Robert L. Heneman
The Ohio State University, Columbus, OH, USA
Previous research suggests that new pay systems (e.g., variable pay and skill based pay) are related to improved employee and organizational performance. Less certain, however, is why these new forms of pay systems work and under what conditions they are likely to work. Current compensation theory does not provide sufficient explanation because it relies upon grand theories, main effects, and commonalities across pay plans. It is argued that the articles in this special issue will advance our understanding of why the new pay systems are effective because they provide research propositions helpful for the study of midrange theories, interaction effects, and differences across types of pay systems.
The demand for new theoretical knowledge regarding pay systems has never been greater. Fundamental changes in the nature of work are leading to a growth in the experimentation with new pay systems by employers (Lawler, Mohrman, & Ledford, 1998). Although there is an increasing body of knowledge showing that the new pay systems are effective, we have relatively little knowledge as to why new pay systems are effective and under what conditions they are likely to be effective. The purpose of this special issue is to provide some new theoretical perspectives to improve our understanding of new pay systems.
The new pay systems take many different forms ranging from variable pay to skill and competency based pay (Heneman & Gresham, 1998). These new forms of pay have arisen in response to fundamental changes in the nature of work (Heneman, Ledford, & Gresham, 2000). These fundamental changes that have impacted pay system development include changes in the nature of the employment relationship, changes in technology, changes in business strategy, changes in organizational structures, and changes in job design. Convincing evidence exists that these new pay systems have a marked impact on employee and organizational performance. This evidence comes from a variety of sources including descriptive reviews (e.g., Mitchell, Lewin, & Lawler, 1990), meta analyses (e.g., Jenkins, Mitra, Gupta, & Shaw, 1998), experimental studies (e.g., Petty, Singleton, & Connell, 1992), and human resource professional program assessments (e.g., McAdams & Hawk, 1994).
While the results of this evaluative stream of research is very encouraging, less encouraging is our ability to explain why these new pay plans work the way that they do and under what circumstances they work best. There are many explanations for why these pay plans should work including main effects for money, participation in decision making, and reinforcement of the business strategy. While the research has shown significant main effects for these variables, the amount of variance explained has not been large. As a result, more complicated explanations have been sought.
A more complicated set of explanations has to do with the study of the interaction between main effects, i.e., ``bundles.'' Although a promising approach, these interaction terms usually fail to reach significance (Gerhart, Trevor, & Graham, 1996). One explanation for the lack of significant findings with regards to interaction terms is that the world of compensation is less complex than believed by scholars, and instead, the world of compensation is a very straightforward one based on main effects only. This explanation does not appear likely given the small magnitude of main effects reported to explain compensation in the literature.
A more likely reason for the failure to observe significant interaction effects has to do with the current state of theory development in compensation. In the absence of a strong theoretical base, it is difficult to specify the most important interaction terms to study. The authors of this special issue have brought forth new theoretical propositions to help guide the study of new pay systems. My belief is that these propositions will help us to identify interaction terms that will explain additional variance above and beyond main effect models.
I am very encouraged by the authors' theoretical propositions in this special issue because many of these are grounded in midrange theories rather than grand theories. A positive feature of grand theories is that they may generalize across subjects, settings, and manipulations. Hence, for example, expectancy theory from psychology and agency theory from economics have been used to study all forms of compensation. As pointed out by Klimoski (1991), however, it is sometimes simply assumed that these grand theories generalize across subjects, settings, and manipulations. While expectancy theory and agency theory have certainly explained some of the variance for some compensation programs, for example, they have certainly not explained all of the variance for all compensation programs.
It is my belief that in order for the study of compensation to further progress, more attention needs to be directed toward midrange theories. By doing so, compensation is not viewed as a homogeneous construct. Different pay plans have different properties as well as commonalities. It is my belief that in compensation research, we have become too focused on grand theory, main effects, and commonalities across pay plans. More research is needed that uses midrange theory, looks for interactions, and is guided by differences in pay plan characteristics. My hope is that the collection of articles in this special issue will provide direction for these efforts.
