Experiments on participation and performance in the workplace

An increasing number of organizations are pursuing employee empowerment programs, expecting that employees’ participation to decision-making enhances workers’ satisfaction and performance. However, quantitative evidence of the effect of workers’ participation on performance is hard to be found. We present two distinct experiments, which we designed to advance the understanding of this topic: one field experiment with workers and one laboratory experiment with students employed in a real job. In the first, we study the effect of participation on the performance of healthcare workers. We run our field experiment in a triplet of care homes, which are managed by the same non-profit organization, but geographically remote. In two of them, we exogenously introduce a recognition scheme. In the first facility, the workers choose the scheme (the bottom-up treatment). In the second facility, the manager introduces the same scheme as chosen in the first facility (the top-down treatment). The third facility is the control group. Our results show that the bottom-up treatment has a significant and positive effect on the quality of workers’ performance (compared to the top-down), as measured by mistakes in the workplace. This difference in quality persists even after the recognition scheme is removed. In the second experiment, we study the effect of participation in a multitasking job. In a controlled environment, the experimental economics laboratory, we enrol students to perform a real data-entry job, whose performance can be measured in terms of speed or accuracy. In this experiment, participation takes the form of a choice over either a flat-rate or a speed-based piece-rate payment scheme. In the baseline treatment, experimental subjects are assigned to one of the payment schemes. In the participation treatment, subjects can choose their payment scheme. In the reference treatment, subjects are shown the availability of either payment scheme and then randomised in the piece-rate or flat-rate scheme. The multitasking theory predicts that, in this setting, pay-for-performance on speed should deviate effort from quality (the non-incentivized task) to speed. We do not find evidence of this theory in our baseline treatment. Instead, the multitasking problem arises under the participation treatment, which has a negative effect on quality. We show that this result cannot be explained by subjects’ self-selection into the piece-rate scheme. By comparing the baseline and reference treatments, we also show that the multitasking problem in the participation treatment does not arise because the subjects are aware of the differences between payment schemes. We conclude discussing conjectures on how to reconcile the findings of these two experiments.

Keywords: participation, incentives, performance, experiment, multitasking, piece rates, reference point.

Luca Savorelli

Lecturer at the University of St Andrews since 2012.

Since 2018 is Director of Teaching of the School of Economics and Finance and has been recently elected in the University Academic Senate.

He is a graduate of Collegio Superiore at the University of Bologna and received his PhD in Economics from the University of Siena in 2011. He has worked on the economics of obesity and published in the top-field Journal of Health Economics and Health Economics. He also published in the field of industrial organization and the economics of patents. Since 2014 he has been studying the effect of workers’ empowerment and participation on performance through a set of field and laboratory experiments. These involved workers across the healthcare, baking, woodworking and data-entry sectors. He was awarded grants for this research by the British Academy and the Scottish Institute for Research in Economics.