while in others Center is involved either as an equal network partner or as a secondary network member.

The nature of Center's R&D projects varies. Some projects are mainly focused on research or clarifying the current state of affairs in some sector, while others focus on developing a certain service for Consortium's member organizations. Typically, projects are formed around a more general challenge or expertise area. The goal is to achieve long-term impacts at the national level by bringing together all the relevant stakeholders. The actual end results and effects of the projects may be hard to determine beforehand or even after project closure.

Center has a long tradition with R&D projects, but the number of projects and the nature of project management practices differ across units. Projects have been unit-specific and highly specialized. Center has a separate R&D unit that supervises the development portfolio. Still, the projects in the R&D portfolio are not limited to the R&D unit but may belong to other units as well. If the project is run by an external project manager, a person from the unit in question is typically designated as responsible for staying in contact between the project and the unit.

Data Collection and Analysis

The primary data was collected through interviews with eleven project managers and six unit managers. The total number of interviewees was 17. Of the interviewees, 12 were female and all had at least a few years of work experience at Center. The interviewees were chosen by the portfolio's managers, and they were expected to be key informants either regarding a particular project or a unit's projects in general.

The selected projects were all complex, and they all involved more than 10 different stakeholders, each with its own interests and expectations towards the project. Consortium member organizations participated in the projects to develop their daily operations, whereas government agencies and public associations pursued broader societal benefits. All the projects were fairly significant in scope, and most included multiple subprojects, for example, concerning different member organizations. However, the core project teams were small, including only a few experts besides the project manager. Due to the complex stakeholder environment, the steering groups of projects were much larger than the actual project team and involved representatives from the different stakeholder organizations. Most of the projects were part of an external stakeholders' broader (funding) program, besides belonging to Center's R&D portfolio.

Each project had its unique content, scope, project team, steering group, duration, and funding structure. All of the 11 projects were considered as successful in that they had achieved their subgoals in line with plans and received positive feedback from stakeholders. Almost all projects were still ongoing at the time of our interviews which took place in 2005. Table 1 includes some information on the projects included in this study.

Table 1: Projects included in the study
(the specific nature of projects is omitted to maintain anonymity)

Project Project Scope Started Planned End
Project 1 Research on service delivery 2001 2006
Project 2 Integration of service offering 2002 2007
Project 3 New tool/system for service delivery 2004 2006
Project 4 New service evaluation tool 2004 2008
Project 5 Evaluation study on a service 2003 2005
Project 6 Developing a service solution 2003 2005
Project 7 Developing strategy in a service area 2004 2007
Project 8 Reorganizing a service delivery 2005 2006
Project 9 Developing a service delivery model to improve service quality 2000 2004
Project 10 Developing a service 2004 2007
Project 11 Service opportunity identification and development 2000 2006

Some of the interviewed project managers were part of Center's permanent personnel, while others were hired on a temporary basis for running the project. Unit managers represented the parent organization's units and central subsidiaries involved in R&D. They participated in selecting project ideas suggested for the portfolio board, and some were active in projects, e.g., by taking part in project steering group meetings or seminars.

Interviews were carried out by two interviewees from our research team, using a semi-structured interview protocol. The questions included descriptive information on the projects and their stakeholder networks, rationale, and practices in project initiation and management, position of the projects in Center's operations, current R&D portfolio and its management, utilizing the results from projects, and improvement needs. In addition to these rather open-ended questions, interviewers prompted with more detailed questions upon need.

The interviews lasted between 52 and 105 minutes, averaging 85 minutes for project managers and 60 minutes for unit managers. All interviews were tape recorded and transcribed. Documents and secondary interviews were used as supportive material. For this study, we used the interview data selectively. Data from project managers primarily centers to the project, its stakeholders, initiation, success criteria, and follow-up, whereas data from unit managers are more about the general role of R&D projects in Center, the unit's specific way of organizing projects, and development needs related to managing projects both at the single project and at the portfolio level.

The data was content analyzed. Analysis codes were developed based on the literature and the first reading of the interview transcripts. In project manager interviews, our analysis centered to different indicators of autonomy (i.e., relationships, activities, ways of working, and other signs associated with the project's autonomy), and organizational and stakeholder influences to autonomy. In unit manager interviews, the focus of analysis was on different organizational enablers and barriers related to autonomy. Quotations fitting with the analysis codes were identified from the data, and some additional codes were created. Each research question was covered through cross-interview analysis by seeking commonalities and differences, illustrative excerpts, and comparisons to prior literature. Besides descriptive results, we will present a framework for understanding context-specific types of autonomy.


Availability and Use of Autonomy

In the interview data, we identified hundreds of indicators of autonomy, and we coded them in line with Gemünden et al. (2005) into three main dimensions: goal defining autonomy, structural autonomy, and resource autonomy. We also sought indicators of social autonomy and other possible forms of autonomy, but they were clearly less frequently identified in the data.

Indicators of goal defining autonomy showed that the 11 project managers all felt that they had had some degree of authority in setting, defining, modifying, and following up the project's goals. Some project managers reported that they themselves had proposed the project idea based on earlier experience and personal interests, negotiated with all relevant stakeholders, and acquired also financial support for the project. According to one project manager, “I was the one who started preparing the proposal. I invited a team of four people to discuss if the idea made any sense and if we should invest in this and start generating real models for helping [a group of clients] better with current resources.” When defining goals, many conscious choices were made with regards to the entire scope of the project. One project manager described her project's choices: “We defined this [project] in such a manner that we rely on earlier research and utilize the results from prior studies, we won't be reinventing the wheel here.” In many interviewees' responses, the goal setting was associated to “us,” a team negotiating the goals. However, this “we” already included not only the project manager, but possible partners within Center and among stakeholders. Furthermore, the projects typically set their own success criteria and measures and followed them up, besides adhering to basic financial and schedule goals requested by the parent organization and stakeholders.

A majority of the interviewees indicated that the project's goals and changes were connected to prior projects and analyses, Center's other operations, stakeholders' broader development programs, wider institutional and legislative priorities, and also the personal interests of the projects' participants. Goals were developed based on the project manager's or sponsor's knowledge about Center's strategies and stakeholders' intentions and interests, and by actively negotiating motives and rationale for the project. Not only did external needs come up in the form of media messages and strategic statements, but project managers were quite active in seeking support and interest through seminars and discussions.

Structural autonomy appeared in the projects, for instance, in the form of the projects' separation from the line organization, their own organizational structures and networks, own ways of operating, unique reporting structures and systems, and project-specific evaluation systems. All projects even seemed to name their steering group in different ways: management team or board, steering committee, or operative committee. Some interviewees felt that their project was separate and different from more “typical” R&D projects in Center's portfolio or in a specific unit of Center. “We do report yearly towards the R&D portfolio, but content wise our project's own follow up is more important. The project has its stakeholders and its own schedule and goals anyway, we cannot lose track of them.” Many interviewees considered it important that the project's unique ways of operating had been approved and were followed in Center and the unique stakeholder network. As one project manager said, “We have established these core teams in [partner organizations], and we have supported their work from the project team. This is our way of working”. Some interviewees felt that “this project is not connected to Center's R&D portfolio in any way,” or “[our sponsor] has not had any effect on the project, besides the funding.”

Despite such claims of autonomy, all project managers associated some indicators of structural autonomy with constant interplay in the project-specific network. For instance, most project managers followed the reporting structure and requirements of both Center and different stakeholders. Besides obediently serving the requests of these parties, the project managers utilized the reporting channels and systems in an active way to pursue the interests of the project. For many projects, this meant that several overlapping reports were made for different purposes, which project managers interpreted as active and positive communication. Many of the connections with Center and the network relied on the project managers' or sponsors' personal connections and earlier experience, in addition to the formal reporting channel.

Project managers described many aspects of resource autonomy in the form of a project's unique competence needs, dedicated funding, dedicated project managers or experts, and access to additional resources upon need. All projects had a quite unique funding structure, although many of them were partly funded by Center's R&D portfolio. Also personnel were project-specific, although largely part-time and nondedicated. Many interviewees mentioned single individuals' names when characterizing the resource status of the project. “We were quite fortunate that this [person] had gained a lot of experience when working on her doctoral thesis and eventually became available to join our project. We together assembled a plan for the whole initiative.” Even if the 11 projects each had only a small project team, all project managers felt that they could negotiate further resources either from Center or through the project's unique stakeholder network. Many of the interviewees mentioned that “I always get expert support when I ask.” Some interviewees were satisfied with the projects' freedom to organize and re-plan their resources upon need. One project manager gave a practical example of resource changes: “There was this one year when we did not have any project meetings because I said that I was too busy with other work.”

It was evident that resource competition took place, and individual employees had to prioritize their time among different projects and activities. Not all projects had received the funding they had requested, which caused delays in project initiation. Instead of satisfying with a decline decision, the project managers had sought replacing funding through stakeholders actively. The project managers used both Center's and stakeholders' resource base in a skilled manner, both in terms of funding and personnel. Because of the part-time nature of project work, resource prioritizations took place at an individual (rather than portfolio) level.

The interview data revealed only a few indicators of social and other forms of autonomy, possibly due to the small project teams and the nondedicated and part-time nature of most project work. Some interviewees took up the issue of trust, individual commitment and enthusiasm, and shared work space as forms of social autonomy. Many interviewees mentioned the lack of information on and sporadic connections to other projects in Center's R&D portfolio, which could be seen as a form of relational or social autonomy at project-level. Also, the low preparedness for possible future changes in the environment may indicate that autonomy can have a temporal dimension, i.e., a project's activities with regards to time may be relevant in uncertain environments.

Although many indicators of autonomy were clearly connected with influence from Center and the project's unique stakeholder network, the interviews revealed that connection with these external parties produced the same kinds of benefits that are usually expected from autonomy. In fact, it was apparent that the 11 projects were able to use the resources available in the network, connect with the surrounding structures, and engage external stakeholders in goal setting, and thereby they generated the kind of commitment, dedication, and support needed for their good progress. Despite apparently moderate degrees of autonomy with the project and high degree of integration with a unique network, the projects seemed to act as if they were autonomous by embedding deeply in the surrounding network.

Organizational Enablers and Barriers of Autonomy

Interviews with the unit managers of Center and its subsidiaries revealed many ways in which the organizational environment both promoted and restricted the projects' autonomy. Table 2 summarizes the enablers and barriers that were identified and grouped according to autonomy's different dimensions.

Table 2: Summary of organizational enablers and barriers to autonomy

          Enablers           Barriers
  • Shared strategy as an overall guideline on what is important for Center and Consortium
  • Unit managers' limited time and attention to projects
  • Unit managers' intent to select the unit's projects and affect their initial goal setting
  • Many unit managers want to ensure that the projects contribute to other activities of their unit
  • Some general guidelines and benchmarking targets for project management
  • Instant legitimacy for projects when approved in R&D portfolio
  • Visibility for projects, e.g. through Center's website and publications
  • Using temporary, hired project managers from outside Consortium, not accustomed to Center's culture and ways of working
  • Center's lack of detailed project management guidelines
  • Some general guidelines and benchmarking targets for project management
  • Responsibilities related to Center's R&D portfolio's practices (e.g. yearly reporting)
  • Responsibilities related to reporting projects' progress and presenting projects in the unit
  • Often a person from the unit is designated as responsible for staying in contact between the project and the unit
  • Monetary resources and other facilities from R&D portfolio
  • Expert support available upon request from Center's permanent employees
  • Support related to general issues, such as project communication, when requested
  • Center's good image helps in convincing partners to commit to projects; and also in publishing project results in the media
  • In some units: lack of interest and resources towards supervising project managers
  • Limited resources to project activities: only a part of project ideas gets chosen for implementation, and limited resourcing to each chosen project may force to partner up dependence on several organizations
  • Separation of project personnel from unit's other personnel
  • Geographical dispersion of project's personnel due to each stakeholder's own premises for its project staff

R&D projects seemed clearly secondary to Center's units' activities, which limited the resources and attention directed towards project activities. Apart from providing the basic conditions for projects, Center's units did not seem to deliberately promote project's autonomy, but rather the units' lack of either resources or interest towards projects resulted as relatively high project autonomy. Some unit managers reported that they and other employees of their unit simply did not have time to get involved in projects, and thus the projects did not receive much attention.

Center provided the projects with monetary resources and other basic facilities and also some overall structure. The selection of projects in Center's R&D portfolio indicated that the projects had received sufficiently high scoring results in important strategic criteria, and thus it also seemed to legitimate the projects. The inclusion of projects in the R&D portfolio indicated some responsibilities for project managers, including the yearly project reporting requirement. Still, in general these procedures were seen effortless. Aside from the practices related to Center's R&D portfolio, and the possible reporting responsibilities and steering groups requested by the external financiers, project managers were often free to choose the organizing principles for their project. Center's lack of common project management processes, standards or guidelines provided the project managers with higher autonomy in terms organizing and managing the projects.

With regards to individual projects, typical unit managers seemed most concerned with project selection and initiation and more specifically, ensuring that the project's overall theme served the interests of their unit, and that the project's initial goals were aligned with those of the units. Even though project selection decisions were formally made in the R&D portfolio board, the unit managers seemed to have a central role in project selection. They preselected among their unit's alternative project initiatives, and often had a good idea on the number of projects they could get chosen for the portfolio. They also had vehicles to promote projects even if they would not be approved centrally. After project selection and initial goal setting, unit managers were typically less concerned with the projects. One unit manager clarified her role: “In the beginning I often participate in the discussions and ideation, and when we seek financing, I review [the project proposals] and finally prioritize the proposals of our unit. If required, I discuss [new projects] with the R&D unit. But otherwise the project managers take care of the projects.” Boundaries were not deliberately built between projects and units' other activities, but rather through the lack of attention towards projects: “Some [project managers] have unfortunately remained too alone with their projects.” As projects represented the fragmented expertise areas of Center, the project manager often was the only person who knew the focus area properly, which made it hard to collaborate with employees of the corresponding unit.

According to the interviews, Center's units differed in the level of emphasis and attention given to R&D projects. Those unit managers who stressed the importance of projects most also seemed to manage and follow the projects more tightly. These unit managers saw projects as heavily related to the other (main) activities of their units: the knowledge generated through the projects provided critical input for supervising the interests of the member organizations of Consortium, and project's results were used either directly or indirectly in providing services to the member organizations. As one unit manager described: “It [R&D activity] is absolutely necessary. If we did not have this kind of activity, we would very soon loose our credibility. It would mean that we would not have this kind of profound, analyzed knowledge and insight of these issues at hand. On the other hand, [R&D activity] serves our service function, for example, in projects we generate publications that [Consortium member organizations] can benefit directly from. It is a very central part of this work, although an annoyingly small part nowadays.”

The projects related to these units that paid more attention to projects could be considered less autonomous than projects of other units, as there were more mechanisms in place to control the projects and to link the projects to other activities of units. However, these projects could also be characterized as more autonomous since the unit managers tried to ensure that projects had the required resources, and the projects received attention they deserved. Mechanisms put in place to guarantee a constant linkage between the projects and the units' other activities included project presentations or reviews in unit's meetings, arranging work space for project managers at the unit's premises, and naming a partner for the project manager among those employed permanently by the unit. “They [project managers] are…hired here as temporary resources….Most often they sit here [in unit's premises] and…they are supervised quite closely.”

All unit managers expressed the need for a separate budget for starting small, unit-specific projects that would be selected and managed inside their unit and would not be a part of the Center's R&D portfolio. Some unit managers indeed had such internal projects. This arrangement provided autonomy for the unit in selecting project topics. In all, unit managers seemed concerned with supervising the interests of their own unit with regards to Center's R&D portfolio and other units. Although the unit managers generally agreed upon the need for systematic principles in managing R&D efforts (e.g., the structures provided by the R&D portfolio), they seemed to heavily resist the centralized project selection and desired for more autonomy. For example, one person critiqued the way of letting temporary employees to generate project proposals for the R&D portfolio: “It cannot be so that temps ideate new jobs for themselves in terms of what would be fun to do.. When we set up new projects, it should occur in connection with the strategic planning of the line organization.”

As Center's resources for R&D projects were in general very limited, the unit managers understood the importance of external resources for investigating and developing issues they felt were worth promoting. Center's good reputation often helped in getting other sponsors committed to the project. “One efficient means of [Center] in this R&D activity is its authoritative position in [Center's field] at national level, it is very clear that even though the R&D budget is not very large, its social effect is much larger, so [Center] gets its voice heard with an investment in many projects”. The drawback in these network projects was that Center did not anymore have the full authority over these projects: “When we have to include these external sponsors, it is of course so that the [project] is being controlled by several actors and… it is no longer just our perspective.” Many partners had authority in decision making and had their own interests and requirements concerning the project.

Center's main role as supervising the interests of Consortium members also seemed to affect Center's role in taking part in network projects. In many projects, rather than being a proactive actor, Center seemed to focus on naming a representative in the project's steering group and making sure that the project's overall direction was in line with Center's and Consortium's interests. Instead of initiating a new project, Center's units might choose to get involved in an existing network of organizations or an inter-organizational taskforce formed over an issue. As one unit manager characterized: “the taskforce has set certain tasks and it pursues a certain goal and you either like the goal or you do not like it.” Through this kind of involvement, Center might not be able to make real difference on the state of affairs, but at least it remained aware of the national discussion on the issue.

Different Types of Autonomy in Projects

All 11 projects' informants reported different indications of autonomy in their projects. However, the projects differed in the dominance of certain autonomy types. Four informants characterized a resource and structure centered autonomy in their projects (projects 1, 3, 9, and 10). In such projects, goal-defining autonomy was clearly less frequently mentioned than access to resources and the project's structural position, and resource autonomy dominated. Three informants characterized a structure-centered autonomy (projects 5, 8, and 11). In these projects, the project's structural position dominated as a form of autonomy, although resource and goal autonomy were visibly present. Two informants described a goal-and structure-centered autonomy (projects 2, 7). In these projects, resource autonomy was less apparent than goal defining and structural autonomy. Also, two informants described a multi-dimensional autonomy (projects 4, 6), where the three main dimensions of autonomy all were present and almost equally dominant in the project.

The projects differed from each other in how many different indicators of autonomy were described by the informants. Projects 3, 4, 6, and 10 reported clearly more indicators than projects 1, 2, and 7.

As shown previously, the project managers' reported indicators of autonomy were tightly linked to the interplay between the project and the parent organization, as well as the project and its unique stakeholder network. As a matter of fact, a clear majority of autonomy indicators appeared in such interactions. Therefore, we further examined whether the projects differed from each other in how autonomy was generated in the projects' context and identified four autonomy types based on the data.

There were only two projects which appeared to report autonomy indicators dominantly without a connection to the parent organization and stakeholders, namely, projects 1 and 5. The parent organization and stakeholder network had their influence on these projects, for instance, as access to information. Many indicators of autonomy were not connected with organizational or stakeholder influence at all, and therefore, the projects seemed more isolated than other projects.

One of the projects (project 3) reported a high number of autonomy indicators, a majority of which appeared through interacting with the parent organization. The project had a tight link with Center as it had been prepared in line with Center's strategies, was strongly resourced and supported by Center's R&D program, and relied directly on the reporting structures and systems of Center. Although the project also had its stakeholder network, the stakeholders' influence was less apparent than that of Center. In this way, the project was privileged in Center's operations, as compared to other projects.

Four projects (2, 4, 7, and 9) reported a high degree of their autonomy indicators in interaction with stakeholders, but not with Center. These projects had strong links to their unique network, and they used the network to identify customers' and sponsors' needs, define goals based on them, align ways of working with those of the stakeholders, and report their results in line with stakeholders' expectations. The projects differed from each other in the reported degree of autonomy and dominant autonomy types, but the stakeholders' influence was more apparent than Center's influence. Therefore, the projects can be considered networked in their autonomy, as compared to other project types.

Also, four projects (6, 8, 10, and 11) reported a majority of autonomy indicators in connection with the parent organization and with stakeholders. These projects used both the parent and the stakeholders in their goal definition, access to funding and resources, tailoring the suitable ways of working for the project, and fulfilling reporting and information expectations. The projects appeared as highly flexible when accommodating the needs of both Center and stakeholders, but at the same time they seemed particularly skilled in also manipulating the environment to their needs. In an odd contrast to the strong influence of stakeholders and Center, the projects at the same time reported fairly high degrees of autonomy and can therefore be considered as embedded.

A summary of these findings is illustrated in Figure 1.

Four project autonomy types in a complex stakeholder environments

Figure 1: Four project autonomy types in a complex stakeholder environments


Our findings have revealed that project autonomy takes different forms when a project is analyzed in interplay with its complex stakeholder environment. The results showed a multitude of ways in which the project takes and uses its autonomy, and how the surrounding organization enables and constrains autonomy.

While lending general support to earlier research that acknowledges both the project's and the parent organization's perspective to autonomy (Gerwin & Moffat, 1997a, 1997b; Hoegl & Parboteeah, 2006), this study raises four major issues for discussion. First, the concept and nature of project autonomy needs to be understood in its context. Second, we draw attention to how the project takes and uses autonomy through its stakeholders and parent organization. Third, the parent organization's role in enabling and constraining autonomy is discussed and connected with project managers' tendencies and aspirations. Finally, the boundary between the project and its network needs to be considered to understand how project autonomy is created, maintained, and purposefully altered.

Nature of Project Autonomy in Context

When analyzing the data, we noticed that it was neither clear nor straightforward to distinguish between the different dimensions of autonomy as indicated by earlier research (Gemünden et al., 2005). Rather, the dimensions of autonomy were interlinked and appeared in different combinations in different projects. It was apparent that each project had shaped its unique autonomy profile to fit with the surrounding stakeholder network, meaning that the nature of autonomy differs across projects.

We also noticed that different indicators of autonomy did not appear in isolation, but rather informants characterized aspects of autonomy in constant interplay with stakeholders and the parent organization. We identified four autonomy types in context: isolated, privileged, networked, and embedded. These autonomy types were not associated with the degree of autonomy or success, but rather showed that different projects use their surrounding network differently for their own benefit. These findings give support to recent studies that propose project autonomy as a central context variable influencing project strategy selection (Artto, Kujala et al., 2007; Artto, Martinsuo et al., 2007), and as a relative, context-dependent factor contributing to organizational learning (Scarbrough et al., 2004).

Earlier research has dominantly treated project autonomy in the context of one parent organization as a project's feature that can possibly given or withdrawn by external authorizing bodies (Gerwin & Moffat, 1997a, 1997b; Hoegl & Parboteeah, 2006). As a contrast, our results suggest that project autonomy in complex stakeholder environments is a feature of the project network relationships (rather than that of the project only). Our results suggest that further research should be directed at investigating the required management practices in each of the different autonomy types.

Use of Autonomy in Context

At the same time as stakeholders had an influence on Center's projects, the projects were able to use their stakeholder network to gain autonomy. Projects purposefully created arenas for stakeholder influence and thereby sought commitment to project goals, gained access to local resources for project execution, and used existing structures and channels to communicate and share the project's deliverables. Similarly, the parent organization influenced the projects and this connection was used to gain autonomy for the projects. When approving the parent organization's strategies, project selection systems, and reporting requirements, the projects simultaneously used the structures to guarantee funding and expertise, managers' support, and communication channels.

This active use of stakeholders creates an evident contrast to prior research on project or team autonomy with just one single parent organization, where withdrawal of autonomy appears as a significant influence (Gerwin & Moffat, 1997a, 1997b). In a complex stakeholder environment, the momentary allowance and withdrawal of project autonomy appeared as relatively insignificant due to the opportunity to constantly renegotiate the project's position both with the parent organization and stakeholders. Project autonomy appeared as an evolving phenomenon, not permanently labeling a project's character but constantly taking shape through new events and negotiations.

Our findings suggest that a project's autonomy per se is not directly associated with success, at least in a complex stakeholder environment. Rather, projects differ in their success factors depending on the nature of autonomy they take and use in their unique context, as suggested also by prior studies (Gemünden et al., 2005; Artto, Martinsuo et al., 2007). Our case projects each had developed their own success recipe based on the autonomy the project was able to maintain and other situational factors. Although all of the projects were to a great degree dependent on their surroundings, the projects were able to use the connections to reach the benefits of autonomy. We suggest that the success factors for projects in complex stakeholder networks differ from those of projects inside one parent organization.

Parent Organization's Role in Enabling Project Autonomy

Interviews with Center's unit managers revealed limited resources, some structural requirements in the portfolio, particular interest in project initiation, and differences in Center's role and the attention given to projects. Many barriers to project autonomy directly lead way for new enablers of autonomy (Scarbrough et al., 2004), thereby supporting the earlier claim on the evolving nature of project autonomy. The units appeared as more concerned with their own autonomy than the projects' autonomy.

Scarbrough et al. (2004) suggested that project autonomy allows the development of practices different from mainstream organizational practices and thereby promotes learning. Center as a rather immature project-based organization enabled such freedom for projects, which appeared to fit with the unique and complex stakeholder networks well. This, however, raises a question about the maturity or standardization of project-based management and its relevance to project autonomy. Even though established project management practices might delimit projects' autonomy, our results indicated that the few agreed-upon project practices could also be used as enablers for project autonomy. Therefore, the use and enablers of project autonomy appear to be connected with the surrounding organization's institutionalized practices. We suggest further research on the relationship between project management maturity or standardization (as the surrounding organization's routine practices) and project autonomy.

In Center's case, the parent organization's passive and secondary role with regards to projects draws attention to the power and authority of a project to manipulate its own autonomy, and thereby also to the role of project managers. Where prior studies emphasize the authority of external managers in influencing team autonomy (Gerwin & Moffat, 1997a, 1997b), our results show that project managers' own desire and readiness for project autonomy can have an important impact in complex stakeholder networks. Center's project managers, for instance, were not discouraged by loss of funding or lack of resources, but they used their network for finding replacements. Hereby our results support prior studies on the novel and context-specific leadership competence requirements in complex projects and programs (Partington, Pellegrinelli, & Young, 2005; Ruuska & Vartiainen, 2005) and add to them the need to match project managers' competence with the nature and availability of autonomy in the project's environment.

This study did not investigate the project managers' profiles and competences or their fit with project autonomy. However, as the different autonomy types all allowed project success, we suggest that different project manager profiles may either promote certain autonomy types or that project manager competences need to be fit with the project's possible autonomy. Future research is suggested concerning the relationship between project manager competences and roles, and project autonomy.

Boundary Management in Autonomy Formation

The projects at Center differed in their isolation and embeddedness with regards to their environment. Yet, activities in all projects repeatedly spanned the boundary between the project and its environment. Project managers and unit managers presented many ways in which the boundary was crossed to reach benefits usually expected from project autonomy. Boundary management appeared as an important and useful way to get additional resources, integrate the project with surrounding structures and systems, commit sponsors and partners to the projects' goals, and share the results of the projects. Both Center and the projects' stakeholders offered many practical, institutionalized tools and systems for boundary management: publication outlets, internet services, a well-functioning culture of using seminars and workshops, reporting tools, and personal contacts.

The results of this study demonstrated a highly permeable boundary between a project and its complex environment, which indeed is typical in uncertain contexts (Leifer & Delbecq, 1978). Project initiation, with a strong influence from both the parent organization and stakeholders, appeared as active time of boundary crossing, thereby supporting prior studies on program initiation (Lehtonen & Martinsuo, 2007; Martinsuo & Lehtonen, 2007). In contrast with Lundin and Söderholm's (1995) theoretical argument, our findings indicate that projects in complex stakeholder networks do not necessarily seek nor achieve such isolation even during project execution, differing from projects operating within one parent organization. Our results suggest that active boundary management is an important means for gaining the desired level of autonomy for a project in complex network environments. We encourage further research concerning the mechanisms and effects of boundary management in negotiating and shaping project's autonomy.


This study has contributed to research on project autonomy by characterizing the use and enablers of autonomy in a complex stakeholder environment, questioning the dominantly isolated and stable concept of project autonomy and suggesting a contextually sensitive and evolving nature for project autonomy. The case study showed evidence on different degrees and types of autonomy and different uses and enablers of autonomy in 11 projects, in constant interplay with their parent organization and their unique stakeholder networks. Based on the results, we have proposed that project autonomy as such is not associated with project success, but rather it is a characteristic of the project's external relationships calling for context-specific management practices and skilled project manager selection. We also showed that both the parent organization and project personnel had means to actively influence project autonomy through engaging in boundary management.

The embedded single-case study causes limitations to the generalizability of the findings. Consortium as a nonprofit sector network is a unique context whose ways of operating may not match those of private-sector consortia or organizations. Our data collection was not purposefully directed at project autonomy that was taken up as a research subject only after data collection. This may have left important questions out of our interview protocol. Also, we are aware that the single-informant data from projects and units has limitations of reliability, but due to the small project teams, we relied on the project managers as key informants. To make the research procedure transparent and its delimitations clear, we have made an effort to characterize the research strategy and data collection and analysis procedure as thoroughly as possible. Further research should test and verify the findings in new contexts.


Artto, K., Kujala, J., Dietrich, P., & Martinsuo, M. (2007). What is project strategy? Paper presented at EURAM European Academy of Management Conference, Paris, France.

Artto, K., Martinsuo, M., Dietrich, P., & Kujala, J. (2007). Project strategy—strategy types and their contents in innovation projects. Paper presented at IRNOP VIII International Research Network on Organizing by Projects Conference, Brighton, UK.

Bailyn, L. (1985). Autonomy in the industrial R&D lab. Human Resource Management, 24(2), 129–147.

Cova, B., Ghauri, P., & Salle, R. (2002). Project marketing: Beyond competitive bidding. Chichester, West Sussex, England: John Wiley & Sons Ltd.

Engwall, M. (2003). No project is an island: Linking projects to history and context. Research Policy, 32(5), 789–808.

Garcia, R., & Calantone, R. (2002). A critical look at technological innovation typology and innovativeness terminology: A literature review. Journal of Product Innovation Management, 19, 110–132.

Gemünden, H. G., Salomo, S., & Krieger, A. (2005). The influence of project autonomy on project success. International Journal of Project Management, 23(5), 366–373.

Gerwin, D., & Moffat, L. (1997a). Authorizing processes changing team autonomy during new product development. Journal of Engineering and Technology Management, 14, 291–313.

Gerwin, D., & Moffat, L. (1997b). Withdrawal of team autonomy during concurrent engineering. Management Science, 43(9), 1275–1287.

Griffin, A. (1997). PDMA research on new product development practices: Updating trends and benchmarking best practices. Journal of Product Innovation Management, 14, 429–458.

Hellgren, B., & Stjernberg, T. (1995). Design and implementation in major investments—a project network approach. Scandinavian Journal of Management, 11(4), 377–394.

Hobday, M. (2000). The project-based organisation: An ideal form for managing complex products and systems? Research Policy, 29, 871–893.

Hoegl, M., & Parboteeah, K. P. (2006). Autonomy and teamwork in innovative projects. Human Resource Management, 45(1), 67–79.

Johansson, S., Löfström, M., & Ohlsson, Ö. (2007). Separation or integration? A dilemma when organizing development projects. International Journal of Project Management, 25, 457–464.

Katz, R. (1982). The effects of group longevity on project communication and performance. Administrative Science Quarterly, 27, 81–104.

Koufteros, X. A., Cheng, T. C. E., & Lai, K.-H. (2006). “Black-box” and “gray-box” supplier integration in product development: Antecedents, consequences, and the moderating role of firm size. Journal of Operations Management, 25, 847–870.

Lampel, J., & Jha, P. P. (2004). Models of project orientation in multiproject organizations. In P. W. G. Morris & J. K. Pinto (Eds.), The Wiley guide to managing projects, (pp. 223–236). London: John Wiley & Sons Inc.

Larson, E. W., & Gobeli, D. H. (1988). Organizing for product development projects. Journal of Product Innovation Management, 5, 180–190.

Leifer, R., & Delbecq, A. (1978). Organizational/environmental interchange: A model of boundary spanning activity. Academy of Management Review, 3, 40–50.

Lehtonen, P., & Martinsuo, M. (2007). Change program initiation: Defining and managing the program—organization interface. Paper presented at EURAM European Academy of Management Conference, Paris, France.

Lundin, R. A., & Steinthórsson, R. S. (2003). Studying organizations as temporary. Scandinavian Journal of Management, 19(2), 233–250.

Lundin, R. A., & Söderholm, A. (1995). A theory of temporary organization. Scandinavian Journal of Management, 11(4), 437–455.

Martinsuo, M., & Lehtonen, P. (2007). Context and action in initiating organizational change programs. Paper presented at IRNOP VIII International Research Network on Organizing by Projects Conference, Brighton, UK.

Partington, D., Pellegrinelli, S., & Young, M. (2005). Attributes and levels of programme management competence: An interpretive study. International Journal of Project Management, 23, 87–95.

Petersen, K. J., Handfield, R. B., & Ragat, G. L. (2005). Supplier integration into new product development: Coordinating product, process, and supply chain design. Journal of Operations Management, 23, 371–388.

Reid, S. E., & de Brentani, U. (2004). The fuzzy front end of new product development for discontinuous innovations: A theoretical model. Journal of Product Innovation Management, 21, 170–184.

Ruuska, I., & Vartiainen, M. (2005). Characteristics of knowledge sharing communities in project organizations. International Journal of Project Management, 23, 374–379.

Scarbrough, H., Swan, J., Laurent, S., Bresnen, M., Edelman, L., & Newell, S. (2004). Project-based learning and the role of learning boundaries. Organization Studies, 25(9), 1579–1600.

Shenhar, A. J. (2001). One size does not fit all projects: Exploring classical contingency domains. Management Science, 47(3), 394–414.

Tatikonda, M. V., & Rosenthal, S. R. (2000). Successful execution of product development projects: Balancing firmness and flexibility in the innovation process. Journal of Operations Management, 18, 401–425.

Thiry, M. (2004). “For DAD”: A programme management life-cycle process. International Journal of Project Management, 22, 245–252.

Turner, J. R., & Müller, R. (2003). On the nature of the project as a temporary organization. International Journal of Project Management, 21, 1–7.

Tushman, M., & Katz, R. (1980). External communication and project performance: An investigation into the role of gatekeepers. Management Science, 26(11), 1071–1085.

Varma, R. (1999). Project selection methods or professional autonomy? Prometheus, 17(3), 269–282.

Waldrop, M. M. (1992). Complexity. New York: Simon & Schuster.

Wheelwright, S. C., & Clark, K. B. (1992). Revolutionizing product development: Quantum leaps in speed, efficiency, and quality. New York: Free Press.

© 2008 Project Management Institute



Related Content

  • Project Management Journal

    Does executive sponsorship matter for realizing project management value? member content locked

    By Chandler, Dawne E. | Thomas, Janice Executive sponsorship is critical to project success; however, the impact of the sponsorship role on project management value and sustainability is not known. Using correlation analysis we examine…

  • Learning from the East African experience member content open

    By Egginton, Bill E. | Fitz-Gerald, Ann M. The paper draws on detailed thematic analyses of the lessons learned data from 53 projects and programs in Ethiopia (24), Kenya (5), Uganda (4), Somalia (2), Sudan (2), Rwanda (2), and 13 other…

  • PM Network

    Under contract member content open

    By Jedd, Marcia Project management's rise in prominence within the corporate world has encouraged organizations to rethink the nature of project contracts. Based on the white paper "An ounce of prevention . . .…

  • Transformation program management member content open

    By Saliunas, Joseph Organizations initiate and implement transformation programs to change the way they use people, processes, and technology. To successfully accomplish such an initiative, however, organizations must…

  • Closing the gap between project management and governance member content open

    By Davis-Muffett, Patricia | Kerr, Keith Competing in today's dynamic global marketplace demands that organizations better govern their project performance to maximize revenue, profits, and value. And the most effective way to accomplish…