With the widespread use of narrative feedback in companies, understanding how such feedback can be valuable for employee performance improvement is important. Drawing on proprietary data from an e-commerce company, we investigate the role of language specificity and causality – two key language characteristics relevant to self-regulation and learning. Our findings suggest that the effects of specificity and causality differ depending on whether the feedback refers to strengths or weaknesses. Specifically, employees are more likely to improve when they receive more specific narrative feedback on their strengths, consistent with employees engaging in more systematic exploration when feedback refers to many desirable behaviors. In contrast, we find that increases in the specificity of narrative feedback on weaknesses can have negative performance consequences, which aligns with unsystematic exploration when the feedback refers to many undesirable behaviors. Furthermore, in line with employees gaining a better understanding of why certain behaviors were undesirable, employees are more likely to improve when narrative feedback on their weaknesses contains a higher degree of language causality. In additional analyses, we find evidence consistent with supervisor characteristics moderating the relation between language causality and employee performance improvement. Overall, this study provides novel empirical insights into the types and characteristics that make feedback valuable.

Many firms have begun granting teams greater autonomy in determining how they organise and report on their work. This study examines how communication influences (dis)honest reporting within such autonomous teams. Through a series of three experiments, we analyse how team members use communication to influence one another and steer their team’s reporting decisions. Our first two experiments provide evidence of an asymmetric effect on honesty: while communication from initially dishonest team members corrupts initially honest team members, the latter fail to discipline the former, resulting in a convergence toward dishonesty. In the third experiment, where we make the adverse consequences of misreporting for the firm’s owner salient to the team members, the convergence toward dishonesty disappears. Collectively, our findings are consistent with social norm theory, which suggests that when communication is present, it can be used to increase team misreporting depending on the activation and salience of different situational cues. Our study contributes to the literature on participative budgeting and dishonesty by revealing the process through which communication can escalate collective dishonesty in teams and identifying ways in which firms can mitigate this effect.

This study examines whether non-wage benefits can facilitate the sorting of employees into different remuneration arrangements within firms. Using data from the largest employer review platform in Germany, we derive four main insights. First, we document that the number and types of benefits employees receive exhibit substantial within-firm variation. Second, we find that women receive more benefits compared to men and that this difference is concentrated in benefits that can be considered family-friendly. Third, analyses that exploit information on employee satisfaction and cross-sectional variation in firm characteristics suggest that the observed gender differences in received benefits are due to differential preferences. Finally, we examine whether, consistent with preference-based sorting, firms offering more variation in benefit arrangements are associated with positive employment outcomes. We find that greater within-firm variability in benefits is associated with higher employee satisfaction and retention. Taken together, our findings provide corroborating evidence for the sorting role of benefits in the design of employee remuneration packages.

This study examines a performance evaluation process that integrates multiple raters and calibration committees-practices increasingly adopted by organizations yet underexplored in academic research. Using proprietary data, we investigate how supervisors aggregate multi-rater assessments into initial performance ratings and how calibration committees adjust these aggregation decisions. Our findings indicate that both supervisors and calibration committees generally follow the principles of information economics. When aggregating multi-rater assessments, supervisors assess the informativeness of different assessments and place more weight on more informative assessments. However, our results also reveal limitations in information processing. Under conditions of high information load, supervisors appear to rely on cognitive shortcuts, engaging less in detailed weighting decisions based on the informativeness of individual multi-rater assessments. Regarding calibration, we find that committees are less likely to adjust supervisors’ aggregation decisions when those decisions are consistent with more informative multi-rater assessments and when supervisors provide well-substantiated justifications, highlighting the complementary roles of multi-rater systems and calibration. Additionally, calibration committees appear to strategically focus on cases where supervisors face a high information load and information sources that may not have been fully incorporated into the initial performance ratings.

New digital technologies allow companies to provide employees with performance feedback in real time. We investigate the causal effects of such real-time feedback on individual effort and the quality of initial (prefeedback) and final (postfeedback) decisions by exploiting a natural quasi-experiment in the setting of professional soccer refereeing. Our results suggest that real-time feedback has differential effects for experienced compared with inexperienced agents. Specifically, we find some evidence that experienced agents decrease their effort under real-time feedback without harming the quality of their initial decisions. This results in better final decisions because real-time feedback allows agents to correct wrong initial decisions upon feedback. In contrast, inexperienced agents increase their effort, but make worse initial decisions. Upon feedback, they manage to compensate for their increased number of wrong initial decisions, but see no improvement in the quality of final decisions.

This study investigates the individual performance consequences of adopting unrewarded group relative performance information (RPI). Using proprietary data from a non-profit clinic, we exploit the staggered feature of the adoption of unrewarded group RPI on a reporting task and find that performance increases by 55.5 % when physicians are provided with group RPI. This overall improvement in performance can be divided into the initial treatment effect (10.5 %) and marginally decreasing improvements in performance (starting from 7.8 % per month) over time. Furthermore, we find significant variation in the improvement of performance, which is dependent on (1) the prior performance rank of the group, (2) the size of the group and (3) the task heterogeneity within the group. In sum, our study implies that group RPI can be effective in eliciting higher individual performance even when performance is not tied to financial incentives, and that the effectiveness of such group RPI is dependent on the conditions under which group RPI is implemented.

Prior research on voluntary disclosures has long debated whether managers tend to withhold bad news. However, these studies have been conducted in settings in which, ex-ante, the trade-off between the potential benefits and the potential costs of withholding information is obscure. In this paper, we study voluntary disclosure choices using a context-rich setting of distressed firms in which potential benefits from withholding news (particularly bad news) are seemingly high, whereas the potential costs are seemingly low. Specifically, we focus on the question of how ‘going concern’ uncertainty affects management earnings forecasts in financially distressed firms. Our results suggest that as financial distress intensifies, there is a lower likelihood and frequency of management earnings forecasts, indicating that managers may be withholding news, particularly bad news, in distressed firm-years. For comparative purposes, we also present results for ‘safe’ firm-years and find that managers have a tendency to disclose bad news as the financial health of the firm worsens.