610 Assignment
jsbfg0001
performancerelated.pdf
Nicola Bellé is assistant professor in the
Department of Policy Analysis and Public
Management at Bocconi University in Milan,
Italy. His research focuses on work motiva-
tion, public employee attitudes, job design,
and leadership.
E-mail: [email protected]
230 Public Administration Review • March | April 2015
Public Administration Review,
Vol. 75, Iss. 2, pp. 230–241. © 2015 by
The American Society for Public Administration.
DOI: 10.1111/puar.12313.
Nicola Bellé Bocconi University, Italy
Abstract: Th is article advances our understanding of the eff ects of monetary rewards on public employee performance and of the contingencies that may moderate these eff ects. In a randomized control-group experiment with nurses working at a local health authority in Italy, performance-related pay (PRP) had a larger eff ect on task performance when the rewards were kept secret than when they were disclosed. Th e negative interaction between PRP and visibility was stronger among participants who were exposed to direct contact with a benefi ciary of their eff orts, which heightened their perception of making a positive diff erence in other people’s lives. Th ese results are consistent with theoretical predictions that monetary incentives for activities with a prosocial impact may crowd out employee image motivation. Th ere were no crowding-out eff ects when a symbolic reward was substituted for the monetary incentive.
Practitioner Points • For activities with a prosocial impact, monetary incentives tend to have a larger performance eff ect when
they are secret rather than disclosed. • Nonmonetary rewards may be immune—or at least less prone—to some of the motivational drawbacks
encountered when using monetary incentives for activities with a prosocial impact. • Public organizations and their managers should take full advantage of nonmonetary incentive options in this
time of budget restraints, which make it impossible to off er bonuses that are large enough to be eff ective.
moderate the eff ectiveness of performance-related pay systems” (Perry, Engbers, and Jun 2009, 44), and this review identifi es only occasional PRP successes, which appear to be associated with particular types of public service industries and certain organizational levels. In particular, the few positive studies tend to be concentrated in the medical context and involve lower organizational levels.
Unfortunately, the literature on the eff ectiveness of contingent pay plans in the public sector is sparse and relies primarily on correlational designs, which are well suited for testing theoretical predictions in a broad range of populations but do not perform particularly well with respect to their internal validity (McGrath 1981). Th e absence of sound experimental
research has precluded rigor- ous causal inferences about the eff ects of monetary rewards on public employee motivation and performance and about the contingencies that may moder- ate these eff ects. Filling this gap in the literature seems particu- larly important because of the
Performance-Related Pay and the Crowding Out of Motivation in the Public Sector: A Randomized Field Experiment
“More pay for better performance” has long been the mantra behind the public sector personnel reforms inspired by New Public Management. Since the late 1970s, the vast majority of countries in the Organisation for Economic Co-operation and Development have adopted performance-related pay (PRP) provisions for government employees (Lah and Perry 2008; OECD 2005). Th e use of monetary incentives in the public sector does not seem to be declining; on the contrary, it seems to have enjoyed a recent resurgence in interest and popularity (Bellé and Cantarelli 2014; Bellé and Ongaro 2014; Perry, Engbers, and Jun 2009).
PRP continues to be adopted by public jurisdictions, although research on its eff ectiveness in the public sector is inconclusive (Bellé 2010; Ingraham 1993; Kellough and Lu 1993; Milkovich and Wigdor 1991; Perry 1986). In fact, the most recent compre- hensive review of studies on this topic reports mixed results and suggests that “a variety of contextual factors appear to
Th e absence of sound experimental research has
precluded rigorous causal infer- ences about the eff ects of mone- tary rewards on public employee
motivation and performance.
Performance-Related Pay and the Crowding Out of Motivation in the Public Sector: A Randomized Field Experiment 231
are held to strict transparency requirements regarding their compen- sation policies, whereas private companies that use monetary incen- tives most successfully tend to rely heavily on pay secrecy (Colella et al. 2007). Furthermore, public institutions face budget constraints and public expectations about the responsible stewardship of resources that make it either legally or politically impossible to off er bonuses that are large enough to be eff ective (Miller and Whitford 2007)—as required by reinforcement theory (Skinner 1969) and expectancy theory (Vroom 1964) and as suggested by experimental research (Gneezy and Rustichini 2000). Th e institutional perspective of this second group of studies is shared by a strand of literature that draws on transaction cost economics to explain the uneven intro- duction of incentives that may be observed across public jurisdic- tions. Building on previous work by Miller and Falaschetti (2001), Dahlström and Lapuente (2010) argue that PRP provisions are more likely to be implemented successfully in countries with a clear separation of interests among politicians and senior civil servants. As with owners of private fi rms, the argument goes, politicians may be tempted to renege ex post on promises of incentives and divert resources to ends that better serve their political goals. If the careers of senior civil servants (i.e., those who actually manage the incen- tive system) directly depend on politicians, senior civil servants will derive direct benefi ts from complying with politicians’ wishes and their temptations for opportunistic defection. As a result, without a clear separation between politicians (i.e., “owners”) and senior civil servants (i.e., “managers”), employees will not believe that promises about incentives are credible and will make only minimum eff orts to achieve incentivized goals.
A third group of studies on the shortcomings of PRP in the public sector points to the motivational diff erences between public sector and private sector employees. Th e strand of literature on the unin- tended motivational eff ects of extrinsic rewards falls into this third group. Th ese studies provide two distinct explanations for why PRP may undermine the eff orts and performance of public employees. A fi rst explanation points to a crowding-out eff ect related to intrinsic motivation (Frey and Jegen 2001; Frey and Oberholzer-Gee 1997). Using factorial survey data from 186 master of business admin- istration students, Weibel, Rost, and Osterloh (2010) show that fi nancial incentives produce two opposing eff ects: they increase extrinsic motivation (price eff ect) and crowd out individuals’ intrinsic motivations by threatening their feelings of autonomy, competence, and/or relatedness (Ryan and Deci 2000). According to Weibel and colleagues (2010), the overall performance impact of PRP—which depends on the relative strength of the price eff ect and the crowding-out eff ect—is likely to be weaker in the pub- lic sector than in the private sector, for two main reasons. First, incentives are typically smaller in the public sector; therefore, the price eff ect tends to be smaller. Second, an abundant literature has shown that public sector employees tend to be more intrinsically motivated compared with private sector workers (e.g., Buelens and Van den Broeck 2007; Cacioppe and Mock 1984; Crewson 1997; DiIulio 1994; Georgellis, Iossa, and Tabvuma 2011; Houston 2000; Jurkiewicz and Massey 1997; Perry 1997); therefore, all other things being equal, fi nancial incentives are more likely to crowd out intrinsic motivation in public organizations compared with private companies because there is more intrinsic motivation in the public sector, and more of it may be destroyed (Weibel, Rost, and Osterloh 2010).
widespread and ever-increasing diff usion of PRP schemes in public organizations.
Both policy makers and scholars have recently urged the study of the performance eff ects of PRP in the context of public administra- tion using stronger research methods—particularly fi eld experi- ments—to illuminate causal paths that have long remained unclear (Bellé 2010; Perry, Engbers, and Yun 2009). We answered this call by conducting a randomized fi eld experiment that investigated the eff ects of monetary incentives on the performance of a group of nurses working for public hospitals in Italy. We examined how these eff ects depend on two conditions that play important roles in the public sector: (1) the transparency or observability of individual rewards (Ariely, Bracha, and Meier 2009; Perry, Engbers, and Yun 2009) and (2) employee perceptions of making a positive diff erence in other people’s lives (Grant 2007, 2008a, 2008b; Grant et al. 2007).
Th e emphasis on PRP and the tendency to lump all incentives into a common category have long combined to obscure the diff er- ences between the eff ects of monetary and nonmonetary rewards. Responding to recent calls to investigate this neglected issue through advanced research designs (Grant and Shin 2012), we replicated our experiment, substituting a symbolic reward for the monetary incen- tive. Th is enabled us to observe whether and how symbolic rewards diff er from fi nancial incentives in terms of their interaction with reward visibility and employee perception of prosocial impact. We believe this comparison may signifi cantly contribute to the extant literature on the use of diff erent incentive options in the public sec- tor (Perry, Mesch, and Paarlberg 2006; Perry and Porter 1982).
In the following sections, we fi rst situate our research in the relevant literature and illustrate our hypotheses. We then describe the experiment we conducted to test these hypotheses and conclude with a discussion of our fi ndings and their implications for theory and practice.
Theoretical Background, Research Questions, and Hypotheses In a research synthesis of 57 empirical studies set in the public sector and conducted between 1977 and 2008, Perry, Engbers, and Yun argue that “performance-related pay continues to be adopted but persistently fails to deliver on its promise” (2009, 46). Scholars have provided three main types of explanations for this persistent failure of PRP plans in the public sector, which contrasts with successful outcomes in private industry. A fi rst group of studies points to the poor technical design of contingent pay plans and the inadequacy of the performance management practices that support the incentive systems. According to this view, PRP provisions might be eff ective if they were better designed and supported by better performance management practices (e.g., better performance appraisal systems), which are often inadequate in public organizations (Egger-Peitler, Hammerschmid, and Meyer 2007; Kessler and Purcell 1992; Marsden and Richardson 1994).
A second type of explanation for the frequent failure of PRP in the public sector involves the fundamental institutional characteristics of public organizations, which cannot be attenuated by simply improving performance management practices. Public organizations
232 Public Administration Review • March | April 2015
the desires to be liked by others (external image motivation) and to respect oneself (introjected image motivation).
An individual’s image is aff ected by how prosocial or greedy he or she is considered to be by other people and how greedy the individual considers himself or herself. In an attempt to maximize their image value, indi- viduals are motivated to engage in behavior that appears prosocial and to refrain from behavior that may be perceived as greedy (Bénabou and Tirole 2006). Th erefore,
fi nancial rewards for activities with a prosocial impact are likely to elicit two opposing eff ects on public employees’ extrinsic motiva- tion: individuals are incentivized to work harder to obtain the monetary reward (price eff ect), but they may refrain from doing so because they are concerned about being considered greedy, which would spoil their social image (crowding-out eff ect related to external image) or because they are afraid of feeling guilty and losing self-esteem (crowding-out eff ect related to introjected image motivation).
Th e contradictory eff ects of price and the crowding out of image motivation eff ects, which occur contemporaneously when prosocial activities are fi nancially incentivized, are apparent in a laboratory experiment conducted by Ariely, Bracha, and Meier (2009) on 161 Princeton University undergraduates with respect to perform- ing a task with a prosocial impact. Th e students had to repeatedly click two keys on a computer keyboard for up to fi ve minutes, and the researchers donated funds to charity according to the number of clicks. A random subgroup of participants who were off ered a performance-contingent monetary reward outperformed those who were not off ered a fi nancial incentive. Among the students who received a bonus, a random subsample whose performance and pay were kept secret tended to outperform those whose performance and pay were publicly observable. In other words, visibility nega- tively moderated the performance eff ect of the monetary reward. Th e results of this “click for charity” experiment were replicated in a similarly designed quasi-experiment with 151 Massachusetts Institute of Technology students cycling on a stationary bike for up to 10 minutes, with researchers donating $1 per mile (Ariely, Bracha, and Meier 2009). Both experiments suggest that fi nancial incentives and reward transparency or observability interact negatively with respect to activities with a prosocial impact, which supports the hypothesis that image motivation is subject to being crowded out.
Th is experimental evidence has potential implications for the motivational eff ects of PRP in the public sector because many public sector jobs have a signifi cant prosocial impact (Grant 2008b). However, to our knowledge, research has yet to investigate whether and to what extent the results of the lab experiments by Ariely, Bracha, and Meier (2009) apply to public sector work environments in the real world. To help fi ll this gap, we formulated and tested the following hypothesis using a randomized true fi eld experiment.
Hypothesis 1: Th e visibility of rewards will moderate the eff ect of monetary rewards on job performance such that monetary rewards will have a stronger eff ect when they are secret and a weaker eff ect when they are visible.
It is important to note here that research on diff erences in attitudes and values between public and private workers is inconclusive. For instance, a cross-sectional study of 549 knowl- edge workers employed in large Canadian organizations found no sectoral diff erences in general values, although participants working in public organization tended to value work that contributes to society more than their private sector counterparts (Lyons, Duxbury, and Higgins 2006). In a research synthesis of 14 empirical studies on value diff erences between public and private managers, Boyne argues that “there seems to be strong evidence of the existence of a public service ethos” (2002, 112–13) but cautions that this statis- tical evidence is limited, and its validity is threatened by serious methodological limitations.
A second explanation for why monetary incentives may under- mine the motivations of public employees focuses on crowding out image motivation, that is, an individual’s tendency to be motivated by “the desire to be liked and respected by others and by one’s self ” (Ariely, Bracha, and Meier 2009, 544). Using longitudinal survey data from the British Household Panel Survey, Georgellis, Iossa, and Tabvuma (2011) conclude that higher extrinsic rewards (i.e., higher wages, more job security and fewer working hours) tend to reduce the propensity of British workers to move from the private to the public sector. Th e authors trace this eff ect to reputational concerns: incentives based on more power and other extrinsic rewards may harm the appeal of the prosocial image of public sector jobs, making them less attractive to individuals who want to appear prosocial. Th is conclusion is supported by the results of a study on volunteer fi refi ghters by Carpenter and Myers (2010). Using a multisource observational research design that included a survey of 205 volunteer fi refi ghters in Vermont, the authors found that those who were paid small stipends were more likely to respond to emergency calls than those who were not off ered any monetary rewards, but the positive association between fi nancial incentives and turnout was weaker for those who had greater image concerns. Th ese two studies, by Georgellis and colleagues (2011) and Carpenter and Myers (2010), provide joint support for theoretical predictions that monetary rewards can crowd out image motivation among public employees. However, the lack of true experimental evidence has so far precluded rigorous causal inferences. Our work aims to take a step toward fi lling this gap in literature.
PRP and Image Motivation Th e construct of image motivation falls under the larger umbrella concept of extrinsic motivation, which Ryan and Deci (2000) defi ne as a continuum with four degrees that workers experience as being progressively less controlled by others and more self-determined (Grant and Shin 2012): (1) external, which is triggered by out- side rewards and punishments; (2) introjected, which is based on internal rewards and punishments, such as guilt and self-esteem; (3) identifi ed, which is based on consistency with a person’s system of values; and (4) integrated, which is based on assimilation into a person’s value system. Th e image motivation construct proposed by Ariely, Bracha, and Meier (2009) taps into the external and the introjected degrees of extrinsic motivation because it is triggered by
In an attempt to maximize their image value, individuals are
motivated to engage in behavior that appears prosocial and to
refrain from behavior that may be perceived as greedy.
Performance-Related Pay and the Crowding Out of Motivation in the Public Sector: A Randomized Field Experiment 233
design, and procedures. We then describe the measures we used for our statistical analyses.
Participants, Design, and Procedures We conducted our experiment with 300 nurses attending a man- datory training program at a Local Healthcare Authority (LHA) in Italy. LHAs are part of Italy’s National Health Service (Servizio Sanitario Nazionale or SSN), a publicly funded national health care system originally modeled after Britain’s National Health Service. As of 2012, public health expenditures accounted for 78.2 percent of total health expenditures in Italy (World Bank 2014). As of February 2014, the SSN comprised 140 LHAs (Ministero della Salute 2014). Every LHA is responsible for providing health care in a specifi c area of the country. Nurses working at LHAs are hired through open competitions and have the status of public employees.
At the beginning of 2011, the LHA where we conducted our experi- ment joined an international cooperation project aimed at strength- ening the capacity of the health care system in a former war zone currently facing a humanitarian emergency. Th e LHA contributes to the project by collecting surgical tools and drugs donated by various organizations (e.g., pharmaceutical companies, public and private hospitals, and nongovernmental organizations), entering them into an inventory, controlling the quality of the products, and assem- bling surgical kits ready for shipment to health care practitioners operating in the target area.
Using a random number generator, the participants were randomly assigned to one of 12 balanced groups, each consisting of 25 units (see table 1). We experimentally manipulated the three independent variables.
Reward. The participants were randomly assigned to one of three types of rewards: fi xed pay (100 nurses), fi xed pay plus a performance-contingent monetary reward (100 nurses), or fi xed pay plus a symbolic reward (100 nurses). The nurses in the fi xed-pay group were informed that the four hours spent on the project would be paid according to their normal hourly wage. The participants in the PRP group were promised an incentive in addition to their fi xed pay according to a decreasing payment format: 1 euro for each of the fi rst 25 surgical kits, 50 cents for each of the next 25 surgical kits, and 25 cents for each surgical kit above 50. We designed this decreasing payment schedule based on Ariely, Bracha, and Meier (2009). The participants were informed that the monetary rewards would be given in the form of gift cards. The participants in the symbolic reward group were informed that in addition to their fi xed pay, the top fi ve performers in their group would be awarded a certifi cate of outstanding contribution by the director of the LHA in a ceremony attended by top management. We designed the symbolic reward manipulation based on Mickel and Barron (2008), who suggest that rewards for high performance and accomplish- ments are more likely to increase motivation when they are awarded in a public ceremony by high-profi le fi gures of authority.
Visibility. The participants were randomly assigned to either a disclosed (150 nurses) or a secret (150 nurses) condition. Nurses in the open condition were told that the individual performance of each participant would be displayed on a bulletin board located in
Th e negative interaction between fi nancial incentives and reward vis- ibility found by Ariely, Bracha, and Meier (2009) was not replicated in a similar laboratory experiment conducted by Bamberger and Belogolovsky (2010) with 139 undergraduate students at an Israeli university. Th e main diff erence between the two studies lies in the lack of prosocial impact of the experimental task in Bamberger and Belogolovsky. Th is suggests the need for an investigation into whether the negative two-way interaction between fi nancial incen- tives and reward visibility is stronger for activities that have a higher perception of prosocial impact. Because external image motivations may be crowded out when there are concerns about being suspected of being motivated by a fi nancial reward for a prosocial activity rather than a sincere desire to do good, it seems reasonable to expect that this eff ect would be stronger for activities that have a higher perception of a prosocial impact. Although Ariely and colleagues (2009) did not test for a three-way interaction of fi nancial incentives, reward visibility, and perceived prosocial impact, the graphs included in their article seem to suggest that the two-way interaction between monetary rewards and visibility is stronger when the receiving charity has an unquestionably positive image than when the receiving char- ity has a more controversial reputation. We formulated and tested the following hypothesis to shed more light on this issue.
Hypothesis 2: A perceived prosocial impact will moderate the interaction between monetary rewards and reward vis- ibility such that there will be a stronger (negative) interaction for activities with a higher perception of prosocial impact. Th us, there will be a three-way interaction between monetary rewards, reward visibility, and perceived prosocial impact.
The Effects of Nonmonetary Rewards on Public Employee Performance Scholars have long recognized that money and closely related tangible rewards aff ect an individual’s motivation diff erently than intangible or symbolic rewards, such as positive feedback or other manifestations of social approval (for a review, see Deci, Koestner, and Ryan, 1999). Unfortunately, the emphasis on PRP and the tendency to lump all incentives into a common category have long obscured the importance of intangible rewards among practitioners and academics (Grant and Shin 2012). Th is is particularly evident in the public administration literature, in which the study of the eff ects of diff erent types of incen- tives on employee behavior has received only scant attention (e.g., Perry, Mesch, and Paarlberg 2006; Perry and Porter 1982). As a step toward fi lling this gap, we replicated the experimental design that we used to test hypotheses 1 and 2 by substituting a symbolic reward for the mon- etary incentive.Bénabou and Tirole argue that “esteem-based incentives can adequately replace material rewards and punishments in spheres in which gaining distinction is the dominant reputational concern” (2006, 1672). Compared with fi nancial incentives, which might have a negative reputation associated with greed and money-oriented behavior, symbolic rewards are less prone to the “overjustifi cation eff ect,” in which extrinsic incentives crowd out prosocial behavior (Lepper, Greene, and Nisbett 1973; Pearce 1983; Titmuss 1970; Upton 1973). Th erefore, we expect that the crowding-out eff ects will be smaller for nonmonetary incentives compared with monetary rewards.
Method In this section, we describe the experiment we conducted to test our hypotheses. We begin by explaining the experiment’s participants,
234 Public Administration Review • March | April 2015
Ta b
le 1
E xp
er im
en ta
l I n te
rv en
ti o n s,
C o n tr
o ls
, an
d P
er fo
rm an
ce b
y G
ro u
p
G ro
u p
A B
C D
E F
G H
I G
K L
O b s.
2 5
2 5
2 5
2 5
2 5
2 5
2 5
2 5
2 5
2 5
2 5
2 5
C o
n d
it io
n
M o
n et
ar y
re w
ar d ( PR
P)
X
X X
X
Sy m
b o lic
r ew
ar d
X X
X X
O b
se rv
ab le
r ew
ar d
X
X
X
X
X
X
B en
efi c
ia ry
c o n ta
ct
X
X
X
X
X
X
μ σ
μ σ
μ σ
μ σ
μ σ
μ σ
μ σ
μ σ
μ σ
μ σ
μ σ
μ σ
P re
-t e st
c o
n tr
o ls
A g
e 4 4 .1
6 1 0 .9
2 4 3 .5
2 1 1 .3
8 4 3 .1
6 1
0 .3
7 4
3 .9
6 1
0 .6
2 4
1 .7
6 1
0 .6
6 4
4 .0
8 1
0 .6
9 4
3 .3
6 1
1 .0
9 4
2 .7
2 1 1 .5
0 4 2 .9
6 1 1 .8
5 4 4 .2
0 1 1 .9
1 4 4 .7
6 1 1 .2
3 4 5 .1
6 1 1 .2
9
Pr o
p o rt
io n f
em al
e 0 .7
2 —
0 .6
8 —
0 .7
2 —
0 .8
4 —
0 .8
0 —
0 .7
2 —
0 .8
4 —
0 .7
6 —
0 .8
0 —
0 .7
6 —
0 .7
2 —
0 .6
8 —
Jo b e
xp er
ie n ce
2 1 .8
4 1 1 .6
1 2 0 .6
8 1 1 .8
7 1 9 .9
2 1
1 .8
6 1
7 .9
6 1
1 .6
4 1
9 .4
4 1
1 .0
6 1
8 .4
8 1
1 .4
6 1
9 .2
8 1
1 .0
3 1
8 .6
4 1 1 .6
7 2 0 .1
2 1 1 .2
9 2 0 .8
0 1 0 .9
1 1 9 .8
0 1 0 .5
0 2 0 .4
0 1 0 .1
0
PS M
4 .6
8 0 .6
8 4 .4
8 0 .7
3 4 .7
2 0
.6 9
4 .5
2 0
.7 3
4 .6
4 0
.8 6
4 .5
8 0
.6 5
4 .6
8 0
.8 1
4 .4
8 0 .7
4 4 .6
4 0 .7
2 4 .7
6 0 .6
9 4 .7
2 0 .7
7 4 .7
6 0 .6
7
Se lf -e
ffi c
ac y
5 .0
4 1 .1
0 4 .9
6 0 .9
8 5 .1
2 1
.0 4
5 .0
0 1
.0 1
4 .9
8 0
.9 9
4 .9
4 1
.0 0
4 .9
6 1
.1 0
5 .0
4 1 .0
8 5 .0
8 1 .1
4 4 .9
6 1 .0
4 4 .8
8 0 .9
7 5 .0
4 1 .0
6
C o
n sc
ie n ti o u sn
es s
4 .7
2 0 .6
9 4 .6
8 0 .6
6 4 .7
6 0
.6 0
4 .8
4 0
.6 3
4 .7
6 0
.5 9
4 .6
8 0
.6 8
4 .8
0 0
.6 2
4 .7
6 0 .6
8 4 .6
8 0 .7
1 4 .7
2 0 .8
0 4 .5
6 0 .6
9 4 .6
0 0 .7
0
In tr
in si
c m
o ti va
ti o n
4 .6
4 1 .1
0 4 .8
4 0 .9
8 4 .7
2 1
.0 1
4 .6
8 1
.0 4
4 .7
6 1
.0 0
4 .7
2 1
.0 8
4 .8
0 1
.1 2
4 .7
2 1 .0
8 4 .6
8 0 .9
8 4 .7
6 0 .9
9 4 .8
4 1 .0
4 4 .6
4 1 .1
4
Eq u it y
se n si
ti vi
ty 3 .8
8 1 .0
1 4 .0
4 1 .0
8 3 .7
2 0
.9 8
4 .0
0 1
.0 7
4 .0
8 1
.1 2
3 .9
2 1
.0 3
4 .0
4 1
.0 4
3 .7
6 1 .0
0 3 .9
6 0 .9
7 4 .0
0 1 .2
4 4 .0
8 1 .1
3 3 .8
8 1 .0
5
P o
st -t
e st
c o
n tr
o ls
Ex tr
in si
c re
w ar
d 2 .3
6 1 .0
8 2 .1
2 1 .1
5 2 .5
2 1
.1 0
2 .2
4 1
.0 8
4 .0
4 1
.2 6
3 .7
2 1
.3 8
4 .1
2 1
.1 7
3 .9
6 1 .4
8 3 .3
6 1 .3
9 3 .8
8 1 .5
1 3 .9
2 1 .2
5 3 .4
4 1 .1
0
R ep
u ta
ti o n al
c o n ce
rn 3 .1
2 1 .3
9 4 .9
6 1 .5
1 2 .9
6 1
.2 5
5 .1
2 1
.4 8
3 .2
0 1
.1 0
4 .8
8 1
.1 0
2 .9
2 1
.2 6
5 .0
0 1 .1
4 3 .0
0 1 .0
8 4 .9
2 1 .1
5 3 .0
4 1 .1
7 5 .1
6 1 .2
5
Pe rc
ei ve
d s
o ci
al im
p ac
t 4 .0
4 1 .2
6 4 .2
0 1 .5
1 4 .9
6 1
.1 6
5 .2
0 1
.1 8
3 .9
6 1
.2 7
4 .0
8 1
.3 9
5 .0
4 1
.1 5
4 .9
2 1 .1
7 3 .9
2 1 .1
0 4 .0
0 1 .1
5 5 .0
0 1 .1
4 5 .1
2 1 .1
2
P e rf
o rm
a n
ce
# S
u rg
ic al
k it s
3 8 .8
4 1 0 .2
2 5 3 .8
0 8 .2
2 4 5 .2
0 1
1 .6
0 6
0 .8
0 1
1 .3
0 6
4 .9
2 9
.1 2
6 9
.4 8
1 1
.0 8
6 1
.8 8
1 3
.0 2
5 3
.3 6
1 1 .3
8 4 6 .3
6 1 5 .2
7 6 1 .2
8 1 2 .1
4 5 4 .6
4 1 0 .4
6 6 5 .6
4 1 2 .1
8
Performance-Related Pay and the Crowding Out of Motivation in the Public Sector: A Randomized Field Experiment 235
Controls. In addition to questions regarding age, gender, and job experience, the pre-experiment questionnaire featured a widely used, fi ve-item version of Perry’s (1996) original scale to measure motivations for public service (Alonso and Lewis 2001; Brewer and Selden 2000; Kim 2005; Pandey, Wright and Moynihan 2008; Wright, Moynihan, and Pandey 2012; Wright and Pandey 2008); three items taken from a scale of effort-performance expectancy to measure self-effi cacy (Sims, Szilagyi, and McKemey 1976; Wright 2007); four items developed by Donnellan et al. (2006) to measure conscientiousness; three items adapted from self-regulation scales developed by Ryan and Connell (1989) to measure the intrinsic motivation of participants with respect to their jobs; and two items taken from a scale of equity sensitivity (Sauley and Bedeian 2000). We included the last control in light of experimental evidence showing that the performance effects of pay secrecy may be moderated by individual tolerance for inequity (Bamberger and Belogolovsky 2010).
Th e post-experiment questionnaire featured six questions intended to assess the eff ects of our experimental manipulations (see appen- dix). It might have been more logical to measure the eff ect of our interventions not after the intervention but before performing the task. Unfortunately, this was not feasible because of logistical and practical constraints. Moreover, although we controlled for reputa- tional concern, we did not measure image motivation, nor did we use a manipulation check of visibility. All of these represent poten- tial limitations of our study.
We adapted two items from Wright (2007) to ascertain whether participants in the PRP condition had higher perceptions that better performance was rewarded compared with their colleagues in the fi xed-pay condition. We used two items adapted from the public self-consciousness scale originally developed by Fenigstein, Scheier, and Buss (1975) to assess whether participants in the open condi- tion were more concerned about what their colleagues might think about their performance while performing the experimental task compared with nurses in the private condition. Finally, we used two items adapted from Grant (2008a) to measure whether nurses who had the opportunity to meet the patient from the target area understood that their eff orts would have a greater prosocial impact than those who did not meet the benefi ciary.
Results Th e means and standard deviations for the key variables by condi- tion are displayed in table 1. As a result of randomization, the 12 groups did not diff er at the .05 level with respect to the participants’ age, gender, years of nursing experience, public service motivation, self-effi cacy, conscientiousness, intrinsic motivation, and toler- ance for inequity. A series of two-sample t-tests indicated that the perception that higher performance was rewarded was lower (p < .001) among participants in the fi xed-pay condition (M = 2.31, SD = 1.10) compared with their colleagues who were off ered the monetary reward (M = 3.96, SD = 1.32) or the symbolic award (M = 3.65, SD = 1.31) in addition to their normal hourly wage. Th e diff erence between nurses in the PRP and in the symbolic reward conditions was not signifi cant at the .05 level. Th e participants in the open condition reported that they were more concerned about what their colleagues might think about their performance while performing the experimental task (M = 5.01, SD = 1.27) compared
the nurses’ staff room. Participants in the secret condition were told that they would be notifi ed of their own level of individual performance but that this information would be anonymous and not publicly disclosed.
Benefi ciary contact. Recent experimental research on relational job design has shown that public employees who meet the benefi ciaries of their efforts may have a greatly heightened awareness of making a positive difference in other people’s lives (Bellé 2013, 2014; Grant 2007, 2008a, 2012; Grant et al. 2007). The participants were randomly assigned to either a benefi ciary-contact (150 nurses) or no-benefi ciary-contact (150 nurses) condition. Participants in the benefi ciary-contact condition had the opportunity to meet a person from the target area who had benefi ted from the surgical kits after being injured in the past by an antipersonnel mine and who had later joined the project staff. During the benefi ciary’s visits, which lasted approximately 15 minutes, he explained how his life was saved by surgical tools similar to those the participants would assemble. The participants in the no-benefi ciary-contact condition did not meet the patient from the target area.
Our experimental manipulations resulted in a between-subjects 3 (reward: none, fi nancial, symbolic) x 2 (visibility: yes, no) x 2 (ben- efi ciary contact: yes, no) design.
Each of the 12 groups attended its own separate session: four sessions in February 2011 for the groups in the fi xed-pay condition, four ses- sions in November 2011 for the groups in the symbolic reward con- dition, and four sessions in August 2012 for the groups in the PRP condition. All sessions were led by the same training coordinator and an assistant, who were blind to the specifi c research hypotheses. Th e 12 sessions were identical except for the specifi c experimental manipulations. In particular, in all sessions, the nurses watched a short video that provided basic information about the project’s aims and instructions on how to assemble the surgical kits. Th e nurses had exactly the same amount of time (three hours) to actually perform the assigned task. Participants in all groups answered two short questionnaires: a pre-experiment questionnaire at the beginning of their shift—after the training coordinator had distributed consent forms and reassured them that their responses would be handled confi dentially and would only be disclosed in aggregate form with- out any identifying personal information—and a post-experiment questionnaire at the end of their shift. Th e pre-test survey featured questions to measure the participants’ standard demographics and baseline attitudes that were relevant to our research question. Th e post-test questionnaire featured items to test whether our experimen- tal manipulations had produced the intended eff ects.
Measures Th e appendix reports the variables and the measures we used in the study.
Performance. We measured performance as the number of surgical kits that each participant assembled correctly during his or her three-hour shift. This metric was meant to capture the participants’ effort and persistence (e.g., Blumberg and Pringle 1982; Gneezy and Rustichini 2000; Grant 2008a; Grant et al. 2007; Schmidt and Hunter 1983) and their ability to maintain their attention and accuracy while performing their job (e.g., Brewer and Brewer 2011).
236 Public Administration Review • March | April 2015
nurses in the open condition (+4.12 surgical kits, p = .095). Th us, introducing PRP nullifi ed the performance eff ect of visibility that appeared in the fi xed-pay condition.
Th e signifi cant three-way interaction terms in tables 2a and 2b indicate that the negative interaction between PRP and visibility was stronger for participants who were exposed to contact with the project benefi ciary compared with nurses who did not meet the patient from the target area (p = .026). Our data suggest that the perceived prosocial impact moderated the interaction between monetary rewards and reward visibility such that there was a stronger (negative) interaction for activities with a higher perception of prosocial impact, which supports hypothesis 2. Figure 2 depicts the three-way interaction of PRP, visibility, and benefi ciary con- tact on job performance. Th e two solid lines represent participants who were introduced to the benefi ciary (higher perceived prosocial impact), whereas the two dotted lines represent participants who did not meet the patient from the target area (lower perceived prosocial impact). We note that the divergence between the line indicating participants in the open condition (hollow markers) and the line representing nurses in the secret condition (solid markers) is greater for participants who met the benefi ciary.
In addition to the results that are directly related to our hypotheses, the ANOVA analyses reported in tables 2a and 2b indicate an over- all negative interaction of PRP and benefi ciary contact, F(1,199) = 28.15, p < .001. Th us, averaging the two levels of visibility (open/ secret), the performance eff ect of incentive pay decreased as per- ceived prosocial impact increased. Because of the negative interac- tion between PRP and benefi ciary contact, meeting the benefi ciary of their eff orts caused a positive performance eff ect for nurses in the fi xed-pay condition (+6.68 surgical kits, p = .011) but had a det- rimental eff ect when there were monetary rewards (–9.58 surgical kits, p < .001).
Table 2b shows that the interaction between PRP and benefi ci- ary contact was negative for both levels of visibility. Although the
with nurses in the private condition (M = 3.04, SD = 1.21, p < .001). Nurses who had the opportunity to meet the patient from the target area perceived a greater prosocial impact of their eff orts (M = 5.04, SD = 1.15) than those who did not meet the benefi ciary (M = 4.03, SD = 1.28, p < .001). Th ese results indicate that our interven- tions produced their intended eff ects.
To explore how monetary rewards, visibility, and benefi ciary contact interacted in determining the nurses’ performance, we conducted a three-way analysis of variance (ANOVA) on data from groups A–H (i.e., those in the fi xed-pay and PRP conditions). Th e results of the three-way ANOVA and coeffi cient estimates of the underlying ordi- nary least squares (OLS) regression model are displayed in tables 2a and 2b, respectively. Th e signifi cant interaction of PRP and visibility in tables 2a and 2b indicates that, averaging across the two levels of benefi ciary contact (yes/no), monetary rewards were less eff ective among participants whose performance and pay were observable by their colleagues compared with nurses in the secret condition, F(1,199) = 31.72, p < .001. Th is fi nding supports hypothesis 1. Th e overall negative interaction between PRP and visibility appears as the crossing of the two lines in fi gure 1.
Figure 1 shows that disclosing performance and pay information had a surprisingly strong eff ect on nurses in the fi xed-pay condition (+15.28 surgical kits, p < .001) but no signifi cant eff ect on those in the monetary reward condition (p = .433). Although PRP greatly enhanced performance in the secret condition (+21.38 surgical kits, p < .001), monetary rewards were only marginally eff ective for
Table 2b OLS Regression Underlying the Three-Way ANOVA for Groups A–H
IVs Coef. SE t p β
PRP 26.08 2.74 9.52 .000 0.91*** Benefi ciary contact 6.36 3.09 2.06 .041 0.22* Visibility 14.96 2.62 5.70 .000 0.52*** PRP * Benefi ciary –9.40 4.44 –2.12 .035 –0.28* PRP * Visibility –10.40 3.89 –2.67 .008 –0.31** Benefi ciary * Visibility 0.64 4.17 0.15 .878 0.02 PRP * Benefi ciary * Visibility –13.72 6.13 –2.24 .026 –0.32* Constant 38.84 2.04 19.01 .000 F(7,192) = 23.69, p < .0001 Root MSE = 10.84 N = 200 R2 = 0.4513
*p < .05; **p < .01; ***p < .001.
Table 2a Three-Way ANOVA for Groups A–H
Source Partial SS df MS F p
Model 18544.915 7 2649.274 22.56 .000 PRP 8128.125 1 8128.125 69.23 .000*** Benefi ciary contact 105.125 1 105.125 0.90 .345 Visibility 2211.125 1 2211.125 18.83 .000*** PRP * Benefi ciary 3304.845 1 3304.845 28.15 .000*** PRP * Visibility 3723.845 1 3723.845 31.72 .000*** Benefi ciary * Visibility 483.605 1 483.605 4.12 .044* PRP * Benefi ciary * Visibility 588.245 1 588.245 5.01 .026* Residual 22543.840 192 117.416 Total 41088.755 199 206.476 Root MSE 10.84 N = 200 R2 = 0.4513 Adjusted R2 = 0.4313
*p < .05; **p < .01; ***p < .001. 30
35
40
45
50
55
60
65
70
75
80
Fixed-Pay Fixed-Pay +
PRP
N . S
u rg
ic al
K it
s
Secret
Open
Figure 1 Two-Way Interaction of Monetary Reward (PRP) and Reward Visibility on Job Performance
Performance-Related Pay and the Crowding Out of Motivation in the Public Sector: A Randomized Field Experiment 237
analysis on data from nurses belonging only to groups A–D and I–N, the groups that were off ered either fi xed pay or fi xed pay plus a symbolic award. Table 3a shows the results of this three-way ANOVA, and table 3b reports the coeffi cient estimates of the under- lying OLS regression model. Unlike the PRP manipulation, there was no overall interaction between the symbolic award and visibility on job performance, F(1,199) = 0.500, p = .480. Th e interac- tion between the symbolic reward manipulation and visibility was similarly insignifi cant at both levels of perceived prosocial impact. Unlike the PRP manipulation, no overall interaction was detected between the symbolic award and benefi ciary contact, F(1,199) = 0.010, p = .913. Th e interaction between the symbolic reward and benefi ciary contact was similarly insignifi cant at both levels of visibility.
Discussion Th is article extends knowledge about the performance eff ects of PRP plans in public institutions and the factors that may moderate these eff ects. In a fi eld experiment with nurses working at public hospitals in Italy, PRP had a larger eff ect on task performance when the rewards were secret than when they were observable. Th e negative interaction between
monetary rewards and visibility was more pronounced among par- ticipants who had been introduced to a benefi ciary of their eff orts
negative interaction between PRP and benefi ciary contact in the open condition may be caused by the crowding out of both external and introjected image motivation, only the latter can be aff ected when rewards are kept secret (lines with solid markers in fi gure 2). Although this explanation seems consistent with our theoretical predictions, our data do not allow any causality claim because we did not directly measure image motivation.
Th e signifi cant three-way interaction terms in tables 2a and 2b indicate that the interaction of PRP and benefi ciary contact was stronger in the open condition than in the secret condition. Th is is represented graphically in fi gure 2, in which divergence between the solid line indicating benefi ciary contact and the dotted line representing no benefi ciary contact is greater for participants in the open condition (hollow markers) than for participants in the secret condition (solid markers).
Table 2a also reports the signifi cant main eff ects of PRP, F(1,199) = 69.23, p < .001, and visibility, F(1,199) = 18.83, p < .001, which indicates that both interventions were associated with an increase in performance when the levels of the other conditions that were experimentally manipulated were averaged. Th ere was no signifi cant main eff ect of ben- efi ciary contact, F(1,199) = 0.90, p = .345. Moreover, table 2a shows a signifi cant interac- tion between benefi ciary contact and visibility when the two levels of monetary reward were averaged, F(1,199) = 4.12, p = .044.
To investigate whether and how the eff ects we observed in the monetary reward manipulation changed when a symbolic reward was off ered, we repeated our three-way ANOVA
Table 3a Three-Way ANOVA for Groups A–D and I–N
Source Partial SS df MS F p
Model 14927.840 7 2132.549 15.910 .000 Symbolic reward 2679.120 1 2679.120 19.980 .000*** Benefi ciary contact 2112.500 1 2112.500 15.760 .000*** Visibility 9968.720 1 9968.720 74.350 .000*** Symbolic * Benefi ciary 1.620 1 1.620 0.010 .913 Symbolic * Visibility 67.280 1 67.280 0.500 .480 Benefi ciary * Visibility 33.620 1 33.620 0.250 .617 Symbolic * Benefi ciary *
Visibility 64.980 1 64.980 0.480 .487 Residual 25741.680 192 134.071 Total 40669.520 199 204.369 Root MSE = 11.5789 N = 200 R2 = 0.3671 Adjusted R2 = 0.3440
*p < .05; **p < .01; ***p < .001.
Table 3b OLS Regression Underlying the Three-Way ANOVA for Groups A–D and I–N
IVs Coef. SE t p β
Symbolic reward 7.52 3.67 2.05 .042 0.26* Benefi ciary contact 6.36 3.09 2.06 .041 0.22* Visibility 14.96 2.62 5.70 .000 0.52*** Symbolic * Benefi ciary 1.92 4.82 0.40 .691 0.06 Symbolic * Visibility –0.04 4.70 –0.01 .993 0.00 Benefi ciary * Visibility 0.64 4.17 0.15 .878 0.02 Symbolic * Benefi ciary * Visibility –4.56 6.55 –0.70 .487 –0.11 Constant 38.84 2.04 19.01 .000 . F(7,192) = 16.48, p < .0001 Root MSE = 11.579 N = 200 R2 = 0.3671
*p < .05; **p < .01; ***p < .001.
Th e negative interaction between monetary rewards and visibility was more pronounced
among participants who had been introduced to a benefi ciary
of their eff orts.
Note: All the slopes of the lines are signifi cant at the .001 level.
Figure 2 Three-Way Interaction of Monetary Reward (PRP), Reward Visibility, and Benefi ciary Contact on Job Performance
30
35
40
45
50
55
60
65
70
75
80
Fixed-Pay Fixed-Pay +
PRP
N . S
u rg
ic al
K it
s
(1) Beneficiary, Open
(2) Beneficiary, Secret
(3) No Beneficiary, Open
(4) No Beneficiary, Secret
238 Public Administration Review • March | April 2015
may be immune—or at least less prone—to some of the motivational drawbacks encoun- tered when using fi nancial incentives for activities with a prosocial impact. Th is result encourages both academics and practitioners to reevaluate the importance of nonmonetary rewards, which has been obscured by the emphasis on PRP. Understanding how public organizations and their managers can take full advantage of nonmonetary incentive options seems imperative in this time of budget
restraints, which makes impossible off ering bonuses that are large enough to be eff ective (Miller and Whitford 2007)—as required by reinforcement theory (Skinner 1969) and expectancy theory (Vroom 1964).
Our article also contributes to the literature on relational job design (Grant 2007) and, more specifi cally, to the research on the eff ects of contact with benefi ciaries on employee performance (Grant 2008a, 2008b; Grant et al. 2007). Our contribution to this strand of literature lies in identifying a boundary condition of the benefi ci- ary eff ect. In fact, our fi ndings suggest that benefi ciary contact has a positive impact on employee performance in the absence of PRP provisions, but not when pay-for-performance schemes are in place.
Limitations Our fi ndings should be interpreted in light of several limitations. In particular, the unique professional nature of nursing and the use of a convenience sample reduce the external validity of our results. One might think that nurses are more inclined to do good than the average public employee, such that the crowding-out eff ects we observed in our experiment would have been weaker or absent in other categories of public sector employees who are less proso- cially oriented. As noted by an anonymous reviewer, our fi ndings may be valid for any helping profession irrespective of its public or private nature. External validity is not a strength of our research design, which aimed to maximize internal validity through random assignment. Th e lack of random sampling is an inherent limitation of most small-scale fi eld experiments such as ours, whose strength lies in their suitability for establishing causality. On this point, we must agree with Wright and Grant, who observe that “the choice of research design refl ects inherent trade-off s between the ability to make causal statements, the ability to generalize those statements to other settings, and the ability of a broader audience to accept and apply them (McGrath 1981). While each attribute is desirable, at best, any single research design can only maximize two of these criteria while falling short on the third” (2010, 692).
Regarding external validity, the use of temporary and experimen- tally induced manipulations of the conditions under study does not permit the automatic generalization of our results to more endur- ing and naturally occurring variations in these conditions. Future research might triangulate our fi ndings using nonexperimental designs, such as longitudinal studies and case studies. Although inferior to other experiments in terms of internal validity (Shadish, Cook, and Campbell, 2002), observational designs may be superior in terms of external validity because they deal “with intact groups and thus [do] not disrupt the existing research setting” (Dimitrov and Rumrill 2003, 160).
to heighten their awareness that their work made a positive diff erence in other people’s lives. No such interaction was observed when we replicated the experiment using a symbolic reward instead of monetary incentives.
Contributions Th e main contribution of this article consists of developing experimental evidence with respect to the ongoing debate about the eff ects of PRP on the motivation and performance of public employees. Although there is abundant circumstantial evidence that PRP has generally fallen short of expectations for improved performance with respect to public sector employees (e.g., Ingraham 1993; Kellough and Lu 1993; Milkovich and Wigdor 1991; Pearce, Stevenson, and Perry 1985; Perry 1986), reliance on correlational designs has thus far precluded rigorous causal infer- ences about the reasons for this poor performance. Explanations for the potential drawbacks of monetary incentives in the public sector (Perry, Engbers, and Jun 2009) include the fi ndings of previous research that indicate a crowding-out eff ect on both the intrinsic motivation (e.g., Weibel, Rost, and Osterloh 2010) and the image motivation (e.g., Carpenter and Myers 2010; Georgellis, Iossa, and Tabvuma 2011) of public employees. We have contributed to this strand of research by adding novel experimental evidence that mon- etary rewards may crowd out employees’ motivation for prosocial behavior (Bénabou and Tirole 2006). Our fi ndings are particularly valuable because they replicate the results of previous laboratory experiments by Ariely, Bracha, and Meier (2009) internationally and with real public employees performing a task that was part of their ordinary job.
A second contribution of our article is to the still limited scholarly research on the eff ects of pay secrecy on employee performance and behavior (Bamberger and Belogolovsky 2010; Colella et al. 2007; Hartmann and Slapničar 2012). Our contribution to this nascent strand of literature is twofold. First, to our knowledge, we provide for the fi rst time fi eld experimental evidence that fi nancial incen- tives tend to be less eff ective when they are disclosed rather than secret. Second, our data show that the negative interaction between PRP and visibility is more pronounced in the case of activities with higher perceived prosocial impact. Th ese fi ndings suggest caution in the use of PRP provisions in public organizations, which are held to strict transparency requirements regarding their compensation poli- cies and often engage in activities with relevant prosocial impact. Our results seem to corroborate previous research that traces the persistent failure of PRP in the public sector, which contrasts with successful outcomes in private industry (Colella et al. 2007), back to fundamental institutional characteristics of public organizations that cannot be attenuated by simply improving performance manage- ment practices (Perry, Engbers, and Yun 2009).
Another contribution of our article is that it sheds light on the dif- ferences between monetary and nonmonetary rewards in terms of their eff ects on the motivation and performance of public employ- ees. Responding to a recent call to address this long-neglected issue (Grant and Shin 2012), we explored whether and how the experi- mental eff ects changed when a symbolic award was substituted for fi nancial incentives. Our data suggest that nonmonetary rewards
Our data suggest that non- monetary rewards may be immune—or at least less
prone—to some of the motiva- tional drawbacks encountered when using fi nancial incentives for activities with a prosocial
impact.
Performance-Related Pay and the Crowding Out of Motivation in the Public Sector: A Randomized Field Experiment 239
Bénabou, Roland, and Jean Tirole. 2006. Incentives and Prosocial Behavior. American Economic Review 96(5): 1652–78.
Blumberg, Melvin, and Charles D. Pringle. 1982. Th e Missing Opportunity in Organizational Research: Some Implications for a Th eory of Work Performance. Academy of Management Review 7(4): 560–69.
Boyne, George A. 2002. Public and Private Management: What’s the Diff erence? Journal of Management Studies 39(1): 97–122.
Brewer, Gene A., and Gene A. Brewer, Jr. 2011. Parsing Public/Private Diff erences in Work Motivation and Performance: An Experimental Study. Special issue, Journal of Public Administration Research and Th eory 21: i347–62.
Brewer, Gene A., and Sally Coleman Selden. 1998. Whistle Blowers in the Federal Civil Service: New Evidence of the Public Service Ethic. Journal of Public Administration Research and Th eory 8(3): 413–39.
Buelens, Marc, and Herman Van den Broeck. 2007. An Analysis of Diff erences in Work Motivation between Public and Private Sector Organizations. Public Administration Review 67(1): 65–74.
Cacioppe, Ron, and Philip Mock. 1984. A Comparison of the Quality of Work Experience in Government and Private Organizations. Human Relations 37(11): 923–40.
Carpenter, Jeff rey, and Caitlin Knowles Myers. 2010. Why Volunteer? Evidence on the Role of Altruism, Image, and Incentives. Journal of Public Economics 94(11–12): 911–20.
Colella, Adrienne, Ramona L. Paetzold, Asghar Zardkoohi, and Michael J. Wesson. 2007. Exposing Pay Secrecy. Academy of Management Review 32(1): 55–71.
Crewson, Philip E. 1997. Public-Service Motivation: Building Empirical Evidence of Incidence and Eff ect. Journal of Public Administration Research and Th eory 7(4): 499–518.
Dahlström, Carl, and Victor Lapuente. 2010. Explaining Cross-Country Diff erences in Performance-Related Pay in the Public Sector. Journal of Public Administration Research and Th eory 20(3): 577–600.
Deci, Edward L., Richard Koestner, and Richard M. Ryan. 1999. A Meta-Analytic Review of Experiments Examining the Eff ects of Extrinsic Rewards on Intrinsic Motivation. Psychological Bulletin 125(6): 627–68.
DiIulio, John D., Jr. 1994. Principled Agents: Th e Cultural Bases of Behavior in a Federal Government Bureaucracy. Journal of Public Administration Research and Th eory 4(3): 277–318.
Dimitrov, Dimiter M., and Phillip D. Rumrill, Jr. 2003. Pretest-Posttest Designs and Measurement of Change. Work: A Journal of Prevention, Assessment and Rehabilitation 20(2): 159–65.
Donnellan, M. Brent, Frederick L. Oswald, Brendan M. Baird, and Richard E. Lucas. 2006. Th e Mini-IPIP Scales: Tiny-yet-Eff ective Measures of the Big Five Factors of Personality. Psychological Assessment 18(2): 192–203.
Egger-Peitler, Isabell, Gerhard Hammerschmid, and Renate Meyer. 2007. Motivation, Identifi cation, and Incentive Preferences as Issues for Modernization and HR Strategies in Local Government—First Evidence from Austria. Paper presented at the Annual Conference of the European Group of Public Administration, Madrid, Spain, September 19–22.
Fenigstein, Allan, Michael Scheier, and Arnold Buss. 1975. Public and Private Self-Consciousness: Assessment and Th eory. Journal of Consulting and Clinical Psychology 43(4): 522–27.
Frey, Bruno S., and Reto Jegen. 2001. Motivation Crowding Th eory. Journal of Economic Surveys 15(5): 589–611.
Frey, Bruno S., and Felix Oberholzer-Gee. 1997. Th e Cost of Price Incentives: An Empirical Analysis of Motivation Crowding-Out. American Economic Review 87(4): 746–55.
Georgellis, Yannis, Elisabetta Iossa, and Vurain Tabvuma. 2011. Crowding Out Intrinsic Motivation in the Public Sector. Journal of Public Administration Research and Th eory 21(3): 473–93.
Future Directions Our experimental results point toward promising directions for future research. In particular, our data suggest deeper investigations of the positive eff ects of individual performance visibility and ben- efi ciary contact on the performance of participants in the fi xed-pay condition. Exploring further how organizations and their managers can use these interventions to enhance their employees’ motivation may be both theoretically and practically relevant. Finding solutions to compensate for the absence of PRP seems particularly important for public institutions because they often face budget constraints and public expectations about the responsible stewardship of resources that make it impossible to off er fi nancial incentives that are large enough as theoretically required (Gneezy and Rustichini 2000; Skinner 1969; Vroom 1964) to be eff ective (Miller and Whitford 2007).
Future research should also explore the motivational mechanisms that mediate and/or moderate the performance eff ects of visibility or transparency and contact with benefi ciaries. With respect to the latter, a recently published study has shown that employees’ public service motivation may act as both a moderator and a mediator. In a randomized control-group experiment with nurses, meeting the benefi ciary of their eff orts had a stronger positive eff ect on those participants who entered the experiment with a stronger baseline motivation for public service. Moreover, benefi ciary contact caused an increase in the public service motivation that partially mediated the positive eff ects on job performance (Bellé 2013). Future stud- ies should investigate how employee motivations mediate and/or moderate the performance eff ects of visibility/transparency using a similar research design. In general, substantive work is required that contributes additional empirical evidence to this nascent and highly promising stream of research.
References Alonso, Pablo, and Gregory B. Lewis. 2001. Public Service Motivation and Job
Performance: Evidence from the Federal Sector. American Review of Public Administration 31(4): 363–80.
Ariely, Dan, Anat Bracha, and Stephan Meier. 2009. Doing Good or Doing Well? Image Motivation and Monetary Incentives in Behaving Prosocially. American Economic Review 99(1): 544–55.
Bamberger, Peter, and Elena Belogolovsky. 2010. Th e Impact of Pay Secrecy on Individual Task Performance. Personnel Psychology 63(4): 965–96.
Bellé, Nicola. 2010. Così Fan Tutte? Adoption and Rejection of Performance-Related Pay in Italian Municipalities: A Cross-Sector Test of Isomorphism. Review of Public Personnel Administration 30(2): 166–88.
———. 2013. Experimental Evidence on the Relationship between Public Service Motivation and Job Performance. Public Administration Review 73(1): 143–53.
———. 2014. Leading to Make a Diff erence: A Field Experiment on the Performance Eff ects of Transformational Leadership, Perceived Social Impact, and Public Service Motivation. Journal of Public Administration Research and Th eory 24(1): 109–36.
Bellé, Nicola, and Paola Cantarelli. 2014. Monetary Incentives, Motivation, and Job Eff ort in the Public Sector: An Experimental Study with Italian Government Executives. Review of Public Personnel Administration. Published electronically on January 24. doi:0734371X13520460.
Bellé, Nicola, and Edoardo Ongaro. 2014. NPM, Administrative Reforms and Public Service Motivation: Improving the Dialogue between Research Agendas. International Review of Administrative Sciences 80(2): 382–400.
240 Public Administration Review • March | April 2015
Gneezy, Uri, and Aldo Rustichini. 2000. Pay Enough or Don’t Pay at All. Quarterly Journal of Economics 115(3): 791–810.
Grant, Adam M. 2007. Relational Job Design and the Motivation to Make a Prosocial Diff erence. Academy of Management Review 32(2): 393–417.
———. 2008a. Th e Signifi cance of Task Signifi cance: Job Performance Eff ects, Relational Mechanisms, and Boundary Conditions. Journal of Applied Psychology 93(1): 108–12.
———. 2008b. Employees without a Cause: Th e Motivational Eff ects of Prosocial Impact in Public Service. International Public Management Journal 11(1): 48–66.
———. 2012. Leading with Meaning: Benefi ciary Contact, Prosocial Impact, and the Performance Eff ects of Transformational Leadership. Academy of Management Journal 55(2): 458–76.
Grant, Adam M., Elizabeth M. Campbell, Grace Chen, Keenan Cottone, David Lapedis, and Karen Lee. 2007. Impact and the Art of Motivation Maintenance: Th e Eff ects of Contact with Benefi ciaries on Persistence Behavior. Organizational Behavior and Human Decision Processes 103(1): 53–67.
Grant, Adam M., and Jihae Shin. 2012. Work Motivation: Directing, Energizing, and Maintaining Eff ort (and Research). In Th e Oxford Handbook of Human Motivation, edited by Richard M. Ryan, 505–19. New York: Oxford University Press.
Hartmann, Frank, and Sergeja Slapničar. 2012. Pay Fairness and Intrinsic Motivation: Th e Role of Pay Transparency. International Journal of Human Resource Management 23(10): 4283–4300.
Houston, David J. 2000. Public-Service Motivation: A Multivariate Test. Journal of Public Administration Research and Th eory 10(4): 713–28.
Ingraham, Patricia W. 1993. Of Pigs in Pokes and Policy Diff usion: Another Look at Pay-for-Performance. Public Administration Review 53(4): 348–56.
Jurkiewicz, Carole L., and Tom K. Massey, Jr. 1997. What Motivates Municipal Employees: A Comparison Study of Supervisory vs. Non-Supervisory Personnel. Public Personnel Management 26(3): 367–77.
Kellough, J. Edward, and Haoran Lu. 1993. Th e Paradox of Merit Pay in the Public Sector: Persistence of a Problematic Procedure. Review of Public Personnel Administration 13(2): 45–64.
Kessler, Ian, and John Purcell. 1992. Performance Related Pay: Objectives and Application. Human Resource Management Journal 2(3): 16–33.
Kim, Sangmook. 2005. Individual-Level Factors and Organizational Performance in Government Organizations. Journal of Public Administration Research and Th eory 15(2): 245–61.
Lah, T. J., and James L. Perry. 2008. Th e Diff usion of the Civil Service Reform Act of 1978 in OECD Countries: A Tale of Two Paths to Reform. Review of Public Personnel Administration 28(3): 282–99.
Lepper, Mark, David Greene, and Richard Nisbett. 1973. Undermining Children’s Interest with Extrinsic Rewards: A Test of the “Overjustifi cation Hypothesis.” Journal of Personality and Social Psychology 28(1): 129–37.
Lyons, Sean T., Linda E. Duxbury, and Christopher A. Higgins. 2006. A Comparison of the Values and Commitment of Private Sector, Public Sector, and Parapublic Sector Employees. Public Administration Review 66(4): 605–18.
Marsden, David, and Ray Richardson. 1994. Performing for Pay? Th e Eff ects of “Merit Pay” on Motivation in a Public Service. British Journal of Industrial Relations 32(2): 243–61.
McGrath, Joseph E. 1981. Dilemmatics: Th e Study of Research Choices and Dilemmas. American Behavioral Scientist 25(2): 179–210.
Mickel, Amy E., and Lisa A. Barron. 2008. Getting “More Bang for the Buck”: Symbolic Value of Monetary Rewards in Organizations. Journal of Management Inquiry 17(4): 329–38.
Milkovich, George T., and Alexandra K. Wigdor. 1991. Pay for Performance: Evaluating Performance Appraisal and Merit Pay. Washington, DC: National Academy Press.
Miller, Gary J., and Dino Falaschetti. 2001. Constraining Leviathan: Moral Hazard and Credible Commitment in Institutional Design. Journal of Th eoretical Politics 13(4): 389–411.
Miller, Gary J., and Andrew B. Whitford. 2007. Th e Principal’s Moral Hazard: Constraints on the Use of Incentives in Hierarchy. Journal of Public Administration Research and Th eory 17(2): 213–33.
Ministero della Salute [Italian Ministry of Health]. 2014. Elenco Aziende sanitarie locali e strutture di ricovero [List of Local Health Units and Hospitals]. http://www.salute.gov.it/portale/documentazione/p6_2_8_1_1. jsp?lingua=italiano&id=13 [accessed July 3, 2014].
Organisation for Economic Co-operation and Development (OECD). 2005. Performance-Related Pay Policies for Government Employees. Paris: OECD.
Pandey, Sanjay K., Bradley E. Wright, and Donald P. Moynihan. 2008. Public Service Motivation and Interorganizational Citizenship Behavior: Testing a Preliminary Model. International Public Management Journal 11(1): 89–108.
Pearce, Jone L. 1983. Job Attitude and Motivation Diff erences between Volunteers and Employees from Comparable Organizations. Journal of Applied Psychology 68(4): 646–52.
Pearce, Jone L., William B. Stevenson, and James L. Perry. 1985. Managerial Compensation Based on Organizational Performance: A Time Series Analysis of the Eff ects of Merit Pay. Academy of Management Journal 28(2): 261–78.
Perry, James L. 1986. Merit Pay in the Public Sector: Th e Case for a Failure of Th eory. Review of Public Personnel Administration 7(1): 57–69.
———. 1996. Measuring Public Service Motivation: An Assessment of Construct Reliability and Validity. Journal of Public Administration Research and Th eory 6: 5–22.
———. 1997. Antecedents of Public Service Motivation. Journal of Public Administration Research and Th eory 7:181–97.
Perry, James L., Trent A. Engbers, and So Yun Jun. 2009. Back to the Future? Performance-Related Pay, Empirical Research, and the Perils of Persistence. Public Administration Review 69(1): 39–51.
Perry, James L., Debra Mesch, and Laurie Paarlberg. 2006. Motivating Employees in a New Governance Era: Th e Performance Paradigm Revisited. Public Administration Review 66(4): 505–14.
Perry, James L., and Lyman W. Porter. 1982. Factors Aff ecting the Context for Motivation in Public Organizations. Academy of Management Review 7(2): 89–98.
Ryan, Richard M., and James P. Connell. 1989. Perceived Locus of Causality and Internalization: Examining Reasons for Acting in Two Domains. Journal of Personality and Social Psychology 57(5): 749–61.
Ryan, Richard M., and Edward L. Deci. 2000. Self-Determination Th eory and the Facilitation of Intrinsic Motivation, Social Development, and Well-Being. American Psychologist 55(1): 68–78.
Sauley, Kerry S., and Arthur G. Bedeian. 2000. Equity Sensitivity: Construction of Measure and Examination of Its Psychometric Properties. Journal of Management 26(5): 885–910.
Schmidt, Frank L., and John E. Hunter. 1983. Individual Diff erences in Productivity: An Empirical Test of Estimates Derived from Studies of Selection Procedure Utility. Journal of Applied Psychology 68(3): 407–14.
Shadish, William R., Th omas D. Cook, and Donald T. Campbell. 2002. Experimental and Quasi-Experimental Designs for Generalized Causal Inference. Boston: Houghton Miffl in.
Sims, Henry P., Jr., Andrew D. Szilagyi, and Dale R. McKemey. 1976. Antecedents of Work-Related Expectancies. Academy of Management Journal 19(4): 547–59.
Skinner, Burrhus F. 1969. Contingencies of Reinforcement. New York: Appleton-Century-Crofts.
Titmuss, Richard. 1970. Th e Gift Relationship: From Human Blood to Social Policy. London: Allen and Unwin.
Performance-Related Pay and the Crowding Out of Motivation in the Public Sector: A Randomized Field Experiment 241
Table A1 Variables and Measurements
Variable (Source) Measurements
Performance Number of surgical kits assembled correctly
Extrinsic reward (Wright 2007) Likert-type item (0 = disagree strongly, 6 = agree strongly); Cronbach’s α = .86 Working hard was recognized. Hard work was adequately rewarded.
Reputational concern (Fenig- stein, Scheier, and Buss 1975)
Likert-type item (0 = disagree strongly, 6 = agree strongly); Cronbach’s α = .81 I worried about making a good impression and proving myself to my coworkers. I was concerned about what my coworkers might think about my performance.
Perceived prosocial impact (Grant 2008a)
Likert-type scale (0 = disagree strongly, 6 = agree strongly); Cronbach’s α = .83 I am very aware of the ways in which the job I performed will benefi t others. I can have a positive impact on others through the effort I put in the project.
Public service motivation (Perry 1996)
Likert-type scale (0 = disagree strongly, 6 = agree strongly); Cronbach’s α = .77 Meaningful public service is very important to me. I am often reminded by daily events about how dependent we are on one another. Making a difference in society means more to me than personal achievements. I am prepared to make enormous sacrifi ces for the good of society. I am not afraid to go to bat for the rights of others even if it means I will be ridiculed.
Self-effi cacy (Wright 2007) Likert-type scale (0 = disagree strongly, 6 = agree strongly); Cronbach’s α = .77 I am confi dent that I can successfully perform any tasks assigned to me in my current job. I can complete the work that is expected of me. I am not as well prepared as I could be to meet all the demands of my job.
Conscientiousness (Donnellan et al. 2006)
Likert-type scale (0 = disagree strongly, 6 = agree strongly); Cronbach’s α = .85 I get chores done right away. I often forget to put things back in their proper place. (R) I like order. I make a mess of things. (R)
Intrinsic motivation (Ryan and Connell 1989)
Likert-type scale (0 = disagree strongly, 6 = agree strongly); Cronbach’s α = .88 My job is fun. I fi nd my job engaging. I enjoy my work.
Equity sensitivity (Sauley and Bedeian 2000)
Likert-type scale (0 = disagree strongly, 6 = agree strongly); Cronbach’s α = .90 It is really satisfying to me when I can get something for nothing at work. (R) It is the smart employee who gets as much as he/she can, while giving as little as possible in return. (R)
Age Years of age
Gender 0 = male, 1 = female
Job experience Years of experience in the fi eld of nursing
Appendix
Upton, William. 1973. Altruism, Attribution and Intrinsic Motivation in the Recruitment of Blood Donors. In American Red Cross, Selected Readings in Donor Motivation and Recruitment, vol. 21, 7–38. Washington, DC: American Red Cross.
Vroom, Victor H. 1964. Work and Motivation. New York: Wiley. Weibel, Antoinette, Katja Rost, and Margit Osterloh. 2010. Pay for Performance in
the Public Sector—Benefi ts and (Hidden) Costs. Journal of Public Administration Research and Th eory 20(2): 387–412.
World Bank. 2014. Health Expenditure, Public (% of Total Health Expenditure). http://data.worldbank.org/indicator/SH.XPD.PUBL [accessed December 8, 2014].
Wright, Bradley E. 2007. Public Service and Motivation: Does Mission Matter? Public Administration Review 67(1): 54–64.
Wright, Bradley E., and Adam M. Grant. 2010. Unanswered Questions about Public Service Motivation: Designing Research to Address Key Issues of Emergence and Eff ects. Public Administration Review 70(5): 691–700.
Wright, Bradley E., Donald P. Moynihan, and Sanjay K. Pandey. 2012. Pulling the Levers: Transformational Leadership, Public Service Motivation, and Mission Valence. Public Administration Review 72(2): 206–15.
Wright, Bradley E., and Sanjay K. Pandey. 2008. Public Service Motivation and the Assumption of Person–Organization Fit: Testing the Mediating Eff ect of Value Congruence. Administration & Society 40(5): 502–21.
Copyright of Public Administration Review is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
