Paying for Performance?

Paying for Performance?

Exploring the links between the factors influencing

pay-for-performance perception and the consequences on employees’

attitudes and behaviour

Margreet Compagner Student number: 1333976 Msc HRM, Faculty of Management and Organization Wielingenplein 5b 3522PC Utrecht Phone: 06-44974110 margreetcompagner@live.nl

First master thesis supervisor: Drs. J. van Polen

Second master thesis supervisor Drs. A.J.E. Schilder

Company supervisors Akzo Nobel

Phillip Zeman, Assistant BU HR Manager

Phone: (0031) 033 467 6335 Ronald Vergouwen, HR Officer

2 ABSTRACT

Paying for performance ?

Exploring the links between the factors influencing pay-for-performance perception and the consequences on employees’ attitudes and behaviour

Like many companies in this globalising world that want to improve performance and productivity, Akzo Nobel implemented performance management as a leading HR tool. The performance appraisal rating, decided upon during the Performance & Development Dialog (P&D Dialog), is determining the individual annual base salary raise, thereby linking performance with pay. However, survey results from 2006 and 2007 indicate that employees do not see a clear link between their performance and pay.

Two research questions are formulated to get to know how and what is determining the extent that employees perceive their performance to be linked with pay (pay-for-performance perception), and the consequences of this perception on employees attitudes and behaviour. These are “What factors are

influencing pay-for-performance perception of employees and to what extent?” and “To what extent is pay-for-performance perception influencing outcomes of commitment, motivation and opportunistic behaviour?”.

First a literature study is conducted, which formed the basis of a research model. Based on the research model, an online questionnaire is developed and sent to the middle managers of Functional Chemicals and Polymer Chemicals in the Netherlands). With 129 completed questionnaires a response rate of 64% is realised.

Results are analysed mainly with help of regression analysis. The results showed that about 50% of this group reached the maximum of their salary scale, which is expected to influence their perception, attitudes and behaviour in different ways. Therefore this research compares the different influence of the factors explaining perception and outcomes for both groups; employees who reached the maximum of their salary scale or above (=>100%) or not (<100%).

The pay-for-performance perception, attitudes and behaviour of employees in the salary position <100% depends primarily on elements in the system that stimulate and recognize performance and development, (appropriate goals, feedback on performance, support and trust of the supervisor) thereby rewarding employees intrinsically. The pay raise itself, the extrinsic reward, only has a marginal influence. This directs the idea that factors that stimulate and recognize performance are the “means” to receive a high performance and thereby come to “the end”, a pay raise. The pay raise is perceived as the outcome and recognition of the employees’ effort.

It is obvious that for all employees it is not pay-for-performance perception that is primarily determining attitudes of commitment, motivation and opportunistic behaviour. This calls up the question whether it is useful then to pay employees for their performance. When considering the outcomes in this perspective, outcomes do suggest that monetary rewards have importance, since they make intrinsic rewards more motivating. Therefore a holistic reward strategy is needed, creating synergy between intrinsic rewards (development and recognition) and extrinsic rewards (monetary reward).

The outcomes also lead to the insight that the salary position and development possibilities are related with each other. People who belief they are not rewarded for their performance (as employees who reached the maximum of their salary scale in general do), also have unmet expectations regarding their development. Outcomes of the employee survey and comments of employees support this idea.

Therefore most important recommendation is to create new alternatives to stimulate mobility and development of middle managers. This will prevent that employees get stuck in their function and have less development and pay raise possibilities, thereby becoming less committed and motivated. Different activities can contribute to this like creating an ideal position time, stimulating horizontal career moves, creating a more coaching role for supervisors, making someone responsible for internal mobility and more transparency in possible job changes.

For employees in the salary position <100%, it is recommended to set more challenging goals and to give more frequent performance feedback, in order to make them more motivated and committed.

4

CONTENTS

1. INTRODUCTION……….. 6

2. THEORETICAL FRAMEWORK……… 9

2.1. The Antecedents Influencing Pay-for-Performance Perception……….. 9

2.1.1 Goal Setting Antecedents……… 11

Appropriate goals... 11

Participation in goal setting………. 11

2.1.2. Performance Appraisal Antecedents ………..11

Performance rating validity……….. 12

Performance feedback... 12

2.1.3. Reward System Antecedents………. 13

Pay raise satisfaction………... 13

Pay raise differences……… 14

Discrepancy expected/actual pay raise………. 14

Equality of input/output ratio of others………... 14

Pay raise communication satisfaction... 15

2.2. Mediating factors Supervisor satisfaction... 15 Job satisfaction………..16 2.3. Outcomes Commitment……….. 17 Motivation………... 17 Opportunistic behaviour………... 18 3. RESEARCH METHOD 3.1. Participants……… 19 3.2. Procedure……… 19 3.3 Measurement of variables……….. 20 3.3.1. Pay-for-performance perception... 20

3.3.2. Goal setting antecedents……….. 20

3.3.3. Performance appraisal antecedents……….. 20

3.3.4. Reward system antecedents……… 21

3.3.5 Mediating factors……….………... 21 3.3.6. Outcomes... 21 3.3.7. Control variables………. 22 3.4. Method of analysis………... 22 4. RESULTS 4.1. Descriptive Statistics………... 23

4.2. Regression Analysis Results: Factors influencing pay-for-performance perception……….. 24

4.2.1 Salary position < 100%... 24

4.3 Regression Analysis Results: Factors influencing commitment, motivation

and opportunistic behaviour………. 26

4.3.1. Salary position < 100%... 27

4.3.2. Salary position >=100%... 29

5. DISCUSSION 5.1. Comparing the two groups……… 31

5.1.1. Comparing the antecedents………. 31

Goal setting………. 31

Performance appraisal……….. 31

Reward system………... 32

5.1.2. Comparing the mediators………. 32

Supervisor satisfaction……….. 32

Job satisfaction ……….. 33

5.1.3. Comparing the outcomes……….. 33

Commitment... 33 Motivation………. 34 Opportunistic behaviour………. 34 5.2. Conclusion………. 34 5.3. Theoretical implications………. 37 5.4. Practical implications……….. 38

5.5. Limitations and directions for future research ……… 41

6 1. INTRODUCTION

In order to reach competitive advantage in a globalising world, developments in business are all focused on continuously improving performance (Beer & Cannon, 2004; Hogg, 2003). A leading HR practice that is most tangible and accepted to increase productivity and firm performance is performance management (Ducharme, Singh & Podolsky, 2005; Huselid, 1995; Lazear, 2000). Most performance management systems are linking performance and pay through three practices; 1.goal setting 2.performance appraisal and 3.reward systems (Cummings & Worley, 2001). Motivation theories state that when monetaryrewards are based upon employee’s performance this will motivate them to reach their goals, thereby contributing to organisational objectives (Banker, Lee, Potter & Srinivasan, 2001; Cummings & Worley, 2001; Merchant & van der Stede, 2003; Rynes, Gerhart & Parks, 2005; St-Onge, 2000). However, money can only be a motivator when employees perceive this link between performance and pay and also perceive this link to be fair (Beer & Cannon, 2004; Campbell, Campbell & Chia, 1998; Heneman & Judge, 2000; Jones, Scarpello & Bergmann, 1999; Vest, Scott & Markham, 1994). This perceived link between performance and pay is called pay-for-performance perception (Heneman et al., 1988; St-Onge, 2000). Perceptual studies indicate that pay-for-performance perception is an important concept, because it is more influencing the motivational effect and the effectiveness of performance management, than objective and demographic measures like salary level, performance rating, tenure, age and gender (Fong & Schaffer 2003; Heneman, Greenberger & Strasser, 1988; St-Onge, 2000; Sweeney & McFarlin, 2005).

Part of the performance management system at Akzo Nobel is the performance appraisal tool called the Performance & Development Dialog (P&D Dialog) which was implemented on January 1, 2005. Thus at the time of research the P&D Dialog has been in use for three years. In the Netherlands and in several other countries, the annual base salary increase is linked with the performance rating determined by performance appraisal. However, results from employee surveys held in 2006 and 2007 (at a global scale), indicate that employees are not satisfied about certain aspects of performance management. More specifically, 62% of the employees worldwide do not see a clear link between their performance and the monetary reward they receive. When specifying these results for the business units Functional and Polymer Chemicals, it appears that with percentages of unfavourable answers of 59% and 65% in the Netherlands respectively, employees in these BU’s have the same difficulty in seeing a relation between performance and pay.

Thus, the survey results indicate that having a formal performance management policy linking performance and pay as such is not enough to create pay-for-performance perception. This creates the need to investigate what factors (antecedents) are influencing the pay-for-performance perception. First research known, of Heneman et al. (1988), indicates that pay-for- performance perception is related with the satisfaction of pay-level and the adequacy of the pay system structure and administration (perception of how the pay system operates). St-Onge (2000) follows this idea by relating the following factors with pay-for-performance perception: the actual link between performance and pay, trust in decision makers, procedural justice perception (fairness of how compensation decisions are made), and distributive outcomes (size of pay increase and performance rating). However, St-Onge (2000) states that these factors only determine 50% of the variation in pay-for-performance perception and that there must be other factors influencing the perception. This leads to the first research question:

However, even if we know what is influencing pay-for-performance perception and how, from an organisational perspective it is more important to know the consequences of this in attitudinal and behavioural terms. Unfortunately, outcomes of pay-for-performance perception in a work environment have not been investigated. This seems to stem from the way pay systems are used in the past. First, pay systems were used for an administrative purpose. Only since the use of pay systems shifted towards a strategic purpose, interest in the consequences on organisational level arose (Heneman, Greenberger & Fox, 2002). This did result in studies on outcomes of pay satisfaction (satisfaction perceived with all dimensions of monetary rewards: pay level, pay raise, the structure/administration and benefits) (Ducharme et al., 2005; Heneman & Schwab, 1985). Studies of pay satisfaction report of attitudinal outcomes as reduced levels of commitment (Bretz & Thomas, 1992; Heneman & Schwab, 1985) , motivation (Kinicki, McKee-Ryan, Schriesheim, Carson, 2002), and opportunistic behavioural outcomes as turnover (DeConinck & Stilwell, 2004; Kinicki et al, 2002), absenteeism (Kinicki et al., 2002; Williams, McDaniel & Nguyen, 2006) and lateness (Kinicki et al., 2002).

Since pay-for-performance is part of the whole package of monetary rewards, and perceptual studies do indicate that pay-for-performance perceptions form a significant determinant of pay satisfaction (Heneman et al., 1988; Fong & Schaffer, 2003; St-Onge, 2000), it can be assumed that pay-for-performance perception is also influencing the attitudinal and behavioural outcomes mentioned above. To keep the research manageable, three outcomes are chosen that are expected to influence organisational effectiveness the most. These outcomes are partly based on arguments of Katz & Kahn (1978) and are 1. organisational commitment 2. motivation and 3. opportunistic behaviour.

Organisational commitment is very important to create attachment of employees with the organization, which is influencing whether the employee will stay or not (Vuuren, Jong, Seydel, 2006), which is of course very important in the current, tight labour market. Commitment also influences organisational effectiveness, because high committed employees are positively influencing the revenues and profits (Towers Perrin, 2005, 2007).

The second factor of Katz & Kahn is job performance. However, based on other studies stating that job performance is a function of many forces, and that mainly the system and not the individual is seen to be responsible for the variance in performance, (Deming, 1986; March & Sutton, 1997; Schneider, Hanges, Smith & Salvaggio, 2003; Siehl & Martin, 1997) it was decided to replace job performance. Job performance is replaced for motivation, since the idea behind linking performance with pay is that it will influence employees’ motivation and thereby their job performance (Cummings & Worley, 2001; Merchant & van der Stede, 2003; Rynes et al., 2005; St-Onge, 2000; Kreitner, Buelens, Kinicki, 2002). Also this is in line with the compensation plan of Akzo Nobel that follows this idea by stating that its compensation should attract and retain the employees needed by stimulating motivation and commitment (Compensation Plan, 2008).

The last outcome, opportunistic behaviour is important because opportunistic activities of employees are focused on maximizing the pay raise through undesired behaviour, for example setting less difficult goals or providing incorrect data. This behaviour will undermine the organisational effectiveness (Katz & Kahn, 1978; Ramaswami, 2002).

8 effectiveness of the organisation. Akzo Nobel needs to know to what extent the pay-for-performance perception is influencing these outcomes, to be able to make a deliberate decision on whether it is worth to intervene in order to improve the pay-for-performance perception. According to Jones et al, (1999) and St-Onge (2000), this will contribute to Akzo Nobel’s ability to “effectively manage the compensation budget and to achieve its compensation goals”. Altogether, this leads to the second research question:

2. To what extent is pay-for-performance perception influencing outcomes of commitment, motivation and opportunistic behaviour?

2. THEORETICAL FRAMEWORK

In this section the theoretical framework of the research is described. An overview of the research model is given in Figure 1 (p.10). In this model variables are divided in three different columns: antecedents, mediators and outcomes. In the first part of the theoretical framework, section 2.1 the antecedents are described, these are factors that, based on other studies, are expected to influence pay-for-performance perception. The antecedents reflected in the research model are representing the three practices of performance management, goal setting, performance appraisal and a reward system, that jointly influence the performance of employees (Figure 2, Cummings & Worley, 2001). By setting goals the desired performance is specified, with a performance appraisal the performance of employees is assessed against these goals, and finally a reward system helps to reward the employee in line with his performance to ensure that desired performance will be repeated.

FIGURE 2: Performance management model

Reward system

Performance appraisal

Performance

In section 2.2. the mediators i.e. the third factors that are expected to influence the relation between the antecedent and pay-for-performance perception will be dscribed. The last part, section 2.3 explains how pay-for-performance perception is expected to influence outcomes in employees attitudes and behaviour, measured as commitment, motivation and opportunistic behaviour. The numbers in the research model correspond with the hypotheses stated in the theoretical framework.

2.1. The Antecedents Influencing Pay- for- Performance Perception

Before starting to discuss the factors that influence how employees perceive the link between performance and pay, it is important to have a clear definition of pay-for-performance perception. A report of Heneman & Schwab (1985) about the Pay Satisfaction Questionnaire gave rise to much research about pay and perception and the idea that employees experience different satisfaction with the different dimensions of pay, pay level, pay raises, benefits and pay structure/administration (system that determines the link between performance and pay). Based on this idea Heneman et al. (1988) developed the concept “pay-for-performance perception” and defined it as ‘the extent to which pay is perceived to be based upon performance’. Other researchers are also calling the same concept instrumentality belief (Scott, Markham, Vest, 1996; Vest et al., 1994). However the construct pay-for-performance perception of Heneman et al., (1988) and St-Onge (2000), is preferred because it is better reflecting that it is all about the perception of employees. With this definition in mind, the character of the factors that have an influence on pay-for-performance perception are discussed.

paid fair for his performance, when reporting a weak link between performance and pay (Heneman & Judge, 2000). According to organisational justice theory employees make a distinction in three different aspects when considering fairness. The first aspect, labelled distributive justice is the fairness of the outcome distribution or allocations, thus “the degree to which employees believe that their pay raise is fair and provides the full amount that they deserve or expect” (Eskew & Heneman,1996; Ramaswami & Singh, 2003, p. 52). The second aspect procedural justice, implicates how fair the processes and procedures are, to come to a distribution decision (Heneman & Judge, 2000). The most recent concept,

interactional justice, focuses on the intangible, interpersonal relation between employee and distribution

decision maker (Fong & Schaffer, 2003; Jones et al., 1999). This leads to the conclusion that if performance management wants to lead to high pay-for-performance perception and improved employee behaviours and attitudes, employees must perceive the amount of pay and procedures to be fair (Eskew & Heneman, 1996) but also the relationship with the supervisor needs to be good.

2.1.1. Goal Setting Antecedents

Individual goals need to be set based on the strategic goals of the organization and need to clarify the duties and responsibilities of the employees’ job, thereby letting employees know how their individual objectives are contributing to the organisational goals (Cummings & Worley, 2001; Noë et al., 2003). The use of goal setting is fundamental for performance management (Lawler, 1990) on the one hand because goals help to focus employees’ behaviour. On the other hand, because goals are at the same time the criteria employees are appraised and paid on (Cummings & Worley, 2001; Heneman et al., 1988). Heneman et al. (1988) state that goal setting is the basis for making pay-for-performance decisions and thereby influences the final for-performance decision made. This makes goal setting the basis of pay-for-performance perception. Locke & Latham (1990) say that considering the management of the process of goal setting, two elements need to be involved 1. appropriate difficulty of the goals and 2. the level of participation.

Appropriate goals

The first aspect when considering an appropriate difficulty of the goals, according to goal setting theory, is the establishment of challenging but realistic goals. Challenging goals can increase the effort expanded to achieve goals, as long as the goals are seen as feasible. Goals that are perceived as too difficult will take away the motivational effect because employees fail to achieve the goals and will give up (Cummings & Worley, 2001). Another aspect of appropriate goals is specifying and clarifying the goals and defining them operationally. For example, increase the sales in the next year with 4%. This will decrease ambiguity about expectations, thereby giving guidance to the employee how to reach the goals (Cummings & Worley, 2001). Since appropriate goals increase the possibility of achieving the goals and thereby the pay raise based on performance, they will also influence the pay-for-performance perception. This brings us to the first hypothesis:

H1: The more employees see the goals as appropriate the greater their pay-for-performance

perception

Participation in goal setting

12 2002; Ramaswami & Singh, 2003; Van Vuuren et al, 2006). Having knowledge and understanding the goals, means knowing what gets measured and how performance is linked with pay, making it easier to achieve the goals (Mulvey et al., 2002). This will very likely positively influence the pay-for-performance perception. Therefore, the second hypothesis is:

H2: The more employees can participate in setting the goals the greater their pay-for-performance

perception

2.1.2. Performance Appraisal Antecedents

Appraising performance serves many purposes and is a very important part of performance management, because performance related pay increases are based upon the performance appraisal ratings (Berger & Berger, 1999; Eskew & Heneman, 1996; Noë et al., 2003, Vest et al., 1994). Through performance appraisal, data is collected to get an overview about the strengths and weaknesses of the individual performance. This appraisal information will primarily be used to make performance based pay decisions and to provide performance feedback. Besides, it can be used for decisions about promotions, training needs and lay-offs (Berger & Berger, 1999; Noe et al., 2003). The performance appraisal is an important link between goal setting and the reward system, therefore employees need to perceive the rating as fair, to increase their pay-for-performance perception. In order to perceive the rating as fair the performance appraisal needs to be 1. valid and 2. employees need to receive feedback about their performance in order to understand the rating and to improve their performance.

Performance rating validity

Since the performance appraisal rating will determine the pay raise in the reward process, it is very important that employees have the perception that supervisors “understand and observe employees work, understand the performance appraisal process and apply it consistently”, Montemayor, (1996:142) named this performance rating validity. The first element implicated by this definition, is that the supervisor should understand and observe employees’ work. The second implication is that the supervisor should

understand the performance appraisal process to be able to apply it consistently. For a consistent

application the supervisor should also be unbiased, which means he should not include personal preferences (Jones et al., 1999; Montemayor, 1996). Heneman (2002) and Miceli et al. (1991) explain that if the supervisor does not stick to the formal performance appraisal standards, the employee will perceive that he is losing personal influence over his performance objectives. Jones et al., (1999) and Vest et al. (1995) add to Montemayor’s definition of performance validity, that the supervisor should use relevant and

accurate information to let employees perceive a relationship between performance and pay. Altogether it

can be concluded that without the supervisor understanding the employees work, a consistent and unbiased appraisal and the use of accurate information, the expectancy perception that the performance is tied with the level of pay will be constrained (Heneman, 2002; Vest et al., 1995). In other words, an invalid performance rating process is expected to decrease the perceived link between performance and pay. Thus:

H3: The more employees perceive the performance rating process as valid the greater their

pay-for-performance perception

Performance feedback

between performance and pay, resulting in a higher pay-for-performance perception (Mulvey et al., 2002; Stajkovic & Luthans, 2001).

Feedback about performance can be outcome or process oriented. Stajkovic & Luthans (2001: 583) state that feedback about outcomes enables employees to reflect their own work, because the supervisor clarifies the tasks and goals and indicates “the discrepancies between the current level of performance and the desired level”. However this self-reflection is useful, Appelbaum, Baily, Berg & Kalleberg (2000) indicate this is not enough to improve performance, since performance is a function of the employees’ abilities, motivation and opportunity to succeed in the specific work context. The employee also needs ‘opportunities’ like support, technology and training to be able to improve the abilities and performance and to succeed in the work context. This is embedded in process feedback, communicating towards the employee how he performed and what action can be undertaken to improve his performance. More specifically, Bartoll (1999) says that a performance improvement plan information needs to be provided of how to make progress to reach the desired goals. Very likely this will improve employees’ performance in the future (Ramaswami & Singh, 2003). The quality of the performance improvement plan will influence how steep the learning curve and the improvement in performance will be.

Together, performance outcome and process feedback inform the employee how well he is performing related to his goals and what activities he can undertake to improve his performance and come closer to achieving the goals. Receiving good feedback will make this process easier and will increase the extent that employees perceive the pay raises to be based on performance. Thus:

H 4: The better performance feedback is perceived by employees, the higher their

pay-for-performance perception

2.1.3. Reward System Antecedents

In the last phase of performance management, performance is rewarded with pay. When using merit pay the purpose is to realign an individual’s base salary based on the performance appraisal rating. This will result in salaries that recognize differences in performance (Berger & Berger, 2000; Somers & Birnbaum, 2000). Figure 3 is reflecting a merit grid, the tool typically used to translate the performance appraisal rating in a pay raise (Noë et al., 2003).

FIGURE 3: Structure of a merit grid

Rating/Position in range

Unsatisfactory Does not meet expectations Meets expectations Exceeds expectations Significantly exceeds <90% ….% …..% …..% ….% …..% >90-<100 ….% …..% …..% …..% …..%

As the merit grid indicates the size of the pay increase is determined by two factors. First, the performance rating of the individual (better performers receive a higher increase) and second the position in range, or compa-ratio. This is an index of the actual pay against the maximum pay of the job level, which maintains the pay structure and prevents that employees exceed the pay level maximum (Noë et al., 2003). For rewards to reinforce the desired performance, employees need to be satisfied with their pay raise. Not only with the absolute pay raise, but also in comparison to others and towhat they expected. Communication about the pay raise procedures plays a great role in this.

Pay raise satisfaction

14 determine the extent that the employee perceives the performance to be linked with pay (Folger & Cropanzano, 1998; Merchant & van der Stede, 2003; Milkovich & Newman, 2005). First, the pay raise must be perceived as valuable, to be able to motivate employees to show extra effort and perform better. Second, the individual pay raise should be large enough to have impact and to create the perception that performance and pay are linked (Merchant& van der Stede, 2003). Merchant & van der Stede (2003) state that this stems from the fact that part of the pay raise decision is influenced by “noise”. Noise results from production uncertainty and measurement errors. When the pay increase is only small, this noise will influence the pay raise for a larger extent than when pay raises are larger, thus a larger pay raise will create less noise. This makes that employees will better see the link between their effort (input) and the pay raise increase (output). As a consequence the pay-for-performance perception will be stronger (Lazear, 1998). Besides the absolute value of pay raises, Folger & Cropanzano (1998) and Milkovich & Newman (2005) state that pay raise satisfaction is influenced through the informative and symbolical value of the pay raise, because it reflects the employee’s value and status to the organization. This leads to the following hypothesis:

H 5: The more employees are satisfied with their pay raise the greater their pay-for-performance

perception

Pay raise differences As explained in the paragraph above about pay raise satisfaction, the pay raise should be large enough,

to avoid the negative effect of “noise”. This also counts for the differences in pay raise between low and high performers. According to Folger & Konovsky (1989) and Heneman et al. (1988), the differences need to be high enough to create a situation in which noise is minimized, in order to have employees who perceive that the pay increase is based on their performance. If the differences are only marginal it is not very likely employees perceive pay-for-performance perception. Lazear (1998) explains that the larger the perceived spread between high and low performers, the more the differences can be related with performance, giving employees more reason to increase their effort.

H6: The more pay raise differences employees perceive between low and high performers, the

more pay-for-performance perception they will perceive

Discrepancy expected actual pay raise

After Schulz & Tanguay (2006) considered expectancy theory, they concluded that employees’ effort in a merit pay system is determined by the extent to which employees expect a pay raise. When employees perceive a strong relation between performance and pay, this indicates that the better they perform, the more pay raise they expect. This will result in more effort of the employee. Without a strong perceived link, a discrepancy between the expected and actually received pay raise is easily created. St-Onge (2000) says that how this influences the pay-for-performance perception, depends on whether it is detrimental or beneficial to the employee. Porter, Greenberger & Heneman (1990) named this the “Matthew effect” Employees who are receiving more than expected (positive inequity), will say that their pay raise is based on performance. Otherwise they have to confess that the payment is not fair (Williams et al., 2006). Employees who are receiving less than they expected (negative inequity), may perceive their payment as not based on performance and as not fair, because they do not like to admit that their performance is not worth a higher payment (Porter et al., 1990; Williams et al., 2006). Therefore, there is a negative influence of the (size of the) inequity on the pay-for-performance perception among employees with negative inequity (St-Onge, 2000).

H 7A: Pay-for-performance perception is negatively influenced when employees are receiving a

However, research indicates that employees with positive inequity, thus receiving more than they expected, do not perceive a more positive link between performance and pay (St-Onge, 2000).

H 7B: Pay-for-performance perception is not influenced when employees are receiving a higher

pay raise than expected

Equality of input/output ratio of others

According to equity theory of Adams (1965) people evaluate the fairness of their pay relative to that of others. They compare their own ratio of perceived outcomes (pay) and perceived inputs (effort, ability, experience) with the ratio of outcomes/inputs of referent others. The more equal the ratios, the more fairness and pay-for-performance perception perceived (Heneman & Judge, 2000; Noe et al, 2003; Schulz & Tanguay, 2006). According to Gerhart & Rynes (2003) the principle of equity theory is very likely to play a role in pay-for-performance perception, since people care more about how the value of their pay raise compared to that of relevant others, than about the absolute amount of rewards they receive (Folger & Corpanzano, 1998). Folger & Corpanzano (1998) support this and explain that pay raise distributions signal an individual’s relative performance and his organisational value and status For employees in leadership positions, pay relative to that of others is even of higher importance (Trank, Rynes, Bretz, 2002) because at higher levels of pay, money is more likely to take on symbolic properties and to become an instrument of social comparison (Stajkovic &Luthans, 2001). This makes clear that in stead of absolute pay comparative pay is much more important (Furnham, 2005; Gerhart & Rynes, 2003).

The people that employees will compare with, (referents others) can be people within the same organization (internal equity) or outside the organization (external equity) (Williams et al., 2006). Since it appeared that internal equity is more important and because it is expected to be very difficult for employees to compare their pay raise based on performance with external others, this research will only focus on internal equity (Hay Group, 2003). Imagine the situation where your own appraisal rating is very positive and as a consequence your pay raise is high, but your colleague who is performing bad, receives the same performance appraisal and pay raise. Very likely your perception of pay-for-performance will dramatically decrease. Thus:

H 8: The more employees perceive the input-output ratio of others to be equal the greater their

pay-for-performance perception

Pay raise communication satisfaction

Montemayor (1996) concluded that in the reward phase of performance management, communication about the pay raise proved to be very important. He defined pay raise communication as “the set of employee perceptions related to input, two-way communication and recourse issues around merit raise allocations” (1996:142). The definition indicates that communication needs to be two-way; indicating that both the manager and employee should have the chance to deliver input and ask questions around the pay raise decision. In addition, Jones et al., (1999: 139) say that the supervisor needs to give justification of the pay raise decision, because this determines the value of the pay raise communication. Last implications of Montemayor’s definition are recourse issues. These are possibilities for the employee to complain about pay-raise decisions (Heneman et al., 2002; Montemayor, 1996). Altogether, pay raise communication gives more information about why a certain pay raise decision is made, making the employees’ expectation about pay raise/merit pay more realistic and better seeing the link between performance and pay. Thus it can be expected that:

H 9: The more satisfied employees are about the communication regarding the pay raise, the

16 2.2. Mediating factors

Mediating factors are establishing how or why an antecedent predicts or causes pay-for-performance perception. They explain the relation between the antecedents and an outcome and are the mechanism through which an antecedents influences the outcome, in this case pay-for-performance perception. Since satisfaction about the supervisor and the job play a big role in performance management and employees’ working life, positive feelings about these subjects are expected to positively influence employees pay-for-performance perception.

Supervisor satisfaction

Expecting supervisor satisfaction to be a mediator indicates that it is assumed that the antecedents cause the extent of satisfaction with the supervisor, and on his turn that supervisor satisfaction causes pay-for-performance perception. This originates from the idea that the supervisor is responsible for applying all three phases in order to relate performance with pay. They have to communicate the design, the performance measure criteria, the performance rating and pay-out to employees, which implicates a lot of interaction between the supervisor and his employees. Therefore employee’s perceptions of the way they are treated by their supervisor may influence their satisfaction with the relation between performance and pay (Fong & Schaffer, 2003; Ramaswami & Singh, 2003; St-Onge; 2000).

H 10: The more employees are satisfied about their supervisor, the greater their

pay-for-performance perception

When considering the antecedents influencing the satisfaction with the supervisor, there are several driving dimensions (Vuuren et al., 2006). A heavily influencing driver is trust in the supervisor, because it reflects the willingness of employees to rely on their supervisor to protect their interests’ (Ramaswami & Singh, 2003: 49). Trust is determined mainly through the perception of employees about the fairness of the procedures to distribute payments (Folger & Konovsky, 1989; Ramaswami & Singh, 2003), because it generates the expectation to be treated fairly (Konovsky & Cropanzano, 1991). Since it is obvious that the supervisor needs to make the performance appraisal and decides about the rating that determines the pay raise in the next phase, it is expected to influence the satisfaction about the supervisor. Therefore the following hypothesis is formulated:

H11: The more employees perceive the performance rating process as valid the more satisfied

they will be with their supervisor

The second dimension influencing supervisor satisfaction is two-way communication with the employee, since the supervisor is the person who is speaking for the organization and representing it, while giving feedback about performance and the related pay raise (Vuuren et al., 2006). Communication provides not only information to the employee, thereby resulting in more realistic expectations. It also shows that employees are valued and supported by the organization (Ramaswami & Singh, 2003; Vuuren et al., 2006). Therefore performance feedback and satisfaction about the communication with reference to the pay raise, are expected to influence the satisfaction with the supervisor. This leads to the following hypotheses:

H 12: The better performance feedback is perceived by employees the more satisfied they will be

with their supervisor

H 13: The more satisfied employees are about the communication regarding the pay raise, the

Job satisfaction

Job satisfaction is expected to be a mediator between the antecedents and pay-for-performance perception. This implies that the antecedents are causing the extent of job satisfaction and that on his turn job satisfaction is influencing pay-for-performance perception (Ducharme & Singh, 2005).

H 14: The more employees are satisfied about their job, the greater their pay-for-performance

perception

Locke (1969) defines job satisfaction as a pleasurable, positive feeling that results from the perception that one’s job fulfils or allows for the fulfilment of an employee’s important job values. Kreitner et al., (2002) explain that this definition indicates that there are three important aspects determining job satisfaction. The first aspect is that job satisfaction is a function of values (what a person desires to obtain). Secondly, different employees have different values. Then the third very important aspect is perception, since it is someone’s perception of the present situation relative to his values that are determining his job satisfaction. Thus it can be expected that giving employees an opportunity to influence the fulfilment of their job values will increase their job satisfaction. Participation in goal setting and performance feedback both allow two way communication and thereby create a way to influence the job values (Vuuren et al., 2006). This leads to the following hypotheses:

H15: The more employees can participate in setting goals, the more they are satisfied with their

job

H 16: The better employees perceive performance feedback, the more satisfied they will be with

their job

The theory above also indicates that job satisfaction is influenced by the perceived fairness of the own situation in comparison with that of others as is supported by a meta-analysis of Witt & Nye (1992). Also the expectancy model of Porter and Lawler (1968) states that the perceived equity of rewards received is influencing job satisfaction. Therefore it is expected:

H 17: The more employees perceive the input-output ratio of others to be equal the more satisfied

they will be with their job

2.3. Outcomes

Commitment

Commitment can have different subjects of focus (job, organization, supervisor) (Somers & Birnbaum, 2000), but since the focus of the research are outcomes influencing organisational effectiveness, this research is focusing on organisational commitment. Organisational commitment is an individual attitude and reflects “the extent that a person identifies himself with and is committed to the organization goals” (Kreitner et al., 2002; Mowday, Steers & Porters, 1979). According to Fiorito, Bozeman, Young & Meurs, (2007:188) there are three factors characterizing organisational commitment; a strong belief in the goals and values of the organization, the willingness to exert extra effort, and a strong desire to maintain membership Based on social exchange theories, perceived organisational support (POS) and psychological contract theory, Fiorito et al., (2007) argue that more favourable exchanges are strengthening the attraction of the employee to the employer and will increase commitment. Research of Fletcher & Williams, (1996) confirms that the link between performance and pay is influencing organisational commitment. Therefore, it can be assumed that the more performance is linked to pay the more favourable an exchange is perceived by the employee, which results in a higher level of organisational commitment.

18 Motivation

Following the definition of Mitchell (1982) motivation “represents those psychological processes that cause the arousal, direction and persistence of actions that are goal-directed”. Thus motivation itself is not behaviour, but it is an attitude that can lead to goal-directed behaviour. Kreitner et al. (2002) explain that the most direct outcomes of motivational behaviour are the effort and persistence that employees show, contributing to job performance. Although performance also depends on the abilities and opportunities (Appelbaum et al., 2000), motivation is a necessary factor for job performance, making motivation an important outcome influencing organisational effectiveness (Kreitner et al., 2002).

There are several motivational theories developed, that contribute to explaining how pay-for-performance perception can influence motivation. Expectancy theory is already explained, indicating that the motivation of an employee is determined by the extent to which there is a visible tie between employee performance and pay, because this determines the expectation that a certain level of effort and performance will result in a valued payment (Kreitner et al., 2002; Schulz & Tanguay, 2006). Equity theory suggests that employees compare their own input-output ratio with that of others (Heneman & Judge, 2000). So together expectancy and equity theory suggest that if employees perceive that better performance is rewarded with higher pay increases, there is a reason to be motivated .

Another motivation theory that can contribute to explaining the influence of pay-for-performance perception is goal-setting theory, explaining how challenging, realistic and specified goals can influence pay-for-performance perception. The model of Locke & Latham (1990) explains that goal setting also stimulates mechanisms that contribute to the motivation of employees. These mechanisms are; 1) directing attention on what is relevant 2) regulating effort 3) increasing persistence (the effort expended on a task over time) and 4) encouraging the development of action plans to achieve the goals. Thus all these mechanisms activated by applying performance management, are the same as the ones stimulating motivation. Therefore it is expected:

H 19: The higher the pay-for-performance perception, the more motivated the employee will be

Opportunistic behaviour

According to Ramaswami (1996: 106) can opportunistic behaviour be defined as “employee activities that further personal interests but that may be harmful to long-term organisational performance”. In a performance management system this indicates that employees will try to maximize the pay raise. Because pay raises are based on the goals achieved as measured in the performance appraisal, it is not surprising that employees will try to influence the goals and measures (Ramaswami, 2002). If goals are measured by output, employees may try to set less difficult goals, smooth performance results to impress the supervisor or distort the data about performance (Ramaswami, 2002). If goals are measured by process, employees may try to provide a better impression about the quantity and quality of their effort by giving invalid data, smoothing reports about daily activities and providing incorrect information.

The chance that employees are showing opportunistic behaviour is decreasing when employees perceive to receive good feedback about their performance, because this makes it easier for them to achieve the goals in a ‘legal’ way (Ramaswami, 2002). Since the satisfaction about the feedback depends on how the supervisor is providing this feedback, supervisor satisfaction is also very important in the consideration of employees to work according to the organization goals or to act to increase self-interests. Because better feedback and more supervisor satisfaction are also expected to lead to a higher pay-for-performance perception this leads to the following hypothesis:

H 20: The higher the pay-for-performance perception the fewer employees will show opportunistic

3. RESEARCH METHOD

3.1 Participants

The sample of employees involved in the questionnaire is the “middle management” of the business units Functional and Polymer Chemicals. This group of people is chosen since a benchmark of the survey results indicated that management perceives a relatively weaker link between performance and pay than non-managers (Siroto, Pulse survey 2007). According to a benchmark with other managers, about 58% of the managers should perceive a strong link between performance and pay, in stead of 41% now. Defined as middle management are managers at the level MM1-MM6, corresponding with functions with 269 – 770 Hay points. At the time of the study, Akzo Nobel Functional and Polymer Chemicals employed 202 middle managers in the Netherlands. 155 Of them answered the questionnaire although there were 26 responses with missing values. This results in 129 useful responses, which is a response rate of 63.8%. The data about demographics and background indicate that 52% of the respondents (N=67) represent Functional Chemicals and 48% (N=62) represents Polymer Chemicals. On average they worked 5.3 years in their present function and 17.0 years at Akzo Nobel. 42,6 % (N=55) had supervisory responsibilities against 57,4% who had not (N= 74). Ultimately, a distinction is made between employees who reached the maximum of their salary level and employees who did not. For this last category the answers almost, no, and do not know are summed up. This shows that 48.8% of the respondents has reached the maximum of their salary level, and the other 51.2% has not.

3.2 Procedure

To obtain insight in the perception of employees about the link between performance and pay, and the consequences of this pay-for-performance perception on employees’ attitudes and behaviour, data was obtained through a questionnaire. This method was chosen, because a questionnaire is suitable to get a general idea of people’s perception. Also the reach of a questionnaire is larger and takes less time. Another important consideration was that an interview would give too much weight to the questionnaire, creating too high expectations on the side of the middle management group about changing the performance management tool. Besides that confidentiality, an important issue to ensure participation and high-quality data for this sensitive subject, was easier to guarantee.

20 3.3 Measurement of Variables

For most of the constructs of the study, scales were copied or adapted from literature or new scales were made of items from different scales. In these items some of the words were changed to let the items fit in the context of Akzo Nobel and the P&D Dialog. In addition, new scale items were made based on insight from literature. The complete list of items used and the sources, to measure each construct is given in appendix 1. If nothing specified, the construct was measured with the mean response on the items. These were mostly stated positively and measured with a five-point Likert scale coded 1= strongly agree 5 = strongly disagree. To prevent that the range of the items was influencing the answers, the order of the items within a scale were randomly changed.

3.3.1. Pay-for-performance perception

The construct pay-for-performance perception, measures the extent that employees perceive a direct link between performance and pay raises and contains five items. The items are a combination from prior measures of Folger & Konovsky (1989) and St-Onge (2000). A sample item is: “My pay increases are based upon how my performance compares with my goals”.

3.3.2. Goal setting antecedents

Appropriate goals - Appropriate goals are challenging, realistic and specific, thus items were made

representing these three categories. Two items are made by the researcher, two items are adapted of Scott (1980) and three items from Ramaswami (2002). Examples for each of categories are; “The goals set for my P&D Dialog are stimulating me to work harder”, “I feel that the objectives expected of me in the P&D Dialog are too high “(scored reverse) and “Performance goals established for me in the P&D Dialog are specific”.

Participation in goal setting - This construct measures the extent the employee perceives to be involved in

setting performance goals. This is reflected in the mean response of a 4-item scale adapted from Teas (1981) and Vroom (1964). Only adaptation made is replacing “work goals” for “P&D dialog goals”. A sample item scale is “I am allowed a high degree of influence in the determination of the goals for the P&D Dialog”.

3.3.3 Performance appraisal antecedents

Performance rating validity - Performance rating validity measures the extent that the employee perceives

the supervisor to apply the performance appraisal process in an unbiased, consistent way and is using accurate and relevant information to understand employees’ work. The concept is measured with an 11-item scale, of which seven 11-items are adapted from Jones et al. (1999) and four from Vest et al. (1995). Item examples are: “My supervisor obtains accurate information about my performance”, “My supervisor allows personal motives or biases to influence performance appraisal ratings” (scored reverse), and “My supervisor uses consistent standards when evaluating my performance”.

Performance feedback – Feedback about performance, measures whether employees perceive to receive

3.3.4. Reward system antecedents

Pay raise satisfaction - The perceived satisfaction with pay raises, measures the satisfaction with the

amount of pay raise, with a 6-item scale. The first four items are measured with a different 5-point Likert scale varying from 1= very satisfied 5 = very unsatisfied (Currall et al. 2005). The last two items are made by the researcher and are measured with the “standard” 5-point Likert scale. Scale item examples are: “How satisfied are you with your most recent pay raise”, “I have the feeling that my pay raise is accurately reflecting my value for the company”.

Discrepancy expected/actual pay- With a 3-item scale the discrepancy between the expected and actual

pay is measured. These items are adapted from Ramaswami & Singh, (2003) and Folger & Konovsky (1989). Negative discrepancy is measured with two items, “The pay raise was less than what I expected based on my performance” and “The pay raise given to me was the full amount I deserved based on my performance” (scored reverse). Positive discrepancy is measured with these items, but then reverse, and an additional item “The size of my last merit increase was more than what I expected”

Equality of input/output ratio of others - To measure the differences perceived in internal equity, a 3-item

scale is used, adapted from Sweeney & McFarlin (2005). A scale item example is: “I feel that the pay raise I earn for my performance is very fair in comparison to others within the company”.

Pay raise communication - The perception of how well the supervisor is communicating about the pay

raise is measured with a 9-item scale, containing items measuring two-way communication and recourse issues. These items are adapted from several studies (Folger & Konovsky, 1989; Moorman, 1991; Ramaswami & Singh, 2003; Jones et al., 1999; St-Onge, 2000). Scale item examples are: My supervisor; “Considered what I said when he made his pay increase decision”, “Explains the reason(s) for the size of my pay raise”

.

3.3.5. Mediating factors

Supervisor satisfaction – To measure the extent that the employee is satisfied with his supervisor, an

6-item scale measuring supervisor trust and supervisor support is used. Three 6-items, adapted from Ramaswami & Singh (2003) use slightly different scales. An example of an item is: “How free do you feel to discuss with your immediate supervisor the problems and difficulties in your job without jeopardizing your position or having it held against you later?” The answer scale varies from 1= completely free 5 = very cautiously. Three items of Jones et al. (1999) are used to measure support, with the standard Likert scale. An item example is: My supervisor; “Represents my pay interests to other management”.

Job satisfaction - The extent that employees are satisfied with their job is measured with a 4-item scale

adapted from Lucas, Parasuraman & Enis (1987). A scale item example is: “I’m really doing something worthwhile in my job”

3.3.6. Outcomes

Motivation - To measure motivation of employees a 5-item scale of Yan & Li (1999) is used. An item

example is: “I feel quit motivated through the current P&D Dialog and the related salary increases, I get what I need from it”.

Commitment – The extent of commitment the employee feels with the organization is measured with a

22 in 2003. An example item is: “I am wiling to put in a great deal of effort beyond that normally expected to help this organization to be successful”

Opportunistic behaviour- To measure the extent that an employee behaves opportunistically a 6-item

scale is used that is adapted from two studies of Ramaswami (2002; 2003). Two scale example items are: “I work on unimportant activities simply because they are evaluated within my P&D Dialog” and “When presenting data to my supervisor”, I generally try to emphasize data that reflects favourably upon me”.

3.3.7. Control variables

There were some variables included that were used as control variables. Considering that the height of the pay raise is influenced by the performance appraisal and the position in the salary scale, the variables; salary position and salary level were included. The surveys at Akzo Nobel also indicate that managers are more positive than non-managers, therefore the variable manager/non-manager was included (Siroto, 2003). Next to that, organisational tenure and job tenure are included, because difference in tenure has been found to be related with pay-for-performance perception (Vest, Scott & Markham, 1994; Vest et al 1995).

3.4 Method of Analysis

To be able to analyze the results SPSS version 15.0 is used. Means, standard deviations, scale reliabilities (Cronbach’s alpha) and Pearson correlations among all variables were calculated. Then separate two-step step-wise regression analyses were performed to test the hypotheses. Results of these regression analyses were separated for employees with a salary position <100%, and for employees with a salary position >=100%. At step 1, the control variables were entered; organizational tenure, job tenure, manager yes/no and salary level low/high. At step 2, the independent variables influencing the dependent variable were entered. This step-wise regression was chosen for, since literature study does not indicate clear expectations about the independent variables influencing the dependent variables the most. With the regression analyses a partial F-test was generated, investigating the significant unique contribution (R²Change) for each variable in the regression model, while controlling for all other variables in the model. A significant R²Change for a variable supports one of the proposed hypotheses or adds more insight into the model.

4. RESULTS

4.1. Descriptive Statistics

In appendix 2 means, standard deviations and reliability estimates among the variables are summarized. Important to note, is that a separation is made between answers of employees who reached the maximum of their salary level and employees who did not. This choice to investigate both groups separately was made, because there is a big difference in the mean of pay-for-performance perception among these groups (3.34- 2.70) (appendix 2). Considering the system used at Akzo Nobel a difference in factors causing pay-for-performance perception could be expected. Besides, about half of the respondents, (48.8%) reached the maximum of the salary scale. To be able to understand the details, figure 4 reflects the matrix used at Akzo Nobel to decide about the pay raise. As the table indicates, the pay raise employees receive, is dependent on the performance rating and the position in the salary scale. At the moment that employees reach the maximum of their scale (100%), they can only receive a small pay raise, when they receive the performance rating “exceeds expectations or significantly exceeds expectations”. Guidelines indicate that 20-45% of the employees should be given this rating. Also then employees cannot grow higher than 115% of their salary level.

FIGURE 4: Merit grid used within Akzo Nobel

Relative salary position

Unsatisfactory Does not meet expectations/ Improvement required

Meets expectations Exceeds expectations Significantly exceeds expectations < 90% 0% % in between 3,5% % in between 6% 90-100% 0% % in between 2,5% % in between 4% 100-115%* -1,0% % in between 0% % in between 2,5%

* 100-115%; only if expectations have been exceeded significantly

To test the internal consistency of the variables the reliability of every variable is computed with Cronbach’s alpha. To increase the reliability, 6 items are deleted. In appendix 1 these deleted items are marked. Altogether, this resulted in scales with Cronbach alpha’s variating from 0.62 till 0.90 (appendix 2), indicating relatively high reliabilities of the scales (Nunally & Bernstein, 1994). Only the variable “Pay differences” had a cronbach’s alpha of .35, therefore this variable is deleted. Therefore hypothesis 6 could not be analyzed and was excluded of further research.

24 4.2. Regression Analyses Results; the factors influencing pay-for-performance perception

In this section all regression results in order to answer the first research question “What factors are influencing pay-for-performance perception and to what extent” will be presented. In section 4.2.1. the question will be answered for the group who not reached the maximum of their salary scale, and in section 4.2.2 for employees who did.

The results of the regression analyses are reflected in appendices 5 to 8 and are separated per group. B weights (=regression coefficients) and R² change are reported only for those control variables and antecedents that accounted for a significant amount of the variance. B weights are commonly used to give an indication of the importance of an individual variable (Fletcher & Williams, 1996). The B weight indicates the amount that the dependent variable will increase/decrease when the independent increases with one point on the scale, when all other variables are remaining the same (Norusis, 2006). A significant R² change figure indicates the amount of unique variance that the variable can explain.

4.2.1. Salary position < 100%

Results from the regression analyses for this group are summarized in appendix 5A. When using pay-for-performance perception as the criterion variable it appears there are three antecedents explaining a significant variance (R²=.41, p <. 05). H1 and H5 are supported: the more employees perceive the goals as appropriate (R²=.11 B = .56 p < .00), and the more satisfied they are about the pay raise (R²= .25 B =. 24, p < .05), the better they will see the relation between performance and pay raise. Next to this a positive discrepancy, is positively influencing pay-for-performance perception, (R²=.05 B = .56 p < .00) in stead of not influencing as stated in hypothesis 7b.

Results failed to support a significant effect for the other hypotheses with reference to the antecedents and pay-for-performance perception. This means more participation in goal setting (H2) a more valid performance rating process (H3), better performance feedback (H4), more satisfaction about the pay raise (5), a more equal input-output ratio compared to others (H8) and more satisfaction about communication regarding the pay raise (H9) are not contributing to more pay-for-performance perception.

Appendix 5A also reflects there are two antecedents explaining a significant variance in supervisor satisfaction (R²=.48, p <. 05). The antecedent explaining the biggest part of this variance is performance rating validity(R²=.43 B= .73, p<.001) supporting H11, that the more employees perceive the performance rating process as valid, the more satisfied they will be with their supervisor. Contrary to the hypothesis, a positive discrepancy between the expected and actual pay raise is also explaining a part of the variance in supervisor satisfaction, although a much smaller part than performance rating validity (R²= .05 B= .21, p<.001). Support for the hypotheses H12 and H13, that feedback and communication regarding pay raise are contributing to satisfaction about the supervisor was not found.

The regression results gave no evidence of antecedents or control variables explaining a significant variance in the job satisfaction. Thus, more participation in setting goals (H15), better performance feedback (H16) and a more equal input-output ratio in comparison with others (H17) are not significantly contributing to job satisfaction.

Altogether figure 5 visualizes what factors are influencing pay-for-performance perception of employees who did reach the maximum of their salary scale. The figure makes clear it is not possible to check for mediating effects, since there is no direct relation between the mediators and pay-for-performance perception, indicating that he requirements of Baron and Kenny (1986) are not met.

FIGURE 5 : Factors influencing pay-for-performance perception( salary position<100%)

Correlation is significant at * 0.05 level / ** 0.01 level / ***0.001 level

4.2.2. Salary position =>100%

Using pay-for-performance perception as the criterion variable, there are two antecedents explaining a significant variance (R²=.34, p<. 05) (Appendix 6A). Results suggest that the more employees perceive the performance rating process as valid the less their pay-for-performance perception, contradictory with what was stated in H3 (R²=.05 B = -.34, p< .05). Employees who are more satisfied with their pay raise are having a greater pay-for-performance perception, supporting H5 (R²=. 29 B = .57, p <.001). Results failed to support a significant effect for the other hypotheses with reference to the antecedents influencing pay-for-performance perception. This means more appropriate goals (H1), more participation in setting the goals (H2), better performance feedback (H4), a discrepancy between the expected and actual pay raise (H7), more equal input-output ratio’s (H8), and more satisfaction about communication regarding

pay raise (H9) are not significantly contribute to pay-for-performance perception.

Three antecedents are explaining a significant and very high amount of variance in supervisor satisfaction (R²= .77, p <0.01) (appendix 6A). Performance rating validity explains the biggest variance, (R²=.61 B =.75, p<.001) supporting H11, that the more employees perceive the performance rating process as valid, the more satisfied they will be with their supervisor. It appears that also a higher pay raise than expected, can increase the satisfaction with the supervisor (R²=.05 B = .29, p<.05), which was not expected. Last influencing antecedent is pay raise communication satisfaction (R²=.11 B=.33, p<.001), supporting H13; employees who are more satisfied about the communication regarding their pay raise, are more satisfied with their supervisor. There are no results supporting H14, that better performance feedback is related with employees who are more satisfied with their supervisor.

Discrepancy expected/ actual pay raise Pay raise satisfaction

26 The regression results show two factors explaining variance in job satisfaction (R²= .25, p <0.001) (appendix 6A). The first is the control variable salary level (R²=.07 B= .19, p<.05). A T-test indicates that employees in higher salary scales are more satisfied with their job with a mean difference of .42. The other factor influencing job satisfaction is performance rating validity (R²= .18 B = .46, P< .001), which was not expected. This indicates that the more the performance rating is valid, the more satisfied employees will be with their job. The expected relations that more participation in goal setting (H15), better performance feedback (H16) and a more equal input-output ratio (H17) would result in more job satisfaction are not supported.

At this point, only the mediators need to be analysed in a regression analysis indicate that they both do not explain a significant variance of pay-for-performance perception, thus not supporting H10 and H14.

Altogether figure 6 visualizes what factors are influencing pay-for-performance perception of employees who did reach the maximum of their salary scale. The figure makes clear it is not possible to check for mediating effects, since there is no direct relation between the mediators and pay-for-performance perception, indicating that he requirements of Baron and Kenny (1986) are not met.

FIGURE 6 : Factors influencing pay-for-performance perception (Salary position =>100%)

4.3. Regression analysis; factors influencing outcomes of commitment, motivation and opportunistic behaviour

Now the factors that influence pay-for-performance perception, in order to answer research question one are analyzed, it is time to analyze the results in order to answer research question two; which factors are influencing the attitude and behaviour of employees.

4.3.1. Salary position <100%

First thing to know is how pay-for-performance perception is influencing commitment, motivation and opportunistic behaviour. Significant results from these three different regression analyses are shown in appendix 7A. It appears that pay-for-performance perception is only causing a significant variance in the

Pay raise communication

satisfaction Discrepancy expected/

actual pay raise Pay raise satisfaction

Supervisor satisfaction Pay-for-performance perception Job satisfaction H3 -.34* H11 +.75*** H5 +.58*** H13 +.34*** Performance rating validity Performance appraisal Mediators Antecedents Pos: Rewar d system +.25** +.46***