We analyze production and capacity utilization strategies in a supply chain where individual components can be made either at a shared in-house manufacturing facility or at dedicated facilities of external subcontractors. The manufacturer and the subcontractor differ in terms of costs, production capacities, rates, and service level capabilities. Using Markov decision process models, we determine the optimal policy and characterize its structure. We derive the set of conditions that partitions the state space into regions and characterize optimal policies in each region. We derive optimal policies for manufacturer and subcontractors under different settings and show that the optimal policy has a multi-index structure in some settings.