Content uploaded by Barry Gerhart
Author content
All content in this area was uploaded by Barry Gerhart on Mar 17, 2015
Content may be subject to copyright.
Pay, Intrinsic Motivation, Extrinsic Motivation, Performance, and Creativity in the Workplace:
Revisiting Long-held Beliefs
Barry Gerhart
University of Wisconsin-Madison
Meiyu Fang
National Central University (Taiwan)
March 2015
Annual Review of Organizational Psychology and Organizational Behavior, 2015, Volume 2 (pre-
publication version)
Final version available May 2015 at: http://www.annualreviews.org/journal/orgpsych or from
authors.
2
In their Annual Review paper, Rynes, Gerhart, and Parks (2005, p. 572) observed that
“Given the importance of pay and performance to employers and employees as well as the
potential for well-designed [pay for performance] PFP1 systems to improve performance, one
would think that research examining PFP would be plentiful in psychology. However, this has
not been the case, particularly in recent years.” It does not appear that this situation has
changed since then. For example, Gupta and Shaw (2014, p. 1) stated that “When we look
across the topics that have been the focus of attention in [human resource management] HRM
and organizational behavior research, we find thousands of studies on employee selection,
performance appraisal, and turnover…By contrast, research on employee compensation is
sporadic and sparse” (Gupta & Shaw, 2014, p. 1).
As another example, in the Handbook of Psychology: Industrial and Organizational
Psychology, published in 2013, there were 26 chapters. None were on compensation. There
was a chapter in the Handbook on motivation (Schmidt, Beck, & Gillespie, 2013), a topic that
could include compensation. However, the chapter, like the study of motivation in psychology
generally, focused primarily on psychological mechanisms and individual differences as
determinants of motivation. Compensation was not discussed. Major topics (as indicated by
major chapter headings) were: Overview of goals and goal processes; expectancies, self-
efficacy, and related concepts; affect; individual differences related to the self and personality;
temporal dynamics; multiple goals and decision making. (See also Latham & Pinder, 2005.)
It is clear that compensation is a major policy lever that organizations use to motivate
employee attraction, performance, and retention (e.g., Lawler, 1971, 1981) and that private
sector organizations in competitive markets routinely use pay for performance (PFP), with
3
individual performance typically playing a key role, especially as one moves to higher job levels
(Gerhart & Fang, 2014; Gerhart, Rynes, & Fulmer, 2009; Shaw, 2014). Organizations that are
currently regarded as highly creative/innovative and successful, such as Google, Facebook
and/or that rely heavily on human capital (e.g., consulting firms, in addition to technology
companies) also give a central role to pay, being among the highest paying companies (e.g.,
Robinson, 2014; Truong, 2014). High pay often allows such companies to have not only
rigorous selection standards, but also rigorous performance standards that employees must
meet to ensure continued employment, advancement, and high pay. Thus, the study of
motivation in the workplace seems to lead to the study of compensation/PFP.
What makes PFP such an interesting and important topic is that “when ‘it works,’ it
seems capable of producing spectacularly good results and when it does not work, it can
likewise produce spectacularly bad results” (Gerhart et al., 2009, p. 253). PFP has been
described as a high risk, high return strategy (Gerhart & Fang, 2014; Gerhart, Trevor, & Graham,
1996) and many scholars have made important contributions by documenting what can go
wrong in using PFP (e.g., Kerr, 1975; Kohn, 1993; Lawler, 1971; Milgrom & Roberts, 1992;
Pfeffer, 1998; Roy, 1952; Sanders & Hambrick, 2007), including: excessive risk-taking, excessive
competition within the firm, focusing too little on performance measures (e.g., quality,
customer service, long-term performance) not explicitly included in the PFP plan, and focusing
too much on and/or gaming/manipulating performance measures (e.g., sales, stock returns)
that are included in the plan. However, as noted, PFP is very widely used. Also, theory and
research both suggest that PFP, whatever its risks and however challenging its successful design
and execution, is central to organizational effectiveness (e.g., Gerhart et al., 2009).
4
In the best case, work on pay (especially PFP) and motivation (including psychological
mechanisms and individual differences) would be integrated to better understand both
literatures. Gerhart and Milkovich (1992) suggested that such research was needed to draw
credible causal inferences regarding observed relationships between compensation/PFP
policies and outcomes. Rynes et al. (2005, p. 573) argued that psychological research on pay
“has much to contribute” as a better understanding of the “psychological mechanisms” that
drive employee motivation and behavior can help “explain employee reactions to pay
plans…and also help identify reasons why [PFP] plans do not always work as intended.”
Why haven’t organizational behavior/psychology scholars devoted more attention to
these topics? Rynes et al. (2005) speculated that, in addition to the concerns about PFP
summarized above, it was because pay had come to be so widely viewed as a negative
influence on motivation, primarily due to three theories: Maslow’s need hierarchy theory,
Herzberg’s motivation-hygiene theory, and Deci and Ryan’s cognitive evaluation theory (CET).
We would add a fourth: the early work on creativity (e.g., Amabile, 1983, 1996) and its
generally negative view of PFP. A theme in the CET and creativity literatures was not only that
PFP was often detrimental to intrinsic motivation, but also that even when PFP produced
positive effect on motivation, it was on extrinsic motivation, which, importantly, was seen as
being of lower “quality” motivation than intrinsic motivation in terms of sustainability and/or
ability to generate key positive outcomes such as creativity and well-being.
Thus, perhaps the lack of attention to compensation is not much of a mystery. CET and
creativity, the literatures that have most sought to understand the role of compensation in
motivation both traditionally concluded that it was detrimental. And, there has been no
5
shortage of management scholars who have picked up on this theme and worked to
communicate it to a broad audience of academics and practitioners, with Harvard Business
Review being a (highly visible) outlet of choice (e.g., Amabile, 1998; Frey & Osterloh, 2012;
Kohn, 1993; Pfeffer, 1998). Scholars in other areas (e.g., economics) have also become
interested in the undermining effects of extrinsic rewards (e.g., Frey & Oberholzer-Gee, 1997;
Frey & Jegen, 2001; Kreps, 1997; Kuntz & Pfaff, 2002; Prendergast, 2008)2 and a recent best-
seller on Amazon (Pink, 2009) communicated the undermining message to an even broader
audience, including the claim that “pay-for-performance schemes…usually don't work and often
do harm” and that one of the “deadly flaws” of PFP is that it “can extinguish intrinsic
motivation.”. Thus, the idea that PFP does not work, or is even harmful (e.g., to intrinsic
motivation and creativity), has been widely communicated.
But, in recent years, there appear to have been major changes in how extrinsic rewards
are viewed in the CET and creativity literatures. CET is now described (Ryan & Deci, p. 70) as a
“subtheory within” self-determination theory (SDT). Importantly, SDT, unlike CET, says that
some types of extrinsic (integrated and identified) motivation behave much like intrinsic
motivation, which is to say they are more self-determining and autonomous than recognized
under CET and, thus, can be high quality forms of motivation. Together with intrinsic
motivation, these are labeled as autonomous motivation under SDT. Other forms of extrinsic
motivation (external and introjected) continue to be viewed as lower quality and less
autonomous and are labeled as controlled motivation in SDT. In parallel to changing views on
motivation, views on rewards have also changed to recognize positive aspects, at least in some
SDT work: “when rewards are administered in an autonomy-supportive climate, they are less
6
likely to undermine intrinsic motivation and, in some cases, can enhance intrinsic motivation”
(Gagné & Deci, p. 354). A recent meta-analysis by Cerasoli et al. (2014) should provide further
support for this argument. They found that the intrinsic motivation-performance relationship
was positive not only in the absence of extrinsic incentives (ρ = .27), but also positive (and,
inconsistent with CET), larger when incentives were in place (ρ = .36).3
We have seen a similar shift in the literature on creativity. An influential earlier view
was, consistent with general CET logic, that “a primarily intrinsic motivation to engage in an
activity will enhance creativity, and a primarily extrinsic motivation will undermine it” (Amabile,
1983, p. 366). However, years later, in their Annual Review paper, Hennessey & Amabile, 2010)
explicitly recognize that extrinsic rewards can have positive consequences for creativity, as do
Zhou and Hoever (2014) in their recent Annual Review paper on creativity.4
Perhaps some of this change in thinking is due in part to seeing that both intrinsic and
extrinsic rewards seem to play a major role at “creative” companies (e.g., Google, Facebook).
To us, the increased recognition that extrinsic rewards such as PFP do not necessarily have
negative effects on motivation and creativity and, in fact, can have positive effects, may be an
opportunity and provide a much needed impetus for greater attention in future research to the
linkages between compensation, motivation, performance, and creativity. As one of the few
theories that does focus on the link between PFP and motivation and its underlying
psychological mechanisms, CET and its successor, SDT, could play a major role in this research.
Therefore, we take a close look at the role of PFP in the CET/SDT and creativity
literatures. We connect and identify common themes in the CET/SDT and creativity literatures,
which tend to be mostly separate. We would like to begin by understanding why these theories,
7
particularly CET, took such a negative view of the role of compensation in motivation. We
would also like to understand the strengths and limitations of these theories, especially SDT, in
moving the literature on compensation, motivation, creativity, and performance forward in a
helpful way. To do so, we focus on what CET (and SDT) research to date can and cannot tell us
about PFP, motivation, performance, and creativity (in the workplace). We identify future
research needs, including the need to: focus more on explaining workplace behaviors
(performance, creativity), incorporate the role of choice (between jobs and between multiple
goals within jobs), and more systematically address the construct validity of motivation
measures, especially as it relates to changes in the conceptualization of motivation from CET,
intrinsic versus extrinsic, to SDT, autonomous (intrinsic motivation plus integrated and
identified extrinsic motivation) versus controlled (introjected and externally regulated extrinsic
motivation) and its propositions regarding how different types of motivation differ in their
quality and thus, their consequences.
We begin by briefly describing of CET and SDT logic and research.
Cognitive Evaluation Theory (CET)
Describing their thinking as being influenced by the work of Heider (1958) and
deCharms (1968), Ryan, Deci, and colleagues have, in cognitive evaluation theory (CET) and
later in self- determination theory (SDT), sought to understand the role of personal causation,
autonomy, and self-determination in work motivation (Ryan & Deci, 2006). CET categorizes
motivation as internal or external: ”the most basic distinction is between intrinsic motivation,
which refers to doing something because it is inherently interesting or enjoyable, and extrinsic
motivation, which refers to doing something because it leads to a separable outcome” (Ryan &
8
Deci, 2000, p. 55). They further see intrinsic motivation as “inherent”: “from the time of birth”
and children “in their healthiest states, are active, inquisitive, curious, and playful, even in the
absence of specific rewards (e.g. Harter, 1978).”
Importantly, this inherent intrinsic motivation can be put at risk by the environment.
They argue that “the maintenance and enhancement of this inherent propensity requires
supportive conditions, as it can be fairly readily disrupted by various nonsupportive conditions.”
Indeed, they go so far as to say that “our theory of intrinsic motivation does not concern what
causes intrinsic motivation…rather, it examines the conditions that elicit and sustain, versus
subdue and diminish, this innate propensity” (Ryan & Deci, 2000, p. 70).5 Compensation is one
such condition.
Under CET, performance-contingent extrinsic rewards may influence intrinsic motivation
via controlling and informational aspects (Ryan et al., 1983). If the task must be performed ”in
some particular way, at some particular time, or in some particular place . . . to receive the
reward, the reward tends to be experienced as controlling” (Ryan et al., 1983, p. 738). If so,
self-determination, and thus intrinsic motivation, will be undermined. In contrast, the
informational aspect of performance-contingent extrinsic rewards is seen as having the
potential for either a negative or positive influence on experienced competence and, thus,
intrinsic motivation. In fact, a positive informational effect seems to be the more typical
expectation (Arnold, 1985; Eisenberger, Rhoades, & Cameron, 1999; Fisher, 1978; Gagné &
Deci, 2005). Therefore, the overall effect of performance-contingent extrinsic rewards on
intrinsic motivation depends on whether the informational, competence-enhancing effect is
positive and strong enough to dominate the expected negative controlling effect.
9
The original CET paradigm used in studies that empirically examined the effect of
performance-contingent rewards on intrinsic motivation can be seen in Deci’s early work. For
example, Deci (1971) randomly assigned 24 introductory psychology students to one of two
groups where they participated in 3 sessions of 13 minute each. They worked on a puzzle and
were each asked to reproduce four puzzle configurations based on drawings. Experimental
group subjects were told at the beginning of the second session that they would be paid based
on how many puzzles they completed. (They were not paid in the first or third session.) In the
control group, subjects were never paid. The experimenter left the room for eight minutes “in
the middle of each session” and told the subjects “you may do whatever you like while I am
gone.” According to Deci, during this free-choice period, subjects could “read magazines, work
on the puzzle, stare around the room, and so on.” The measure of intrinsic motivation was “the
amount of time during the 8-minute free choice situation spent working on the puzzle.”
We re-produce Table 1 from Deci (1971) in our Exhibit 1. Deci’s main focus was on the
fact that the mean number of seconds spent on the puzzles during the free-choice period
increased from Session 1 to Session 3 for the control group, but decreased for the experimental
group. Deci’s interpretation is that paying the experimental group during the Session 2 thirteen
minute session undermined their subsequent intrinsic motivation.
Several reviews over the years have examined the body of evidence produced under this
paradigm (Deci, Koestner, & Ryan, 1999; Deci & Ryan, 1985; Eisenberger & Cameron 1996;
Rummel & Feinberg 1988; Tang & Hall 1995; Wiersma 1992). We focus here on the most
recent and comprehensive review, the Deci et al. (1999) meta-analysis, which, as summarized
10
by Ryan and Deci (2000, p. 70), “confirmed…that all expected tangible rewards made
contingent on task performance do reliably undermine intrinsic motivation.”
Deci et al.’s (1999) meta-analysis examined as outcomes two measures of intrinsic
motivation: free-choice behavior (as in the Deci, 1971 study described above) and self-reported
interest (interest/enjoyment). Deci et al. summarized results from 128 studies and found
consistent evidence of an undermining effect of contingent rewards on intrinsic motivation. In
the case of performance-contingent rewards, d = -.28 (95 % confidence interval of -.38 to -.18)
for the free-choice measure of intrinsic motivation and d = -.01 (95 % confidence interval of -.10
to +.08) for the self-reported interest measure of intrinsic motivation. Although the importance
of an effect size depends on the context, Cohen’s benchmarks for the effect size are sometimes
used: d >.20 “small”; d >.50 “medium”; d >.80 “large”. As such, d = -.28 for the free-choice
measure of intrinsic motivation is a small effect size. Recall also that the effect size for self-
reported interest/enjoyment was essentially zero.
Applying CET to the Workplace. Although not everyone agrees with the findings of the
Deci et al. (1999) meta-analysis (e.g., Eisenberger, Pierce, & Cameron, 1999), to simplify and
streamline our discussion, we take the Deci et al. findings as a starting point. That allows us to
focus on the applicability of CET (and SDT) to the workplace.
One major limitation of the Deci et al study (and general CET and later SDT paradigms)
becomes immediately apparent upon examining a standard definition of work motivation: “a
set of energetic forces that originate both within and as well as beyond an individual’s being, to
initiate work-related behavior, and to determine its form, direction, intensity, and duration”
(Pinder, 1998, p. 11). No “work-related behavior” was included in the Deci et al. meta-analysis.
11
The only behavior examined was what people did in their free time. “The difference between
measuring task behavior during a free-time session as opposed to measuring task performance
during the experimental session is important” (Wiersma, 1992, p. 104). Similarly, Locke and
Bartol (2000, p. 108) argued that “what people do during the time they are not being paid is of
no central importance” in work settings. So, what does CET research (and other research) tell
us about the effect of PFP not just on intrinsic motivation alone, but on key work-related
behaviors like performance and creativity?
Performance
Intrinsic Motivation ≠ Performance. Studying only the effect on intrinsic motivation of
PFP is not sufficient, given that performance is typically of great interest and it, in turn, is likely
a function of both intrinsic and extrinsic motivation (Cerasoli, Nicklin, & Ford, 2014; Fang, 1997;
Gerhart & Milkovich, 1992; Ledford, Gerhart, & Fang, 2013; Locke & Latham, 1990). Perhaps
the sole focus on intrinsic motivation in CET stemmed from the focus of the CET, SDT, and
creativity literatures on the quality of motivation (with intrinsic being seen as higher quality).
In Deci (1971), performance could, for example, have been measured by the time (less
time would indicate better performance) to solve the puzzles during the actual 13 minute on-
task experimental sessions. Better yet, the number of puzzles solved during on-task time could
have been measured. Neither types of data were reported. However, Deci makes an interesting
observation on his results in Table 1: “As one would expect, when the external rewards were
introduced to the experimental people during Time 2, their motivation increased.” Indeed, the
experimental group spent 26 % more time on the puzzles during the free-choice period in the
middle of Time 2 than they did in the middle of Time 1. As such, Deci seems to acknowledge
12
that performance-contingent rewards, while reducing intrinsic motivation, may have actually
increased total motivation and thus, by implication, possibly performance.
Importantly, however, performance has not typically been an outcome variable of
interest in the CET (and SDT) research program. The Deci et al. (1999) meta-analysis included
no results on performance.
Hamner and Foster (1975, p. 402, quoted in Wiersma, 1992) observed that “Whereas
Deci examined the performance vigilance after the contingent reward time period was over
(i.e., during a ‘free’ period), both expectancy theory and reinforcement theory models predict
performance during the reward period itself.” Deci, Koestner, & Ryan (1999) do not seem to
disagree, stating that “seldom has it been suggested that performance while the reward
contingency is in effect represents a measure of intrinsic motivation.” Likewise, Deci, Koestner,
& Ryan (2001, p. 7) state that in their 1999 meta-analysis, they “included only studies that
assess intrinsic motivation after the rewards had been clearly terminated, because while the
reward is in effect participants' behavior reflects a mix of intrinsic and extrinsic motivation.”
Deci et al. (1999, p. 657) state that “There is no lack of agreement between our viewpoint and
that of the operant and neo-operant theorists about the power of rewards to control
behavior…indeed, CET specifically proposes that it is because people are controlled by rewards
that they become less intrinsically motivated. Any lack of agreement concerns the unintended
consequences of rewards being used to control behavior.” Thus, extrinsic motivation seems to
have been viewed almost as a nuisance factor in studying intrinsic motivation. Of course, in the
workplace, what employees do while “the reward contingency is in effect” (i.e., behaviors such
as performance and creativity) is of great interest to most organizations.
13
Given the lack of attention to performance as an outcome in the CET literature,
implications regarding the performance consequences of PFP in the workplace based on CET
(e.g., Kohn, 1993; Pfeffer, 1998, Pink, 2009) must be viewed with great caution. Wiersma (1992,
Table 2) provides a compelling look at the difference between focusing only on intrinsic
motivation (the free-time measure) versus also looking at performance as an outcome of
extrinsic incentives. In 17 studies using a combined sample size of 865 subjects, he found a
negative (i.e., detrimental) effect (d = -.50) of extrinsic rewards on intrinsic motivation (free-
choice behavior), consistent with CET.
However, Wiersma also found 11 studies (N = 729) that examined the influence of
extrinsic rewards on performance. The effect was positive (d = +.34). As such, either the
positive effects of extrinsic rewards on performance via higher extrinsic motivation dominated
the negative effects via intrinsic motivation or the free-time measure of intrinsic motivation is
simply not relevant to performance (which is what happens during work time, not free time).
We also computed the sample size weighted d for the subset of 5 studies (N = 300) from
Wiersma’s Table 2 that reported both intrinsic motivation and performance effect sizes. The
intrinsic motivation d = -.71 whereas the performance d = +.49.
Finally, it is important to note that other research has examined the effect of PFP on
performance and has done so in both laboratory and workplace settings. (For reviews, see
Gerhart & Rynes, 2003 and Gerhart et al., 2009.). For example, a meta-analysis by Jenkins,
Mitra, Gupta, and Shaw (1998) found, based on 41 studies and 2,773 employees, that financial
incentives positively related to quantity of performance (mean r = +.32, which converts to d =
+.68).6 The d = .68 is much (2.4 times) larger in absolute magnitude than the effect size
14
reported by Deci et al. (1999) of performance-continent rewards on (free-choice measures of)
intrinsic motivation. Jenkins et al. also reported that the mean effect size in the field/workplace
settings (r = .46, d = 1.04) was roughly twice as large as the mean effect size based on
laboratory studies (r = .23, d = .47). They also used type of task as a moderator. The effect size
for tasks having more intrinsic interest (r = .33, d = .72) was nearly identical to the effect size in
less interesting tasks (r = .34, d = .72), which appears to conflict with the traditional CET/SDT
view that extrinsic rewards are best used for boring tasks.
Performance and PFP: Definition Matters. In addition to the need to study not only
intrinsic motivation, but also extrinsic motivation and performance, it is important to also
consider how different definitions/types of performance and PFP may influence findings and
conclusions. PFP is defined to include any plan where pay depends on performance, with the
specific PFP plan being a function of two performance measurement decisions (Gerhart et al.,
2009): degree of emphasis on results or behaviors and on individual or aggregate (i.e., group,
unit, or organization) level of analysis. Thus, for example, merit pay uses behaviors at the
individual level, whereas an individual incentive or sales commission uses results at the
individual level. Profit-sharing and stock plans use results at the aggregate level of analysis.
Meta-analytic evidence (Bommer, Johnson, Rich, & Podsakoff, Mackenzie, 1995)
indicates that r = .317 between results-based and behavior-based performance measures at the
individual level. (See also Heneman, 1986.) We also know that the shape and variance of
individual level performance distributions can be quite different, depending on whether results
or behaviors are used (Beck, Beatty, & Sackett, in press; O’Boyle, & Aguinis, , 2012). Thus, it
possible that the generally positive effects of PFP, which are based primarily on results-based,
15
individual level performance measures, could be weaker or stronger if behavior-based and/or
aggregate level performance measures were to be used. Clearly, the field has moved toward a
greater focus on performance (e.g., profits, productivity, shareholder return) measured at the
supra-individual/aggregate level and in terms of results (e.g., Arthur, 1994; Becker & Gerhart,
1996; Gerhart & Milkovich, 1990; Kim & Ployhart, 2014; Ployhart, Van Iddekinge, & MacKenzie,
2011). Theory and research on motivation will likely have a greater impact on practice to the
degree it likewise complements a focus on individual level motivation and performance
behaviors with an emphasis on performance in terms of results and/or at the aggregate level.
Creativity
Another important work behavior is creativity. To the degree that organizations must
increasingly compete based on knowledge and innovation, employee creativity may become
more important in formulating and executing strategies and achieving competitive advantage.
In her seminal work on creativity, Amabile (1983, p. 366; see also Amabile, 1996, p. 15) stated
that “a primarily intrinsic motivation to engage in an activity will enhance creativity, and a
primarily extrinsic motivation will undermine it. Hennessey and Amabile (1998, p. 675) further
stated that In her influential 1996 book, Amabile (p. 15) similarly stated that a “general
principle” is that “Intrinsic motivation is conducive to creativity, but extrinsic motivation is
detrimental.” In a commentary in the American Psychologist in 1998 on Eisenberger and
Cameron (1996), Hennessey and Amabile (1998, p. 675) acknowledged that there were “very
specific situations under which [extrinsic] reward can have either no impact or even a positive
impact on intrinsic motivation and creativity,” but they closed their commentary by saying:
“working for [extrinsic] reward…can be damaging to both intrinsic motivation and creativity.”
16
Thirty years later, these arguments continue to have influence. Forgeard and
Mecklenburg (p. 255), citing work by Amabile, state that “an extensive body of literature”
shows that “the main motivator of creative behavior is…intrinsic interest and enjoyment…that
intrinsic motivation enhances creativity, and extrinsic motivation can harm creativity insofar as
it decreases intrinsic motivation.” In the management literature, Amabile’s work also
influenced thinking. For example, Pfeffer (1998, p. 116) claimed that "extrinsic rewards
diminish intrinsic motivation" and "large extrinsic rewards can actually decrease performance in
tasks that require creativity and innovation" (See also Hunter, Cushenberry, & Friederich, 2012).
Yet, this negative view of the role of extrinsic motivation in creativity has begun to be
re-visited by Amabile herself, as well as by other creativity scholars. For example, in their
Annual Review chapter, Hennessey and Amabile (2010, p. 581) state that:
When investigations of…extrinsic constraints began about 30 years ago….High
levels of extrinsic motivation were thought to preclude high levels of intrinsic
motivation; as extrinsic motivators and constraints were imposed, intrinsic
motivation (and creativity) would necessarily decrease. Now…hundreds of
investigations later, most researchers…have come to appreciate the many
complexities of both motivational orientation and extrinsic motivators,
particularly expected reward….rewards can actually enhance intrinsic motivation
and creativity when they confirm competence, provide useful information in a
supportive way, or enable people to do something that they were already
intrinsically motivated to do. These boosting effects are most likely when initial
levels of intrinsic motivation are already strong (Amabile 1993).
Hennessey and Amabile make three key points: extrinsic rewards do not necessarily
undermine creativity and may actually enhance it, their positive effects are most likely when
17
intrinsic motivation is already high, and a positive effect of PFP may come via positive effects on
perceived competence.
Let’s begin with the potential positive effects of reward. Even in early work on extrinsic
motivation and creativity, one could sense an unease with categorically saying that PFP was
detrimental to creativity. For example, Hennessey and Amabile (1998, p. 675), while continuing
to view extrinsic rewards as generally detrimental to creativity, acknowledged that “when
working adults feel that incentive systems signal the value of their contribution, their
motivation and creativity of performance can be enhanced…,” a view that seems consistent
with the eventual recognition by SDT that PFP, if instrumental for achieving personal goals,
could positively influence creativity. Similarly, although Amabile (1998, p. 84) argued that
“Because monetary rewards make people feel as if they are controlled, such a tactic probably
won’t work,” she nevertheless acknowledged: “At the same time, not providing sufficient
recognition and rewards for creativity can spawn negative feelings within an organization.
People can feel used, or at least under-appreciated, for their creative efforts. And it is rare to
find the energy and passion of intrinsic motivation coupled with resentment” (p. 84). (This point
seems to say that employees generally feel inequitably treated if extrinsic rewards do not
correspond to their performance and inequity will negatively influence motivation and
creativity. In other words, it seems the argument is that employees generally prefer PFP. We
will later provide evidence to support this idea.) Shalley and Gilson (2004, p. 42) also noted
that “if creativity is positively evaluated but never rewarded, it may be that the employee is
given a mixed message and thus may or may not decide to continue trying to be creative.”
George (2008, p. 445) observed that “there might be very real sources of extrinsic motivation in
18
organizations” and that these “are not clearly and necessarily negative influences on creativity.”
(We observe that these arguments seem to treat extrinsic rewards as something that can get in
the way of creativity if not managed correctly. It is not clear whether extrinsic rewards are seen
as able to motivate creativity. This view seems similar in some ways to Hertzberg’s motivation-
hygiene theory logic.)
Hennessey and Amabile’s (2010) second point goes further and says that extrinsic
rewards may have a positive effect on creativity and, even more noteworthy, that such a
positive effect is actually more (not less) likely when intrinsic motivation is already high. This
idea seems to be at odds with not only CET, but also the broader SDT, which both focus on how
environmental factors like extrinsic rewards can “thwart” intrinsic motivation (Ryan & Deci,
2000). Latham (2007, p. 107; see also Locke & Latham, 1990) have argued that “it seems
unlikely that the needs for self-determination and competence can be wellsprings of human
motivation and, at the same time, be so fragile that their effects are negated by the most
common of life’s exigencies” as sometimes seems to be the case under CET and SDT.7
Hennessey and Amabile’s (2010) seem to agree that intrinsic motivation (and its expected
consequences such as creativity) are not so “fragile” and no so easily subdued and diminished
as argued by Ryan and Deci (2000). In fact, they argue that when intrinsic motivation is already
strong, wise use of PFP can strengthen it and creativity. SDT acknowledges that PFP can have a
positive effect (if internalized in the form of either “integrated regulation” or “identified
regulation”—see below), but again, continues to emphasize that PFP can also put intrinsic
motivation and creativity at great risk if it leads to less positive forms of extrinsic motivation
(“external regulation” or “introjected regulation”—see below).
19
Third, Hennessey and Amabile remind us that it is important to not forget about the
other pathway (in addition to autonomy) to intrinsic motivation under CET: the potentially
positive competence-enhancing informational aspect. Learned industriousness (Eisenberger,
1992), “focuses on the informational aspects of rewards…that [guide] goal-directed behavior
and, thus, increase creative performance” (Byron & Khazanchi, 2012, p. 809). In addition, as we
will see, Kanfer (1990) and later Gagné and Deci (2005) have recognized that performance-
contingent rewards can have a positive influence on autonomy.
What does the evidence say about the effect of using contingent (on creativity)
rewards? A recent meta-analysis by Byron and Khazanchi (2012) found that in 34 experimental
studies, the use of contingent (on creative performance) extrinsic incentives resulted in a
sizeable positive (not negative) effect on creative performance (Hedges g = .62). In eight
experimental studies, the effect size for contingent extrinsic rewards was again positive, but
much smaller (Hedges g = .07). (This major difference in effect size between experimental and
non-experimental studies serves to reinforce the need for caution in assessing how what is
observed in the laboratory will generalize to work organizations.)
Before one concludes that extrinsic incentives have a strong positive effect on creativity in work
organizations, one must look more closely at the creativity measures used in the laboratory
studies (and field studies). As an example, Eisenberger and Rhoades (2001) assessed creativity
in their experimental studies by asking 5th and 6th grade students and college students to write
titles for stories such as the “popcorn” story. The story titles were then coded/scored on a
creativity scale. The popcorn story asks subjects to imagine that they are“tiny golden kernel[s]
of popcorn lying in the bottom of a frying pan…snuggled up close to each other” and how they
20
go from feeling “cramped, uncomfortable, steaming hot, sweating, dizzy” to “Suddenly, you,
the popcorn kernel, feel yourself exploding, bursting…”
Although it is interesting to know that creativity in writing titles for the popcorn
story is higher under creativity contingent rewards, it is may not be sufficient evidence
to recommend to organizations that they can enhance creativity (and innovation) by
using PFP in this way. What about research outside of the lab? As noted, the
nonexperimental studies yield much smaller, positive effect sizes for PFP on creativity.
Here too, an examination of the measures is useful. It appears that the typical way to
measure creativity is to use a supervisory rating (For example, Eisenberger and Aselage
(2009), supervisors were asked to rate each employee on items including: “this
employee generates creative ideas” and “This employee takes a creative approach to
solving problems.”) Evidence of discriminant validity of the creativity rating from an
ordinary supervisory rating of performance is not provided. Yet, in our experience, a
rating of creativity/innovation behavior collected from supervisors correlated highly (r =
.65, corrected r = .72) with overall performance rating collected from supervisors (Fang,
1997, Table 4-9). Thus, it would be helpful to see better evidence that (a) in the jobs
being studied, employees have the latitude to display different levels of creativity and
do, and (b) that supervisors can assess creativity with sufficient independence from
other aspects of performance.
Another issue with the typical nonexperimental study of creativity included in
the Byron and Khazanchi meta-analysis is the level of analysis. All the studies appear to
have been conducted at the individual level of analysis and use employee perceptions of
21
PFP. Yet, there is no reason to expect that employees working in the same job in the
same unit or company work under different PFP systems. Thus, such designs would
appear to be incapable of telling us whether different organization-level or unit-level
PFP policies hinder or enhance creative behavior in organizations. What is needed is a
multilevel design that includes variance both between employees within organizations
and variance between organizations in PFP and other factors thought to influence
creativity. However, the validity concerns with supervisory performance ratings do not
disappear when aggregated to the organization level because while the temptation is to
use such average creativity ratings to capture between-organization differences, we
know that ratings vary between organizations for reasons other than true performance
or creativity differences. We also note that none of the studies included in the Byron
and Khazanchi (2012) meta-analysis appear to study innovation, which, in an
organization setting, is quite different and arguably more complex and of more applied
interest than creativity, especially as typically measured in research to date (Baer, 2012;
Montag, Maertz, & Baer, 2012). Creativity can be defined as “the development of novel,
potentially useful ideas” whereas innovation occurs only when those ideas “are
successfully implemented at the organization or unit level” Shalley, Zhou, & Oldham,
2004).
Given these definitions, one can see that there is much more to innovation in
organizations than just the “first step” (Shalley et al.) of creativity. There is little organization-
level research on how (non-executive) pay strategy affects employee innovation behavior and
22
outcomes. The employee attributes, work design, and pay policies needed to foster innovation
may be different, at least in part, from those that foster creativity.
Self-Determination Theory (SDT)
Extrinsic Motivation is No Longer Always Bad
Under self-determination theory (SDT), extrinsic rewards and motivation are, in sharp
contrast to CET, not always bad: “when rewards are administered in an autonomy-supportive
climate, they are less likely to undermine intrinsic motivation and, in some cases, can enhance
intrinsic motivation” (Gagné & Deci, 2005, p. 354, emphasis added. As we have seen, there has
also been a similar major and parallel change in thinking regarding the role of rewards in the
creativity literature.
Instead of focusing on intrinsic motivation versus extrinsic motivation and their different
consequences for experienced autonomy/control, SDT now, as Exhibit 2 shows, primarily
focuses on the distinction between autonomous (self-determined) and controlled (non self-
determined) motivation. “An important aspect of SDT is the proposition that extrinsic
motivation can vary in the degree to which it is autonomous versus controlled” (Gagné & Deci,
2005, p. 334). Exhibit 2 also shows how the various forms of motivation are arrayed in this
respect. Controlled extrinsic motivation (external regulation and introjected regulation)
corresponds to the more traditional, negative view of extrinsic motivation under CET, and
occurs when an activity is not inherently interesting (i.e., is not intrinsically motivating) and thus
requires an external reward contingency.8
STD, however, recognizes that “other types of extrinsic motivation result when a
behavioral regulation and the value associated with it have been internalized” (p. 334).
23
Internalization takes place when “external regulation of behavior is transformed into an
internal regulation and thus no longer requires the presence of an external contingency” (p.
334) to motivate the behavior and the result is that “people identify with the value of a
behavior for their own self-selected goals,” which leads to people experiencing “greater
freedom and volition because the behavior is more congruent with their personal goals and
identities.” In other words, this form of extrinsic motivation “is characterized not by the person
being interested in the activity for its own sake, but rather because the activity is perceived as
being instrumentally important for personal goals” (Gagné & Deci, p. 335). When this
identification is integrated “with other aspects of oneself,” it becomes more fully internalized
and “truly autonomous or volitional” and is called integrated regulation. The next most
internalized form of extrinsic motivation is identified regulation.
To make these motivation constructs more concrete, Exhibit 3 shows how they have
actually been measured.9 In Gagné et al. (2010), for example, a sample identified regulation
(autonomous extrinsic motivation) item is being motivated “Because this job fulfills my career
plans.” In contrast, a sample external regulation (controlled extrinsic motivation) item is
“Because this job affords me a certain standard of living.” Others have used controlled extrinsic
motivation items that are more negative. For example, Grant et al. (2011) used items such as
“So my parents and mentors won’t get mad at me” and “Because that’s what I’m supposed to
do.” Other examples of items used to measure external regulation extrinsic motivation are
“Because that’s something others (e.g., parents, friends, etc.) force me to do” (Vansteenkiste et
al. 2009) and “Because I’ll get in trouble if I don’t (Ryan & Connell, 1989). As we can see, the
24
degree of support for quality of motivation hypotheses will depend importantly on how the
different types of motivation are defined and measured (i.e., how positively or negatively).
In essence, it appears that SDT now says that performance-contingent pay does not
undermine intrinsic motivation if pay is “instrumentally important for other goals.” For years,
CET and those who sought to apply it to the workplace have consistently warned against the
use of performance-contingent pay and sometimes still do.10 Now, after many years, it appears
that there is recognition among some SDT scholars that pay might be helpful to employees for
achieving their personal goals and, thus, pay (including PFP) is not necessarily a negative.
Well…yes. That is consistent with other work on motivation. For example, Lawler (1971)’s
Figure 2-1 explicitly identifies the instrumentality of pay for achieving a wide range of needs
and he stated (p. 26) that “a given amount of pay derives its importance from its perceived
associations with the six types of needs mentioned by Maslow [1954]” and that “The evidence
rather clearly suggests that pay can be instrumental for the satisfaction of a variety of needs”
(p. 33). In any case, we commend SDT scholars for recognizing the need to make fundamental
conceptual changes and moving the theory in a direction that makes it more realistic and useful
in the workplace. Work by Gagné (Gagné & Deci, 2005; Gagné & Forest, 2008) has been especially
helpful in this regard.Why did SDT change from CET’s almost uniformly negative view of
performance-contingent extrinsic rewards? Gagné and Deci (2005, p. 356) acknowledge that
“Many… found [CET] of limited use with respect to promoting performance and satisfaction in
work organizations.”11 They believe that SDT “provides a fuller and more useful approach” for
understanding motivation in workplace settings. The SDT acknowledgment (at least as
described in Gagné & Deci) that, under SDT, rewards, depending on how administered, may
25
actually have a net positive effect on autonomous motivation (at a minimum, via their effect on
integrated and/or identified regulation forms of extrinsic motivation), is consistent with
arguments made earlier by others (see Fang & Gerhart’ 2012 review) regarding the potential for
positive effects on intrinsic motivation (using CET logic rather than SDT logic). For example,
Kanfer (1990) argued that “the controlling features of evaluative contingencies are likely to be
less salient due to widespread beliefs about the appropriateness of such evaluations in the
workplace” (p. 89). (See also Staw, 1977.) In this vein, research at the individual level
(Eisenberger and Aselage 2009; Eisenberger, Rhoades, & Cameron, 1999; Fang & Gerhart, 2000)
has found that employees who perceive stronger PFP perceive more, not less, autonomy and
intrinsic motivation. Research at the organization level further demonstrates that organizations
having a stronger PFP systems also have employees with higher levels of perceived autonomy
(DeVaro and Kurtulus, 2010; Fang & Gerhart, 2012). In other words, in the workplace, PFP and
autonomy tend to go hand in hand. SDT now seems to recognize this fact.
An important step in the evolution of thinking regarding the controlling aspects of
rewards in the workplace under SDT can be found in the Deci et al. (1999) meta-analysis and
review. (See Fang & Gerhart, 2012.) They concluded that extrinsic rewards “are more
detrimental for children than for college students” (p. 656) and that “This set of findings has
never been predicted before . . . so we can only speculate about what might be occurring”.
Deci et al. (1999, p. 656) suggest that “college students have greater cognitive capacity for
separating the informational and controlling aspects of rewards and are also more accustomed
to operating with performance goal orientations, so they may be more ready to interpret
rewards as indicators of their effective performance than as controllers of their behavior.”
26
In addition, the positive effects of performance-contingent rewards on competence
information were perhaps given too little weight in CET (Fang & Gerhart, 2012). Early on,
Harackiewicz, Manderlink and Sansone (1984) argued that performance-contingent rewards
provide a tangible symbol of achievement (a ‘cue value’) that can intensify the affective
significance and importance of accomplishment, making competency information more salient.
Other work also suggested that employee perceptions mattered and that employees might not
always distinguish as sharply between intrinsic and extrinsic aspects of motivation as
envisioned in CET (Brief & Aldag, 1977; Dyer & Parker, 1975) and that some aspects could be, as
later termed in SDT, autonomous. Dyer and Parker, for example, respondents often saw both
intrinsic and extrinsic aspects of outcomes such as achievement, recognition, prestige, and
advancement, consistent perhaps with SDT’s new-found forms of extrinsic motivation
(identified and regulated forms) that are instrumental for achieving goals.
Eisenberger and Cameron (1996) proposed learned industriousness theory, which,
consistent with arguments summarized above (see also Bandura, 1986), sees monetary rewards
as potentially competence-enhancing. Importantly, they also argue that “when a previously
unavailable reward is made contingent on performance, the reward may be experienced as
providing increased freedom of choice” (p. 1161).
We believe that the revised view of extrinsic rewards found in SDT is much more
consistent with the view that many employees have: they wish to succeed in their jobs and
their careers for both intrinsic and extrinsic reasons and these may not be distinct in their
minds. Many of the extrinsic reasons are very positive: they wish to have economic security
and the freedom that comes with it that allows them to choose how they spend their time, as
27
well as financial wherewithal to help others, whether it be to provide a comfortable and secure
life for their family and/or having the time and/or economic wherewithal to give to
philanthropy. As SDT emphasizes, people respond positively to autonomy and choice and
financial success may help make these outcomes more possible. As previously noted, empirical
evidence also indicates that autonomy and PFP covary in the workplace.
Quality of Motivation
So, where does this leave us? Under SDT, the intrinsic-extrinsic dichotomy is less
prominent than under CET. Now, under SDT, some extrinsic motivations are similar enough to
intrinsic motivation to be combined into a single category, autonomous motivation. The
(negative) remnants of the older CET view of extrinsic motivation are now combined into the
controlled motivation category. As noted, Gagné and Deci (2005) argue that a key difference
between SDT and most other motivation theories is its focus on quality (not just quantity) of
motivation. Specifically, in explaining how SDT (e.g., Ryan & Deci, 1996, p. 37; Ryan & Deci,
2000) is different from other work motivation theories such as goal-setting (e.g., Locke &
Latham, 1990), Gagné & Deci (2005), relying on work by Sheldon and Elliot (1999) and Sheldon,
Ryan, Deci, and Kasser (2004), state that “no attention is given to the fact that different goal
contents and different types of regulation of goal pursuits lead to different qualities of
performance” (p. 341). Ryan and Deci (2000, p. 69) argue that the more autonomous the
motivation, the higher its quality and the more “authentic” it is, which means that people “have
more interest, excitement, and confidence, which in turn is manifest…as enhanced
performance, persistence, and creativity” (emphasis added).
28
Similarly, Gagné and Deci (2005) contend that goal-setting theory does “not
differentiate the concept of performance in order to examine differences between the types of
goals and regulations that predict algorithmic versus heuristic performance. In contrast, SDT
proposes that autonomous motivation and intrinsic goals are better predictors of effective
performance on heuristic tasks (Vansteenkiste et al., 2004). Not only does quality of motivation
have different consequences, it requires different strategies to enhance (or at least not subdue
and diminish): “Strategies focused on optimizing the psychological need satisfactions
associated with active engagement of various tasks within specific performance settings thus
offer important alternatives to the use of rewards and other social controls to motivate
behavior” (Deci et al., 1999, p. 659).
The idea that intrinsic motivation is of higher quality than extrinsic motivation is also
found in the creativity literature. (Creativity would presumably be viewed as a form of heuristic
rather than algorithmic performance under SDT.) In the words of Amabile (1998, p. 78): “Not
all motivation is created equal (emphasis added). An inner passion to solve the problem at
hand leads to solutions far more creative than do external rewards such as money. This
component [is] called intrinsic motivation” (emphasis in original).
Challenges for SDT
Address Construct Validity Issues and Quality of Motivation. Returning to Exhibit 3, we
can further examine the content of key measures of extrinsic motivation and intrinsic
motivation over the years, as well as the content of later measures, which typically focus on
autonomous and controlled motivation. Consider the important early measure of intrinsic
motivation and extrinsic motivation developed by Harter (1981). That measure had two
29
significant attributes that influenced the path of research. First, it used a forced choice format.
Respondents (students) had to choose whether they were intrinsically or extrinsically motivated
(cite). One result was a tendency to view intrinsic motivation and extrinsic motivation as
opposites and not as complementary/able to occur simultaneously (Lepper et al., 2005).
Second, the definition of extrinsic motivation implied by the actual items used in the measure
make clear that it is either an overly narrow construct definition or a deficient measure because
the measure focuses only on the negative aspects of extrinsic motivation. As can be seen in
Exhibit 3, the items in Harter’s measure classifies as extrinsically motivated those students who
prefer easy work and who are dependent on the teacher for direction. As can also be seen
from Exhibit 3, Ryan and Connell’s (1989) measure also included only negative aspects of
extrinsic motivation. Clearly, extrinsic motivation in terms of “freedom of choice” (Eisenberger
and Cameron, 1996, p. 364) and the later logic of SDT, that certain types of extrinsic motivation
can be internalized and instrumental in achieving valued goals, was not incorporated.
Not surprisingly, when extrinsic motivation is defined and measured in terms of only
negative aspects (coercion, guilt, shame, avoiding trouble, preference for taking the easy path
rather than learning), that part of motivation will indeed prove itself to be of lower quality (in
the sense that higher levels predict more negative outcomes) than more positive aspects of
motivation. As now recognized by some SDT scholars, extrinsic rewards can be negative and/or
controlling, but clearly (some) extrinsic rewards can also be positive and autonomy-enhancing.
What happens when extrinsic rewards are not defined/measured as entirely negative?
Krishamurthy et al. (2014) studied open source software programmers, who perhaps tend not
to be thought of as extrinsically motivated. Extrinsic motivation items included “Working on an
30
open source project increases my opportunities for a better job.” Under SDT, such items would
be referred to as identified or integrated regulation and would be considered a form of
(extrinsic) autonomous motivation. This aspect of extrinsic motivation had a sizeable positive
correlation with intrinsic motivation (corrected r = .52), suggesting that open source software
programmers can be dually motivated (Amabile et al., 1994), making an undermining effect of
extrinsic motivation less likely. It also consistent with our earlier discussion that employees
tend to perceive intrinsic and many aspects of extrinsic motivation together.
In fact, the evidence suggests that the positive (more autonomous) aspects of extrinsic
motivation may not be empirically distinct from intrinsic motivation. Exhibit 4 summarizes the
corrected correlations between intrinsic motivation and the multiple forms of extrinsic
motivation now recognized under SDT based on four recent studies. We do see that intrinsic
motivation is distinct from external regulation extrinsic motivation. (In the case of introjected
extrinsic motivation, it depends on the measure/study). But, intrinsic motivation correlates
strongly, between .69 and .75, with integrated extrinsic motivation and between .64 and .80
with identified extrinsic motivation. As such, there is not much evidence of discriminant
validity. (One might be tempted to argue that these motivation dimensions are distinct because
the corrected correlation is less than 1.0. However, even correlations (uncorrected or
corrected) between different measures of the same construct (i.e., convergent validity) are
rarely if ever 1.0._Of course, SDT now combines intrinsic motivation with these two forms of
extrinsic motivation. So, in that sense, SDT has it right when it combines them into a single
construct called autonomous motivation.
31
Yet, under SDT, integrated extrinsic motivation is supposed to be more autonomous
than identified extrinsic motivation and both of these are specified as being less autonomous
than intrinsic motivation. Based on this initial evidence, these hypothesized differences are not
evident as of yet. Again, introjected regulation too does not “act” as would be expected under
SDT. Thus, the remaining quality of motivation logic under SDT, which Gagné and Deci (2005)
emphasize as something unique to SDT, may not be empirically supported.
Additional evidence regarding (the lack of) discriminant validity comes when one
compares intrinsic motivation and the more autonomous forms of extrinsic motivation
(integrated and identified) in terms of their correlations with outcome variables for evidence of
the differential prediction/quality of motivation hypothesis. As Exhibit 5 shows, the patterns of
correlations, in actuality, look strikingly similar. On the plus side, such similarity again suggests
that the SDT re-conceptualization of motivation (where some aspects of extrinsic motivation
combine with intrinsic motivation to form autonomous motivation) is more consistent with
how employees perceive intrinsic and extrinsic motivation (i.e., as often occurring together).
Less positive is that we can now clearly see that the CET focus for so many years on extrinsic
motivation as an almost exclusively negative, poor quality for of motivation was largely a result
of defining and measuring extrinsic motivation in overly narrow (negative) terms.
Further, as noted, although SDT now combines integrated extrinsic motivation and
identified extrinsic motivation with intrinsic motivation to form autonomous motivation, SDT
still, nevertheless, seems to wish to distinguish between the quality of extrinsic and intrinsic
motivation in some cases.12 For example, Gagné and Deci’s (2005) Proposition 1 states that:
“Autonomous extrinsic motivation will be more effective in predicting persistence on
32
uninteresting but effort-driven tasks, whereas intrinsic motivation will be more effective in
predicting persistence on interesting tasks.” (Presumably, the latter are more important and
impactful.) They note further that “Exactly how intrinsic motivation versus well-internalized
extrinsic motivation [i.e., integrated and identified regulation] will be differentially predictive in
the workplace is still to be determined, but it is an important issues…” Based on the evidence
we have examined (see Exhibits 4 and 5), there does not (at least thus far) appear to be
support for this remaining quality of motivation (differential prediction) proposition under SDT.
A second construct validity issue is social desirability. Lepper et al. (2005), in a school
setting, reported that intrinsic motivation scale correlated .42 with a measure of social
desirability, whereas their extrinsic motivation scale (composed of mostly externally regulated
and/or introjected aspects, see Exhibit 3) correlated -.24 with social desirability, a very large
difference (.66) in magnitude. That intrinsic motivation appears to be more socially desirable
than extrinsic motivation may be consistent with work on reward preferences, where the
importance of monetary rewards to respondents is thought to be understated (relative to
“higher order” rewards such as interesting and challenging work) when they are asked directly,
perhaps due to social desirability. (See Rynes et al., 2004 for a review.)13
If Quality of Motivation Matters: How to Facilitate Internalizing Extrinsic Motivation. If
the evidence were to support the SDT hypothesis that internalized extrinsic motivation is higher
quality, future research would then need to better understand the degree to which and how
extrinsic motivation can be internalized. Sheldon, Turban, Brown, Barrick, and Judge (2003), as
well as by Bloom and Colbert (2011) provide an overview of how compensation and other HR
practices might be designed to facilitate internalization. To the degree that employees already
33
(without any intervention) naturally perceive intrinsic, integrated extrinsic, and identified
extrinsic motivations as occurring in combination, such intervention would be less important.
(See our earlier discussion.) Nevertheless, the fact that much extrinsic motivation appears to be
internalized could alternatively reflect ongoing management practices in organizations to
support such internalization and it is possible that there would be a significant return on
investment from enhancing such programs, at least for some organizations and some jobs.
Incorporate Person Characteristics. Lepper et al. (2005) reported that intrinsic
motivation correlated positively with grade point average and achievement test score (r = .34, r
= .27, respectively), while extrinsic motivation correlated negatively with both (r = -.23, r = -.32,
respectively). The differences in correlations (.57 and .59) are large. Thus, a person’s
motivation profile may not be independent of their ability. The role of ability is rarely if ever
discussed in the CET/SDT literatures. Ability may be a source of omitted variable bias in field
settings (or in small sample size experiments where randomization may not result in equivalent
groups) if higher ability persons work in more complex, interesting jobs and are also more
intrinsically motivated. Those with higher ability may also make their jobs broader in scope
(Morgeson, Delaney-Klinger, & Hemingway), which may again influence intrinsic motivation.
The broader issue is that CET/SDT does not give much attention to the P part of the B = (P,E)
equation. Motivation, its level and causes, may vary by person (e.g., Barrick, Mount, & Li,
2013). Although Amabile, Hill, Hennessey, and Tighe (1994), for instance, developed the Work
Preference Inventory, which can be used to measure individual differences in intrinsic
motivation and extrinsic motivation orientation, it has been little used in CET/SDT.
34
Incorporate the Role of Choice: Goal Choice and ASA/Sorting. Work motivation can be
defined in terms of choices (Vroom, 1964). Although CET and SDT are, in a sense, all about the
importance of choice (i.e., autonomy), upon closer inspection, the treatment of choice is
narrow in conceptual terms and almost non-existent in empirical terms. On the conceptual
side, CET/SDT seem to focus primarily on the current job and how much autonomy the person
has within that job. But, as discussed below, there is much more to choice in the workplace.
Empirically, the only choice receiving much attention in CET seems to have been the choice of
what to do during free time (non-work time).
Goal Choice. Goals play a major role in motivating and directing behavior (Locke &
Latham, 2002) and/or in self-regulation (Bandura, 1997; Lord, Diefendorff, Schmidt, & Hall,
2010). In the typical CET laboratory experiment, the goal is given. In the workplace, by
contrast, employees must often choose which goal or goals to prioritize. Limited cognitive
resources likewise require prioritization and choice (Kanfer & Ackerman, 1989). Employers
generally wish to motivate particular goal choices. The equal compensation principle (Milgrom
& Roberts, 1992, p. 228; see also the “multitasking” literature, Prendergast, 1999) states that “If
an employee’s allocation of time or attention between two different activities cannot be
monitored by the employer, then either the marginal rates of return to the employee must be
equal, or the activity with the lower marginal rate of return receives no time or attention.”
Similarly, in psychology, Lawler (1971) earlier observed that “If an employee is not evaluated in
terms of an activity, he will not be motivated to perform it” (p. 171). (See Gerhart & Rynes,
2003 for a review.) In the creativity literature, a similar concern is that organizations may ask
employees to be (more) creative, but if they continue to reward other, more traditional
35
behaviors and not creativity, they may not see much creativity: “if creativity is a role
expectation, it should be rewarded appropriately” (Shalley et al., 2004, p. 40).
Schmidt and DeShon (2007) document that goal-performance discrepancy is an
important factor influencing which goal receives attention/priority. However, they also found
that once an incentive is introduced for achieving one goal, but not the other goal, the goal-
performance discrepancy becomes less important and the incentive becomes the major
influence on goal choice. Similarly, one study (Wright, George, Farnsworth, & McMahan, 1993)
observed in a laboratory setting that introducing a monetary incentive plan increased effort
toward the goals covered in the plan, but decreased effort toward other (e.g., prosocial helping
behaviors) goals. (See also related work by Bergeron, Shipp, Rosen, and Furst, 2013).
Sorting/Attraction-Selection-Attrition/ASA. Employees (and employers) also make job
choices. The standard CET paradigm randomly assigns subjects to performance-contingent pay
conditions or to control groups, ensuring that, on average, subjects’ PFP preferences are
independent of (i.e., not matched, mis-matched, to) their assigned pay condition. In contrast,
assignment of employees to work organizations (and their PFP systems) is not random (Fang &
Gerhart, 2012). Sorting (Lazear, 2000; Gerhart & Rynes, 2003) and attraction-selection-attrition
(ASA, Schneider, 1987) models describe the nonrandom (systematic) nature of the matching
process. To the degree such matching takes place, “there is less probability of a mismatch
between worker preferences for [performance-contingent pay] and the actual [pay] system
that covers them” (Fang & Gerhart). If indeed mis-matches are less likely in the workplace
(than in the laboratory) due to ASA/sorting processes, then a detrimental effect of PFP on
intrinsic motivation would also be less likely observed in the workplace. Those who experience
36
a loss of intrinsic motivation under PFP would presumably gravitate to jobs/organizations
having PFP policies that are a better fit with their preferences.
For matching to unfold, there must be sufficient employee movement. So, how much
employee movement is there? Based on data from the U.S. Bureau of Labor Statistics (BLS)
(Exhibit 6), a lot. Using three different snapshots, annual quit rates range from 19 to 29
percent and total separations range from 42 to 53 percent per year. These data apply only to
external movement. There is also substantial internal movement in many organizations as
employees change jobs (e.g., promotions, lateral moves) or as the content of their job evolves.
Is this movement nonrandom and consistent with the operation of ASA/sorting
processes? If so, we should see significant variance in employee attributes between
organizations relative to within organizations. Schneider et al (1998) found that 24 percent of
the variance in employee personality (using the four Myers–Briggs type indicator personality
variables) occurred between (could be explained by) organizations. Fang and Gerhart (2012)
focusing specifically on motivation-related traits (extrinsic motivation orientation, intrinsic
motivation, and internal work locus of control), found that 19 percent of the variance occurred
between organizations. Further, Fang and Gerhart reported that extrinsic motivation
orientation and internal work locus of control were higher in organizations that used PFP.
What about matching based on employee performance? Laboratory evidence suggests
that high performers are much more likely to choose PFP over fixed pay (Cadsby, Song, &
Tapon, 2007; Dohmen & Falk, 2011). In addition, field work shows that high performing
employees are more likely than others to quit when the pay-performance link is weak (Lazear,
2000; Nyberg, 2010; Salamin and Hom 2005; Trevor et al., 1997). Thus, an organization with
37
weak PFP would be expected to disproportionately lose its high performers, keep its lower
performers, and replace departing high performers with more low performers. Lazear (2000),
for example, observed that after the introduction of an incentive plan at a company, employee
productivity increased by 44%. However, when he looked only at the subsample of employees
who were there both before and after the incentive plan was implemented, their average
productivity had increased by 22%, thus accounting for only one-half of the total 44% increase.
What explained the other one-half? Less productive workers were more likely than more
productive workers to leave after the incentive plan was implemented. Even more important,
those hired after the incentive plan was put in place were more productive. Lazear referred to
the increase in productivity that occurred among employees there before and after the change
as an incentive effect and the increase due to the change in the workforce (Gerhart &
Milkovich, 1992) as a sorting effect (Gerhart & Rynes, 2003). (Note that the Deci et al. (1999)
and Jenkins et al. meta-analyses discussed earlier capture incentive effects, not sorting effects.)
As we saw earlier, Eisenberger and Cameron (1996, p. 364) argued almost 20 years ago
that “when a previously unavailable reward is made contingent on performance, the reward
may be experienced as providing increased freedom of choice.” Further insight into the role of
choice comes from a meta-analysis by Patall, Cooper, and Robinson (2008). Consistent with
CET/SDT, they found greater choice resulted in higher autonomy and intrinsic motivation. Of
particular interest, they also found that “reward has no impact on the effectiveness [in
enhancing intrinsic motivation] of having been given choice. This finding suggests that as long
as individuals have some control over the reward, it is not perceived as controlling, and the
positive effect of choice on motivation remains” (pp. 295-296). Also, of great interest was their
38
finding that the positive effect of choice on intrinsic motivation was much larger (more than
twice as large) for adults (d = .55) compared to children (d = .25). This finding suggests that
perceived causality perceptions of adults may be less fragile and less environment-dependent
than those of children, consistent with Deci et al.’s (1999, p. 656) suggestion that “college
students have greater cognitive capacity [than children] for separating the informational and
controlling aspects of rewards and are also more accustomed to operating with performance
goal orientations, so they may be more ready to interpret rewards as indicators of their
effective performance than as controllers of their behavior.” To the degree that is, PFP may be
experienced as more of an opportunity and more autonomy-enhancing than controlling.
Incorporate Equity Considerations. We also know that most employees have a strong
sense of equity. That has a number of implications. First, employee satisfaction is highly
sensitive to social comparisons, which are typically measured with a question such as (Williams
et al., 2006): “Compared with those working in similar jobs in other organizations [or another
basis of comparison], your pay is much worse, somewhat worse,” and so forth, with a high
score representing a positive comparison (i.e., much better).” In fact, the correlation with pay
satisfaction is .56 for internal (same organization) social comparisons and .57 for external
(different organization).
Second, although such a question does not address how an individual chooses the
comparison standard, we do know that performance is viewed by employees as the most
important factor that should determine pay among those in similar jobs. In studies that ask
respondents what criteria should be used or that they would use to allocate rewards, holding
the job constant, the consistent finding is that employee performance emerges as most
39
important (Dyer et al., 1976; Fossum & Fitch, 1985; Giaccobe-Miller et al., 2003; Sherer et al.,
1987; Zhou & Martocchio, 2003). In addition, the Williams et al. meta-analysis also reports that
employee pay satisfaction is strongly and positively related to employees’ perception of the
degree to which their organization uses PFP (r = .31).
The above evidence, which demonstrates the primary role of performance in
determining whether employees see pay as equitable, would seem to point to the need to use
PFP to achieve perceived equity. Given the sorting evidence we have seen, the use of PFP
would appear to be even more important for perceptions of equity among high performing
employees. Therefore, although CET’s focus was on the detrimental effect of PFP on intrinsic
motivation, the literatures we have just examined, in contrast, suggest that not using PFP may
be what causes the most serious problems when it comes to motivation, including intrinsic
motivation. Similarly, as we saw earlier, Amabile (1998, p. 84) argued that “it is rare to find the
energy and passion of intrinsic motivation coupled with resentment.” Finally, recent evidence
suggests that performance distributions for many occupations have a positive skew (e.g., power
law form), meaning that a small number of employees may create a disproportionate amount
of value through very high performance (Aguinis &O’Boyle, 2014) and thus, equitable pay may
require corresponding large pay differentials.
Future Research
Based on our review, we have identified several areas where we think future research
would be most helpful. First, a fundamental premise of both the creativity and (especially) SDT
literatures is that different types of motivation (intrinsic versus extrinsic, autonomous versus
controlled) differ in their quality (i.e., they have differential predictive power with respect to
40
key work outcomes). However, our reading is that empirical support for this (central) premise
is currently lacking and is an area for future research. If support continues to be lacking, it will
be necessary to determine if that is because of theoretical shortcomings (e.g., different types of
motivation are actually equivalent/fungible) or because of empirical limitations (e.g., the need
to re-formulate measures to better match construct definitions, the need to choose different
organization and or occupation settings where quality of motivation differences are most likely
to be of consequence). Without support for the quality of motivation logic, the value and
uniqueness of SDT would be seriously undermined. To the degree that quality of motivation
logic is supported, a related research need would be to examine the degree to which (and how)
organizations can influence the degree to which extrinsic motivation can be internalized.
Second, echoing earlier calls (e.g., Gerhart & Milkovich, 1992; Rynes et al. 2005), we
hope to see more integration of the compensation/PFP and motivation/psychological processes
literatures. Most organizations use PFP and give a major role to individual performance
(Gerhart & Fang, 2013). However, PFP can take many forms. Our impression is that the
CET/SDT and, to a lesser extent, the creativity literature, often envision incentives/PFP in terms
of piece rate plans where employees often perform simple, narrow, well-defined, repetitive
tasks to receive closely linked rewards. Yet, PFP in work organizations takes many forms and
piece rates are one of the least used forms. By far, the more common forms of PFP are annual
merit pay and promotion programs where performance is often (formally) evaluated only once
per year. Compensation payouts and promotion decisions tied to these performance
evaluations are typically on a similar timeline. As such, the magnitude of the relationship
between individual pay and individual performance often becomes clear only over the course of
41
multiple years as higher performers, on average, experience greater pay growth (e.g., Trevor et
al., 1997). Indeed, rather than pay and performance being so strongly linked that employees
feel controlled, the more likely situation in most organizations and occupations is that
employees question whether pay and performance are related at all (e.g., Milkovich et al.,
2014). Thus, the idea that PFP typically exerts a controlling influence is open to question. For
many organizations, the bigger issue is to strengthen the PPF link and communicate its
existence to employees. In any case, one avenue of research, consistent with SDT logic, would
be to determine how PFP plans can be designed to maximize motivation (incentive and sorting)
effects, which would include studying which plans are most conducive to perceived autonomy
and competence. The concept of reward “salience” (Cerasoli et al., 2014) may be of use here.
At a general level, we can extend our earlier classification of PFP plans to include not only
emphasis on results versus behaviors and individual versus aggregate level of analysis, but also
emphasis on short-term versus long-term performance measures, as well as process
dimensions such as participation and communication. Future research might assess the degree
to which these different PFP design characteristics influence different motivation outcomes
(e.g., as defined under SDT), as well as performance, creativity, and other key outcomes.
Third, SDT seems to continue to hypothesize (Cerasoli et al., 2014; Gagné & Deci, 2005;
Ryan & Deci, 2000) that PFP is a good idea for boring, routine tasks, but still potentially risky or
at least superfluous/irrelevant for interesting, complex, creative work. This argument seems to
flow from the SDT (and earlier creativity literature) argument regarding quality of motivation.
However, as noted, Jenkins et al. (1993) found no difference in the effects of incentives on
performance as a function of how intrinsically motivating the task was. The studies available to
42
Jenkins et al. were somewhat limited in terms of being done in the workplace and using jobs
high in intrinsic interest. Perhaps different samples should be studied also. In addition,
Hennessey and Amabile (2010, p. 581) have proposed a competing hypotheses, which states
that the boosting effects [of PFP] are most likely when initial levels of intrinsic motivation are
already strong.” Future research that examines whether PFP is irrelevant/detracts from
intrinsic motivation and performance in intrinsically motivating jobs versus “boosts” the
positive effects of intrinsic motivation would be valuable.
Fourth, future research on the role of PFP in creativity is needed. In addition to the
issues discussed above, we would like to see more work on innovation, rather than on creativity
alone. Related to that, we would like to see more work on creativity/innovation move beyond
the individual level of analysis and beyond sole reliance on aggregated supervisory ratings of
creativity. We need work on how PFP, motivation, and other psychological mechanisms
contribute to tangible measures of innovation at the group/organization levels. Fifth, as we
have noted, we would like to see greater recognition of the role of ASA/sorting, goal choice,
and equity in studying PFP, motivation, creativity, and performance in the workplace.
Finally, country differences are always a potentially important contextual factor in
human resources, including compensation and motivation, which can reflect differences in
(Milkovich et al., 2014) national culture, economic systems, regulation, and other institutions
(e.g., labor union strength). Thus far, most research on country differences in compensation,
PFP, and motivation has focused on national culture. Our own work (Gerhart & Fang, 2005;
Gerhart, 2008; Milkovich et al., 2014, Chapter 16; Rabl et al., 2011) indicates that, while there
are national culture-based differences, the similarities are often as (or more) noteworthy.
43
Thus, we encourage future work in this area (as in all areas) to focus on effect size magnitude
(in addition to statistical significance) to fully grasp the nature of differences and similarities.
Conclusion
Traditionally, the CET/SDT and creativity literatures viewed extrinsic rewards as
detrimental to performance and creativity and, similarly, viewed extrinsic motivation as being
of lower quality than intrinsic motivation. We have provided an analysis of how and why the
negative view of extrinsic incentives/PFP has shifted. We have also examined how SDT has
evolved from CET and its attempt to re-cast some types of extrinsic motivation as autonomous,
but still as of lower quality than intrinsic motivation. We have talked about the need for future
research on this quality of motivation topic and in other important areas, which we feel will
provide a better understanding of how PFP/extrinsic rewards influence performance, creativity,
and the role that motivation and other psychological processes play in this process.
44
References
Aguinis H, O’Boyle E. 2014. Star performers in twenty-first century organizations. Pers. Psychol. 67:313–
50
Amabile TM. 1983a. The Social Psychology of Creativity. New York: Springer-Verlag
Amabile TM. 1983b. The social psychology of creativity: a componential conceptualization. J. Personal.
Soc. Psychol. 45:357–77
Amabile TM. 1993. Motivational synergy: toward new conceptualizations of intrinsic and extrinsic moti-
vation in the workplace. Hum. Resour. Manag. Rev. 3:185–201
Amabile TM. 1996. Creativity in Context. Boulder, CO: Westview Amabile TM. 1998. How to kill
creativity. Harv. Bus. Rev. 76(5):76–87
Amabile TM, Hill KG, Hennessey BA, Tighe EM. 1994. The work preference inventory: assessing intrinsic
and extrinsic motivational orientations. J. Appl. Psychol. 66:950–67
Arnold HJ. 1985. Task performance, perceived competence, and attributed causes of performance as
determinants of intrinsic motivation. Acad. Manag. J. 28:876–88
Arthur JB. 1994. Effects of human resource systems on manufacturing performance and turnover. Acad.
Manag. J. 37:670–87
Baer M. 2012. Putting creativity to work: the implementation of creative ideas in organizations. Acad.
Manag. J. 55(5):1102–19
Bandura A. 1985. Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs,
NJ: Prentice-Hall
Bandura A. 1997. Self-Efficacy: The Exercise of Control. New York: Freeman
Barrick MR, Mount MK, Li N. 2013. The theory of purposeful work behavior: the role of personality,
higher- order goals, and job characteristics. Acad. Manag. Rev. 38:132–53
Bartol KM, Locke EA. 2000. Incentives and motivation. In Compensation in Organizations: Progress and
Prospects, ed. S Rynes, B Gerhart, pp. 104–47. San Francisco: New Lexington Press
Beck JW, Beatty AS, Sackett PR. 2014. On the distribution of job performance: the role of measurement
characteristics in observed departures from normality. Pers. Psychol. 67:531–66
Becker BE, Gerhart B. 1996. The impact of human resource management on organizational
performance: progress and prospects. Acad. Manag. J. 39:779–801
Benita M, Roth G, Deci EL. 2014. When are mastery goals more adaptive? It depends on experiences of
autonomy support and autonomy. J. Educ. Psychol. 106:258–67
Bergeron DM, Shipp AJ, Rosen B, Furst SA. 2013. Organizational citizenship behavior and career
outcomes: the cost of being a good citizen. J. Manag. 39(4):958–84
45
Bloom M, Colbert AE. 2011. An integration and extension of intrinsic motivation theories: the role of
core affect. Res. Pers. Hum. Resour. Manag. 30:73–114
Boiché J, Stephan Y. 2014. Motivational profiles and achievement: a prospective study testing potential
mediators. Motiv. Emot. 38(1):79–92
Bommer WH, Johnson JL, Rich GA, Podsakoff PM, Mackenzie SB. 1995. On the interchangeability of
objective and subjective measures of employee performance: a meta-analysis. Pers. Psychol. 48:587–
605
Brief AP, Aldag RJ. 1977. The intrinsic-extrinsic dichotomy: toward conceptual clarity. Acad. Manag. Rev.
2:496–500
Byron K, Khazanchi S. 2012. Rewards and creative performance: a meta-analytic test of theoretically
derived hypotheses. Psychol. Bull. 138:809–30
Cadsby CB, Song F, Tapon F. 2007. Sorting and incentive effects of pay-for-performance: an
experimental investigation. Acad. Manag. J. 50:387–405
Cerasoli PC, Nicklin JM, Ford MT. 2014. Intrinsic motivation and extrinsic incentives jointly predict per-
formance: a 40-year meta-analysis. Psychol. Bull. 140:980–1008
deCharms R. 1968. Personal Causation. New York: Academic
Deci EL. 1971. Effects of externally mediated rewards on intrinsic motivation. J. Personal. Soc.
Psychol.18(1):105–15
Deci EL, Koestner R, Ryan RM. 1999. A meta-analytic review of experiments examining the effects of
extrinsic rewards on intrinsic motivation. Psychol. Bull. 25:627–68
Deci EL, Koestner R, Ryan RM. 2001. Extrinsic rewards and intrinsic motivation in education:
reconsidered once again. Rev. Educ. Res. 71(1):1–27
Deci EL, Ryan RM. 1985. Intrinsic Motivation and Self-Determination in Human Behavior. New York:
Plenum
Deci EL, Ryan RM. 2013. The ombudsman: Do CEOs’ aspirations for wealth harm stockholders?
Interfaces 43(6):593–95
DeVaro J, Kurtulus FA. 2010. An empirical analysis of risk, incentives, and the delegation of worker
authority. Ind. Labor Relat. Rev. 63:641–61
Dohmen T, Falk A. 2011. Performance pay and multidimensional sorting: productivity, preferences, and
gender. Am. Econ. Rev. 101(2):556–90
Dyer L, Parker D. 1975. Classifying outcomes in work motivation research: an examination of the
intrinsic- extrinsic dichotomy. J. Appl. Psychol. 60:455–58
Dyer L, Schwab DP, Theriault RD. 1976. Managerial perceptions regarding salary increase criteria.
Pers.Psychol. 29:233–42
Eisenberger R. 1992. Learned industriousness. Psychol. Rev. 99:248–67
46
Eisenberger R, Aselage J. 2009. Incremental effects of reward on experienced performance pressure:
positive outcomes for intrinsic interest and creativity. J. Organ. Behav. 30:95–117
Eisenberger R, Cameron J. 1996. Detrimental effects of reward: reality or myth? Am. Psychol. 51:1153–
66 Eisenberger R, Pierce WD, Cameron J. 1999a. Effects of reward on intrinsic motivation—negative,
neutral, and positive: comment on Deci, Koestner, and Ryan (1999). Psychol. Bull. 125(6):677–91
Eisenberger R, Rhoades L. 2001. Incremental effects of reward on creativity. J. Personal. Soc.
Psychol.81:728–41
Eisenberger R, Rhoades L, Cameron J. 1999b. Does pay for performance increase or decrease perceived
self- determination and intrinsic motivation? J. Personal. Soc. Psychol. 77:1026–40
Fang M. 1997. A study of work motivation: the influence of organizational variables and individual char-
acteristics on work motivation and outcomes. PhD Thesis, Univ. Cornell
Fang M, Gerhart B. 2000. Does pay for performance diminish intrinsic interest? A workplace test using
cognitive evaluation theory and the attraction-selection-attrition hypothesis. Work. Pap. Univ.
Wisconsin, Madison
Fang M, Gerhart B. 2012. Does pay for performance diminish intrinsic interest? Int. J. Hum. Resour.
Manag.23:1176–96
Fisher CD. 1978. The effects of personal control, competence, and extrinsic reward systems on intrinsic
motivation. Organ. Behav. Hum. Decis. Process. 21:273–78
Forgeard MJC, Mecklenburg AC. 2013. The two dimensions of motivation and a reciprocal model of the
creative process. Rev. Gen. Psychol. 17:255–66
Fossum JA, Fitch MK. 1985. The effects of individual and contextual attributes on the sizes of
recommended salary increases. Pers. Psychol. 38:587–602
Frey BS, Jegen R. 2001. Motivation crowding theory. J. Econ. Surv. 15:589–611
Frey BS, Oberholzer-Gee F. 1997. The cost of price incentives: an empirical analysis of motivation
crowding- out. Am. Econ. Rev. 87(4):746–55
Frey BS, Osterloh M. 2012. Stop tying pay to performance. Harv. Bus. Rev. 90:51–52
Gagné M, Deci EL. 2005. Self-determination theory and work motivation. J. Organ. Behav. 26:331–62
Gagné M, Forest J. 2008. The study of compensation systems through the lens of self-determination
theory: reconciling 35 years of debate. Can. Psychol. 49(3):225–32
Gagné M, Forest J, Gilbert MH, Aubé C, Morin E, Malorni A. 2010. The motivation at work scale:
validation evidence in two languages. Educ. Psychol. Meas. 70:628–46
George JM. 2008. Creativity in organizations. Acad. Management Ann. 1:439–77
Gerhart B. 2008. Compensation and national culture. In Global Compensation: Foundations and
Perspectives, ed. LR Gomez-Mejia, S Werner, pp. 142–57. London: Routledge
47
Gerhart B, Fang M. 2005. National culture and human resource management: assumptions and
evidence. Int. J. Hum. Resour. Manag. 16(6):971–86
Gerhart B, Fang M. 2013. Pay for (individual) performance: issues, claims, evidence and the role of
sorting effects. Hum. Resour. Manag. Rev. 24:41–52
Gerhart B, Milkovich GT. 1990. Organizational differences in managerial compensation and financial
performance. Acad. Manag. J. 33:663–91
Gerhart B, Milkovich GT. 1992. Employee compensation: research and practice. In Handbook of
Industrial and Organizational Psychology, ed. MD Dunnette, LM Hough, pp. 481–570. Palo Alto, CA:
Consulting Psychologists Press
Gerhart B, Rynes S. 2003. Compensation: Theory, Evidence, and Strategic Implications. Thousand Oaks,
CA: Sage
Gerhart B, Rynes SL, Fulmer IS. 2009. Pay and performance: individuals, groups, and executives. Acad.
Manag. Ann. 3:251–315
Gerhart B, Trevor C, Graham M. 1996. New directions in employee compensation research. In Research
in Personnel and Human Resource Management, ed. GR Ferris, pp. 143–203. Greenwich, CT: JAI Press
Giacobbe-Miller JK, Miller DJ, Zhang W, Victorov VI. 2003. Country and organization-level adaptation to
foreign workplace ideologies: a comparative study of distributive justice values in China, Russia and the
United States. J. Int. Bus. Stud. 34:389–406
Grant AM, Berry JW. 2011. The necessity of others is the mother of invention: intrinsic and prosocial
motivations, perspective-taking, and creativity. Acad. Manag. J. 54:73–96
Grant AM, Nurmohamed S, Ashford SJ, Dekas K. 2011. The performance implications of ambivalent
initiative: the interplay of autonomous and controlled motivations. Organ. Behav. Hum. Decis. Process.
116(2):241–51
Gupta N, Shaw JD. 2014. Employee compensation: the neglected area of HRM research. Hum. Resour.
Manag. Rev. 24(1):1–4
Hamner WC, Foster LW. 1975. Are intrinsic and extrinsic rewards additive: a test of Deci’s cognitive
evaluation theory of task motivation. Organ. Behav. Hum. Perform. 14(3):398–415
Harackiewicz JM, Manderlink G, Sansone C. 1984. Rewarding pinball wizardry: effects of evaluation and
cue value on intrinsic interest. J. Personal. Soc. Psychol. 47:287–300
Harter S. 1978. Effectance motivation reconsidered: toward a developmental model. Hum. Dev. 1:661–
69 Harter S. 1981. A new self-report scale of intrinsic versus extrinsic orientation in the classroom:
motivational and informational components. Dev. Psychol. 17:300–12
Heider F. 1958. The Psychology of Interpersonal Relations. New York: Wiley
Heneman RL. 1986. The relationship between supervisory ratings and results-oriented measures of
performance: a meta-analysis. Pers. Psychol. 39:811–26
48
Hennessey BA, Amabile TM. 1998. Reward, intrinsic motivation, and creativity. Am. Psychol. 53(6):674–
75
Hennessey BA, Amabile TM. 2010. Creativity. Annu. Rev. Psychol. 61:569–98
Herzberg F. 1968. One more time: How do you motivate employees? Harv. Bus. Rev. 46(1):53–62
Hunter ST, Cushenbery L, Friedrich T. 2012. Hiring an innovative workforce: a necessary yet uniquely
challenging endeavor. Hum. Resour. Manag. Rev. 22(4):303–22
Jenkins DG, Mitra A, Gupta N, Shaw JD. 1998. Are financial incentives related to performance? A meta-
analytic review of empirical research. Appl. Psychol. 83:777–87
Kanfer R. 1990. Motivation theory and industrial and organizational psychology. In Handbook of
Industrial and Organizational Psychology, ed. MD Dunnette, LM Hough, pp. 75–170. Palo Alto, CA:
Consulting Psychologists Press
Kanfer R, Ackerman PL. 1989. Motivation and cognitive abilities: an integrative/aptitude-treatment in-
teraction approach to skill acquisition. Appl. Psychol. 74(4):657–90
Kerr S. 1975. On the folly of rewarding A, while hoping for B. Acad. Manag. J. 18:769–83
Kim Y, Ployhart RE. 2014. The effects of staffing and training on firm productivity and profit growth
before, during, and after the Great Recession. J. Appl. Psychol. 99:361–89
Kohn A. 1993. Why incentive plans cannot work. Harv. Bus. Rev. Sept. 1993, pp. 54–63
Kreps DM. 1997. Intrinsic motivation and extrinsic incentives. Am. Econ. Rev. 87:359–64
Krishnamurthy S, Ou S, Tripathi AK. 2014. Acceptance of monetary rewards in open source software
development. Res. Policy 43:632–44
Kunz AH, Pfaff D. 2002. Agency theory, performance evaluation, and the hypothetical construct of
intrinsic motivation. Account. Organ. Soc. 27(3):275–95
Latham GP. 2007. Work Motivation: History, Theory, Research, and Practice. Thousand Oaks, CA: Sage
Latham GP, Pinder CC. 2005. Work motivation theory and research at the dawn of the twenty-first
century.Annu. Rev. Psychol. 56:485–516
Lawler EE. 1971. Pay and Organizational Effectiveness: A Psychological View. New York: McGraw-Hill
Lawler EE. 1981. Pay and Organization Development. Reading, MA: Addison Wesley
Lazear E. 2000. Performance pay and productivity. Am. Econ. Rev. 90:1346–61
Ledford GE Jr, Gerhart B, Fang M. 2013. Negative effects of extrinsic rewards on intrinsic motivation:
more smoke than fire. WorldatWork J. 22(2):17–29
Lepper MR, Corpus JH, Iyengar SS. 2005. Intrinsic and extrinsic motivational orientations in the
classroom: age differences and academic correlates. J. Educ. Psychol. 97(2):184–96
49
Lepper MR, Greene D, Nisbett RE. 1973. Undermining children’s intrinsic interest with extrinsic rewards:
a test of the “overjustification” hypothesis. J. Personal. Soc. Psychol. 28:129–37
Locke EA, Latham GP. 1990. A Theory of Goal Setting and Task Performance. Englewood Cliffs, NJ:
Prentice-Hall
Locke EA, Latham GP. 2002. Building a practically useful theory of goal setting and task motivation. Am.
Psychol. 57:705–17
Locke EA, Latham GP. 2009. Has goal setting gone wild, or have its attackers abandoned good
scholarship? Acad. Manag. Perspect. 23:17–23
Lord RG, Diefendorff JM, Schmidt AM, Hall RJ. 2010. Self-regulation at work. Annu. Rev. Psychol. 61:543–
68
Milgrom P, Roberts J. 1992. Economics, Organization, and Management. Englewood Cliffs, NJ: Prentice
Hall Milkovich GT, Newman JM, Gerhart B. 2014. Compensation. Boston: Irwin/McGraw-Hill
Montag T, Maertz CP, Baer M. 2012. A critical analysis of the workplace creativity criterion space. J.
Manag. 38(4):1362–86
Moran CM, Diefendorff JM, Kim T, Liu Z. 2012. A profile approach to self-determination theory
motivations at work. J. Vocat. Behav. 81:354–63
Morgeson FP, Delaney-Klinger K, Hemingway MA. 2005. The importance of job autonomy, cognitive
ability, and job-related skill for predicting role breadth and job performance. J. Appl. Psychol. 90:399–
406
Nyberg AJ. 2010. Retaining your high performers: moderators of the performance-job satisfaction-
voluntary turnover relationships. J. Appl. Psychol. 95:440–53
O’Boyle EH, Aguinis H. 2012. The best and the rest: revisiting the norm of normality of individual
performance. Pers. Psychol. 65:79–119
Patall EA, Cooper H, Robinson JC. 2008. The effects of choice on intrinsic motivation and related
outcomes: a meta-analysis of research findings. Psychol. Bull. 134(2):270–300
Pfeffer J. 1998. Six dangerous myths about pay. Harv. Bus. Rev. May 1998, pp. 109–19
Pinder CC. 1998. Work Motivation in Organizational Behavior. Englewood Cliffs, NJ: Prentice Hall
Pink D. 2009. Drive: The Surprising Truth about What Motivates Us. New York: Penguin
Ployhart R, Van Iddekinge C, MacKenzie W. 2011. Acquiring and developing human capital in service
contexts: the interconnectedness of human capital resources. Acad. Manag. J. 54:353–68
Prendergast C. 1999. The provision of incentives in firms. J. Econ. Lit. 37(1):7–63
Prendergast C. 2008. Intrinsic motivation and incentives. Am. Econ. Rev. 98:201–5
50
Rabl T, Jayasinghe M, Gerhart B, Kühlman TM. 2011. How much does country matter? A meta-analysis of
the HPWP systems-business performance relationship. Presented at Proc. Acad. Manag. Annu. Meet.,
71st, Aug. 12–16, San Antonio, TX
Robinson M. 2014. The best-paying companies in America. Business Insider, April 16
Roy D. 1952. Quota restriction and gold bricking in a machine shop. Am. J. Sociol. 57:427–42
Rummel A, Feinberg R. 1988. Cognitive evaluation theory: a meta-analytic review of the literature. Soc.
Behav. Personal. 16:147–64
Ryan RM, Connell JP. 1989. Perceived locus of causality and internalization: examining reasons for acting
in two domains. J. Personal. Soc. Psychol. 57:749–61
Ryan RM, Deci EL. 2000. Self-determination theory and the facilitation of intrinsic motivation, social
development, and well-being. Am. Psychol. 55(1):68–78
Ryan RM, Deci EL. 2006. Self-regulation and the problem of human autonomy: Does psychology need
choice, self-determination, and will? J. Personal. 74:1557–86
Ryan RM, Mims V, Koestner R. 1983. Relation of reward contingency and interpersonal context to
intrinsic motivation: a review and test using cognitive evaluation theory. J. Personal. Soc. Psychol.
45:736–50
Rynes SL, Gerhart B, Minette KA. 2004. The importance of pay in employee motivation: discrepancies
between what people say and what they do. Hum. Resour. Manag. Rev. 43:381–94
Rynes SL, Gerhart B, Parks L. 2005. Personnel psychology: performance evaluation and pay for
performance. Annu. Rev. Psychol. 56:571–600
Salamin A, Hom PW. 2005. In search of the elusive U-shaped performance-turnover relationship: Are
high performing Swiss bankers more liable to quit? J. Appl. Psychol. 90:1204–16
Sanders WG, Hambrick DC. 2007. Swinging for the fences: the effects of CEO stock options on company
risk taking and performance. Acad. Manag. J. 50:1055–78
Schmidt AM, Beck JW, Gillespie JZ. 2013. Motivation. In Handbook of Psychology: Industrial and Orga-
nizational Psychology, ed. NW Schmitt, S Highhouse, I Weiner, 12:311–40. Hoboken, NJ: Wiley
Schmidt AM, DeShon RP. 2007. What to do? The effects of discrepancies, incentives, and time on
dynamic goal prioritization. J. Appl. Psychol. 92(4):928–41
Schneider B. 1987. The people make the place. Pers. Psychol. 40:437–53
Schneider B, Smith DB, Taylor S, Fleenor J. 1998. Personality and organizations: a test of the
homogeneity of personality hypothesis. J. Appl. Psychol. 83:462–70
Shalley CE, Gilson LL. 2004. What leaders need to know: a review of social and contextual factors that
can foster or hinder creativity. Leadersh. Q. 15:33–53
Shalley CE, Zhou J, Oldham GR. 2004. The effects of personal and contextual characteristics on creativity:
Where should we go from here? J. Manag. 30:933–58
51
Shaw JD. 2014. Pay dispersion. Annu. Rev. Organ. Psychol. Organ. Behav. 1:521–44
Sheldon KM, Elliot AJ. 1999. Goal striving, need-satisfaction, and longitudinal well-being: the self-
concordance model. J. Personal. Soc. Psychol. 76:482–97
Sheldon KM, Kasser T. 1995. Coherence and congruence: two aspects of personality integration. J.
Personal.Soc. Psychol. 68(3):531–43
Sheldon KM, Ryan RM, Deci EL, Kasser T. 2004. The independent effects of goal contents and motives on
well-being: it’s both what you pursue and why you pursue it. Personal. Soc. Psychol. Bull. 30:475–86
Sheldon KM, Turban DB, Brown KG, Barrick MR, Judge TA. 2003. Applying self-determination theory to
organizational research. Res. Pers. Hum. Resour. Manag. 22:357–94
Sherer PD, Schwab DP, Heneman HG. 1987. Managerial salary-raise decisions: a policy-capturing
approach.Pers. Psychol. 40(1):27–38
SHRM (Soc. Hum. Resour. Manag.). 2014. Job satisfaction and engagement survey. Alexandria, VA.
http://www.shrm.org/Research/SurveyFindings/Documents/14-
0028%20JobSatEngage_Report_FULL_FNL.pdf
Staw BM. 1976. Intrinsic and Extrinsic Motivation. Morristown, NJ: General Learning Press
Tang SH, Hall VC. 1995. The overjustification effect: a meta-analysis. Appl. Cogn. Psychol. 9:365–404
Tremblay MA, Blanchard C, Villeneuve M, Taylor S, Pelletier LG. 2009. Work extrinsic and intrinsic
motivation scale: its value for organizational psychology research. Can. J. Behav. Sci. 41(4):213–26
Trevor C, Gerhart B, Boudreau JW. 1997. Voluntary turnover and job performance: curvilinearity and the
moderating influences of salary growth and promotions. J. Appl. Psychol. 82:44–61
Truong A. 2014. Google tops this list of companies that offer the best pay and perks. FastCompany. May
23. http://www.fastcompany.com/3031005/most-innovative-companies/google-tops-this-list-of-
companies- that-offer-the-best-pay-and-per
Vansteenkiste M, Sierens E, Soenens B, Luyckx K, Lens W. 2009. Motivational profiles from a self-
determination theory perspective: the quality of motivation matters. J. Educ. Psychol. 101:671–88
Vroom VH. 1964. Work and Motivation. New York: Wiley
Wiersma UJ. 1992. The effects of extrinsic rewards in intrinsic motivation: a meta-analysis. J. Occup.
Organ. Psychol. 65:101–14
Williams ML, McDaniel MA, Nguyen NT. 2006. A meta-analysis of the antecedents and consequences of
pay level satisfaction. J. Appl. Psychol. 91(2):392–413
Wright PM, George JM, Farnsworth R, McMahan GC. 1993. Productivity and extra-role behavior: the
effects of goals and incentives on spontaneous helping. J. Appl. Psychol. 78:374–81
Zhou J, Hoever IJ. 2014. Research on workplace creativity: a review and redirection. Annu. Rev. Organ.
Psychol. Organ. Behav. 1:333–59
52
Zhou J, Martocchio JJ. 2001. Chinese and American managers’ compensation award decisions: a
comparative policy-capturing study. Pers. Psychol. 54:115–45
53
Exhibit 1. Mean Number of Seconds Spent Working on the
Puzzle During the Eight-Minute Free Choice Sessions (Deci,
1971)
Session 1
Session 2
Session 3
Experimental (n = 12)
248
314
199
Control (n = 12)
214
206
242
54
Exhibit 2. Types of Motivation and their Degree of Self-Determination
Source: Ryan RM, Deci EL. 2000. Self-determination theory and the facilitation of intrinsic
motivation, social development, and well-being. Am. Psychol. 55(1): 68-78
55
Exhibit 3. Content (sample items) from Measures of Extrinsic and Intrinsic Motivation
Study
Setting
Controlled or Extrinsic
Motivation Item Content
Autonomous or Intrinsic
Motivation Content
Krishamurthy et
al. (2014)
Workplace
Extrinsic Motivation
Working on an open source
project increases my
opportunities for a better job.
Intrinsic Motivation
Writing open source software
programs is fun.
Boiche´ &
Stephan (2014)
College
Students
Extrinsic Motivation
Introjected Regulation
To prove to myself that I am
capable of completing my
college degree
External Regulation
Because with only a high-school
degree I would not find a high-
paying job later on.
Autonomous Motivation
Intrinsic Motivation
Because I experience pleasure
and satisfaction while learning
new things.
Identified Regulation
Because I think that a college
education will help me better
prepare for the career I have
chosen’.
Moran et al.
(2012)
Workplace
Controlled Motivation
External Motivation
Because my boss wants me to
do it.
Introjected Motivation
Because I would feel guilty if I
did not do well.
Autonomous Motivation
Intrinsic Motivation
Because the work is fun.
Integrated Motivation
Because my work is a big part of
who I am.
Identified Motivation
Because I believe the work is
valuable.
Grant et al.
(2011), Study 1
[adapted from
Ryan & Connell,
1989 scale]
Workplace
Controlled Motivation
So my parents and mentors
won’t get mad at me.
Because that’s what I’m
supposed to do.
Because I don’t want others to
be mad at me.
Autonomous Motivation
Because I enjoy the process.
Because it’s fun.
Because I enjoy doing it.
Grant et al.
(2011), Study 2
Workplace
Controlled Motivation
Autonomous Motivation
56
[adapted from
Ryan & Connell,
1989 scale]
Because I need to earn money.
Because I need to pay bills.
Because I need the income.
Because it’s fun.
Because I enjoy it.
Gagné et al.
(2010)
Workplace
Controlled Motivation
External Regulation
Because this job affords me a
certain standard of living.
Introjected Regulation
Because my work is my life and I
don’t want to fail.
Autonomous Motivation
Intrinsic Motivation
Because I enjoy this work very
much.
Identified Regulation
Because this job fulfills my
career plans.
Tremblay et al.
(2009)
Workplace
Work Self-Determination = low
External Regulation
For the income it provides me.
Introjected Regulation
Because I want to be very good
at this work, otherwise I would
be very disappointed.
Amotivation
I don’t know why, we are
provided with unrealistic
working conditions
Work Self-Determination = high
Intrinsic Motivation
For the satisfaction I experience
when I am successful at doing
difficult tasks.
Integrated Regulation
Because it has become a
fundamental part of who I am.
Identified Regulation
Because it is the type of work I
have chosen to attain certain
important objectives.
Vansteenkiste et
al (2009)
High School
and College
Students
Controlled Motivation
External Regulation
Because that’s something others
(parents, friends, etc.) force me
to do.
Introjected Regulation
Because I would feel ashamed if
I didn’t study.
Autonomous Motivation
Intrinsic Motivation
Because I enjoy doing it.
Identified Regulation
Because I want to learn new
things
Lepper et al.
(2005)
School
Extrinsic Motivation
Easy Work
I like to learn just what I have to
in school.
Pleasing teacher
I read things because the
teacher wants me to.
Intrinsic Motivation
Challenge
I like difficult schoolwork
because I find it more
interesting.
Curiosity
I read things because I am
interested in the subject.
57
Dependence on teacher
If I get stuck on a problem I ask
the teacher for help.
Independent Mastery
I like to do my schoolwork
without help.
Sheldon & Kasser
(1995)
College
Students
Controlled Motivation
Autonomous Motivation
Each participant generates 10 goal strivings and then rated them
“as to how much they pursued them for each of four reasons”:
External
Because somebody else wants
you to or because you’ll get
something from somebody if
you do.
Introjected
Because you would feel
ashamed, guilty, or anxious if
you didn’t strive for this.
Each participant then rated each
striving “as to how much it
helped take them toward
possible futures in six culturally
endorsed value domains”:
Intrinsic
Purely because of the fun and
enjoyment
Identified
Because you really believe it is
an important goal to have—you
endorse it freely and
wholeheartedly.
Sheldon & Kasser
(1995)
College
Students
Extrinsic Motivation Coherence
Intrinsic Motivation Coherence
Each participant “then rated each striving as to how much it helped
take them toward possible futures in six culturally endorsed value
domains:”
Financial success: having a job
that pays very well and having a
lot of nice possessions.
Fame and recognition: being
known and admired by many
people.
Physical appearance: looking
good and being attractive to
others.
Self-acceptance and personal
growth: being happy and having
a meaningful life.
Intimacy and friendship: having
many close and caring
relationships with others.
Societal contribution: working
to help make the world a better
place.
58
Amabile, Hill,
Hennessey, &
Tighe (1994)
Workplace
and College
Students
Extrinsic Motivation
I am strongly motivated by the
[grades] [money] I can earn.
Intrinsic Motivation
I enjoy trying to solve complex
problems.
Ryan & Connell
(1989)
School
External Perceived Locus of
Causality (Controlled)a
External
Because I’ll get in trouble if I
don’t.
Introjected
Because I’ll feel ashamed of
myself if I don’t.
Internal Perceived Locus of
Causality (Autonomous)a
Intrinsic
Because it’s fun.
Identification
Because I want to learn new
things.
Harter (1981)
School
Extrinsic Motivation
Intrinsic Motivationb
Preference forc:
Easy work
Pleasing teacher/getting grades
Dependence on teacher
Reliance on teacher’s judgment
External criteria
Challenge
Curiosity/interest
Independent mastery
Independent judgment
Internal criteria
Note: Unless otherwise noted, a single item is a sample item. Underlined words are labels of sub-
dimensions. “School” means respondents were students in elementary school or middle school. Where
factor loadings are reported, the sample item is the item having the largest factor loading on that
dimension.
aRyan and Connell (1989) did not use the construct labels of “controlled and “autonomous.” We added
them to the table because such labels are used in later research using this item content.
bGiven its influence on later scales, we include all items in Harter’s scale.
cAccording to Harter (1981, p. 302): The child is first asked to decide which kind of kid is most like him or
her and then asked whether this is only sort of true or really true for him or her. As such, the scale has
a forced choice aspect.
59
Exhibit 4. Corrected Correlations between Intrinsic Motivation and Four Types of
Extrinsic Motivation, Workplace Settings
Tremblay et al.
(2009)
Gagné et al.
(2010)
Moran et al.
(2012)
Boiché and
Stephan (2014)
Extrinsic
Motivation Type
Correlation of Intrinsic Motivation with:
Integrated
Extrinsic
Motivation
.69
.75
Identified
Extrinsic
Motivation
.70
.80
.64
.77
Introjected
Extrinsic
Motivation
.63
.43
.33
.68
Externally
Regulated
Extrinsic
Motivation
.11
.13
.13
-.18
Note: For Tremblay et al. and Moran et al., we used their reported coefficient alphas to
correct for attenuation in both variables. In the case of the Gagné et al. study, we use
the reported correlation between latent variables as the corrected correlation. For
Boiché and Stephan (2014), exact coefficient alphas were not reported, only they ranged
from .69 to .85 for the motivation scales. Thus, we assumed an alpha of .77 for all
variables in their study.
60
Exhibit 5. Correlations of SDT Motivation Types with External Variables
Autonomous Motivation
Controlled Motivation
Intrinsic
Motivation
Extrinsic
Motivation
Integrated
Regulation
Identified
Regulation
Introjected
Regulation
External
Regulation
Tremblay et al. (2009):
Job Satisfaction
.46
.45
.40
.34
.02
Organization
Commitment
.41
.37
.32
.32
.13
Turnover Intention
-.47
-.36
-.35
-.26
-.03
Work Strain
-.06
-.12
-.08
-.01
.10
Gagné et al. (2010):
Job Satisfaction
.58
¤
.53
.27
.13
Affective organizational
commitment
.59
.64
.38
-.18
Turnover Intention
-.26
-.27
-.12
-.03
Well-being
.54
.43
.14
-.09
Psychological Distress
-.48
-.34
-.06
.20
Need for autonomy
.55
.60
.36
.17
Need for competence
.25
.27
.08
.09
Need for relatedness
.51
.52
.07
.01
61
Exhibit 6. Annual Employee Movement Rates (%), U.S. Labor Market
Year
Total
Separations
Quits
Layoffs/
Discharges
Other
Separations
Hires
Unemployment
Rate
2013
42.2
22.8
16.4
3.0
44.3
7.4
2010
40.5
19.2
18.4
2.9
41.7
9.6
2005
50.7
29.0
18.8
3.0
52.7
5.1
Source: Job Openings and Labor Turnover Survey, U.S. Bureau of Labor Statistics
The JOLTS survey design is a stratified random sample of 16,000 nonfarm business and government
establishments. The sample is stratified by ownership, region, industry sector, and establishment
size class. The establishments are drawn from a universe of over 9.1 million establishments
compiled by the Quarterly Census of Employment and Wages (QCEW) program which includes all
employers subject to state unemployment insurance laws and federal agencies subject to the
Unemployment Compensation for Federal Employees program.
62
Exhibit 7. Suggested Future Research Directions
1. Quality of Motivation. Do Different types of motivation (extrinsic, intrinsic; autonomous,
controlled) differentially predict effectiveness outcomes (e.g., performance, creativity)?
2. Integrate the motivation and compensation/PFP literatures. Determine the degree to which
different PFP program design characteristics (e.g., frequency of evaluation/reward, intensity
of performance monitoring/measurement), as well as definition of performance (e.g., results
versus behaviors, individual versus unit/organization level of analysis) influence employee
motivation, perceived autonomy, perceived competence, performance, and creativity.
3. Assess the degree to which the influence of PFP on effectiveness outcomes depends on how
intrinsically motivating the job is.
4. Examine the role of choice. How does PFP influence goal choice (e.g., among tasks that vary
in intrinsic interest). What is the role of PFP, attraction-selection-attrition/sorting in matching
people to jobs and to what degree does that reduce the likelihood of ongoing mis-matches
(and their consequences) between employee motivation preferences and the rewards
attached to their jobs?
5. Be aware of the influence of national differences in context.
63
Notes
1 We will use the terms “pay” and “compensation” interchangeably. Pay for performance (PFP)
refers to any pay program where pay depends on performance. Performance can be measured
using results (e.g., physical output, productivity, profit) and/or behaviors (e.g., typical
performance appraisal dimensions) and it can be measured at the individual and/or aggregate
(e.g., organization) level. (We provide examples of PFP in the section on “Performance.”)
However, when we use the term PFP here, we assume that individual performance plays a
significant role in determining pay. In other literatures we will discuss, terms like performance-
contingent pay are similar, but are often used to describe individual PFP where performance is
a results-based measure. The term extrinsic rewards is also used in other literatures. It
typically is used in a way similar to the performance-contingent pay term.
2CET logic has been used in economics to develop “motivation crowding” theories, under which
providing monetary incentives may not increase overall motivation (and its behavioral
consequences) because the higher extrinsic motivation may diminish (“crowd out”) the intrinsic
motivation. However, it is very important to note that the crowding out literature does not, for
the most part, focus on workplace motivation, performance, and creativity. Indeed, Frey and
Jergen (2001, p. 590) very explicitly say that the crowding-out effect specifically occurs “when a
previously non-monetary relationship is transformed into an explicitly monetary one.” That is
not the focus of our article.
3 The Cerasoli et al. study is much different from the Deci et al. (1999) meta-analysis. Only 6 of
183 effect sizes in Cerasoli et al. are from experiments where intrinsic motivation was
manipulated. The remaining 177 effect sizes are from what they term “correlational” designs.”
In contrast, Deci et al. (p. 635) used only experiments and excluded field studies (i.e., those with
a “correlational design”). Thus, whereas Deci et al. summarized experiments, where subjects
working under incentives versus not working under incentives, were presumed to be equivalent
across studies due to random assignment, the Cerasoli et al. meta-analysis primarily compares
non-equivalent subjects from different studies, some of whom work under incentives, some
who do not..
4 Although not our main focus here, we believe it is useful to point out that recent reviews of
the creativity literature have also become more cautious regarding the long-held view that
intrinsic motivation has consistent positive effects on creativity (Grant & Berry, 2011; Zhou &
Hoever, 2014).
5 See also the over-justification effect (Lepper, Greene, & Nisbett, 1973), based on attribution
theory, which suggests that people who receive extrinsic rewards for performing an interesting
activity attribute the cause of their behavior to the extrinsic reward, thus discounting their
interest in the activity as the cause of their behavior.
6 There was no negative effect of incentives on quality of performance (r = .08, d = .16, n.s.),
Thus, extrinsic rewards were associated with higher quantity with no cost to quality.
7 It seems likely that the fragility of intrinsic motivation under CET and SDT reflects the fact that
much of the empirical research has been on the intrinsic motivation of children. Deci et al.
(1999) observed that children seem more sensitive than adults (college students) to the
controlling aspect of contingent rewards. Likewise, in our subsequent discussion of the effect
64
of choice on intrinsic motivation, we will also see that choice is more strongly positively related
to intrinsic motivation among children than among adults.
8 Introjected regulation is more autonomous than externally regulated extrinsic motivation and
“has been taken in by the person but has not been accepted as his or her own” (Gagné & Deci,
2005, p. 334). Here, behavior is regulated/motivated by internal sources: ego protection, guilt,
and so forth.
9 Ryan and Connell (1989, p. 750) gave the following definitions. “External reasons were those
where behavior is explained by reference to external authority, fear of punishment, or rule
compliance. Introjected reasons were framed in terms of internal, esteem-based pressures to
act, such as avoidance of guilt and shame or concerns about self- and other-approval.
Identifications were captured by reasons involving acting from one's own values or goals, and
typically took the form of "I want." Finally, and where applicable, we included intrinsic reasons
for action where the behavior is done simply for its inherent enjoyment or for fun.”
10See our earlier footnote containing recent quotes from SDT scholars (e.g., Deci & Ryan, 2013),
which seem to continue to take a negative view of extrinsic rewards, much like that found
under CET.
11Among those raising concerns over the years are: Bartol and Locke, 2000; Eisenberger &
Cameron, 1996; Gerhart & Milkovich, 1992; Latham & Locke, 1990; Latham, 2007; Locke &
Latham, 2009; Pinder, 1998; Wiersma, 1992.
12 One can still, at times, in reading the SDT literature, see CET’s negative view of extrinsic
rewards, which continue to be seen as a major potential threat to intrinsic motivation. For
example, Benita, Roth, and Deci (p. 260) in 2014 stated that “Substantial research has shown
that events such as the use of rewards, deadlines, threats, surveillance, and pressuring
language tend to be experienced as controlling and thus to undermine autonomous regulation
resulting in poorer performance and greater ill being (Ryan & Deci, 2000).” Note the other
things “rewards” are grouped with. See also Deci and Ryan (2013) for a number of statements
indicating continued unease with monetary rewards.
13Even when asked directly, employees tend to identify extrinsic rewards as very important. For
example, based on roughly 600 employees surveyed annually from 2004 through 2013 by the
Society for Human Resource Management (SHRM), the 3 (out of 19 job attributes) most often
rated as “very important” over the 10-year period were, in order, job security (60 %), benefits
(59 %), and compensation/pay (59 %), all extrinsic rewards (SHRM, 2014.
