A principal selects a group of agents to execute a collection of tasks
according to a specified order. Agents, however, have their own individual
ordering preferences according to which they wish to execute the tasks. There
is information asymmetry since each of these ordering preferences is private
knowledge for the individual agent. The private nature of the priorities of the
individual agents (adverse selection) leads to the effort expended by the
agents to change from the initial preferred priority to the realized one to
also be hidden as well (moral hazard). We design a mechanism for selecting
agents and incentivizing the selected agents to realize a priority sequence for
executing the tasks that achieves socially optimal performance in the system,
i.e., maximizes collective utility of the agents and the principal. Our
proposed mechanism consists of two parts. First the principal runs an auction
to select some agents to allocate the tasks, based on the ordering preference
they bid. Each task is allocated to one agent. Then, the principal rewards the
agents according to the realized order with which the tasks were performed. We
show that the proposed mechanism is individually rational and incentive
compatible. Further, it is also socially optimal under linear cost of ordering
preference modification by the agents.