The effects of organic and mechanistic control in exploratory and exploitative innovations

G Model YMARE-494; No. of Pages 20

ARTICLE IN PRESS Management Accounting Research xxx (2013) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Management Accounting Research journal homepage: www.elsevier.com/locate/mar

The effects of organic and mechanistic control in exploratory and exploitative innovations Mika Ylinen a,∗ , Benita Gullkvist b,1 a b

Vaasan ammattikorkeakoulu, University of Applied Sciences, Raastuvankatu 31-33, FI-65100 Vaasa, Finland Hanken School of Economics, Handelsesplanaden 2, FI-65100 Vaasa, Finland

a r t i c l e

i n f o

Keywords: Management control systems Exploitation Exploration Innovation project Performance Tension

a b s t r a c t This study investigates the indirect effects of mechanistic and organic types of control on project performance acting through innovativeness in exploratory and exploitative innovation projects. It also examines the interaction effect of these controls on performance. The research model is empirically tested with survey data from 119 projects in various project organizations, using Partial Least Squares (PLS) with controls for the size of the project and task uncertainty. The results illustrate that organic control, acting through innovativeness on project performance is an important form of control in exploratory innovations, and also enhances performance in exploitative innovations. In addition, the results indicate that the interaction effect of organic and mechanistic control types enhances performance in both exploratory and exploitative innovation projects, suggesting a complementary effect. The findings are discussed in relation to theory and their managerial implications. © 2013 Elsevier Ltd. All rights reserved.

1. Introduction Scholars have long considered innovation a major determinant of organizational long-term performance (e.g., Bisbe and Otley, 2004; Kanter, 2001) and an effective management of innovation projects is a challenge facing today’s organizations (e.g., Jansen et al., 2006; Tushman and O’Reilly, 1996). Empirical studies investigating the innovation–performance relationship have also suggested that the relationship’s strength is moderated by the type of innovation (Calantone et al., 2010). As an innovation project is the most widespread vehicle for organizing and managing innovation activities (Chiesa et al., 2009; Martino, 1995), this study takes exploratory and exploitative innovation projects as its unit of analysis. Exploratory (radical) innovations cause fundamental, revolutionary

∗ Corresponding author. Tel.: +358 40 011 3998. E-mail addresses: mika.ylinen@puv.fi (M. Ylinen), benita.gullkvist@hanken.fi (B. Gullkvist). 1 Tel.: +358 (0)40 3521727.

changes in technology and represent clear departures from existing practice (Ettlie et al., 1984) by developing new products and services for emerging customers or markets and pursuing new knowledge. In contrast, exploitative (incremental) innovations are other changes in products and processes, which are generally less significant or which do not introduce considerable novelty (OECD, 2004) as they extend existing products and services for existing customers and build on existing knowledge (Benner and Tushman, 2003). Previous research has asserted that control mechanisms exert differing influences on exploratory and exploitative innovations (e.g., Benner and Tushman, 2003; Davila et al., 2009b; Hill and Rothaermel, 2003), but empirical studies examining such relationships have produced mixed results (Cardinal, 2001; Damanpour, 1991; Dewar and Dutton, 1986; Ettlie et al., 1984; Jansen et al., 2006). For example, results by Cardinal (2001) at the organizational level show that input, behavior, and output control enhance exploratory (radical) innovation, and input and output controls enhance exploitative (incremental) innovation, and

1044-5005/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.mar.2013.05.001

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20 2

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

Cardinal concluded that incremental and radical innovations should not be managed differently. Conversely, results by Jansen et al. (2006) at an organizational unit level indicate that centralization negatively affects exploratory innovation; formalization positively influences exploitative innovation; and connectedness (social relations among unit members) appears to be an important antecedent of both exploratory and exploitative innovations. Thus, the issue of whether exploratory and exploitative innovations require different control mechanisms remains largely unresolved. Examining these innovations separately, but within the same empirical study, offers a means to analyze whether project controls differ across innovation projects. Drawing on the classification in Chenhall (2003), this study adopts the concepts of the mechanistic control (MC) and organic control (OC) forms of project control mechanisms to represent two opposing forms of control. Mechanistic project controls rely on formal rules, standardized operating procedures and routines, whereas organic project controls are more flexible, responsive, involve fewer rules and standardized procedures and tend to be richer in data (Chenhall, 2003). Organic project control as used here reflects two important characteristics: (i) informal control reflecting norms of cooperation, communication and emphasis on getting “things done”, and (ii) open channels of communication and free flow of information between project manager and subordinates (Burns and Stalker, 1961). Prior studies (Burns and Stalker, 1961) maintain that a formal management control system (MCS) supports the periodic execution of the same routines in organizations where changes are small or non-existent. Empirical evidence also confirms this (e.g., Ouchi, 1979). In this regard, mechanistic forms of project controls would appear to be of little relevance to the innovation process associated with high level of uncertainty. These limitations proposed for the traditional MCS have, however, been questioned and proved unfounded in more recent studies, as researchers find that these systems may be important in providing the discipline to help manage uncertainty, and show that there is also a need for formal MCSs in uncertain settings, such as project environments (see e.g. Abernethy and Brownell, 1999; Bisbe and Otley, 2004; Cardinal, 2001; Davila et al., 2009a). Furthermore, Adler and Borys (1996), distinguishing between coercive and enabling bureaucracies, found that an MCS may be instrumental to innovation, and Simons (1995) that an interactive systems concept can play an explicit role in sparking innovation around strategic uncertainties. Thus, for the most part recent empirical evidence indicates that innovation processes may gain from the presence of an MCS. More recent studies have also suggested that opposing control mechanisms should be implemented simultaneously to foster innovativeness and performance (e.g., Chenhall and Morris, 1995; Henri, 2006; Lewis et al., 2002; Sheremata, 2000). Despite prior studies, scholars claim that there is little systematic evidence of potential indirect effects or whether the effects of one form of control are governed by the level of simultaneous reliance on another form of control (Abernethy and Brownell, 1997; Malmi and Brown, 2008).

Moreover, although scholars generally agree that innovation contributes to firm performance and that the understanding of innovation and control issues requires a unit of analysis other than the organizational level (e.g. Davila et al., 2009b), there are few accounting studies that have investigated the relevance of MCSs in project environments (Chenhall, 2008). In projects resembling temporary matrix organizations that draw on resources from many functions and are characterized by a high level of uncertainty (Tatikonda and Rosenthal, 2000a), project managers may face issues managing the dynamics of their project teams. That is because innovation and development require a high degree of flexibility in the structural and communication processes (Burns and Stalker, 1961; Van de Ven, 1986) as well as efficiency. Therefore, drawing on Dougherty (1996), it is suggested that a focus on the relationships between project controls, innovativeness and performance at the project level permits a more thorough treatment of the particular project controls acting at this level and will likely produce greater stability in the proposed relationships. Therefore, the objective of this study is to examine the effects of mechanistic and organic forms of control on project performance through innovativeness in exploratory and exploitative innovation projects. Innovativeness or innovative accomplishments are here defined very broadly to include any policy, structure, method or process, product or market opportunity that the project manager perceives to be new (Kanter, 1983; Zaltman et al., 1973). In comparison, innovation in addition to novelty also comprises commercialization and implementation of accomplishments (e.g., Dewar and Dutton, 1986). Adopting the approach introduced by Gupta and Govindarajan (1984), project performance was measured by comparing actual project performance and a priori expectations rather than measuring it on an absolute scale. By assessing project performance relative to targets and other projects, the effects of strategic choice on project performance are indirectly controlled for. The current research develops a conceptual model and tests it through PLS analysis on a sample of 119 projects, divided into two sub-samples: exploitative and exploratory settings. Previous studies (e.g., Bisbe and Otley, 2004; Jansen et al., 2006) suggest an indirect positive effect of an organic form of control on performance through innovativeness in exploratory and a similar effect brought about by a mechanistic form of control in exploitative projects. Moreover, prior research (e.g., Chenhall and Morris, 1995; Henri, 2006; Lewis et al., 2002) indicates that performance within different innovation projects can be enhanced by the effects of combined use of organic and mechanistic project control. Although prior research on opposing control forces in exploratory innovation settings does exist (e.g., Lewis et al., 2002; Sheremata, 2000), empirical research reporting on the indirect and interaction effects of opposing forms of project control in both exploratory and exploitative innovative project settings was not found. Thus, this study contributes to literature by extending prior research in MCSs (Chenhall and Morris, 1995; Bisbe and Otley, 2004; Henri, 2006; Jørgensen and Messner, 2009; Mundy, 2010)

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

to another level of analysis—the project level. In particular, and contrary to the findings of Bisbe and Otley (2004) at the organizational level, this study reports indirect effects of organic project controls on performance via innovativeness in exploratory settings. This study also offers empirical evidence of existing tension, resulting from the joint use of mechanistic and organic forms of control that influence performance in exploratory and exploitative innovation project settings, and thus clarifies how mechanistic and organic forms of control interact at the project level. Furthermore, examining project controls separately in exploratory and exploitative innovation projects makes it possible to study possible differences and thus enhances our understanding of the role of MCS in two different innovation settings. The remainder of this paper is structured as follows: Section 2 provides a brief overview of extant literature, resulting in the formulation of the hypotheses. Section 3 presents the design of the empirical survey study conducted to collect data. Section 4 reports the tests of the hypothesis. Sections 5 and 6 conclude the paper with a discussion of the findings and their implications as well as limitations and directions for future research. 2. Literature review and development of the hypotheses The increasing number of innovation projects in today’s business has highlighted the need for research on the applicability of MCSs for managing innovations (e.g., Davila, 2000; Davila et al., 2009b). Such systems can facilitate coordinating and controlling the project process and stimulating dialog and idea generation to enhance innovativeness and project performance (Davila et al., 2009b; Dougherty, 1996; Lewis et al., 2002). Scholars emphasize that exploitative and exploratory innovations may need different management (Rogers, 1995; Van de Ven et al., 1999) as they require different degrees of change, most likely stemming from a different mix of environmental, organizational, managerial, and structural forces. For example, Van de Ven et al. (1999) argue that different degrees of novelty need to be managed differently and maintain that structural variables decreasing the degree of radical (exploratory) innovation may simultaneously increase the degree of incremental (exploitative) innovation. Furthermore, both exploratory and exploitative innovations generate knowledge, but for different purposes, and the extent of new knowledge differ (Un, 2010). Dewar and Dutton (1986) argue that exploratory (radical) innovations contain a high degree of new knowledge, whereas exploitative (incremental) innovations build on a low degree of new knowledge. Moreover, especially in the early stages, radical change creates a high degree of uncertainty in the exploratory project as well as the whole organization, whereas the level of uncertainty is regarded as much lower in incremental (exploitative) innovations (e.g., Chiesa et al., 2009). Information processing theory suggests that an increasing level of uncertainty will increase the need for information and coordination of the project efforts (Olson et al., 1995). It has been suggested that MCSs are effective tools for managing uncertainty,

3

because they provide the information necessary to close the information gap between the information required for a particular task and that already available (Davila, 2000; Tushman and Nadler, 1978). As the degree of uncertainty will also vary with innovation type, it is likely that the need for information as well as for coordination and control will differ in different types of innovations (Davila, 2000; Olson et al., 1995). Whereas some researchers dismiss the use of formal controls as totally incompatible with innovation (e.g., Ouchi, 1977) there are those who advocate more organic forms of control—meaning a fluid style of management dealing with issues as they emerge—to foster creativity and improvisation (e.g., Dougherty, 1992). At the other end of the scale, are those who prescribe more mechanistic forms of control like disciplined planning as a way to focus on and speed project efforts (e.g., Wheelwright and Clark, 1992; Zirger and Maidique, 1990). For example, Chiesa et al. (2009) found that project managers, in response to the greater uncertainty inherent in exploratory projects, appeared to be more determined to adopt boundary and interactive control mechanisms, a leaning manifested in more frequent internal meetings. Simons’ framework (1995) emphasizes the importance of the interactive use of formal MCS in ensuring the success of innovation initiatives, but does not clearly distinguish between different, conceptually distinct, types of potential effects of the use of MCS, nor is such research known of in regard to project innovation and performance. Furthermore, the more recent findings of Bisbe and Otley (2004) relating to an organizational level do not support the positive effect of an interactive use of MCS on innovation. They found that the relationship between innovation and performance was moderated by the style of use of the MCS. Proponents of either organic or mechanistic forms of control have, however, also recognized the need for a balance between the competing roles of controlling the achievement of predetermined goals while simultaneously enabling employees to explore alternative options to solve problems and develop new suggestions (e.g. Ahrens and Chapman, 2004; Chenhall and Morris, 1995; Simons, 1995). Lewis et al. (2002) suggest that managing tensions between flexibility and discipline may provide a key to high project performance, and argue for a blend of organic and mechanistic forms of control, or as they put it, “emergent and planned approaches”. In project environments with wellknown operations, where stability and conformity are appreciated, tension may, however, be less useful and disturb well-established organizational routines, roles and internal processes (Henri, 2006). While there is a growing body of empirical research on the tension between the use of mechanistic (diagnostic) and organic forms of MCSs (e.g., Ahrens and Chapman, 2004; Chenhall and Morris, 1995; Henri, 2006; Jørgensen and Messner, 2009; Marginson, 2002; Mundy, 2010; Widener, 2007), previous studies have largely been conducted at the organizational level. The current research aims to increase understanding of the use of project controls in innovation environments and to examine two potential effects of MCSs. To do so, a conceptual model is developed. The proposed research model is illustrated in Fig. 1 and explained below.

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20 4

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

Organic control (OC)

Innovation type (Degree of newness)

Innovativeness

Mechanistic control (MC)

Control variables

Project performance

MC x OC

Direct effect ____ Moderating effect -----Fig. 1. Theoretical model. Direct effect (—), moderating effect (-----).

Regarding the direct effects of OC on project performance, prior research provides limited empirical evidence that may be divided into two views. The first view maintains that OC, implying an increasing flow of information, higher transfer of knowledge as well as shared interpretations of project goals and the conduction of project tasks (Ayers et al., 1997; Turner and Makhija, 2006), will provide project members with the flexibility and opportunity to adjust their outcome so as to increase the market value of the product or service provided (Rijsdijk and van den Ende, 2011). The second view argues for a negative effect on project performance, as OC is seen to lead to unfocused efforts among project members, resulting in the loss of innovative design features, which would hamper the success of a new product or service in particular (Ayers et al., 1997). Regarding the direct effects of MC on project performance, prior studies have also found contradictory results. For example, one line of research (Dvir et al., 2003; Shenhar et al., 2002) has found that a high level of formality and detailed project planning regarding schedules, responsibilities, budgets and goals is especially important for complex projects involving high uncertainty and will increase project performance, whereas Song and MontoyaWeiss (1998) found that detailed project planning could have negative effects on the outcome of highly innovative projects. Further, Tatikonda and Montoya-Weiss (2001) argue that the positive impact on project operational outcomes is evident irrespective of the technical uncertainty faced by the product or service. While the research model contemplates the possibility that the use of some form of OC and MC might directly influence project performance, no formulation of hypotheses for the potential direct effects was developed, as neither prior research evidence nor the theoretical development would provide clear arguments for a potential direct effect. Consistent with reasoning by Bisbe and Otley (2004), the potential direct effect is also expected to be relatively small and the major proportion of the potential relationship between the use of an MCS and performance is proposed

to come indirectly through innovativeness rather than through a direct effect. The direct effect is, however, controlled for in the model. 2.1. The organic form of project control in exploratory innovations Dent (1990) proposes that control systems can stimulate and foster curiosity and experimentation. In particular, the more OC mechanisms build on communication and decision processes that are participative, flexible and open, and therefore that can enhance opportunities to identify problems or new ideas (Kamm, 1987; Morse and Lorsch, 1970). Organic forms of control are also needed to ensure that individuals are motivated to participate in creative decision making and to provide the free flow of ideas that is essential for developing entrepreneurial strategies (Burns and Stalker, 1961) or helping focus attention on strategic issues (Simons, 1995). Scholars (e.g., Montoya-Weiss and Calantone, 1994; Olson et al., 1995) have also suggested that OC may be best suited for project environments, because decentralization, autonomy and empowerment may lead to conflict resolution and effective decision making, whereas more formalized, mechanistic approaches may inhibit the diffusion of ideas among project team members. Providing an organic form of control entails informal communication and the free flow of information, which also enables project team members to explore alternative ways to solve problems (e.g., Chenhall, 2003; Wageman, 2001) and develop new suggestions (Wilson et al., 2007), which in turn form a basis for knowledge generation and foster innovativeness. We suggest that this is especially important in exploratory projects, characterized by high uncertainty and radical change. Moreover, organic forms of project control might provide the means for project managers to indicate to team members that innovativeness is highly valued and consistent with project aims. Therefore, it is proposed that the extent of the use of organic project

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

control in exploratory innovation projects will be positively related to innovativeness and eventually to project performance. Accordingly, the following hypothesis is proposed: H1a. There is a positive effect of OC on innovativeness in exploratory innovations. In contemporary management research, empirical studies have repeatedly confirmed a positive relationship between, for example, product innovativeness and project performance in terms of new product success and failure (e.g., Calantone et al., 2010; Kleinschmidt and Cooper, 1991; Song and Montoya-Weiss, 1998). In addition to its direct effect on performance, Van de Ven and Polley (1992) emphasize that learning during the innovation process generates absorptive capacity (defined by Cohen and Levinthal (1990) as the capability to identify, assimilate, and apply knowledge) through which competitive advantages (Zahra and George, 2002) are generated. Although exploratory innovations in the early stages of the innovation process, might have little or no economic impact (Popadiuk and Choo, 2006), the effect of innovativeness on performance has generally been regarded as positive, and mainly long-term (e.g., Lewis, 2000; Tushman and O’Reilly, 1996). Researchers (e.g., Droge et al., 2008) argue that for example innovating firms with unique knowledge, capabilities and superior products/services should record high performance. Likewise, it is here proposed that exploratory projects, drawing on new ideas and developing new products, processes and so forth, will contribute to strong project performance. The following hypothesis is proposed: H1b. There is a positive effect of innovativeness on project performance in exploratory innovations. In summary, if the organic form of project control can be linked to innovativeness and innovativeness can be linked to project performance, then the use of OC can be expected to have implications for project performance through an induced increase in innovativeness. As a result, this study anticipates OCs acting through innovativeness having an indirect effect on project performance in exploratory innovation. 2.2. The mechanistic form of project control in exploitative innovations Mechanistic control, in its various forms, provides formal rules and regulations for project team members, which may enhance project efficiency (Lewis et al., 2002; Shenhar and Dvir, 1996). Milestones based on schedules or budgets help ensure teams are more aware of scarce resources (Wheelwright and Clark, 1992) and guide decision making regarding resource allocation (Lewis et al., 2002). In addition, a mechanistic form of control, particularly a more directive one, may also define project boundaries and thus reduce the probability of wasteful explorations and costly errors (McDonough and Barczak, 1991). Whereas previous research has traditionally associated MCSs with mechanistic organizations (Burns and Stalker, 1961), and suggested that they are of little or no use in research and development settings (Davila et al., 2009a), more recent studies provide

5

empirical evidence of the need for MCs in uncertain situations such as project environments (e.g., Cardinal, 2001; Davila et al., 2009a). Contemporary research on the effects of MC on innovation has indicated a positive effect of mechanistic organizational structures (Calantone et al., 2010; Jansen et al., 2006). Jansen et al. maintain that rules and procedures might not be as detrimental to innovation efforts as was previously assumed. In exploitative innovations, MC can facilitate the generation of proposals to improve existing routines; make existing knowledge and skills more explicit, and diminish possible variance through improvements in existing processes and outcomes (Benner and Tushman, 2003). Furthermore, MC can be used to codify best practice, making it easier to adopt more efficiently (Zander and Kogut, 1995). Chiesa et al. (2009) reported that managers in exploitative projects, less affected by uncertainty, relied more on formalized control systems throughout the project compared to managers in exploratory projects. In summary, high levels of MC are likely to produce inertial forces and a focus on exploitation (e.g., Cardinal, 2001) as well as aid planning and restrain excessive innovation (Chenhall, 2003). Therefore, drawing on the above findings, a positive effect of MC on innovation in exploitative innovation is proposed. The above analysis generates the following hypothesis: H2a. There is a positive effect of MC on innovativeness in exploitative innovations. Regarding the relationship between innovativeness and performance in exploitative environments, scholars have suggested that exploitative innovations engender less uncertainty and build on firm synergies and existing knowledge, through which performance is generated (e.g., Calantone et al., 2010). In addition, although the returns from exploitative innovations may be considered shortterm, they are typically positive and more predictable than those from exploratory innovations (Menguc and Auh, 2008). Accordingly, this study also hypothesizes that exploitative innovations will have a positive effect on project performance, as formally stated in the following hypothesis: H2b. There is a positive effect of innovativeness on project performance in exploitative innovations. To sum up, if the use of MC can be associated with innovativeness and innovativeness can be related to project performance, then the use of MC can be expected to affect project performance through inducing increased innovativeness. Therefore, MC is proposed to have an indirect effect on project performance through innovativeness in exploitative innovations. 2.3. Combination of project controls Scholars maintain that control mechanisms may behave as either substitution or complementary controls (Fisher, 1995). This means that control mechanisms can be effective and achieve the same desired outcome individually, in other words, one control may be substituted for another (a substitution effect) or two or more controls may reinforce

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20 6

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

each other in pursuit of the desired outcome (complementary controls), making both controls necessary to achieve the desired results (Widener, 2004). Milgrom and Roberts (1995) illustrated analytically that control mechanisms can be complementary, as an increase in the emphasis on one control component increases the benefits gained from increasing the emphasis on another control. Investigating controls in combination is therefore important because a control may add to, detract from, or multiply the effects of another. A combination can also bring about equifinality, in which different combinations of control practices produce equivalent results (Anderson and Dekker, 2005). While researchers agree on the general assumption that MCSs are inter-dependent (e.g., Anderson and Dekker, 2005; Milgrom and Roberts, 1995; Widener, 2007), empirical studies have found evidence of both complementary and substitutional activity. In the LOC framework (Simons, 1995) control of business strategy is achieved by integrating the four levers of beliefs systems, boundary systems, diagnostic control systems, and interactive control systems. Simons suggests that these four levers create tension as follows: the beliefs and interactive control system create positive energy, and the boundary and diagnostic control systems create negative energy. He also shows how opposing forces such as diagnostic and interactive control systems may complement each other over time in the implementation of strategic change, and balance inherent organizational tension. Widener (2007) provides empirical evidence on the relations among the various control systems in the LOC framework and finds complementary effects. She reports that when enterprises emphasize their belief system, they also emphasize each of the three other control systems. Henri (2006) finds empirical evidence for managers using performance measures in both a diagnostic and interactive way and of that use resulting in a desirable state of dynamic tension enhancing organizational capabilities and performance. Drawing on the rationale of positive and negative energy (Henri, 2006; Simons, 1995), the organic form of control, stimulating the generation of new ideas, represents a positive force when OC is used to expand opportunity seeking and learning within and throughout the project. It represents a negative force when used to mold team members’ actions to predetermined goals and curb what is deemed excessive innovation activity. Likewise, the mechanistic form of project control, including monitoring, coordinating and focusing on pre-set goals and correcting performance deviations from those goals, may represent both positive and negative forces. The focus on deviations may provide a negative or positive feedback signal to the project manager, and the possible corrective actions undertaken will be either positive or negative, i.e. the reverse of the feedback signal to correct the innovation process. Thus, OC and MC have different purposes, but could be used simultaneously in the project environment to manage inherent tension within the project. This is exemplified as follows. While MCs, providing formal rules and regulations for project team members, are useful in assisting planning and curbing excessive innovation (Chenhall, 2003), team members might increasingly come to view those same rules

and regulations as rigid and repressive (Dougherty, 2006; Abernethy and Brownell, 1997). Consequently, knowledge generation will suffer (Amabile et al., 2005), which would have a negative effect on team innovation and project performance. Similarly, providing OC entails informal communication and free flow of information, which would enable project team members to explore alternative options to solve problems (e.g., Chenhall, 2003; Wageman, 2001) and develop new suggestions (Wilson et al., 2007). Nevertheless, if only OCs are provided and at the same time little or no mechanistic control is applied, the team members may utilize the available freedom too expansively and the increased search for alternative opportunities may reduce the level of coordination within the project team (Sheremata, 2002). That in turn can lead to unfocused and ineffectual efforts which would not promote project performance. Thus, when considering the above separate effects of organic and mechanistic project controls, it appears likely that a combination of opposing forms of project control would yield a balanced set of forces enhancing and impeding project performance. Although both OC and MC may yield effects when applied individually, and the combined use of OCs and MCs is likely to create either negative or positive tension (conflict, paradox) due to the different controls being contradictory (Lewis, 2000), we consider the simultaneous implementation of opposing OC and MC necessary to enhance project performance, because the combination should create a positive synergy that enhances project performance. For example, an opposing form of control is called for to hinder the negative effect of relying solely on OC. Mechanistic forms of control, through coordinated action plans, will have an indirect positive effect by preventing the exploration of alternative undesired options and ideas, something that could otherwise escalate into negative spirals (Argyris and Schön, 1983; Sundaramurthy and Lewis, 2003), and ultimately make the project team unmanageable. These direct and indirect effects of OC and MC create the necessary imbalance of forces that is believed to favor innovativeness, and also make project performance more focused and effectual. Therefore, mechanistic and organic forms of control are believed to mutually reduce the undesired side effects of the respective opposing control mechanism and enable the positive effects of the respective opposite strategy to be fully realized to create synergy. For this reason, a multiplicative (rather than an additive) effect of these two types of project control on project performance is posited in this study. Following previous research, which have has mostly found complementary effects of control combinations (e.g., Anderson and Dekker, 2005; Widener, 2007), this is considered a likely outcome in this study, too. For example, in exploitative innovation projects, where MC is essential for controlling and coordinating innovation projects and keeping them on track, OC could fulfill a complementary role by assisting MC through improving flexibility and enabling the adaptability required when dealing with unforeseen events (Turner and Makhija, 2006). Thus, the complementary role of OC would strengthen the effect of MC and help achieve the desired project performance outcomes. Likewise, in

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

exploratory projects, MC could provide additional benefits on top of OC to enhance project performance. Despite this reasoning, the technical tests of the relationships around the two control dimensions in this study follow an interaction approach, which will permit either relation. Thus, what is tested is the more general proposition that the two project controls used jointly produce a positive synergy effect on project performance. In summary, in this study OC and MC are believed to mutually reduce the undesired side effects of the respective control mechanism and thus enable the positive effects of the respective opposite strategy to be fully realized to create synergy. Positive effects of tension might be fully realized in project environments where simultaneous requirements for both control mechanisms exist. This is proposed to hold true for both exploratory and exploitative innovations. Therefore, the third hypothesis of this study is as follows: H3. There is a positive interaction effect of OC and MC on project performance in exploratory and exploitative innovations. The research model was controlled for perceived task uncertainty and project characteristics such as size, here measured as the number of employees working on the project and the duration of the project. Previous research has consistently shown that innovation projects are characterized by considerable uncertainty and ambiguity, and that exploratory innovations show a higher level of uncertainty throughout the development phase than exploitative projects (e.g., Chiesa et al., 2009). Prior studies have also used the level of uncertainty to distinguish between exploratory and exploitative innovation projects. Furthermore, previous research has found team size to be related to cohesiveness and internal communication for groups (e.g., Ancona and Caldwell, 1992). Scholars have also found that larger groups show greater potential for team conflict (Williams and O’Reilly, 1998), which may result in reduced team performance (De Dreu and Weingart, 2003). In addition, project duration is used here as a proxy for project size, because long-term projects will necessitate extensive use of tangible and intangible resources. 3. Research design The research model was tested with a stratified sample of project organizations randomly extracted from the database of the Project Management Association of Finland. True random sampling, requiring the a priori establishment of the population from which a sample could be drawn, was not possible as detailed and complete public data on projects or project managers were not available. Thus, there is a danger that the sample will not represent the broader population. Data were collected by a survey questionnaire administered to 240 project managers from 145 different project organizations. Project managers of innovative projects were initially targeted with a personally addressed email providing response instructions and guaranteeing anonymity. The initial request was supported by one reminder email, following which, 119 usable responses were received, a

7

final response rate of approximately 50%. The respondents were asked to classify their project type as either exploratory or exploitative. Of the projects, 68 represented exploratory innovation projects and 51 exploitative innovation projects. The relatively high number of exploratory innovation projects in this study may be explained by the fact that the majority of the responding project managers worked in high-tech industries. A comparison of means on all the measured variables was undertaken to test for potential non-response bias by comparing the mean responses of the questionnaires received prior to the reminder email to those received after the reminder email. A two-sample t-test revealed no significant differences, indicating that non-response bias is unlikely to affect the results. The project organizations in this study represent a variety of industries including information technology (28 projects), telecommunications (20 projects), consulting (11 projects), automation (9 projects), electronics (9 projects), metal industry (7 projects), construction (7 projects), engineering (6 projects), banking and financial services (6 projects) and miscellaneous (11 projects). Descriptive statistics indicate the following data for explorative innovation projects and exploitative innovation projects respectively: the average previous project experience of the project manager was 8.61 years and 8.82 years; the average project budget was EUR 1.11 million for the explorative projects and EUR 1.47 million for the exploitative; the average project duration was 15.21 months and 11.25 months; and finally, the average number of employees working on an explorative project was 14 as against 10 employees for an exploitative one. Variables were measured using multiple indicators and operationalized through multi-item constructs on Likerttype scales. Established and reliable scales for measuring were used, with a modification to fit the present research context. To check on the relevance of these measures, the web-questionnaire was pre-tested on five academics and four project managers with expertise in the areas of management control and project management. Final measures were then developed and refined. The measurement instrument is provided in Appendix A and the descriptive statistics in Table 1. Degree of innovation newness (Innovation type) was measured using a dummy variable where different degrees of newness were condensed into the following two definitions: exploratory (radical) innovations aim to produce fundamental changes in the products, services, processes or activities of the organization and represent clear departures from existing practice; exploitative (incremental) innovations are projects that aim to produce only minor changes and thus require less departure from existing practices (Dewar and Dutton, 1986; Ettlie et al., 1984). Organic form of control (OC) was operationalized as a three-item construct modified from the instruments developed by Van der Stede (2001) and Chenhall and Morris (1995). The adopted items are considered to capture the communicative and interactive project control mechanisms and expected to measure the level of use of informal and face-to-face meetings as well as the free flow of information as one form of OC. The items were measured on a

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

ARTICLE IN PRESS

G Model YMARE-494; No. of Pages 20

M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

8

Table 1 Descriptive statistics on research variables. N = 119

No. of items used

Theoretical range

Min

Max

Median

Mean

SD

Organic control Mechanistic control Innovativeness Project performance

3 4 3 3

1–7 1–7 1–7 1–7

1.00 1.00 1.00 1.00

7.00 7.00 7.00 7.00

6.00 5.00 5.00 5.00

5.55 4.45 4.44 5.06

1.53 1.73 1.23 1.14

Exploratory (n = 68)

Minimum

Maximum

Median

Mean

SD

Organic control (OC) Mechanistic control (MC) Innovativeness Project performance Combination (OC × MC) Project duration No. of people Task uncertainty (a)

1.00 1.00 1.00 1.00 1.00 3.00 2.00 1.00

7.00 7.00 7.00 7.00 49.00 50.00 80.00 7.00

6.00 5.00 5.00 5.00 25.00 12.00 8.00 4.00

5.69 4.40 4.67 5.16 25.65 15.21 13.65 3.78

1.54 1.68 1.22 1.19 12.67 11.00 15.89 1.40

Exploitative (n = 51) Organic control Mechanistic control Innovativeness Project performance Combination (OC × MC) Project duration No. of people Task uncertainty (a)

1.00 1.00 1.00 2.00 1.00 2.00 3.00 1.00

7.00 7.00 7.00 7.00 49.00 42.00 50.00 7.00

6.00 5.00 4.00 5.00 24.00 9.00 7.00 3.00

5.37 4.50 4.13 4.94 24.70 11.25 10.33 3.53

1.50 1.82 1.17 1.07 12.84 8.22 8.72 1.31

(a) Reverse coded.

seven-point Likert scale, where 1 reflected “strongly disagree” and 7 “strongly agree”. The mechanistic form of control (MC) was operationalized as a four-item variable, measuring the extent to which the project manager has a detailed interest in specific project performance line-items in an evaluation and does not tolerate deviations from interim project performance targets. The instrument was developed based on previous research by Van der Stede (2001). The items were scored on a sevenpoint Likert scale with 1 reflecting “strongly disagree” and 7, “strongly agree”. Innovativeness (INN) was operationalized as a threeitem variable and measured using a modification of the job-performance instrument of Shields et al. (2000). Respondents were provided with the following definition of innovativeness: “Innovative accomplishments are defined here very broadly to include any policy, structure, method or process, product or market opportunity that you as the manager of the project perceived to be new”, and asked to indicate the level of the project team innovativeness (the number of innovations) on a seven-point Likert scale ranging from 1 indicating “extremely low” to 7 reflecting “extremely high”. Performance of development project (PERF) was measured using subjective self-assessment of a project’s performance relative to predetermined targets and other similar projects. Performance was operationalized as a multi-item construct capturing the most important dimensions relative to project performance targets (standards), other project teams and overall assessment. These items are a modification of the job-performance construct of Shields et al. (2000). The level of performance of the project team was scored on a seven-point Likert scale ranging from 1 indicating “extremely low” to 7 reflecting

“extremely high”. While the degree to which subjective self-ratings correspond to objective performance measures is debatable, evidence from prior research has shown that managers’ subjective self-ratings of performance and objective measures are highly correlated (see e.g., Shields et al., 2000). Similar approaches at the project level have been used by Sheremata (2002). The combined effect of MC and OC is operationalized as a product term between organic and mechanistic forms of control. Following previous research (e.g., Poppo and Zenger, 2002; Siggelkow, 2002), the interaction effects method for testing complementarity between control combinations and project performance was used. Prior research suggests that there will be evidence of complementarity between control combinations and project performance if the coefficient of the interaction term is positive. In the context of this study, a significant positive interaction effect between OC and a given type of MC would suggest that they are complementary and a negative one would suggest that they are substitutional. The control variable project size was operationalized as two project characteristics—the number of full-timeequivalent workers assigned to the project and the project duration measured in months. The second control variable, task uncertainty was operationalized as project task uncertainty resulting from the characteristics of the project team’s work process or set of value activities. Task uncertainty (TU) was measured using the three items representing task difficulty (task analyzability) of the measurement instrument developed by Withey et al. (1983). The items were anchored with “to a small extent” (1) and “to a great extent” (7), but the scale was revised in the analysis to reflect that the highest numbers accorded with the higher perceived TU.

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

4. Results The model was estimated using a structural equation modeling approach. The path analytic modeling technique of PLS was chosen because of the predictive nature of this study (Jöreskog and Wold, 1982) and because it is a method that can provide unbiased estimates with small sample sizes (Falk and Miller, 1992). The Smart PLS software (Ringle et al., 2005) was used in this study. In accordance with common procedure, the model was analyzed and interpreted in two stages (Barclay et al., 1995; Hulland, 1999): the first involved the assessment of the reliability and validity of the measurement model and the second, the assessment of the structural model. This was done to ensure that the constructs’ measures were reliable and valid before assessing the nature of the relations between the constructs (Barclay et al., 1995; Hair et al., 1998; Hulland, 1999). For the measurement model, each construct was modeled to be reflective. As the reflective measurement model requires, the indicators of each construct are highly inter-correlated (test of crossloadings—Table 4) and the indicators share a common theme, which makes them interchangeable (Chin, 1998). Causality is thus assumed to flow from each construct to its indicators. All items in the analysis are univariate normal, but some data were missing at the item level. As the amount of missing data was very small and the data missing completely at random, the missing data were replaced using the mean imputation approach (Hair et al., 1998).2 The PLS analysis of the current research contains multiplicative interaction terms, which are developed following the procedure outlined in Chin et al. (2003). Because interaction terms increase the potential for multicollinearity, all items reflecting the predictor and moderator constructs were standardized (m = 0; s2 = 1) (Aiken and West, 1991; Chin et al., 2003; Cronbach, 1987; Drazin and Van de Ven, 1985; Venkatraman, 1989). Doing so minimizes the degree of multicollinearity among the variables and improves the interpretability of the results. We tested for interaction effects using the product indicators procedure suggested by Chin et al. (2003). The statistical power of PLS in analyzing interaction effects with a product indicator approach has been confirmed using Monte Carlo simulation (Chin et al., 2003; Goodhue et al., 2007), and a recent study by Goodhue et al. (2007) suggests that significant interaction results revealed using the PLS product indicator approach are reliable.

9

ability. All items suggest good indicator reliability with few indicator loadings below 0.7. Convergent validity can be assessed by examining composite reliability (CR) and average variance extracted (AVE) measures (Fornell and Larcker, 1981; Hulland, 1999). As shown in Table 2, the CR values indicate good consistency of all constructs bar one, when using the threshold of 0.7 (Nunnally, 1978). The proportion of total variance in a construct that is extracted by the component set of indicator variables (AVE) should be greater than 0.50 (Chin, 1998; Hair et al., 1998). This is true for all but one variable, which is 0.38. Therefore, most measures indicate good convergent validity. Discriminant validity is evaluated by comparing the AVE of each construct and the variance shared between such constructs and other constructs in the model. Furthermore, the cross-loadings of indicators on other constructs should be minimal. Table 3 shows the correlations between different constructs in the lower left, off-diagonal elements of the matrix, and the square root of the AVE value calculated for each of the constructs along the diagonal (marked in bold). For adequate discriminant validity, as occurs in this analysis, the diagonal elements should be significantly greater than the off-diagonal elements in the corresponding rows and columns (Fornell and Larcker, 1981; Hulland, 1999). In addition, a matrix of cross-loadings was constructed to test discriminant validity on the item level (Table 4). All items demonstrated higher loadings on their associated constructs (marked in bold) compared with their crossloadings. While cross-loadings derived from this procedure will inevitably be higher than those derived from typical exploratory factor analysis (Gefen and Straub, 2005), the cross-loading differences were much higher than the suggested threshold of 0.1 (Gefen and Straub, 2005). The above assessment suggests that all scales behaved reliably in both groups, demonstrated satisfactory convergent and discriminant validity and exhibited adequate psychometric properties. In addition, the correlation analysis in Table 3 indicates a strong relationship between the two variables OC and MC. The control variable Task Uncertainty correlates strongly with the dependent variable PERF. A high level of TU is considered a typical feature of the development project environment, and has also been found to relate to project performance in previous studies (e.g., Tatikonda and Rosenthal, 2000b). 4.2. Common method bias

4.1. Measurement model Table 2 illustrates the measurement model parameters for both subgroups. Regarding item reliability, Fornell and Larcker (1981) suggest that loadings of indicators on latent constructs greater than 0.7 are sufficient to establish reli-

2 An examination of means, standard deviations, and correlations before and after the replacement revealed only minor differences. Further, the proposed research model maintains high robustness even after observations with missing variables were dropped.

As with all self-reported data, there is a potential for common method bias (CMB) resulting from multiple sources such as consistency motif and social desirability (Podsakoff et al., 2003; Podsakoff and Organ, 1986). Several actions were taken to address the potential threat of CMB. First, a Harmon one-factor test (Podsakoff and Organ, 1986) was conducted on the three variables in the theoretical model including MC, OC, INN and PERF. The results from this test showed that four factors are present and the most covariance explained by one factor was approximately 28%, indicating that CMB is not a likely contaminant of the results. Second, following Podsakoff et al.

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20 10

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

Table 2 Estimation of the measurement model parameters. Original sample

Sample mean

Standard deviation

Exploratory projects (n = 68) Organic control (OC) OC1 OC2 OC3

t-statistic

0.801 0.76 0.561

0.781 0.747 0.54

0.102 0.105 0.156

7.814 7.253 3.607

Mechanistic control (MC) MC1 MC2 MC3 MC4

0.775 0.839 0.672 0.705

0.710 0.775 0.616 0.656

0.210 0.154 0.182 0.166

3.696 5.447 3.700 4.260

Innovativeness Inn1 Inn2 Inn3

0.872 0.754 0.930

0.873 0.742 0.932

0.032 0.074 0.013

27.017 10.242 71.082

Project performance Perf1 Perf2 Perf3

0.874 0.836 0.874

0.879 0.833 0.873

0.026 0.043 0.070

33.753 19.360 12.533

Control variable TU1 TU2 TU3

0.710 0.793 0.755

0.693 0.773 0.733

0.136 0.117 0.133

5.225 6.745 5.672

Exploitative projects (n = 51) Organic control (OC) OC1 OC2 OC3

0.538 0.719 0.565

0.473 0.668 0.544

0.231 0.248 0.272

2.324 2.901 2.078

Mechanistic control (MC) MC1 MC2 MC3 MC4

0.523 0.800 0.754 0.765

0.536 0.760 0.737 0.718

0.163 0.127 0.122 0.125

3.204 6.280 6.175 6.124

Innovativeness Inn1 Inn2 Inn3

0.615 0.898 0.966

0.666 0.827 0.885

0.223 0.146 0.167

2.761 6.140 5.770

Project performance Perf1 Perf2 Perf3

0.926 0.821 0.934

0.924 0.825 0.932

0.014 0.042 0.012

65.866 19.516 76.626

Control variables TU1 TU2 TU3

0.789 0.909 0.714

0.777 0.903 0.712

0.085 0.044 0.098

9.299 20.590 7.304

AVE

CR

0.511

0.754

0.563

0.836

0.731

0.890

0.742

0.896

0.567

0.797

0.375

0.639

0.517

0.807

0.706

0.875

0.801

0.923

0.653

0.848

Loadings of indicators on latent constructs (original sample), 0.7 or above indicates good indicator reliability. AVE: average variance extracted, 0.5 or above indicates good convergent reliability. CR: composite reliability, 0.7 or above indicates good convergent reliability.

(2003) and Williams et al. (2003), a common method factor was included in the PLS model. The indicators of all constructs were associated reflectively with the method factor. Next, each indicator variance explained by the principle construct and by the method factor was computed.3 As shown in Appendix B (Table B1), the results demonstrate that the average substantively explained variance of the indicators was 0.669, while the average method based variance was 0.006. The ratio of substantive variance to

3 This study employed the analytical procedure previously used by Liang et al. (2007) and Hsieh et al. (2008).

method variance was about 111:1. In addition, all method factor loadings bar one are non-significant. Given the small magnitude and insignificance of method variance, evidence collectively suggests that common method bias is not a significant issue in this study. 4.3. Structural model The second step in the PLS analysis is the estimation of the specified structural equations. As PLS does not make any assumption about data distribution, the statistical significance of examined paths is evaluated by nonparametric techniques. Bootstrapping with replacement

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

11

Table 3 Discriminant validity coefficients. Exploratory projects (n = 68) OC

MC

Inn.

Perf.

Duration

No.of people

TU

Organic control (OC) Mechanistic control (MC) Innovativeness (INN) Project performance (PERF) Project duration No. of people Task uncertainty (TU)

0.715 0.493** 0.357** 0.220 −0.131 0.096 −0.054

0.751 0.093 0.243 −0.217 0.172 −0.256

0.855 0.429** 0.061 0.032 −0.041

0.862 −0.020 −0.053 −0.331**

N/A 0.304* 0.196

N/A 0.203

0.753

Exploitative projects (n = 51) Organic control (OC) Mechanistic control (MC) Innovativeness (INN) Project performance (PERF) Project duration No. of people Task uncertainty (TU)

0.612 0.437** 0.157 0.353* −0.003 0.221 −0.141

0.719 0.087 0.276 −0.023 0.336** −0.391**

0.841 0.268 0.129 0.036 0.014

0.895 −0.215 0.158 −0.422**

N/A 0.157 −0.049

N/A −0.361**

0.808

The square root of the AVE value for each of the constructs along the diagonal (in bold). Correlations between different constructs in the lower leftoff-diagonal elements of the matrix. * p < 0.050 (two-tailed test). ** p < 0.010 (two-tailed test).

Table 4 Cross-loadings. Exploratory projects OC

MC

Inn

Perf.

Duration

No. people

TU

0.801 0.760 0.561 0.318 0.404 0.224 0.442 0.225 0.214 0.424 0.224 0.180 0.162 −0.131 0.097 −0.073 0.039 −0.110

0.530 0.204 0.287 0.775 0.839 0.672 0.705 0.061 0.031 0.124 0.269 0.161 0.193 −0.217 0.172 −0.162 −0.182 −0.234

0.348 0.208 0.173 −0.064 0.152 0.014 0.098 0.872 0.754 0.930 0.313 0.474 0.325 0.061 0.032 −0.097 0.097 −0.128

0.156 0.253 −0.009 0.210 0.219 0.029 0.152 0.354 0.295 0.430 0.874 0.836 0.874 −0.020 −0.053 −0.199 −0.290 −0.246

−0.085 −0.114 −0.094 −0.181 −0.201 −0.215 −0.101 0.006 0.053 0.086 −0.017 0.042 −0.077 1.000 0.304 −0.051 0.196 0.253

−0.024 0.166 0.123 −0.037 0.189 0.146 0.216 0.017 0.036 0.030 −0.013 −0.058 −0.069 0.304 1.000 0.108 0.123 0.228

−0.088 −0.026 0.055 −0.334 −0.213 −0.029 −0.076 −0.013 −0.036 −0.050 −0.378 −0.128 −0.345 0.196 0.203 0.710 0.793 0.755

Exploitative projects 0.538 OC1 0.719 OC2 0.565 OC3 0.138 MC1 0.413 MC2 0.166 MC3 0.456 MC4 0.088 Inn1 0.102 Inn2 Inn3 0.175 0.316 Perf1 0.284 Perf2 0.346 Perf3 −0.003 Duration People 0.221 −0.154 TU1 −0.088 TU2 −0.110 TU3

0.351 0.200 0.316 0.523 0.800 0.754 0.765 −0.110 0.106 0.075 0.264 0.221 0.254 −0.024 0.336 −0.322 −0.313 −0.318

0.154 0.085 0.087 −0.054 0.118 0.008 0.109 0.615 0.898 0.966 0.313 0.143 0.239 0.129 0.036 −0.301 −0.390 −0.324

0.060 0.265 0.247 0.121 0.239 0.214 0.194 −0.021 0.177 0.305 0.926 0.821 0.934 −0.215 0.158 −0.097 −0.068 0.047

−0.130 0.071 −0.022 0.216 −0.065 −0.026 −0.059 0.096 0.174 0.089 −0.238 −0.061 −0.244 1.000 0.157 0.051 0.104 −0.136

0.085 0.045 0.260 0.350 0.276 0.259 0.163 −0.071 0.028 0.041 0.078 0.246 0.134 0.157 1.000 −0.288 −0.314 −0.272

−0.065 −0.009 −0.182 −0.151 −0.364 −0.366 −0.194 0.200 0.046 −0.014 −0.395 −0.352 −0.386 −0.049 −0.361 0.789 0.909 0.714

OC1 OC2 OC3 MC1 MC2 MC3 MC4 Inn1 Inn2 Inn3 Perf1 Perf2 Perf3 Duration People TU1 TU2 TU3

Discriminant validity on item level indicated as higher loadings on associated constructs compared to cross-loadings.

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20 12

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

Table 5 PLS structural model results: path coefficients, t statistics, R2 and Q2 . Exogenous variables

Organic control (OC) → Perf Organic control (OC) → INN Mechanistic control (MC) → Perf Mechanistic control (MC) → INN Innovativeness (INN) → Perf OC × MC → Perf Control variables: Project size: duration Project size: no. of people Task uncertainty R2 (Perf) R2 (INN) Q2 (Perf) Q2 (INN)

Exploratory

Exploitative

Stage I

Stage II

Stage I

Stage II

−0.004 (0.050) 0.411 (4.862)*** 0.160 (1.702)

0.030 (0.377) 0.411 (4.636)*** 0.146 (1.533)

0.267 (2.445)** 0.147 (1.316) −0.031 (0.521)

0.254 (2.729)** 0.148 (1.345) 0.015 (0.256)

−0.110 (1.084)

−0.110 (1.140)

0.023 (0.222)

0.022 (0.213)

0.402 (4.239)***

0.413 (4.442)*** 0.218 (2.716)**

0.270 (2.554)**

0.233 (2.277)** 0.341 (4.501)***

0.060 (0.909) −0.056 (0.934) −0.274 (3.147)*** 0.303 0.136 −0.001 0.083

0.027 (0.454) −0.053 (0.881) −0.279 (3.162)*** 0.348 0.136 0.009 0.083

−0.269 (4.262)*** −0.009 (0.192) −0.417 (5.162)*** 0.391 0.025 0.094 0.016

−0.249 (4.238)*** −0.003 (0.060) −0.404 (4685)*** 0.505 0.025 0.092 0.016

Each cell reports the path coefficient (t-value). Stage I reports test results without interaction. Stage II reports test results with interaction. ** p < 0.010 (one-tailed tests). *** p < 0.001 (one-tailed tests).

using 500 samples (generated from the original dataset) was undertaken, as suggested by Chin (1998). Table 5 presents the structural model and outlines the results of the hypothesis testing. The structural model was independently tested for the exploratory and exploitative project subgroups. All control variables were included in the research model from stage 1, which represents the model without the interaction variable. The control variable measuring project size by number of people was not found to be significant at any stage of the analysis, whereas the duration of project variable was significant in exploitative projects in both stages. Furthermore, the control variable Task Uncertainty was found to have a significant effect on PERF in both subgroups and at both stages. Although no hypotheses were proposed for the direct effects of OC and MC on project performance, the results will be briefly reported. First, the test of the initial research model without the mediator variable INN indicates direct effects of OC on performance in exploratory (ˇ = 0.240, p < 0.05) and exploitative innovation (ˇ = 0.326, p < 0.01), but non-significant effects of MC (ˇ = 0.104, p > 0.05 and ˇ = −0.010, p > 0.05, respectively). Second, after including the mediator INN, the evidence indicates that there is a significant positive direct effect of OC on PERF in exploitative innovation projects (ˇ = 0.267, p < 0.01), whereas there is a significant direct positive relationship between MC and PERF in exploratory innovation projects (ˇ = 0.160, p < 0.05) (Stage I in Table 5). This would imply that OC enhances

project performance in exploitative settings by increasing the shared understanding and common values of project goals and, as a consequence, are likely to increase project efficiency and performance. In addition MC, including the focus on detailed performance evaluation and target deviations, ensures discipline and defines project boundaries, and is therefore likely to make the team aware of scarce resources, to reduce costly errors and thereby safeguard project performance in exploratory innovation projects (Chenhall and Morris, 1995). The results in Table 5 (Stage I) also indicate a significant path between OC and INN (ˇ = 0.411, p < 0.01) in exploratory innovations. Thus, OC seems to be useful in driving project innovativeness in exploratory projects, where innovation is highly valued and generation of new knowledge is more important. Moreover, the evidence indicates a significantly positive effect of INN on PERF in exploratory projects (ˇ = 0.402, p < 0.001, Table 5). Contrary to expectations, no evidence of a significant positive relationship between MC and INN was found in the exploitative innovation subgroup (ˇ = 0.023, p > 0.05, Stage I in Table 5). The empirical evidence, however, indicates a significant positive effect of INN on PERF (ˇ = 0.270; p < 0.01, Table 5). This positive significant relation between INN and PERF confirms the strong relationship between innovativeness and performance in both types of innovation, something also noted in previous studies.

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

ARTICLE IN PRESS

G Model YMARE-494; No. of Pages 20

M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

13

Table 6 Two-way interactions of OC and MC mean scores of PERF and INN. Performance

Exploratory projects

Exploitative projects

Mechanistic

Organic

Low

High

Innovation Organic

Low

High

Mechanistic

Low

High

Low

High

A1 ¯ = 4.783 X S.D. = 1.033 n = 20 B1 ¯ = 5.716 X S.D. = 0.921 n=8

C1 ¯ = 5.429 X S.D. = 1.134 n=7 D1 ¯ = 5.406 X S.D. = 0.948 n = 23

A3 ¯ = 4.457 X S.D. = 0.795 n = 14 B3 ¯ = 5.143 X S.D. = 0.766 n=7

C3 ¯ = 5.222 X S.D. = 1.328 n=6 D3 ¯ = 5.266 X S.D. = 0.807 n = 14

A2 ¯ = 4.050 X S.D. = 1.078 n = 20 B2 ¯ = 5.517 X S.D. = 0.696 n=8

C2 ¯ = 4.429 X S.D. = 0.789 n=7 D2 ¯ = 4.942 X S.D. = 0.925 n = 23

A4 ¯ = 4.156 X S.D. = 0.935 n = 14 B4 ¯ = 4.286 X S.D. = 1.177 n=7

C4 ¯ = 3.389 X S.D. = 1.405 n=6 D4 ¯ = 4.043 X S.D. = 0.725 n = 14

It was hypothesized that OC and MC respectively would affect project performance through innovativeness. Stage II in Table 5 shows that OC in exploratory innovation has a positive effect on INN (ˇ = 0.411; p < 0.001), which in turn has a positive effect on PERF measured through INN (ˇ = 0.402; p < 0.001). Further, the results indicate in exploitative innovation a non-significant positive effect of MC on INN (ˇ = 0.023; p > 0.05) and a positive effect of INN on PERF (ˇ = 0.233; p < 0.01, Stage II in Table 5). Together, the results in exploratory innovation fulfill the conditions (Preacher et al., 2007) to consider that INN mediates the relationship between OC and PERF, but not in exploitative innovation. More specifically, the results in exploratory innovation support a pattern of partial mediation (Mathieu and Taylor, 2006), as the effect of OC appears to still exist after including INN. Thus, these results suggest that the association between OC and PERF is partially accounted for by INN, but that OC most likely also has a direct effect on project performance in exploratory innovation. Further, following previous practice (Sarkar et al., 2001) the indirect effect was calculated as a multiplication of the statistically significant indirect effects on and from innovativeness. The indirect effect for exploratory innovations was derived by multiplying the significant path coefficients, OC–INN 0.411 and INN–PERF 0.402, which generated an indirect and total effect of 0.165. The indirect effect for exploitative innovation could not be analyzed as the path between MC–INN was not significant. Thus, the results reveal that only the indirect effect of OC on project performance through innovativeness can be analyzed in this study. The statistical significance of the indirect effect of OC on project performance through innovativeness was tested using a bootstrapping procedure with 1000 bootstraps. Compared to common techniques such as the three-step procedure of Baron and Kenny (1986) or the Sobel test, the bootstrapping procedure has been used more recently (e.g., ˇ 2012; Hall, 2011) and is suggested Bisbe and Malagueno, to be more appropriate for small samples not following a

multivariate normal distribution, like the sample in this study (e.g., Preacher and Hayes, 2004, 2008; MacKinnon et al., 2004). Following the procedure of Preacher and Hayes (2008, pp. 883–884), for each of the 1000 bootstraps the estimated coefficients for each direct path were multiplied in order to calculate an estimated coefficient for the indirect effect. The 1000 indirect effect coefficients were then rank-ordered and analyzed. The bootstrapped 95% confidence intervals did not record a zero in exploratory projects, indicating a significant indirect effect whereas in exploitative projects the bootstrapped 95% confidence intervals included zero, indicating a non-significant indirect effect. Thus, the results of the undertaken procedure provide support for Hypothesis 1, but not Hypothesis 2. In addition to the indirect effects, the results of the test for the combined effect of MC and OC indicate a significant positive interaction effect (ˇ = 0.218; p < 0.01) for the exploratory project subgroup and a significant positive interaction effect (ˇ = 0.341; p < 0.001) for the exploitative projects, when project size and uncertainty are controlled for (Table 5). Thus, Hypothesis 3 was supported. To shed some light on the nature of the interactions, “low” and “high” subgroups (A, B, C and D) were created for OC and MC, and the differences in the PERF and INN mean scores for each of the four subgroups in exploratory and exploitative projects respectively were examined (Table 6). As expected, the lowest mean scores of performance PERF and innovativeness INN occur where both OC and MC are low (cells A1, A2, A3, but with the exception of A4, instead C4). The highest PERF occurs in exploratory projects where OC is high and MC is low (cell B1) and in exploitative settings where both OC and MC are high (cell D3). The highest INN occurs in both exploratory and exploitative projects where OC is high and MC is low (cell B2 and B4). An ANOVA test of the differences in mean scores between the subgroups revealed statistically significant differences in exploratory innovations for INN (3.913, p = 0.013) and in exploitative innovations for PERF (3.586, p = 0.02). The Scheffè test for unplanned multiple comparisons, however,

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

14

Combined effect MC x OC

Organic control (OC)

0.411***/0.147

-0.004/0.267**

0.218**/0.341***

0.402***/0.270**

Innovativeness

-0.110/0.023

Project performance

0.160*/-0.031

Mechanistic control (MC)

***p <0.001, **p <0.010 Results for exploratory projects are marked in bold and exploitative projects in italic. Fig. 2. Results of PLS analysis. *** p < 0.001, ** p < 0.010. Results for exploratory projects are marked in bold and exploitative projects in italic.

indicates marginally significant differences in exploratory projects only between cells A2–B2 and A2–D2 for INN (p = 0.055 and p = 0.058 respectively) and in exploitative projects between cells A3 and D3 for PERF (p = 0.072). In summary, this analysis indicates that project performance in both projects is lowest where both OC and MC are lowest, and highest where OC is high and MC is either low (in exploratory projects) or high (in exploitative projects). In a similar way, innovativeness is lowest where OC is low and MC is either low (exploratory projects) or high (exploitative projects), and INN is highest with high OC and low MC in both projects. These results suggest that OC plays an important role in both exploratory and exploitative projects, whereas MC has a minor role. Finally, as the objective of PLS is to maximize variance explained rather than fit, prediction-orientated measures, such as R-square (R2 ), were used to evaluate the PLS models (Chin, 1998). The R2 for the main effects and interaction models are presented in Table 5. To see how much predictive value the interaction terms add to the model, the R2 from the main model without interactions (Stage 1) was compared with the R2 from the interaction effects model (Stage 2). The additive explanatory power of the interaction model was determined by calculating Cohen’s f2 effect size measure (Chin et al., 2003; Cohen, 1988). The results show that the interaction constructs have a total effect size f of 0.23 in exploitative innovation projects and 0.07 in exploratory innovation projects, which indicates that the inclusion of the interaction terms does improve the explanatory power of the model. The predictive validity of the parameter estimates can be assessed by use of a cross-validated redundancy index or a Stone–Geisser Q2 -test (Geisser, 1974; Stone, 1974). As PLS models lack an index for goodness of fit statistics, Tenenhaus et al. (2005) and Vandenbosch (1996) argue that besides the reliability and validity of constructs, the significance of variance explained and positive Q2 s for all but one of the constructs provide sufficient evidence of model fit. The results in Table 5 suggest that the model has predictive relevance.

5. Discussion Overall, the results of this study emphasize the importance of OC being the main form of control in both exploratory and exploitative innovations. The effect of OC is however somewhat different, as in exploratory innovation projects the focus of OC appears to be on enhancing project performance through innovativeness, whereas in exploitative innovations, OC drives project performance. The findings also indicate that MC is not without value, not when used individually, but rather if used in combination with OC to enhance project performance. MC appears not to drive innovativeness, but instead to be associated with performance mainly in exploratory innovation settings. Further, MC appears to interact with organic forms of control to enhance project performance in both innovation contexts. The findings are shown in Fig. 2 and will be discussed in more detail below. First, although not hypothesized, the direct effects were tested and analyzed. In exploitative settings the effect of OC is manifest on project performance, whereas in exploratory innovations OC works best indirectly through innovativeness. This may be explained by the differences between the two innovation types, related to the different need for information in exploratory and exploitative innovations deriving from a different level of novelty and uncertainty as well as a different need for new knowledge. The importance of OC in project environments in this study is consistent with the findings of previous studies by Jansen et al. (2006) conducted on the organizational level. There, pursuing exploratory innovation was more effective in dynamic environments, and centralization was found to have a negative effect on a unit’s exploratory efforts. Furthermore, while this study registers a slightly positive direct effect of MC on performance in exploratory projects, overall MC appears not to make much of a direct and individual contribution to innovativeness or project performance. The insignificance of MC in innovation projects has also been suggested in early research (e.g., Ouchi, 1977). It therefore appears as if mere mechanistic

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

approaches such as detailed project planning including budgets and timetables do not generate forces and a focus on increasing performance in innovative project settings. On the other hand, the results also indicate that MC may not be as detrimental to exploratory innovation efforts as some previous studies have found (e.g., Song and Montoya-Weiss, 1998). In exploratory innovations, with high levels of uncertainty, radical change and strong emphasis on new knowledge generation, it appears that OC works best indirectly through innovativeness. One possible explanation is that OC stimulates idea generation, flexibility, and interaction among team members and thus enhances innovativeness, which is of vital importance in exploratory innovations, rather than project completion and overall project performance. In exploitative settings, with regard to MC, there is no evidence of any indirect effects of MC on project performance through innovativeness. Third, the results suggest that performance is positively affected by tension emerging from the combined use of OC and MC in both exploratory and exploitative innovation projects. These results extend the findings of Henri (2006), who in an ex post analysis revealed that dynamic tension might have a direct positive and significant impact on performance, in particular for firms facing high environmental uncertainty and with values encompassing flexibility. The results are also in line with Lewis et al. (2002), who argued that meeting tenuous demands for innovation and efficiency requires a blend of “emergent” (organic) and “planned” (mechanistic) approaches, and that managing tensions between flexibility and efficiency may prove key to strong project performance. Combining opposing forms of control such as OC and MC enables the desired direct and indirect effects of OC and MC to simultaneously yield high positive levels of both. Thus, MC used jointly with OC appears to create a tension that enhances project performance in both exploratory and exploitative innovations. This combined effect is particularly useful in exploitative settings, where MC interacts with OC to enhance project performance. Therefore, the tension created in exploratory and exploitative innovation projects has complementary effects. This finding is interesting and also consistent with more recent theorizing that MC is not without value in uncertain environments. In summary, OC appears to be the project manager’s chosen tool for managing project tensions. The results support an indirect effect of OC on project performance through innovativeness in exploratory projects (H1), but do not support indirect effects of MC on project performance through innovativeness in exploitative settings (H2). The interaction effect of OC and MC was supported (H3) in both innovation environments. Thus, it is the conclusion of the current research examining project controls in two different innovation types that exploratory and exploitative innovation projects may indeed benefit from different types of management controls. 6. Conclusion The aim of this study was to investigate the indirect effects of mechanistic and organic forms of control

15

on project performance through innovativeness as well as the interaction effect of those controls on project performance in exploratory and exploitative innovation projects. Overall, the results support the importance of OC, showing it acting through innovativeness on project performance in exploratory innovations and enhancing project performance in exploitative innovations. The results also suggest that the interaction effect of OC and MC enhances performance in both exploratory and exploitative innovation projects, and support a complementary effect. This study contributes to literature in three ways. Firstly, this study draws on prior research in management accounting and control (Chenhall and Morris, 1995; Henri, 2006; Jørgensen and Messner, 2009; Mundy, 2010) to extend those to another level of analysis, the project level rather than the organizational level. While accounting research has studied control mechanisms in other organizational contexts, the project level has largely been forgotten (Chenhall, 2008), although scholars argue it to be more appropriate for studying control systems in innovation settings (Davila et al., 2009b). Secondly, this study offers empirical evidence of the impact of indirect effects of project controls on project performance. By distinguishing between exploratory and exploitative innovative project settings, this study finds indirect effects of organic project controls on performance via innovativeness in exploratory settings. These findings are seen as extending the findings of Bisbe and Otley (2004) at the organizational level. This study also finds empirical evidence of tension on project performance resulting from a combined use of mechanistic and organic forms of control in both exploratory and exploitative innovation project settings. As the results indicate, the tension caused by a combination of project control mechanisms appears to be complementary, but not dependent on the degree of innovation newness. Although prior research on opposing control forces in exploratory innovation settings exists, empirical research comparing the effects of combined use of two opposing forms of project control in both exploratory and exploitative innovative project settings is non-existent. This study also has important managerial implications. First, it emphasizes the important role of organic communication processes in enhancing project performance through innovativeness in exploratory innovation settings. It also shows that project managers can identify how the different forms of control they already use or intend to use can most benefit innovation by combining those controls with their respective complementary opposite coordination mechanisms. The variety of possible combinations of different forms of MCS or style of use of MCS provides the project manager with a pool of options. A project manager who is informed about the desirable and undesirable effects and who pays them due attention, can enhance performance and innovativeness. However, combining opposing control strategies may not be easy in practice within the same social system, and evidence from prior studies also indicates that practitioners often do not combine opposing action strategies (e.g. Sheremata, 2002).

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20 16

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

Some limitations of this study need to be noted. As survey research, the study is vulnerable to the typical weaknesses relating to validity and reliability of items and tests, although best practice was followed in the development and pre-testing of the instrument (Van der Stede et al., 2005). One of the obvious limitations of this study is the small sample size and the fact that its sample was not strictly randomly selected; meaning any inferences from the results must be drawn cautiously. The relatively high number of exploratory projects compared to exploitative projects in this study differs from a normal innovation setting, and may suggest a bias. The combined use of organic (OC) and mechanistic (MC) forms of project control was also examined using a product term, and a complete and reliable test for complementarity between the two forms of control and project performance would include incorporating appropriate bi-directional links into the model, which was not possible in this study. In addition, the limited measurement of mechanistic and organic forms of control and the use of a categorical (and not continuous) measure of innovation type would suggest the results should be interpreted with caution. Furthermore, it would be wise to be cautious when interpreting the statistical associations as causal relationships, owing to the cross-sectional nature of the study. Although project managers are considered the best judges of their own actions (Brownell, 1995), a further limitation is the use of self-assessed use ratings. Previous research has, however, found a significant positive correlation between superiors’ subjective rating and objective measures of subordinate performance (Bommer et al., 1995; Furnham and Stringfield, 1994). It is also important to note that the effects obtained in the different runs of the model show high robustness and the results of the preliminary and final analyses are consistent. It is suggested that more accounting studies on control and coordination mechanisms in project environments would be worthwhile. Studies examining MCS from different innovation perspectives—such as other types or stages of innovation—may find different outcomes or allow replications that make these results generalizable. The tension arising from the simultaneous use of various forms of control or various styles of use of control mechanisms may change over time, and enhance project performance differently in the early stages of project development than they do during late development and commercialization. Thus, studies using longitudinal data might reveal whether project managers adjust their actions to contextual changes, and if so, how (Lewis et al., 2002; Pennings, 1992). In addition, qualitative methodologies may provide new insights into the ways in which project managers in different kinds of organizations and project settings reinforce and manage tension in practice. Moreover, the personality traits of the project manager are likely to be of considerable importance when implementing and using opposing forms of MCS simultaneously and well worth investigation in future research. Acknowledgments The authors would like to thank two anonymous reviewers, Professor Robert H. Chenhall and the

participants of the MONFORMA conference in Melbourne in November, 2010 for their very insightful and helpful comments.

Appendix A. Measurement instrument A.1. Organic form of control (OC) Level of agreement with the statements (1 = strongly disagree, 7 = strongly agree). OC1 . I call my project team subordinates in to discuss project performance deviations in face-to-face meetings. OC2 . My own project team subordinates and I often discuss and resolve project performance issues together informally. OC3 . I place considerable emphasis on open channels of communication and the free flow of information between myself and my subordinates. A.2. Mechanistic form of control (MC) Level of agreement with the statements (1 = strongly disagree, 7 = strongly agree). Detail of performance evaluation and importance of project performance target deviations. MC1 . I judge my project team performance with performance measures that explain in detail project performance variances on a line-by-line basis. MC2 . I am not only interested in how well my project team achieves the overall project performance targets, but I also evaluate the extent to which my project team is on target in each of the project performance lineitems. MC3 . I attach a great deal of importance to interim project performance target deviations from budgeted performance and project milestones. MC4 . I require my project team subordinates to report the actions taken to correct causes of deviation from the interim project performance targets.

A.3. Innovativeness (INN) Innovative accomplishments are defined here very broadly to include any policy, structure, method or process, product or market opportunity that you as the manager of the project perceived to be new. Anchored by 1 = extremely low, and 7 = extremely high. INN1 . The level of my project team innovation performance (number of innovations) measured relative to project performance targets (standards). INN2 . The level of my project team innovation performance (number of innovations) relative to other project teams measured innovation performance working on the same type of project.

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

17

Table B1 Common method bias analysis. Item

Substantive factor loading (R1)

R12

Common method factor loading (R2)

R22

Mechanistic control (item 1) Mechanistic control (item 2) Mechanistic control (item 3) Mechanistic control (item 4) Organic control (item 1) Organic control (item 2) Organic control (item 3) Project Performance (item 1) Project Performance (item 2) Project Performance (item 3) Innovativeness (item 1) Innovativeness (item 2) Innovativeness (item 3) Task Uncertainty (item 1) Task Uncertainty (item 2) Task Uncertainty (item 3) Project Duration (item 1) No. People in Project (item 1) Average

0.714** 0.639** 0.895** 0.727** 0.700** 0.823** 0.521** 0.817** 0.866** 0.929** 0.910** 0.829** 0.858** 0.789** 0.841** 0.699** 1.000 1.000 0.801

0.510 0.408 0.801 0.529 0.490 0.677 0.271 0.667 0.750 0.863 0.828 0.687 0.736 0.623 0.707 0.489 1.000 1.000 0.669

−0.034 0.180* −0.162 −0.017 0.068 −0.083 0.027 0.074 −0.034 −0.040 −0.107 −0.017 0.111 0.001 0.043 −0.049 0.000 0.000 −0.002

0.001 0.032 0.026 0.000 0.004 0.007 0.001 0.005 0.001 0.002 0.011 0.000 0.012 0.000 0.002 0.002 0.000 0.000 0.006

* **

p < 0.05. p < 0.01.

INN3 . The level of my project team measured innovation performance. A.4. Project performance (PERF) Level of performance of your (1 = extremely low, 7 = extremely high).

project

team

Perf1 . The level of my project team measured performance relative to my performance targets (standards). Perf2 . The level of my project team measured performance relative to other project teams measured performance working on the same kind of project. Perf3 . The level of my project team measured performance. A.5. Task uncertainty (TU) Level of agreement with the statements (1 = to a small extent, 7 = to a great extent). The scale was revised in the analysis, that is, a larger number now indicates a greater uncertainty. TU1 . To what extent was there a clearly understood way of doing the major types of work normally encountered in your project? TU2 . To what extent was there a clearly defined body of knowledge of subject matter, which could guide the work done in your project? TU3 . To do the work of your project, to what extent could personnel actually rely on established procedures and practices? Appendix B. Table B1.

References Abernethy, M.A., Brownell, P., 1997. Management control systems in research and development organizations: the role of accounting, behavior and personnel controls. Accounting, Organization and Society 22 (3–4), 233–248. Abernethy, M.A., Brownell, P., 1999. The role of budgets in organizations facing strategic change: an exploratory study. Accounting, Organizations and Society 24 (3), 189–204. Adler, P.S., Borys, B., 1996. Two types of bureaucracy: enabling and coercive. Administrative Science Quarterly 41 (1), 61–89. Ahrens, T., Chapman, C.S., 2004. Accounting for flexibility and efficiency: a field study of management control systems in a restaurant chain. Contemporary Accounting Research 21 (2), 271–301. Aiken, L.S., West, S.G., 1991. Multiple Regression: Testing and Interpreting Interactions. Sage, Newbury Park, CA. Amabile, T.M., Barsade, S.G., Mueller, J.S., Staw, B.M., 2005. Affect and creativity at work. Administrative Science Quarterly 50 (3), 367–403. Ancona, D.G., Caldwell, D.F., 1992. Demography and design: predictors of new product team performance. Organization Science 3 (3), 321–341. Anderson, S.W., Dekker, H.C., 2005. Management control for market transactions: the relation between transaction characteristics, incomplete contract design and subsequent performance. Management Science 51 (12), 1734–1753. Argyris, C., Schön, D., 1983. Organizational Learning. Addison-Wesley, MA. Ayers, D., Dahlstrom, R., Skinner, S.J., 1997. An exploratory investigation of organizational antecedents to new product success. Journal of Marketing Research 34 (1), 107–116. Barclay, D., Higgins, C., Thompson, R., 1995. The Partial Least Squares (PLS) approach to causal modeling: personal computer adoption and use as an illustration. Technology Studies 2 (2), 285–324. Baron, R.M., Kenny, D.A., 1986. The moderator-mediator variable distinction in social psychological research: conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology 51 (6), 1173–1182. Benner, M.J., Tushman, M.L., 2003. Exploitation, exploration, and process management: the productivity dilemma revisited. Academy of Management Review 28 (2), 238–256. ˇ R., 2012. Using strategic performance measurement Bisbe, J., Malagueno, systems for strategy formulation: does it work in dynamic environments? Management Accounting Research 23 (4), 296–311. Bisbe, J., Otley, D., 2004. The effects of the interactive use of management control systems on product innovation. Accounting, Organizations and Society 29 (8), 709–737. Bommer, W.H., Johnson, J.L., Rich, G.A., Podsakoff, P.M., MacKenzie, S.B., 1995. On the interchange ability of objective and subjective measures of employee performance. Personnel Psychology 48, 587–605.

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20 18

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

Brownell, P., 1995. Research Methods in Management Accounting. Coopers and Lybrand. Burns, T., Stalker, G.M., 1961. The Management of Innovation. Tavistock, London. Calantone, R.J., Harmancioglu, N., Droge, C., 2010. Inconclusive innovation “returns”: a meta-analysis of research on innovation in new product development. Journal of Product Innovation Management 27 (7), 1065–1081. Cardinal, L., 2001. Technological innovation in the pharmaceutical industry: the use of organizational control in managing research and development. Organization Science 12 (1), 19–36. Chenhall, R.H., 2003. Management control systems design within its organizational context: findings from contingency-based research and directions for the future. Accounting, Organizations and Society 28 (2–3), 127–168. Chenhall, R.H., 2008. Accounting for the horizontal organization: a review essay. Accounting, Organizations and Society 33 (4–5), 517–550. Chenhall, R.H., Morris, D., 1995. Organic decision and communication processes and management accounting systems in entrepreneurial and conservative business organizations. Omega: International Journal of Management Science 23 (5), 485–497. Chiesa, V., Frattini, F., Lamberti, L., Noci, G., 2009. Exploring management control in radical innovation projects. European Journal of Innovation Management 12 (4), 416–443. Chin, W.W., 1998. The partial least square approach for structural equation modeling. In: Marcoulides, G.A. (Ed.), Modern Methods for Business Research. Lawrence Erlbaum Associates. Chin, W.W., Marcolin, B.I., Newsted, P.R., 2003. A partial least squares latent variable modeling approach for measuring interaction effects: results from a Monte Carlo simulation study and an electronic–mail emotion/adoption study. Information Systems Research 14 (2), 189–217. Cohen, J., 1988. Statistical Power Analysis for the Behavioral Sciences. Lawrence Erlbaum Associates, Hillsdale, NJ. Cohen, W.M., Levinthal, D.A., 1990. Absorptive capacity: a new perspective on learning and innovation. Administrative Science Quarterly 35 (1), 128–152. Cronbach, L.J., 1987. Statistical tests for moderator variables: flaws in analyses recently proposed. Psychological Bulletin 102 (3), 414– 417. Damanpour, F., 1991. Organizational innovation: a meta-analysis of effects of determinants and moderators. Academy of Management Journal 34 (3), 555–590. Davila, T., 2000. An empirical study on the drivers of management control systems’ design in new product development. Accounting, Organizations and Society 25 (4–5), 383–409. Davila, A., Foster, G., Li, M., 2009a. Reasons for management control systems adoption: insights from product development systems choice by early-stage entrepreneurial companies. Accounting, Organizations and Society 34, 322–347. Davila, A., Foster, G., Oyon, D., 2009b. Accounting and control, entrepreneurship and innovation: venturing into new research opportunities. European Accounting Review 18 (2), 281–311. De Dreu, C.K.W., Weingart, L.R., 2003. Task versus relationship conflict and team effectiveness: a meta-analysis. Journal of Applied Psychology 88 (4), 741–749. Dent, J.F., 1990. Strategy, organization and control: some possibilities for accounting research. Accounting, Organizations and Society 15 (1–2), 3–25. Dewar, R.D., Dutton, J.E., 1986. The adoption of radical and incremental innovations: an empirical analysis. Management Science 32 (11), 1422–1433. Dougherty, D., 1992. Interpretive barriers to successful product innovation in large firms. Organization Science 3 (2), 179–202. Dougherty, D., 1996. Organizing for innovation. In: Clegg, S.R., Hardy, C., Nord, W.R. (Eds.), Handbook of Organization Studies. Sage, Thousand Oaks, CA, pp. 424–439. Dougherty, D., 2006. Organizing for innovation in the 21st century. In: Clegg, S.R., Hardy, C., Lawrence, T.B., Nord, W.R. (Eds.), The Sage Handbook of Organization Studies. Sage Publications, London, pp. 598– 617. Drazin, R., Van de Ven, A.H., 1985. Alternative forms of fit in contingency theory. Administrative Science Quarterly 30 (4), 514–539. Droge, C., Calantone, R., Harmancioglu, N., 2008. New product success: is it really controllable by managers in highly turbulent environments? Journal of Product Innovation Management 25 (3), 272–286. Dvir, D., Raz, T., Shenhar, A.J., 2003. An empirical analysis of the relationship between project planning and project success. International Journal of Project Management 21, 89–95.

Ettlie, J.E., Bridges, W.P., O’Keefe, R.D., 1984. Organization strategy and structural differences for radical versus incremental innovations. Management Science 30 (6), 682–695. Falk, R.F., Miller, N.B., 1992. A Primer for Soft Modeling. The University of Akron Press, Akron, OH. Fisher, J.G., 1995. Contingency-based research on management control systems: categorization by level of complexity. Journal of Accounting Literature 14, 24–53. Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research 18 (1), 39–50. Furnham, A., Stringfield, P., 1994. Congruence of self and subordinate ratings of managerial practices as a correlate of superior evaluation. Journal of Occupational and Organizational Psychology 67, 57–67. Gefen, D., Straub, D., 2005. A practical guide to factorial validity using PLS-Graph: tutorial and annotated example. Communications of the Association for Information Systems 16 (5), 91–109. Geisser, S., 1974. A predictive approach to the random effect model. Biometrika 61 (1), 101–107. Goodhue, D., Lewis, W., Thompson, R., 2007. Statistical power in analyzing interaction effects: questioning the advantage of PLS with product indicators. Information Systems Research 18 (2), 211– 227. Gupta, A.K., Govindarajan, V., 1984. Business unit strategy, managerial characteristics, and business unit effectiveness at strategy implementation. Academy of Management Journal 27, 25–41. Hall, M., 2011. Do comprehensive performance measurement systems help or hinder managers’ mental model development? Management Accounting Research 22 (2), 68–83. Hair, J.F., Anderson, R.E., Tatham, R.L., Black, W.C., 1998. Multivariate Data Analysis. Prentice-Hall. Henri, J.F., 2006. Management control systems and strategy: a resourcebased perspective. Accounting, Organizations and Society 31 (6), 529–558. Hill, C.W.L., Rothaermel, F.T., 2003. The performance of incumbent firms in the face of radical technological innovation. Academy of Management Review 28 (2), 257–274. Hsieh, P-O.J.J., Rai, A., Kell, M., 2008. Understanding digital inequality: comparing continued use behavioural models of the socioeconomically advantaged and disadvantaged. MIS Quarterly 32 (1), 97–126. Hulland, J., 1999. Use of partial least Squares (PLS) in strategic management research: a review of four recent studies. Strategic Management Journal 20 (2), 195–204. Jansen, J.J.P., van den Bosch, F.A.J., Volberda, H.W., 2006. Exploratory innovation, exploitative innovation, and performance: effects of organizational antecedents and environmental moderators. Management Science 52 (11), 1661–1674. Jørgensen, B., Messner, M., 2009. Management control in new product development: the dynamics of managing flexibility and efficiency. Journal of Management Accounting Research 21 (1), 99–124. Jöreskog, K.G., Wold, H., 1982. The ML and PLS techniques for modeling with latent variables: historical and comparative aspects. In: Wold, H., Jöreskog, K. (Eds.), Systems Under Indirect Observation: Causality, Structure, Prediction (Vol. I). North-Holland, Amsterdam, pp. 263–270. Kamm, J.B., 1987. An Integrative Approach to Managing Innovation. Lexington Books, Lexington. Kanter, R., 1983. The Change Masters. Simon & Schuster, New York. Kanter, R., 2001. Evolve! Succeeding in the Digital Culture of Tomorrow. Harvard Business School Press, Boston. Kleinschmidt, E.J., Cooper, R., 1991. The impact of product innovativeness on performance. Journal of Product Innovation Management 8, 240–251. Lewis, M.W., 2000. Exploring paradox: toward a more comprehensive guide. Academy of Management Review 25 (4), 760–776. Lewis, M.W., Welsh, M.A., Dehler, G.E., Green, S.G., 2002. Product development tensions: exploring contrasting styles of project management. Academy of Management Journal 45 (3), 546– 564. Liang, H., Saraf, N., Hu, Q., Xue, Y., 2007. Assimilation of enterprise systems: the effect of institutional pressures and the mediating role of top management. MIS Quarterly 31 (1), 59–87. MacKinnon, D.P., Lockwood, C.M., Williams, J., 2004. Confidence limits for the indirect effect: distribution of the product and resampling methods. Multivariate Behavioral Research 39 (1), 99–128. Malmi, T., Brown, D.A., 2008. Management control systems as a package: opportunities, challenges and research directions. Management Accounting Research 19 (4), 287–300.

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

Marginson, D., 2002. Management control systems and their effects on strategy formation at the middle-management levels: evidence from a U.K. organization. Strategic Management Journal 23 (11), 1019–1031. Martino, J., 1995. Research and Development Project Selection. Wiley and Sons, New York, NY. Mathieu, J.E., Taylor, S.R., 2006. Clarifying conditions and decision points for mediational type inferences in organizational behavior. Journal of Organizational Behavior 27, 1031–1056. McDonough III, E.F., Barczak, G., 1991. Speeding up new product development: the effects of leadership style and source of technology. Journal of Product Innovation Management 8 (3), 203–211. Menguc, B., Auh, S., 2008. Innovation and knowledge creation: the asymmetric moderating role of market orientation on the ambidexterity–firm performance relationship for prospectors and defenders. Industrial Marketing Management 37 (4), 455–470. Milgrom, P., Roberts, J., 1995. Complementarities and fit: strategy, structure, and organizational change in manufacturing. Journal of Accounting and Economics 19 (2–3), 179–208. Montoya-Weiss, M.M., Calantone, R., 1994. Determinants of new product performance: a review and meta-analysis. Journal of Product Innovation Management 11 (5), 397–417. Morse, J.J., Lorsch, J.W., 1970. Beyond theory. Harvard Business Review 48 (May–June), 61–68. Mundy, J., 2010. Creating dynamic tensions through a balanced use of management control systems. Accounting, Organizations and Society 35 (5), 499–523. Nunnally, J.C., 1978. Psychometric Theory. McGraw-Hill, New York. OECD, 2004. Oslo Manual. The Measurement of Scientific and Technological Activities. Proposed Guidelines for Collecting and Interpreting Technological Innovation Data. European Commission, Eurostat, Available at: http://www.oecd.org/dataoecd/35/61/2367580.pdf Olson, E.M., Walker, O.C., Ruekert, R.W., 1995. Organizing for effective new product development: the moderating role of product innovativeness. Journal of Marketing 59 (1), 48–62. Ouchi, W.G., 1977. The relationship between organizational structure and organizational control. Administrative Science Quarterly 22 (1), 95–113. Ouchi, W.G., 1979. A conceptual framework for the design of organizational control mechanisms. Management Science 25, 833–848. Pennings, J.M., 1992. Structural contingency theory: a reappraisal. In: Staw, B.M., Cummings, L.L. (Eds.), Research in Organizational Behavior, 14. JAI Press, Greenwich CT, pp. 267–309. Podsakoff, P.M., MacKenzie, S.B., Lee, J.-Y., Podsakoff, N.P., 2003. Common method biases in behavioral research: a critical review of the literature and recommended remedies. Journal of Applied Psychology 88 (5), 879–903. Podsakoff, P.M., Organ, D.W., 1986. Self-reports in organizational research: problems and prospects. Journal of Management 12 (4), 531– 544. Popadiuk, S., Choo, C.W., 2006. Innovation and knowledge creation: how are these concepts related? International Journal of Information Management 26 (4), 302–312. Poppo, L., Zenger, T., 2002. Do formal contracts and relational governance function as substitutes or complements? Strategic Management Journal 23, 707–725. Preacher, K.J., Hayes, A.F., 2004. SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods 36 (4), 717–731. Preacher, K.J., Rucker, D.D., Hayes, A.F., 2007. Assessing moderated mediation hypotheses: theory, method, and prescriptions. Multivariate Behavioral Research 42, 185–227. Preacher, K.J., Hayes, A.F., 2008. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods 40 (3), 879–891. Rijsdijk, S.A., van den Ende, J., 2011. Control combinations in new product development projects. Journal of Product Innovation Management 28, 868–880. Ringle, C.M., Wende, S., Will, A., 2005. SmartPLS, 2.0 (beta), Available at: http://www.smartpls.de. Hamburg, Germany. Rogers, E., 1995. Diffusion of Innovations. Free Press, New York. Sarkar, M.B., Echambadi, R., Cavusgil, S.T., Aulakh, P.S., 2001. The influence of complementarity, compatibility, and relationship capital on alliance performance. Academy of Marketing Science Journal 29 (4), 358–373. Shenhar, A.J., Dvir, D., 1996. Toward a typological theory of project management. Research Policy 25 (4), 607–632. Shenhar, A.J., Tishler, A., Dvir, D., Lipovetsky, S., Lechler, T., 2002. Refining the search for project success factors: a multivariate, typological approach. R&D Management 32, 111–126.

19

Sheremata, W.A., 2000. Centrifugal and centripetal forces in radical new product development under time pressure. Academy of Management Review 25 (2), 389–408. Sheremata, W.A., 2002. Finding and solving problems in software new product development. Journal of Product Innovation Management 19 (2), 144–158. Shields, M.D., Kato, Y., Deng, J., 2000. The design of control systems: test of direct and indirect – effect models. Accounting, Organizations and Society 25 (2), 185–202. Siggelkow, N., 2002. Misperceiving interactions among complements and substitutes: organizational consequences. Management Science 48 (7), 900–916. Simons, R., 1995. Levers of Control: How Managers Use Innovative Control Systems to Drive Strategic Renewal. Harvard Business School Press, Boston, MA. Song, X.M., Montoya-Weiss, M., 1998. Critical development activities for really new versus incremental products. Journal of Product Innovation Management 15, 124–135. Stone, M., 1974. Cross-validatory choice and assessment of statistical predictions. Journal of the Royal Statistical Society, Series B 36, 111–133. Sundaramurthy, C., Lewis, M., 2003. Control and collaboration: paradoxes of governance. Academy of Management Review 28 (3), 397–415. Tatikonda, M.V., Montoya-Weiss, M.M., 2001. Integrating operations and marketing perspectives of product innovation. The influence of organizational process factors and capabilities on development performance. Management Science 47, 151–172. Tatikonda, M.V., Rosenthal, S.R., 2000a. Successful execution of product development projects: balancing firmness and flexibility in the innovation process. Journal of Operations Management 18 (4), 401–425. Tatikonda, M.V., Rosenthal, S.R., 2000b. Technology novelty, project complexity, and product development project execution success: a deeper look at task uncertainty in product innovation. IEEE Transactions on Engineering Management 47 (1), 74–87. Tenenhaus, M., Vinzi, V.E., Chatelin, Y., Lauro, C., 2005. PLS path modelling. Computational Statistics and Data Analysis 48 (1), 159–205. Turner, K.L., Makhija, M.V., 2006. The role of organizational controls in managing knowledge. Academy of Management Review 31 (1), 197–217. Tushman, M., Nadler, D., 1978. Information processing as an integrating concept in organizational design. Academy of Management Review 3 (3), 613–624. Tushman, M.L., O’Reilly, C.A., 1996. Ambidextrous organizations: managing evolutionary and revolutionary change. California Management Review 38 (4), 8–30. Un, C.A., 2010. An empirical multi-level analysis for achieving balance between incremental and radical innovations. Journal of Engineering and Technology Management 27 (1–2), 1–19. Wageman, R., 2001. How leaders foster self-managing team effectiveness: design choices versus hands-on coaching. Organization Science 12 (5), 559–577. Van de Ven, A.H., 1986. Central problems in the management of innovation. Management Science 32 (5), 590–607. Van de Ven, A., Andrew, H., Polley, D.E., Garud, R., Venkataraman, S., 1999. The Innovation Journey. Oxford University Press, New York. Van de Ven, A.H., Polley, D., 1992. Learning while innovating. Organization Science 3 (1), 92–116. Van der Stede, W.A., 2001. Measuring tight budgetary control. Management Accounting Research 12 (1), 119–137. Van der Stede, W.A., Young, S.M., Chen, C.X., 2005. Assessing the quality of evidence in empirical management accounting research: the case of survey studies. Accounting, Organizations and Society 30 (7–8), 655–684. Vandenbosch, M.B., 1996. Confirmatory compositional approaches to the development of product spaces. European Journal of Marketing 30 (3), 23–46. Venkatraman, N., 1989. The concept of fit in strategy research: toward verbal and statistical correspondence. Academy of Management Review 14 (3), 423–444. Wheelwright, S.C., Clark, K.B., 1992. Revolutionizing Product Development: Quantum Leaps in Speed, Efficiency, and Quality. Free Press, New York. Widener, S.K., 2004. An empirical investigation of the relation between the use of strategic human capital and the design of the management control system. Accounting, Organizations and Society 29 (3–4), 377–399. Widener, S.K., 2007. An empirical analysis of the levers of control framework. Accounting, Organizations and Society 32 (7–8), 757–788. Williams, K.Y., O’Reilly, C.A., 1998. Demography and diversity in organizations: a review of 40 years of research. In: Staw, B.M., Sutton, R.I.

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001

G Model YMARE-494; No. of Pages 20 20

ARTICLE IN PRESS M. Ylinen, B. Gullkvist / Management Accounting Research xxx (2013) xxx–xxx

(Eds.), Research in Organizational Behavior, 20. JAI Press, Greenwich, CT, pp. 77–140. Williams, L.J., Edwards, J.R., Vandenberg, R.J., 2003. Recent advances in causal modeling methods for organizational and management research. Journal of Management 29 (6), 903–936. Wilson, J.M., Goodman, P.S., Cronin, M.A., 2007. Group learning. Academy of Management Review 32 (4), 1041–1059. Withey, M., Daft, R.L., Cooper, R.H., 1983. Measures of Perrow’s work-unit technology: an empirical assessment and a new scale. Academy of Management Journal 26 (1), 45–63.

Zahra, S.A., George, G., 2002. Absorptive capacity: a review, reconcepualisation, and extension. Academy of Management Review 27 (2), 185–203. Zaltman, G., Duncan, R., Holbek, J., 1973. Innovations and Organizations. Wiley, New York. Zander, U., Kogut, B., 1995. Knowledge and the speed of the transfer and imitation of organizational capabilities: an empirical test. Organization Science 6 (1), 76–92. Zirger, B.J., Maidique, M.A., 1990. A model of new product development: an empirical test. Management Science 36 (7), 867–883.

Please cite this article in press as: Ylinen, M., Gullkvist, B., The effects of organic and mechanistic control in exploratory and exploitative innovations. Manage. Account. Res. (2013), http://dx.doi.org/10.1016/j.mar.2013.05.001