Relational effects of relative performance information: The role of professional identity

Description
Prior accounting studies on relative performance information (RPI) have mainly considered
the effects of RPI on employees’ performance. This study extends this stream of research by
investigating relational effects of RPI that capture RPI recipients’ attitudes towards RPI provision.
Specifically, we examine the conditions under which RPI has a positive influence on
perceived organizational support (POS). This is important because research on organizational
behaviour has demonstrated that POS is an antecedent of desirable job-related attitudes
such as organizational commitment, job involvement, and extra-role performance.
The results of a dyadic survey in a hospital setting suggest that the effect of RPI on POS
is conditional not only on the perceived usefulness of information and relative performance
but also on the professional identification of RPI recipients. Our findings indicate that RPI
provision is especially suitable for employees who possess a stronger social identification
with their peer groups. Our study adds a relational perspective to the hitherto mostly informational
and motivational focus in studying social comparison mechanisms linked to RPI
systems.

Relational effects of relative performance information:
The role of professional identity
Matthias D. Mahlendorf
a,b,?
, Fabian Kleinschmit
b
, Paolo Perego
c
a
Frankfurt School of Finance & Management, Sonnemannstr. 9-11, 60314 Frankfurt am Main, Germany
b
WHU – Otto Beisheim School of Management, Institute of Management Accounting and Control, Burgplatz 2, 56179 Vallendar, Germany
c
RSM Erasmus University, Department of Accounting & Control, Burgemeester Oudlaan 50, 3062PA Rotterdam, The Netherlands
a b s t r a c t
Prior accounting studies on relative performance information (RPI) have mainly considered
the effects of RPI on employees’ performance. This study extends this stream of research by
investigating relational effects of RPI that capture RPI recipients’ attitudes towards RPI pro-
vision. Speci?cally, we examine the conditions under which RPI has a positive in?uence on
perceived organizational support (POS). This is important because research on organiza-
tional behaviour has demonstrated that POS is an antecedent of desirable job-related atti-
tudes such as organizational commitment, job involvement, and extra-role performance.
The results of a dyadic survey in a hospital setting suggest that the effect of RPI on POS
is conditional not only on the perceived usefulness of information and relative performance
but also on the professional identi?cation of RPI recipients. Our ?ndings indicate that RPI
provision is especially suitable for employees who possess a stronger social identi?cation
with their peer groups. Our study adds a relational perspective to the hitherto mostly infor-
mational and motivational focus in studying social comparison mechanisms linked to RPI
systems.
Ó 2014 Elsevier Ltd. All rights reserved.
Introduction
The effects of relative performance information (RPI) on
employee performance have been studied in the account-
ing literature under the assumption of two underlying
mechanisms. First, economic theory posits that RPI ?lters
out common uncertainty from compensation contracts,
thereby improving risk sharing and increasing employee
effort (Frederickson, 1992; Holmstrom, 1982). Second,
social psychology theory assumes that RPI encourages
social comparisons among employees, which in turn moti-
vates higher learning, effort, and performance (Festinger,
1954; Frederickson, 1992; Hannan, Krishnan, & Newman,
2008; Hannan, McPhee, Newman, & Tafkov, 2013; Tafkov,
2013). Neither strand of accounting research investigates
how the provision of RPI in?uences the employees’ atti-
tudes towards their organizations. In other words, prior
accounting studies have exclusively focused on how RPI
provision affects employees’ productivity but not on how
employees feel about RPI provision. RPI may increase effort
and task-related performance, but evidence is lacking on
whether RPI has positive or negative consequences for
employees’ attitudes at work.
By drawing on insights from social psychology, in this
paper we add a relational perspective to the hitherto
mostly informational and motivational focus on the effects
of RPI. We examine how RPI provision affects an employ-
ee’s attitude towards the organization by focusing on per-
ceived organizational support (POS). The POS concept
captures employees’ general views about whether their
employers care about employee welfare and valuehttp://dx.doi.org/10.1016/j.aos.2014.05.001
0361-3682/Ó 2014 Elsevier Ltd. All rights reserved.
?
Corresponding author at: Frankfurt School of Finance & Management,
Sonnemannstr. 9-11, 60314 Frankfurt am Main, Germany. Tel.: +49 (69)
154008 837.
E-mail address: [email protected] (M.D. Mahlendorf).
Accounting, Organizations and Society 39 (2014) 331–347
Contents lists available at ScienceDirect
Accounting, Organizations and Society
j our nal homepage: www. el sevi er. com/ l ocat e/ aos
employee contributions (Rhoades & Eisenberger, 2002).
Prior studies in organizational behaviour have shown that
POS is a signi?cant antecedent of desirable job-related atti-
tudes, including organizational affective commitment, job
involvement, and extra-role performance (Baran,
Shanock, & Miller, 2012; Rhoades & Eisenberger, 2002).
The objective of this study is to investigate the conditions
under which RPI provision positively affects POS. Consis-
tent with the tenets of Tesser’s (1988) self-evaluation
maintenance model, we ?rst examine how the perceived
usefulness of information and RPI recipient’s relative per-
formance affect the relation between RPI provision and
POS. We then investigate how professional identi?cation
affects such a relationship. Prior research in accounting
assumes that the use of RPI in settings where individuals
perceive similarities between themselves and their peer
group leads to enhanced effort levels (Hannan et al.,
2013, 2008; Tafkov, 2013). However, the factors that drive
such social comparisons have not been explicitly exam-
ined. We posit that professional identity in?uences per-
ceived similarities among employees and thus
signi?cantly interacts with RPI provision in affecting indi-
vidual variations in POS.
We test our hypotheses using data from a dyadic survey
containing pair-wise information (including RPI) from the
hospital’s administrative heads and the chief physicians
of the orthopaedics departments in 107 hospitals in
Germany. In line with our expectations, we ?nd that the
provision of RPI has a positive effect on POS if the RPI reci-
pient perceives the information more useful for her or his
job and she or he has relatively good performance. More-
over, the results suggest that the effect of RPI on POS is
conditional on the professional identi?cation of RPI recipi-
ents. The positive interaction between RPI and professional
identi?cation on POS indicates that RPI is especially suit-
able for (professional) employees who possess a strong
social identi?cation with their peer group. These condi-
tional effects hold in the presence of other types of perfor-
mance information provision (those that are not captured
by RPI in our setting), as well as in the presence of a num-
ber of individual- and organization-level control variables.
Our empirical analysis is further robust to different speci-
?cations of RPI provision that capture either the timing
(frequency) or content (scope) of the RPI released to the
chief physicians in our sample. Our main contention in
this research is that POS stems from a social comparison
process, which employees develop as a reaction to RPI.
RPI provision can therefore exert an important relational
effect on (professional) employees’ attitudes towards their
organization that previous accounting research has not
considered.
The contribution of this paper is twofold. First, we
advance research on RPI effectiveness. Our study explicitly
addresses POS as a key affective reaction to social compar-
isons, thereby complementing previous ?ndings on RPI
effectiveness that primarily drew on economic theory.
More speci?cally, we explicitly integrate two theoretical
perspectives that build on social comparison theory (the
self-evaluation maintenance model and social identity the-
ory). This approach allows us to examine how RPI recipi-
ents’ attitudes towards their organization vary across
individuals who perceive different levels of similarities
between themselves and their peer groups. Second, by
illustrating how RPI can positively in?uence chief physi-
cians’ POS, we contribute to research on managing and
controlling professionals. This is particularly relevant, as
most of the previous management accounting studies con-
ducted in professional organizations primarily focused on
the potential con?icts and dysfunctional outcomes of
bureaucratic forms of management control (Abernethy &
Stoelwinder, 1990a, 1990b, 1995; Comerford &
Abernethy, 1999; Wallace, 1995). By showing how RPI
effectiveness also depends on relational effects in the
workplace, our study provides relevant insights into how
(professional) organizations can improve their perfor-
mance measurement and control systems.
The next section introduces the theoretical background
of the study and develops the hypotheses. We then
describe the research method, the sample selection and
the variable measurement. Next, we report the results of
our empirical analysis. The last section discusses our
?ndings, limitations and directions for future research.
Background and hypothesis development
Literature review on RPI and objectives of this study
We begin by brie?y reviewing the two theoretical per-
spectives on RPI developed in prior accounting studies,
namely agency theory and social comparison theory.
Agency theory posits that relative performance evaluation
(RPE), which is RPI linked to compensation contracts for
performance evaluation, helps the principal ?lter out com-
mon uncertainty from an agent’s contract (Frederickson,
1992; Holmstrom, 1982). An agent’s performance (out-
come) is dependent on her or his effort (input), idiosyn-
cratic uncertainty, and common uncertainty, which is
caused by a random natural state. Measuring the perfor-
mance of multiple agents allows the principal to ?lter out
common uncertainty (noise). Therefore, basing an agent’s
compensation on her or his relative performance with
respect to other agents can improve risk sharing between
the agent and the principal and provide an economic incen-
tive for the agent to increase effort. Extant archival empir-
ical literature investigates the prevalence and the bene?ts
of RPE incentive contracting in reducing common uncer-
tainty. However, this literature has been limited mainly to
investigating whether CEOs and top executives are com-
pensated as if their performance is evaluated relative to
competitors’ performance, and it provides mixed results
(e.g., Aggarwal & Samwick, 1999; Antle & Smith, 1986).
Cichello, Fee, Hadlock, and Sonti (2009) is one of the few
studies conducted at non-executive levels. It provides par-
tial evidence that ?rms conduct RPE by ?ltering out indus-
try performance when choosing to dismiss divisional
managers. The lack of conclusive evidence prompted
research that investigates potential explanations for the
discrepancy between RPE theory and the corresponding
empirical evidence. This literature identi?es issues inher-
ent in the implementation of RPE, such as appropriate peer
groups, variability in the de?nition of relative performance
332 M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347
and model misspeci?cation (Gong, Li, & Shin, 2011). The lit-
erature attributes the limited empirical support for RPE to
contextual factors, such as industry concentration and ?rm
governance, which may in?uence RPE effectiveness
(Albuquerque, 2009; Dikolli, Hofmann, & Pfeiffer, 2013).
An emerging stream of experimental research focuses
on the effects of RPI in the absence of performance incen-
tives, with the objective of isolating the behavioural effects
of RPI and thus indirectly testing the underlying tenets of
agency theory. According to social comparison theory,
individuals strive to outperform others (Festinger, 1954;
Frederickson, 1992). Therefore, RPI that provides insights
about a person’s abilities and skills in relation to others,
stimulates greater effort and a motivation to outperform
those with whom one is being compared (Festinger,
1954; Frederickson, 1992). The motivational effect of RPI
(Hannan et al., 2013) thus refers to social comparison
mechanisms that induce a psychological incentive to exert
effort. It focuses on the content of RPI in terms of estab-
lished goals by comparing individual performance with
that of others (Festinger, 1954; Locke & Latham, 1990,
2002). Hannan et al. (2008) ?nd that precise relative per-
formance feedback improves the performance of partici-
pants compensated under an individual incentive scheme
but worsens the performance of participants compensated
under a tournament incentive scheme. Murthy and Schafer
(2011) investigate whether providing RPI feedback and
framing the feedback in a positive (good job) or negative
(poor job) manner induces performance improvements in
a repetitive task. Their results reveal a signi?cant positive
effect of providing RPI and positively framed feedback.
Hannan et al. (2013) examine how RPI affects employee
performance and allocation of effort across tasks in a
multi-task environment. They show that RPI induces both
motivation and effort distortion effects and that both
effects are magni?ed when RPI is public rather than pri-
vate. Tafkov (2013) predicts and ?nds that RPI positively
affects performance under two compensation contracts
(?at-wage or individual performance-based), and this posi-
tive effect is greater under an individual performance-
based contract than it is under a ?at-wage contract. He also
predicts and ?nds that, although both public and private
RPI have a positive effect on performance, the effect is
greater when RPI is public.
Prior accounting studies that rely on informational
(common uncertainty reduction) and motivational
(effort-inducing) motives have not, however, devoted any
attention to the effect of RPI on employees’ attitude
towards their organization. In other words, prior account-
ing research exclusively focuses on howemployees’ perfor-
mance is affected when RPI is provided or contracted upon
but not on how RPI in?uences employees’ affective states
or feelings. While RPI is potentially effective in increasing
employee task-related performance, there is no evidence
regarding the effect of RPI on employees’ job-related atti-
tudes. In this study, we explore relational responses of
RPI recipients by focusing on how favourably employees
feel their organizations treat them. Two reasons motivate
our choice. First, it is theoretically important to investigate
whether social comparisons in?uence job-related atti-
tudes, independent of the monetary incentives to which
these comparisons can be contractually linked. Recent
studies in labour and behavioural economics attempt to
disentangle motivational versus affective responses
evoked by information feedback, either relying upon
experimental (e.g., Eriksson, Poulsen, & Villeval, 2009;
Kuhnen & Tymula, 2012) or ?eld data (e.g., Azmat &
Iriberri, 2010; Vidal & Nossol, 2011). The accounting liter-
ature thus far seems to have neglected the growing knowl-
edge on RPI and similar feedback mechanisms. Second, a
research focus on employees’ attitudes towards their orga-
nization that results from RPI provision in the workplace
has relevant practical implications in determining whether
social comparisons increase or decrease morale. As positive
attitudes also have bene?cial effects on employees’ well-
being (Rhoades & Eisenberger, 2002), we posit that more
knowledge on RPI consequences provides relevant insights
about how organizations can improve their management
control systems.
In this study, we explore employee attitudes associated
with RPI provision by relying on the assumptions from
social exchange theory that relationships in the workplace
are built upon the exchange of impersonal resources (e.g.,
fringe bene?ts and information) and socio-emotional
resources (e.g., approval and respect) (Aselage &
Eisenberger, 2003; Cropanzano & Mitchell, 2005;
Eisenberger, Armeli, Rexwinkel, Lynch, & Rhoades, 2001;
Eisenberger, Cummings, Armeli, & Lynch, 1997; Foa &
Foa, 1980). Social exchange theory relies on the norm of
reciprocity (Blau, 1964; Coyle-Shapiro & Shore, 2007;
Cropanzano & Mitchell, 2005; Gouldner, 1960), whereby
recipients of supportive treatment from others are
expected to reciprocate such treatment (Eisenberger
et al., 2001; Gouldner, 1960).
1
To appraise the exchange
relationship between the RPI recipient and the organization,
we rely on the concept of perceived organizational support
(POS) (Cropanzano & Mitchell, 2005; Eisenberger et al.,
1986; Levinson, 1965). The notion of POS captures the idea
that employees form a general view about whether their
employer ‘‘values their contributions and cares about their
well-being’’ (Rhoades & Eisenberger, 2002, p. 698); in other
words, whether employees perceive their organization as
supportive and helpful. The concept of POS is deeply rooted
in social psychology research (Aselage & Eisenberger, 2003;
Baran et al., 2012; Rhoades & Eisenberger, 2002). In their
meta-analysis, Rhoades and Eisenberger (2002) document
that POS is positively associated with organizational com-
mitment, in-role performance, and extra-role performance,
while it is negatively correlated with turnover intentions
and withdrawal behaviours.
Social exchange relationships in general and the forma-
tion of POS in particular imply that employees collect and
interpret a great deal of job-related information obtained
1
Social exchange theorists assume that the reciprocation of such
positive treatments is due to felt obligations that arise from implicit
psychological contracts (Aselage & Eisenberger, 2003; Eisenberger et al.,
2001). Prior studies document that managers and employees reciprocate
with loyalty and appreciation when the organization provides support and
resources, such as relevant information and feedback (Blau, 1964; Coyle-
Shapiro & Shore, 2007; Cropanzano & Mitchell, 2005; Eisenberger,
Huntington, Hutchison, & Sowa, 1986; Eisenberger et al., 2001; Gouldner,
1960; Levinson, 1965; Rhoades & Eisenberger, 2002).
M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347 333
via informal interactions with coworkers and observations
of peers, and via formal, performance measurement sys-
tems provided by the ?rm, such as RPI. The POS level
may thus depend on the social comparison standards used
by an individual who receives feedback and performance
rank information (even in cases when performance is not
formally linked to incentives). Whether RPI provision
exerts a positive or negative relational effect on POS may
be contingent on different factors. In the social psychology
literature, Tesser’s self-evaluation maintenance (SEM)
model (Tesser, 1988; Tesser, Millar, & Moore, 1988) pro-
vides an important extension to Festinger’s (1954) basic
social comparison theory. The SEM model considers the
role of emotions and affective consequences in social com-
parison processes, assuming that individuals have a strong
motivation to maintain a positive self-image depending on
the type of comparisons they make with others (Achee,
Tesser, & Pilkington, 1994; Tesser, Pilkington, & Mcintosh,
1989). The SEM model has proven to be robust in showing
the maintenance of an individual self-evaluation after a
comparison depending upon three factors: (1) the rele-
vance of the comparison domain (highly self-relevant ver-
sus not relevant to one’s identity); (2) the comparison
direction (upward versus downward comparison); and
(3) the psychological closeness of the compared person
(similar/friend versus different/stranger). In this paper,
we investigate the applicability of the SEM model as a
broader adaptive state in interpersonal contexts to exam-
ine relational effects (measured by variations in POS) fol-
lowing social comparisons induced by RPI provision. We
draw upon the SEM model to study how RPI affects POS
depending on three variables: (1) the perceived usefulness
of RPI (to capture the self-relevance of the comparison
dimension); (2) the relative performance of the RPI recipi-
ent (to capture the direction of comparison); and (3) the
professional identi?cation of the RPI recipient (to capture
the degree of closeness or peer group similarity in affecting
the comparison process). We develop the hypotheses for
each of these three variables next.
Theory and hypothesis development
In the search for constant performance improvement,
individuals rely on comparisons with others to analyse
processes and make appropriate decisions. Such compari-
sons stimulate learning by relying on the experience of
others and compensate for internal shortcomings by rais-
ing awareness of external performance standards (Baines
& Lang?eld-Smith, 2003; Chenhall & Lang?eld-Smith,
1998). RPI facilitates decision-making by providing a refer-
ence point regarding what may be considered a suf?cient
level of aspiration for a certain task and an indication of
the intensity of effort necessary for its achievement
(Festinger, 1954; Suls, Martin, & Wheeler, 2002;
Vidyarthi, Liden, Anand, Erdogan, & Ghosh, 2010). As RPI
contains information on the results of others, it assists in
setting goals with respect to dif?culty and direction. RPI
also provides employees with cues regarding the attempts
and corresponding success of others. Therefore, as RPI
serves as a potential means to decide what to achieve
and how to achieve it, it reduces complexity when
choosing among alternative goals and indicates the level
of effort required to accomplish them. Furthermore, RPI
enables its recipients to obtain a better understanding of
their capabilities and to discover their strengths and weak-
nesses. This empowers individuals to construct a more
complete self-image and answer important questions such
as ‘‘Who am I?’’ and ‘‘What can I do?’’ Thus, RPI allows
employees to experience self-awareness through the pro-
cess of self-evaluation and self-enhancement (Wood,
1989, 1996). Consequently, RPI increases self-knowledge
by encouraging individuals to consider their own quali?ca-
tions, relative to those of others. In sum, we argue that RPI
has the potential to facilitate decision-making. In turn, this
potential bene?t should be reciprocated by how an
employee feels she or he is treated by the organization
providing RPI.
According to the SEM model, the information that an
individual requires to perform an activity or make a
decision is considered relevant if the activity or decision
is central to the individual’s self-concept and if the individ-
ual strives to perform the activity well or successfully
make the decision (Tesser, 1988). Relevance of the compar-
ison domain hence represents the extent to which the indi-
vidual strives for competence on a speci?c performance
dimension. The SEM model predicts that a comparison
with another’s performance poses a larger threat or chal-
lenge to an individual when the relevance of a performance
dimension and the self-de?ning importance of the activity
are high, compared to when the relevance is low. Similarly,
it can be expected that an employee feels supported by
RPI provision only if she or he perceives the feedback
information as useful (i.e., highly self-relevant) to carry
out an activity or make a decision. Alternatively, an
employee who thinks RPI is useless (i.e., the information
about a performance dimension has low relevance to an
individual’s self-de?nition) may feel annoyed by receiving
RPI. Such comparison mechanisms are apparent from
past ?eld studies conducted in professional organizations.
For example, studies of healthcare settings have shown
that the relationship between medical professionals
and administrative functions tends to be con?icted as a
result of divergent work ideologies and opinions on the
perceived usefulness of performance measurement sys-
tems (Abernethy & Stoelwinder, 1990a, 1990b, 1995;
Comerford & Abernethy, 1999; Wallace, 1995). In line with
these arguments, we posit that:
H1. The provision of RPI has a negative effect on POS if the
RPI recipient perceives the information provided by her or
his organization as less useful but has a positive effect if
the RPI recipient perceives the information as more useful
for her or his job.
From the onset of social comparison research in social
psychology, the preferred direction of comparison has been
a controversial theoretical issue (Corcoran, Crusius, &
Mussweiler, 2011; Suls & Wheeler, 2000). Both positive
and negative reactions can be triggered by the provision of
RPI or other forms of performance feedback (Festinger,
1954; Tesser, 1988; Tesser et al., 1988). Current develop-
ments in the literature underscore that individuals compare
334 M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347
themselves with others who are similar (lateral compari-
sons), better (upward comparisons), or worse (downward
comparisons). With whom individuals choose to compare
themselves has important consequences for performance
rankings such as RPI and depends largely on the motive that
underlies the comparisons (Corcoran et al., 2011).
One well established ?nding from the social compari-
son literature is the distinction between contrast effects
and assimilation effects (Smith, 2000). The contrast effect
describes a situation in which the individual seeks to dis-
tance herself or himself from the comparison target, while
the assimilation effect occurs if the individual strives to
connect with the comparison target. The literature sug-
gests that the contrast condition is characterized by indi-
viduals who experience positive affective consequences
when making comparisons with others who are worse off
(downward comparisons), in combination with negative
affective reactions when making comparisons with refer-
ents who perform better (upward comparisons) (Buunk &
Gibbons, 2007; Corcoran et al., 2011; Greenberg, Ashton-
James, & Ashkanasy, 2007). Simply put, comparing oneself
to those who are worse off enhances one’s self-esteem,
reduces anxiety, and generates positive affective reactions,
all of which are essential for self-enhancement. Conversely,
comparing oneself to others who are better off prompts an
unfavourable self-image.
2
Little ?eld or archival research
conducted in organizations examines social comparisons
and the role of moderating factors that induce contrast or
assimilation effects. The literature is thus not conclusive
on the preferred direction of comparisons, with both upward
and downward comparisons having their ‘‘ups and downs’’.
3
To the best of our knowledge, no prior study has exam-
ined how directional social comparison processes induced
by the provision of RPI relates to POS levels. However, a
few prior studies have examined the directional effects of
social comparisons on affective commitment, of which
POS remains a signi?cant antecedent (Rhoades &
Eisenberger, 2002). More speci?cally, Brown, Ferris,
Heller, and Keeping (2007, p. 63) argue that, because affec-
tive commitment is thought to develop through a social
exchange mechanism, ‘‘contrast effects from downward
comparisons set a lower reference point against which to
evaluate one’s current situation’’, thus enhancing an
employee’s perception that she or he receives positive
experiences from an organization. Their results, which
are based on a large and occupationally diverse sample of
991 employed adults, indicate that upward (downward)
social comparison is signi?cantly negatively (positively)
related to affective commitment. This ?nding supports a
contrast, not an assimilation, effect.
We similarly posit that RPI recipients with good relative
performance compared to inferior referents (downward
comparisons) will bene?t from RPI, because it provides
an enhanced self-view. Consequently, these individuals
will feel supported by an organization that provides RPI.
In contrast, we argue that RPI recipients with relatively
poor performance compared to superior others (upward
comparisons) may feel threatened by a ranking system,
and thus will have a negative affective reaction and
reduced POS. Thus, any self-improvement motive that
may be associated with upward comparisons (assimilation
effect) would be outweighed by feelings of inferiority and
jealousy (contrast effect) that negatively impact POS (cf.,
Buunk et al., 2005; Eddleston, 2009). Accordingly, we posit
that:
H2. The provision of RPI has a negative effect on POS if the
RPI recipient has relatively poor performance but a positive
effect if the RPI recipient has relatively good performance.
In addition to the effects associated with the relevance
of the comparison domain and the comparison direction,
it is also important whether the comparison group matters
for the individual receiving RPI (Tesser, 1988). Prior
research shows that the use of RPI in settings where indi-
viduals perceive similarities between themselves and their
peer group leads to higher effort, with the underlying
rationale that a social comparison should have a greater
effect on an agent’s ‘‘cognitive processes as the source of
the social in?uence increases in saliency’’ (Frederickson,
1992, p. 653). However, the underlying mechanisms of
peer group similarities have been largely ignored in
accounting research. Research has instead focused on RPI
effects through implicit approaches (i.e., manipulating
RPI provision in laboratory settings or examining
RPE-based executive compensation contracts in archival
studies). This study aims at explicitly exploring social
in?uences and similarity of comparison agents by drawing
upon social psychology insights.
According to social identity theory, individuals perceive
fellow group members to be similar and to hold similar
values and goals (Ashforth & Mael, 1989; Brewer &
Gardner, 1996). As a result of social identi?cation, individ-
uals order their social worlds into in-groups and out-groups
(Ashforth & Mael, 1989; Turner, Hogg, Oakes, Reicher, &
Wetherell, 1987). The individual then uses the reference
group as a relevant standard against which one’s current
circumstances can be evaluated. Our study draws on social
identity theory, because it allows us to account for an indi-
vidual’s affective reaction regarding the value of RPI,
depending on how similar and alike the individual consid-
ers those with whomshe or he is compared. We speci?cally
investigate social comparisons in a professional setting
and focus on the concept of professional identi?cation,
which is de?ned as perceived similarity to a group of
peer professionals (Mael & Ashforth, 1992). We propose
that the identi?cation of professional managers (chief
2
Even though upward comparisons typically have negative effects (e.g.,
shame and fear of sanctions), Wheeler (1966) suggested that an individual
may attempt to reaf?rm her or his self-esteem by engaging in upward
comparisons, because such comparisons encourage learning from those
who are more skilled and successful (Collins, 1996, 2000; Corcoran et al.,
2011). A series of theoretical models tested in experimental settings
examine the conditions under which downward or upward comparisons
evoke contrast versus assimilation effects (for comprehensive reviews, see
Buunk, Zurriaga, Peiro, Nauta, & Gosalvez, 2005; Corcoran et al., 2011; Suls
& Wheeler, 2000).
3
Buunk et al. (2005) posit and provide evidence that contrast effects are
more likely among individuals who perceive their working climates as
highly competitive and who place great emphasis on individual rankings in
a referent group. In a similar vein, Eddleston (2009) ?nd that competitive-
ness of a manager’s group is a signi?cant moderator in the way social
comparisons affect managers’ turnover intentions and career satisfaction.
M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347 335
physicians, in our setting) with their professional group
affects their attitudes towards comparative feedback infor-
mation provided by their organization. We posit that RPI
can be viewed as supportive if the recipient strongly iden-
ti?es with the peer group with which she or he is compared.
Our assertion that individuals have a natural drive to
evaluate their abilities in reference to those of others stems
from Festinger (1954), who described the circumstances
that might facilitate and evoke comparisons. In alignment
with Wood’s ‘‘similarity hypotheses’’ (Wood, 1989, p.
231), we argue that professionals who strongly (weakly)
identify with their professional group will possess a high
(low) perceived similarity with those about whom they
receive RPI. In turn, it can be assumed that professional
identi?cation will render RPI more salient and valuable.
RPI saliency increases for recipients with strong profes-
sional identi?cation because RPI reduces subjective uncer-
tainty. That is, comparing one’s own decisions and actions
to those of similar and relevant others ampli?es the impor-
tance of these comparisons (Tesser & Campbell, 1980). We
therefore argue that the provision of RPI about members of
an individual’s identi?ed group will help to construct the
prototypical image of this group and will in turn facilitate
the self-categorization process (Hogg & Terry, 2000). Con-
sistent with the SEM model, we posit that RPI exerts a posi-
tive effect on POS only if the peers, and consequently their
performance, are salient to the individual’s self-de?nition
(Tesser & Campbell, 1980). We therefore expect that:
H3. The provision of RPI has a negative effect on POS if the
RPI recipient has low professional identi?cation but a
positive effect if the RPI recipient has high professional
identi?cation.
Overall, we expect that the effects of RPI may vary
depending on professional identi?cation. Furthermore,
the effects may vary even if other moderators, such as rel-
ative performance and usefulness, are held constant (i.e.,
are controlled for).
Method
Sample and survey
We collected data in a professional organizational set-
ting, obtaining pair-wise responses fromthe administrative
head and the chief physician of the orthopaedics depart-
ment in German hospitals. Furthermore, we collected data
from publicly available sources for all hospitals in our
sample. These data included ownership type, number of
beds, and teaching hospital status, which we use as control
variables. Inthis study, the administrative headis part of the
executive board and is responsible for all administrative
activities, such as ?nancial planning and control within
the hospital. This includes the responsibility to design
and operate performance measurement and control sys-
tems. However, chief physicians perform managerial and
professional roles in their departments. While they are phy-
sicians by professional training, it is common for them to
take on additional managerial responsibilities during their
careers as employed professionals. As chief physicians of a
department, they normally divide their time between pro-
fessional tasks (e.g., performing surgeries) and managerial
tasks (e.g., chairing department meetings).
We chose this setting for several reasons. It provides an
environment where activities are likely to arise from impli-
cit contracts, interpersonal relations and social compari-
sons (Eldenburg, 1994; Hannan et al., 2013), which are
the focal points of this investigation. It allows us to identify
and segregate the information provider (the administrative
head) from the information recipient (the chief physician).
This setting provides an environment where all informa-
tion recipients could identify with their comparison group,
namely the medical profession. The setting also ensures
similar identi?cation targets for all respondents. By choos-
ing chief physicians as respondents, we follow previous
studies on professionals in management accounting and
organizational behaviour that examined physicians as
research subjects, as the medical profession is one of the
most established and widespread (e.g., Abernethy &
Stoelwinder, 1995; Comerford & Abernethy, 1999;
Hekman, Bigley, Steensma, & Hereford, 2009; Hekman,
Steensma, Bigley, & Hereford, 2009). Finally, the manage-
ment accounting literature highlights that the relationship
between professionals and bureaucratic organizations
tends to be con?icted as a result of differing work ideolo-
gies (Abernethy & Stoelwinder, 1990a, 1990b, 1995;
Comerford & Abernethy, 1999; Wallace, 1995). Thus, it is
of particular interest to investigate the conditions under
which information provided by an organization such as
RPI evoke positive affective responses by professionals.
We rely on internet-based, ?eld surveys to collect the
data. As we investigate the relationship between RPI pro-
viders and recipients, we divided the data collection into
two inquiries. This dyadic design ensured that we con-
tacted the most appropriate respondent regarding the phe-
nomena in question, and it reduced the effects of common
method bias (Podsakoff, MacKenzie, Lee, & Podsakoff,
2003; Spector, 2006). The ?rst survey,
4
which we addressed
to the hospital administrative heads, sought to gather data
on the internal reporting system of RPI. The second survey,
which we addressed to the chief physicians of the orthopae-
dics departments, was designed to assemble data regarding
perceptions of and attitudes towards the employer. We
included only the orthopaedics medical specialization to
assure that all respondents to the second survey had similar
socialization experiences during their professional educa-
tions and performed medical treatments with a similar level
of task complexity and task interdependence. The decision
to contact the chief physicians was made because, if RPI
were provided by the employing hospital, the head of the
department would usually receive that type of information
and incorporate it into management decisions. For physi-
cians at lower levels in the hierarchy, this is not always
4
Prior to administering the survey, we conducted ?fteen ?eld interviews
in seven hospitals. Five of them were conducted with physicians, nine with
employees and managers of the administration, and one with an employee
of the nursing service. Insights from the interviews, which lasted one hour
on average, were used to understand more about the processes involved
when providing RPI in hospitals and to ensure proper sampling and
respondent selection.
336 M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347
the case. We collected survey data from the orthopaedics
department primarily to reduce the variance in unobserved
context variables and for three additional reasons. First,
orthopaedics is a widespread medical specialization, ensur-
ing a suf?ciently large population. Second, in comparison
to other medical specializations (e.g., cardiology) orthopae-
dists have more discretion when planning their surgeries,
which allows them to more easily incorporate the informa-
tion provided into their daily decisions. Third, as the inter-
views revealed, orthopaedists are often considered typical
representatives of the medical profession.
We selected hospitals from a database containing infor-
mation on 2072 of the 2083 hospitals (as of 2008) located
in Germany (Destatis., 2013). We identi?ed and selected
647 hospitals that reported information on orthopaedic
treatments for 2008. This ensured that a second respon-
dent from the orthopaedics department was available if
the administrative head had participated in the ?rst sur-
vey. We designed and administered the survey following
the recommendations by Dillman, Smyth, and Christian
(2009). We discussed early versions of the questionnaire
with each target group during the expert interviews. Fur-
ther, we relied on established constructs from previous
studies to measure the variables. We pre-tested the ?nal
versions with three members of each target group, none
of whom participated in the ?nal study. Finally, we trans-
lated all survey items into German, followed by a reverse-
translation to compare with the original wording.
We sent a pre-notice letter by regular mail consisting of
a personalized and hand-signed cover letter explaining our
motivation, an invitation to participate and an assurance of
con?dentiality, as well as a ?yer providing background
information on the project, the researchers involved, and
their previous projects. Soon after the pre-notice, we sent
the initial invitation via personalized e-mail containing a
reference to our pre-notice letter and a link to the web-
based survey. The initial invitation was followed by a tele-
phone call and two follow-up e-mails. As incentives, we
offered the administrative heads a detailed report about
the study and an invitation to participate in a workshop,
and we offered the chief physicians a detailed report about
the study and a €10 donation to a charity of their choice. We
administered the ?rst survey to 596 hospitals for which we
had contact details for the administrative head. This ?rst
questionnaire was completed by 204 participants, resulting
in a response rate of 34%. We administered the second sur-
vey to 184 chief physicians of the orthopaedics depart-
ments of those hospitals that participated in the ?rst
survey. Of those 184, 110 responded, which yielded a
response rate of 60%. We dropped three observations with
missing values regarding the dependent variable, leaving
a dyadic dataset of 107 pairs from both questionnaires.
Variable measurement
The administrative heads completed the ?rst survey
which included questions concerning, among other
aspects, the provision of RPI and the relative performance
of the orthopaedics department (PERF) (see Appendix).
From the second survey, which was completed by the chief
physician of the orthopaedics department, we obtained
perceived usefulness of information (USEFUL), professional
identi?cation (PI), and perceived organizational support
(POS) from the second survey. Table 1 shows the descriptive
statistics, Cronbach’s alpha and composite reliability scores
(Bagozzi & Yi, 1988). For the main analysis, we used factor
scores for all multi-itemmeasures if not indicated otherwise.
5
POS captured the chief physicians’ overall perceptions of
being valued, cared for, and supported by their hospitals using
the 8-item, short-form survey developed by Eisenberger et al.
(1986). We asked the respondents to indicate the extent of
their agreement using a 5-point Likert scale (1 = strongly dis-
agree; 5 = strongly agree). Composite reliability of POS is 0.90
and the Cronbach’s alpha 0.89, indicating internal reliability.
We examined two variables that captured the provision
of RPI, namely RPI_FREQ and RPI_SCOPE. We adopted scales
from Eldenburg (1994) and Pizzini (2006) by asking the
administrative heads how frequently (RPI_FREQ) the chief
physicians of the orthopaedics department were provided
with information for benchmarking their department’s per-
formance against that of other departments. RPI_FREQ was
measured on a 7-point scale (1 = not reported; 2 = annu-
ally; 3 = semi-annually; 4 = quarterly; 5 = monthly; 6 =
weekly; 7 = daily) along four dimensions: medical perfor-
mance (e.g., Case-Mix-Index,
6
number of cases), ?nancial
performance (e.g., earnings, expenditures, etc.), quality of
care (e.g., patient satisfaction, adherence to quality stan-
dards, etc.), and ef?ciency (e.g., patient length of stay, depart-
ment occupancy, etc.). A factor analysis revealed satisfactory
convergent validity for all four dimensions. We obtained a
composite reliability and a Cronbach’s alpha of 0.82. To cal-
culate RPI_SCOPE, we summed the number of different
dimensions on which RPI was reported.
7
Thus, RPI_SCOPE
had a value of 0 if no RPI was reported, a value of 1 if RPI
was reported on one dimension (e.g., medical performance),
and up to a value of 4 if RPI was reported along all four
dimensions (i.e., medical performance, ?nancial perfor-
mance, quality of care, and ef?ciency).
PERF captured the relative performance of the orthopae-
dics department for whichthe chief physicianwas responsi-
ble. PERF was based on responses provided by the
administrative headof the hospital. This performance rating
included nine dimensions (earnings per case; adherence to
budget; pro?t margin; expenditures per case; overall qual-
ity of care; patient satisfaction; average length of stay; aver-
age occupancy; and reputation). We asked the respondents
to indicate the relative performance using a 5-point Likert
scale (1 = below average; 5 = above average). Composite
reliability and the Cronbach’s alpha of PERF were both 0.88.
We used the variable measuring usefulness of informa-
tion (USEFUL) to elicit whether chief physicians perceived
the information provided by the organization as useful.
We measured USEFUL with 16 items addressing the same
5
In an untabulated robustness check, we replicated our main analyses
with equally weighted means of multi-item measures, instead of factor
scores. The results were substantially similar.
6
The Case-Mix-Index represents a surrogate for the average level of
illness severity at the hospital and departmental levels (cf., Eldenburg,
1994). Hospitals use this as a multiplier in the reimbursement scheme.
7
We select a cutoff value of 3 regarding frequency (i.e., we counted an
RPI dimension as being reported if it was provided at least every six months
or more frequently).
M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347 337
four aspects that were used to capture RPI provision: medical
performance (e.g., Case-Mix-Index, number of cases); ?nan-
cial performance (e.g., earnings, expenditures, etc.); quality
of care (e.g., patient satisfaction, adherence to quality stan-
dards, etc.); and ef?ciency (e.g., patient length of stay, depart-
ment occupancy, etc.). The items were adapted from Legris,
Ingham, and Collerette (2003) and used a 5-point Likert scale
(1 = strongly disagree; 5 = strongly agree). Composite reliabil-
ity and the Cronbach’s alpha of USEFUL were 0.93.
We measured professional identi?cation (PI) of chief
physicians with their peer-group using the 5-item identi?-
cation scale from Hekman, Bigley et al. (2009) and
Hekman, Steensma et al. (2009), which examined a similar
professional setting. Respondents were asked to indicate
how strongly they agreed or disagreed with survey state-
ments using a 5-point Likert scale (1 = strongly disagree;
5 = strongly agree). A factor analysis revealed suf?cient
factor loadings for all ?ve items. The composite reliability
of PI was 0.78 and the Cronbach’s alpha 0.77.
The questionnaires further included a number of control
variables. We controlled for physicians’ age (AGE) because
older individuals are more likely to identify with their profes-
sion (Ashforth & Mael, 1989; Riketta, 2005) and are also more
susceptible to perceptions of favourable treatment by their
organization (Rhoades & Eisenberger, 2002). In their meta-
analysis on POS, Rhoades and Eisenberger (2002) document
that stressors have a signi?cant negative impact on POS. In
a hospital context, an important stressor is a high occupancy
rate, as it is associated with overtime work and less time with
patients. Thus, we included the occupancy rate (OCCUP) as a
control variable measured in percent. Another stressor that is
relevant for POS (Rhoades & Eisenberger, 2002) is role ambi-
guity (ROLAMB). We used the 6-item instrument from Rizzo,
House, and Lirtzman (1970) that was also used in previous
accounting studies (Burney, Henle, & Widener, 2009;
Burney & Widener, 2007; Maas & Matejka, 2009). The instru-
ment includes (reversed) questions such as ‘‘I feel certain
about how much authority I have’’ and ‘‘I know exactly what
is expected of me’’. Composite reliability of this measure was
0.78 and the Cronbach’s alpha was 0.75. As job autonomy can
have a positive impact on POS (Rhoades & Eisenberger,
2002), we included it to control for possible confounding
effects. Autonomy (AUTON) was measured using the ‘‘self-
determination’’ dimension from Spreitzer’s (1995) scale on
psychological empowerment that has been used in prior
accounting studies (Hall, 2008). It consists of items such as
‘‘I have signi?cant autonomy in determining how I do my
job’’. Composite reliability of AUTON was 0.90 and the Cron-
bach’s alpha was 0.91.
Similar to Hannan et al. (2013), in this study we aimed to
capture the effects of RPI while controlling for potential con-
founding results caused by variable monetary incentives. We
therefore included a survey itemin which we asked the chief
physician the maximum percentage of her or his bonus rel-
ative to the ?xed salary (INCENT). The average share of var-
iable pay in our sample amounted to 11.68%, signalling that
?xed pay is more important than variable pay across a large
section of German hospitals. Furthermore, as it can be
assumed that a higher proportion of private patients is ben-
e?cial for chief physicians by allowing them to generate
additional personal income (Strehl, 2002), we controlled
for the proportion of private patients (PRIV_PAT) in a hospi-
tal. This control variable aligns with the argument that
favourable job conditions, such as compensation level, posi-
tively in?uence POS (Rhoades & Eisenberger, 2002).
Hutchison and Garstka (1996) have provided evidence
that overall communication intensity has a positive effect
on POS. To ensure that the observed effects were unambig-
uously related to RPI provision, we controlled for two
variables (FREQ and SCOPE) as proxies for the overall com-
munication intensity of information provided by the orga-
nization. The measures for these constructs were identical
to RPI_FREQ and RPI_SCOPE, except that the survey items
asked for information ‘‘about their departments’’ instead
Table 1
Descriptive statistics and psychometric quality assessment.
Variables Theoretical range Empirical range Mean SD Alpha CR
POS 1.00–5.00 1.00–4.88 3.20 0.79 0.89 0.90
RPI_SCOPE 0.00–4.00 0.00–4.00 2.40 1.50 NA NA
RPI_FREQ 1.00–7.00 1.00–5.50 3.42 1.23 0.82 0.82
PERF 1.00–5.00 1.89–4.89 3.56 0.59 0.88 0.88
USEFUL 1.00–5.00 1.25–5.00 3.00 0.75 0.93 0.93
PI 1.00–5.00 1.00–5.00 3.25 0.84 0.78 0.77
AGE 36.00–63.00 52.42 6.10 NA NA
OCCUP 64.00–120.00 86.27 9.54 NA NA
ROLAMB 1.00–5.00 1.00–4.17 2.27 0.61 0.78 0.75
AUTON 1.00–5.00 1.00–5.00 3.84 0.85 0.91 0.90
INCENT 0.00–100.00 0.00–100.00 11.68 15.55 NA NA
PRIV_PAT 0.00–100.00 4.00–65.00 16.34 9.05 NA NA
FREQ 1.00–7.00 2.75–5.75 4.46 0.60 0.50 0.39
SCOPE 0.00–4.00 2.00–4.00 3.54 0.64 NA NA
GOALCOM 1.00–5.00 1.00–5.00 3.57 0.67 0.79 0.78
NONPROFIT 0.00–1.00 0.00–1.00 0.51 0.50 NA NA
PUBLIC 0.00–1.00 0.00–1.00 0.42 0.50 NA NA
Note: CR = composite reliability; NA = not applicable. The descriptive statistics of multi-item scales (see Appendix) refer to mean scores. For the main
results, we use factors scores with a mean of 0 and a standard deviation of 1.
POS perceived organizational support, RPI_SCOPE scope of relative performance information, RPI_FREQ frequency of relative performance information, PERF
relative performance, USEFUL usefulness of information, PI professional identi?cation, OCCUP occupancy rate, ROLEAMB role ambiguity, AUTON autonomy,
INCENT percentage of bonus relative to ?xed salary, PRIV_PAT proportion of private patients, FREQ frequency of information, SCOPE scope of information,
GOALCOM goal communication, NONPROFIT/PUBLIC indicator variables taking the value of 1 for for-pro?t/public hospitals and 0 otherwise.
338 M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347
of information ‘‘which compares their department’s per-
formance with the performance of other departments’’.
Composite reliability of the FREQ measure was 0.50 and
the Cronbach’s alpha was 0.39. Because these values were
rather low and indicated a lack of internal consistency for
the FREQ measure, we conducted a robustness check that
included the four dimensions of FREQ separately. The
results (untabulated) were qualitatively similar. Moreover,
we controlled for goal communication to exclude the pos-
sibility that the speci?c effects of RPI provision re?ected
the impact of communicating broader goals within a hos-
pital (GOALCOM). The four survey items to measure this
variable addressed the same dimensions as RPI_FREQ and
RPI_SCOPE, except that we elicited responses concerning
goal communication instead of RPI provision. Composite
reliability of GOALCOM was 0.79 and the Cronbach’s alpha
was 0.78. Finally, ownership type is an important context
variable in hospital settings, because private hospitals face
different institutional pressures and exhibit different cost
structures than non-pro?t and public hospitals
(Holzhacker, Krishnan, & Mahlendorf, 2014). Thus, we
controlled for ownership type by including the dummy
variables NON_PROFIT (whether a hospital is a for-pro?t
or a not-for-pro?t organization) and PUBLIC (whether a
hospital is a public or privately owned organization).
8
Table 2 presents the correlation matrix among the
variables. The average variance extracted (AVE) for the
latent constructs exceeded the recommended criterion
of 0.50 for all variables (Fornell & Larcker, 1981). To
assess whether the discriminant validity is adequate,
AVE should exceed the variance shared between the
intended construct and other constructs in the model
(i.e., the squared correlation between two constructs).
The diagonal elements in Table 2 satisfy this condi-
tion. The correlations between POS and RPI_FREQ (0.17)
and RPI_SCOPE (0.20) were signi?cant at p < 0.10 and
p < 0.05 respectively. All p-values reported in the paper
are two-tailed.
Results
Multivariate analyses
We conduct multivariate regression analysis with full
information maximum likelihood estimation (Raykov,
2005; Schafer & Graham, 2002) and robust standard errors
(Huber, 1967; White, 1980) to test our hypotheses. We use
STATA software version 13. To improve interpretability of
the results, we mean center all continuous independent
variables (Dalal & Zickar, 2012).
As explained in the measurement section, we have two
proxies for the provision of RPI, which are labelled
RPI_SCOPE and RPI_FREQ. We begin with the analysis of
RPI_SCOPE (see Table 3). In both models, the dependent
variable is POS. Model 1 shows the results with main
effects only. Model 2 includes the interaction effects. The
interaction of RPI_SCOPE and USEFUL is signi?cant
(p < 0.10), indicating that the effect of RPI_SCOPE on
POS becomes more positive for higher values of USEFUL.
Table 2
Correlation matrix.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. POS 0.73
2. RPI_SCOPE 0.20
**
NA
3. RPI_FREQ 0.17
*
0.92
***
0.73
4. USEFUL 0.32 0.05 0.00 0.67
5. PERF 0.01 0.15 0.15 0.04 0.67
6. PI 0.07 0.07 0.02 0.12 À0.19 0.66
7. INCENT 0.03 0.06 0.06 À0.02 0.03 0.00 NA
8. PRIV_PAT 0.24
**
0.07 0.02 0.20
*
À0.04 À0.12 0.03 NA
9. OCCUP À0.14 0.06 0.05 À0.06 0.11 À0.13 0.33
***
0.21
**
NA
10. AGE À0.16 À0.08 À0.07 À0.23
**
À0.07 0.14 0.17 À0.08 À0.09 NA
11. ROLAMB À0.53
***
À0.10 À0.11 À0.34
***
0.08 À0.16 À0.05 À0.22
**
À0.08 À0.17
**
0.62
12. AUTON 0.49
***
0.43
***
0.44
***
0.23
**
0.13 À0.05 0.07 0.16 À0.09 0.04 À0.37
***
0.88
13. GOALCOM À0.08 À0.14 À0.11 0.10 0.05 À0.07 0.24
**
À0.01 À0/09 À0.01 À0.01 0.02 0.70
14. FREQ 0.04 0.15 0.24
**
À0.15 0.03 0.06 0.01 0.04 À0.01 À0.08 À0.03 0.15 0.09 0.53
15. SCOPE 0.07 0.31
***
0.27
***
À0.07 0.19
*
À0.12 0.08 À0.00 À0.14 À0.07 0.06 0.19
*
0.05 0.44
***
NA
16. NONPROFIT À0.08 À0.18
*
À0.16 0.08 À0.14 0.09 À0.14 0.14 0.06 0.03 À0.01 À0.10 À0.00 À0.15 À0.01 NA
17. PUBLIC 0.02 0.14 0.13 À0.05 0.17 À0.14 À0.01 À0.12 0.03 À0.07 0.02 0.13 À0.04 0.09 0.01 À0.86
***
NA
Note: Elements on the diagonal represent the square root of the Average Variance Extracted (AVE). Off-diagonal elements are Pearson correlations among
variables. We used factor scores for multi-items scales. NA = not applicable. All signi?cance levels are two-tailed p-values.
POS perceived organizational support, RPI_SCOPE scope of relative performance information, RPI_FREQ frequency of relative performance information, PERF
relative performance, USEFUL usefulness of information, PI professional identi?cation, OCCUP occupancy rate, ROLEAMB role ambiguity, AUTON autonomy,
INCENT percentage of bonus relative to ?xed salary, PRIV_PAT proportion of private patients, FREQ frequency of information, SCOPE scope of information,
GOALCOM goal communication, NONPROFIT/PUBLIC indicator variables taking the value of 1 for for-pro?t/public hospitals and 0 otherwise.
*
Signi?cant at the 10% level.
**
Signi?cant at the 5% level.
***
Signi?cant at the 1% level.
8
In additional empirical tests (untabulated), we analysed the potential
impact of eight more control variables. These variables included tenure of
the chief physician, department size (measured as the number of beds),
Case-Mix-Index of the department, length of stay in the department,
hospital size (measured as the number of beds), teaching status of the
hospital, level of information detail (measured on the same four dimen-
sions as RPI provision), and administrative educational background of the
chief physician. These variables were not signi?cant and did not in?uence
our results. Thus, we dropped them from the ?nal analysis to improve the
ratio of cases to independent variables (Tabachnick & Fidell, 2007, p. 123).
M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347 339
Moreover, the interaction of RPI_SCOPE and PERF is sig-
ni?cant (p < 0.05). Thus, the better the relative perfor-
mance (PERF) is, the more positive is the effect of
RPI_SCOPE on POS. Finally, the interaction between
RPI_SCOPE and PI is signi?cant (p < 0.05). Thus, for higher
values of PI, the effect of RPI_SCOPE on POS is more posi-
tive than it is at lower levels of PI. As Model 2 also
includes the interactions with PERF and USEFULNESS,
the results suggest that PI has an impact on the relation-
ship between RPI_SCOPE and POS in addition to these two
other variables.
Table 4 presents the results for our second proxy for
RPI provision (RPI_FREQ). As in Table 3, the dependent
variable is POS. Model 1 shows the results with main
effects only, while Model 2 includes the interactions.
The signi?cance (p < 0.10) of the ?rst interaction term
suggests that the effect of RPI_FREQ on POS becomes more
positive with increasing values of USEFUL. Likewise, the
second interaction is signi?cant (p < 0.10), with the effect
of RPI_FREQ on POS becoming more positive with increas-
ing values of PERF as expected. The third interaction coef-
?cient referring to RPI_FREQ and PI is also signi?cant
(p < 0.05). This result suggests that PI has an effect on
the relationship between RPI_FREQ and POS after control-
ling for possible moderators of this relationship, namely
PERF and USEFUL.
Simple slope analyses
Overall, the results presented in Tables 3 and 4 ?t with
our hypotheses because they suggest that the effect of RPI
on POS becomes more positive (negative) with higher
(lower) values of the interaction variables USEFUL, PERF
and PI. To obtain a more detailed understanding of the
interaction effects presented in Tables 3 and 4, we analyse
the simple slopes (i.e., partial derivatives) (e.g., Artz,
Homburg, & Rajab, 2012). Speci?cally, we graphically rep-
resent and test the signi?cance of the simple slopes of the
effects of RPI_SCOPE and RPI_FREQ on POS at different levels
of the interaction variables. Fig. 1 shows the simple slopes,
with the ‘‘low’’ and ‘‘high’’ values of the moderators refer-
ring to values calculated for À/+ one standard deviation
from the mean values of USEFUL, PERF and PI respectively.
Panel 1 in Fig. 1 shows the simple slopes for the interac-
tion effect between RPI provision and USEFUL. The left pic-
ture depicts the effects of RPI_SCOPE (according to the
model speci?cation in Table 3, Model 2), while the right pic-
turepresents the effects of RPI_FREQ(accordingtothe model
speci?cation in Table 4, Model 2). The simple slope analysis
suggests that individuals feel supported by receiving RPI if
they consider the information provided by the organization
useful. In contrast, RPI provision exerts a negative effect on
POS for those individuals who consider the information less
Table 3
Regression analysis results – RPI_SCOPE.
Dependent variable: Perceived organizational support (POS)
Model 1 (main effects only) Model 2 (including all interaction effects)
Constant 0.553 (2.09)
**
0.305 (1.17)
RPI_SCOPE À0.048 (À0.82) À0.001 (À0.03)
USEFUL À0.004 (À0.04) À0.023 (À0.22)
PERF 0.008 (0.09) 0.012 (0.15)
PI 0.051 (0.68) À0.003 (À0.04)
INCENT 0.003 (0.74) 0.006 (1.46)
PRIV_PAT 0.015 (1.37) 0.016 (1.65)
OCCUP À0.014 (À1.65) À0.021 (À2.20)
**
AGE À0.045 (À3.36)
***
À0.045 (À3.82)
***
ROLAMB À0.393 (À4.90)
***
À0.405 (À4.23)
***
AUTON 0.372 (3.92)
***
0.374 (4.81)
***
GOALCOM À0.160 (À1.97)
**
À0.197 (À2.32)
**
FREQ À0.097 (À1.10) 0.001 (0.01)
SCOPE 0.036 (0.25) À0.195 (À1.38)
NONPROFIT À0.604 (À2.17)
**
À0.328 (À1.15)
PUBLIC À0.540 (À1.82)
*
À0.387 (À1.42)
RPI_SCOPE x USEFUL 0.103 (1.80)
*
RPI_SCOPE x PERF 0.143 (2.59)
**
RPI_SCOPE x PI 0.109 (2.41)
**
Number of observations 107 107
R
2
0.51 0.57
Note: The table reports nonstandardized coef?cients, z-values in parentheses. Robust standard errors (Huber–White sandwich estimators) are used. All
signi?cance levels are two-tailed. We used factor scores for multi-items scales.
POS perceived organizational support, RPI_SCOPE scope of relative performance information, RPI_FREQ frequency of relative performance information, PERF
relative performance, USEFUL usefulness of information, PI professional identi?cation, OCCUP occupancy rate, ROLEAMB role ambiguity, AUTON autonomy,
INCENT percentage of bonus relative to ?xed salary, PRIV_PAT proportion of private patients, FREQ frequency of information, SCOPE scope of information,
GOALCOM goal communication, NONPROFIT/PUBLIC indicator variables taking the value of 1 for for-pro?t/public hospitals and 0 otherwise.
*
Signi?cant at the 10% level.
**
Signi?cant at the 5% level.
***
Signi?cant at the 1% level.
340 M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347
useful. We calculate signi?cance levels (untabulated) at the
highest and lowest observed values of the moderator vari-
ables in the sample. For RPI_SCOPE the positive and negative
simple slopes are signi?cant (p < 0.05). For RPI_FREQ, only
the negative slope is signi?cant (p < 0.01) at a conventional
level. These results support H1.
Panel 2 in Fig. 1 exhibits the interaction of RPI provision
with PERF. The left picture depicts the effects of RPI_SCOPE
(according to the model speci?cation in Table 3, Model 2),
while the right picture presents the effects of RPI_FREQ
(according to the model speci?cation in Table 4, Model 2).
Both ?gures showthat RPI provision exerts a negative effect
on POS in the case of weak relative performance, but a posi-
tive effect in the case of good relative performance. Signif-
icance tests of the simple slopes suggest that the effect of
RPI_SCOPE is negatively signi?cant (p < 0.05) at the lowest
value of PERF in the sample and positively signi?cant
(p < 0.05) at the highest value of PERF. For RPI_FREQ, the
effects are signi?cant at p < 0.10. These results support H2.
Finally, Panel 3 shows the effect of RPI provision on POS
for individuals with low PI compared to individuals with
high PI. The left picture depicts the effects of RPI_SCOPE
(according to the model speci?cation in Table 3, Model
2), while the right picture presents the effects of RPI_FREQ
(according to the model speci?cation in Table 4, Model 2).
RPI has a negative effect at low levels of PI and a positive
effect at high levels of PI. For RPI_SCOPE, the negative slope
at the lowest sample value of PI is signi?cant at p < 0.05.
The positive slope at the highest sample values of PI is
signi?cant at p < 0.10. For RPI_FREQ, the positive slope is
signi?cant at p < 0.05 and the negative slope is signi?cant
at p < 0.01. Thus, H3 is also supported. In sum, the simple
slopes analysis indicates that the type of interactions
observed in all three panels is disordinal (Cohen, Cohen,
West, & Aiken, 2003; Luft & Shields, 2003). RPI provision
can have bene?cial or harmful relational effects depending
on the levels of perceived usefulness of RPI (USEFUL), rela-
tive performance (PERF), and professional identi?cation
(PI) perceived by the RPI recipient.
Additional analyses
To address potential selection effects, we also examine
the mean differences in performance between participants
and non-participants of the second survey to ensure that
our empirical ?ndings are not causedby a self-selectionbias
of hospitals or orthopaedics departments with above aver-
age performance. In the ?rst survey, which was addressed
to the administrative head, we ask for subjective
performance ratings regarding the hospital overall (10
performance items)
9
and the orthopaedics department in
Table 4
Regression analysis results – RPI_FREQ.
Dependent variable: Perceived organizational support (POS)
Model 1 (main effects only) Model 2 (including all interaction effects)
Constant 0.548 (2.06)
**
0.278 (1.01)
RPI_FREQ À0.089 (À1.13) À0.053 (À0.69)
USEFUL À0.011 (À0.10) À0.037 (À0.35)
PERF 0.009 (0.10) 0.008 (0.10)
PI 0.048 (0.64) 0.010 (0.14)
INCENT 0.003 (0.74) 0.006 (1.54)
PRIV_PAT 0.014 (1.35) À0.178 (À1.68)
*
OCCUP À0.014 (À1.70)
*
À0.020 (À2.31)
**
AGE À0.045 (À3.37)
***
À0.045 (À3.81)
***
ROLAMB À0.393 (À4.94)
***
À0.390 (À4.05)
***
AUTON 0.381 (4.03)
***
0.405 (5.24)
***
GOALCOM À0.161 (À1.96)
**
À0.213 (À2.55)
**
FREQ À0.086 (À0.99) À0.024 (À0.29)
SCOPE 0.026 (0.18) À0.129 (À0.92)
NONPROFIT À0.598 (À2.16)
**
À0.314 (À1.08)
PUBLIC À0.548 (À1.80)
*
À0.270 (À0.90)
RPI_FREQ x USEFUL 0.146 (1.81)
*
RPI_FREQ x PERF 0.158 (1.94)
*
RPI_FREQ x PI 0.154 (2.32)
**
Number of observations 107 107
R
2
0.51 0.57
Note: The table reports nonstandardized coef?cients, z-values in parentheses. Robust standard errors (Huber–White sandwich estimators) are used. All
signi?cance levels are two-tailed. We used factor scores for multi-items scales.
POS perceived organizational support, RPI_SCOPE scope of relative performance information, RPI_FREQ frequency of relative performance information, PERF
relative performance, USEFUL usefulness of information, PI professional identi?cation, OCCUP occupancy rate, ROLEAMB role ambiguity, AUTON autonomy,
INCENT percentage of bonus relative to ?xed salary, PRIV_PAT proportion of private patients, FREQ frequency of information, SCOPE scope of information,
GOALCOM goal communication, NONPROFIT/PUBLIC indicator variables taking the value of 1 for for-pro?t/public hospitals and 0 otherwise.
*
Signi?cant at the 10% level.
**
Signi?cant at the 5% level.
***
Signi?cant at the 1% level.
9
These 10 items comprise ratings at the hospital level regarding
earnings per case, adherence to the budget, pro?t margin, administrative
expenses, expenditures per case, overall quality of care, patient satisfaction,
average length of stay, average occupancy, and reputation. These items
were measured with a 5-point Likert scale (ranging from 1 = well below
average to 5 = well above average).
M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347 341
particular (nine performance items measuring PERF as
described in the variable measurement section). We compare
the 107 hospitals for which we obtained responses from the
chief physicians in the second survey with the non-responding
hospitals. Speci?cally, we compare the overall performance
rating of the hospital and the performance rating of the
orthopaedics department between respondents and non-
respondents and do not ?nd signi?cant differences.
To distinguish the speci?c effects of RPI from the
more general effects of the communication of business-
related information, our main analyses includes the control
variables FREQandSCOPEtorepresent informationprovided
by the organization and GOAL to represent information
about communicating organizational goals. While our main
analysis includes only these variables as main effects, we
also replicate the analysis by including the interaction
effects between PI and the three variables related to more
general communication characteristics. The results remain
stable withthe inclusionof the additional interactionterms.
Furthermore, we check the robustness of our measure-
ment of RPI_FREQ by applying two transformations. The
?rst transformation consists of a recoding into the actual
Panel 1: Simple slopes for Usefulness (USEFUL) interactions (H1)
Panel 2: Simple slopes for Performance (PERF) interactions (H2)
Panel 3: Simple slopes for Professional Identification (PI) interactions (H3)
Note: In all three panels, the simple slopes are depicted for +/? one standard deviation from the
mean values of the moderators. Similarly, Low RPI represents minus one standard deviation from
the mean of RPI and High RPI represents plus one standard deviation from the mean.
Fig. 1. Simple slopes (marginal effects) of the interactions.
342 M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347
frequency of RPI per year, ranging from 0 (no RPI provision
at all) to 365 (daily provision of RPI). When substituting
our initial measurement of RPI_FREQ in Table 3 Model 2
with this recoded measure, the interactions become insig-
ni?cant. This suggests that the effects of RPI_FREQ might be
nonlinear. For the second transformation, we take the log-
arithm of our recoded measure. We then repeat the analy-
sis and ?nd that all interactions turn signi?cant (p < 0.10),
which provides further support that the effect of RPI_FREQ
is indeed nonlinear.
Discussion and conclusions
The literature on RPI has primarily focused on two
effects. First, relying on agency theory, RPI has been inves-
tigated with an emphasis on reducing common uncertainty
from agents’ contracts under different incentive schemes
(Frederickson, 1992; Holmstrom, 1982). Second, drawing
from social comparison theory, prior studies have explored
the motivational effects of RPI (Festinger, 1954;
Frederickson, 1992; Hannan et al., 2008, 2013). This study
takes a different approach by exploring the in?uence of RPI
provision on employees’ attitudes towards the organiza-
tion, which we refer to as the relational effect of RPI.
By relying on insights from social psychology, we inves-
tigate the conditions under which RPI recipients develop a
perception that they are supported, encouraged and cared
for by their employers. We posit that variations in POS
occur under speci?c conditions. First, it appears that the
relevance of the comparison domain triggered by another’s
performance (captured by the perceived usefulness of
information) determines the relative importance of a social
comparison process. Individuals tend to experience higher
levels of POS in situations where RPI provides them more
useful, self-relevant comparisons. Second, individuals per-
ceive RPI provision as supportive if their relative rankings
score high above par, whereas RPI turns unsupportive in
case of poor performance. Our results point to a contrast
effect. Comparing individuals to superior others (i.e.,
upward comparisons) can result in negative affective reac-
tions to the extent that the comparison threatens individ-
ual self-esteem and undermines the relational ties with the
organization. On the contrary, comparing individuals to
inferior others (i.e., downward comparisons) can result in
positive affective reactions that enhance the perception
of receiving bene?ts from RPI provision. Thus, our ?ndings
show that perceived usefulness of information and relative
performance are signi?cant interaction variables regarding
the relationship between RPI provision and POS. This ?nd-
ing is in line with the self-evaluation model elaborated by
Tesser (1988) as an extension to Festinger’s (1954) social
comparison theory.
Our study further documents that professional identity
in?uences social comparison processes among employees
and interacts with RPI provision in affecting individual vari-
ations in POS in addition to the other interacting effects.
These conditional effects emerge while controlling for types
of performance information provision other than RPI in our
setting, as well as includingavarietyof individual- andorga-
nization-level control variables. Furthermore, our analysis is
robust to different speci?cations of RPI provision intended
to capture either the timing (RPI frequency) or content
(RPI scope) of the information released to the chief physi-
cians in our sample. We thus provide evidence that the per-
ceived similarity to those about whom one receives
comparative information, which is captured by the degree
of professional identi?cation, is relevant in making social
comparisons effective (Festinger, 1954; Wood, 1989). This
relational effect indicates that the provision of RPI as a feed-
back mechanism affects the social ties between individuals
and their employers, which subsequently impacts behav-
iour. Our results further suggest that the provision of RPI
does not necessarily exclusively entail dysfunctional conse-
quences. The disordinal interactions predicted in our
hypotheses and supportedin our empirical analysis demon-
strate that RPI effects are conditional on contingent
variables, among which professional identity appears to be
a signi?cant moderator.
This study contributes to accounting research on RPI in
two ways. First, we examine POS as a key reaction to social
comparisons, thereby complementing previous ?ndings on
RPI effectiveness that primarily drew on economic theory.
More speci?cally, we incorporate social identity theory to
examine how RPI recipients’ attitudes towards their
employers vary across settings in which individuals per-
ceive different levels of similarities between themselves
and their peer groups. We thus contribute to the under-
standing of the functioning of RPI in organizational settings
by revealing a relational effect associated with RPI provi-
sion. This relational effect complements previous ?ndings
by providing insights from social psychology theories on
responses to social comparison information. Second, by
illustrating how RPI can positively in?uence chief physi-
cians’ POS, we contribute to research on managing and con-
trolling professionals. Most of the previous management
accounting studies conducted in professional organizations
primarily focused on the potential con?icts and dysfunc-
tional outcomes of bureaucratic forms of management con-
trol systems (Abernethy&Stoelwinder, 1990a, 1990b, 1995;
Comerford &Abernethy, 1999; Wallace, 1995). Our ?ndings
show that RPI systems have the potential to strengthen the
relationship between professionals and their employer,
since providing feedback and relative ranks ful?ls informa-
tion needs related to decision-making, internal and external
justi?cations and the general human need to self-evaluate.
Thus, our ?ndings suggest that RPI can have bene?cial
effects in a professional setting, despite potential profes-
sional-bureaucratic con?icts.
Our ?ndings also provide insights for management
accountants interested in the design of effective perfor-
mance measurement and internal reporting systems. Our
results are particularly informative regarding the potential
social ties of employees towards groups within or outside
the company that might affect their behavioural responses
to RPI. Depending on the knowledge that organizations
possess regarding their employees’ professional identi?ca-
tion, management accountants can better decide whether
to provide RPI and to whom.
This study is subject to several limitations. The survey
data do not allow us to distinguish whether the RPI systems
contain information on other departments in the same
hospital or similar departments in other hospitals, thus
M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347 343
addingnoisetoour measurement of RPI. Futurestudies might
attempt to explore potential differences between internal
and external RPI. Further, while we make claims concerning
department managers in general, our study is limited to
chief physicians employed in a hospital setting with a dual
responsibility of managing their department and being an
expert intheir medical ?eld. As such, the impact of RPI might
be in?uenced by the fact that chief physicians have rela-
tively little experience in managing a department compared
to conducting their professional activities, even though we
control for positional tenure. This might in?uence chief
physicians’ perceptions of RPI on their department’s perfor-
mance, because this information might primarily cater to
their need to acquire management knowledge and reduce
uncertainty. Further, while we specify and test alternative
models to assess potential issues with model speci?cation,
we acknowledge that our research design suffers the typical
limitations of cross-sectional, non-experimental data.
Disentangling possible effects caused by multiple respon-
sibilities (managerial versus professional) and RPI on differ-
ent levels (individual versus departmental) provides rich
opportunities for future studies. This also applies to possible
effects caused by the source providing the RPI, which might
be distal (administrators providing RPI about professionals,
as in our setting) or proximal (professionals providing RPI
about other professionals). In addition, prior research indi-
cates that the diffusion of RPI (public versus private) might
act as a moderator on its possible effects (Hannan et al.,
2013), an angle which our paper does not examine. In sum-
mary, while this studyadds tothe literature onRPI andman-
aging professionals, further research on the source, the
content, the diffusion and the recipients of RPI is warranted.
Acknowledgments
Weare grateful toMikeShields (the Editor) andtwoanon-
ymous journal referees for their helpful and constructive
comments. In addition, we would like to thank Utz Schäffer
and Jürgen Weber for their support throughout the project
and acknowledge Martin Artz, Steffen Giessner, Igor
Goncharov, Frank Hartmann, and Daan van Knippenberg
for their feedback on earlier drafts of this paper. Further,
we would like to thank Markus Arnold, Gary Cunningham,
Lynn Hannan, Stephan Kramer, Jochen Rehring, and all par-
ticipants in the workshops at Michigan State University,
RSM Erasmus University, and WHU – Otto Beisheim School
of Management, and participants at the Management
Accounting Research Group Conference 2011, the Annual
Conference for Management Accounting Research 2012
and the Annual Conference of the European Accounting
Association 2012 for their valuable suggestions.
Appendix A. Measurement instruments
The instruments that were included in the question-
naires are presented below. Survey items that are reversely
coded contain the symbol ‘(RC)’.
First survey: questionnaire submitted to the hospi-
tal’s administrative head.
Frequency of RPI provision (RPI_FREQ)
How often does your administration report information
to your chief physicians (e.g., the head of the
orthopaedics department) which compares their
department performance with the performance of
other departments (benchmarking information)?
Seven-point scale: 1 = not reported, 2 = annually, 3 =
semi-annually, 4 = quarterly, 5 = monthly, 6 =
weekly, 7 = daily
1. Benchmarking concerning medical performance
indicators (e.g., Case-Mix-Index, number of cases,
etc.)
2. Benchmarking concerning ?nancial performance
indicators (e.g., earnings, expenditures, etc.)
3. Benchmarking concerning quality of care
indicators (e.g., patient satisfaction, adherence to
quality standards, etc.)
4. Benchmarking concerning ef?ciency indicators
(e.g., patient length of stay, department occupancy,
etc.)
Performance of orthopaedics department (PERF)
For the following statements, please focus on the
orthopaedics department - respectively the
department providing orthopaedic services. How do
you rank this department’s performance in
comparison to comparable departments in other
hospitals?
Five-point scale: from 1 = below average to 5 = above
average
1. Earnings per case
2. Adherence to budget
3. Pro?t margin
4. Expenditures per case
5. Overall quality of care
6. Patient satisfaction
7. Average length of stay
8. Average occupancy (use of capacity)
9. Reputation
Frequency of information provision (FREQ)
How often do your chief physicians (e.g., the head of the
orthopaedics department) receive information by the
administration about their department concerning . . .
Seven-point scale: 1 = not reported, 2 = annually, 3 =
semi-annually, 4 = quarterly, 5 = monthly, 6 =
weekly, 7 = daily
1. . . . medical performance indicators (e.g., Case-
Mix-Index, number of cases, etc.)?
2. . . . ?nancial performance indicators (e.g., earnings,
expenditures, etc.)?
3. . . . quality of care indicators (e.g., patient
satisfaction, adherence to quality standards, etc.)?
4. . . . ef?ciency indicators (e.g., patient length of stay,
department occupancy, etc.)?
344 M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347
Communication of goals (GOALCOM)
To what extent does the administration communicate
goals to the chief-physicians (e.g., the head of the
orthopaedics department) concerning . . .
Five-point scale: from 1 = no extent to 5 = great extent
1. . . . medical performance indicators (e.g., Case-
Mix-Index, number of cases, etc.)?
2. . . . ?nancial performance indicators (e.g., earnings,
expenditures, etc.)?
3. . . . quality of care indicators (e.g., patient
satisfaction, adherence to quality standards, etc.)?
4. . . . ef?ciency indicators (e.g., patient length of stay,
department occupancy, etc.)?
Second survey: questionnaire submitted to the chief
physician heading the orthopaedics department.
Perceived organizational support (POS)
To what extent do you agree with the following
statements?
Five-point scale: from 1 = strongly disagree to 5 =
strongly agree.
1. My hospital cares about my opinion.
2. My hospital really cares about my well-being.
3. My hospital strongly considers my goals and
values.
4. Help is available from my hospital when I have a
problem.
5. My hospital would forgive an honest mistake on
my part.
6. If given the opportunity, my hospital would take
advantage of me. (RC)
7. My hospital shows very little concern for me. (RC)
8. My hospital is willing to help me if I need a special
favor.
Usefulness of information (USEFUL)
The provision of medical performance information (e.g.,
Case-Mix-Index or case numbers) . . .
Five-point scale: from 1 = strongly disagree to 5 =
strongly agree.
1. . . . increases my productivity.
2. . . . improves the quality of the work I do.
3. . . . gives me greater control over my work.
4. Overall, I ?nd the provision of medical information
useful in my job.
The provision of ?nancial performance information (e.g.,
earnings, expenditures, etc.) . . .
Five-point scale: from 1 = strongly disagree to 5 =
strongly agree.
1. . . . increases my productivity.
2. . . . improves the quality of the work I do.
3. . . . gives me greater control over my work.
4. Overall, I ?nd the provision of ?nancial
information useful in my job.
The provision of quality of care information (e.g., patient
satisfaction, adherence to quality standards, etc.) . . .
Five-point scale: from 1 = strongly disagree to 5 =
strongly agree.
1. . . . increases my productivity.
2. . . . improves the quality of the work I do.
3. . . . gives me greater control over my work.
4. Overall, I ?nd the provision of quality of care
information useful in my job.
The provision of ef?ciency information (e.g., patient
length of stay, department occupancy, etc.) . . .
Five-point scale: from 1 = strongly disagree to 5 =
strongly agree.
1. . . . increases my productivity.
2. . . . improves the quality of the work I do.
3. . . . gives me greater control over my work.
4. Overall, I ?nd the provision of ef?ciency
information useful in my job.
Professional identi?cation (PI)
To what extent do you agree with the following
statements?
Five-point scale: from 1 = strongly disagree to 5 =
strongly agree.
1. When someone criticizes doctors, it feels like a
personal insult.
2. When I talk about doctors, I usually say ’we’ rather
than ’they’.
3. When someone praises doctors, it feels like a
personal compliment.
4. Medicine’s successes are my successes.
5. If a story in the media criticized doctors, I would
feel embarrassed.
Role ambiguity (ROLAMB)
To what extent do you agree with the following
statements?
Five-point scale: from 1 = strongly disagree to 5 =
strongly agree.
1. I feel certain about how much authority I have.
2. I have clear, planned goals and objectives for my
jobs.
3. I know that I have divided my time properly.
4. I know what my responsibilities are.
5. I know exactly what is expected of me.
6. I receive a clear explanation of what has to be
done.
Job autonomy (AUTON)
To what extent do you agree with the following
statements?
Five-point scale: from 1 = strongly disagree to 5 =
strongly agree.
1. I have signi?cant autonomy in determining how I
do my job.
2. I can decide on my own how to go about doing my
work.
3. I have considerable opportunity for independence
and freedom in how I do my job.
M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347 345
References
Abernethy, M. A., & Stoelwinder, J. U. (1990a). Physicians and resource
management in hospitals: An empirical investigation. Financial
Accountability & Management, 6(1), 17–31.
Abernethy, M. A., & Stoelwinder, J. U. (1990b). The relationship between
organisation structure and management control in hospitals: An
elaboration and test of Mintzberg’s professional bureaucracy model.
Accounting Auditing and Accountability Journal, 3(3), 18–33.
Abernethy, M. A., & Stoelwinder, J. U. (1995). The role of professional
control in the management of complex organizations. Accounting,
Organizations and Society, 20(1), 1–17.
Achee, J., Tesser, A., & Pilkington, C. (1994). Social-perception – A test of
the role of arousal in self-evaluation maintenance processes. European
Journal of Social Psychology, 24(1), 147–159.
Aggarwal, R. K., & Samwick, A. A. (1999). Executive compensation,
strategic competition, and relative performance evaluation: Theory
and evidence. Journal of Finance, 54(6), 1999–2043.
Albuquerque, A. (2009). Peer ?rms in relative performance evaluation.
Journal of Accounting & Economics, 48(1), 69–89.
Antle, R., & Smith, A. (1986). An empirical-investigation of the relative
performance evaluation of corporate-executives. Journal of Accounting
Research, 24(1), 1–39.
Artz, M., Homburg, C., & Rajab, T. (2012). Performance-measurement
system design and functional strategic decision in?uence: The role of
performance-measure properties. Accounting, Organizations and
Society, 37(7), 445–460.
Aselage, J., & Eisenberger, R. (2003). Perceived organizational support and
psychological contracts: A theoretical integration. Journal of
Organizational Behavior, 24(5), 491–509.
Ashforth, B. E., & Mael, F. (1989). Social identity theory and the
organization. Academy of Management Review, 14(1), 20–39.
Azmat, G., & Iriberri, N. (2010). The importance of relative performance
feedback information: Evidence from a natural experiment using high
school students. Journal of Public Economics, 94(7–8), 435–452.
Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation
models. Journal of the Academy of Marketing Science, 16(1), 74–94.
Baines, A., & Lang?eld-Smith, K. (2003). Antecedents to management
accounting change: A structural equation approach. Accounting,
Organizations and Society, 28(7–8), 675–698.
Baran, B. E., Shanock, L. R., & Miller, L. R. (2012). Advancing organizational
support theory into the twenty-?rst century world of work. Journal of
Business and Psychology, 27(2), 123–147.
Blau, P. M. (1964). Exchange and power in social life. New York, NY: Wiley.
Brewer, M. B., & Gardner, W. (1996). Who is this ‘‘we’’? Levels of collective
identity and self representations. Journal of Personality and Social
Psychology, 71(1), 83–93.
Brown, D. J., Ferris, D. L., Heller, D., & Keeping, L. M. (2007). Antecedents
and consequences of the frequency of upward and downward social
comparisons at work. Organizational Behavior and Human Decision
Processes, 102(1), 59–75.
Burney, L., Henle, C. A., & Widener, S. K. (2009). A path model examining
the relations among strategic performance measurement system
characteristics, organizational justice, and extra- and in-role
performance. Accounting, Organizations and Society, 34(3–4), 305–321.
Burney, L., & Widener, S. K. (2007). Strategic performance measurement
systems, job-relevant information, and managerial behavioral
responses - role stress and performance. Behavioral Research in
Accounting, 19, 43–69.
Buunk, B. P., & Gibbons, F. X. (2007). Social comparison: The end of a
theory and the emergence of a ?eld. Organizational Behavior and
Human Decision Processes, 102(1), 3–21.
Buunk, B. P., Zurriaga, R., Peiro, J. M., Nauta, A., & Gosalvez, I. (2005). Social
comparisons at work as related to a cooperative social climate and to
individual differences in social comparison orientation. Applied
Psychology-an International Review, 54(1), 61–80.
Chenhall, R. H., & Lang?eld-Smith, K. (1998). The relationship between
strategic priorities, management techniques and management
accounting: An empirical investigation using a systems approach.
Accounting, Organizations and Society, 23(3), 243–264.
Cichello, M. S., Fee, C. E., Hadlock, C. J., & Sonti, R. (2009). Promotions,
turnover, and performance evaluation: Evidence from the careers of
division managers. The Accounting Review, 84(4), 1119–1143.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple
regression/correlation analysis for the behavioral sciences. Mahwah, NJ:
Lawrence Erlbaum Associates.
Collins, R. L. (1996). For better or worse: The impact of upward social
comparison on self-evaluations. Psychological Bulletin, 119(1),
51–69.
Collins, R. L. (2000). Among the better ones: Upward assimilation in social
comparison. In J. Suls & L. Wheeler (Eds.), Handbook of social
comparison: Theory and research (pp. 159–172). New York: Kluwer/
Plenum Press.
Comerford, S. E., & Abernethy, M. A. (1999). Budgeting and the
management of role con?ict in hospitals. Behavioral Research in
Accounting, 11, 93–110.
Corcoran, K., Crusius, J., & Mussweiler, T. (2011). Social comparison:
Motives, standards, and mechanisms. In D. Chadee (Ed.), Theories in
Social Psychology. Oxford, UK: Wiley-Blackwell.
Coyle-Shapiro, J. A. M., & Shore, L. M. (2007). The employee-organization
relationship: Where do we go from here? Human Resource
Management Review, 17(2), 166–179.
Cropanzano, R., & Mitchell, M. S. (2005). Social exchange theory: An
interdisciplinary review. Journal of Management, 31(6), 874–900.
Dalal, D. K., & Zickar, M. J. (2012). Some common myths about centering
predictor variables in moderated multiple regression and polynomial
regression. Organizational Research Methods, 15(3), 339–362.
Destatis (2013). Gesundheit – Grunddaten der Krankenhäuser.
Retrieved 24. 02. 14.
Dikolli, S. S., Hofmann, C., & Pfeiffer, T. (2013). Relative Performance
Evaluation and peer-performance summarization errors. Review of
Accounting Studies, 18(1), 34–65.
Dillman, D. A., Smyth, J. D., & Christian, L. M. (2009). Internet, mail, and
mixed-mode surveys: The tailored design method (3rd ed.). Hoboken, NJ:
John Wiley & Sons Inc.
Eddleston, K. A. (2009). The effects of social comparisons on managerial
career satisfaction and turnover intentions. Career Development
International, 14(1), 87–110.
Eisenberger, R., Armeli, S., Rexwinkel, B., Lynch, P. D., & Rhoades, L. (2001).
Reciprocation of perceived organizational support. Journal of Applied
Psychology, 86(1), 42–51.
Eisenberger, R., Cummings, J., Armeli, S., & Lynch, P. D. (1997). Perceived
organizational support, discretionary treatment, and job satisfaction.
Journal of Applied Psychology, 82(5), 812–820.
Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Perceived
organizational support. Journal of Applied Psychology, 71(3), 500–507.
Eldenburg, L. (1994). The use of information in total cost management.
The Accounting Review, 69(1), 96–121.
Eriksson, T., Poulsen, A., & Villeval, M. C. (2009). Feedback and incentives:
Experimental evidence. Labour Economics, 16(6), 679–688.
Festinger, L. (1954). A theory of social comparison processes. Human
Relations, 1, 117–140.
Foa, E., & Foa, U. (1980). Resource theory: Interpersonal behavior as
exchange. In K. J. Gergen, M. S. Greenberg, & R. H. Willis (Eds.), Social
exchange: Advances in theory and research (pp. 77–94). New York, NY:
Plenum Press.
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models
with unobservable variables and measurement error. Journal of
Marketing Research, 18(1), 39–50.
Frederickson, J. R. (1992). Relative performance information: The effects
of common uncertainty and contract type on agent effort. The
Accounting Review, 67(4), 647–669.
Gong, G. J., Li, L. Y., & Shin, J. Y. (2011). Relative performance evaluation
and related peer groups in executive compensation contracts. The
Accounting Review, 86(3), 1007–1043.
Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement.
American Sociological Review, 25(2), 161–178.
Greenberg, J., Ashton-James, C. E., & Ashkanasy, N. M. (2007). Social
comparison processes in organizations. Organizational Behavior and
Human Decision Processes, 102(1), 22–41.
Hall, M. (2008). The effect of comprehensive performance measurement
systems on role clarity, psychological empowerment and managerial
performance. Accounting, Organizations and Society, 33(2/3),
141–163.
Hannan, R. L., Krishnan, R., & Newman, A. H. (2008). The effects of
disseminating relative performance feedback in tournament and
individual performance compensation plans. The Accounting Review,
83(4), 893–913.
Hannan, R. L., McPhee, G. P., Newman, A. H., & Tafkov, I. D. (2013). The
effect of relative performance information on performance and effort
allocation in a multi-task environment. The Accounting Review, 88(2),
553–575.
Hekman, D. R., Bigley, G. A., Steensma, H. K., & Hereford, J. F. (2009a).
Combined effects of organizational and professional identi?cation on
the reciprocity dynamic for professional employees. Academy of
Management Journal, 52(3), 506–526.
346 M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347
Hekman, D. R., Steensma, H. K., Bigley, G. A., & Hereford, J. F. (2009b).
Effects of organizational and professional identi?cation on the
relationship between administrators’ social in?uence and
professional employees’ adoption of new work behavior. Journal of
Applied Psychology, 94(5), 1325–1335.
Hogg, M. A., & Terry, D. J. (2000). Social identity and self-categorization
processes in organizational contexts. Academy of Management Review,
25(1), 121–140.
Holmstrom, B. (1982). Moral hazard in teams. The Bell Journal of
Economics, 13(2), 324–340.
Holzhacker, M., Krishnan, R., & Mahlendorf, M. D. (2014). The impact of
changes in regulation on cost behavior. Contemporary Accounting
Research (in press).http://dx.doi.org/10.1111/1911-3846.12082.
Huber, P. J. (1967). The behavior of maximum likelihood estimates under
nonstandard conditions. Paper presented at the Proceedings of the ?fth
Berkeley symposium on mathematical statistics and probability.
Hutchison, S., & Garstka, M. L. (1996). Sources of perceived organizational
support: Goal setting and feedback. Journal of Applied Social
Psychology, 26(15), 1351–1366.
Kuhnen, C. M., & Tymula, A. (2012). Feedback, self-esteem, and
performance in organizations. Management Science, 58(1), 94–113.
Legris, P., Ingham, J., & Collerette, P. (2003). Why do people use
information technology? A critical review of the technology
acceptance model. Information & Management, 40(3), 191–204.
Levinson, H. (1965). Reciprocation: The relationship between man and
organization. Administrative Science Quarterly, 9(4), 370–390.
Locke, E. A., & Latham, G. P. (1990). A theory of goal setting & task
performance. Englewood Cliffs, NJ: Prentice-Hall Inc.
Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of
goal setting and task motivation: A 35-year odyssey. American
Psychologist, 57(9), 705–717.
Luft, J., & Shields, M. D. (2003). Mapping management accounting:
Graphics and guidelines for theory-consistent empirical research.
Accounting Organizations and Society, 28(2–3), 169–249.
Maas, V. S., & Matejka, M. (2009). Balancing the dual responsibilities of
business unit controllers: Field and survey evidence. The Accounting
Review, 84(4), 1233–1253.
Mael, F., & Ashforth, B. (1992). Alumni and their alma mater: A partial test
of the reformulated model of organizational identi?cation. Journal of
Organizational Behavior, 13(2), 103–123.
Murthy, U. S., & Schafer, B. A. (2011). The effects of relative performance
information and framed information systems feedback on
performance in a production task. Journal of Information Systems,
25(1), 159–184.
Pizzini, M. J. (2006). The relation between cost-system design, managers’
evaluations of the relevance and usefulness of cost data, and ?nancial
performance: An empirical study of US hospitals. Accounting,
Organizations and Society, 31(2), 179–210.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003).
Common method biases in behavioral research: A critical review of
the literature and recommended remedies. Journal of Applied
Psychology, 88(5), 879–903.
Raykov, T. (2005). Analysis of longitudinal studies with missing data
using covariance structure modeling with full-information maximum
likelihood. Structural Equation Modeling, 12(3), 493–505.
Rhoades, L., & Eisenberger, R. (2002). Perceived organizational support: A
review of the literature. Journal of Applied Psychology, 87(4), 698–714.
Riketta, M. (2005). Organizational identi?cation: A meta-analysis. Journal
of Vocational Behavior, 66(2), 358–384.
Rizzo, J. R., House, R. J., & Lirtzman, S. I. (1970). Role con?ict and
ambiguity in complex organizations. Administrative Science Quarterly,
15(2), 150–163.
Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of
the art. Psychological Methods, 7, 147–177.
Smith, R. (2000). Assimilative and contrastive emotional reactions to
upward and downward social comparisons. In J. Suls & L. Wheeler
(Eds.), Handbook of social comparison: Theory and research
(pp. 173–200). New York: Kluwer Academic/Plenum.
Spector, P. E. (2006). Method variance in organizational research: Truth or
urban legend? Organizational Reserach Methods, 9(2), 221–232.
Spreitzer, G. M. (1995). Psychological empowerment in the workplace:
Dimensions, measurement, and validation. Academy of Management
Journal, 38(5), 1442–1465.
Strehl, R. (2002). In M. Arnold, J. Klauber, & H. Schellschmidt (Eds.),
Reformbedarf bei der Chefarzt- und Hochschullehrervergütung in
Krankenhäusern und Universitätsklinika. Stuttgart: Schattauer.
Suls, J., Martin, R., & Wheeler, L. (2002). Social comparison: Why, with
whom, and with what effect? Current Directions in Psychological
Science, 11(5), 159–163.
Suls, J., & Wheeler, L. (2000). Handbook of social comparison: Theory and
research. New York: Kluwer Academic/Plenum.
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5 ed.).
Boston: Pearson.
Tafkov, I. D. (2013). Private and public relative performance information
under different compensation contracts. The Accounting Review, 88(1),
327–350.
Tesser, A. (1988). Toward a self-evaluation maintenance model of social
behavior. In L. Berkowitz (Ed.). Advances in Experimental
Social Psychology (Vol. 21, pp. 181–227). New York, NY: Academic
Press.
Tesser, A., & Campbell, J. (1980). Self-de?nition: The impact of the relative
performance and similarity of others. Social Psychology Quarterly,
43(3), 341–347.
Tesser, A., Millar, M., & Moore, J. (1988). Some affective consequences of
social comparison and re?ection processes: The pain and pleasure of
being close. Journal of Personality and Social Psychology, 54(1),
49–61.
Tesser, A., Pilkington, C. J., & Mcintosh, W. D. (1989). Self-evaluation
maintenance and the mediational role of emotion – The perception of
friends and strangers. Journal of Personality and Social Psychology,
57(3), 442–456.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S.
(1987). Rediscovering the social group: A self-categorization theory.
Oxford: Blackwell.
Vidal, J. B. I., & Nossol, M. (2011). Tournaments without prizes:
Evidence from personnel records. Management Science, 57(10),
1721–1736.
Vidyarthi, P. R., Liden, R. C., Anand, S., Erdogan, B., & Ghosh, S. (2010).
Where do I stand? Examining the effects of leader–member exchange
social comparison on employee work behaviors. Journal of Applied
Psychology, 95(5), 849–861.
Wallace, J. E. (1995). Organizational and professional commitment in
professional and nonprofessional organizations. Administrative
Science Quarterly, 40(2), 228–255.
Wheeler, L. (1966). Motivation as a determinant of upward comparison.
Journal of Experimental Social Psychology, 1(Supplement 1), 27–31.
White, H. (1980). A heteroskedasticity-consistent covariance matrix
estimator and a direct test for heteroskedasticity. Econometrica,
48(4), 817–838.
Wood, J. V. (1989). Theory and research concerning social
comparisons of personal attributes. Psychological Bulletin, 106(2),
231–248.
Wood, J. V. (1996). What is social comparison and how should we study
it? Personality and Social Psychology Bulletin, 22(5), 520–537.
M.D. Mahlendorf et al. / Accounting, Organizations and Society 39 (2014) 331–347 347

doc_812127257.pdf