To motivate effort, organizations commonly use budget-based tangible rewards (e.g., gift cards, merchandise) in lieu of or in addition to cash rewards and they can distribute tangible rewards to employees either directly (employees are given merchandise directly) or indirectly (via a redeemable points program). In conjunction with various budget-based financial rewards, employees can receive feedback about how they performed relative to other employees. However, employees can intentionally misstate their expected performance (i.e., create budget slack) when participating in the budgeting process, impairing the usefulness of budgets for planning and motivation. This dissertation investigates the effects of different types of budget-based rewards (cash, tangible, or redeemable points) on budget slack creation and performance, and whether relative performance information [RPI] moderates these effects. As predicted, results from an experiment completed by 166 undergraduate students show that participants eligible to earn redeemable points create less slack (i.e., set more difficult performance budgets) than those eligible for cash or direct tangible rewards. Further, RPI provides participants with a descriptive norm that slack creation is socially acceptable, resulting in more slack. Although I do not find support for the predicted indirect relationship between reward type or RPI on performance via their effects on budget slack, I do find that the provision of RPI has a direct positive effect on performance. Finally, supplemental analysis shows that those provided with RPI and cash rewards outperform all others. These results suggest that firms choosing to provide budget-based tangible rewards and allowing employees to participate in the budgeting process should consider using a redeemable points system rather than providing rewards directly to eligible employees. Further, before deciding whether to provide RPI to employees, firms should weigh the positive direct effects of RPI on performance against its negative effects on budget slack creation. Last, if a firm does choose to provide employees with RPI because of its positive effects on employee effort, firms may be well-advised to offer employees budget-based cash rewards instead of budget- based tangible rewards or budget-based points rewards.

In this study, employees are given autonomy in effort allocation across two tasks - complex and simple tasks, where the return to the organization is significantly higher for the complex task requiring high skill than for the simple task requiring low skill. An unavoidable feature of multi-task settings is that effort expended on one task detracts from effort that can be expended on another task. This effort trade-off among tasks becomes problematic when the returns from different tasks are unequal, with important consequences for a firm's overall performance. The design of management accounting control systems in such multi-task setting is difficult because organizations have to achieve multiple objectives: to improve productivity on both simple and complex tasks (i.e., performance) and to direct employee effort to more complex tasks given that the complex tasks are more valuable to firms (i.e., effort allocation). In a laboratory experiment, I examine the effects of two motivational mechanisms, financial compensation and relative performance information (RPI), on employee performance and effort allocation between simple and complex tasks. I find that the effects of RPI and financial compensation are independent such that each motivational mechanism affects performance and effort allocation separately. In addition, I find that the effects of RPI or financial compensation depend on whether a worker is a top performer or a bottom performer. Also, findings demonstrate that the effects of these motivational mechanisms on employee effort allocation and performance depend on the complexity of the task. Future research studies and managers who design incentive systems should consider the implementation of different types of incentives for different performer levels. Organizations should consider the degree of complexity of the tasks that workers must perform in multi-task settings.