Forecasting Demand and Measuring Price Elasticity

The basic tools of economics and statistics are applied to water‐demand forecasting and price‐elasticity measurement by the East Bay Municipal Utility District, Oakland, Calif. Little expertise is required to produce good forecasting results with time‐series models, which can also yield meaningful elasticity estimates if price increases are significant and the data are sufficiently disaggregated. Pooled time‐series and cross‐sectional models are more demanding in their structure and data requirements but often provide better estimates of the impact of price variables than simple time‐series analyses.