An optimal decentralisation policy is a way to maintain a solution to the problem of the planner under a maximum possible protected sphere. To enquire into the nature of an optimal decentralisation policy is to understand the nature of deviations between private decisions and the corresponding public decisions at the optimum. Intuitively, even if the gradual attainment of all conditions for an optimal decentralisation policy is deemed impossible, one could hope that the application of piecemeal policy over a part of the state of society would improve social welfare. Nevertheless the application of piecemeal policy may even reduce social welfare. This happens because, once a condition for optimal decentralisation policy is violated, the remaining conditions are no longer desirable in general. Thus current social ideals cannot be used as blueprints for social action; and the correction principle becomes irrelevant unless it is applied all at once.