Designing Compensation Mix
Designing appropriate compensation is very difficult because they are so complex and related to many different theoretical perspectives. Pay mix (i.e., base pay and variable pay) is one of the major factors that we should consider in designing compensation systems. In order to determine the appropriate pay mix, we should consider (1) equity issues, (2) effectiveness of financial incentives or pay-for-performance, and (3) the amount of risk to be shared between management and employees (e.g., principal-agent relationship).
In order to design effective compensation, we should consider mainly three equity issues. They are (1) internal equity, (2) external equity and (3) employee equity (e.g., Milkovich & Newman, 1999). Internal equity refers to the internal consistency with regards to the relative job worth. People should be compensated based on the worth of jobs that they perform. External equity means that people should be paid for the same level if they perform the same jobs across organizations. Internal and external equity are closely related to the design of base pay. Employee equity focuses on the issue of employee input and output. That is, people with high performance should be rewarded as such. Employee equity might be related to the variable pay or pay-for-performance rather than fixed pay.
The issue of pay mix should be related to all of these thee equity issues. If the pay mix is perceived unfair by employees, compensation systems can not be effective. Brown & Huber (1992) examined the effect of introducing variable pay while reducing base pay. They found that both pay outcome satisfaction and pay process satisfaction decreased after the introducing of variable pay, even if the mean expected pay level was not decreased. This research suggests that decreasing base pay could harm perceived fairness of the compensation systems.
The other issue on equity is the performance measurement in determining the variable pay. If pay-for-performance is introduced without appropriate performance measurement systems (i.e., performance appraisal), it may harm the perceived fairness of the compensation. Research shows that if performance appraisal is used for administrative purpose, the rating tend to inflate (Murphy & Cleveland, 1995). If this happens, rewarding based on the employee performance is quite difficult. This difficulty may cause ambiguous decision-making in determining actual pay level, which may harm both procedural and distributive justices.
The effectiveness of financial incentives
We should also consider the effectiveness of financial incentives. Financial incentives include both individual level (e.g., individual bonus) and group level (e.g., profit sharing and gain sharing). Also, financial incentives can be short-term (e.g., annual bonus) or long-term (e.g., stock option). We should understand the mechanism in which financial incentives motivate people and increase subsequent performance.
In terms of the overall effectiveness of financial incentives, there have been controversies regarding whether financial incentives are effective or detrimental for employee motivation and performance. Operant conditioning or reinforcement theory basically supports the effectiveness of financial incentives. Especially, schedule of reinforcement is important in order to implement financial incentives effectively (e.g., Latham & Huber, 1992). Expectancy theory (e.g., Vroom, 1964) suggests that financial incentives increase motivation because they increase instrumentality (relationship between performance to valued outcomes). Goal setting theory suggests that financial incentives increase the goal acceptance and commitment, which leads to high performance.
On the other hand, cognitive evaluation theory (e.g., Deci & Lyan, 1985) suggests that financial rewards may be detrimental because such external control may reduce the individual's sense of control, which in turn reduce the intrinsic motivation. However, Eisenberger & Cameron (1996) suggest that this effect is rather limited and conclude that financial incentives can increase learned industriousness (e.g., willingness to exert efforts) which increases performance. Empirical research shows that financial incentives can increase performance. For example, Jenkins et al.'s (1996) meta analysis suggests that pay-for-performance increase performance quantity. Field study conducted by Wagner et al also showed that the introduction of financial incentive has a long-term effect on plant productivity.
Differences between individual based incentives and group-based incentives are also important in considering the effectiveness of performance-based rewards. Dematteo et al. (1998) suggest that group rewards do not have an independent effect on group performance. Rather, they suggest that other factors such as task, group, organizations should also be considered. Empirical study by Wageman (1995) shows that if tasks are interdependent, group-based rewards work better than individual-based rewards. On the other hand, if tasks are independent, individual-based rewards work better than group-based rewards.
Other researchers suggest that pay-for-performance can decrease cooperative behaviors because people may become selfish. For example, if the specific performance standards are set as criteria for pay-for-performance, they might tend to reduce other behaviors such as OCBs that are not related to the performance standards. Empirical research shows that if people are not oriented to the organizational goals, the introduction of pay-for-performance actually reduces members' OCBs.
Issues on social loafing and free riding are also serious matters for performance-based pay, especially for group-based rewards such as gain sharing (Cooper, 1992). Welbourne et al.'s (1995) research suggests that perceived fairness is the key for increasing mutual monitoring among employees, which may reduce free riding.
In sum, it can be said that performance-based pay increase employee motivation and performance. However, there are various conditions that change the degree of effectiveness of performance-based pay.
Risk sharing perspective
Introducing performance-based pay to the compensation package increases the risk exposure for employees. One of the theories that considers such issue is agency theory (e.g., Eisenhardt, 1990). Agency theory focuses on the principal-agent contracts, in which agents tend to pursue their self-interests that can harm the principal's benefits. If principals can monitor the agent behaviors, behavioral based rewards might be appropriate. However, if principals cannot monitor agent behaviors, outcome-based reward can align agent interests with those of principals. Thus agents will act on behalf of principles. Performance-based pay appears to be consistent with agency theory because employees as agents are rewarded if they produce the outcomes that align with the management or stockholders' (i.e., principals') interests.
However, agency theory suggests that if environment is uncertain, the performance-based rewards should not work well, because it assigns too much risk to the employees. Because employees are considered to be risk aversive, they may withdraw efforts if both environment and rewards are so uncertain. In fact, empirical research shows that if environment is uncertain, organizations are less likely to use performance-based rewards and that the organizations that use such rewards decrease their performance.
This line of research gives us a lot of insights about the pay mix. That is, if environment is uncertain, the large amount of variable pay in the total compensation package may not be effective. Rather, relatively large amount of fixed or base pay can provide "insurance" for the employees, which can increase employees' motivation and performance. Effective wage theory supports this perspective. That is, paying above market rate keeps employee motivated because they do not want to lose such jobs.
Conclusion and future research
In conclusion, the issue of performance based pay versus fixed pay versus a combination is very complex and difficult. As I reviewed the literature above. Only one or two theories or models are not enough in order to understand this issue, and more comprehensive understanding through the wide range of theories and research is required. Especially, there might be a lot of boundary conditions that determine the actual effectiveness of performance-based pay and fixed pay. We should look at each condition carefully in the actual cases of designing compensation mix.
Perhaps we still do not know a lot about compensation mix issues. For example, the time-range of performance-based pay (long-term vs. short-term) is one of the less studied area. Also, we should consider different types of employees (full-time, part-time, contingent workers, professionals, etc.) and different levels of employees (executives, managers, lower employees, etc.). Research on these areas will increase our understanding of the compensation mix.