Data for experiment reported on in the manuscript "The Effects of Reporting Frequency and Reporting Structure on Employee Effort"

<p dir="ltr">The data were collected in the CREED laboratory. The main task consists of four periods of 320 seconds during which the participants in the employee role perform a real effort task. This real effort task is the “Word Encryption task with Double Randomization” (‘WEDR task’) developed by Benndorf, Rau, and Sölch (2019) based on a similar task used by Erkal, Gangadharan, and Nikiforakis (2011). In this task, participants encrypt three-letter words into numbers using a provided encryption table. The encryption table contains all letters of the alphabet in random order, and a random number between 100 and 1,000 corresponding to each letter. For the three letters that form a word, employees need to fill out the corresponding number from the encryption table. After participants enter a correctly encrypted word, the computer generates a new word and a new encryption table. Participants cannot proceed to the next word until they correctly enter the current word.</p><p dir="ltr">We chose the WEDR task over similar real effort tasks used in previous studies because it minimizes the opportunities for learning behavior in repeated settings and because performance on this task is not substantially affected by participants’ ability (Benndorf et al. 2019). We opted for a real effort task as opposed to a chosen effort task to enhance experimental realism (Charness, Gneezy, and Henderson 2018). However, to alleviate concerns about experimenter control loss in real effort tasks, we opted for a hybrid approach similar to for example Chan (2018) and Gächter, Huang, and Sefton (2016). In each of the four work periods, employees lose one point out of an endowment of 300 points for every second spent working on the encryption task. Every period starts with a twenty-second grace period in which work is costless. After these twenty seconds, a stop button appears on the employees’ screen. Clicking this button immediately stops the encryption task. The seconds remaining at the moment the stop button is clicked are transformed into points. Points, in turn, determine the payoff of the participants from the experiment. Specifically, one point is worth 0.5 euro cents. Thus, for example, by clicking the stop button immediately as it appears, a participant can secure 300 × 0.5 = 150 euro cents (i.e., €1.50) in each work period. Employees who decide to stop working sit idle until the period ends and the experiment proceeds.</p><p dir="ltr">The role of the manager in the experiment is limited to assigning a discretionary reward to each employee in their company after the final (fourth) work period. For each employee, managers select a reward anywhere between 0 and 3,000 points. To eliminate reneging temptations (Baiman and Rajan 1995; Bol 2008), the bonus selected by the managers is not deducted from the managers’ payoff but is paid by the experimenters. Immediately after the managers assign their discretionary rewards, employees are informed about their manager’s reward decision. As emphasized in the instructions handout, employees are only informed about their own reward, and never learn the reward, effort, or performance of the other employee(s) in the company.<a href="#_ftn1" target="_blank">^[1]</a></p><p dir="ltr">We manipulate reporting frequency by varying how frequently managers are informed about the performance of their employees. In the low reporting frequency condition, managers are only informed about the aggregate performance of their employees after the fourth and final period. In the high reporting frequency condition managers are informed about the performance of their employees after each of the four work periods. The reports that the managers receive after each period list the employees’ performance in that period and the aggregate performance over the previous periods.</p><p dir="ltr">Span of control is manipulated by varying the number of employees who report to a specific manager. In the narrow (wide) span of control condition, each manager observes the performance and assigns a reward to two (five) employees. Managers’ payoff depends on the output of the employees in their company. In the narrow span of control condition, in which managers have two employees, each word correctly decoded by an employee contributes twenty points to their manager’s total. In the wide span of control condition, in which managers have five employees, each word correctly decoded by an employee contributes eight points to their manager’s total. Thus, the total expected number of manager points does not vary between conditions. However, an individual employee’s contribution to this total is inversely proportional to the span of control.</p><p dir="ltr">Results from manipulation checks included in the post-experimental questionnaire suggest that both the reporting frequency manipulation and the span of control manipulation were salient to participants. First, we check the reporting frequency manipulation by asking participants how often their performance was reported to their manager: “continuously”, “after each period”, “only after the final period”, or “never”. Out of the 58 participants in the low frequency condition, 54 (93.1 percent) answer this question correctly and out of the 64 participants in the high frequency condition, 53 (82.8 percent) answer this question correctly. The span of control manipulation is checked by asking employee-participants to recall the number of employees (including themselves) who reported to their manager. Out of the 62 participants in the narrow span of control condition, 61 (98.4 percent) gave the correct answer and out of the 60 participants in the wide span of control condition, 57 (95.0 percent) gave the correct answer. We conclude that most participants were adequately able to recall the specifics of their condition. Since all participants also passed the comprehension checks in advance of the experiment, we test our hypotheses using the full sample.<a href="#_ftn2" target="_blank">[2]</a></p><p dir="ltr">In addition to their earnings from the task, employees and managers also earn a fixed salary, which amounts to 300 points for the employees and 600 points for the managers. In summary, the payoff functions are as follows (payoffs expressed in points):</p><p dir="ltr">Employee Payoff = 300 + (1,200 – seconds worked) + bonus (1)</p><p dir="ltr">Manager Payoff = 600 + piece-rate × number of correctly</p><p dir="ltr">encrypted words by the company’s employees (2)</p><p dir="ltr">where the bonus for the employee is at the discretion of the manager and lies between 0 and 3,000 and the piece-rate for the manager is 20 in the narrow span of control condition and 8 in the wide span of control condition.</p><p dir="ltr">We measure employee effort, the dependent variable, as the number of costly seconds that employees work on the task in the aggregate of the four periods (i.e., the total number of seconds used minus the 80 “free” seconds from the four grace periods).This measure is consistent with Baiman (1982), who defines effort as a construct that is controllable by the agent, creates negative utility for the agent, and results in an increase in expected output.</p><p><br></p>