Management Control Systems and Perceived

Stress in a Public Service Organization

UPPSALA UNIVERSITY

Department of Business Studies

Master Thesis

Spring Semester 2012

Author: Claes Palm

Supervisor: Lars Frimanson

Date of submission: 2012-05-25

Management Control Systems and Perceived Stress in a Public Service Organization

Claes Palm

Keywords: Perceived Stress, Management Control, Employee Control, Feedback, Coping

Organizational stress research has demonstrated the role of employee control in buffering the

impact of work demands, such as time pressure, on stress (Häuser, Mojzisch, Niesel

& Schulz-Hardt, 2010; van der Doef & Maes, 1999). Public service organizations have

increased their management control (Diefenbach, 2009; Hood, 1991). It has further been

suggested that some types of management control come at the expense of employee control

(Karasek & Theorell, 1990). In the context of a public service organization, this study

examines how management control and employee control relate to stress.

Stress is the single most common cause of long term sick leave in Sweden, resulting in high

expenses for the Swedish society (Sverigesradio.se, 2012). Stress is related to, for example,

job dissatisfaction, accidents and absenteeism which lead to high costs for organizations (e.g.,

Cooper & Cartwright, 1984; Ongori & Agolla, 2008).

In the last two-three decades, public organizations have gone through transformations to be

more transparent, installing more auditing, accounting and visualizing technologies.

Employees are more explicitly evaluated for their performance and effectiveness. In other

A popular notion is that an employee that experiences low control

together with high demand is more likely to perceive stress.

Management control has been intensified in public service

organizations after New Public Management reforms, which is

presumed to come at the expense of employee control. This study

examined how management control systems, as a package and as

specific components, are related to perceived stress. 130 subordinates

in a Swedish public service organization completed self-report

measures. A multiple regression analysis gave support for the

hypotheses that work demand is positively and feedback from

superior is negatively related to stress. No support was found for the

hypotheses that employee control, feedback from the information

system and formality by performance evaluation should be negatively

related to stress. It is suggested that management control systems can

serve as support for the employees´ efforts of coping with the

demands.

2

words, management control has been intensified in public organizations. These trends have

been called the New Public Management (NPM) (Blomgren, 2007; Hood, 1991; Shore &

Wright, 1999).

Management control can be exercised through a range of various devices and systems,

summarized as ”management control systems” (MCS), which purpose is to ensure that the

employees behave and take decisions in line with the objectives and the overall strategy, or

rather that the deviations from these are kept as low as possible (Anderson & O`Reilley, 1981;

Green & Welsh, 1984; Malmi & Brown, 2008). It has been suggested that the components of

a MCS should be studied as a package, and not in isolation, when examining the relation to

employee behavior. The reason is that the components are interconnected, their effect tend to

depend on one another (Malmi & Brown, 2008; Otley, 1980). But there are certain aspects

which are central in MCSs, such as the focus on continuous feedback and performance

evaluation (Hood, 1991; Malmi & Brown, 2008).

Increased stress due to NPM reforms has mainly been explained as a result of cost reductions

and organizational change (Härenstam, Bejerot, Leijon, Scheele & Waldenstrom, 2004). The

purpose of this study is to extend the research by examining how increased management

control in a public service organization relates to stress. MCS are studied as a package

considered to be related to employee control and perceived stress. Furthermore, three specific

MCS components and their relation to perceived stress is examined; feedback from superior,

feedback from the information system and formality of performance evaluation.

Literature review

There are many definitions of stress (Cox, 1993; Hobfoll, 1989), implying that one has to be

clear of what position one adhere to. Lazarus and Folkman (1984) made a review of

contemporary stress research. They mean that many of the definitions can be categorized as

either stimulus or response definitions, which in turn have caused semantic difficulties.

Stimulus definitions focus on external events that are seen as normatively stressful, as for

example, natural disasters or the death of a close relative. The problem with the stimulus

definitions is that there are large variations in personal reaction to those events that are not

extreme. These definitions would therefore limit stress to extreme events and, for the purpose

of this paper, would be useless as stress here is studied in relation to MCS (i.e. everyday life

events) in organizations.

3

Lazarus and Folkman (1984) mean that the other types of definitions, response definitions,

often are used in biology and medicine research. These focus on an internal “state of stress” or

behavior, for example that the individual is “reacting with stress” or being “distressed”. Such

a definition makes it difficult to identify in advance if an event is a stressor or not, as one has

to wait for the person´s reaction to make a judgment. Another problem is that responses can

faulty be considered to be states of stress when they are not. Furthermore, both stimulus and

response definitions tend to depend on one another with circular reasoning. For an event to be

a stressor, it has to result in a stress response, and for something to be a stress response it has

to be preceded by a stressor.

Making an analogy with disease and illness in contemporary epidemiological research,

Lazarus and Folkman (1984) suggest another definition. People do not become ill or die

merely from virus and bacteria, but by being vulnerable to them. Thus, when defining stress

one also has to take into account the characteristics of the person, her or his vulnerability.

This person-environment relationship is mediated by the processes of cognitive appraisal.

Cognitive appraisal is here the conscious or unconscious process of evaluating the

environment with regard to his or her resources. Lazarus and Folkman´s (1984) definition of

stress can be summarized as “a particular relationship between the person and the

environment that is appraised by the person as taxing or exceeding his or her resources and

endangering his or her well-being” (p. 19). This definition is used in the present study.

This definition suggests an operationalization that assesses the individuals’ perceived stress

with self-report measurements. How unpredictable, uncontrollable and to what extent the

person perceive that they do not have sufficient resources in relation to the demands are

common experiences of stress (Bandura, 1989; Cohen & Williamson, 1988; Cox, 1993;

Hobfoll, 1989). The relationship between the environment and the person is therefore

appraised by the subject himself or herself – only people themselves can answer if they

perceive that they, for example, do not have the resources to meet the demands (Cohen,

Kamarck & Mermelstein, 1983).

The cognitive appraisal process can be divided into primary and secondary appraisal. What

the person is evaluating is how demanding or threatening an event is (primary appraisal) and

how well one can handle or cope with this event (secondary appraisal). In the secondary

appraisal, coping can mean to alter the event itself, the situation, his or her own behavior or

emotions and cognitions. For example, if a situation at work would be appraised as

4

demanding, with several projects that suddenly have become shorter deadlines, coping

behaviors can range from being external such as prioritizing tasks, to be internally directed

such as altering cognitions or to meditate (Lazarus & Folkman, 1984). To perceive that one

has the freedom to act, to have control, is necessary for the former (Karasek, 1979; Lazarus &

Folkman, 1984).

Employee control and stress

An influential model that has been used in the last decades in organizational stress research is

Karasek´s (1979) Job Demand and Control model (the JDC-model) (for review see Häuser et

al., 2010; van der Doef & Maes, 1999). The model predicts when an employee is likely to

experience strain, a mental and physical result of stress. Job demand is conceptualized as the

pace of work and the time to complete tasks. Job control is conceptualized by two

components, decision authority and skill discretion. The former is the freedom to decide over

one’s own work tasks and the latter concerns the possibility to use a variety of skills, to learn

new skills in relation to the work and the possibility to show creativity. Strain is predicted to

occur when the demands are high and the job control is low. Job control is supposed to reduce

the effect of these stressors because it implies the possibility to cope with them. For example,

the employee can take a break whenever he or she wants or has the freedom to decide how to

organize the tasks.

In this study, a more specific definition of control is used which does not include skill

discretion as in Karasek´s model (1979). Here, control is considered as the freedom, the

decision authority, one has to decide how the work should be done and what tasks that should

be done in relation to the goals. This more focused conceptualization of control fits the

purpose of this paper. It is easier to visualize how employee control, in this more focused

sense, can be decreased if management control increases.

In laboratory studies, this type of control has been shown to have strong stress-buffering

effects (for review see Miller, 1978). In the second cognitive appraisal (Lazarus, 1991), mere

beliefs of this type of control can have effect on stress (Miller, 1978). In Miller´s (1978)

review of studies of control and stress, participants in laboratory studies who had in advance

been given the freedom to decide if or when to stop an aversive event, experienced less

arousal than those who had not been given this decision authority. They experienced less

arousal even if they did not make use of their granted freedom to stop. Miller explained these

observations with his minimax hypothesis. The participant know that his or her instrumental

5

control “minimizes the maximum duration, intensity and/or frequency of the aversive event”

(p. 295), and attributes the relief to an internal source instead of an external source which

would be less stable.

In this way, employee control differs from skill discretion. Skill discretion can be argued not

to be job control per se (Wall et al., 1996). van der Doef, Maes and Diekstra (2000) measured

separately the effect of this focused type of control as a buffer on the stressor demand (time

pressure) and skill discretion, and also as an aggregate of them both (as in Karasek´s (1979)

original model) and found buffering effect on stressors in all conditions. This suggests that the

narrow focus of control, leaving out skill discretion, can adequately be considered when

studying stress in organizations.

Karasek and Theorell (1990) meant that increased management control implies decreased

employee control (although their definition included skill discretion). Management control in

this study is considered as all those behaviors, procedures and technologies that are used to

control the employee (Malmi & Brown, 2008; Simons, 1987). The combined practices of the

MCS package could be considered to decrease the type of employee control as defined in this

study. For example, all those administrative controls (Simons, 1987) that follow from

increased accountability (Blomgren, 2007; Hood, 1991) could be considered to interfere with

the employee control of how to execute the tasks. When demands are heightened, as in

increasing time pressure, this restricted situation entails less space for coping (Lazarus &

Folkman, 1984).

Some studies have not found support for that employee control is related to stress (for review

see Häuser et al., 2010). Warr (1994) argued that the relationship between employee control

and well-being is non-linear with an inverted U-shape. Control has only positive effects on

mental health up to a certain level, too high control is negative for the employee’s mental

health as it means higher responsibility, uncertainty and more difficult decision making. In a

similar vein, Eriksson (1991) meant that high control can be stressful for employees who in

their role and position at work often need to take decisions and reconsider decisions. This

“decision stress” is more likely to be experienced by officers than by workers. Still, to the

author´s knowledge, only partial or weak support has been found for Warr´s (1994) notion

that the relationship between employee control and mental health should be curvilinear (e.g.,

Jeurissen & Nyklicek, 2001). Further, in the present study, the organization in focus has

implemented tighter MCS after NPM reforms (Hood, 1991), which could mean that the

6

employee´s autonomy is restricted to the degree that it seldom reaches the level argued in by

Eriksson (1991).

How the employees perceives their own control varies with their own cognitive appraisals

(Lazarus & Folkman, 1984). The totality, or the package, of all the management controls is

reflected in the subjective conceptualization of employee control. It captures the total effect,

including eventual interactions of all the intensified MCS on the employees own subjective

control. This is in line with Malmi and Brown´s (2008) reasoning that MCS should be studied

as a package if one wants to understand its effects on behavior and cognitions. To the extent

that employee control is decreased or increased, partly as a function of increased management

control, perceived stress is affected. From this reasoning, and together with the well replicated

notion that work demand is related to stress (Häuser et al., 2010; van der Doef & Maes, 1999),

two hypotheses can be deducted;

Hypothesis 1: Work demand is positively related to stress.

Hypothesis 2: Employee control is negatively related to stress.

Feedback and stress

This paper also examines the relation between stress and some specific components of MCS –

two types of feedback and performance evaluation. First, the relation between feedback and

stress will be examined.

The research regarding how the MCS component feedback is related to stress is scarce and

inconclusive. In the later version of Karasek´s JDC-model, support was added as an additional

variable and considered a stress buffer (Karasek & Theorell, 1991). This variable is

operationalized by items concerning support from co-workers, and also feedback from

superiors (Häuser et al., 2010). As support is measured as an aggregate of them both, the pure

effect of feedback from superior is difficult to outline. However, the relation between

feedback and stress is in this paper understood from separate studies that when combined can

be argued to suggest a mediating role of goal clarity, coping and employee control, between

feedback and stress.

To give feedback is a crucial part of management control. It is a corrective action by

management to ensure that the employees are performing in the direction of the goals

(Anderson & O`Reilley, 1981; Green & Welsh, 1984). Goals that are specific and clear are

related to higher task performance than informal goals, or ”do your best”-goals (Locke &

7

Latham, 2002; Verbeeten, 2008). It has also been suggested that feedback is necessary in

conjunction with specific goals to increase performance (Hirst & Lowy, 1990).

Elovainio and Kivimäki´s (1996) study on stress among nurses demonstrated that goal clarity,

as in clear awareness of the goals of one’s own work tasks, the work unit’s goals and the

organizational goals, was related to lower stress levels. Kahn, Wolfe, Quinn, Snoek and

Rosenthal (1964) defined role ambiguity partly as what the employee experiences when he or

she has a lack of clarity over his or her duties, responsibilities and rights, and means-end

knowledge (knowledge of how to reach the goals). They found that role ambiguity was related

to job tension, lower satisfaction and lower self-confidence.

Sawyer (1992) studied the antecedents of clarity regarding goals and processes. Task

feedback from superior was positively related to process clarity, and recognition (for work

performed well) was positively related to goal clarity. In addition, recognition was also

positively related to job satisfaction. This suggests that feedback from the superior concerning

how to perform the task and recognition for when work has been done well is negatively

related to stress.

Elovainio and Kivimäki (1996) reasoned that goal clarity has a stress buffering effect because

it means that the employee understands the underlying causal links of increasing or decreasing

demands. How this understanding of the causal links behind the demands relates to stress can

also be understood from the cognitive appraisal phases and coping. As the MCS components

of feedback are tightly connected to what has to be done in the organization (Malmi & Brown,

2008) and thus the demands, one can assume that these have effect on each phase of the

employee’s cognitive appraisal process.

As feedback is a corrective action (Anderson & O`Reilley, 1981; Green & Welsh, 1984) and

as specific goals lead to more efficient behavior (Locke & Latham, 2002), the demands that

the employee is evaluating in the primary appraisal (Lazarus & Folkman, 1984) may be lower

than if the goals had been unclear. The employee would spend less time and resources on

tasks that are irrelevant in relation to the goals, his or her efficiency would be increased which

avoids that work tasks pile up, causing unnecessary time-pressure to deadlines.

In the second appraisal (though occurring at the same time as the first), the person interprets

his or her own resources and to what extent he or she is able to cope with the demands

(Lazarus & Folkman, 1984). The skills one has in relation to the tasks are important in

8

relation to stress (Karasek & Theorell, 1979). Bandura (1989) demonstrated that just the mere

belief of problem-solving skills, that one can cope with the demands, means less stress. He

termed this belief as “self-efficacy”. As feedback is bringing clarity to the goals (Sawyer,

1992) and how to reach them (Kahn et al., 1964), one can assume that this could make it

easier for the employee to structure, plan and schedule the tasks, thus increasing the self-

efficacy. In this way feedback could be assumed to facilitate coping.

As mentioned, the totality of the MCS components that has been intensified after NPM

reforms (Hood, 1991; Malmi & Brown, 2008) has been argued to lower the employees control

(Karasek, 1991). As the MCS component feedback is a corrective action (Anderson &

O`Reilley, 1981; Green & Welsh, 1984), it means that an employee that is working on a task

might be interfered. Thus, the freedom of how and what tasks that should be done could be

decreased. Still, from the positive effects of management control as feedback, this decrease in

freedom should not be that large that it takes away the positive effects of goal clarity and

coping. This leads to the third hypothesis;

Hypothesis 3: Feedback from superior is negatively related to stress.

Feedback from information systems

Given these theories that feedback increases goal clarity and coping, which in turn is

negatively related to stress, one can infer that feedback should be negatively related to stress.

Should there be any differences in terms of the relation with stress if the feedback comes from

another source; directly from the computerized information system? To the author´s

knowledge, also research for this MCS component and its relation to stress is scarce.

Key performance indicators (KPI) purpose is to direct attention to critical success factors and

to hold employees accountable for performance in some specific areas (Malmi & Brown,

2008; Ittner & Larcker, 1998; Sanchez & Robert, 2010). One study that relates KPIs to stress

was found, a study by Eriksson and Fernholm (2011). They suggested that KPIs can be

perceived as increasing work demand, thus suggesting a positive relation to stress. However,

this was only the case for those employees that were new in the organization.

Studies suggest that KPIs should be adequate in relation to the objectives and overall goals.

They have to be updated when situations change to be kept relevant. If not, their effect in

directing employee behavior will not be adequate (Adler, 1999; Beatham, Anumba, Thorpe,

& Hedges, 2004). As it is clarity over the particular means-end knowledge and the overall

9

goals that are negatively related to stress (Elovainio & Kivimäki, 1996; Kahn et al., 1964),

this study measures this type of feedback concerning its relevance, hence how the employee

perceives that the feedback helps her or him in relation to the goals and tasks.

Are employees making active use of the feedback from the information system? Wilson

(1997) means that people can differ in information-seeking behavior in regards to stress.

Krohne (1989) made a distinction between two classes of coping strategies; vigilance and

cognitive avoidance. “It is postulated that people can be habitually distinguished according to

how frequently they employ strategies of either one or the other class in situations of threat”.

That is, they tend to differ in the “frequent use of vigilant strategies on the one hand, with

frequent use of cognitive avoidant strategies on the other hand”. Vigilance can be explained as

“intolerance of uncertainty or negative surprise”, which make them inclined to extensive

monitoring, and cognitive avoidance as “intolerance of emotional arousal”, which means that

they try to avoid situational cues that could trigger their stress in the first place (p. 236-238).

This implies that some employees avoid the feedback from the information system; hence for

these employees this feedback will not increase the goal clarity or facilitate adequate coping

behaviors. Still, if the employee find the feedback relevant, and believe that the information

helps her or him, they would be more inclined to use the system, not to avoid it. In terms of

behaviorism, those actions that are followed by positive consequences are reinforcing – the

person is more likely to conduct the same actions again (Passer & Smith, 2004; Skinner,

1953). Therefore, this study measures to what extent the employee perceives that the

information system feedback helps him or her. This leads to the fourth hypothesis;

Hypothesis 4: Feedback from the information system is negatively related to stress.

Formality of performance evaluation

The third specific MCS component and its relation to perceived stress is formality of

performance evaluation. To be evaluated on formal measures, rather than informal, reflects

the tendency of public organizations to use objective measures when dispersing accountability

(Blomgren, 1997; Hood, 1991). In the particular organization in focus, the employee is

evaluated in a meeting with a superior each 6 weeks. The employee is evaluated on the goals

that were set in a meeting 6 weeks earlier. Also for this MCS component, the research is

scarce on its relation to stress.

10

Formal goals have the advantage over informal (“do your best”) goals as they are more

specific. As mentioned, specific goals lead to better performance and are negatively related to

stress as they increase goal clarity (Elovainio & Kivimäki, 1996; Hartmann & Slapničar,

2009; Locke & Latham, 2002). In this way, formality of performance evaluation can be

argued to be negatively related to stress.

Another question concerns how the evaluation is related to stress. Would there be a difference

to be evaluated with formal rather than informal measures? A study by Hartmann and

Slapničar (2009) demonstrated that employees that were evaluated more on formal terms

perceived the evaluation procedure as more fair and had greater trust in their superiors.

Another study by Elovainio, Kivimäki and Helkama (2001) suggested that perceptions of

procedural justice and fairness have a moderating effect of employee control on stress. Thus,

formality of performance evaluation, if combined with employee control, could be assumed to

be negatively related to stress. This leads to the fourth hypothesis;

Hypothesis 4: Formality of performance evaluation is negatively related to stress.

In sum, in a public service organization MCS are argued to lower the employee´s perception

of control, in terms of freedom to decide what task and how these tasks should be performed.

Still, specific MCS components, namely feedback from superior, feedback from the

information system and formality of performance evaluation are argued to increase goal

clarity, which facilitates the employee´s effort to cope with the demands. Hence, the

hypotheses were that work demand should be positively related to stress. Further, employee

control, feedback from supervisor, feedback from information system and formality of

performance evaluation should be negatively related to perceived stress.

Method

Case selection and participants

The data had been collected as part of the research project ”För Hälsa” (~”For Health”) at

Uppsala University in 2011. A public service organization was chosen after

Ekonomistyrningsverket (the Swedish National Financial Management Authority) had been

asked to list public organizations that were tightly controlled. The second organization on this

list was selected after the first declined participation. Two offices in Stockholm belonging to

this organization chose to participate with the incentive to gain insights on how to improve

11

work conditions and productivity. The two offices were identical in regard to work procedures

and objectives.

A total of 160 employees from ten departments completed questionnaires, of which 30

employees were excluded from the analysis based on the criteria that they were either

supervisors or administrative personel. A free health examination was offered as

compensation for participating. The remaining 130 subordinates (104 women, 20 men and 6

of unknown gender due to missing values) had almost similar roles and objectives and

reported to a total of 10 supervisors. The age distribution had been collected with a 5-point

Likert scale (M = 2.4, SD = 1.0); 1 = 18-34 years (20.8%), 2 = 35-44 years (34.6%), 3 = 45-54

years (29.2%), 4 = 55-64 years (14.6%) and 5 = > 64 years (0.8%). The mean tenure was 11.9

years (SD = 11.2).

Data collection and instruments

Data had been collected during a total of six days at the two offices. The participants

completed a test battery containing 96 Likert-type questions, which took about 30 minutes to

complete. As this study is part of a larger study, not all the instruments will be discussed. Four

instruments were used in this study;

Perceived stress.

The Perceived Stress Scale (PSS) was developed by Cohen et al. (1983) to measure the

”degree to which situations in one's life are appraised as stressful. The PSS items were

designed to tap the degree to which respondents found their lives unpredictable,

uncontrollable, and overloading” (p. 387). A global measure of perceived stress was chosen

because this study examines the global impact of MCS as a package in relation to perceived

stress, and because the employees spend a large proportion of their days at work exposed to

these MCS. ”The PSS does not tie appraisal to particular situations; it is sensitive to the

nonoccurence of events as well as to ongoing life circumstances, to stress resulting from

events occurring in the lives of friends and relatives, and to expectations concerning future

events” (Cohen & Williamson, 1988, p. 34).

The 10-item version of the PSS (PSS-10) (Cohen & Williamson, 1988) was used. The

participant answered on a 5-point Likert scale: 1 (never), 2 (almost never), 3 (sometimes), 4

(fairly often) and 5 (very often), on questions such as “In the last month, how often have you

been upset because of something that happened unexpectedly?”. The instrument has shown

12

predictive validity to the occurrence of psychological and physiological symptoms and

visiting of health services. Further, it has shown discriminant validity to scales measuring

psychological disorders and distress. That is, people ranking high on PSS-10 do not

necessarily score high on distress scales (Cohen & Williamson, 1988).

Work demand, employee control and feedback from superior.

The Quality Work Competence (QWC) questionnaire was designed by Anderzén and Arnetz

(2005) to assess organisational and employee well-being. This instrument consists of 11

subscales, of which three are in focus in this study. These were renamed Work Demand,

Employee Control, and Feedback from Superior (for the original names, see Appendix). The

participant answered on a 4-point Likert-scale: 1 (Never), 2 (Seldom), 3 (Sometimes) and 4

(Often) on questions adressing, for example, if the employee perceives that he or she has

”sufficient time to execute tasks” (Work Demand), ”decision latitude for deciding how work

should be done” (Employee Control) and if he or she receive ”feedback from supervisor when

tasks have been done well” (Feedback from Superior).

Feedback from the information system.

The Information System Feedback scale was originaly developed by Simon, Guetzkow,

Kozmetsky and Tyndall (1954), and has been modified to fit this study by adding one

question. The scale addresses to what extent the employee finds that the information system

gives useful feedback. And also, to what extent the employee perceives that the information

system directs his or her attention to important matters in the work procedures. The

participant answered on a 4-point Likert scale: 1 (Strongly agree), 2 (Agree), 3 (Some what

disagree) and 4 (Strongly disagree).

Formality of performance evaluation.

The Formality of Performance Measurement scale (in this study renamed Formality of

Performance Evaluation scale) was designed by Hartmann and Slapničar (2009) to assess the

employee´s perception of to what extent his or her performance is evaluated by the manager

in informal and formal terms. This scale was adjusted to reflect the work characteristics of the

case organization. That is, the participant is first asked to distribute a total of 100 points to

four different performance areas that he or she perceives that the manager finds important

when evaluating him or her; efficiency (e.g., case throughflow, results), customer contacts

(e.g., customer treatment, progress meetings with customers´ personal cases), external

cooperation (e.g., with public employment services, caregivers) and quality (e.g., judgements,

13

improvement of processes). The answers on the other items (see below) are then weighted

with this point distribution over each individual participant. To weight after several different

performance areas is in line with Malmi and Brown´s (2008) notion that MCS should be

studied as packages in their effect on behavior and cognition.

The items come in two groups of four items each. A mean score with weighting of these four

items are then calculated. In addition, these two values are then combined as a mean of them

both, which is the variable Formality of Performance Evaluation. For example, the first item

addresses “How do you perceive that your immediate superiors evaluate your performance,

personal (Informal) / objective (Formal) evaluation, for efficiency? Mark more to the left (less

formal) or right (more formal) on the scale”. The participant answered on a 5-point Likert-

scale from 1 = Less formal (“When judging my performance, my superior uses his (her)

personal judgment of my performance”) to 5 = More formal (“When judging my

performance, my superior relies on objective information from the information system”), for

the performance areas efficiency, customer contacts, external cooperation and quality.

Factor analysis

As this study uses variables with items from different instruments, some variables´ items were

similar to the items of one another, meaning that there was lack of discriminant validity

between the constructs. For example, the PSS-10 scale (Cohen et al., 1988) includes items

that address how uncontrollabe the participant find situations in his or her life, which makes it

vulnerable to be ”intertwined” with the construct Employee Control from the QWC

questionnaire (Anderzén & Arnetz, 2005). To solve this and to achieve discriminant validity,

exploratory factor analysis was performed in SPSS. The dataset was first scrutinized for

potential outliers; two were excluded due to error in coding. All the items from the five a

priori constructs were then added to the exploratory factor analysis with varimax rotation.

Concerning the items from the Formality of Performance Evaluation scale, these were added

after weighting (see above).

As expected, some items from the a priori separate constructs loaded on the same factors.

These items were deleted completely from the analysis. Further, items that were loading under

0.32 were excluded from their constructs (as proposed by Tabachnik and Fidell, 2001).

Furthermore, factors with eigenvalues greater than one were retained, and only the first six

factors which corresponded to the variables that the items were presumed to measure (as

proposed by Kaiser, 1960). New constructs were then manually computed and are presented

14

in Table 1 together with the retained items. All items, including those that were excluded from

the factor analysis, are presented in the appendix. The reliability estimates (Cronbach´s

Alpha) ranged from 0.65 to 0.84.

The eventual problem of common method variance, that can imply biases in behavioural

research that uses self-reports (see Podsakoff and Organ, 1986), was addressed by the factor

analysis. According to Podsakoff and Organ (1986), "if a substantial amount of common

method variance is present, either (a) a single factor will emerge from the factor analysis, or

(b) one "general" factor will account for the majority of the covariance in the independent and

criterion variables" (p. 536). See Table 1, none of the factors are constituting more than 50%

of the total variance. Thus, eventual common method variance was considered to be of no

concern in the data.

Inferential data analysis

To test the hypotheses 1 to 5, a multiple regression analysis was conducted with Perceived

Stress as the dependent variable and Work Demand, Employee Control, Feedback from

Superior, Information System Feedback and Formality of Performance Evaluation as

independent variables. Since studies have shown that age (e.g., Cohen & Williamson, 1988)

and tenure (e.g., Bradley, 2007) are related to stress, these where controlled for by adding

them to the first step of the multiple regression analysis. The independent variables were

added in the second step. An alpha level of 0.05 was used. Missing values, if any, were

handled by excluding cases list-wise.

The hypotheses are reflected in the following regression equation:

Perceived Stress = β0 + β1Word Demand - β2Employee Control - β3Feedback from Superior -

β3Feedback from Information System - β3Formality of performance Evaluation + ε

15

Table 1. Factor analysis with eigenvalues (E) and % of variance (V) for each factor and factor

loadings and communalities for the retained items.

Factor and item

Factor

loading Communality

Factor 1: Perceived Stress (E = 4.08, V = 22.67 %)

In the last month, how often have you been upset because of

something that happened unexpectedly?

0.57 0.50

In the last month, how often have you felt confident about your

ability to handle your personal problems?

0.52 0.40

In the last month, how often have you felt that things were

going your way?

0.74 0.55

In the last month, how often have you found that you could not

cope with all the things that you had to do?

0.65 0.50

In the last month, how often have you been able to control

irritations in your life?

0.77 0.64

In the last month, how often have you felt that you have had

control over things?

0.81 0.71

Factor 2: Information System Feedback (E = 2.72, Variance = 15.13 %)

From the information system I can understand if I do a good

job 0.80 0.74

From the information system I can understand if I do a bad job 0.79 0.73

The information system directs my attention to relevant task

related problems

0.83 0.76

The information from the systems helps me to analyze how

task related problems should be addressed

0.84 0.77

Factor 3: Formality of Performance Evaluation (E = 1.89, Variance = 10.47 %)

Formality of Performance Evaluation 1*

0.89 0.81

Formality of Performance Evaluation 2*

0.80 0.76

Factor 4: Work Demand (E = 1.43, Variance = 7.96 %)

Do you have time to plan you work tasks in advance? 0.81 0.72

Do you have enough time to complete your work tasks? 0.77 0.74

Factor 5: Feedback from Superior (E = 1.11, Variance = 6.18 %)

Do you get clear word directives from immediate supervisor? 0.85 0.76

Do you get feedback from your immediate supervisor when

tasks have been done well?

0.80 0.77

Factor 6: Employee Control (E = 1.09, Variance = 6.07 %)

Do you have decision latitude for deciding how your work

should be done?

0.79 0.75

Do you have decision latitude for deciding what tasks should

be done? 0.90 0.78

* Weighted group of items - see heading Formality of performance evaluation p.12-13.

16

Results

Means, standard deviations and skewness estimates of the variables are presented in Table 2.

The mean value of Feedback from Superior is high, 3.39 (maximum possible value 4), which

indicates that the employees in general perceive that they often get feedback from their

superior. This also means that the score distribution for that variable is negatively skewd.

Further, the variable Formality of Performance Evaluation is relatively high, 3.59 (maximum

possible value 5), indicating that the employees perceive that they are evaluated by their

superior more on formal than informal terms.

Bivariate correlations are presented in Table 3. Work Demand is positively related to

Perceived Stress, and Employee Control and Feedback from Superior are negatively related to

Perceived Stress. Further, the results indicate that Feedback from Superior is positively and

Formality of Performance Evaluation is negatively related to Employee Control. Furthermore,

Information System Feedback is indicated to be positively, but weakly related to Feedback

from Superior.

Table 2. Means (standard deviation) and skewness (standard error) for all variables.

Variable Mean (SD) Skewness (SE)

Perceived Stress 2.44 (0.65) 0.13 (0.21)

Work Demand 1.93 (0.74) 0.64 (0.21)

Employee Control 2.79 (0.79) -0.31 (0.21)

Feedback from Superior 3.38 (0.64) -1.20 (0.21)

Information System Feedback 2.22 (0.67) -0.16 (0.21)

Formality of Performance Evaluation 3.59 (0.77) -0.31 (0.21)

Table 3. Bivariate correlations between all variables.

1 2 3 4 5 6

1. Perceived Stress 1

2. Work Demand

0.38** 1

3. Employee Control -0.19* -0.29** 1

4. Feedback from Superior -0.28** -0.28** 0.23** 1

5. Information System Feedback -0.09 -0.17* 0.09 0.19** 1

6. Formality of Performance Evaluation 0.16 0.11 -0.29** -0.09 0.11 1

* p ˂ 0.05, ** p ˂ 0.01

Before interpreting the results from the multiple regression analysis, it was first examined if

multicollinearity had to be taken into consideration. According to Hair et al. (1998), tolerance

values of 0.1 or below and variance of inflation factor (VIF) values of 10 or above indicate

17

high multicollinearity. The tolerance values ranged from 0.44 to 0.89 and the VIF values

ranged from 1.16 to 2.28. Thus, when interpreting the results, multicollinearity does not need

to be considered.

As can be seen in Table 4, the variable Work Demand is positively related to Perceived

Stress, and Feedback from Superior is negatively related to Perceived Stress. None of the

other variables are significantly related to stress. Thus, support is only found for Hypothesis 1

and Hypothesis 3. The standardized regression equation is summarized as:

Perceived Stress = (0.29 × Word Demand) + (- 0.21 × Feedback from Superior) + ε

Table 4. Summary of multiple regression for variables predicting Perceived Stress (n = 127).

Independent variables B SE B β T R² Adj. R² F

Step 1

0.03 -0.01 0.16

(Constant)

2.41 0.16

14.83**

Demographic variables

Age

0.03 0.09 0.01 0.04

Tenure

0.03 0.01 0.05 0.37

Step 2

0.2 0.15 4.15**

(Constant)

1.23 0.38

3.20**

Demographic variables (Controls)

Age

-0.02 0.08 -0.03 -0.22

Tenure

0.01 0.01 0.12 1.02

Independent variables

Work Demand

0.25 0.08 0.29 3.1**

Employee Control

-0.05 0.08 -0.06 -0.6

Feedback from Superior

-0.21 0.1 -0.21 -2.25*

Information System Feedback -0.003 0.08 -0.003 -0.04

Formality of Performance Evaluation 0.07 0.07 0.08 0.93

* p < 0.05, ** p < 0.01

Discussion

The purpose of this study was to examine the relation between MCS as a package and stress

in a public service organization. It was hypothesized that work demand should be positively

related to stress and employee control negatively related to stress. Further, it was

hypothesized that the specific MCS components feedback from superior, feedback from

information system and formality of performance evaluation should be negatively related to

18

stress. The results gave only support for that work demand is positively and feedback from

superior is negatively related to stress.

That work demand was positively related to stress is in line with numerous of studies that

have replicated this finding. Increasing time pressure often goes hand in hand with perceived

stress (Häuser et al., 2010; van der Doef & Maes, 1999), also in this public service

organization.

The result that employee control was not related to stress was unexpected. Eriksson (1991)

and Warr (1994) meant that too high control, implying high responsibility and decision stress,

leads to perceived stress. If that was the case, this could have obscured the results. A

suggestion for future studies is to include measures of decision stress and other negative

effects of very high control to examine if such aspects should be accounted for when studying

control and stress in public service organizations.

As predicted, feedback from superior was negatively related to stress. Studies have previously

shown that the support dimension that was added in the later versions of Karasek´s JDC-

model is negatively related to stress. This support dimension was an aggregate of both

different kinds of support, such as emotional support from colleagues, and of feedback from

superior. This study extends the research by showing that also feedback as a standalone

construct is negatively related to stress. That feedback should be related to stress was inferred

from separate studies showing that feedback is positively related to goal clarity (Sawyer,

1992), and other studies showing that goal clarity is negatively related to stress (Elovainio &

Kivimäki, 1996). The present study extended the research by showing that feedback from

superior is directly related to stress.

As feedback is a corrective action (Anderson & O`Reilley, 1981; Green & Welsh, 1984),

increasing feedback means higher management control. Karasek and Theorell (1990) meant

that employee control should decrease with management control. And with decreased

employee control, in combination with high demands, would mean higher stress. The result of

this study does not fully support this notion. Feedback from superior was negatively related to

stress. Further, the results from the bivariate correlations suggest that employee control

increases when feedback from superior increases. A suggestion for future studies is to

examine how specific MCS components are related to employee control.

19

It was also argued that feedback could facilitate for the employees to cope with the demands.

As the goals get clearer, the employees get more efficient and as such they might perceive a

higher self-efficacy, which is negatively related to stress (Bandura, 1989). This suggests a

distinction between job control in terms of freedom for deciding how and what tasks that

should be done, and employee control as the belief of being able to perform efficiently, handle

problems and so forth.

That feedback from a different source, from the information system, was not related to stress

was unexpected. As with feedback from superior, the negative relation to stress was deducted

from the reasoning that it should increase goal clarity and facilitate coping. The reason for the

non-significant result may stem from that employees have different information seeking

behaviors, some avoid the information and some seek it to facilitate coping (Krohne, 1989;

Wilson, 1997). Still, the present study measured not only how relevant they found the

feedback, but also if it helped them to solve procedural problems. For those that found it

useful for solving problems, thus facilitating coping (Lazarus & Folkman, 1984), it should

serve as a positive reinforcer increasing the likelihood that the behavior is repeated (Skinner,

1953; Passer & Smith, 2004). Hence, eventual differences in information seeking behavior

could be assumed to have been accounted for in the method of this study. That is, those

employees that scored high on the items that addressed the positive problem solving usage of

the feedback should likely be the same who actively use the system. However, a suggestion

for future studies is to measure the information seeking behavior of the employees to outline

its effects.

That formality of performance evaluation was not related to stress was also unexpected. The

hypothesis was derived from the notion that specific goals lead to higher efficiency (Locke &

Latham, 2002) and goal clarity, which should facilitate coping (Elovainio & Kivimäki, 1996;

Bandura, 1989). An explanation for why it was not related to stress can be to compare it with

feedback from superior. Both these hypotheses were partly derived from the concept of goal

clarity. Feedback from superior is a feedback that is given continuously during the daily work

processes; the goals that get clear are more particular, addressing the everyday work processes

(Kahn et al., 1964). The performance evaluation meetings took only place each 6 weeks

which suggests that the goals cannot be as specific concerning the work processes as feedback

which is given continuously. This suggests that the frequency of feedback, how often it is

given and when, matters in its relation to stress. Future research could address this by

longitudinal studies where the frequency of performance evaluation meetings is manipulated.

20

In addition, it would be interesting to examine if goal clarity mediates the role between

formality of performance evaluation and perceived stress.

In sum, this study extended the research of management control, employee control and the

relation to stress. Management control through the MCS component feedback from superior is

negatively related to stress. This suggests that interventions of increasing feedback from

superiors can have positive effects on reducing stress in organizations. The notions that

management control increases at the expense of employee control, and that this would

increase stress, is partly questioned. When management control is meant as feedback from

superior, management control is negatively related to perceived stress.

References

Adler, R. (1999). Management Accounting: Making it world class. Butterworth-Heinemann (Oxford and

Boston).

Anderzén, I., & Arnetz, B. (2005). The Impact of a Prospective Survey-Based Workplace Intervention Program

on Employee Health, Biologic Stress Markers, and Organizational Productivity. Journal of Occupational &

Environmental Medicine, 47, 671–682.

Anderson, J. C., & O`Reilley, C. A. (1981). Effects of an organizational control system on managerial

satisfaction and performance. Human Relations, 34, 491-501.

Bandura, A. (1989). Human agency in social cognitive theory. American Psychologist, 44, 1175–1184.

Beatham, S., Anumba, C., Thorpe, T., & Hedges, I. (2004). KPIs: a critical appraisal of their use in construction,

Benchmarking, 11, 93-117.

Blomgren, M. (2007). The drive for transparency: organizational field transformations in Swedish healthcare.

Public Administration, 85, 67-82.

Bradley, G. (2007). Job tenure as a moderator of stressor-strain relations: A comparison of experienced and new-

start teachers. Work & Stress: An International Journal of Work, Health & Organisations, 21, 48-64.

Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and

Social Behavior, 24, 4, 385–396.

Cohen, S., & Williamson, G, M. (1988). Perceived Stress in a probability sample of the united states. In S.

Spacapan & S. Oskamp (Eds.), The social psychology of health. Newbury Park, CA: Sage.

Cooper, C. L., & Cartwright, S. (1994). Healthy Mind; Healthy Organization - A Proactive

21

Approach to Occupational Stress, Journal of Human Relations, 47 (1), 455-471.

Cox, T. (1993). Stress research and stress management: putting theory to work. HSE contract research report, 61.

Diefenbach, T. (2009). New public management in public sector organizations: the dark sides of managerialistic

enlightenment. Public Administration, 87, 4, 892–909.

Eisenhardt, K., M. (1985). Control: Organizational and Economic Approaches, Management Science, 31, 134-

149.

Elovainio, M., & Kivimäki, M. (1996). Occupational stresses, goal clarity, control, and strain among nurses in

the Finnish health care system. Research in Nursing and Health, 19, 517-524.

Elovainio, M., Kivimäki, M., & Helkama., K. (2001). Organizational justice evaluations, job control, and

occupational strain. Journal of Applied Psychology, 86, 418-424.

Eriksson, B., & Fernholm, C. (2011). Ekonomiska styrsystem och stress. (Unpublished bachelor thesis).

Department of Business Studies, Uppsala.

Eriksson, N. (1991) Arbetskrav, egenkontroll och socialt stöd – komponenter i den psykosociala arbetsmiljön i

Furåker, B. (1991) Arbetets villkor. Studentlitteratur, Lund.

Greenberger, D., S. & Strasser, S. (1986). Academy of Management Review, 11, 164-177.

Hair, J., F., Anderson, R., E., Tatham, R., L., & Black, W., C. (1998). Multivariate Data Analysis, Prentice-Hall,

Englewood Cliffs, NJ.

Hartmann, F., & Slapničar, S. (2009). How formal performance evaluation affects trust of subordinate managers

in their superior. Accounting, Organizations and Society, 34, 727–725.

Hirst, M., K., & Lowy, S., M. (1990). The linear additive and interactive effects of budgetary goal difficulty and

feedback on performance. Accounting, Organizations and Society, 15, 425-436.

Hood, C. (1991). A public management for all seasons? Public Administration, 69, 1, 3–19.

Hobfoll, S., E. (1989). Conservation of resources: A new attempt at conceptualizing stress. American

Psychologist, 44, 513−524.

Härenstam, A., Bejerot, E., Leijon, O., Scheele, P., & Waldenstrom, K. (2004). Multilevel analyses of

organisational change and working conditions in public and private sector. Journal of Work and Organizational

Psychology, 13, 305−343.

Häusser, J., Mojzisch, A., Niesel, M., & Schulz-Hardt, S. (2010). Ten years on: A review of recent research on

the Job Demand–Control (–Support) Model and psychological well-being. Work and Stress, 24, 1–35.

Ittner, C., D., & Larcker, D., F. (1998). Innovations in Performance Measurement: Trends and Research

Implications. Journal of Management Accounting Research, 10, 205-238.

22

Jeurissen, T., & Nyklicek, I. (2001). Testing the Vitamin Model of job stress in Dutch health care workers. Work

& Stress, 15, 254-264.

Kahn, R., L., Wolfe, D, M., Quinn, R, P., Snoek, J., D., & Rosenthal, R., A. (1964). Organizational stress:

Studies in role conflict and ambiguity. Oxford, England: John Wiley.

Karasek, R., A. (1979). Job demands, job decision latitude, and mental strain: implications for job redesign.

Administrative Science Quarterly, 24, 285–307.

Karasek, R., Theorell, T. (1990). Healthy Work: Stress, Productivity and the Reconstruction of Working Life.

New York: Basic Book.

Krohne, H., W. (1989). The concept of coping modes: relating cognitive person variables to actual

coping behaviour. Advances in Behavioural Research and Theory, 2, 235-249.

Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.

Lazarus, R. S. (1991). Progress on a cognitive motivational-relational theory of emotion. American Psychologist,

46, 819-834.

Locke, E. A., & Latham, G. P. (2002). Building a Practically Useful Theory of Goal Setting and Task

Motivation: A 35-Year Odyssey. American Psychological Association, 57, 705–717.

Malmi, T., & Brown, D.A. (2008). Management control systems as a package — opportunities, challenges and

research directions. Management Accounting Research, 19, 287–300.

Miller, S. M. (1978). Controllability and human stress: Method, evidence and theory. Behaviour Research and

Therapy, 17, 287-304.

Otley, D. (1980). The contingency theory of management accounting: achievement and prognosis. Accounting,

Organizations and Society, 5 (4), 413-428.

Passer, M. W., & Smith, R. E. (2004). Psychology. Ther science of mind and behavior. 2nd edition. New York:

McGraw-Hill.

Podsakoff, P. M., & Organ, D.W. (1986). Self-reports in organizational research: Problems and prospects.

Journal of Management, 13, 419–441.

Sanchez, H., & Robert, B. (2010). Measuring Portfolio Strategic Performance Using Key Performance

Indicators. Project Management Journal, 41, 64-73.

Sawyer, J. E. (1992). Goal and process clarity: Specification of multiple constructs of role ambiguity and a

structural equation model of their antecedents and consequences. Journal of Applied Psychology, 77, 130-142.

Shoer, C., & Wright, S. (1999). Audit culture and anthropology: Neo-liberalism in british higher education. The

Journal of the Royal Anthropological Institute, 5, 557-575.

23

Simon, H. A., Guetzkow, H., Kozmetsky, G., & Tyndall, G. (1954). Centralization vs. decentralization in

organizing the controller's department, New York: Controllership Foundation, Inc.

Simons, R. (1987). Accounting control systems and business strategy: an empirical analysis. Accounting,

Organizations and Society, 12, 357–374.

Skinner, B. F. (1953). Science and human behavior. New York: Macmillan.

Sverigesradio.se: Stressen kostar samhället miljarder (2012). Retrieved at 24/05 2012 from

http://sverigesradio.se/sida/artikel.aspx?programid=78&artikel=4965910.

Tabachnick, B. G., & Fidell, L. S. (2001). Using Multivariate Statistics. Boston: Allyn and Bacon.

van der Doef, M., & Maes, S. (1999). The job demand-control (-support) model and psychological well-being: A

review of 20 years of empirical research. Work & Stress, 13, 87-114.

Verbeeten, F., H., M. (2008). Performance management practices in public sector organizations. Accounting,

Auditing & Accountability Journal, 21, 427-454.

Wall, T. D., Jackson, P. R., Mullarkey, S., & Parker, S, K., (1996). The demands control model of job strain: a

more specific test. Journal of Occupational and Organisational Psychology, 69, 153–66.

Warr, P. (1994). A conceptual framework for the study of work and mental health. Work & Stress, 8, 84–97.

Wilson, T., D. (1997). Information behaviour: an interdisciplinary perspective. Information Processing and

Management, 33, 551–572.

24

Appendix

Table 5. PSS-10 (Cohen et al., 1983). All items.

1. In the last month, how often have you been upset because of something that happened

unexpectedly?

2. In the last month, how often have you felt that you were unable to control the important

things in your life?

3. In the last month, how often have you felt nervous and ”stressed”?

4a. In the last month, how often have you felt confident about your ability to handle your

personal problems?

5a. In the last month, how often have you felt that things were going your way?

6. In the last month, how often have you found that you could not cope with all the things

that you had to do?

7. In the last month, how often have you been able to control irritations in your life?

8a. In the last month, how often have you felt that you were on top of things?

9. In the last month, how often have you been angered because of things that happened

that were outside of your control?

10. In the last month, how often have you felt difficulties were piling up so high that you

could not overcome them?

a. Scored in the reverse direction.

Table 6. The Quality Work Competence (QWC) questionnaire (Anderzén & Arnetz, 2005).

Summary of all the items for the variables used in this study. Original QWC variable names

in parentheses.

Work Demand (Work Tempo)

1. Time for planning work duties in advance

2. Sufficient time to execute tasks

3. Time to consider how tasks have been carried out

4. Time to consider how work processes could be improved in your department

Employee Control (Participatory Management)

1. Latitude for deciding how work should be done

2. Latitude for deciding what tasks should be done

3. Sufficient influence in relationship to responsibilities

4. Opportunity to influence workplace decisions

5. Opportunity to comment on the information received from immediate supervisor

Feedback from Superior (Feedback)

1. Clear work directives from immediate supervisor

2. Feedback from supervisor when tasks have been done well

3. Feedback from supervisor when tasks have been done poorly

25

Table 7. Information System Feedback (Simon et al., 1954). All items including one

additional item that was added in this study.

1. From the information system I can understand if I do a good job

2. From the information system I can understand if I do a bad job

3. The information system directs my attention to relevant task related problems

4.

The information from the systems helps me to analyze how task related problems should

be addressed

Table 8. The Formality of Performance Evaluation Scale. Original name: The Formality of

Performance Measurement Scale (Hartmann & Slapničar, 2009).

Performance area weights

1. Efficiency (e.g., *)

2. Contacts with clients (e.g., *)

3. External collaboration (e.g., *)

4. Quality (e.g., *)

Level of agreement with the following statements:

Less formal Scale Formal

1 2 3 4 5

1. When judging my

performance, my

superior uses

his/her personal

judgment of my

performance

When judging my

performance, my

superior relies on

objective information

from the information

system

2. When judging my

performance, my

superior discusses

my performance in

qualitative terms

When judging my

performance, my

superior expresses my

performance in

quantitative terms

(rates my performance

on quantitative rating)

* Examples cannot be given due to confidentiality.