Social Comparison, Employee Attrition and Productivity: Evidence from a Field Experiment at Workplace
How to cite
Liang, Yucheng; Wu, Hugh Xiaolong; Liu, Shannon X.. 2019. "Social Comparison, Employee Attrition and Productivity: Evidence from a Field Experiment at Workplace." AEA RCT Registry. AEARCTR-0004281. https://doi.org/10.1257/rct.4281-2.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
In many workplaces, newly-employed workers are usually assigned to teams where most teammates are more senior to them. As a result, new workers may be relatively well-informed about the current performance of senior workers, but not so much about the past performance of senior workers and the current performance of other new workers. This differential visibility of co-workers' performance may affect the well-being, productivity, and attrition of new workers through its interaction with two behavioral phenomena. First, new workers may set their reference point at the high level of the current performance of the senior workers, resulting in high stress and high attrition. Second, new workers may misattribute the current high performance of senior workers to their innate ability rather than the outcome of learning, and hence have a dim outlook about their own future performance.
In a field experiment at a leading multi-national spa chain, we study how information about co-workers' performance affects workers' well-being, productivity, and attrition. First, we study how workers' outlook about their future performance and their reference points for performance comparison depend on their knowledge about their co-workers' current and past performance. Then, we provide workers with information about their co-workers' performance. In Treatment 1 (T1), we provide information to each worker about the current performance of co-workers with similar work experience. In Treatment 2 (T2), we provide information to each worker about the performance trajectory of their senior co-workers. We measure the treatment effects on the workers' beliefs about their own future productivity, reference points, stress levels, productivity, attrition, and pro-social behaviors.
Intervention
We send biweekly messages to workers through an app.
T1: For each worker, we select her coworkers in the same region who join the firm in the same three-month window. In each message, we randomly select one such coworker and send his/her performance information of the last month through the app.
T2: We select the workers who have worked at the firm for at least 12 months ("senior workers"). For each worker, in each message, we randomly select one senior coworker in the same region and send his/her performance information in his/her 1st, 3rd, 6th, 12th work month as well as the performance of the last month through the app.
Experimental design
We conduct the experiment at a large spa chain in China. The chain has around 160 stores. About 5000 spa workers participate in the experiment. Before the intervention, we use a workplace info system app to conduct surveys on the workers to elicit their beliefs about their own future performance, the current performance of their coworkers with similar work experience, and the performance history of their senior coworkers. In the first stage of the intervention, we randomize all the stores into 3 groups. All workers in Group 1 (40 stores) undergo Treatment 1; all workers in Group 2 (40 stores) undergo Treatment 2; the rest of the stores (control group) do not receive intervention at this stage. Throughout the intervention, we track the outcome variables listed above.
Design details
Cited by
No citation data available for this registration yet.