We consider a benchmark static incentive scheme, i.e. a per unit subsidy, that induces a monopoly to produce a target output level. We show that the same output level can be achieved by a continuum of dynamic subsidy rules based on a performance indicator. The subsidy rules require only local information. The present value of the subsidies paid under anyone of our dynamic schemes is smaller than the amount paid under the static subsidy. Moreover, each of the dynamic subsidy rules results, at each moment, in a lower per unit subsidy than the static subsidy. The subsidy rate depends on a state variable that reflects the monopolist's history of performance. This variable depreciates over time, therefore requiring a permanent effort of the monopolist to maintain it at an optimal level. In an example with a linear demand, the subsidy costs of inducing efficiency are reduced by almost fifty per cent.We show that the cost of sorting and the network effects jointly determine the rate of participation of consumers in the process of recycling. The dominant producer of virgin material takes into account the recycling activities when it makes its pricing decision. The network effects can create multiplicity of steady-state equilibria. The government can improve welfare by influencing equilibrium selection.
