Results control and introjected motivation: the case of academia Open Access

This paper focuses on the motivational effects of results control. To this end, we draw upon Merchant and van der Stede (2017) to define results control as a type of control system that focuses on achieving desired outcomes or results rather than prescribing how those outcomes should be achieved. To accomplish this, a results control system typically involves setting performance targets or objectives for a particular control object (such as individuals, teams or business units), monitoring the outcomes in relation to such objectives and rewarding those outcomes that meet or beat expectations.

When taking an interest in the motivational effects of this particular type of control, previous Management Control (MC) studies have typically drawn upon Self-Determination Theory (SDT) to discuss how and why the setting of targets, monitoring of outcomes and incentivising performances can trigger different types of motivation (for overviews, see e.g. Englund and Gerdin, 2025; Van der Hauwaert et al., 2022; Wibbeke and Lachmann, 2020). Moreover, as some types of motivation – especially those that SDT refers to as autonomous types of motivation – have been found to have positive effects on the performance and well-being of individuals (Ryan and Deci, 2017), most efforts have been devoted to studying how results control can trigger such autonomous types of motivation. Specifically, many studies have considered how results controls can be designed and used so as to satisfy the three basic psychological needs of individuals that SDT points out, namely to feel competent, related and autonomous (see e.g. Groen et al., 2017; Holderness et al., 2020; Pfister and Lukka, 2019; Van der Hauwaert et al., 2022). The reason for this is that these needs are assumed by SDT to function as the very mechanism through which individuals come to internalise the controls and thus become autonomously motivated.

Notwithstanding the many important insights generated by this literature regarding such basic psychological needs and internalisation processes, though, rather scarce attention has been devoted to the type of internalisation processes that characterise what SDT refers to as introjected motivation. When introjectedly motivated, people have internalised the expectations set up by a control system but only in a partial way. That is, they have taken in the expectations and are motivated by them, but they have not really come to identify with them or integrated them as part of who they are (e.g. Ryan and Deci, 2000a, 2017). The result, as suggested by SDT, is that people come to behave in line with the expectations, but in doing so they typically experience feelings of inner pressure and tension. Or, as suggested by Ryan and Deci (2000a, p. 62), “Introjection describes a type of internal regulation that is still quite controlling because people perform such actions with the feeling of pressure in order to avoid guilt or anxiety or to attain ego-enhancements or pride”.

Indeed, some MC studies have mentioned this type of partial internalisation and introjected motivation as one of many types of motivation pointed out by SDT (see e.g. Bouten and Hoozée, 2022; Kunz and Linder, 2012; Pfister and Lukka, 2019; Sutton and Brown, 2016). In a few cases, introjected motivation has also been theorised in relation to results controls. For example, Guo et al. (2012) studied whether linking incentives to only outcome measures, or to both driver and outcome measures, would affect the propensity of individuals to be introjectedly motivated. Empirically, though, they found no differences between these two types of incentive systems. In a similar manner, Chen et al. (2020) hypothesised and found, among other things, that the use of diagnostic control systems shifts focus away from autonomous motivation and instead increases the relative importance of controlled motivation (where introjected motivation constitutes one such controlled type of motivation).

Apart from these very few studies, though, there has been no real focus on the relationship between results control and introjected motivation in the MC literature. Arguably, this is surprising, for at least two reasons. Firstly, this literature has argued that a results control system can hold qualities that not only support but also thwart one or more of the needs. For example, it has pointed to how results control thwarts the need for competence or relatedness when it provides negative or unreliable feedback (e.g. Kunz and Linder, 2012; van der Kolk et al., 2019). Similarly, the need for autonomy can be thwarted when outcome metrics are oriented towards particular tasks or when such metrics build on hypothesised relationships between drivers and outcomes (e.g. Guo et al., 2012; Pfister and Lukka, 2019; Siverbo, 2023). In such cases, SDT gives us reason to expect less than full internalisation and thus more controlled types of motivation (including introjected motivation). Despite this, though, the study of how such need thwarting qualities of results control can lead to introjected motivation rather than how they hinder the development of autonomous motivation has not been an empirically prioritised area in the existing literature. Secondly, according to SDT introjected motivation is associated with marked negative long-run effects on various outcome variables, including work performance and well-being (Ryan and Deci, 2017). To avoid such negative long-run effects, it should thus be highly important to better understand when internalisation of a results control system leads to introjected rather than autonomous types of motivation. Again, though, this is a matter that has attracted limited attention in the existing MC literature.

Based on this, the purpose of the current paper is to explore when and why results control supports and thwarts the basic psychological needs of individuals to feel competent, related and autonomous in such ways that they only partially internalise the expectations of the control system and thus become introjectedly motivated. This research focus grew out of an empirical study of researchers at Swedish universities who were exposed to results control in the sense that there were clear performance targets that were systematically followed up and incentivised. In fact, some preliminary and rather explorative analyses of the empirical material from this study suggested that the results control system had contributed to make researchers introjectedly motivated in the sense that they clearly wanted what the control system wanted from them at the same time as they experienced feelings of inner pressure and tension in doing so. Based on these very preliminary findings, we wanted to explore what it was about the control system that seemed to lead to this type of motivation. Specifically, two emergent research questions guided our further empirical endeavour:

Overall, our exploration of these two research questions resulted in the following findings. To begin with, we found that the existing control system both supported and thwarted the needs among researchers to feel competent, related and autonomous. It did so through a quality referred to here as a “conditional logic”, which refers to how the control system conditioned the type of competence-, relatedness- and autonomy-support that it offered on how well researchers performed in relation to the expectations set up by the control system. As a result, the needs were pitted against each other in the sense that one or more needs could only be met at the cost of another need. Moreover, we found that this conditional logic, in turn, functioned as an important mechanism for bringing about introjected motivation among the researchers under study. Importantly, though, and drawing upon the distinction in the SDT literature between need-supporting/need-thwarting qualities of the work environment and how individuals actually experience that their needs are satisfied/frustrated by such qualities (Deci et al., 2017), our findings suggest that the individual researchers constituted a very active part in bringing about and upholding this mechanism. In fact, researchers seemed to have developed a coping strategy where they, despite being critical of the results control system and its underlying logic, willingly complied with the underlying expectations of the control system so as to reconcile or align such expectations with their needs. However, while such willing compliance resulted in that some aspects of their needs were satisfied it at the same time meant that other aspects of their needs were frustrated. Consequently, their willing compliance was typically associated with feelings of doubt and inner tension which constitutes a hallmark of introjected motivation. In the paper, we explore how researchers balanced between these conflicting experiences of the control system’s conditional logic and how such balancing, in turn, is important for understanding how and why a results control system can lead to introjected motivation.

When taken together, these findings arguably contribute to the existing literature in two main ways. To begin with, they contribute to the MC literature on control systems and motivation. Again, while existing MC literature has identified both need-supporting and need-thwarting qualities of results control, it has almost exclusively modelled and theorised how such qualities relate to full internalisation and thus to autonomous motivation (e.g. Bouten and Hoozée, 2022; Pfister and Lukka, 2019; Sutton and Brown, 2016). In contrast to this, we focus on and identify a quality of control systems that is likely to lead to only partial internalisation and, thus, to people developing introjected motivation. Moreover, while most of the existing MC literature has seen such qualities as a theoretical mechanism that more or less automatically explains the relationship between a control system and particular types of motivation, the coping strategy identified in the current study suggests that this relationship is considerably more complex than that.

The conditional logic identified and theorised also contributes to the more general SDT literature elaborating on what makes people develop introjected motivation. Indeed, conditioning in the form of conditional regard (e.g. Assor et al., 2004; Haines and Schutte, 2023; Ryan and Deci, 2017) and contingent approval (e.g. Niemiec et al., 2008, 2010) has already been suggested as an important explanation of introjected motivation. That is, it has been suggested that when people experience that they have to think or behave in particular ways to be approved by or achieve positive regard from (important) others, it means that they will have to give up some of their autonomy to satisfy their need for relatedness and competence. The result, it is suggested, is partial internalisation and introjected motivation (e.g. Bartholomew et al., 2010; Haines and Schutte, 2023). Arguably, our findings add to this explanation in two different ways. Firstly, our notion of a conditional logic suggests that the type of conditioning underlying introjected motivation may not only be about the giving up of autonomy to secure relatedness and competence; it can also be about the giving up of some aspects of one’s autonomy to gain or maintain other such autonomy aspects. In the current study, this was found to be the case when researchers willingly gave up some autonomy by complying with the requirements of the control system despite being critical of these very requirements. However, through doing so, they were able to gain autonomy by securing their time to do research and their freedom from interference from superiors. Secondly, when studying how various forms of conditioning can lead to introjected motivation, the existing literature has had a largely cross-sectional and correlational focus (Haines and Schutte, 2023; McAnally and Hagger, 2024; Otterpohl et al., 2021) which arguably makes it hard to reveal the underlying dynamics through which people make sense of, and handle, any conditioning elements in their environments. In contrast to this, our qualitative interview study allowed deeper insights into such dynamics, suggesting the importance of human agency and intentional considerations when trying to understand the relationship between conditioning elements in the environment and the development of a particular type of motivation.

The remaining parts of the paper are organised as follows. In Section 2, we discuss some key aspects of SDT and how they relate to the current study. In particular, we discuss the notion of introjected motivation and how it can be seen as driven by elements in the environment that support and thwart the basic psychological needs at the same time, thereby resulting in a form of partial internalisation of the “introjects”. We also discuss what we have learned thus far about how results control can support and thwart these needs. In Section 3, we discuss our methods for data collection and analysis, while Section 4 outlines the results from the empirical study. In Section 5, we discuss conclusions, contributions and suggestions for future research.

SDT is a psychological theory of human motivation, development and wellness (Ryan and Deci, 2000a, 2000b). For the purposes of the current paper, two key aspects of the theory are particularly relevant. Firstly, SDT focuses on types, rather than just amount, of motivation. That is, it suggests that besides being more or less motivated, people can also have different types of motivation, ranging from being amotivated to being driven by different types of extrinsic motivation and intrinsic motivation (Ryan and Deci, 2017). As suggested above, this paper focuses on one of these types of motivation, namely, introjected motivation which constitutes an extrinsic and controlled type of motivation according to SDT (Deci et al., 1994; Niemiec et al., 2008).

Secondly, SDT addresses how contextual elements, including control systems and their underlying expectations, can be expected to foster such different types of motivation depending on how they support or thwart people’s basic psychological needs to feel competent, related and autonomous (Niemiec et al., 2006; Ryan and Deci, 2017). When it comes to introjected motivation, SDT suggests this type of motivation is neither driven by the needs being fully supported nor by them being fully thwarted, but rather by a combination of these – i.e. when one or more of the needs are supported while the other(s) are thwarted (e.g. Deci et al., 2017). As argued in the SDT literature, this is typically the case in the type of environments where people are able to achieve need satisfaction only to the extent that they display certain types of behaviour or achieve certain types of outcomes. That is, when the contextual elements require individuals to behave in certain ways or achieve certain outcomes to be perceived as competent or to achieve positive regard (e.g. Assor et al., 2004; Deci et al., 2017). In this paper, we will refer to such conditional regard or contingent approval as the contextual elements relying on a conditional logic [¹].

Below, we will elaborate on these two key aspects of the theory and how they relate to the current study, followed by a discussion of what we currently know about the need supportive and need-thwarting characteristics of results control.

SDT differentiates between different types of motivation along a continuum, including intrinsic motivation, two types of autonomous motivation (identified and integrated), two types of controlled motivation (external and introjected) and amotivation (see Figure 1). The different types of motivation represent different degrees of internalisation and, as a result, different degrees of perceived self-determination (Ryan and Deci, 2000a; Ryan and Deci, 2017). Put differently, they represent differences in the extent to which individuals have “taken in” and integrated values and expectations into their sense of self, ranging from full internalisation (integrated and identified motivation) to partial (introjected motivation) and no internalisation at all (external motivation and amotivation).

As suggested in Figure 1, introjected motivation constitutes a controlled extrinsic type of motivation. It is controlled in the sense that compared to the autonomous types of extrinsic motivation – where people do things because they identify with them or see them as integrated parts of who they are (see right-hand side of Figure 1) – introjectedly motivated people feel pressured to behave in certain ways or display certain types of characteristics (Ryan and Deci, 2017; Ryan et al., 2021). Importantly, though, and in contrast to when people feel pressured out of external contingencies such as other-administered rewards or punishments (see the other controlled type of extrinsic motivation referred to as “external regulation” on the left-hand side of Figure 1), introjectedly motivated individuals engage with activities because of an internalised sense of compulsion (Donald et al., 2020; Howard et al., 2021; Otterpohl et al., 2019). That is, the experienced pressure is internalised and thus comes from within the individual (Ryan et al., 2021).

As suggested by van den Broeck et al. (2021, pp. 243-44) and others (see e.g. Deci et al., 1994; La Guardia, 2009; Niemiec et al., 2008), this type of motivation is typically evident when people do things “out of ego-involvement” or out of “contingent self-esteem”. In such cases, people will reward or punish themselves through self-related emotions such as pride or guilt/shame. Put differently, they will experience that they must, or should, do certain things either to feel better about themselves or avoid feeling bad about themselves (see also Assor et al., 2004; Bartholomew et al., 2011b; Howard et al., 2021). To explain how contextual elements, including the design and use of control systems, may lead to people developing introjected motivation, SDT suggests that when such contextual elements thwart the basic psychological needs of individuals the latter ones are more prone to develop controlled types of motivation than when the contextual elements are supportive of such needs (Ryan and Deci, 2017).

The line of reasoning is as follows. According to SDT, individuals have three basic psychological needs, namely, to feel competent, related and autonomous, and depending on whether these are met or not, the type of motivation that individuals experience for, or develop in relation to, a particular task or a particular setting will vary. Competence relates to the need of individuals to understand what they do and having the feeling that they can succeed at doing it. Such feelings are typically related to the ways in which people feel that they can master the different forms of knowledge, abilities and values that are necessary to succeed in a particular social context. A form of competence that is needed for individuals to be able to interact effectively with the environment (Abeysekera and Dawson, 2015). Relatedness, in turn, refers to the need of individuals to feel that they belong, and are connected, to (important) others (see e.g. Ryan and Deci, 2020). As suggested in the literature, such feelings of connectedness typically involve feelings among individuals that they are being respected and cared for by others (Ryan and Deci, 2000a). Autonomy, finally, relates to the need of individuals to feel that they do what they do because of their own choice (e.g. Ryan and Deci, 2000a; Sheldon and Krieger, 2007). According to extant literature, such feelings of self-determination are typically related to the experience of psychological freedom in the sense that people feel that they are in control of their own actions. Put differently, being autonomous tends to provoke a feeling that the motivational force comes from the inside, and hence, that the perceived locus of causality is internal (Ryan and Deci, 2017).

When these three basic psychological needs are supported by the contextual elements (such as a control system), extrinsically motivated activities can become more autonomously motivated as individuals internalise the extrinsic motives and start to identify with and integrate them as part of who they are (cf. a movement towards the two autonomous types of motivation in Figure 1, referred to as “Identified” and “Integrated”).

In the literature, competence supportive elements that facilitate internalisation typically revolve around things that allow individuals to feel that they are capable and effective in their interactions with their environment. Generally speaking, such elements typically offer clear expectations, appropriate challenges and informational feedback (Ryan and Deci, 2017). Clear expectations help people understand what is expected of them which, in turn, reduces confusion and uncertainty. When goals or tasks are clear, people are more likely to feel confident in their ability to succeed (Haerens et al., 2013; Ryan and Deci, 2017). In a similar manner, appropriate challenges – i.e. neither too easy challenges as these lead to boredom nor too hard challenges as these lead to frustration – help promote a feeling of competence by allowing people to stretch their abilities while still experiencing success (Deci and Ryan, 2000). Finally, through regular and informative feedback people can track their progress which, in turn, allows them to recognise growth, correct mistakes and feel a sense of mastery (Ryan and Deci, 2017; Slemp et al., 2021).

Relatedness supportive contextual elements, in turn, are those that foster a sense of belonging, connection and mutual respect among individuals. In short, environments where people feel cared for, understood, and valued by others (Ryan and Deci, 2020). The premise is, as suggested by Ryan and Deci (2017, p. 447) that “People are more willing to internalize ideas and inputs from people to whom they feel connected”. Elements in the environment that tend to elicit such feelings include where people are allowed to be involved in, for example, decision-making, group work or joint activities. The premise is that such participation and inclusion reinforce their sense of being valued members of the group. Moreover, when superiors or other group members pay careful attention to, and take “sincere interest in the person, being curious and engaged in his or her experiences”, this will arguably reinforce such feelings (Ryan and Deci, 2017, p. 447). Other ways to elicit such feelings include where important others show that they are prepared to dedicate their time and energy into issues that matter for an individual, showing them that they are available in case of need, and showing honesty and reliability (e.g. Haerens et al., 2013; Ryan and Deci, 2017).

Finally, when it comes to autonomy-supportive contextual elements in this literature, these typically revolve around the three factors suggested by Deci et al. (1994, p. 119), namely, “providing a meaningful rationale, acknowledging the behaver’s feelings, and conveying choice”. A common denominator of these factors is that they all help support an individual’s feeling of being self-determined or that the motivational force comes from the inside. For example, a meaningful rationale helps individuals understand why a request or task is valuable or necessary which, in turn, can be expected to facilitate internalisation. Moreover, in doing so, they should be more prone to experience the related actions as self-endorsed (Gagné and Deci, 2005; Ryan and Deci, 2017; Vansteenkiste et al., 2018). In a similar manner, showing respect for and interest in the perspective of the individual, allowing them to participate and have a saying in what to do and how to do it, is critical to feeling volitional. Not least this is the case when people feel frustrated over, or disinterested in, a particular task or demand (Vansteenkiste et al., 2019). As suggested by Ryan and Deci (2017, p. 445), opportunities “to provide meaningful inputs and to actively participate (or not) in establishing and monitoring goals is critical to feeling volitional” (Ryan and Deci, 2017, p. 445). Finally, and not least, the feeling of self-endorsement can be fostered through “affordances of choice and encouragement of self-regulation” (Ryan and Deci, 2017, p. 12). The premise is, of course, that when being able to choose, people will experience increased degrees of self-determination.

In contrast to this, SDT suggests that when the three basic psychological needs are thwarted by the contextual elements, those who are intrinsically motivated from the beginning can become more controlled in their motivation through crowding-out processes (cf. a movement from right to left in Figure 1; Ryan and Deci, 2017), while those who are already controlled in their motivation can be expected to remain so.

Generally speaking, many of the elements that are seen as need thwarting in the literature constitute the very opposite of the need-supporting elements outlined above [²]. For example, when it comes to competence, Aelterman et al. (2019, pp. 500-501) point to how the need for structure and clear expectations in a learning context can be thwarted when people in control adopt what they refer to as an awaiting approach, as this typically leads to unclear or even contradictory requirements and expectations for learners (see also Perlman, 2013). In the school context that they studied, they suggest this can lead to that “students may experience the learning environment as confusing, and may feel incapable and uncertain as to how to proceed”. In a similar context, Haerens et al. (2016, p. 63) point to how teachers can “create confusion among students by exerting an illogical and incoherent structure when introducing tasks”. Moreover, such confusion can also be triggered by the giving of ambiguous feedback or destructive criticism (Haerens et al., 2016). In fact, Bartholomew et al. (2018, p. 52) pointed to how being “‘criticized’ in front of their peers […] may increase students’ fears about failing” and make them doubt their own capabilities, while Ryan and Deci (2017, p. 156) concluded that those aspects of negative feedback that diminish the perceived competence among individuals will in fact decrease their intrinsic motivation (see also Bartholomew et al., 2010; Gonzales et al., 2024).

In a similar manner, need-thwarting environments can also frustrate people’s need to feel related. For example, and in contrast to relatedness supportive elements, such environments include where important others ignore individuals and their perspectives, remain unengaged in them, or contribute to a cold or unfriendly climate (see e.g. Ryan and Deci, 2017; Perlman, 2013). Along these lines, for example, Haerens et al. (2016, p. 63) stated that “the need for relatedness can be thwarted when teachers create an emotionally cold learning environment for students. Uninvolved teachers are unfriendly, and even reject or exclude (some) students”.

Finally, SDT suggests that contextual elements tend to thwart the need for autonomy when they are controlling in the sense that they “pressure” people into thinking, feeling and behaving in particular ways (e.g. Assor et al., 2004). For example, this is the case when very specific and preconceived ways of thinking and behaving are imposed upon people (Bartholomew et al., 2011a, 2011b) or when they are done so in an intrusive (Gonzales et al., 2024) or very rigid and inflexible way (Slemp et al., 2018). In a school context, for example, Perlman (2013, p. 415) pointed to how “a controlling environment uses instructional aspects that focus on external factors (e.g., guilt or rewards), are strict within their communication (e.g., deadlines and guilt), ignore children who struggle and attempt to demonstrate power, and pressure them to complete certain tasks”. In contrast to autonomy-supportive environments, controlling environments thus leave little room for choice; they typically bypass the viewpoint of the individual being controlled, and they do not focus on the rationale for why something should be done in a particular way (Aelterman et al., 2019; De Meyer et al., 2014; Haerens et al., 2016; Vansteenkiste and Ryan, 2013).

Importantly, to foster introjected motivation, existing SDT-based literature suggests that environments have to be both need supportive and need thwarting in the senses just referred to above. The theoretical argument for this is that the need for autonomy must be thwarted in some way for introjected motivation to arise, as this is what hinders people from internalising more fully the environmental elements. At the same time, there must be need supportive elements for internalisation to take place at all. Hence, while people must experience that their autonomy is thwarted, they must concurrently experience that their competence and/or relatedness is supported. As repeatedly suggested in the SDT literature, this is typical of the type of controlling environments where various forms of conditioning are prevalent, such as conditional regard and contingent approval (e.g. Assor et al., 2004; Haines and Schutte, 2023; Niemiec et al., 2008, 2010). Generally speaking, and again, such conditioning refers to the practice of offering regard, approval or attention only when people behave in certain desired ways and withholding it when they do not. For example, in a parenting context, conditional regard means that “parents provide more attention and affection when their children act or are as the parents desire and provide less attention and affection when the children do not act or are not as they desire” (Ryan and Deci, 2017, p. 334; see also Assor et al., 2004).

Generally speaking, this type of conditioning relies on a contingent “if-then-logic” where some individuals are typically pressured into behaving in a certain way to gain or avoid losing the affection of others (Gonzales et al., 2024; Sarmah et al., 2022). In terms of the basic psychological needs, this essentially means two things. On the one hand, it means that even if the environment is essentially controlling, not all needs are thwarted. On the contrary, those who experience conditional regard or contingent approval are typically offered support for their need for competence and relatedness as long as they are able to live up to the expectations of others. On the other hand, though, it means that they are never able to satisfy all three needs at once, because the need for autonomy is pitted against other needs (in particular, the need for relatedness). The premise is that to maintain their relatedness, individuals “must comply, irrespective of the fit of the demands with their own interests or with the values they may develop over time” (Ryan and Deci, 2017, p. 339).

As argued by Assor et al. (2004, p. 54) and others (Haines and Schutte, 2023; Ryan and Deci, 2017), this type of “conditional affection represents a prototypic context for promoting introjection” because of how the contingent esteem offered by others “can be readily transformed into the contingent self-esteem that underlies introjected regulation”. People learn, and thus internalise, what it takes to gain or maintain their relationships with others at the cost of their own self-determination. A type of internalisation that is often accompanied by the type of feelings associated with introjected motivation, namely, internal pressure, cognitive conflict and contingent self-esteem (Assor et al., 2004; Bartholomew et al., 2018; Ryan and Deci, 2017).

As suggested in the introduction, the MC literature has had a rather scant focus on how control systems can foster introjected motivation in the ways discussed in the SDT literature (Englund and Gerdin, 2025). Indeed, some studies refer to this type of motivation as such, pointing to how it can result in that organisational members “perform activities, for example, to avoid shame” (Bouten and Hoozée, 2022, p. 6), to “avoid guilt or anxiety or to attain self- and other-approval” (Pfister and Lukka, 2019, pp. 347-348). However, such claims are typically based on general references to SDT as such. When it comes to exploring empirically how and why MC systems can result in such behaviour, though, the existing literature provides very limited insights. Notwithstanding this, though, the existing MC literature has provided important insights into how and why MC systems can affect the basic psychological needs of individuals, not least when it comes to supporting such needs but also to some extent when it comes to their thwarting. Below, we discuss these insights as a general theoretical background to the current study.

When it comes to the basic psychological need for competence, existing literature has mainly pointed to how results control can support this need through providing different types of feedback on how individuals and groups perform in relation to expectations. Sometimes, focus is on the information provided and rewards decided by the systems, and sometimes, more on how managers use such information to provide verbal feedback to employees. Starting out with the design of such systems, several studies have pointed to how results-based information allows individuals a form of feedback on how they perform in relation to targets and also in relation to others (see e.g. Groen et al., 2017; Pfister and Lukka, 2019; Van der Hauwaert et al., 2022). For example, Groen et al. (2017, p. 55) suggested that when employees are rewarded in (non)monetary terms based on performance information, they get information “about how well they do relative to others and relative to targets, and receiving such information may help satisfy the need for competence and hence increase autonomous motivation”. In a similar manner, Pfister and Lukka (2019, p. 361) pointed out that “Employees’ psychological need for competence can be further reinforced through positively promoting feedback on task effectiveness and related forms of recognition”. The premise is that such information may not only boost the confidence of individuals more generally (Pfister and Lukka, 2019) but also allow them to become better at doing their job (Van der Hauwaert et al., 2022). When it comes to the more particular use of such information by managers, some scholars have emphasised how this may have positive effects on the need for competence, while others have identified potentially negative consequences. For example, Groen et al. (2017, p. 55), concluded that “The need for competence is satisfied during the substantive discussion and evaluation of performance, because in such discussions performance metrics are seen as informational”. In a similar manner, De Baerdemaeker and Bruggeman (2015, p. 4) reasoned that when managers are allowed a part in strategic control discussions, they can get “the feeling that they are better able to control their environment” and, hence, feel more competent (see also Mazbayeva et al., 2022).

In contrast, it has to some extent also been discussed how the design and use of results control can thwart the need for competence. In particular, this has been suggested when the targets set up by a results control system are perceived as so challenging that they become more or less impossible to reach (Lau and Roopnarain, 2014; Sutton and Brown, 2016; Van der Hauwaert and Bruggeman, 2015) or that they are set up in ways that increase the uncertainty or feelings of unfairness among those who are being controlled (see e.g. Groen, 2018; Siverbo, 2023). Moreover, the need for competence can also be frustrated when the control system provides performance information related to such targets that is either demeaning (e.g. Lau and Roopnarain, 2014; Pfister and Lukka, 2019) or unreliable (Kunz and Linder, 2012) or when it provides evidence of poor results or failure (Van der Kolk et al., 2019). Along these lines, Kunz and Linder (2012, p. 601) stressed that when there is a low reliability of the information provided by the system, it “cannot inform employees about their performance, and thereby undermines the need for competence”, while Van der Kolk et al. (2019, p. 9) stressed that while feedback information can indeed trigger feelings of competence and of being in control, it can also lead to “feelings of incompetence”, especially “when specific results are not achieved”.

When it comes to the basic psychological need for relatedness, existing literature has mainly pointed to how results control can support this need. Oftentimes, this is done with reference to how the system is used by managers and to organisational members’ participation in its development and use. However, there are a few scholars who point more generally to design aspects of the system, such as when it is designed in such a way that it “facilitates and stimulates communication with other people in the organization […] since this can contribute] to a work context that provides a sense of belongingness and connectedness, which is expected to satisfy managers’ need for relatedness” (Van der Hauwaert et al., 2022, p. 4). As suggested by Van der Hauwaert and Bruggeman (2015), this can be the case when the control system is designed so that local information is linked to information about the company as a whole or when managers are allowed to make necessary changes to the system.

Most common, though, is to emphasise how social interactions surrounding the control system can help satisfy the basic psychological need for relatedness. For example, Groen et al. (2017, pp. 55-56) suggested that “The need for relatedness is satisfied through the social interaction that takes place during the evaluation [by means of performance information]”, while Hoozée and Ngo (2018) pointed to how “Positive interactions with colleagues in discussions about the factors influencing costs could trigger managers’ sense of relatedness because they feed the sense of group belongingness” (Hoozée and Ngo, 2018, p. 752). The premise is that such control-oriented social interactions not only have the potential to enhance the internal communication between different parties, but also to create “a climate of shared effort” (De Baerdemaeker and Bruggeman, 2015, p. 4), to stimulate “an atmosphere of caring for and being recognized by others” (Hoozée and Ngo, 2018, p. 752), and to demonstrate that different views are important (Groen, 2018). Importantly, though, as Kunz and Linder (2012) suggested, this requires that the control system information is reliable, because otherwise people will have difficulty feeling connected and develop trust in important others.

Finally, when it comes to the basic psychological need for autonomy, some scholars stress that a focus on outcomes allows for a certain freedom when it comes to the initiation, design and execution of actions leading to such outcomes (Pfister and Lukka, 2019; Sutton and Brown, 2016; Van der Hauwaert et al., 2022). It has also been stressed that such metrics can provide people with a meaningful rationale for doing their work, thereby granting them a sense of personal relevance and autonomy (Van der Hauwaert et al., 2022). In a similar manner, it has been pointed to how combining outcome metrics with more task-oriented metrics can support the need for autonomy. For example, Guo et al. (2012, p. 215) emphasised that driver measures can be perceived as autonomy-supportive in this way because they “are more controllable and thus less noisy than outcome measures”, while Speklé et al. (2017, p. 76) argued that “if managers use a diverse set of performance measures to capture the key performance areas of the business unit, then managers experience more freedom, autonomy, and opportunity to do their job”.

In contrast to this, though, it has also been stressed that outcome and driver metrics can be highly thwarting of the need for autonomy. Reasons for this include that people can feel restricted or limited when being “obliged to work to the performance measures” (Siverbo, 2023, p. 2; see also van der Kolk et al., 2019), not least when such metrics are oriented towards particular tasks (Pfister and Lukka, 2019) or when they “define hypothesized causal links between drivers and outcomes [as this leaves] less flexibility and freedom for managers to pursue alternative value creating actions” (Guo et al., 2012, p. 215). Moreover, when such metrics are linked to monetary and non-monetary rewards, the autonomy-thwarting effects can be further strengthened. The premise is that such incentives include an oppressive or controlling element (see e.g. Groen et al., 2017; Guo et al., 2014), in the sense that they limit the freedom of people through making, for example, bonuses or promotions conditional upon certain actions or achievements.

To avoid the autonomy-thwarting effects, though, several researchers point to how result controls, regardless of their particular design, can be used in ways that make them supportive of individuals’ need for autonomy. According to the existing literature, examples of such use include where performance evaluations are “employee driven” in the sense that employees are able to affect whether, and how often, evaluations are to take place (Holderness et al., 2020; Sutton and Brown, 2016), or that they are entrusted with a form of “self-monitoring” (Liboriussen et al., 2021). Other examples include where performance information is used in a holistic (Sutton and Brown, 2016) or interactive (Matsuo et al., 2021; Speklé et al., 2017) way, whereby focus becomes less on particular numbers and deviations, and more on “searching out and learning about new opportunities for improvement” (Matsuo et al., 2021, p. 268); a focus which is expected to increase individuals’ feelings of psychological empowerment and autonomy.

To summarise, existing MC literature has argued and to some extent shown that results control systems can be designed and used in ways that both support and thwart people’s basic psychological needs to feel competent, related and autonomous. Again, though, this has almost exclusively been done with the purpose of identifying qualities of the results control system that support or hinder the development of autonomous motivation. In contrast, it is still unclear why such qualities lead to people developing introjected motivation. It is to this question that we turn in the current study.

The empirical material presented in the paper is based on a large research project focusing on results control and its impact on researchers from Swedish universities [³]. Arguably, the Swedish context is interesting in this sense, given how Swedish universities are increasingly governed by means of various forms of Performance Measurement Systems (PMS). For example, nowadays Swedish universities are partly funded by the government based on how they perform in relation to other universities on two indicators, namely, one that measures the number of publications and citations, and one that measures the amount of external funding attracted (Hammarfelt et al., 2016). This has not only resulted in a general increase in the pressure to publish in peer-reviewed international journals, in particular those included in Web of Science (as normalised citation scores from Web of Science are used in the indicators). It has also resulted in that most universities have developed similar PMSs and incentive systems at the local level. In fact, at all focal universities under study in this paper, there were explicit targets, follow-ups of and incentives linked to “external research grants”, “number of publications” and “number of citations”.

The data drawn upon in the current paper mainly consist of transcripts from interviews with 20 researchers. Given our overall purpose, we started out by interviewing researchers from one of the top 10 universities in Sweden, based on the Times Higher Education World University Ranking (which is founded on the type of performance indicators referred to above). Within this university – henceforth referred to as University “1” – respondents from two different scientific fields were approached, namely: Natural science and Engineering and Humanities and Social Sciences. Also, in terms of seniority, we intentionally selected researchers with varying levels of tenure, ranging from assistant to full professors. The reason for this design was that, based on extant literature, we expected that individual academics’ experiences of how they are governed can vary depending on, for example, the setting in which they work, the faculty to which they belong and their academic position (see e.g. Chatelain-Ponroy et al., 2018; Guarini et al., 2020). To find out whether such experiences differed between universities, we also contacted a number of researchers from another university ranked top 20 in Sweden on the Times Higher Education World University Ranking (referred to as University “2” below) and from one ranked top 5 (referred to as University “3” below). The reason for considering their ranking position was that, apart from getting insights on possible differences between universities, we also wanted to see whether the governing practices differed between higher- and lower-ranked universities. In total, another six interviews were conducted at these universities, resulting in a total of 20 interviews altogether. Table 1 provides more detailed information on each respondent.

The interviews were semi-structured in the sense that we used an interview guide to structure our conversations, but depending on how each conversation proceeded we did not necessarily ask each interviewee the exact same questions. The interview guide was developed based on previous accounting- and higher education literature focusing on the use of results control to govern academic research(ers). Based on this guide, the interviews focused on three themes, namely, what it means to be a researcher, the existence and use of results control in academia (with a particular focus on PMSs) and if/how such PMSs affect them and their work as researchers. Within each such theme, the interview guide contained several rather “open-ended” and general questions. For example, related to the first theme, we asked: “What does it mean to be successful as a researcher in your view?” and “Is it important to be successful as a researcher?”, and depending on their answers, we asked related follow-up questions and asked them to elaborate or to exemplify. All interview questions included in the interview guide are available in  Appendix.

When we started the sub-project reported upon in the current paper, we did not depart from SDT as such. Rather, our interest in this particular theory and the type of questions it brings with it, was something that emerged during the empirical study. In fact, as we started to analyse the empirical material, a theme of motivation and results control emerged. Essentially, this resulted in that we, on the one hand, took an interest in how our respondents were managed based on their (quantified) performances and, on the other hand, if and how such performance management could be linked to different types of motivation.

Two insights made during the early parts of our analysis were particularly important for our further analytical work. Firstly, we could not find any obvious systematic differences when it came to how the respondents experienced the control systems in use. On the contrary, they expressed a surprisingly similar view on the growing importance of rather specific performance targets and measures in academia, regardless of the university or faculty they represented, their gender or their academic position. Secondly, we noted that although our respondents did seem to identify themselves with certain aspects of such performance targets and measures and thus displayed some tendencies of autonomous motivation, we found considerably more evidence of introjected motivation in the material. In fact, while most of our respondents seemed to act largely in line with what the control system wanted from them – which could be a sign of autonomous motivation – such actions were typically accompanied by a number of aspects associated with introjected motivation, namely, feelings of pressure, cognitive conflicts and a high awareness of how they were perceived by others through their performances (cf. subsection 2.2). Examples of codings capturing such aspects included expressions about the pressure they felt to constantly develop themselves to be able to perform: “we’re fed with this [information and expectations on how to improve] all the time, […] I could go to meetings several times every week if I wanted to, each with the purpose to develop myself in various ways. It’s [meetings] about applying for research funds and becoming better in all sorts of ways” (I2). Or the perceived pressure to act in certain ways so as to look good through their numbers: “It has changed somewhat [what motivates people]. It has to do with the incentives, that the personal career has become more important. When I got my PhD [many years ago] you never heard anyone saying: ‘that will look good on my CV’. That’s a new type of saying for me, and I think it has to do with the incentives [of the control system]” (I1).

Based on these two insights, we decided not to focus on potential differences in the empirical material. Instead, we decided to work with all 20 interviews collectively and to use this material as a means for conceptualising and theorising the potential relationship between results control and introjected motivation [⁴]. With such an emerging focus, we sorted out all passages that had already been coded as “introjected motivation”, (re)read them and looked for instances where the respondents linked the control system to either of the basic psychological needs for competence, relatedness or autonomy. For competence, we coded all passages where the control system was seen as providing different types of feedback on how they performed in relation to expectations, and thus, constructed them as researchers in particular ways (e.g. as successful or not, as academically junior or senior, etc.). In some cases, the control system was referred to in a positive manner, in the sense that it supported or highlighted their competence and sometimes it was referred to a more negative manner. We marked all instances where the feedback had a positive character with a “C⁺” and all instances where it had a negative character with a “C⁻”, through “scribbling and doodling” in the margins of our interview transcripts (Bazeley, 2020, p. 138). An example of the former was where the control system was seen as crucial when it came to promotions, because then “journal articles are worth the most […] if you have a CV with a sufficient number of them, then you’re a good researcher” (I2). In contrast, the control system could also provide the opposite type of feedback, “because you’re always made aware that you’re not good enough” (I14).

For relatedness, we coded all passages where the control system was linked to, and affected, their feelings of belongingness and connectedness with others within their organisation. When the control system was perceived as supportive of, and contributing towards, such feelings we used the code “R⁺” [e.g. “when someone gets a pile of research money, it’s acknowledged by management” (I10)] and when it hindered or aggravated them, we used the code “R⁻” [e.g. “you don’t want to ‘be’ the paper that has three citations or none. You want to reach a point where you become relevant [to others]” (I10)].

For autonomy, finally, we coded all passages where respondents linked the control system to their ability to control their own actions or to actions being experienced as volitional and self-endorsed. Just as was the case with competence and relatedness, the control system was sometimes referred to in a positive manner in this sense, such as where it was seen to allow for, or support, this type of autonomy [e.g. “it gives you many degrees of freedom” (I15)]. We marked all such instances with the code “A⁺”. In many cases, though, the control system was seen as limiting their autonomy as they felt that they “had to” adapt to what the system wanted from them. For example, this was expressed in terms of “Right now it’s about funding and visibility and that’s the ways it is […] you just have to accept it” (I5), or that “I have to keep acquiring further qualifications so that I can get more research funding, and then I can’t jump on every bandwagon […] it’s too risky” (I15). We marked all such instances with the code “A⁻”.

While these codings provided a form of overall pattern when it came to how results control seemed linked to the basic psychological needs and, further, to introjected motivation, it was through further scrutinising these codings that the main empirical insights of this paper emerged. A first insight relates to the type of qualities of results control that provide support for individuals’ needs to feel competent, related and autonomous. In fact, through our further analyses of the empirical instances related to each individual code, it became obvious that it was not only a question of whether the control system supported or thwarted a particular need or not. Instead, it seemed as if it was more important how this was done. For example, when reading the empirical instances for the two competency-codes (C^+,−), we found that when our respondents received feedback on their performances, such feedback was based on the very specific performance criteria set up by the control system, such as publishing articles in certain outlets, attracting external research funds from particular high-esteem funders, etc. In fact, how individual researchers performed in relation to these performance criteria seemed to largely condition the type of need support they were offered. Based on this, we developed the quality that we refer to as a conditional logic of the results control system, which seemed particularly important for understanding the type of need support that can lead to introjected motivation (see subsection 4.2). As we did so, we also re-read and consulted the SDT literature, which has provided important insights into similar concepts, including conditional regard (e.g. Assor et al., 2004; Haines and Schutte, 2023; Ryan and Deci, 2017) and contingent approval (e.g. Niemiec et al., 2008, 2010). In our analyses, we both drew upon and contrasted our empirical findings to this literature.

A second insight relates to how the conditional logic was turned into introjected motivation by the researchers. We write “by the researchers” because our further scrutiny of the codings showed that the introjected motivation among researchers did not constitute an automatic effect of how they experienced the results control system. On the contrary, it seemed as if their motivation was more a result of how they reacted to, and tried to handle, the type of feedback that they got from the results control system. Put differently, it seemed as if their motivation grew from their attempts to “cope with” the conditional logic of the control system, which on the one hand allowed them to secure a certain sense of need satisfaction and on the other hand resulted in feelings of internal pressure and tension (not least because of how their need for autonomy was thwarted). In our theorising of the empirical material below, we acknowledge researchers’ active involvement in turning the need for support of the control system into their own motivation in this sense, through the notion of a coping strategy referred to as “willing compliance” (see subsection 4.3). Also, when it came to this insight, an important step in our analytical work consisted of consulting the existing SDT literature. As suggested above, though, considerably less work has been done in this literature on how individuals actually interpret and handle conditional elements in their environments, due to the cross-sectional and correlational nature of much of this research (cf. Haines and Schutte, 2023; McAnally and Hagger, 2024).

In Section 4, we present and empirically substantiate these two insights that emerged through working back and forth between the empirical material and the extant literature (cf. Ahrens and Chapman, 2006; Bazeley, 2020).

In this section, we present the findings from our empirical analyses. We start out by describing the results of control systems in use at the universities (subsection 4.1). In a following section, we show how these systems held certain qualities that both supported and thwarted the basic psychological needs of the researchers, with a particular focus on how it did so by means of a conditional logic (subsection 4.2). In a final section, we discuss and illustrate how this conditional logic, in turn, contributed to only a partial internalisation of the control systems and thus to the researchers being introjectedly motivated (subsection 4.3).

As suggested previously, the empirical context under study was a context where the design and use of results control had come to play a prominent role in the sense that there were clear performance targets related to the number of research publications, the number of citations and to the amount of external research funding attracted. Moreover, the targets were systematically followed up by means of several performance measures and there were several monetary and non-monetary incentives linked to the outcomes of such follow-ups [⁵].

Starting out with the performance targets, our respondents pointed to how “the measurement systems we have today, they focus on the number of publications, international publications and to be cited” (Interviewee no. 13, henceforth abbreviated I13), “over the last five years there has become this enormous pressure, that you’re measured on the number of citations you have and the number of publications” (I6), “you need to publish […] and above all you should attract funding from these bigger funders” (I3). Moreover, they explained how there were clear “signals from above [… in terms of] numerical targets” (I1) for each area. For example, related to external funding, several respondents from one of the universities referred to a formal target “to double the amount of external funding that we attract” (I8), while “for publications” (I12) and “the number of citations” (I6), the targets were expressed more in terms of a general increase in these.

All three areas were also systematically followed up, reported and discussed in various settings. For example, it was referred to how research leaders or administrators would “collect everything that we have done this year” (I4), “at some point during the autumn you report on your publications, your conferences, presentations, applications for funding” (I13), “you account for everything the research group has done once a year […]. It’s external funding, publications and so on” (I17), and then “we talk about it in our different meetings, we follow up on […] like the last time, it was individual researchers and how much we publish, the number of research applications we have submitted […] and in the yearly activity report of course, where we report the number of citations, publications and so on” (I9).

Finally, the reported performance numbers also functioned as the basis for different types of rewards, both monetary and non-monetary ones. Examples of the former included references to “a publication bonus” (I9), where “you get a certain amount of money based on what you’ve written, and the amount of money also depends on which journal it is” (I18), and the fact that “those who have published more get more paid, so of course it’s a salary incentive” (I3). Examples of the latter included references to how performance numbers would underly the ranking of candidates for a position (e.g. “articles are valued the most”, I20) and the development of individual careers (“to become a docent you have to publish articles, that’s the only thing that counts”, I4). Moreover, such numbers also determined or affected who would be celebrated within the universities: “there are these informal structures [that guide] what should be celebrated”, which means that “when someone has received external funding or released a publication, it is noted in the internal newsletter” or by “the head of department […], where she can say that ‘this year, it is gratifying to see that [the department] has been successful’, and then she mentions those who have attracted external funding” (I2).

While the performance targets, follow-ups and incentives described above suggest the results control systems were designed to function as external regulators of the researchers’ behaviour, our findings strongly suggest that these external regulators had, in turn, been at least partially internalised by the researchers. That is, they had become part of the inner motivational structure of the researchers, to the extent that their behaviour no longer seemed oriented towards any particular external pressure or to reach, for example, a publication bonus or salary increase. Instead, the pressure to strive for, and pursue, what the system wanted from them seemed to come from within themselves. Below, we will show how this can be understood against the background of how the results control systems not only served as a basis for various forms of rewards (which again is typical of control systems functioning as external regulators) but also for supporting the basic psychological needs of the researchers, which according to SDT is a theoretical prerequisite for individuals to internalise this type of external regulators (cf. Ryan and Deci, 2000a, 2017). However, we will also show that such need support was not enough for full internalisation to take place because it was offered by means of a conditional logic.

As suggested in Section 2, existing MC literature has pointed to how the type of results control system depicted above can be expected to be internalised by individuals when its design and/or use support their basic psychological needs to feel competent, related and autonomous. Specifically, it has been suggested that this will be the case when it supports the need for competence through providing different types of feedback on how individuals and groups perform in relation to expectations. The premise is that such feedback can, apart from providing a direct verification of their competency, function as a means for learning and further development of their competence (see e.g. Groen et al., 2017; Pfister and Lukka, 2019; Van der Hauwaert et al., 2022). Moreover, it has been suggested that the likelihood that results control will be internalised increases when it, for example, is designed in such a way that it “facilitates and stimulates communication with other people in the organization [… as this can contribute] to a work context that provides a sense of belongingness and connectedness” (Van der Hauwaert et al., 2022, p. 4; see also Groen et al., 2017; Sutton and Brown, 2016). Finally, the likelihood of internalisation also increases due to results control’s focus on outcomes rather than pre-specified behaviours, as this allows for a certain autonomy or freedom when it comes to the initiation, design and execution of actions leading to such outcomes (Pfister and Lukka, 2019; Sutton and Brown, 2016; Van der Hauwaert et al., 2022).

On the one hand, and generally speaking, our findings largely corroborate these previous insights that results control can help satisfy all three basic psychological needs and thus contribute towards internalisation. Again, the performance measures referred to above, clearly worked to provide targets for the researchers under study and through following up on, and providing incentives linked to, such targets, it was evident that the control systems not only worked to provide numbers that visualised researchers and their work but also that such numbers clearly affected how researchers perceived themselves and their competencies. For example, it was referred to how the performance indicators provided this type of competence-relevant feedback as they worked as a “measure of whether a person is competent within the area and [whether they] can be seen as successful” (I8).

In a similar manner, our findings point to how the numbers served as a basis for the relatedness-supporting type of communication referred to by Van der Hauwaert et al. (2022). As a result, they allowed for individual researchers to become part of and included in particular groups or categories of researchers (e.g. “successful researchers” and “full professors”) and also to become accepted by important others (e.g. through being nominated to certain tasks or getting appointed to certain committees). As suggested by the following example, this helped to create a dialogue and a sense of belongingness to a larger academic community: “it’s good for keeping track of what we do within our subject, and to acknowledge what everyone is doing. Otherwise, it easily becomes this power game and an unbalance when it comes to who are allowed to be seen and heard. This [i.e. the communication around the numbers] hopefully means that everyone’s situation is brought to the fore, that we become aware of what we publish, the projects we have, that we can come together in applying for funding, and so on” (I9).

Finally, our findings also point to how internalisation was enabled by the perception of results control as a form of control that allows for a certain form of freedom and autonomy. The reason, as explained by the researchers, was that although the control system provided certain performance targets to aim for, there were still plenty of room for manoeuvring when it came to how, when and why to achieve such targets: “as long as they [i.e. university management] can see it in their systems, that is the only thing they care about. They don’t take an interest in what I do, and I’m grateful for that, that they don’t interfere in how I do things” (I14).

On the other hand, though, our findings also show that the just-described need-supporting qualities of results control systems worked in a very particular way, namely, through a conditional logic. That is, these systems departed from the idea that to be seen as competent, to be accepted and trusted among peers, and to be allowed the type of manoeuvring space referred to above, researchers had to prove themselves by delivering “good numbers”. Put differently, the ways in which the results control systems helped satisfy the psychological needs was neither automatic nor unconditional. On the contrary, the offered need support was only for those who were able to perform well according to the performance criteria set up by the systems. The result of this was that the control system not only supported the needs but also threatened to thwart them, in particular for those researchers who were unable to live up to its expectations.

For example, this became very clear when our respondents talked about what it meant to be a competent or successful researcher, as this was highly conditional on being able to deliver what the systems wanted from them: “it’s only publications that count” (I16), “you need a certain volume of research output” (I14) or that you need “an extensive publishing” (I3), “preferably in high-ranked journals” (I17). Or, as suggested by yet other respondents: “the measurement systems we have today, they focus on the number of publications, international publications and to be cited” (I13), “over the last five years there has become this enormous pressure, that you’re measured on the number of citations you have and the number of publications” (I6), and that “publication indexes, the number of publications and funding are growing more and more important” (I5). Hence, competency was defined in a very particular way, and it was primarily through aligning themselves with this definition that the researchers could secure competence support from the control system.

Similar observations were made when it came to the need for relatedness, namely that the numbers largely conditioned who would be recognised, entrusted with assignments and allowed access to certain meetings. That is, individual researchers tended to be recognised and appreciated more depending on how they performed according to the performance measures. Consequently, it was only those who were able to live up to the control system’s performance criteria that could secure such relatedness support. Along these lines, for example, several of our respondents suggested that you are recognised as an individual first and foremost when you are “granted prestigious money” (I3), “if you get external money instead of internal” (I12) or “if you get published in the top three journals” (I10). When that happens, “you get more status” (I3), “you get a good reputation” (I12). Or as put by another respondent: “It’s simple, you just rise […]. That kind of colleagues [i.e. those who have ‘good numbers’] get an enormous amount of respect” (I17). “It’s like, when you get funding, you move up one notch, when you get something published, you move up another. Maybe if your results get some media attention, you move up another notch” (I2).

Finally, it also seemed that the autonomy of researchers was contingent on the type of numbers that could be provided, in the sense that the ability to provide good numbers granted them a form of freedom to design and conduct their research without interference from superiors. This was sometimes referred to in more general terms, like: “I think you have to prove yourself. It’s always like that, if you prove yourself, then you get the freedom to do more” (I19). In other cases, it was linked to, for example, the ability to publish articles: “the more articles you publish, the better you’re off, not least as a professor, but also as an associate professor” (I7) or to the ability to attract external funding (or both). The premise is, as suggested by one of the respondents, that “funding depends on measurable success. So those who show that they can publish will get funding” (I5), and “when you’re good at attracting money to the university and to research, it almost gives you free reins to do whatever you like […]. I mean, the vice chancellor and the central [administration], then they say that: ‘if you’re good, we won’t interfere that much” (I19).

To summarise thus far, the empirical findings show that the results control systems certainly helped support the needs of researchers to feel competent, related and autonomous. However, it did so by means of a conditional logic, which meant that researchers had to perform well to achieve such support, where performing well typically meant to outperform others when it came to publishing articles in well-renowned journals, accumulating citations and attracting external funds for research. The result of this conditional logic was that there was a constant threat of need thwarting, in particular for those who were not able to live up the expectations. Below, we discuss how this conditional logic resulted in introjected motivation among the researchers.

Our analyses suggest that the type of conditional logic outlined and empirically substantiated in the previous section had made researchers introjectedly motivated. Two related types of empirical observations underly this conclusion. Firstly, we noticed that even though most of the researchers were highly critical of the results control system in principle, they were still very numbers-oriented in the sense that they tried to achieve what the control system wanted them to achieve. Importantly though, this did not seem to be grounded in feelings of having to do so because of any external force or requirements, which would have been the case if they were externally regulated by the control system. Rather, through reconciling or aligning their behaviour with the expectations of the control system, they were able to satisfy at least parts of their needs to feel competent, related and autonomous. A form of willing compliance, as it were, that came from within.

Secondly, and related, we found that such an intentional and rather instrumental numbers-orientation was typically associated with feelings of internal pressure and tension. This, we suggest, can be understood in terms of how their willing compliance meant that they had to give up some aspects of their needs to the benefit of other aspects. The result, it seems, was a partial internalisation where researchers never really identified themselves with the expectations underlying the control systems. Below, we illustrate these two interrelated aspects of the introjected regulatory process found in the current study and show how they are grounded in the conditional logic underlying the results control systems.

As suggested above, the first aspect suggesting that researchers had become introjectedly motivated by the results control systems relates to their strong numbers-orientation, despite being highly critical of such an orientation in principle. In fact, many of the respondents not only had clear doubts about the suitability and meaningfulness of measuring the performances of academic researchers by means of a few performance measures, but also of tying e.g. recruitments, promotions and research funding so strongly to such measurements. For example, it was suggested that “we’re victims of this […] New Public Management” (I4) where the “quantifiable” has become more important “than the research questions that people pursue” (I9); “The only thing that counts is what you have published, unfortunately” (I15) which has resulted in that people “only write stuff to get published” (I7). Other things are “not recognized […] it’s just articles” (I3). As a consequence of this, they pointed to how research is increasingly “becoming instrumental” (I9). It becomes a form of “mass production” (I20) or “industry” (I9) where “you get the wrong focus” (I13) in the sense that “publishing [per se] becomes the main focus” (I13), and when that is the case, “you never tell the whole story. Instead, you divide it [into multiple papers, because] you are occupied with these fragments that you try to publish” (I4). Or, as concluded by another respondent: “The measurement criteria favor this type of behavior; salami slicing or recycling” (I13), where you “exploit the empirical material to the extreme” (I10).

Despite such severe critique though, and as indicated already in some of these critical accounts, researchers were still very oriented towards providing “good numbers”. For example, it was admitted that nowadays it is very easy “not to think about the content [of the article] anymore”, but rather to think about “how do we get this published; can we do something on this?” (I4) or “how much do I have to include in this paper to get it published?” (I16). Then, “[you] slice up the material. […] because then you win some, like you get another publication […] I do that myself” (I4), because “it is better to publish something that isn’t that good, than not publishing” (I14), “especially when your publication list is short, then it’s like you publish whatever and in whatever journal, just to make the list [of publications] a bit longer” (I2). Or, in relation to the issue of how to improve the number of citations: “you can always manipulate the data, by making self-citations […]. Or you can publish something just to say something controversial so as to force them to cite you” (I5). As suggested by several respondents, this becomes particularly pertinent when “you’re being assessed all the time [… because then] you don’t try to pursue a common goal. Instead, we compete with one another when it comes to [research] money and everything” (I19). It becomes a form of competition where it easily becomes a question of highlighting yourself (I9), “because all individuals want to be recognised when they have done something well” (I12).

Again, our analyses suggest that this numbers-orientation was not an effect of researchers trying to obtain a particular (external) reward – such as a publication bonus or a salary increase. In fact, we found no clear examples where the respondents seemed directly motivated by, or tried to “chase”, such external incentives when working on, for example, a particular publication or an application for research funding. Instead, our analyses suggest their willing compliance should be understood as a form of coping strategy that allowed them to satisfy their needs to feel competent, related and autonomous. The premise is that when the researchers were able to provide numbers that met or exceeded the expectations set by the control systems, they typically received validation of their competency, approval of their place in the academic community and increased privileges in the form of research autonomy, all of which contributed towards a perceived sense of need satisfaction within the ruling conditional logic.

In the empirical material, we could see this in terms of how the numbers allowed them to secure this type of need satisfaction. For example, it was referred to how “it becomes an acknowledgement, like when you hear a ‘ding’ [i.e. when you receive a signal that someone has cited your work] you know that you exist, that someone has seen your work and thought that ‘this is important’, ‘let’s cite this’, ‘this is truly relevant’” (I10). In a form of competence- and relatedness-boosting way, the numbers thus became “a form of recognition […] and you get really happy, it’s super exciting” (I13), “it’s a home run, like, someone thinks I am worth financing [or publishing] […] you get so happy” (I18), “when you live up to the expectation, like, ‘I am good enough for this’” (I20). Or, as suggested by yet other respondents: “of course you want to be portrayed as serious and reputational, as someone whose work counts [when seen through metrics] like quotation-rates and impact factors” (I5), so “if the Head of department sees this [i.e. rising numbers], you’re lucky” (I20) because “it gives you status” in relation to others (I3). Moreover, in an autonomy-boosting way, they knew that this type of recognition and status would increase their autonomy, because “if you prove yourself, then you get the freedom to do more” (I19).

The second aspect, suggesting that the conditional logic underlying the results control systems had made researchers intrinsically motivated, relates to the strong feelings of inner pressure and tension expressed by our respondents. In the empirical material, this could be traced to the fact that once you have realised and accepted that the main way to secure the type of need satisfaction referred to above is through adopting a strong numbers-orientation and “prove yourself” by providing good numbers, you essentially give up certain aspects of yourself. That is, you secure certain aspects of your basic psychological needs at the cost of other aspects of these very needs. As a result, we could see how the coping strategy that was essentially intended to secure need satisfaction was also inevitably marked by tensions, insecurity and doubt. Or, as summarised by one of the respondents: when everyone knows that “it’s all about putting your best foot forward and have as many publications as possible [then] you end up in a competitive situation where you start to worry about how good others are compared to you” (I1).

When it comes to the need for competency, we could see how the conditional logic led to such worrying, because of how it provided competence support first and foremost when researchers performed well according to the control system. The premise is that when individuals are approved and validated as good researchers only when they perform in certain ways, they may feel that their worth is tied solely to their achievements. This can, in turn, be expected to provoke various negative feelings, including doubts about whether they are good enough, feelings of constant pressure to “prove themselves”, and to stress and frustration when failing to do so. For example, this was the case when performance feedback made it obvious to an individual that “I don’t have a lot of publications at all” (I10) or that “I have to apply for more funding and write more articles, because I’m really bad at that” (I2), then it made “you feel small […] because, you are always competing, competing in everything, and, when someone else is able to boost themselves, then it’s like, […] you feel small” (I15). Or as suggested by another respondent, “it’s when you know that you haven’t been good at producing, when you haven’t been a machine at publishing […], then it is stressful” (I2) because, “then we’re at the inner core, at the self […] and then you doubt yourself, like ‘am I good enough?’, because I don’t have that many publications” (I10). As suggested by several respondents, this type of self-doubt was driven by the conditional logic underlying the control systems and the inner stress and tension that this tended to engender. Or in the words of our respondents: when “you’re evaluated and exposed to competition all the time” (I20) and when “there aren’t any real rest periods during the year anymore” (I17), then it can be quite “stressful to be around [high performers] because they’re breathing down my neck” (I15). Or as vividly explained by another respondent: it becomes like a “honey trap, and when you look too deep in the jar, when you push yourself too hard, then it’s not unusual to get burn-out syndromes in academia” (I1).

In a similar manner, we could see that when it came to the need for relatedness, such a numbers-orientation was associated with researchers feeling that they had to earn their space in relation to others, which, in turn, typically fostered feelings of pressure and tension among them. The premise is that when you have a basic psychological need to feel related to a larger research community, the sheer risk of providing poor numbers suddenly becomes a threat to this. Or, as expressed by one of the respondents: “We’re only humans […] we want to be appreciated by the university, by the head of school, and like […] […] we want to be considered a good subject here at the university, just like everyone else” (I2). However, to do so, they knew that they had to be “visible on the arena” (I1), and because “they [i.e. university management] are only interested in the CVs and the list of publications” (I5) they felt that they had to chase external funding for research, gather citations and make sure they were “the one who publishes” (I9).

Again, as long as they could do so, they testified to how they were typically recognised and appreciated more which, in a sense, helped to satisfy their need for relatedness. As exemplified by one of the respondents who had just recently secured an external grant for research: “It felt good, everyone gave me a pat on the back and told me how good I was” (I12). However, in cases where individual researchers were unable to prove themselves by means of numbers, the inner pressure and stress clearly shined through: “You can’t put forward an article every third year and say, ‘now I’m published’” (I2), because then they “would start to wonder: ‘what is that person doing?’” (I19). Or, as expressed by yet another respondent: “Still, if I just sat for four years and […] nothing came out, if I just had a lot of high-risk projects and didn’t receive any research grants, then I think they would have a discussion with me about what I do as a researcher and how I contribute to the research environment” (I9).

In a similar manner, it was obvious that this inner pressure and tension affected their views on, and relationships to, their peers: deep down, you know that “it is important to be happy when things go well for someone”, but due to the competitive climate, you become “quite bad at confirming each other” (I14), and “you get like… cynical” (I15). From such a perspective, we could see how at least some researchers questioned other people’s motives, such as where they were seen as “alpha males” who become “all narcissistic: ‘look at me, look at what I’ve done’” (I5), and why other researchers were recognised but not them: “There are some [researchers], where you wonder: ‘what are they doing during the days?’ […] I mean, I have to work really hard to reach the next level and then you see someone already on that level who underperforms according to me, and then you […] it’s not good, but you start to see that person, thinking ‘well, you never do anything, you just talk and talk’. So, you easily find yourself holding a grudge against people” (I15).

Finally, and not least, we could see how the conditional logic also led to such feelings of tension and pressure related to the need for autonomy. Arguably, this is what can be expected because a conditional logic inevitably means that people are pressured to perform in certain ways which, in turn, can easily evoke feelings of autonomy erosion and inauthenticity. The premise is that when you feel that you have to behave in certain ways or achieve certain things to help satisfy your needs, this can lead to feelings of external control rather than self-directed choice or to feelings of being torn between pleasing others and following your own path. As a result, your sense of autonomy or authenticity risks being diminished because you feel compelled to prioritise others’ expectations over your own desires and interests.

In the empirical material, such tensions were evident in different ways. For example, several respondents referred to how the performance measures “narrow down what one can do” (I2), that “you are never able to choose freely” (I10), because “you are always constrained” (I5), always forced “to adapt to what is seen as plausible” (I4). One reason for this was that the performance criteria per se were seen as largely given and, as such, they tended to constrain the room for manoeuvring: “it’s tough, because it’s [i.e. the criteria] not something that we’re in control of” (I8). Or, as pinpointed by yet another respondent, in relation to the conditional logic: “you have a lot of degrees of freedom, but you always think that ‘it has to result in something’. So, it’s like you’re free to do what you want, but it has to result in something very specific” (I15).

To summarise, we find that many of the researchers under study had internalised the expectations underlying the control system and, as a consequence, tried to achieve what the system wanted from them. However, we also find that they had not necessarily done so because they identified with the importance of, for example, producing a large number of articles, accumulating evermore citations or attracting more and more external funds. Instead, they seemed to have internalised the expectations as a way of securing at least some aspects of their basic psychological needs. However, as this inevitably meant that other aspects of their needs were frustrated, it resulted in feelings of internal pressure and tension. That is, because they did not necessarily identify with their own behaviour, as it were, they experienced it as an “internalised must” or something that they had to do so as to earn their competency, relatedness or autonomy.

In this paper, we have taken an interest in how results control affects the motivation of individuals through supporting and thwarting their basic psychological needs. However, rather than focusing on when this leads to full internalisation, which has been the main focus of the existing literature, we concentrated on a situation in which results control supported and thwarted the psychological needs in such a way that the expectations of the control systems were only partially internalised, and hence, resulted in introjected motivation. Two emerging research questions guided our empirical and analytical work: What qualities of the results control systems helped support and thwart the need for researchers to feel competent, related and autonomous? and How and why – if at all – did such qualities lead to only partial internalisation and, thus, to introjected motivation? Below, we highlight our key conclusions in relation to these research questions, outline how the conclusions contribute to our understanding of results control and (introjected) motivation, and suggest avenues for future research based on study limitations.

Starting out with the first research question, it is important to remember that in cases where results control has resulted in introjected motivation, it must mean that people have internalised the expectations underlying the control systems. That is, they must have taken in and come to follow the expectations out of inner reasons, even in situations where the expectations are no longer upheld or reinforced by any external force (because the opposite would mean that they were, in fact, externally regulated). This, in turn, must also mean that people have experienced the results of control systems as at least partially supportive of their needs to feel competent, related and autonomous, as such need support is what drives internalisation in the first place (e.g. Ryan and Deci, 2000b, 2017). At the same time, though, it cannot be that people have experienced the results control system as fully supportive of their needs, as this would result in full internalisation and autonomous motivation. Instead, SDT suggests that for introjected motivation to arise, at least the need for autonomy must be thwarted in some sense.

As detailed above, our empirical findings largely corroborate these theoretical arguments. For example, it was clear that results control supported the need for competence through providing informational feedback on how researchers performed in relation to targets and in relation to other researchers. It also supported the need for autonomy through allowing a certain room for agency when it came to how, when and why to achieve the performance targets set up by the control systems. Importantly, however, it was also evident that results control thwarted other aspects of these very needs. For example, this was the case with their need for autonomy, as researchers felt “forced” into certain types of behaviour when it came to conducting research and publishing the results thereof.

To account for how results control can support and thwart the basic psychological needs at the same time in this way, the current study suggests the notion of a conditional logic. This notion emphasises the following. To begin with, it points to how a results control system can clearly tie in with, and support, the needs of individuals to feel competent, related and autonomous through recognising and rewarding those who meet or beat its expectations. In fact, and as detailed above, such a promise of need support when performing well according to the system, was offered to each and everyone, regardless of how they had performed in the past. Importantly, though, regardless of whether researchers were able to enjoy such support or not, our findings suggest this type of contingent approval is destined to result in less than full need satisfaction. We found two interrelated reasons for this. Firstly, we saw how the conditioning as such was perceived as autonomy frustrating because of how it forced the researchers in certain directions. Secondly, and related, we saw how such feelings of autonomy frustration were reinforced by the fact that many researchers were highly critical of the type of direction pointed out by the control system. Based on this, a first key conclusion is that a conditional logic constitutes an important quality for understanding how results control systems can support and thwart the basic psychological needs of individuals at the same time.

Arguably, the conditional logic also forms an important basis for understanding the second research question, namely, how and why results control systems can lead to partial internalisation and introjected motivation. The premise is that because the need for autonomy will always be frustrated to some extent by a conditional logic, even for those individuals who are “rewarded” for living up to its expectations, individuals cannot be expected to endorse and fully internalise such a logic. On the contrary, they can only be expected to partially internalise the logic and, thus, develop introjected motivation. Importantly, however, our findings strongly suggest that this should not be taken to mean that introjected motivation is caused, or determined, by the conditional logic underlying the control system. On the contrary, and as emphasised above, our findings illustrate how the relationship between results control and introjected motivation worked through a form of coping strategy, where researchers willingly complied with the expectations of the control systems as a way of satisfying their own needs to feel competent, related and autonomous, despite being highly critical of the control systems as such. That is, even if it meant that they had to give up some aspects of their needs, they still did so, as a means to secure other aspects of these very needs.

An important reason for this, we suggest, is that even in situations where individuals feel that the embracement of external expectations requires some sort of compliance on their behalf, they may justify or rationalise such compliance when it helps to secure a sense of need satisfaction. For example, when they face external expectations that are experienced as somewhat alien to themselves, they may still align themselves with, and internalise, those expectations if they feel that in doing so, they may secure other aspects of their needs, such as a sense of self-worth and mastery (i.e. competence). For the same reasons, we suggest that individuals may come to internalise the external expectations when they perceive that in doing so, they are able to strengthen their connection to significant others (relatedness) or extend their own room for manoeuvre (autonomy).

In the empirical study, we saw how this type of self-endorsed or willing compliance made researchers want what the control system wanted from them, because of how it allowed them to satisfy at least parts of their needs. However, we could also see how it nurtured feelings of inner pressure and tension, because of the adaptations and compromises that it required from them. Based on this, our second key conclusion is that when exposed to a conditional logic, people may develop “willing compliance” as a way of coping with the logic. However, while such a coping strategy can clearly help satisfy parts of the psychological needs of individuals, it will always be marked by feelings of inner tensions and cognitive conflicts as other aspects of the needs have to be sacrificed. As a result, only partial internalisation will take place, thereby fostering introjected motivation. To understand if – and if so, how and why – a conditional logic results in partial internalisation and introjected motivation, it is thus important to consider the processes through which individuals handle such conflicting experiences of results control systems.

Arguably, the findings provided above are interesting and important in their own right, not least because of their potential practical implications. More specifically, they provide several important insights into how (not) to design and use results control in university settings. Even though a conditional logic may seem rather subtle and harmless when designing and using a results control system, our findings show how it can constitute a clear threat to more autonomous and intrinsic types of motivation. In fact, when using performance metrics and numbers to not only decide on tangible issues such as salaries, rewards and promotions but also on who is to be recognised and socially approved, individuals can come to develop a sense of worth that is not really tied to who they are as individuals but to what they perform. Moreover, and as suggested by our respondents, there is a large risk involved in tying such a notion of worth to the meeting or beating of externally defined, and rather narrow, expectations. The premise is that a narrow focus on quantifiable outputs such as the number of publications, citations and research grants won will inevitably incentivise rather narrow and short-term performances at the cost of more exploratory or risky research. From a motivational perspective, this also risks leading to that researchers are drawn from their original motivations for entering academia or that they become amotivated altogether. To avoid this, our findings suggest universities should neither be governed by means of a narrow focus on very specific performance metrics nor by means of a conditioning of other types of outcomes on how researchers perform in relation to such metrics.

Apart from this, and as discussed below, our findings also offer important contributions to the existing MC literature on results control and motivation and to the SDT literature on controlling environments and introjected motivation.

Our first theoretical contribution relates to the qualities of results control systems that make people internalise their underlying expectations. This is an area of research that has already received some attention in the existing SDT-based MC literature. As discussed in more detail above, this literature has suggested and found that those qualities of results control systems that support the basic psychological needs of individuals will enable internalisation while those qualities that thwart the needs will hinder internalisation (see e.g. De Baerdemaeker and Bruggeman, 2015; Groen et al., 2017; Pfister and Lukka, 2019; Van der Hauwaert et al., 2022). Notwithstanding the many important insights made regarding these types of qualities, though, they almost exclusively focus on what it is about results control that leads, or does not lead, to full internalisation. An important reason for this is arguably that within the MC community, it has been considered an especially important research task to better understand how control systems can contribute towards autonomous motivation and the expected positive effects on individuals’ work performance and well-being thereof (Ryan and Deci, 2017).

Adding to this, the current study draws attention to a different type of quality that can also contribute to internalisation of a results control system and its underlying expectations, namely a conditional logic. Again, this type of quality refers to the ways in which the results control systems defined the performances of researchers in a particular way and then conditioned the type of need support they would get depending on how they performed in relation to this definition. In the current case, we could see how this type of quality clearly contributed towards internalisation among the researchers because of the need support that it offered. Importantly, though, and in contrast to the type of qualities identified in the existing literature, the conditional logic led to the type of partial internalisation that characterises introjected motivation. The reason, we suggest, is that although the conditional logic offered needs support, it did so in a manner that concurrently led to need thwarting, not least because of how it emphasised competence and relatedness at the cost of autonomy. In fact, it de-emphasised the types of feelings often associated with full internalisation and autonomous motivation, such as personal relevance and congruence. Instead, it emphasised feelings of self-awareness and critical reflection among individuals. That is, the control systems made them attentive to how they were perceived in the eyes of others, and that they had to perform in certain ways to sustain their self-worth and earn approval from themselves and others. In short, it encouraged the development of an inner motivation grounded in feelings of having to act in particular ways so as “to avoid guilt or anxiety or to attain ego-enhancements or pride” (Ryan and Deci, 2000a, p. 62).

Our second, and related, contribution to the MC literature concerns the very mechanism whereby the expectations expressed by results control systems are internalised and, thus, result in certain types of motivation. In the existing literature, this is a question that has received rather scant focus. In fact, the whole internalisation process that is assumed to follow from the needs being satisfied by certain qualities of the control system has been left rather unproblematised. Instead, MC scholars have often drawn upon SDT in rather general terms to suggest that the satisfaction of the three needs as such “will encourage internalisation” (De Baerdemaeker and Bruggeman, 2015, p. 4) or that “internalization will be fostered when people’s three basic psychological needs […] are supported” (Bouten and Hoozée, 2022, p. 1867; see also Sutton and Brown, 2016; Pfister and Lukka, 2019). Moreover, and perhaps as an effect of this, many MC scholars have treated the satisfaction of the basic psychological needs as a pure theoretical mechanism – i.e. as a mechanism that explains the expected relationship(s) between a control system and different types of motivation – without really studying them, or the ensuing internalisation process, empirically (but see e.g. Pfister and Lukka, 2019; Van der Hauwaert et al., 2022, for exceptions).

Largely in contrast to this, our findings point to the importance of focusing on how individuals experience and handle the need support offered by the control system when trying to understand the types of motivation they develop. In fact, while the conditional logic was important in the current case, it did not automatically lead to a certain type of internalisation or to a certain type of motivation. It was not deterministic in the sense that it forced the researchers into a certain type of motivation. On the contrary, it offered need support, where the very point of the logic was to leave it open to the individual researcher how to handle this offer. In fact, through doing so, it created a form of manoeuvring space for the individual researcher, not only in the sense that they were able to choose the type of research that they wanted to conduct, how to conduct it, and with whom, but also in the sense that it became possible for them to have an impact on their own need satisfaction. It made it clear to them that they were in control of their own need satisfaction, as long as they managed to publish in well-renowned journals, accumulate citations from other researchers and bring in external funding for research. However, apart from creating such an appealing manoeuvring space, the conditional logic also responsibilised researchers for moving in certain ways within this space. That is, it encouraged them to internalise the responsibility for aligning their behaviour with the performance targets set up by the control systems, through tying these targets to the basic psychological needs of the individual. As a result, it made sure that it was in the interest of the individual researcher to “want what the system wanted from them”, because that was what it would take for the researchers to secure their own need satisfaction.

When considering that most of the researchers were highly critical of the control system in principle, the conditional logic thereby placed them in a situation where they had to balance between their critical views on the control system and the things they knew that they “needed to do” to satisfy their own needs. Again, in the current case, we saw how this resulted in a form of willing compliance where the researchers became highly numbers-oriented in their strive towards enhancing their self-worth and improving their relative position in their research community. A form of coping strategy where they willingly aligned their attitudes and efforts with the expectations placed upon them, even without direct oversight, at the same time as they struggled with feelings of doubt and inner tensions due to how the strategy meant that they had to give up some aspects of their needs.

Judging from the empirical material, it seemed as if an important reason for why the conditional logic resulted in introjected motivation in the current case – where the researchers’ efforts to secure need satisfaction seemed to prevail over those aspects of their needs that they gave up – was the potential for success that the control system offered them. That is, despite a very high awareness of the difficulties associated with, for example, publishing in well-renowned journals or attracting research funding from particular high-esteem funders, the sheer possibility for everyone to do so and the fact that many of them were also able to progress in their academic career through doing so, seemed to outweigh the critique that they had against the control system. Arguably, though, this does not always have to be the case. On the contrary, it may very well be that a conditional logic makes it clear to individuals that they do not have what it takes to succeed within such a system or that they feel that they are not really prepared to comply with the expectations as such. When that is the case, previous SDT-based research has shown that people may not develop introjected motivation but may become amotivated altogether (Ryan and Deci, 2000a; Ryan and Deci, 2017; Vansteenkiste et al., 2023). The premise is that in such cases, it does not really matter that the conditional logic offers potential need support as people will feel that what is being offered is not acceptable as such or within practical reach.

When taken together, these findings add further nuance to the existing literature on the relationship between qualities of control systems and different types of motivation, as they point to the importance of understanding how individuals turn the need support offered by a control system into actual experiences of need satisfaction and certain types of motivation. In fact, in the current case the “effects” of the results control system qualities seemed to constitute much more of an active accomplishment than has been stressed in both the existing MC literature and in much of the more general SDT literature (see subsection 5.2.2 for a further discussion of this); effects that were produced and reproduced as individual researchers balanced between their inherent desire to satisfy some aspects of their basic psychological needs at the cost of other aspects of their needs.

Arguably, the findings from the current study also contribute to the more general understanding of what makes people develop introjected motivation. Again, the SDT literature has argued and provided rather compelling evidence that controlling environments – i.e. environments where people are pressured into thinking or behaving in particular ways – make people develop controlling types of motivation because of the need-thwarting characteristics of such environments (Assor et al., 2004; Bartholomew et al., 2010; Ryan and Deci, 2017). However, when it comes to the controlled type of motivation focused on in this paper – i.e. introjected motivation – this literature also stresses that the controlling environments cannot be altogether need-thwarting for introjection to occur. The reason being that for this type of motivation to arise, some sort of need support is required. That is, to internalise the “introjects”, people need to experience that at least some of their needs are being satisfied. This is also the reason that introjected motivation has tended to be explained as a result of a particular type of controlling environment, namely when it is characterised by various forms of conditioning (Bartholomew et al., 2010; Haines and Schutte, 2023; Niemiec et al., 2008, 2010). The premise is that although such conditioning always means that people will have to give up some of their autonomy, there is something to be gained in the form of relatedness or competence support (Assor et al., 2004; Bartholomew et al., 2010; Haines and Schutte, 2023).

Arguably, our findings add to this literature in two different ways. Firstly, the notion of a conditional logic underlying the control system suggests that the type of conditioning associated with introjected motivation may not only be about the giving up of autonomy to secure relatedness or competence, as stressed in the SDT literature. On the contrary, it can be about the giving up of some aspects of one’s needs to gain or maintain other aspects of these very needs more generally. An example of this was when researchers willingly complied with the requirements of the control system despite being highly critical of these very requirements. Importantly, though, and again, through doing so, they were not only able to increase their chances of getting more resources and time for research – i.e. complying with the external demands so as to be able to keep doing what they wanted to do – but they were also able to secure a sense of freedom from the interference of their superiors. That is, through proving that they could meet or beat the expectations communicated by the control system, they experienced that university management was prepared to grant them a form of trust; “if you’re good, we won’t interfere that much” (I19). Hence, in the current case, complying with what the control system wanted from the researchers was not only about them foregoing “autonomy in order to get the approval and affection of the socializing agent” (Niemiec et al., 2010, p. 180), but also about foregoing some aspects of their autonomy to secure a broader sense of autonomy. Put differently, they gave up some of their autonomy to create a form of psychological space where they were able to follow their inner drive to do research and where there was freedom from hierarchical surveillance.

Arguably, how willing compliance can be used as a strategy to maintain this broader sense of autonomy is a less well-explored area in the SDT literature. One reason for this may be that while many researchers have conducted studies on controlling environments and introjected motivation within schools and families, studies of employees within work contexts are considerably rarer (for similar arguments, see Gonzales et al., 2024; Sarmah et al., 2022; Slemp et al., 2018). Arguably, though, in work organisations, it seems reasonable that the type of coping strategy found in the current study – i.e. where researchers gave up some of their autonomy to maintain other aspects of their autonomy – could be as common as them trying to secure approval or a positive regard from others just to uphold certain social relationships at work. At least, we suggest, this could be the case when it comes to the relationships that employees have to their superiors, as such relationships are not necessarily oriented towards upholding feelings of relatedness but could equally well function as a means to secure one’s freedom from interference and control; “if I make sure to get along with my superiors and show them that I perform well, it could increase the chances of being left alone”.

Secondly, and related, our finding that the relationship between a control system and the development of a particular type of motivation needs to be understood as a highly non-deterministic one adds to previous SDT-studies of how conditional need support may lead to introjected motivation. The premise is that most of this previous research has had a cross-sectional and correlational focus (e.g. Assor et al., 2004; Haines and Schutte, 2023; Otterpohl et al., 2021; Vansteenkiste and Ryan, 2013). That is, much of this research has – often because of the ethical and methodological challenges of studying controlling environments and introjected motivation – relied upon a research design where individuals are asked to fill in one or more questionnaires at a particular point in time. As a consequence, this makes it hard to reveal the micro-dynamics that go on as people experience their environments in certain ways and develop certain types of motivation. Again, as argued and shown in the current study, such dynamics typically involve large degrees of human agency and intentional considerations rather than being pre-determined. In fact, and as detailed above, our findings show that although the conditional logic was important in the current case, it was not deterministic in the sense that it automatically led to a certain type of internalisation or to a certain type of motivation. On the contrary, our findings show how the experienced conditioning of their need support was handled through a form of coping strategy where researchers balanced between their critical views on the control system and the things they knew that they “needed to do” to satisfy their own needs. A coping strategy that, again, not only meant that their competence and relatedness needs were satisfied at the cost of some aspects of their autonomy, but also that they were able to secure other aspects of their autonomy.

Based on the fact that the findings in the current paper are founded on an empirical study of a limited number of researchers at three Swedish universities, we see several important avenues for future research. A general, and perhaps the most urgent, avenue is that more research is needed on when and why results control (and other control systems) tend to result in introjected motivation. The premise is that although some tentative answers to this question have been proposed in the current paper, this is, to our knowledge, one of the first attempts to explore this rather broad and complex question more systematically. Based on this, several issues seem worthy of further attention. One issue relates to the empirical context. Again, the current study was conducted in an academic setting, which could be important for understanding how and why our respondents perceived, and were affected by, the results control systems in particular ways. To further explore the importance of context in this sense, future studies could fruitfully study other types of empirical settings and other types of work tasks to see whether these affect the tendency towards only partial internalisation and, as a result, introjected motivation. Another issue relates to the type of qualities of results control that can contribute towards introjected motivation. In the current study, we identify one such important quality in the form of a conditional logic, which was found important for why researchers internalised the control systems in certain ways. Arguably, though, there may be other qualities of results control systems that also tend to result in partial internalisation. Another important avenue for future research is thus to explore other qualities of results control systems than the one addressed in the current paper. Finally, this study points to the importance of acknowledging how individuals experience and actively seek to cope with the need for support offered by control systems. In our study, the numbers-oriented coping strategy referred to as “willing compliance” was deemed important for understanding how and why results controls came to foster introjected motivation. Based on this, we suggest that an important avenue for future research is to examine the prevalence and characteristics of other coping strategies to better understand the motivational effects of different types of controls.

The authors gratefully acknowledge the useful support of the editor Thomas Ahrens, two anonymous reviewers, and participants at the 28th Nordic Workshop in Management Accounting (NWES XXVIII) in Gothenburg, Sweden.

Introduction to the interview – background and demographic questions

• Could you briefly describe who you are, for how long you have been working here at X, and what your main duties are?

Theme 1: Research- and researcher-oriented questions

• What does your research situation look like?

  • How much time do you devote to research?

• What does it mean to you to be a researcher?

  • What is important when it comes to research, as you see it?

• What does it mean to be a successful researcher?

• How successful do you feel as a researcher?

  • Is it important to feel successful?

Theme 2: Results control – Targets, performance measures and incentives

• Could you say something about the kind of tools that are used to follow up on, and evaluate, research here at your work unit and at the university as such?

  • Where/how are good research performances made visible?

  • Are bibliometric measurements used?

  • Are qualitative evaluations used?

  • Who makes the evaluations?

  • How often are they made?

  • What are the main units of evaluation (individuals, groups, research environments and departments)?

• How are the performance evaluations used?

  • Who is the main user of the evaluations?

  • How are your performances communicated to you?

  • What do you think is the main ambition underlying the performance evaluations (what do you think that the evaluators want to achieve)?

  • What kind of researcher would you say the PMSs encourage or reward?

• What are the pros and cons of this type of evaluation as you see it?

• What kind of research is conducted at your department/in your field of research?

  • What are typical types of research outlets in your field?

  • Has this changed over time?

• What does the market for research funding look like for your type of research?

  • What do you have to do to be able to compete for such money?

• If you get funding for research, what type of expectations does this lead to?

  • What happens if you do not live up to those expectations?

• Have you received external funding when applying?

  • How did that make you feel?

  • Was this acknowledged in any way, at the department or at the university?

  • What is your view on such an acknowledgement?

• If a researcher applies for funding and does not succeed, how is that person perceived by others at the university?

  • How are you perceived if you are successful and bring in a lot of external funding?

  • What happens if you give up applying for funding altogether?

• What is your view on the fact that quite a large share of funding for research is allocated based on competition between researchers in Sweden?

• Do you worry about whether you publish enough or attract enough funding for research?

• Do you often think about how your colleagues perform when it comes to research funding, publications, etc.?

Theme 3: Effects of targets, performance measures and incentives on research(ers)

• Would you say that the type of control that we have discussed now affects the kind of research that you conduct? How and why?

• What kind of researcher do you think one becomes when exposed to this type of control?

• How does it (if at all) affect your work situation/environment?

  • Your work motivation?

  • Your stress level?

  • Competition among colleagues/groups?

  • The degree of individualisation?

• Does it affect the meaning of an ideal researcher/university? How and why?