Author: Michelle van Hooff Student number: 10663932
Thesis supervisor: Anouar El Haji M.Sc. Business Studies University of Amsterdam Date final submission: October 2014
An output-based reward system: fair or not?
Master’s Thesis
Abstract
This thesis explores to what extent knowledge sharing and absorptive capacity determine fairness perceptions about output-based compensation schemes. It also investigates whether these effects are moderated by social preferences, named Social Value Orientation (SVO).
This study conducts a vignette experiment. It uses four vignettes in the form of a two-by-two factorial design that present a real-life situation. A sample of 151 individuals is asked to read the vignette and answer to what extent they perceive the outcome described in the vignette as fair.
The results indicate that the difficulty of knowledge sharing is an important determinant in the fairness perceptions of output-based reward systems. If inequality in output between workers is caused by differences in awareness of knowledge, this situation is perceived more fair when this specific knowledge is difficult to share between colleagues. The recipient’s absorptive capacity is not found to have a significant effect on fairness perceptions. In general, individuals with a prosocial orientation perceive the bonus system to be less fair compared to individualists. However, SVO does not moderate the effect of the variables ‘difficulty of knowledge sharing’ and ‘absorptive capacity’ on fairness perceptions. The findings may help businesses that seek to create fair organizations by imposing formal reward systems designed to distribute organizational resources equally.
Keywords: Fairness perceptions, output-based reward system, knowledge sharing,
absorptive capacity, equity, compensation theory, social value orientation (SVO), bonus system.
Contents
ABSTRACT ... 2
1. INTRODUCTION ... 4
2. LITERATURE REVIEW ... 6
2.1IMPORTANCE OF HUMAN RESOURCES ... 6
2.2FAIRNESS PERCEPTIONS ... 7
2.2.1. Output-based reward systems ... 8
2.3KNOWLEDGE SHARING ... 9
2.4THE ABSORPTIVE CAPACITY OF THE KNOWLEDGE RECIPIENT ... 10
2.5SOCIAL VALUE ORIENTATIONS ... 11
2.6RESEARCH QUESTION ... 13
3. CONCEPTUAL FRAMEWORK AND HYPOTHESES ... 14
3.1KNOWLEDGE SHARING AND FAIRNESS PERCEPTIONS ... 15
3.2ABSORPTIVE CAPACITY AND FAIRNESS PERCEPTIONS ... 17
3.3SOCIAL VALUE ORIENTATIONS ... 18
4. RESEARCH DESIGN ... 20
4.1METHODOLOGY ... 20
4.2OPERATIONALIZATION ... 21
4.2.1. Independent variables ... 21
4.2.2. Moderator ... 24
4.2.3. Demographic and control variables ... 25
4.3SAMPLE ... 26 5. RESULTS ... 27 5.1GENERAL DESCRIPTIVES ... 27 5.2CORRELATIONS ... 30 5.3REGRESSIONS ... 32 6. DISCUSSION ... 37
6.1INTERPRETATION OF THE RESULTS ... 37
6.1.1. Main findings ... 37
6.1.2. Theoretical implications ... 41
6.1.3. Managerial implications ... 42
6.2LIMITATIONS & FURTHER RESEARCH ... 43
6.3CONCLUSION ... 44
REFERENCES ... 46
1. Introduction
In today’s rapidly changing economic environment, the concept of motivating human resources is often discussed in the strategy literature. Human resources and especially knowledge is viewed as a key strategic and competitive resource by organizations (Becker & Gerhart, 1996; Cohen & Levinthal, 1990; Grant, 1996; Ipe, 2003; O’Neill & Adya, 2007). To monetarily satisfy and specifically motivate these human resources, reward systems are implemented in many organizations. Therefore, the optimization of these reward systems that stimulate individual performance improvements is an essential research topic.
Studies show the strong positive role that fairness perceptions play in the ability to motivate workers. Specifically, compensation and justice research show that the fairness of the pay allocation process (procedural justice) has a direct influence on the individual performance of workers (Scarpello & Jones, 1996). Organizations constantly strive for optimal motivation and performance of their employees and are considering that performance is affected by fairness. It is therefore important for businesses to understand which factors determine fairness perceptions (Lea, 1994). The knowledge about these factors is of vital importance in building optimal reward systems, which inspire workers and at the same time are perceived as fair by every individual in the organization.
When efforts are fairly rewarded in a social exchange, employees are motivated to meet higher job demands (Janssen & Janssen, 2000). However, organizations have many reasons for implementing reward systems that are mainly based on output, without taking effort into consideration (Cravens et al., 1993). Output-based reward systems are found to be a popular and effective compensation method. Because fairness perceptions have a positive influence on job performance, it is interesting and relevant to study which factors determine fairness perceptions in individuals with regard to particular reward systems. This thesis investigates whether the difficulty of sharing productive knowledge among individuals
influences the fairness perceptions about output-based reward systems. Furthermore, we study whether the recipients’ ability to recognize, assimilate, and apply new information has an effect on fairness perceptions of individuals in the particular reward systems.
Leung & Bond (1984) found that people’s socio-emotional orientation could be an important determinant in fairness perceptions of earned rewards. Individuals vary in their motivations when evaluating different resource allocations between themselves and another person, which is defined as Social Value Orientation (SVO) (McClintock, 1978; Murphy, Ackermann, & Handgraaf, 2011). The experiment in this thesis investigates whether different SVOs result in different fairness perceptions in the same context or situation. It expects that SVO is one of the factors that have an effect on the fairness perceptions of humans in organizations. The findings could help managers improve the designing of pay-for-performance mechanisms that encourage employee motivation and in turn have a positive effect on organizational performance.
The next chapter consists of a review of the literature and ends with a research question. The third chapter introduces the conceptual framework and the hypotheses. In the fourth chapter the research design is presented. The results are shown in the fifth chapter. The final chapter consists of a discussion and conclusion.
2. Literature review
This section discusses the literature related to human resources, fairness perceptions, knowledge sharing and social value orientation. The first part of the literature review points to the importance of human resources, especially knowledge workers. The second part discusses fairness perceptions and their role in output-based reward systems. The third part of this section summarizes the theory about knowledge sharing, and factors that might influence knowledge distribution in organizations. The final part consists of an analysis of social value orientation and introduces the research question.
2.1 Importance of human resources
The resource-based view (RBV) and related capability-based theories are the most important and influential frameworks in understanding strategic management. Much attention in the RBV literature has been paid to non-human resources, whereas researchers have been claiming for a long time that human resources are neglected as an important organizational asset which are essential in understanding performance (Barney & Wright, 1997). Researchers in strategic organization argue that individuals matter and that micro-foundations are important in recent strategic organization research (Felin & Foss, 2005). Micro-foundations can be defined as organizational behavior which is based on individual action and interaction (Abell, Felin, & Foss, 2008; Felin & Hesterly, 2007; Felin & Foss, 2005; Gavetti, 2005).
A number of scholars and commentators have argued that physical resources are no longer the most important asset but that the economy is driven by technologies based on knowledge, information production and distribution (Dahlberg & de Haan, 2013; Powell & Snellman, 2004). Based on this evolution, knowledge workers became more important
expertise, education and experience, and the primary purpose of their jobs involves the creation, distribution, or application of knowledge (Davenport, 2013). Attracting, motivating and retaining knowledge workers have become more and more important (Kim & Mauborgne, 1998; Ramírez & Nembhard, 2004). Firms try to understand what motivates them and focus on progressive human resource management practices (Coff, 1997; Delaney & Huselid, 1996; Drucker, 1999; Horwitz, Heng, & Quazi, 2003). In motivating knowledge workers, fairness perceptions are found to play an important role affecting the behavior of many people (Fehr & Schmidt, 1999).
2.2 Fairness perceptions
Fairness perceptions are found to be an important determinant in worker’s behavior (Janssen, 2001; Konovsky & Pugh, 1994; Moorman, 1991; Organ & Moorman, 1993; Steers, 1991). Theories of social and interpersonal justice have been applied to understand behavior in organizations (Greenberg & Greenberg, 1990). The term organizational justice refers to perceptions of fairness within organizations (Beugré, 1998). Two factors that need to be considered in forming fairness judgments are the outcome received (distributive justice) as well as the procedure used to arrive at that outcome (procedural justice) (Anderson & Patterson, 2008; Williams, 1999).
Several scholars argue that perceived fairness in the allocation of outcomes and procedures used to make these distributions are an important factor for motivating individuals (Beugré, 1998). This is supported by research which shows that there is a significant positive relationship between procedural justice and intrinsic motivation (Zapata-Phelan, Colquitt, Scott, & Livingston, 2009). When people feel that the strategic decision-making processes are fair, they are more likely to show unique performance improvements based on a higher level of motivation (Zapata-Phelan et al., 2009). As a consequence, organizations often implement
formal procedures like reward systems that promote the fair distribution of valued resources (Jerald Greenberg, 2003).
2.2.1. Output-based reward systems
Compensation systems like monetary rewards are used to motivate individual performance improvements (Baker, Jensen, & Murphy, 1988; Heinrich, 2007). Specifically the structure and the fairness perceptions of compensation systems influence employees’ motivation, which in turn affects organizational performance (Bartol & Srivastava, 2002; Heinrich, 2007).
Research on fair reward distribution often generates interest in the distinction between equity and equality (Jerald Greenberg, 1978). The equity norm of Adams (1965) prescribes that the ratio of one’s reward to one’s inputs must be equal to the rewards/inputs ratio of others (Adams, 1965). Inequity could exist for a person if he perceives that the ratio of his outcomes to inputs and the ratio of other’s outcomes to inputs are unequal (Goodman & Friedman, 1971). To the contrary, the equality norm expresses that individuals must receive an equal share of the reward being allocated, irrespective of their contributions (van Yperen, van den Bos, & de Graaff, 2005). Outcomes refer to rewards such as (bonus) pay or job status, which a person receives for performing his job (Goodman & Friedman, 1971). Inputs represent the contributions the person brings to the job, such as age, education, and physical effort (Goodman & Friedman, 1971). Output is for example the number of products sold, the number of interviews completed or pages proof read (Goodman & Friedman, 1971).
A sizable section of the job market explicitly pay workers for their individual outputs using bonus pay, commissions, or piece rate contracts, which results in more wage inequality among workers. For example, if a manager believes total output is very important, the
They implement control systems that are mainly based on output and involve little managerial involvement with salespeople. It relies on straightforward, objective results measures (like sales, divisional profits or production numbers) and the use of compensation methods that shift risks to the salesperson (i.e. commission or bonus) (Oliver & Anderson, 1994).Although there are many disadvantages when it comes to objective measures, in practice it is more popular compared to subjective measures (Baker et al., 1988). Subjective measures are often not preferred because they contain a great deal of ambiguity as well as room for people to disagree (Baker et al., 1988; Cropanzana, Bowen, & Gilliland, 2007).
In an output based reward system, the output of individual workers is influenced by many factors, like effort, education or access to certain knowledge. Information sharing in the organization can play a considerable role, because it has the ability to lower barriers between knowledge workers by improving access to information, which in turn can affect output (Hendriks, 1999). Previous research found four major sets of factors that are likely to influence the difficulty in knowledge sharing: The type of source, the sort of knowledge being transferred, the context in which the transfer takes place, or the recipient (Szulanski, 1996).
2.3 Knowledge sharing
As mentioned before, knowledge and thus knowledge workers are seen as the fundamental resource of today’s organizations (Ipe, 2003). Knowledge workers become even more important when they share their knowledge with the organization. Therefore, businesses are focusing their attention on the way expertise is spread in the workplace and on those groups of people that acquire, process and apply this know-how (Scarbrough, 1999). Knowledge sharing can be defined as the willingness of individuals or groups in an organization to directly or indirectly share the expertise they have acquired or created with others (Bock,
Zmud, Kim, & Lee, 2005). In a practical sense, knowledge sharing cannot be forced in an organization. It can only be encouraged and facilitated (Bock et al., 2005). Knowledge resides within individuals and the movement of individual knowledge to organizational knowledge is basically dependent on employees’ knowledge-sharing behaviors (Bock et al., 2005). Sharing knowledge might be very difficult for workers. This difficulty can be caused by factors that influence knowledge sharing, like the nature of knowledge, the opportunities to share and the culture of the work environment (Ipe, 2003). Firms desiring to promote knowledge sharing should create facilitative work contexts (Bock et al., 2005).
This thesis aims to understand whether individuals perceive an output-based bonus system to be fair and which factors determine these fairness perceptions. One of the starting points in the discussion is the belief that people perceive the output-based reward system to be more fair if every individual in the organization has access to the same information, where knowledge sharing between individuals is of significant importance. However, there are more factors involved in the formation of fairness judgments. For instance, if knowledge is shared and every individual has access to the same know-how, the discussion about how individuals in an organization use this information is another dominant topic, known as absorptive capacity.
2.4 The absorptive capacity of the knowledge recipient
Absorptive capacity is a broad set of organizational member skills needed to deal with the tacit component of transferred knowledge and the need to modify it (Mowery & Oxley, 1995; Park, Suh, & Yang, 2007). Cohen & Levinthal (1990) define absorptive capacity as the ‘ability to recognize the value of new external information, assimilate it, and apply it to commercial ends’ (Cohen & Levinthal, 1990). Focusing on the recipient of knowledge, a lack
poor transfer of knowledge (Goh, 2002). Scientific literature often suggests that the absorptive capacity of the recipient is the most significant determinant of internal knowledge transfers in multinational companies (Minbaeva et al., 2003).
Absorptive capacity can be further classified into three components: First, the capacity of a knowledge user for recognizing and understanding external knowledge, which is strongly related to a prior knowledge base. When the person already has related knowledge, it helps to absorb new knowledge, because this increases both the ability to memorize new information, and the ability to recall it and put into practice. Secondly, the user’s capacity to assimilate new knowledge is the user’s ability to internalize this into his own task environment. The third component consists of applying the information to the task (Cohen & Levinthal, 1990; Minbaeva et al., 2003; Nonaka & Nishiguchi, 2001; Park et al., 2007). Even if an organization consists of individuals with the ability to learn quickly, their ability to use the absorbed knowledge will be low if employees’ motivation is low or absent (Baldwin, Magjuka, & Loher, 1991; Minbaeva et al., 2003). In the many years since it was introduced, the concept of absorptive capacity has been used in a lot of peer-reviewed academic papers, due to the unique perspective that it provides (Lane, Koka, & Pathak, 2006). However, the knowledge we already have about the influences of absorptive capacity is still very rudimental (Lane et al., 2006). The large number of papers and broad range of using the concept of absorptive capacity raises important concerns about the state of the literature on this subject.
2.5 Social Value Orientations
People are often faced with social dilemmas, situations in which their immediate self-interest is at odds with what is best for their relationships or community (Balliet, Parks, & Joireman, 2009). Social dilemmas apply to a wide range of real-world problems; they exist within dyads, small groups, and society at large. For example, they may occur in situations as diverse
as deciding who washes the dishes, whether to conserve community resources, donating to charity, or voting in an election. Studies on the motivations that underlie interdependent decision behavior have a long history and these motivations have been referred to by a variety of names, including: social preferences, social motives, welfare trade-off ratios, and Social Value Orientation (SVO) (Murphy et al., 2011).
Social value orientation (SVO) is defined by McClintock (1978) as pre-existing
preferences for certain patterns of outcomes for oneself and other(s), on perceptions of behavior in social dilemmas (Van Lange & Liebrand, 1991). SVOs are individual differences of how people evaluate outcomes for themselves and others in interdependent situations (Anderson & Patterson, 2008). Most people can be classified as being either prosocial or proself (van Dijk, De Cremer, & Handgraaf, 2004; Van Lange & Kuhlman, 1994; Van Lange, 1999). Prosocials seek to maximize joint outcomes and equality while proself individuals (individualists) tend to strive for maximizing their own outcome, regardless of other’s outcome. Finally, people with a competitive orientation are motivated to maximize their own outcome at the expense of others (Anderson & Patterson, 2008).
Research about social dilemma’s has been largely disconnected from research on social justice and fairness, so the question about how these are related has never been completely answered (Beersma & de Dreu, 2003). However, there has been a growing interest in the role of individual differences in fairness considerations. Anderson & Patterson (2008) argued that prosocials follow the principles of equity theory in forming distributive fairness judgments. They demonstrated that prosocials perceived an equal outcome as more fair than either an unfavorable outcome or a favorable outcome (Anderson & Patterson, 2008). The fairness perceptions of proself individuals tended to follow the principles of self-interest theory: Given the same inputs as a comparison other, they saw a favorable outcome as fairer than an unfavorable outcome. These findings confirm that SVO affects the evaluation of
outcomes in the formation of fairness judgments, which is an influential starting point in the field.
2.6 Research Question
Despite many attempts at creating workplaces in which everyone is treated fairly, formal plans often fall short of creating conditions that are uniformly regarded as fair (Jerald Greenberg, 2003). The question that still exists is why there is such a limited success of pay for performance programs. Although compensation research has demonstrated the important role of fairness perceptions in the evaluation of outcomes, it has not been able to completely explain which factors determine the level of fairness in compensation decision-making. Also the role of social dilemmas on social justice and fairness has never completely been answered. No pay for performance system can be considered to be most fair and most motivating under all circumstances. It is interesting to study the conditions under which a particular system is perceived as more fair and more motivating than others.
The purpose of the present study is to analyze fairness of third-party observers (who vary in background) to an equitable allocation of rewards when the ‘difficulty of knowledge sharing’ and ‘absorptive capacity of the recipient’ are varied. Furthermore, it tests SVO as a possible moderator in the relationship. The main research question that is addressed in this thesis is:
To what extent do ‘difficulty of knowledge sharing’ and ‘absorptive capacity of the recipient’ determine fairness perceptions of output-based compensation systems and is this relationship influenced by SVO?
3. Conceptual Framework and Hypotheses
The theory of distributive justice can be applied to the case of two or more persons, each of whom receives his rewards from a third party: an employer. This thesis proposes a model that explains the fairness perceptions of knowledge workers about a bonus system that is based on output measures and excludes effort. For example: Person A puts obviously more effort into his work compared to person B. Despite that, person B sells more products because he owns a productivity-increasing strategy. This is valuable knowledge because using this strategy he is able to sell more products without having to increase his effort. Based on this higher output, person B will receive a higher bonus compared to person A.
The model assumes that an individual’s perceptions of fairness in a knowledge organization is partly determined by the difficulty of productivity-increasing knowledge sharing and by the absorptive capacity of the recipient. It is expected that this relationship will be moderated by the social value orientation (SVO) of the individual knowledge worker. Figure 1 explains the relationship between the difficulty of productivity-increasing knowledge sharing (low or high), absorptive capacity of the recipient (low or high) and fairness perceptions. An individual’s social value orientation (SVO) is presented as moderator in the relationship.
Independent variables Dependent variable
Moderator
Figure 1: Conceptual model
3.1 Knowledge sharing and fairness perceptions
Research about value distribution and distributive behavior is dominated by equity theory of Adams (1965) (Leung & Bond, 1984). This theory is especially important for the study of compensation. Designing a compensation plan is an important task for most companies because it helps determine the motivation of the workers and the long-term profitability of the organization (Coughlan & Sen, 1989). Adam’s equity (1965) theory is important in this study because effort is assumed to be one of the most manageable inputs; and wages, especially piece-rate, are very tractable outcomes (Weick, 1966).Equity involves a ratio of outcome to input as perceived by the person. A person will feel inequity if he expends much effort for low outcomes (high/low) while his co-worker gets high outcomes for the same amount of effort (high/high); but, he will be even more bothered if the co-worker obtains his high outcomes with low inputs (low/high). As already mentioned in the literature review, the
equity norm prescribes that the ratio of one’s reward to one’s inputs must be equal to the
rewards/inputs ratio of others (Anderson & Patterson, 2008; van Yperen et al., 2005). In Difficulty of productivity-‐ increasing knowledge
sharing
Fairness perceptions SVO Absorptive capacity of the recipient
economic situations, allocators distribute rewards equitably because they believe equitable allocations promote productivity (Berkowitz, 1966). When the rewards allocated are proportional to each individual’s contributions, the allocation is (according to the equity norm) perceived as fair (van Yperen et al., 2005). In comparison to an equal division (in which everyone receives exactly the same salary), an equitable division is only perceived more fair when effort is presented as the cause of the performance difference (van Yperen et al., 2005).
Economic models of compensation generally assume that higher performance requires greater effort, while this is not always necessarily the case (Baker et al., 1988). For example, possession of productivity increasing knowledge (for example a certain sales strategy) leads to more performance without additional effort. If this capacity-increasing strategy is difficult to share, differences in performance will exist between colleagues that are not caused by a contrast in effort but by diversity in the awareness of this certain strategy.
However, it is assumed that the greater the difference between ratios, the greater the inequity and the pressures to reduce it (Weick, 1966). If the productivity increasing knowledge (which allows for differences in performance) is difficult to share with co-workers, it is hard to reduce this inequity in the organization. This means that it is really hard for the worker who owns the strategy to share it with colleagues. Even if the worker has the intention to share his strategy with colleagues, it might be very hard or even impossible because of its difficulty of sharing. However, if this knowledge can easily be shared (but is not shared), inequity between workers could have been reduced. Since it is assumed that differences in ratios lead to greater pressure to reduce them, this situation is expected to be perceived less fairly by co-workers. This can be hypothesized as:
H1: A bonus system is considered more fair if productive knowledge is difficult to share with others (under the assumption that they do not share), compared to when it is easy to share with others.
3.2 Absorptive capacity and fairness perceptions
As already mentioned, an equitable division is perceived more fair if effort is presented as the cause of the performance difference. However, when talent causes differences in performance, equality rather than equity is perceived as the most fair distribution rule (van Yperen et al., 2005). Awareness of productive knowledge might cause performance differences between workers if this knowledge is not shared. However, if all individuals in the organization were aware of the strategy, differences in absorptive capacity could exist. The recipient of knowledge can have a high absorptive capacity (which means that he is able to use the knowledge in practice) or a low absorptive capacity (which means that he is not able to use the knowledge in practice).
If the recipient of knowledge has a low absorptive capacity, differences in outcome (inequality) could not have been reduced if the productivity-increasing knowledge had been shared, because the recipient is not capable of using the knowledge in practice. So whether the productivity-increasing knowledge is shared or not, if the recipient has a low absorptive capacity variation in outcomes still exists because of the differences in talent. However, if the recipient had a high absorptive capacity, sharing the knowledge could have reduced inequalities. Assuming that the knowledge is not shared, this situation (high absorptive capacity) is therefore perceived as less fair. This can be hypothesized as:
H2: A bonus system is considered less fair if the recipient has a high level of absorptive capacity, compared to when the recipient has a low level of absorptive capacity.
The greater the difference in ratios between co-workers, the more pressure there is to reduce it. Ratios are included in equity theory and can be explained as a ratio of outcome to input as perceived by the person. In practice the distribution of rewards is perceived less fair if the opportunities for reducing inequity are not completely exploited. So if the productivity increasing strategy is not shared but could easily be shared (assuming that the recipient of this knowledge could have used this knowledge in practice), inequity will exist while this could have been prevented by distributing the knowledge. Therefore, it is expected that individuals perceive this situation as less fair, which leads to the following hypothesis:
H3: Assuming that the recipient of knowledge has a high level of absorptive capacity, the bonus system is considered as less fair if the knowledge can easily be shared compared to when the knowledge is difficult to share.
3.3 Social value orientations
An individual evaluates his outcome-income ratio in any given situation relative to the outcome-input ratio of a similar colleague. This evaluation is based on the individual’s values or internalized normative beliefs (Adams, 1963; Weick & Nesset, 1968). People vary in their motivations when evaluating different resource allocations between themselves and another person (Murphy et al., 2011). This socio-emotional orientation can be an important factor in their fairness perception of distributive norms (Leung & Bond, 1984). As mentioned before, prosocials seek to maximize joint outcomes and equality (Murphy et al., 2011; Van Lange, 1999). A key finding from the motivational approach is that individuals with a prosocial value orientation (those who value their own and others’ outcomes equally) collaborate more easily than individuals with an individualistic/proself (those who value their own outcomes only) or
competitive (those who value doing well at the other’s expense) value orientation (De Dreu & Boles, 1998).
Inequality in outcomes might exist if productivity-increasing knowledge increases individual outcome in an organization, but this knowledge is not shared with all colleagues. Owners of this piece of knowledge will achieve a higher outcome compared to others, and therefore a higher bonus. In such a situation of inequity, prosocial individuals are expected to feel a stronger pressure to reduce the inequity ratio (because they strive for equality), compared to proself individuals (individualists). Thus the following hypothesis:
H4: Compared to individualists, prosocials consider the bonus system less fair if the knowledge can easily be shared (but is not shared) compared to when the knowledge is difficult to share.
Assuming that the recipient of productivity increasing knowledge has the ability to use this know-how in practice, prosocials consider the situation as less fair compared to individualists when this knowledge is not shared. This is caused by the fact that prosocials wish to enhance equality (Anderson & Patterson, 2008). Sharing the knowledge could have reduced inequality (whether this is favorable or not), but this opportunity is not exploited. Accordingly, I propose the following hypothesis:
H5: Compared to individualists, prosocials consider the bonus system as less fair if the recipient has a high level of absorptive capacity (under the assumption that the knowledge is not shared) compared to when the recipient has a low level of absorptive capacity.
4. Research design
The following section discusses the design of the study. The first paragraph explains the method and research instruments (vignette technique & survey questionnaire), which are used to collect data. The second paragraph focuses on a detailed operationalization of the variables. Finally, the last paragraph discusses the participants of the study.
4.1 Methodology
This research will be done using a quantitative vignette study, because it is a valid way of conducting experimental research. A quantitative vignette study always consists of two components:
-‐ A vignette experiment as the core element
-‐ A traditional survey for additional respondent-specific characteristics, which are used as covariates in the analysis of vignette data (Atzmüller & Steiner, 2010).
First, a scenario in the form a written vignette is presented to the respondents. In a vignette study, short descriptions of situations or persons are used and shown to respondents in the form of a survey to be able to interpret their opinion about these situations (Atzmüller & Steiner, 2010). It is a short, carefully constructed description of a few persons in a given situation, representing a systematic combination of characteristics (Atzmüller & Steiner, 2010). Secondly, respondents are asked to complete a survey questionnaire in which the SVO of the individuals can be measured, combined with questions about their personal background.
social media have been used to distribute the questionnaire. In the vignette, individuals are asked to take the role of a salesperson in the presented organization and answer whether they perceive the treatment of the salesperson as fair or not based on the scenario. The respondents are asked to answer the question from the perspective of a colleague in the same organization. The independent variables are varied in a two by two factorial design, which creates a total of four versions of the questionnaire. Respondents are randomly allocated to one of the four vignettes.
Figure 2. Vignette Design
4.2 Operationalization
This part of the research design section discusses how the independent variables are manipulated to measure their effect on the dependent variable. It also explains how the moderation effect is measured to find its effect on the relationship between the independent and dependent variables. In the end, it discusses the demographic and control variables.
4.2.1. Independent variables
The vignettes, which are developed for this research, are used to measure the influence of the Level of absorptive capacity
Low High
Difficulty of knowledge sharing
Low
High
Low/Low Vignette 1
High/Low Vignette 3
Low/High Vignette 2 2
High/High Vignette 4
independent variables (difficulty of knowledge sharing & absorptive capacity) on the dependent variable (fairness perception).
Vignette 1 Vignette 2 Vignette 3 Vignette 4
Imagine, you work for an insurance company. You and your colleagues have a relatively low fixed salary, but you all have the possibility to receive a bonus based on the number of insurance policies you sell each month. One of your colleagues, Adam, has a strategy, which helps him to sell 10% more policies compared to his colleagues. Therefore, he receives a bonus every month. Another colleague, Ben, isn’t aware of this strategy. He works the same number of hours and puts the same effort in his work, but sells fewer insurance policies.
(In this part of the vignette the variables are manipulated) Adam decides not to share the strategy with Ben.
Difficulty of sharing knowledge Low High The strategy can easily be shared between colleagues, and if Adam shares the strategy with Ben, it won’t affect his own bonus. The strategy can easily be shared between colleagues, and if Adam shares the strategy with Ben, it won’t affect his own bonus. The strategy is difficult to share between colleagues (because it takes a lot of time to explain) and if Adam shares the strategy with Ben, it won’t affect his own bonus. The strategy is difficult to share between colleagues (because it takes a lot of time to explain) and if Adam shares the strategy with Ben, it won’t affect his own bonus. Level of absorptive capacity
(ability to use the knowledge in practice)
Low High Low High
If Ben would be aware of the strategy, he would not be able to use the strategy (because he does not have the capacity to put it in practice) so it would not help Ben to get the
If Ben would be aware of the strategy, he would be able to use the strategy and also get the bonus.
If Ben would be aware of the strategy, he would not be able to use the strategy (because he does not have the capacity to put it in practice) so it would not help Ben to get the
If Ben would be aware of the strategy, he would be able to use the strategy and also get the bonus.
There are four vignettes that are randomly distributed to the respondents (see table 1). All four vignettes describe the same situation except that the independent variables ‘difficulty of knowledge sharing’ (high or low) and ‘ability to use knowledge in practice’ (high or low) are manipulated in order to influence respondents’ opinion to which extent they perceive that the bonus system treats the salespeople fairly. In each of the vignettes, the independent variables are changed in order to understand their influence on the dependent variable.
Independent variables:
-‐ Difficulty of knowledge sharing (low – high) -‐ Ability to use knowledge in practice (low – high)
Each respondent is randomly assigned one of the 4 vignettes, which describe a situation about 2 colleagues in an insurance company. The participant is asked to read it from the perspective of being another colleague in the same organization. For example, the respondents who were assigned to vignette 1 were asked to read the following scenario:
Imagine, you work for an insurance company. You and your colleagues have a relatively low fixed salary, but you all have the possibility to receive a bonus based on the number of insurance policies you sell each month. One of your colleagues, Adam, has a strategy, which helps him to sell 10% more policies compared to his colleagues. Therefore, he receives a bonus every month. Another colleague, Ben, isn’t aware of this strategy. He works the same number of hours and puts the same effort in his work, but sells fewer insurance policies.
The strategy can easily be shared between colleagues, and if Adam shares the strategy with Ben, it won’t affect his own bonus. If Ben would be aware of the strategy, he would not be able to use the strategy (because he does not have the capacity to put it in practice) so it would not help Ben to get the bonus.
All four treatments assume that effort is on the same level; all employees put an equal amount of effort into their work. One of the colleagues receives a higher payment (bonus) because he possesses certain productivity-increasing knowledge that helps him to increase his output. In the particular organization, bonus payments are only determined by the output of individual salespersons. The situation assumes that this productivity increasing knowledge (which is owned by the colleague) is not shared with others.
The first two vignettes assume that this knowledge (strategy) is easy to share with colleagues. The variable ‘level of absorptive capacity’ (which describes whether the recipient would have been able to use the knowledge in practice) is varied (low-high).
Vignette three and four describe a similar situation, but these assume that the critical knowledge is difficult to share between colleagues. The variable ‘level of absorptive capacity’ (which describes whether the recipient would have been able to use the knowledge in practice) is again varied (low-high).
After reading the story, the participant is asked to what extent he/she considers the situation fair. A 4-point scale is used to measure fairness considerations subjectively: ‘very fair’, ‘acceptable’, ‘unfair’ and ‘very unfair’ (Kahneman, Knetsch, & Thaler, 1986).
4.2.2. Moderator
After the respondents finished the vignette, a survey questionnaire was used to determine the SVO of each respondent. Individual differences in SVO can be defined as pre-existing preferences for certain patterns of outcomes for oneself and other(s) (Van Lange & Liebrand, 1991). Within the SVO framework it is assumed that people vary in their motivations or goals when evaluating different resource allocations between themselves and another person (Murphy et al., 2011). With the SVO of each respondent this study tests whether the relationship is moderated between the independent variables and the dependent variable.
There are several different measurement methods to measure the SVO of each individual respondent. This study uses the SVO Slider Measure (Murphy et al., 2011), because it yields high resolution output, is efficient and is easy to use. The SVO Slider Measure is able to detect the most prevalent SVO individual differences, which allows for an evaluation of rank orders of social preferences (Murphy et al., 2011).
The SVO Slider Measure has six primary slider items (see figure 3). The respondent is asked to imagine that he is randomly paired with another person who he will not know and who will not know who he is. The task of the respondent is now to allocate money between himself and the other person. The respondent has to indicate the allocation he prefers the most. The decision maker evaluates each of the slider items sequentially and indicates for each one his most preferred joint distribution.
Figure 3. One of the primary slider items (Murphy et al., 2011).
4.2.3. Demographic and control variables
The third part of the questionnaire was used to determine the background of each respondent, in terms of demographic variables (gender, age, country of residence) and control variables (educational background, level of education, work experience and experience with reward systems). These demographic and control variables are added to measure whether they influence the dependent or independent variables.
4.3 Sample
All participants were approached online. They were reached through social media and E-mail. After starting the survey they were requested to carefully read the scenario and fill out the questionnaire. It took on average five minutes to finish the survey. To test whether they had carefully read the scenario and whether they had paid serious attention towards answering the questions, the survey included two control questions at the end. These were included to filter out only the respondents who carefully read the situation in the vignette.
5. Results
5.1 General descriptives
Initially, the sample of this study consisted of 234 respondents. In total, 413 respondents started the survey, but only 234 actually finished it. The dataset did not contain any missing values, because the respondents were forced to answer all questions to be able to continue to the next page of the questionnaire. However, some respondents seemed to have difficulties in understanding what was asked of them when they were asked to divide money between themselves and someone else. They came up with unusual answers, which indicated that they did not understand the question. Therefore, a filter variable was used to filter only the respondents who came up with valid answers, and to disregard the outliers. The filter was also used to select only the respondents who answered the two control questions at the end correctly, because this was a very important condition for this experiment. With both these filters, the actual analysis was conducted with a sample of only 151 respondents (65% of all 234 respondents).
Most of these respondents are female (72.2%) and are living in the Netherlands (59.6%). Furthermore, most participants were either a student (53.6%) or an employee (34.4%), of which 48.3% had a bachelor’s degree as the highest achieved educational level. Business & Economics was the sector in which most of the respondents (27.8%) were studying or had studied. By using the SVO Slider Measure (Murphy et al., 2011) it is measured that 76.8% of the sample is prosocial, 22.5% is individualist and only 0.7% is competitor (see table 2). The average age of the respondents is 28.37 (SD = 9.761).
All respondents were randomly assigned to one of the four vignettes. However, because only the respondents who finished the questionnaire were included in the study, the division of the respondents to the vignettes was not equal anymore. For example, if the
respondents who were assigned to vignette 1 had a high drop-out rate, the database ends up with less respondents for this vignette compared to the other vignettes (see table 2).
N Percentage
Gender Male 42 27.8%
Female 109 72.2%
Country of residence Netherlands 90 59.6%
United Kingdom & Ireland 56 37.1%
Other 5 3.3%
Actual position Student 81 53.6%
Employee 52 34.4%
Self-employed 10 6.6%
Unemployed 1 0.7%
Other 7 4.6%
Highest education achieved Grammar school 2 1.3%
High school 27 17.9% Vocational/technical school 8 5.3% Bachelor degree 73 48.3% Master degree 35 23.2% Doctoral degree 2 1.3% Professional degree 3 2.0% Other 1 0.7%
Educational background Business & Economics 42 27.8%
Law 5 3.3% Medicine 7 4.6% Psychology 25 16.6% Architecture 3 2.0% Other 69 45.7%
Full-time work experience 0-1 year 86 57.0%
1-5 years 19 12.6%
5-10 years 12 7.9%
10-20 years 22 14.6%
more than 20 years 12 7.9%
Experience in bonus system Yes 46 30.5%
No 105 69.5% Vignette Vignette 1 35 23.2% Vignette 2 52 34.4% Vignette 3 28 18.5% Vignette 4 36 23.8%
Social value orientation Prosocial 116 76.8%
Competitor 1 0.7%
Individualist 34 22.5%
5.2 Correlations
The following table (table 3) presents the Pearson correlation analysis, which defines whether there are interactions between the different variables. This analysis tests whether there are parallels between the independent variables (difficulty of knowledge sharing & absorptive capacity), the dependent variable (fairness perceptions), the moderator (SVO), and the control variables (gender, age, education level, work experience). The table also indicates whether these correlations are significant.
To be able to do the analysis, some variables are transformed into dummy variables: Gender (1=male), Age, Education level, Experience with bonus system, Nr of years work experience, SVO & Fairness.
M SD 1 2 3 4 5 6 7 8 1. Difficulty of sharing (1= high diff) 0,424 0,496 2. Absorptive capacity (1=high AC) 0,583 0,495 -0,035 3. Gender (1=male) 0,280 0,450 -0,114 0,016 4. Age (1=age>30) 0,278 0,450 -0,084 0,136 -0,022 5. Education level (1=highly educated) 0,753 0,433 ,174* -0,142 0,047 -0,160 6. Experience with bonus system (1=yes) 0,300 0,462 -0,044 -0,024 ,167* ,328** 0,003 7. Nr of years work experience (1= >5 years) 0,305 0,462 -0,073 0,152 -0,026 ,842** -0,156 ,343** 8. SVO (1=prosocial) 0,768 0,423 0,027 -0,019 -,184* ,166* 0,023 -0,080 0,159 9. Fairness (1=fair) 0,556 0,498 ,200* -0,107 ,227** -0,011 0,085 0,099 -0,017 -,175* * Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed).
Table 3: Correlations
The correlation matrix shows that fairness is positively correlated with difficulty in sharing (r = 0.200, p < 0.05). This means that respondents perceive the bonus system as more fair if the knowledge is difficult to share, compared to when it is easy to share. The matrix also shows that fairness is positively correlated with gender (r = 0.227, p < 0.01), which
means that males generally perceive the bonus system to be more fair compared to females. Another notable relationship is the correlation between SVO and gender (r = -0.184, p <
0.05). This correlation supports the idea that females are more prosocial compared to males.
The correlation matrix also indicates the positive relationship between SVO and age (r =
0.166, p < 0.05), which means that older respondents are more prosocial compared to
younger respondents. Another interesting observation is the positive correlation between SVO and fairness (r = -0.175, p < 0.05), which means that prosocial respondents perceive the bonus system to be more fair compared to individualists or competitors.
5.3 Regressions
To test the hypotheses, a multiple linear regression was conducted. A multiple linear regression analysis is an extension of a simple linear regression analysis, used to assess the association between two or more independent variables and a single continuous dependent variable. The regression analysis consists of six models. Table 4 shows the multiple linear regression models that are used to test the five hypotheses. In all regression models, fairness perceptions is the dependent variable. A high score in fairness perception means that the respondent perceives the situation as very fair, and a low score in fairness perceptions means that the respondent perceives the situation as unfair.
The first model consists of only the control variables. In order to test hypothesis 1 & 2, the variables ‘’Difficult to share’’ and ‘’High absorptive capacity’’ are added to Model 1. Figure 4 shows that the groups with high sharing difficulty conditions have higher scores in fairness compared to the groups with low sharing difficulty conditions. Model 2 in table 4 supports this by showing that in a situation with high sharing difficulty conditions, fairness perceptions will be significantly higher compared to a situation with low sharing difficulty
Figure 4.
Figure 5 indicates that the groups who read the vignettes with low absorptive capacity conditions have higher score in fairness compared to the groups who read the vignettes with high absorptive capacity conditions. However, model 2 shows that in a situation in which a recipient has a high absorptive capacity, fairness perceptions are not significantly lower compared to a situation in which a recipient has a low level of absorptive capacity (B =
-0.089, p = 0.273). Therefore, hypothesis 2 is not supported.
Figure 5. 2 2,25 2,5 2,75 Low difficulty of sharing High difficulty of sharing
Fairness Low absorptive
capacity High absorptive capacity 2 2,25 2,5 2,75 Low absorptive capacity High absorptive capacity
Fairness High difficulty of
sharing
Low difKiculty of sharing
In model 3, the variable ‘’difficulty x absorptive’’ is added, in order to test the third hypothesis. According to hypothesis 3 it is assumed that if the recipient of knowledge has a high level of absorptive capacity, the bonus system shows higher scores in fairness in the group with high sharing difficulty condition compared to the group with low sharing difficulty condition. This variable is a combination of the two independent variables difficult to share and absorptive capacity and represents all the responses for vignette 2 (low difficulty of sharing and high absorptive capacity of recipient). The group with the low difficulty of sharing condition has a lower mean score in fairness compared to the high difficulty of sharing condition. However, the regression analysis does not indicate a significant effect (B =
0.023, p = 0.889), so hypothesis 3 is not supported.
Before testing the interaction effect of SVO, the variable ‘Prosocial’ is added in model 4 to test the effect of SVO on fairness perceptions. The model indicates that it has a significant negative effect, which means that prosocials consider the bonus system overall less fair compared to individualists or competitors (B = -0.193, P = 0.047). Hypothesis 4 and 5 propose a moderating effect of SVO on fairness perceptions. In order to test this moderating effect, two interaction-variables are added. Hypothesis 4 states that the differences in mean scores in fairness for the groups with high difficulty of sharing conditions and low difficulty of sharing conditions are moderated by SVO. To test hypothesis 4, the interaction-variable ‘prosocial x difficulty’ is created and will be tested using the regression analysis of model 5. The regression model shows that there is no significant interaction-effect (b = -0.189, p =
0.305), which does not support hypothesis 4.
In model 6, the interaction-variable ‘prosocial x absorptive’ is added in order to test hypothesis 5. The hypothesis states that the differences in mean scores in fairness for the groups with the high absorptive capacity condition and the groups with low absorptive
capacity condition are moderated by SVO. Model 4 shows that there is again no significant interaction-effect (b = -0.169, p = 0.373) which rejects hypothesis 5.
Model 1 Model 2 Model 3
B Sig. B Sig. B Sig.
(Constant) 0.42 0.000 0.400 0.000 0.406 0.001
Gender 0.226 0.014** 0.260 0.004** 0.259 0.005**
Age 0.007 0.967 0.025 0.879 0.028 0.863
Education level 0.080 0.399 0.022 0.816 0.023 0.805 Years experience -0.045 0.782 -0.034 0.833 -0.037 0.819 Experience in bonus system 0.095 0.321 0.091 0.335 0.089 0.347 Difficult to share 0.234 0.004** 0.221 0.083* High absorptive capacity -0.089 0.273 -0.099 0.356
Difficulty x absorptive -0.023 0.889 Prosocial Prosocial x difficulty Prosocial x absorptive R square 0.062 0.122 0.122 F 1.909 2.815 2.228 Sig. 0.096 0.009 0.016 N 149 149 149
Dependent variable: Fair/Unfair dummy
Model 4 Model 5 Model 6
B Sig. B Sig. B Sig.
(Constant) 0.545 0.000 0.474 0.000 0.470 0.003
Gender 0.23 0.012** 0.244 0.008** 0.223 0.015**
Age 0.048 0.764 0.065 0.689 0.055 0.732
Education level 0.033 0.726 0.039 0.675 0.031 0.736 Years experience -0.017 0.912 -0.024 0.881 -0.021 0.895 Experience in bonus system 0.072 0.445 0.067 0.473 0.076 0.419 Difficult to share 0.238 0.004** 0.67 0.023** 0.240 0.003** High absorptive capacity -0.095 0.240 -0.094 0.245 0.036 0.830
Difficulty x absorptive Prosocial -0.193 0.047** -0.114 0.355 -0.095 0.511 Prosocial x difficulty -0.189 0.305 Prosocial x absorptive -0.169 0.373 R square 0.146 0.147 0.148 F 3.018 3.047 3.072 Sig. 0.004 0.003 0.003 N 149 149 149
6. Discussion
This thesis aims to test whether ‘difficulty of productivity-increasing knowledge sharing’ and ‘absorptive capacity of recipient of knowledge’ are important factors in the determination of fairness perceptions about output-based reward systems. It also tests whether this effect is moderated by SVO. In this section, the interpretation and implications of the results are presented. First, the interpretations of the main findings are discussed. The second part consists of the limitations of this experiment and suggestions for further research. This chapter will end with some concluding remarks.
6.1 Interpretation of the results
6.1.1. Main findings
Difficulty of productivity increasing knowledge sharing & absorptive capacity of recipients
The first hypothesis expects that individual’s fairness perceptions about an output-based reward system are partly determined by difficulty of knowledge sharing in the organization. In this study it is assumed that some individuals in the organization have productivity-increasing knowledge, which is not shared with all workers. This results in differences in output and therefore differences in outcome (salary) without having to increase effort for the ones who own this knowledge. The knowledge (which we assume is not shared) might be very difficult to share or might be easy to share. This hypothesis expects that this difficulty of sharing influences fairness perceptions about reward systems. The second hypothesis expects that these fairness perceptions are also partly determined by the absorptive capacity of the recipient. Differences in absorptive capacity can lead to differences in output and outcome without having to increase effort.
Van Yperen et al. (2005) found that an equitable division is only perceived as fair when effort is presented as cause of the performance difference. It does not take into account
any other inputs except effort. In addition, the equity theory of Adams (1965) states that such a division is perceived as fair as long as the ratio of individual’s contributions and outputs are equal to this ratio by others. Contributions consist of elements such as knowledge, training or task proficiency. The assumption in Adams’ equity theory is that the greater the differences between ratios, the greater the inequity and the pressures to reduce it. This is in line with the first result of this experiment. The result of the first hypothesis show that the output-based reward system is considered more fair if knowledge is difficult to share with others (assuming that it is not shared) compared to when it is easy to share with others. This means that if it is easy to share (but is it not shared) opportunities for reducing differences are not exploited, a situation that is perceived less fair. This finding points to the pressure of individuals to reduce inequity ratios. So people value access to the same information as an important factor in reducing these inequity ratios.
However, this experiment found that the absorptive capacity of the recipient does not influence the fairness perceptions of individuals. Inequities caused by differences in absorptive capacity of individual recipients are not found to determine fairness perceptions about an output-based reward system. This is in the same line of reasoning as the results obtained by Weiner and Kukla (1970), Kaplan and Swant (1973), and Rest, Nierenberg, Weiner, and Heckhausen (1973) who argue that allocators base their decisions more on differences between recipient’s performance than on differences between individual ability or absorptive capacity. Task activity could be seriously disrupted if an allocator distributes rewards and resources solely on the basis of individual ability or absorptive capacity (Berkowitz, 1966). Such a policy fails to provide greater reinforcement and resources for individuals whose actual contributions were greatest (Berkowitz, 1966). This could be perceived as an unfair procedure with respect to high performers, which might be the explanation of the results of the experiment in which no differences in fairness were found
when individual absorptive capacity was manipulated. So whether the recipient of knowledge is able to use the knowledge or not does not influence on the fairness perceptions of individuals about reward systems. In the third hypothesis it is expected that assuming that the recipient of knowledge has a high level of absorptive capacity, the reward system is then considered less fair if the knowledge can easily be shared compared to when the knowledge is difficult to share. No significant differences in fairness were found which indicates that adding this information about absorptive capacity of the recipient ensures that the effect of ‘difficulty of knowledge sharing’ on fairness perceptions disappears. This result again supports the finding that absorptive capacity has no influence on fairness perceptions in an output-based reward system.
Social value orientation
Based on research by van Lange & Liebrand (1991), Murphy et al. (2011), Anderson & Patterson (2008), McClintock (1978), this experiment expects that an individual’s social value orientations moderate the effects found in the first hypotheses. They found that an individual’s socio-emotional orientation can be an important factor in their fairness perceptions or distributive norms. This means that in the pay-for-performance system presented in this experiment, it is expected that individualists have different fairness perceptions compared to prosocials. The results of the first hypothesis show that the reward system is considered more fair if knowledge is difficult to share with others (assuming that it is not shared) compared to when it is easy to share, which points to the fact that individuals perceive it more fair if opportunities for reducing differences are exploited. According to Murphy et al. (2011) & van Lange (1999), prosocials seek to maximize joint outcomes and equality. Therefore, the fourth hypothesis expects that this effect will become more substantial if the individual has a prosocial value orientation, compared to when the individual
is an individualist or competitor. The results of this experiment show that there is no significant interaction, which indicates that the expectation that SVO make the effect even stronger is not supported. This means that in comparison to individualists, prosocials are not found to be more sensitive to knowledge sharing (access to the same knowledge/information) as an important factor in their fairness perceptions about payments in an output based reward system.
The results of the second hypothesis show that individuals do not perceive any differences in fairness of the reward system when absorptive capacity of the recipient is manipulated. So whether the recipient of knowledge (and rewards) is able to use the information does not have any significant influence on the fairness perceptions of the reward system. The fifth hypothesis states that this relationship is positively moderated by SVO, which means that in comparison to individualists, prosocial individuals perceive the reward system to be less fair if the recipient has a high absorptive capacity (still assuming that knowledge is not shared) compared to when the recipient has a low absorptive capacity. This points to the stronger willingness of prosocial individuals to reduce inequity between individuals. There was no significant relationship found between absorptive capacity and fairness perceptions, an also SVO is not found to have a significant moderation effect. This means that compared to individualists, prosocials do not consider the output-based reward system to be less fair if the recipient has a high absorptive capacity compared to when the recipient has a low absorptive capacity.
Overall, this research shows that SVO has a direct effect on fairness perceptions. It shows that prosocials perceive the bonus system overall less fair in comparison to individualists. This is in line with research of Anderson & Patterson (2008), who found a significant effect of SVO in the formation of fairness judgments when evaluating outcome information. However, the results of this experiment indicate that SVO has no significant
influence on the fairness perceptions of reward systems determined by difficulty of knowledge sharing and absorptive capacity of individuals.
6.1.2. Theoretical implications
The first main contribution to theory is that access to the same information by means of knowledge sharing is an important factor, which partly determines fairness perceptions for output-based reward systems. Compensation research found the important role of fairness perceptions on job performance, but it has not been able to explain which factors determine the level of fairness in compensation decision making. The connection between knowledge sharing and fairness perceptions of reward systems has never been made explicit and has not been studied before. The present study refers to the process of knowledge sharing, based on research by Hendriks (1999). He argued that knowledge sharing in organizations can play a considerable role because it has the ability to lower barriers between workers, which in turn affects the output of every individual. Therefore, this thesis focused on determining whether knowledge sharing and specifically difficulty of knowledge sharing plays an important role in the fairness perceptions of output-based reward systems.
The second main contribution to the theory is the finding that the absorptive capacity of an individual worker does not determine fairness perceptions in output-based reward systems. It has already been argued that allocators more often base their decisions on differences in performance between individuals than on differences between each individuals absorptive capacity. However, this decision is made by managers but it has never been studied what the effect of this decision is on the fairness perceptions of workers about the system. This experiment indicates that absorptive capacity of individuals does not have any influence on the fairness perceptions of workers about such formal output-based reward systems.
