In this paper, we examine how retail store managers reduce their sales activity in response to target ratcheting. We find that managers with favorable sales performance in the first three quarters reduce their sales activity in the final quarter. We also document that managers who engage in sales reducing activities enhance their likelihood of meeting their next-year sales target, which is based on their current sales. That is, managers who reduce their sales activity in the final quarter are more likely to beat their next-year sales targets than managers who refrain from reducing their final-quarter sales.

Using a sample of 140 managers, we investigate the use of various performance metrics in determining the periodic assessment, bonus decisions, and career paths of business unit managers. We show that the weight on accounting return measures is associated with the authority of these managers, and we document that both disaggregated measures (expenses and revenues), and nonfinancial measures play a greater role as interdependencies between business units increase. The results suggest separate and distinct roles for different types of performance measures. Accounting return measures are used to create the proper incentives for managers with greater authority, while disaggregated and nonfinancial measures are employed in response to interdependencies.

We explore how a supervisor's hierarchical rank affects the extent to which employees’ targets reflect their past performance. Literature documents that supervisors do not fully ratchet targets for past performance, arguably because the commitment not to penalize successful employees with more difficult targets alleviates the severity of the ratchet effect. We argue that commitment is less credible in organizational hierarchies where a middle manager sets employees’ targets. Using data from an organization comprised of three hierarchical layers, we consistently find that a middle manager's exposure to performance pressure is positively associated with the ratcheting of the employees’ targets. Moreover, we show that management at headquarters reduces a middle manager's performance pressure when most of her employees missed their targets in the previous period. Overall, the results imply that the hierarchical rank is an important determinant of the credibility of a supervisor's commitment to deemphasize past performance in target setting.

We study principals’ use of performance information to allocate human capital development resources to underperforming agents. We use proprietary data from a retail firm that sets uniform, noncalibrated performance targets to ensure consistent quality and customer experience across all stores. Unlike calibrated targets, uniform, noncalibrated targets do not account for heterogeneity in local conditions. Given that human capital development is costly, we predict and find that principals do not support all agents who underperform on noncalibrated targets. Instead, they use relative performance information and concentrate support on underperforming agents who outperform their geographical peers. In contrast, for calibrated targets, we find no evidence that principals rely on relative performance; agents who underperform calibrated targets are generally eligible for human capital support. Consistent with our assumption that relative performance is informative about returns to human capital investment, we find that support for underperforming agents who outperform their peers improves performance.

We investigate two determinants of two choices in the control system of divisionalized firms, namely decentralization and use of performance measures. The two determinants are those identified in the literature as important to control system design: (1) information asymmetries between corporate and divisional managers and (2) division interdependencies. We treat decentralization and performance measurement choices as endogenous variables and examine the interrelation among these choices using a simultaneous equation model. Using data from 78 divisions, our results indicate that decentralization is positively related to the level of information asymmetries and negatively to intrafirm interdependencies, while the use of performance measures is affected by the level of interdependencies among divisions within the firm, but not by information asymmetries. We find some evidence that decentralization choice and use of performance measures are complementary.

We study theoretically and empirically the relation between altruism and incentive contract design. Theoretically, we extend Fischer and Huddart (2008) to investigate how social norms reinforce managers’ altruistic preferences, thus affecting the optimal contract design related to incentive strength and performance measurement. Empirically, we draw on the notion of an organization’s work climate to capture managers’ altruistic preferences. Using data collected from a sample of 557 managers, we find that in a work climate where managers are mostly out for themselves, firms have lower pay-for-performance sensitivity and place a greater weight on aggregate performance measures. In addition, respondents report that they engage more in undesirable actions that are unproductive and costly to firm owners. In contrast, in a work climate where managers care about others (including peers in their organizational unit), firms place lower weights on aggregate performance measures. At the same time, respondents report that they supply more effort and engage less in undesirable actions.

We show that favoritism biases subjective evaluations and that the presence of a third party can mitigate this bias. Using archival data from professional ski jumping, we find that, controlling for objective performance, judges favor athletes of their own nationality and athletes who have a compatriot on the panel. We predict and provide evidence that in-person observation by an audience is associated with lower levels of favoritism compared with third-party observation via mediated communication. We contribute to the accounting literature by highlighting how in-person observation by a third party can reduce the likelihood that favoritism biases subjective evaluations. Data availability: The data used in this study are publicly available from open-access sources.