Journal of Organizational Behavior J. Organiz. Behav. 26, 331���362 (2005) Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/job.322 Self-determination theory and work motivation MARYLENE ` GAGNE1* �� AND EDWARD L. DECI2 1Department of Management, John Molson School of Business, Concordia University, Montreal, Quebec, Canada 2Department of Clinical and Social Sciences in Psychology, University of Rochester, Rochester, New York, U.S.A. Summary Cognitive evaluation theory, which explains the effects of extrinsic motivators on intrinsic motivation, received some initial attention in the organizational literature. However, the sim- ple dichotomy between intrinsic and extrinsic motivation made the theory difficult to apply to work settings. Differentiating extrinsic motivation into types that differ in their degree of autonomy led to self-determination theory, which has received widespread attention in the education, health care, and sport domains. This article describes self-determination theory as a theory of work motivation and shows its relevance to theories of organizational behavior. Copyright # 2005 John Wiley & Sons, Ltd. Introduction Building on Vroom���s (1964) expectancy���valence theory of motivation, Porter and Lawler (1968) pro- posed a model of intrinsic and extrinsic work motivation. Intrinsic motivation involves people doing an activity because they find it interesting and derive spontaneous satisfaction from the activity itself. Extrinsic motivation, in contrast, requires an instrumentality between the activity and some separable consequences such as tangible or verbal rewards, so satisfaction comes not from the activity itself but rather from the extrinsic consequences to which the activity leads. Porter and Lawler (1968) advocated structuring the work environment so that effective performance would lead to both intrinsic and extrinsic rewards, which would in turn produce total job satisfaction. This was to be accomplished by enlarging jobs to make them more interesting, and thus more intrin- sically rewarding, and by making extrinsic rewards such as higher pay and promotions clearly contin- gent upon effective performance. Implicit in this model is the assumption that intrinsic and extrinsic rewards are additive, yielding total job satisfaction. Porter and Lawler���s model, Vroom���s theory, and other expectancy���valence formulations generated considerable research, much of which confirmed and refined aspects of the approach (see Mitchell, Received 23 October 2003 Copyright # 2005 John Wiley & Sons, Ltd. Revised 2 June 2004 Accepted 8 January 2005 * Correspondence to: Marylene ` Gagne, �� Department of Management, GM 503-49, John Molson School of Business, Concordia University, 1455 de Maisonneuve W., Montreal, Quebec, Canada, H3G 1M8. E-mail: mgagne@jmsb.concordia.ca

1974). However, one strand of research concerning the additivity of intrinsic and extrinsic motivation was potentially problematic and controversial. Specifically, early studies testing the additivity hypoth- esis found that tangible extrinsic rewards undermined intrinsic motivation whereas verbal rewards enhanced it (Deci, 1971), thus implying that intrinsic and extrinsic motivation can be both positively and negatively interactive rather than additive. Based on several early experiments, cognitive evalua- tion theory (CET Deci, 1975 Deci & Ryan, 1980) was proposed to explain the effects of extrinsic motivators on intrinsic motivation. Cognitive Evaluation Theory Cognitive evaluation theory suggested first that external factors such as tangible rewards, deadlines (Amabile, DeJong, & Lepper, 1976), surveillance (Lepper & Greene, 1975), and evaluations (Smith, 1975) tend to diminish feelings of autonomy, prompt a change in perceived locus of causality (PLOC) from internal to external (deCharms, 1968 Heider, 1958), and undermine intrinsic motivation. In con- trast, some external factors such as providing choice about aspects of task engagement tend to enhance feelings of autonomy, prompt a shift in PLOC from external to internal, and increase intrinsic motiva- tion (Zuckerman et al., 1978). CET further suggested that feelings of competence as well as feelings of autonomy are important for intrinsic motivation. Studies showed that optimally challenging activities were highly intrinsically motivating (e.g., Danner & Lonky, 1981) and that positive feedback (Deci, 1971) facilitated intrinsic motivation by promoting a sense of competence when people felt responsible for their successful per- formance (Fisher, 1978 Ryan, 1982). Further, negative feedback which decreased perceived compe- tence was found to undermine both intrinsic and extrinsic motivation, leaving people amotivated (Deci & Ryan, 1985a). Underlying these CET propositions was the assumption that people need to feel autonomous and competent, so social-contextual factors that promote feelings of autonomy and competence enhance intrinsic motivation, whereas factors that diminish these feelings undermine intrinsic motivation, leav- ing people either controlled by contingencies or amotivated. Spirited debate ensued concerning both the undermining effect and CET (e.g., Calder & Staw, 1975 Deci, 1976 Deci, Cascio, & Krusell, 1975 Scott, 1975), leading to numerous laboratory experiments and field studies intended to support, refine, extend, or refute the undermining effect and CET. Even- tually, a meta-analysis of 128 laboratory experiments confirmed that, whereas positive feedback enhances intrinsic motivation, tangible rewards significantly undermine it (Deci, Koestner, & Ryan, 1999). The Deci et al. (1999) meta-analysis also confirmed CET hypotheses that specified limiting conditions to the undermining effect. Namely, it showed that when rewards were given independent of specific task engagement (as might be the case with a salary) or when the rewards were not anticipated (as might be the case with unexpected bonuses), tangible extrinsic rewards did not undermine intrinsic motivation. Additionally, as found by Ryan, Mims, and Koestner (1983), when rewards were contingent on high- quality performance and the interpersonal context was supportive rather than pressuring, tangible rewards enhanced intrinsic motivation relative to a comparison condition with no rewards and no feed- back. Notably, however, these performance-contingent rewards did lead to lower intrinsic motivation than a control group that got positive feedback comparable to that conveyed by the rewards. Still, the Deci et al. meta-analysis pointed to possible ways to use rewards without having detrimental effects. As noted, the undermining of intrinsic motivation has been controversial from the time it first appeared in the literature (Deci, 1971), and even though the Deci et al. (1999) meta-analysis showed 332 M. GAGNE �� AND E. L. DECI Copyright # 2005 John Wiley & Sons, Ltd. J. Organiz. Behav. 26, 331���362 (2005)

definitively that tangible rewards undermine intrinsic motivation, recent theories of work motivation have still failed to accept the robustness of the findings. For example, Kehr (2004) suggested that rewards would not undermine intrinsic motivation if they did not deactivate implicit motives related to task enjoyment. However, the statement was pure speculation, and no empirical support for the spec- ulation was provided. Furthermore, support for the hypothesis that expected, tangible rewards admi- nistered engagement-contingently or completion-contingently would undermine intrinsic motivation is so strong that, if Kehr���s theoretical speculation were correct, it would mean that these types of rewards must, in fact, deactivate implicit enjoyment motives. Research in line with CET has also shown that contingent, tangible rewards and other extrinsic fac- tors such as competition and evaluations can be detrimental to outcomes such as creativity, cognitive flexibility, and problem solving which have been found to be associated with intrinsic motivation (e.g., Amabile, Goldfarb, & Brackfield, 1990 McGraw, 1978). For example, McGraw and McCullers (1979) found monetary rewards to decrease cognitive flexibility in problem solving, and Erez, Gopher, and Arzi (1990) showed that monetary rewards decreased performance on a complex task with difficult goals. The problems with CET as a theory of work motivation The undermining of intrinsic motivation by extrinsic rewards and the CETaccount of that phenomenon received attention in the organizational literature in the 1970s and early 1980s, leading Ambrose and Kulik (1999) to refer to CET as one of seven traditional theories of motivation in organizations. None- theless, there are several reasons why that attention soon waned. First, most studies that tested CET were laboratory experiments rather than organizational studies. Second, it was difficult to incorporate CET propositions into the prevalent behavioral and expectancy��� valence approaches. Third, and more practically, many activities in work organizations are not intrin- sically interesting and the use of strategies such as participation to enhance intrinsic motivation is not always feasible. Fourth, most people who work have to earn money, so using monetary rewards as a central motivational strategy seems practical and appealing. Fifth, CET seemed to imply that managers and management theorists would have to focus on one or the other���that is, either on promoting intrin- sic motivation through participation and empowerment while minimizing the use of extrinsic factors or, alternatively, on using rewards and other extrinsic contingencies to maximize extrinsic motivation while ignoring the importance of intrinsic motivation. In 1985 Ryan, Connell, and Deci first presented a differentiated analysis of extrinsic motivation using the concepts of internalization, which directly addresses the last of the above critiques of CET and also has implications for some of the others. Internalization refers to ���taking in��� a behavioral regulation and the value that underlies it. The Ryan et al. theorizing, which explains how extrinsically motivated behavior can become autonomous, together with research on individual differences in caus- ality orientations (Deci & Ryan, 1985b), led to the formulation of self-determination theory (SDT) (Deci & Ryan, 1985a, 2000 Ryan & Deci, 2000), which incorporated CET but is much broader in scope. In this paper, we present SDT, review the research on which it was based, compare it to other work motivation theories, lay out a research agenda, and discuss its relevance for organizational behavior and management. Self-Determination Theory Central to SDT is the distinction between autonomous motivation and controlled motivation. Autonomy involves acting with a sense of volition and having the experience of choice. In the words SELF-DETERMINATION THEORY AND WORK MOTIVATION 333 Copyright # 2005 John Wiley & Sons, Ltd. J. Organiz. Behav. 26, 331���362 (2005)

of philosophers such as Dworkin (1988), autonomy means endorsing one���s actions at the highest level of reflection. Intrinsic motivation is an example of autonomous motivation. When people engage an activity because they find it interesting, they are doing the activity wholly volitionally (e.g., I work because it is fun). In contrast, being controlled involves acting with a sense of pressure, a sense of having to engage in the actions. The use of extrinsic rewards in the early experiments was found to induce controlled motivation (e.g., Deci, 1971). SDT postulates that autonomous and controlled moti- vations differ in terms of both their underlying regulatory processes and their accompanying experi- ences, and it further suggests that behaviors can be characterized in terms of the degree to which they are autonomous versus controlled. Autonomous motivation and controlled motivation are both inten- tional, and together they stand in contrast to amotivation, which involves a lack of intention and moti- vation. Extrinsic motivation and the autonomy continuum Intrinsically motivated behavior, which is propelled by people���s interest in the activity itself, is proto- typically autonomous. However, an important aspect of SDT is the proposition that extrinsic motiva- tion can vary in the degree to which it is autonomous versus controlled. Activities that are not interesting (i.e., that are not intrinsically motivating) require extrinsic motivation, so their initial enact- ment depends upon the perception of a contingency between the behavior and a desired consequence such as implicit approval or tangible rewards. Within SDT, when a behavior is so motivated it is said to be externally regulated���that is, initiated and maintained by contingencies external to the person. This is the classic type of extrinsic motivation and is a prototype of controlled motivation. When externally regulated, people act with the intention of obtaining a desired consequence or avoiding an undesired one, so they are energized into action only when the action is instrumental to those ends (e.g., I work when the boss is watching). External regulation is the type of extrinsic motivation that was considered when extrinsic motivation was contrasted with intrinsic motivation. Other types of extrinsic motivation result when a behavioral regulation and the value associated with it have been internalized. Internalization is defined as people taking in values, attitudes, or regulatory structures, such that the external regulation of a behavior is transformed into an internal regulation and thus no longer requires the presence of an external contingency (thus, I work even when the boss is not watching). However, although most theories of internalization view it as a dichotomy���that is, a reg- ulation either is external to the person or has been internalized���SDT posits a controlled-to-autono- mous continuum to describe the degree to which an external regulation has been internalized. The more fully it has been internalized, the more autonomous will be the subsequent, extrinsically moti- vated behavior. According to SDT, internalization is an overarching term that refers to three different processes: introjection, identification, and integration. A regulation that has been taken in by the person but has not been accepted as his or her own is said to be introjected and provides the basis for introjected regulation. With this type of regulation, it is as if the regulation were controlling the person. Examples of introjected regulation include contingent self- esteem, which pressures people to behave in order to feel worthy, and ego involvement, which pres- sures people to behave in order to buttress their fragile egos (deCharms, 1968 Ryan, 1982). Introjected regulation is particularly interesting because the regulation is within the person but is a relatively con- trolled form of internalized extrinsic motivation (e.g., I work because it makes me feel like a worthy person). Being autonomously extrinsically motivated requires that people identify with the value of a beha- vior for their own self-selected goals. With identified regulation, people feel greater freedom and voli- tion because the behavior is more congruent with their personal goals and identities. They perceive the 334 M. GAGNE �� AND E. L. DECI Copyright # 2005 John Wiley & Sons, Ltd. J. Organiz. Behav. 26, 331���362 (2005)

cause of their behavior to have an internal PLOC���that is, to reflect an aspect of themselves. If nurses strongly value their patients��� comfort and health and understand the importance of doing their share of the unpleasant tasks for the patients��� well-being, the nurses would feel relatively autonomous while performing such tasks (e.g., bathing patients), even though the activities are not intrinsically interesting. The fullest type of internalization, which allows extrinsic motivation to be truly autonomous or voli- tional, involves the integration of an identification with other aspects of oneself���that is, with other identifications, interests, and values. With integrated regulation, people have a full sense that the beha- vior is an integral part of who they are, that it emanates from their sense of self and is thus self-deter- mined. If integrated, the nurses would not only identify with the importance of the activities for maintaining their patients��� comfort and health, but regulation of the activities would be integrated with other aspects of their jobs and lives. Thus, the profession of nurse would be more central to their iden- tity, they would be more likely to act in ways that are consistent with caring for people more generally, and they could come to appreciate the importance of doing uninteresting activities. Integrated regulation is theorized to represent the most developmentally advanced form of extrinsic motivation, and it shares some qualities with the other type of autonomous motivation, namely, intrin- sic motivation. Integrated regulation does not, however, become intrinsic motivation but is still con- sidered extrinsic motivation (albeit an autonomous form of it) because the motivation is characterized not by the person being interested in the activity but rather by the activity being instrumentally impor- tant for personal goals. In short, intrinsic motivation and integrated extrinsic motivation are the two different types of autonomous motivation (with identified extrinsic motivation being relatively autonomous). It is important to note that the SDT model of internalization is not a stage theory and does not sug- gest that people must invariantly move through these ���stages��� with respect to particular behaviors. Rather, the theory describes these types of regulation in order to index the extent to which people have integrated the regulation of a behavior or class of behaviors. As such, SDT proposes that, under opti- mal conditions, people can, at any time, fully integrate a new regulation, or can integrate an existing regulation that had been only partially internalized. To summarize, SDT posits a self-determination continuum (see Figure 1). It ranges from amotiva- tion, which is wholly lacking in self-determination, to intrinsic motivation, which is invariantly self- determined. Between amotivation and intrinsic motivation, along this descriptive continuum, are the four types of extrinsic motivation, with external being the most controlled (and thus the least self- determined) type of extrinsic motivation, and introjected, identified, and integrated being progressively more self-determined. Assessing intrinsic and extrinsic motivation Intrinsic motivation and each type of extrinsic motivation are reflected in different reasons for behav- ing, and these reasons provide a means for assessing the types of motivation (Ryan & Connell, 1989). The Ryan and Connell approach has spawned a family of questionnaires that involve asking partici- pants why they would do particular behaviors that are relevant to the situation being researched. Then, participants are presented with various reasons for doing the behaviors that reflect intrinsic motivation or one of the types of extrinsic motivation. Participants rate the degree to which each is true for them. Examples of external reasons are doing the behavior to get a raise or so the boss won���t be upset, whereas examples of introjected reasons are behaving to avoid guilt or to feel worthy. Identified and integrated reasons involve behaving because people personally value the behavior and have fully accepted its importance for their self-selected goals and their well-being. Intrinsic motivation involves SELF-DETERMINATION THEORY AND WORK MOTIVATION 335 Copyright # 2005 John Wiley & Sons, Ltd. J. Organiz. Behav. 26, 331���362 (2005)