The Effects of Incentives on Workplace Performance: A Meta-analytic Review of Research Studies1
Steven J. Condly University of Central Florida Richard E. Clark University of Southern California Harold D. Stolovitch, Ph.D. HSA Learning & Performance Solutions, LLC
ABSTRACT A meta-analytic review of all adequately designed field andlaboratory research on the use of incentives to motivate performance is reported. Of approximately 600 studies, 45 qualified. The overall average effect of all incentive programs in all work settings and on all work tasks was a 22% gain in performance. Team-directed incentives had a markedly superior effect on performance compared to individually-directed incentives. This effect was not influenced by thelocation of the study (business, government, or school), the competitive structure of the incentive system (programs where only the highest performers get incentives versus programs where everyone who increased performance receives incentives), the type of study (whether the study was a laboratory experiment or a field study), or the performance outcome (quality, quantity, or both). In thesestudies, money was found to result in higher performance gains than non-monetary, tangible incentives (gifts, travel). More research is needed on the relative cost-benefit of cash and gift incentives, and the way different types of tangible incentives are selected. Long-term programs led to greater performance gains than shorter-term programs did, and somewhat greater performance gains were realizedfor manual than for cognitive work. Explanations based on cognitive psychological principles accompany each of the analyses.
Meta-analysis is a relatively new statistical procedure that allows us to summarize the results of many different experiments conducted on a specific topic by different researchers at different points in time. It is vastly superior to previous “box score” methods ofsummarizing experimental results. Box score methods of summarizing research involve 46
counting the positive, negative, or no results of studies on a topic, summing the results and reporting the number of studies “in favor or opposed” to a generalization such as “incentives boost performance.” A number of research design experts such as Jenkins, Mitra, Gupta, and Shaw (1998) haveprovided compelling evidence that the box score method often gives misleading information about Performance Improvement Quarterly
the results of many experiments. Meta-analytic approaches were developed to overcome these problems and allow researchers to estimate the effect size of treatments such as incentive systems as a percentage of a standard deviation change in performance due to the strategybeing investigated. Expressing the impact of a performance improvement strategy as a percentage of a standard deviation allows us to quickly estimate the percentage increase in performance that can be expected from the strategy that was studied. These estimates have been shown to be highly accurate, provided that well designed studies are utilized for the meta-analysis. This report summarizes ourattempt to perform a comprehensive meta-analysis of incentive experiments conducted in the field and in the laboratory.2
Plan of the Study
Financial incentives, and their use and misuse, have long been the focus of researchers and practitioners dedicated to maximizing human performance. What has been lacking, however, is a sound methodological review of the incentives literature. Jenkins etal. (1998) provided the first meta-analysis of the financial incentives literature. To be included in their meta-analysis, studies had to: (a) be between 1960 and 1996; (b) be empirical in nature; (c) not involve self-report measures; (d) focus on incentives geared toward the individual, not groups or organizations; (e) have a control or premeasure comparison groups; (f) focus on monetary...