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: nicola.belle@unibocconi.it

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).