Forecasting demand for secondary materials: A case study of waste paper

The authors examine how best to forecast the demand for a resource, in this case waste paper, using three methods. The first is based on long or short-term GDP forecasts, the second is a basic or modified Holt-Winters model, and the third is an input-output model, which can incorporate various scenarios to allow for the effects of technical change. All three techniques are superior to time-series projection. As data availability increases, the application of a combination of these methods would improve the policy response to resource shortage.