Multiple tasks and political organization

Consider an environment such as a political election where a principal requires the completion of multiple tasks, but an agent can only be rewarded with a hire/fire decision rather than an endogenously chosen monetary payment. When the principal hires a single agent to perform multiple tasks, the agent allocates effort between the tasks inefficiently. I demonstrate that, even though hiring multiple agents completely mitigates this effort distortion problem, the principal is still better off hiring a single agent if the (exogenous) rewards for the tasks are sufficiently different. In contrast to similar results in the multi-task literature, this finding is not driven by risk aversion or noise. Rather, it is a direct result of the restricted contract space inherent to the environment.