A meta-analytic view of performance-based pay systems

To date, research on compensation systems has relied heavily on single studies, while giving little consideration to well-conceived, cumulative methods to grasp a better view of performance based pay (PBP) plans. Using the meta-analytic procedure of Hunter and Schmidt (1990, 2004), this dissertation examined the influence of reward type, research design, and study setting on performance-based pay systems. Two key strengths of meta-analysis are (a) estimating a true population parameter from a collection of individual studies of a related topic and (b) identifying intervening variables that may explain mixed results found in such studies. This paper exploited these strengths by adding to a slowly emerging body of knowledge of PBP systems based on meta-analysis. In general, the results show merit pay appears to be successful in motivating employee performance, incentive pay and merit pay comparably influence performance motivation, and study setting (field, laboratory, work simulation) differentially affect PBP systems. In addition, the study explored various moderator effects on merit pay systems, which heretofore had not been investigated meta-analytically. Major findings of these supplemental analyses show (a) the length of performance evaluation period has no effect on merit pay increases, (b) performance ratings has little impact on the pay-performance relationship, (c) the pay-performance relationship grows stronger over time, and (d) management occupations have stronger pay-performance correlations than non-managers. The final section of this dissertation provides suggestions for PBP plan implementation and evaluation, which should be of interest to compensation practitioners. It also proposes ideas for future PBP research, as well as study limitations.