Subjective Stress, Incentives, and Autonomy
How to cite
Shvartsman, Elena; Dohmen, Thomas. 2018. "Subjective Stress, Incentives, and Autonomy." AEA RCT Registry. AEARCTR-0003532. https://doi.org/10.1257/rct.3532-1.0
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 this laboratory experiment, we intend to analyze how work arrangements that couple employee autonomy and performance incentives affect occupational stress. According to the Job Demand-Control model (Karasek, 1979, Administrative Science Quarterly, p. 285-308), individuals may experience detrimental health effects when job demands are not met by a sufficient level of decision latitude (i.e., autonomy). Yet, it is not clear whether granting autonomy to employees who face incentivized performance targets may impair their well-being. In such incentive schemes, autonomy, e.g., in terms of self-determined working time, could induce workers who strive for reaching their performance goals to provide effort beyond an optimal level. This effect should be particularly pronounced if individuals face uncertainty about the level of their performance. We use a real-effort task in a laboratory setting, in which we manipulate the extent of autonomy, certainty, and incentives, to analyze which working conditions are perceived as particularly stressful.
Experimental design
Our design comprises six treatment arms, which vary by the extent of certainty about one’s own performance, the level of working time autonomy and the incentive structure. In all treatments, subjects work on a real-effort task adopted from Fliessbach et al. (2007, Science, p. 1305-1308). We ran a calibration study to select a work sequence, which permits us to target an ex-ante performance rate.
In the four main treatment arms, we incentivize performance by offering a bonus (as opposed by a fixed pay in the two other treatments) for reaching an absolute performance. We manipulate the extent of performance uncertainty by either continuously providing performance feedback or by only informing subjects at the end of the experiment about their performance. Since the level of difficulty of the working task is held constant across treatments and set such that the performance target is feasible according to the ex-ante calibrated rate, our research design enables us to elicit how individuals’ stress and effort are affected by exogenous changes in working time autonomy. We do so by contrasting a fixed working time regime with an autonomy regime. In the latter, subjects can stop working at any time or extend working time by foregoing monetary compensation. Subjects are allowed to read magazines when not working.
During the experiment, we elicit measures of subjective stress and well-being, the appraisal of a situation’s stressfulness, and track our subjects’ working time and working regime choices. Furthermore, we measure participants’ demographics, personality traits, and preferences, such as risk attitudes, uncertainty aversion, and their willingness to avoid workplace stress, in order to assess whether the individual stress response to working conditions is heterogeneous and whether this heterogeneity is partly determined by these measures.
Design details
Cited by
No citation data available for this registration yet.