Abstract
This article estimates a new model of residential electricity demand. It differs from previous work in two ways. First, we use individual monthly billing data in a pooled time-series/cross- section framework. Second, we use an engineering/thermal-load technique to model the household space-heating technology. We estimate the model using data from the Pacific Northwest and use the results to analyze three conservation measures: a price increase, a reduction in thermostat settings, and improvement of insulation levels. We find average rates of return for insulation upgrades of 4.9% for ceilings and 8.3% for walls.