Incentives and Burnout: Dynamic Compensation Design With Effort Cost Spillover

Abstract
Employee burnout is a significant issue that has long plagued firms. Salespeople are particularly susceptible to burnout due to the high-pressure, boundary-spanning nature of their role as well as their performance-driven compensation. The prevalence of burnout is an indication that the costs of work-related effort (such as fatigue) not only drive a salesperson’s utility and choices in the present, but carry over into the future. The single-period principal-agent model typically used to study sales force compensation design cannot fully account for this, as it effectively treats both the firm and salesperson as myopic. Thus, we incorporate this ‘effort cost spillover’ effect in a dynamic, two-period principal-agent model, with the salesperson’s effort cost in the second period increasing in both her second-period effort and her first-period effort. We use this model to explore the optimal design of the salesperson’s compensation plan over time and to consider the connection between burnout and plan design. Our model allows a forward-looking firm to account for the cumulative effect of effort on the salesperson. As a result, we find that the firm prefers to offer weaker incentives in the first period than a single-period model would suggest, inducing less effort from the salesperson in order to decrease her cost of effort (and likelihood of burning out) in the future. If both the firm and the salesperson are forward-looking, the firm achieves its best possible outcome by committing to the salesperson’s contract for both periods in advance. In that case, the salesperson earns a negative expected surplus in the first period, which is then recovered with a positive surplus in the second. If the firm is unwilling or unable to commit to a long-term contract, that best possible outcome is not always achievable. Moreover, the firm’s equilibrium strategy may be to induce the salesperson to burn herself out (working so hard in the first period that she cannot be profitably employed in the second) and quit, even when she cannot be replaced in the second period and the best possible outcome is achieved by employing her in both periods. Joint with Rob Waiser and Jean-Pierre Benoît.