Long Time Versus Steady State Optimal Control